Understanding the Genetics of Potato Tuber Calcium and its Implications in Breeding for Improved Quality By Cinthya Zorrilla Cisneros A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Plant Breeding and Plant Genetics) at the UNIVERSITY OF WISCONSIN-MADISON 2013 Date of final oral examination: 08/26/13 The dissertation is approved by the following members of the Final Oral Committee: Jiwan P. Palta, Professor, Horticulture Michael Havey, Professor (USDA), Horticulture Shelley Jansky, Associate Professor (USDA), Horticulture John Bamberg, Professor (USDA), Horticulture Shawn Kaeppler, Professor, Agronomy
262
Embed
Understanding the Genetics of Potato Tuber Calcium and its ...
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Understanding the Genetics of Potato Tuber Calcium and its Implications in Breeding for Improved Quality
By Cinthya Zorrilla Cisneros
A dissertation submitted in partial fulfillment of the requirements for the degree of
Doctor of Philosophy (Plant Breeding and Plant Genetics)
at the UNIVERSITY OF WISCONSIN-MADISON
2013
Date of final oral examination: 08/26/13 The dissertation is approved by the following members of the Final Oral Committee: Jiwan P. Palta, Professor, Horticulture Michael Havey, Professor (USDA), Horticulture Shelley Jansky, Associate Professor (USDA), Horticulture John Bamberg, Professor (USDA), Horticulture Shawn Kaeppler, Professor, Agronomy
i
GENERAL ABSTRACT
Calcium is a nutrient that plays an important signaling and structural roles in plants. The goal of
this research was to understand the relationship between tuber calcium and tuber quality using
two approaches. The first approach was the development of two reciprocal populations
segregating for specific gravity, yield, chip quality, internal quality, common scab, and tuber
calcium at the tetraploid level. This was accomplished by crossing Atlantic and Superior, two
cultivars that differ for all these traits. The broad-sense heritabilities and genetic correlations for
these traits were evaluated using data from field performances during 2009 to 2012 seasons.
Hollow heart incidence and incidence and severity of pitted scab were negatively correlated to
tuber calcium. Calcium was also negatively correlated to specific gravity, yield, and chip
lightness; and positively correlated. Quantitative trait loci that control tuber calcium, tuber
quality and pitted scab tolerance in the Atlantic x Superior population were identified. The
inheritance of tuber calcium has an important genetic component and is controlled by several
QTL located throughout the genome. The second approach, in the present investigation was to
study the effects of the expression of the calcium vacuolar antiporter CAX1 from Arabidopsis
under the control of the CaMV35S and the cdc2a promoters in the potato cultivar Atlantic. This
cultivar has low tuber calcium and poor tuber quality. Ou results suggest that an increased
transport of calcium into the vacuoles of these transgenic lines caused calcium deficiency
symptoms, compromised plant heath and increased tuber defects by a reduction of apoplastic
calcium and cell wall damage. These deficiency symptoms were partially ameliorated under
higher calcium treatments. Our results show that the calcium stored in the vacuoles of these
transgenics is in the form of calcium oxalate crystals which trap calcium and make it unavailable
to maintain proper membrane and cell wall functions. The new knowledge generated about the
ii
genetics and physiology of tuber calcium and its relationship with tuber quality could help in the
breeding effort to produce potato cultivars with improved tuber internal quality and scab
tolerance.
iii
DEDICATION
This dissertation is dedicated
to my family: Percy, Raymi and Illary;
to my parents Pedro and Marta;
and to my siblings Gina and Daniel.
for your love and support that gave me the strength to finish this journey.
iv
ACKNOWLEDGEMENTS
Thanks to Dr. Jiwan Palta for the opportunity to study the PhD. at UW Madison.
Thanks to all the professors in my Committee: Shelley Jansky, John Bamberg, Mike Havey and Shawn Kaeppler for the constructive criticism of my work. Thanks for allocating some of your
precious time to read my manuscript and listen to me.
Thanks to the current and past members of the Palta lab, especially to Felix Navarro, Sandra Vega, Kyle Rak, Justin Schabow, Zienab Fawsi, Amr Hassan, Nesse Okut, Fei Li, Vladimir
Chernov, and Young Suk Chung that collaborated with me to record the phenotypic data.
Thanks also to every person in the Agronomy, Botany, Horticulture and Plant Pathology departments that ever lend me a reagent, a piece of equipment, or a tool.
Thanks to all the staff at Walnut street greenhouses as well as the Agricultural Research stations at Hancock, Rhinelander, and Arlington.
Thanks to Kendall Hirschi from the Baylor College of Medicine at Texas A&M University for sharing the CAX1 constructs and the positive controls.
Thanks to Christine Hackett from the James Hutton Institute for her support with the TetraploidMap software.
Thanks to the SolCAP project for selecting my population for genotyping with the SNP chip, especially to Joseph Coombs and David Douches.
Thanks to the USDA National Institute of Food and Agriculture and the HATCH grant by the University of Wisconsin, College of Agricultural and Life Sciences that partially funded this
research.
v
TABLE OF CONTENTS
ABSTRACT………………………………………………………………………………... i
DEDICATION…………………………………………………………………………...... iii
ACKNOWLEDGEMENTS………………………………………………………………... iv
TABLE OF CONTENTS…………………………………………………………..………. v
CHAPTER 1: General Introduction and Research Objectives………………………….…. 1
GENERAL INTRODUCTION……………………………………………………………. 1
RESEARCH OBJECTIVES……………………………………………………………….. 17
BIBLIOGRAPHY………………………………………………………………………….. 18
CHAPTER 2: The Atlantic x Superior reciprocal populations segregating for yield, specific gravity, tuber calcium, internal tuber quality, chip quality and common scab: opportunities to study the genetics of these traits at the tetraploid level………………..…..
32
ABSTRACT………………………………………………………………………………. 32
INTRODUCTION………………………………………………………………………… 35
MATERIAL AND METHODS…………………………………………………………… 39
RESULTS AND DISCUSSION…………………………………………………………… 42
CONCLUSIONS..………………………………………………………………………… 54
BIBLIOGRAPHY…………………………………………………………………………. 55
TABLES…………………………………………………………………………………… 62
FIGURES………………………………………………………………………………… 67
CHAPTER 3: Correlations and heritabilities of tuber quality, pitted scab and tuber calcium: Implications for selection of potatoes with improved tuber quality…...…………
69
ABSTRACT……………………………………………………………………………..... 69
vi
INTRODUCTION………………………………………………………………………… 70
MATERIAL AND METHODS…………………………………………………………… 73
RESULTS AND DISCUSSION…………………………………………………………… 78
CONCLUSIONS…………………………………………………………………………… 94
BIBLIOGRAPHY…………………………………………………………………………. 96
TABLES…………………………………………………………………………………… 102
FIGURES………………………………………………………………………………… 108
CHAPTER 4: Mapping QTL for Tuber Calcium, Tuber Quality and Pitted Scab in a Tetraploid Population of Potato (Solanum tuberosum) derived from Atlantic x Superior…
113
ABSTRACT……………………………………………………………………………… 113
INTRODUCTION………………………………………………………………………… 114
MATERIAL AND METHODS…………………………………………………………… 117
RESULTS AND DISCUSSION…………………………………………………………… 123
CONCLUSIONS…………………………………………………………………………… 143
BIBLIOGRAPHY…………………………………………………………………………. 144
TABLES…………………………………………………………………………………… 151
FIGURES………………………………………………………………………………… 159
CHAPTER 5: Over-expressing the Vacuolar Antiporter CAX1 in the Potato Cultivar Atlantic: Phenotype of the Transformed Clones and Implications to Understand the Role of Calcium on Tuber Quality and Plant Health……………………………………………
177
ABSTRACT……………………………………………………………………………… 177
INTRODUCTION………………………………………………………………………… 178
MATERIAL AND METHODS…………………………………………………………… 184
vii
RESULTS AND DISCUSSION…………………………………………………………… 190
SUMMARY AND CONCLUSIONS……………………………………………………… 200
BIBLIOGRAPHY…………………………………………………………………………. 203
TABLES…………………………………………………………………………………… 209
FIGURES…………………………………………………………………………………. 214
CHAPTER 6: General Discussion and Conclusions……………………………………… 233
DISCUSSION…………………………………………………………………………….. 233
CONCLUSSIONS………………………………………………………………………… 244
BIBLIOGRAPHY…………………………………………………………………………. 249
1
CHAPTER 1
General Introduction and Research Objectives
GENERAL INTRODUCTION
Role of calcium at the cellular and whole plant level
Calcium has been established as regulator of plant growth and development because changes
in cellular Ca2+ acting through Ca2+-modulated proteins and their targets regulate an
astonishing variety of cellular processes (Bush 1995, Reddy and Reddy 2004, Harper et al.
2004, Hepler 2005). Calcium concentrations vary within cell compartments. Cytoplasmic
Ca2+ concentrations are tightly regulated at 100-200nM but in the organelles can be in the µM
to mM range. The vacuoles are important repositories of Ca2+ with concentrations in the
millimolar range (Gilroy et al. 1993). Calcium also plays important structural roles in plants.
Nevertheless there is variability between plant species in the response of supplemental
calcium under salinity (Cramer 2004). Recently a protective effect of calcium that limited the
impact of salinization on metabolic ripening process and induced plant salt tolerance has been
demonstrated by evaluating the tomato proteome (Manaa et al. 2013).
Calcium has been shown to play an important role in freezing injuries (Minorsky 1985, Arora
and Palta 1988, Palta 1996, Piotrowska 1998, Palta 2013). Calcium is reduced in the plasma
membrane when subjected to freezing injury initiating the injury process (Arora and Palta
1988). Calcium induces the expression of cold acclimation genes (Monroy and Dhindsa 1995,
Knight et al. 1996). Some of the approaches to reduce chilling injury in fruits and vegetables
includes the post-harvest use of calcium application (Wang 1993).
Calcium is also a primary sensory molecule in response to heat stress (Saidi et al. 2011,
Mittler et al. 2012). Reduced thermo-tolerance was observed when the extracellular calcium
was artificially reduced in various plant species (Gong et al. 1998, Liu et al. 2005, Wu and
Jinn 2010) and moss (Saidi et al. 2009). In potato the impact of heat stress is mitigated by
increasing soil applied calcium (Tawfik et al. 1996, Kleinhenz and Palta 2002). Calcium
influx into the cytoplasm in response to heat stress is regulated by plasma membrane cyclic
nucleotide gated calcium channels (Finka et al. 2012). Calcium has also been involved in the
protection against heat stress-induced oxidative stress (Larkindale and Knight 2002).
In addition calcium has a role in senescence (Poovaiah 1979). For instance senescence of
several plant tissues and organs has been shown to be retarded by calcium (Poovaiah and
Leopold 1973, Ferguson et al. 1983, Chéour et al. 1992, among others). Calcium regulates
senescence by regulating antioxidant enzyme activity (Sairam et al. 2011).
4
Calcium also has a role in biotic stress signaling. The calcium-mediated pathogen defense
programs have been recently reviewed by Ma and Berkowitz (2012). The plant defense
responses are triggered by increased cytosolic Ca2+. Recent reports have found conserved and
unique responses of calcium regulated genes to biotic and abiotic stresses (Narsai et al. 2013,
Yang et al. 2013). Higher tuber calcium has been shown to provide protection against soft-rot
caused by Pectobacterium (McGuire and Kelman 1984, 1986).
Candidate genes to increase calcium content in crops
There are numerous proteins of diverse function that interact with calcium (Boudsocq and
Sheen 2010). Therefore it is a complex system to study and a single gene may not be
responsible of regulating potato tuber calcium. The efforts to genetically modify plants for
increasing calcium content to improve stress tolerance and nutritional value are shortly
reviewed below. Calreticullin (CRT) is a conserved protein that has several functions in
plants (Jia et al. 2009) including intracellular Ca2+ homeostasis and Ca2+-dependant signal
pathways (Gelebart et al. 2005), molecular chaperone activity in the endoplasmic reticulum
(Denecke et al. 1995, Gelebart et al. 2005, Williams 2006), control of cell adhesion (Johnson
et al. 2001, Opas et al. 1996), immune system and apoptosis (Waterhouse and Pinkoski 2007)
and wound healing and pathogenesis (Qiu and Michalak 2009). The CRT protein is a
signalling molecule localized in the cytoplasm and the endoplasmic reticulum (Corbett and
Michalak 2000). Three CRT have been identified in Arabidopsis. CRT1a and CRT1b are
mainly related to general protein folding events whereas CRT3 is involved in pathogen
responses (Christensen et al. 2008, 2010). One approach to increase calcium concentration
has been to over-express CRT. The over-expression of CRT resulted in increased calcium
content and increased tolerance to abiotic stress in wheat (Jia et al. 2008). Another approach
5
is to alter calcium transporters. The H+/ Ca2+ antiporter 1 designated as CAX1 is a tonoplast
calcium antiporter was identified in Arabidopsis thaliana by suppressing yeast mutants
defective in vacuolar Ca2+ transport (Hirschi et al. 1996). Several CAX antiporters have been
identified in Arabidopsis with different ion specificities such as CAX2 that transports heavy
metals (Hirschi et al. 2000), CAX3 that transports Ca2+ mainly in roots (Manohar et al. 2011)
and CAX4 (Cheng et al. 2002) among others. The over-expression of a short version of the
Cation Exchanger 1 (sCAX1) has been found to increase calcium content in Arabidopsis
(Hirschi et al. 1996), potato (Park et al. 2005a), tomato (Park et al. 2005b) and carrots
(Morris et al. 2008). However, the increase in calcium content in the potato tuber was not
significant for the purpose of improving the nutritional value of potatoes for human
consumption (Park et al. 2005a). The expression of sCAX1 was shown to reduce apoplastic
Ca2+ levels which increased membrane leakiness and increased blossom end defect in tomato
(de Freitas et al. 2011). These studies suggest that the calcium transported into vacuoles by
CAX1 is unavailable in the apoplast. Coexpression of CRT1 resulted in a significant decrease
in Ca2+ deficiency symptoms in both tomato and tobacco without the addition of
supplemental Ca2+ (Wu et al. 2012). This effect of CRT on sCAX1-expressing lines might
have been caused by its calcium storage release function observed in Arabidopsis (Wyatt et
al. 2002). These studies suggest that the genetic control of calcium stored in organs is
controlled by several genes with complex interactions.
The potato plant
Potato is the most important non-grain crop in the world. The most commonly cultivated
potato species is the autotetraploid Solanum tuberosum with a basic number of 12
6
chromosomes. The size of the potato genome is 844 Mb and 86% of this genome has been
sequenced and assembled (Potato Genome Sequencing Consortium 2011). In addition, a
study of the transcriptome of potato found tissue-specific gene expression and genes with
tissue or condition restricted expression (Massa et al. 2011). Linkage disequilibrium decay to
r2<0.1 evaluated in a set of cultivated potato was approximately 5.3 Kb (D’Hoop et al. 2010).
This characteristic has been considered as promising because the required marker density is
not a limiting factor to do association mapping (Björn et al. 2010).
Why study calcium in potato?
Potato tubers have some important characteristics that make potato a good model to study the
role of calcium in plants. The transpiration of potato tubers is low because they are under the
soil and therefore the movement of water and calcium to the tubers is low (Palta 1996).
Potatoes are rich in potassium and phosphorus but are rather poor sources of sodium and
calcium (Lampitt and Goldenberg 1940). Calcium taken up by the main root system is
transported into the plant foliage but not into the tubers and only the calcium taken in by
tuber roots and stolon roots is transported into the tuber (Kratzke and Palta 1986). Phenotypic
variation for the amount of calcium has been observed in wild and cultivated potatoes
(Bamberg et al. 1993, Karlsson et al. 2006). Tuber calcium concentrations between 119.2 -
295.1 μg/g dry weight were reported in cultivated potatoes grown in soils without
supplemental calcium whereas this concentrations increased to 142.9 - 338.4 μg/g dry weight
after application of supplemental calcium (Karlsson et al. 2006). These characteristics make
the study of calcium in potato tubers an interesting subject.
7
Breeding potato chipping varieties
Potato chips are very popular snack foods in the western diet. Yearly about 30 million MT of
potatoes or almost 10% of the global potato crop are converted into consumer products,
mainly in the European Union and North America where one to two thirds of the daily potato
consumption is in processed form such as French fries and potato chips (Keijbets 2008).
Processed potato consumption in the United States has steadily increased over the past five
decades from 11.48 kg per person in 1960 to an estimated 37 kg per person based on potato
fresh weight (Lucier and Dettmann 2007, Keijbets 2008). The increasing demand of this
snack food has been accompanied by the growth of the potato chip industry and thus
increasing the need for varieties with high yields that are suitable for chipping. In the last 50
years significant progress has been made towards breeding varieties with better chipping
quality containing higher tuber solids, lower levels of reducing sugars and lighter chip color.
Historically, the development of Lenape has been identified as a landmark for the
development of modern chipping varieties with higher tuber solids (Love et al. 1998).
Snowden, Pike and Atlantic are the most popular varieties used for chip production (Lucier
and Dettmann 2007). However chipping varieties still have some problems with internal
defects such as hollow heart, brown center, internal brown spot and black spot bruise. Chip
quality of potatoes is measured mainly by chip color. Chip color is affected by several factors
including the composition of the tubers (Rodriguez-Saona and Wrostald 1997). Chip color is
also affected by the storage temperature and age of the tubers that trigger the breakage of
starch into reducing sugars (Hyde and Morrison 1964). A high amount of reducing sugars
will turn into darker chips due to a chemical reaction called Maillard reaction which is a non-
enzymatic browning (Danehy 1986). In addition enzymatic browning or darkening of cut
8
potatoes is a result of the oxidation of various organic compounds (Makower 1964, Matheis
1983, Friedman 1997) and is considered a negative trait that reduces chip quality.
Evaluation of cytoplasmic inheritance in reciprocal populations of potato
Studies trying to understand cytoplasmic inheritance in potato have been performed in
reciprocal populations. Differences between reciprocal populations of potato have been
reported for male sterility and other traits. De la Puente and Peloquin (1968) studied
cytoplasmic male sterility in potato in relation to the tetraploid Groups Andigena and
Tuberosum. In addition, large reciprocal differences were observed in when clones of Group
Phureja and Group Stenotomum were reciprocally crossed to Group Tuberosum haploids.
These populations differed for several characteristics including tuber initiation, tuber set, vine
senescence, tuber yield, flowering, and male fertility generated by the difference in
photoperiod. Cytoplasmic inheritance was considered one of the possible explanations for
this difference in photoperiod reaction (Sanford and Hanneman 1979). A report of significant
yield differences between reciprocal populations of potato was observed when parents of
opposite maturities were crossed. The higher-yielding reciprocal always had the higher-
yielding parent as the maternal parent (Sanford and Hanneman 1982). Differences in chip
color performance between reciprocal populations were observed in diploid populations
(Lauer and Shaw 1970, Jakuczun and Zimnoch-Guzowska 2004) but not in tetraploid
populations (Coffin et al. 1988, Ehlenfeldt et al. 1990, and Pereira et al. 1993).
9
Internal defects of potato
Absence of internal defects is critical for acceptable tuber quality for potato processing.
Internal tuber defects such as hollow heart, brown center, internal brown spot and black spot
bruise can reduce the value of potatoes; and thus cause huge losses to growers (Tucker 2013).
Current popular chipping varieties namely Atlantic (Webb et al. 1978) and Snowden suffer
from poor internal quality. Consequently, breeding efforts are needed to generate new chip
varieties with reduced defects and good tuber quality.
Hollow heart is a physiological disorder characterized by a star-like or irregularly-shaped
cavity in the pith region of the tuber (Levitt 1942). This defect arises when growing
conditions abruptly change during the season or when the potato plants recover too quickly
after a period of environmental or nutritional stress (Rex and Massa 1989). After recovery the
tubers begin to grow rapidly that results in the tuber pith necrosis or tissue pulling apart
leaving a void in the center (Hutchinson 2003). Brown center is another common internal
defect of the potato tuber pith characterized by a region of cell necrosis which results in
brown tissue that frequently precedes the development of hollow heart (Hutchinson 2003).
Some methods have been developed to detect hollow heart without destroying the tuber
including an acoustic impact method that measures the resonant frequencies of the tubers that
detected hollow heart with an R2=0.97 (Elbatawi 2008), and X-rays analysis (Finney and
Norris 1978). However, these methods are useful for the industry, but they do not prevent
losses for the growers.
Another severe physiological disorder called internal brown spot (IBS) in the mid-western
US and called internal heat necrosis (IHN) in the eastern US is a non-pathogenic internal
necrosis. The expression of symptoms varies significantly by genotype-environment
interactions (Sterret and Heninger 1997, Yencho et al. 2008). This defect is characterized by
10
brown spots or blotches that first appear toward the apical end of the tuber parenchyma but in
severe cases may involve most of the parenchyma (Henninger et al. 2000). The application of
calcium to plants in a nutrient solution was shown to decrease internal brown spot (Olsen et
al. 1996, Ozgen et al. 2006). However, some contradictory data have been found in other
studies that indicated that neither lime nor gypsum were effective to reduce internal heat
necrosis (Sterret and Henninger 1991).
Blackspot bruise is a physiological disorder that results from the mechanical stress during
harvesting and handling (Baritelle et al. 2000, Lærke et al. 2002). This type of lesion is
normally recognised as a 1-2 mm zone with a bluish-grey to black colored region beneath the
skin without visible cell wall fractures (Hughes 1980). This defect is not detected either
visually or chemically on the surface (Baritelle et al. 2000). The color of blackspots results
from polyphenol oxidase mediated oxidation of phenols to the black pigment melanin
(Matheis 1987). The disruption of intracellular membranes is an immediate effect of the
impact; and the consequent contact between the polyphenol oxidase located in the
amyloplasts and its substrates located in the vacuole may be the explanation for the
development of blackspots (Lærke et al. 2000). The susceptibility to bruising has been related
with high specific gravity (Baritelle and Hyde 2003). Therefore the structural properties of
the tuber are crucial for its resistance to blackspot formation caused by impact. In addition
recent studies suggest that the incidence of bruise goes down dramatically once the tuber
calcium concentration is above 200 μg/g (Karlsson et al. 2006).
11
Common scab of potato
Another disease that affects tuber quality is common scab caused by several species of the
bacterial genus Streptomyces. that belong to the Actinomycetes (Bjor and Roer 1980). The
lesions caused by Streptomyces spp. in potato tubers range from superficial to deep cavities
on the potato surface (Loria et al. 1997). The characterization of strains that cause pitted scab
show that they are a homogenous group with high cellulolytic and proteolytic activities
(Faucher et al. 1995). Higher incidences of common scab have been observed in soils with
higher pH (Blodgett and Cowan 1935). Tubers of scab susceptible cultivars have been found
to contain more reducing sugars in the tuber peel (Goto 1981). Resistance to common scab is
usually tested in fields where scab is known to occur frequently and a natural inoculum is
available (Goth et al. 1993, Mishra and Srivstava 2001). No cultivars are known to be
immune to common scab and differences in resistance and susceptibility are quantitative
(McKee 1958, Harrison 1962, Scholte and Labruyere 1985). The most practical method of
reducing losses caused by common scab is to use resistant varieties and cultural practices that
make conditions unfavorable for scab development (Loria 2001, Haynes et al. 2010).
Tuber calcium and its relationship with internal defects of potato and disease tolerance
Calcium has a very important nutrient that has signaling and structural roles in plants.
Transport of calcium is associated with water flow and therefore transpiration (Gilliham et al.
2011). Tubers are structures that do not transpire as much as the foliage (Baker and Moorby
1969) and thus can experience calcium deficiency even at optimal calcium levels in the soil.
Localized tissue calcium deficiencies are implicated as mechanisms initializing cell death and
tissue necrosis leading to internal brown spot and hollow heart in potatoes (Bangerth 1979,
12
Collier et al. 1980, Olsen et al. 1996, Palta 1996, Levitt 1942, Arteca et al. 1980).
Physiological studies have been conducted that demonstrate that in-season fertilization with
calcium results in an increase in tuber calcium (Clough 1994). This increase in tuber calcium
was correlated to a decrease in the incidence of internal defects such as blackspot bruise
(Karlsson et al. 2006) hollow heart and internal brown spot (Clough 1994, Kleinhenz et al.
1999, Ozgen et al. 2006) internal brown spot as well as sub-apical necrosis (Tzeng et al.
1986).
A good supply of calcium contributes to reduce favorable conditions for pathogen
development. For example pathogen populations of Pseudomonas solanacearum decreased in
tomato stems with increased calcium concentrations (Yamazaki and Hoshina 1995). Also a
negative correlation has been observed between calcium and severity of bacterial wilt
induced by Ralstonia solanacearum in tomato (Jiang et al. 2013). Potato and tomato are
related species and therefore we can expect similar effects of calcium on potato pathogens.
McGuire and Kelman (1984, 1986) demonstrated reduced severity of soft rot caused by
Pectobacterium carotovorum and improved storage quality of potatoes with increased
calcium concentration. Subsequent studies by Flego et al. (1997) demonstrated that an
increase in extracellular calcium concentration in the plant repressed pehA expression a pectic
enzyme-encoding gene by the pathogen. Another disease that has been studied in relation to
calcium is the common scab of potato. Studies of the correlation between tuber calcium and
the incidence of common scab have shown contradictory results. Whereas some of them have
indicated a positive correlation (Davis et al. 1974), others showed no correlation (Blodgett
and Cowan 1935, Lambert and Manzer 1991). The discrepancy in their results may be caused
by differences the type of calcium product applied and the time of application.
13
Approaches to identify quantitative trait loci (QTL)
Mapping quantitative trait loci (QTL) is a procedure that helps breeders to identify regions in
the chromosomes that are associated with simple and quantitative traits (Collard et al. 2005).
The identification of QTL allows breeders to detect regions associated with traits and in some
cases the identification of the actual genes that control traits. QTL mapping is usually
performed in diploid populations coming from F2 or recombinant inbred lines (Collard et al.
2005, Broman and Sen 2009). F2 populations also called intercross are generated by selfing
the F1 of two inbred lines so that all marker classes AA AB and BB will be found in the
population. Recombinant inbred lines (RIL) are generated by crossing two inbred strains
followed by repeated selfing or sibling mating to create a new inbred line whose genome is a
mosaic of the parental genomes (Broman and Sen 2009). Population size is an important
factor that affects QTL detection. A small population size can cause the over-estimation of
QTL effects and the loss of power to detect QTL (Xu 2003). Another important factor on
QTL detection is heritability; analytical methods locate QTL with poor precision (10–30 cM)
unless the heritability of an individual QTL is high (Kearsey and Farquhar 1998). The largest
benefit from the application of MAS would be observed for traits that exhibit low heritability
(Beavis 1998). Unfortunately, there is little power to identify markers linked to QTL or
accurately estimate their effects on traits with low heritability (Beavis 1994, 1998).
There are several methods to identify QTL in a linkage map; these methods can be classified
in single-QTL and multiple-QTL mapping methods. Interval mapping is the most popular
single-QTL mapping method that estimates the location of a QTL relative to its flanking
markers. A disadvantage of interval mapping is that QTL outside the interval under
consideration could lead to false positive or negative results (Rodriguez-Zas et al. 2002). An
approach that allows a more precise detection of QTL by accounting for the effects of
14
neighboring QTL is called composite interval mapping. Composite interval mapping (Zeng
1993, Jansen 1993) involves the use of interval mapping to locate a QTL between a pair of
markers and the use of multiple regression to estimate the effects of the QTL using a selected
set of markers as cofactors (Bernardo 2010). These cofactors, also called covariates, are a
predetermined number of markers that reduce the residual sum of squares of the multiple
regression models that best predict the phenotype (Broman and Sen 2009). The covariates are
searched outside a predetermined window size. The optimal window size depends on the
linkage disequilibrium of the mapping population so that the larger the number of breakpoints
the smaller the window size should be (Broman and Sen 2009). The advantages of using
composite interval mapping compared to interval mapping are an increase in the precision of
the detection by reducing the residuals detection of loci with modest effect and separation of
linked QTL (Zeng 1993, Jansen 1993). The use of near markers as covariates is a useful
exploratory strategy; however it turns a multidimensional search into a single dimensional
search that can overestimate the precision of the QTL detected. The choice of the number of
covariates used is a critical step in the process because too few or too many covariates can
cause a loss of power to detect QTL (Broman and Sen 2009). Currently there are several
methods that use a multiple QTL mapping approach such as multiple interval mapping (Kao
et al. 1999) and Bayesian QTL mapping (Ball 2001). Multiple QTL mapping can be
accomplished by using different model selection procedures from complete additivity to
multi-way interaction models. The models used to fit the QTL model include the Haley-Knott
regression extended Haley-Knott regression and multiple interval mapping; and different
model search approaches such as forward selection backward elimination and stepwise model
search (Broman and Sen 2009). Another approach is association mapping that uses an
association mapping panel that represents the historical recombination and natural genetic
15
diversity instead of the classical population structures used for interval mapping. Association
mapping enables researchers to use modern genomic technologies to exploit natural diversity
using genomic data (Zhu et al. 2008).
Mapping QTL in tetraploid potato populations
Since the first high density map of tomato and potato was constructed (Tanksley 1992) many
studies attempting to map QTL in potatoes have been published (Douches and Freyre 1994,
Menéndez et al. 2002, among others). Most of these studies involve diploid populations. The
densest diploid map is the RH-DH ultra-high density map constructed with 10000 AFLP
markers developed by van Os et al. (2006). This map has been used for example for high
resolution mapping of the H1 locus harboring resistance to the cyst nematode Globodera
(Bakker 2004). The methodology and software for tetraploid linkage maps have been
developed more slowly compared to diploid maps due to the complex inheritance in
tetraploid genomes (Hackett et al. 2007). New statistical approaches to deal with these
complex models of segregation and software that uses these approaches have been developed
in the last decade. Methodologies to perform interval mapping has been developed for
tetraploids (Hackett et al. 2001, Cao et al. 2005, Li et al. 2011) and implemented in
TetraploidMap (Hackett et al. 2007). This software was developed to deal with simple
sequence repeats (SSR) and amplified fragment length polymorphisms (AFLP) but can be
adapted to use with single nucleotide polymorphisms (SNP). Furthermore, the availability of
genome-wide genotyping tools such as the Illumina Infinium Bead Chip developed by the
SolCAP that evaluates simultaneously 8303 SNP (Hamilton et al. 2011) is allowing the
generation of high density maps. The advantages of using SNP markers for tetraploid
mapping are that polymorphisms can be detected ideally in every position of the genome and
16
the possibility to detect dosage. Recently a tetraploid map with 3839 mapped SNP markers
was constructed for the Stirling x 12601ab1 population (Hackett et al. 2013) which is the
densest tetraploid potato map to this date. Previous QTL maps in tetraploid populations
include those developed by Meyer et al. (1998) for late blight (Phytophtora infestans (Mont.)
de Bary), Simko et al. (2004) for Verticillium dahliae, Sagredo et al. (2009) for Colorado
potato beetle (Leptinotarsa decemlineata [Say]), Bradshaw et al. (2008) for yield agronomic
traits and quality traits and by McCord et al. (2011a, b) for agronomic traits and internal heat
necrosis. All these populations are generated by bi-parental crosses. Therefore the QTL
detected are mostly relevant to those particular populations. The association mapping
approach is much more powerful to detect genetic variants across natural populations.
However, it is complex in tetraploids because the genotyping system used should distinguish
among alleles and quantify the allele copy number. Association mapping using genomic data
is starting to become a reality for tetraploid potato. A study by Urbany et al. (2011) found
several markers associated with bruising and enzyme discoloration using SSR and markers
targeting candidate genes. Also a recent study by Uitdewilligen et al. (2013) demonstrated the
accuracy of genotyping-by sequencing (GBS) of a large collection of autotetraploid potato
cultivars using next-generation sequencing.
Marker assisted selection in potato
One of the applications of the knowledge generated by QTL mapping is the development of
markers for marker assisted selection (MAS). MAS in potato offers great opportunities to use
currently available genetic data (Barone 2004) but it has not been extensively exploited. Most
of the reports of markers for MAS in potato are related to pathogen resistance (Pineda et al.
1993, Hämäläinen et al. 1997, Oberhagemann et al. 1999, Colton et al. 2006, Gebhardt 2006,
17
Śliwka et al. 2010). Some cases are related to tuber quality (Freyre and Douches 1994, Li et
al. 2013). Lopez-Pardo et al. (2013) reported the successful application of MAS to select for
resistance to PVY and Globodera pallida. Li et al. (2013) indicate that their MAS
experiments to select for tuber quality were only partially successful probably due to GxE
affecting marker-trait associations. Only the Pain1-8c marker showed a consistent positive
effect on chip quality in two years of evaluation. These studies suggest that there are some
benefits that can be expected from the application of MAS in potato but these markers have
to be tested in several environments and genetic backgrounds. The cost-effectiveness of the
large scale application of these markers for MAS has been discussed recently by Slater et al.
(2013). This study determined that MAS could be applied cost-effectively in the second
clonal generation for all models currently employed in potato breeding.
The advance in QTL mapping analysis genotyping methods and the availability of the potato
genome sequence promise to generate accelerated progress on potato breeding and genetics in
the coming years.
RESEARCH OBJECTIVES
The objectives of the research presented in this thesis include: (a) the generation of reciprocal
tetraploid populations that segregate for tuber calcium, yield, specific gravity, internal
defects, chip quality, and tolerance to common scab; (b) the evaluation of the genetic
correlation between tuber calcium and internal tuber defects, tuber calcium and chip quality
traits and tuber calcium and common scab; (c) the identification of promising clones with
good chipping quality, at least as good as Atlantic, that have improved internal tuber quality
18
compared to Atlantic (select for an Atlantic replacement); (d) the identification of QTL
regions that control tuber calcium, yield, specific gravity, internal quality, chip quality and
pitted scab as well as markers that could potentially be used for marker assisted selection; (e)
the investigatation of the role of CAX1, a vacuolar Ca+2 antiporter, in tuber calcium uptake,
plant health and tuber quality.
BIBLIOGRAPHY
Arora, R. and J.P. Palta. 1988. In vivo perturbation of membrane-associated calcium by freeze-thaw stress in onion bulb cells: Simulation of this perturbation in extracellular KCl and alleviation by calcium. Plant Physiology 87: 622-628. Arteca, R.N., B.W. Poovaiah, and L.K. Hiller. 1980. Electron microprobe and neutron activation analysis for the determination of elemental distribution in hollow heart potato tubers. American Journal of Potato Research 57:241-247. Baker, D.A. and J. Moorby. 1969. The transport of sugar, water, and ions into developing potato tubers. Annals of Botany 33: 729-741. Bakker, E., U. Achenbach, J. Bakker, J. van Vliet, J. Peleman, B. Segers, S. van der Heijden, P. van der Linde, R. Graveland, R. Hutten, H. van Eck, E. Copoolse, E. van der Voosen, J. Bakker and A. Goverse. 2004. A high-resolution map of the H1 locus harbouring resistance to the potato cyst nematode Globodera rostochiensis. Theoretical and Applied Genetics 109: 146-152. Ball, R.D. 2001. Bayesian methods for quantitative trait loci mapping based on model selection: Approximate analysis using the Bayesian information criterion. Genetics 59: 1351–1364. Bamberg, J.B., J.E. Palta, L.A. Peterson, M. Martin, and A.R. Krueger. 1993. Screening tuber-bearing Solanum (potato) germplasm for efficient accumulation of tuber calcium. American Journal of Potato Research 70: 219-226. Bangerth, F. 1979. Calcium-related physiological disorders of plants. Annual Review of Phytopathology 17: 97-122. Baritelle, A., G. Hyde, R. Thornton, and R. Bajema. 2000. A classification system for impact-related defects in potato tubers. American Journal of Potato Research 77: 143-148.
19
Baritelle, A.L. and G.M. Hyde. 2003. Specific gravity and cultivar effects on potato tuber impact sensitivity. Postharvest Biology and Technology 29: 279-286. Barone, A. 2004. Molecular marker-assisted selection for potato breeding. American Journal of Potato Research 81: 111-117. Bassett, P.A. 1980. The effect of soil salinity and calcium levels on the growth of Bromus mollis in the Camargue, France. Oikos: 353-358. Beavis, W.D. 1994. The power and deceit of QTL experiments: lessons from comparative QTL studies. In Proceedings of the Forty-Ninth Annual Corn and Sorghum Industry Research Conference, 250-266. Washington, DC: American Seed Trade Assoc. Beavis, W.D. 1998. QTL analyses: power, precision, and accuracy in molecular dissection of complex traits, ed A.H. Paterson, 145-162. Boca Raton, FL: CRC. Bernardo, R. 2010. Breeding for quantitative traits in plants. Second Edition. Minnnesota, USA: Stemma Press. Bjor, T., and L. Roer, 1980. Testing the resistance of potato varieties to common scab. Potato Research 23: 33-47. Blodgett, F.M. and E.K. Cowan. 1935. Relative effects of calcium and acidity of the soil on the occurrence of potato scab. American Potato Journal 12: 265-274. Boudsocq, M., and J. Sheen. 2010. Stress signaling II: calcium sensing and signaling. In Abiotic Stress Adaptation in Plants, 75-90. Springer Netherlands. Bradshaw, J.E., C.A. Hackett, B. Pande, R. Waugh, and G.J. Bryan. 2008. QTL mapping of yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Theoretical and Applied Genetics 116: 193–211. Broman, K.W., and S. Sen. 2009. A Guide to QTL mapping with R/qtl. New York: Springer. Bush, D.S. 1995. Calcium regulation in plant cells and its role in signaling. Annual review of plant biology 46: 95-122. Cao, D., B.A. Craig and R.W. Doerge. 2005. A model selection-based interval-mapping method for autopolyploids. Genetics 169: 2371-2382. Cheng, N.H., J.K. Pittman, T. Shigaki and K.D. Hirschi. 2002. Characterization of CAX4, an Arabidopsis H+/cation antiporter. Plant Physiology 128: 1245-1254. Cholewa, E., and C.A. Peterson, 2004. Evidence for symplastic involvement in the radial movement of calcium in onion roots. Plant Physiology 134: 1793-1802. Clough, G.H. 1994. Potato tuber yield, mineral concentration, and quality after calcium fertilization. Journal of the American Society of Horticultural Science 119: 175-179.
20
Coffin R, MK Keenan, R Yada, G Tai, M Slavnik, and G Johnston. 1988. Inheritance of genetic factors controlling chip processing quality after tuber storage at low temperatures [Abstract]. American Journal of Potato Research 65:474. Collard, B.C.Y., M.Z.Z. Jahufer, J.B. Brouwer and E.C.K. Pang. 2005. An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142: 169-196. Collier, G.F., D.C.E. Wurr, and V.C. Huntington. 1980. The susceptibility of potato varieties to internal rust spot in the potato. Journal of Agricultural Science Cambridge 94: 407-410. Colton, L. M., H.I. Groza, S.M. Wielgus and J. Jiang. 2006. Marker-assisted selection for the broad-spectrum potato late blight resistance conferred by gene Rb derived from a wild potato species. Crop Sciences 46: 589-594. Conn, S., and M. Gilliham. 2010. Comparative physiology of elemental distributions in plants. Annals of Botany 105: 1081-1102. Corbett, E.F., and M. Michalak. 2000. Calcium, a signaling molecule in the endoplasmic reticulum?. Trends in Biochemical Sciences 25: 307-311. Cosgrove, D. 2005. Growth of the plant cell wall. Nature Reviews 6: 850-861. Cramer, G.R. 2004. Sodium-calcium interactions under salinity stress. In Salinity: environment-plants-molecules, eds. A. Läucli and U. Luttge, 205-227. Netherlands: Springer. Chéour, F., J. Arul, J. Makhlouf, and C. Willemot. 1992. Delay of membrane lipid degradation by calcium treatment during cabbage leaf senescence. Plant Physiology 100: 1656-1660. Christensen, A., K. Svensson, S. Persson, M. Michalak, J. Jung, M. Michalak, S. Widell, and M. Sommarin. 2008. Functional characterization of Arabidopsis Calreticulin1a; a key alleviator of endoplasmic reticulum stress. Plant Cell Physiol 49: 912-924.
Christensen, A., K.Svensson, L.Thelin, W. Zhang, N. Tintor , D. Prins and S. Persson. 2010. Higher plant calreticulins have acquired specialized functions in Arabidopsis. PLoS One 5: e11342. D’Hoop, B.B., M.J. Paulo, K. Kowitwanich, M. Sengers, R.G. Visser, H.J. van Eck, and F.A. van Eeuwijk. 2010. Population structure and linkage disequilibrium unravelled in tetraploid potato. Theoretical and Applied Genetics 121: 1151-1170. Danehy, J.P. 1986. Maillard reactions: nonenzymatic browning in food systems with special reference to the development of flavor. In: Advances in food research 30, ed. C.O. Chichester, 77-138. Florida: Academic Press Inc.
21
Davis, J.R., G.M. McMaster, R.H. Callihan, J.G. Garner, and R.E. McDole. 1974. The relationship of irrigation timing and soil treatments to control potato scab. Phytopathology 64: 1404-1410. De la Puente, F. and S.J. Peloquin. 1968. Male fertility of selected 24 chromosome S. tuberosum hybrids. American Journal of Potato Research 45:436-437. Demarty, M., C. Morvan, and M. Thellier. 1984. Calcium and the cell wall. Plant, Cell & Environment 7: 441-448. Denecke, J., L.E. Carlsson, S.Vidal, A.S. Hoglund, B. Ek, M.J. van Zeijl, K.M. Sinjorgo, and E.T. Palva. 1995. The tobacco homolog of mammalian calreticulin is present in protein complexes in vivo. Plant Cell 7: 391-406. Douches, D.S., and R. Freyre. 1994. Identification of genetic factors influencing chip color in diploid potato (Solanum spp.). American Potato Journal 71: 581-590. Elbatawi, I.E. 2008. An acoustic impact method to detect hollow heart of potato tubers. Biosystems Engineering 100: 206-213. Ehlenfeldt M.K., D.F. Lopez-Portilla, A.A. Boe, and R.H. Johansen. 1990. Reducing sugar accumulation in progeny families of cold chipping potato clones. American Journal of Potato Research 67: 83-91. Epstein, E. 1972. Mineral nutrition of plants. Principles and Perspectives. New York: John Wiley and Sons. Faucher, E., E. Paradis, C. Goyer, N.C. Hodge, R. Hogue, R.E. Stall and C. Beaulieu. 1995. Characterization of Streptomycetes causing deep-pitted scab of potato in Quebec, Canada. International Journal of Systematic Bacteriology 45: 222-225. Ferguson, I.B., C.B. Watkins, and J.E. Harman. 1983. Inhibition by calcium of senescence of detached cucumber cotyledons effect on ethylene and hydroperoxide production. Plant Physiology 71: 182-186. Finka, A., A.F.H. Cuendet, F.J. Maathuis, Y. Saidi, and P. Goloubinoff. 2012. Plasma membrane cyclic nucleotide gated calcium channels control land plant thermal sensing and acquired thermotolerance. The Plant Cell Online 24: 3333-3348. Finney, E.E. and K.H. Norris. 1978. X-ray scans for detecting hollow heart in potatoes. American Potato Journal 55: 95-105. Flego, D., M. Pirhonen, H. Saarilahti, T.K. Palva, and E.T. Palva. 1997. Control of virulence gene expression by plant calcium in the phytopathogen Erwinia carotovora. Molecular Microbiology 25: 831-838. de Freitas, S.T., M. Padda, Q. Wu, S. Park, and E.J. Mitcham. 2011. Dynamic alternations in cellular and molecular components during blossom-end rot development in tomatoes
22
expressing sCAX1, a constitutively active Ca2+/H+ antiporter from Arabidopsis. Plant
Physiology 156: 844-855. Freyre, R., and D.S Douches.1994. Development of a model for marker-assisted selection of specific gravity in diploid potato across environments. Crop Science 34: 1361-1368. Fricke, W., R.A. Leigh, and A.D. Tomos. 1994. Concentrations of inorganic and organic solutes in extracts from individual epidermal, mesophyll and bundle-sheath cells of barley leaves. Planta 192: 310-316. Friedman, M. 1997. Chemistry, biochemistry, and dietary role of potato polyphenols. A review. Journal of Agricultural and Food Chemistry 45: 1523-1540. Gebhardt, C., D. Bellin, H. Henselewski, W. Lehmann, J. Schwarzfischer, and J. P.T. Valkonen. 2006. Marker-assisted combination of major genes for pathogen resistance in potato. Theoretical and Applied Genetics 112: 1458–1464. Gelebart, P., M. Opas M. Michalak. 2005. Calreticulin, a Ca2+-binding chaperone of the endoplasmic reticulum. International Journal of Biochemistry and Cell Biology 37: 260-266. Gilroy, S., P.C. Bethke and R.L. Jones. 1993. Calcium homeostasis in plants. Journal of Cell Science 106: 453-461. Gilliham, M., M. Dayod, B.J. Hocking, B. Xu, S.J. Conn, B.N. Kaiser, and S.D. Tyerman. 2011. Calcium delivery and storage in plant leaves: Exploring the link with water flow. Journal of Experimental Botany 62: 2233-2250. Gong, M., A.H. van der Luit, M.R. Knight, and A.J. Trewavas. 1998. Heat-shock-induced changes in intracellular Ca2+ level in tobacco seedlings in relation to thermotolerance. Plant Physiology 116: 429-437. Goto, K. 1981. The relationship between common scab severity and reducing sugar contents in the peel of potato tubers. Potato Research 24: 171-176. Goth, R W, Haynes K G, Wilson D R. 1993. Evaluation and characterization of advanced potato breeding clones for resistance to scab by cluster analysis. Plant Disease 77: 911-914. Hackett, C.A., J.E. Bradshaw, and J.W. McNicol, 2001. Interval mapping of QTLs in autotetraploid species. Genetics 159: 1819-1832. Hackett, C.A., I. Milne, J.E. Bradshaw, and Z. Luo. 2007. TetraploidMap for Windows: linkage map construction and QTL mapping in autotetraploid species. Journal of Heredity 98: 727-729 Hackett, C. A., K. McLean and G.J. Bryan. 2013. Linkage analysis and QTL mapping using SNP dosage data in a tetraploid potato mapping population. PloS One 8: e63939. Hämäläinen, J. H, K.N. Watanabe, J.P.T. Valkonen, A. Arihara, R.L. Plaisted, E. Pehu, L.
23
Miller, and S.A. Slack. 1997. Mapping and marker-assisted selection for a gene for extreme resistance to potato virus Y. Theoretical and Applied Genetics 94: 192–197. Hamilton, J.P., C. N. Hansey, B. R. Whitty, K. Stoffel, A. N. Massa, A. Van Deynze, W.S. De Jong, D.S. Douches, and C.R. Buell. 2011. Single nucleotide polymorphism discovery in elite North American potato germplasm. BMC Genomics 12: 302-312 Hanson, J.B. 1960. Impairment of respiration ion accumulation and ion retention in root tissue treated with ribonuclease and ethylenediamine tetraacetic acid. Plant Physiology 35: 372–379. Harrison, M.D. 1962. Potato russet scab, its cause and factors affecting its development. American Journal of Potato Research 39: 368-387. Harper, J.F., G. Breton, and A. Harmon. 2004. Decoding Ca2+ signals through plant protein kinases. Annual Review of Plant Biology 55: 263-288. Haynes, K.G., L.A. Wanner, C.A. Thill, J.M. Bradeen, J. Miller, R.G. Novy, L. Jonathan, D. Whitworth, L. Corsini, and B.T. Vinyard. 2010. Common scab trials of potato varieties and advanced selections at three US locations. American Journal of Potato Research 87: 261-276. Henninger, M.R., S.B. Sterret and K.G. Haynes. 2000. Broad-sense heritability and stability of internal heat necrosis and specific gravity in tetraploid potatoes. Crop Science 40: 977–984. Hepler, P.K. 2005. Calcium: A central regulator of plant growth and development. The Plant Cell 17: 2142-2155. Hirschi, K.D., R.G. Zhen, K.W. Cunningham, P.A. Rea, and G.R. Fink. 1996. CAX1, an H+/Ca2+ antiporter from Arabidopsis. Proceedings of the National Academy of Sciences 93: 8782-8786. Hirschi, K.D., V.D. Korenkov, N.L. Wilganowski, and G.J. Wagner. 2000. Expression of Arabidopsis CAX2 in tobacco. Altered metal accumulation and increased manganese tolerance. Plant Physiology 124: 125-134. Hirschi, K.D. 2004. The calcium conundrum. Both versatile nutrient and specific signal. Plant Physiology 136: 2438-2442. Hughes, J.C. 1980. Potatoes I: Factors affecting susceptibility of tubers to damage. Span 23:65-75. Hutchinson, C.M., J.M. White, D.M. Gergela, P.A. Solano, K.G. Haynes, R. Wenrich, and C.S. Lippi. 2003. Performance of chip processing potato varieties in northeastern Florida. HortTechnology 13: 706-711.
24
Hyde, R. B., and J.W. Morrison. 1964. The effect of storage temperature on reducing sugars, pH, and phosphorylase enzyme activity in potato tubers. American Journal of Potato Research 41: 163-168. Jakuczun, H. and E. Zimnoch-Guzowska. 2004. Inheritance of glucose content in tubers of diploid potato families. American Journal of Potato Research 81: 359-370. Jansen, R.C. 1993. Interval mapping of multiple quantitative trait loci. Genetics 135: 205-211. Jia, X.Y., C.Y. Xu, R.L. Jing, R.Z. Li, X.G. Mao, J.P. Wang and X.P. Chang. 2008. Molecular cloning and characterization of wheat calreticulin (CRT) gene involved in drought-stressed responses. Journal of Experimental Botany 59: 739-751. Jia, X.Y., L. H. He, R.L. Jing, and R.Z. Li. 2009. Calreticulin: conserved protein and diverse functions in plants. Physiologia Plantarum 136: 127-138. Jiang, J.F., J.G. Li, and Y.H Dong. 2013. Effect of calcium nutrition on resistance of tomato against bacterial wilt induced by Ralstonia solanacearum. European Journal of Plant Pathology 136: 547-555. Johnson, S., M. Michalak, M. Opas, P. Eggleton. 2001. The ins and outs of calreticulin: from the ER lumen to the extracellular space. Trends Cell Biol 11: 122-129. Kao C.H., Z.B Zeng, and R.D. Teasdale. 1999. Multiple interval mapping for quantitative trait loci. Genetics 152: 1203-1216. Karley, A.J., R.A. Leigh, and D. Sanders. 2000. Where do all the ions go? The cellular basis of differential ion accumulation in leaf cells. Trends in Plant Science 5: 465-470. Karlsson B.H., J.P. Palta, and P.M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. Hort Science 41: 1213-1221. Kleinhenz, M.D., J.P. Palta, C.C. Gunter, and K.A. Kelling. 1999. Impact of source and timing of calcium and nitrogen applications onatlantic'potato tuber calcium concentrations and internal quality. Journal of the American Society for Horticultural Science 124: 498-506. Kleinhenz, M.D. and J.P. Palta. 2002. Root calcium modulates the response of potato plants to heat stress. Physiologia Plantarum 115: 111-118. Kearsey, M.J. and A.G.L. Farquhar. 1998. QTL analysis in plants; where are we now?. Heredity 80: 137-142. Keijbets, M.J.H. 2008. Potato processing for the consumer: Developments and future challenges. Potato Research 51: 271-281. Kirkby, E.A., and D.J. Pilbeam. 1984. Calcium as a plant nutrient. Plant, Cell & Environment 7: 397-405.
25
Knight, H., A.J. Trewavas, and M.R. Knight. 1996. Cold calcium signaling in Arabidopsis involves two cellular pools and a change in calcium signature after acclimation. The Plant Cell Online 8: 489-503. Kratzke, M.G. and Palta, J.P. 1986. Calcium accumulation in potato tubers: Role of the basal roots. HortScience 21: 1022-1024.
Lampitt, L.H., and N. Goldenberg. 1940. The potato as food. Chemistry & Industry 1940: 748–761.
Lærke, P.E., J. Christiansen, M.N. Andersen, and B. Veierskov. 2002. Blackspots bruise susceptibility of potato tubers during tuber growth and storage determined by two different test methods. Journal of the Science of Food and Agriculture 80: 1332-1338. Lambert, D. H., and F. E. Manzer.1991. Relationship of calcium to potato scab. Phytopathology, 81: 632-636. Larkindale, J., and M.R Knight. 2002. Protection against heat stress-induced oxidative damage in Arabidopsis involves calcium abscisic acid ethylene and salicylic acid. Plant Physiology 128: 682-695. Lauer, F. and R. Shaw. 1970. A possible genetic source for chipping potatoes from 40 F storage. American Journal of Potato Research 47: 275-278. Legge, R.L., J.E. Thompson, J.E. Baker, and M. Lieberman. 1982. The effect of calcium on the fluidity and phase properties of microsomal membranes isolated from postclimacteric Golden Delicious apples. Plant and Cell Physiology 23: 161-169. Levitt, J. 1942. A historical study of hollow heart potatoes. American Journal of Potato Research 19: 134-143. Li, J., K. Das, G. Fu, C.Tong, Y. Li, C. Tobias and R. Wu. 2011. EM algorithm for mapping quantitative trait Loci in multivalent tetraploids. International Journal of Plant Genomics. Article ID 216547. doi:10.1155/2010/216547. Li, L.,E. Tacke, H.R. Hofferbert, J. Lübeck, J. Strahwald, A.M. Draffehn, B. Walkemeier, and C. Gebhardt. 2013. Validation of candidate gene markers for marker-assisted selection of potato cultivars with improved tuber quality. Theoretical and Applied Genetics 126: 1039-1052. Liu, J.X., and S.H. Howell. 2010. Endoplasmic reticulum protein quality control and its relationship to environmental stress responses in plants. The Plant Cell Online 22: 2930-2942. Liu, H.T., D.Y. Sun, and R.G. Zhou. 2005. Ca2+ and AtCaM3 are involved in the expression of heat shock protein gene in Arabidopsis. Plant, Cell & Environment 28: 1276-1284.
26
Lopez-Pardo, R., L. Barandalla, E. Ritter and J.I. Ruiz de Galarreta. 2013. Validation of molecular markers for pathogen resistance in potato. Plant Breeding 132: 246-251. Loria, R., R.A. Bukhalid, B.A. Fry, and R.R. King. 1997. Plant pathogenicity in the genus Streptomyces. Plant Disease 81: 836-846 Loria, R. 2001. Diseases caused by bacteria: Common scab. In Compendium of potato diseases ed. W.R. Stevenson R. Loria G.D. Franc and D.P. Weingartner 14–15. St. Paul: American Phytological Society. Love, S.L., J.J. Pavek, A. Thompson-Johns, and W. Bohl. 1998. Breeding progress for potato chip qQuality in North American cultivars. American Journal of Potato Research 75: 27–36. Lucier, G. and R. Dettmann. 2007. Vegetables and Melons Outlook. United States Department of Agriculture. Economic Research Service, VGS, 320, 49. Lynch, J., G.R. Cramer, and A. Läuchli. 1987. Salinity reduces membrane-associated calcium in corn root protoplasts. Plant Physiology 83: 390-394. Ma, Y., and G.A. Berkowitz. 2012. Biotic Stress Signalling: Calcium-Mediated Pathogen Defence Programmes. In: Plant Stress Physiology. Sergey Shabala Ed. Cambridge USA.CAB International. 291. Makower, R.U. 1964. Effect of Nucleotides on Enzymic Browning in Potato Slices. Effect of Nucleotides on Enzymic Browning in Potato Slices. Plant Physiology 39: 956-959. Manohar, M., T. Shigaki, and K.D. Hirschi. 2011. Plant cation/H+ exchangers (CAXs): biological functions and genetic manipulations. Plant Biology 13: 561-569. Manaa, A., M. Faurobert, B. Valot, J.P. Bouchet, D. Grasselly, M. Causse, and H.B Ahmed. 2013. Effect of salinity and calcium on tomato fruit proteome. Omics: A Journal of Integrative Biology 17: 338-352. Marinos, N.G. 1962. Studies on submicroscopic aspects of mineral deficiencies. I. Calcium deficiency in the shoot apex of barley. American Journal of Botany 49: 834-841. Massa, A.N., K.L. Childs, H. Lin, G.J. Bryan, G. Giuliano, and C.R. Buell. 2011. The transcriptome of the reference potato genome Solanum tuberosum Group Phureja clone DM1-3 516R44. PLoS One 6(10): e26801. Matheis, G. 1983. Enzymatic browning of foods. Zeitschrift für Lebensmittel-Untersuchung und Forschung 176: 454-462. Matheis, G. 1987. Polyphenol oxidase and enzymatic browning of potatoes (Solanum tuberosum). II. Enzymatic browning and potato constituents. Chemie Mikrobiologie Technologie der Lebensmittel 11: 67-70.
27
McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011a. Linkage mapping and QTL analysis of agronomic traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Crop Science 51: 771-785. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011b. QTL mapping of internal heat necrosis in tetraploid potato. Theoretical and Applied Genetics 122: 129–142. McGuire, R.G., and A. Kelman. 1984. Reduced severity of Erwinia soft rot in potato tubers with increased calcium content. Phytopathology 74: 1250-1256. McGuire, R. and A. Kelman. 1986. Calcium in potato tuber cell walls in relation to tissue maceration by Erwinia carotovora pv. atroseptica. Physiology and Biochemistry 7: 401-406. McKee, R.K. 1958. Assessment of the resistance of potato varieties to common scab. Potato Research 1: 65-80. Menéndez, C.M., E. Ritter, R. Schäfer-Pregl, B. Walkemeier, A. Kalde, F. Salamini, and C. Gebhardt. 2002. Cold sweetening in diploid potato: Mapping quantitative trait loci and candidate genes. Genetics 162: 1423-1434. Meyer, R.C., D. Milbourne, C.A. Hackett, J.E. Bradshaw, J.W. McNichol, and R. Waugh. 1988. Linkage analysis in tetraploid potato and association of markers with quantitative resistance to late blight (Phytophthora infestans). Molecular and General Genetics 259: 150-160. Mishra, K. K. and J.S. Srivastava. 2001. Screening potato cultivars for common scab of potato in a naturally infested field. Potato Research 44: 19-24. Mittler, R., A. Finka, and P. Goloubinoff. 2012. How do plants feel the heat?. Trends in Biochemical Sciences 37: 118-125. Minorsky, P.V. 1985. A heuristic hypothesis of chilling injury in plants: A role for calcium as the primary physiological transducer of injury. Plant Cell & Environment 8: 75-94. Monroy, A.F., and R.S. Dhindsa. 1995. Low-temperature signal transduction: induction of cold acclimation-specific genes of alfalfa by calcium at 25 degrees C. The Plant Cell Online 7: 321-331. Morris, J., K.M. Hawthorne, T. Hotze, S.A. Abrams, and K.D. Hirschi. 2008. Nutritional impact of elevated calcium transport activity in carrots. Proceedings of the National Academy of Sciences 105: 1431-1435. Narsai, R., C. Wang, J. Chen, J. Wu, H. Shou, and J. Whelan. 2013. Antagonistic, overlapping and distinct responses to biotic stress in rice (Oryza sativa) and interactions with abiotic stress. BMC genomics 14: 93. doi:10.1186/1471-2164-14-93.
28
Oberhagemann, P., C. Chatot-Balandras, R. Schäfer-Pregl, D. Wegener, C. Palomino, F. Salamini, and C. Gebhardt. 1999. A genetic analysis of quantitative resistance to late blight in potato: Towards marker-assisted selection. Molecular Breeding 5: 399-415. Olsen N.L., L.K. Hiller, and L.J. Mikitzel. 1996. The dependence of internal brown spot development upon calcium fertility in potato tubers. Potato Research 39: 165-178. Oparka, K.J., and H.V. Davies. 1988. Subcellular localisation of calcium in potato tubers. Potato Research 31: 297-304. Opas, M., M. Szewczenko-Pawlikowski, G.K. Jass, N. Mesaeli, and M. Michalak. 1996. Calreticulin modulates cell adhesiveness via regulation of vinculin expression. Journal of Cell Biology 135: 1913-1923. van Os, H., S. Andrzejewski, E. Bakker, I. Barrena, G.J. Bryan, B. Caromel and van H.J. Eck. 2006. Construction of a 10,000-marker ultradense genetic recombination map of potato: providing a framework for accelerated gene isolation and a genomewide physical map. Genetics 173: 1075-1087. Ozgen S., B.H. Karlsson, and J.P. Palta. 2006. Response of potatoes (cv. ‘Russet Burbank’) to supplemental calcium applications under field conditions. Tuber calcium yield and incidence of brown spot. Journal American Potato Research 83: 195-204. Palta, J.P. 1996. Role of calcium in plant responses to stresses: Linking basic research to the solution of practical problems. HortScience 31: 51-57. Palta, J.P. 2013. Merging physiological and genetic approaches to improve abiotic stress resistance. In Combating climate change: An agricultural perspective, eds. M.S. Kang and S.S. Banga, 237-270. Boca Raton, US: CRC Press. Park S., T. Kang, C. Kim, J. Han, S. Kim, R.H. Smith, L.M. Pike, and K.D. Hirschi. 2005a. Genetic manipulation for enhancing calcium content in potato tuber. Journal of Agricultural and Food Chemistry 53: 5598-5603. Park S., N.H. Cheng, J.K. Pittman, K.S. Yoo, J. Park, R.H. Smith, and K.D. Hirschi. 2005b. Increased calcium levels and prolonged shelf life in tomatoes expressing Arabidopsis H/Ca2 transporters. Journal of Agricultural and Food Chemistry 139: 1194-1206. Pereira, A.D.S., R.H. Coffin, R.Y. Yada and V.S. Machado. 1993. Inheritance patterns of reducing sugars in potato tubers after storage at 12C and 4C followed by reconditioning. American Potato Journal 70: 71-76. Pineda, O., M.W. Bonierbale, R.L. Plaisted, B.B. Brodie and S.D. Tanksley. 1993. Identification of RFLP markers linked to the H1 gene conferring resistance to the potato cyst nematode Globodera rostochiensis. Genome 36: 152-156.
29
Piotrowska, G. 1998. The role of calcium ions in freezing injuries of winter oilseed rape leaves. Effects of calcium channel blockers and mimesis by calcium ionophore. Acta Physiologiae Plantarum 20: 257-261. Poovaiah, B.W. 1979. Role of calcium in ripening and senescence. Communications in Soil Science and Plant Analysis 10: 83-88. Poovaiah, B.W. and A.C. Leopold. 1973. Deferral of leaf senescence with calcium. Plant Physiology 52: 236-239. Potato Genome Sequencing Consortium. 2011. Genome sequence and analysis of the tuber crop potato. Nature 475: 189-195. Preston, R.D. 1979. Polysaccharide formation and cell wall function. Annual Review of Plant Biology. 30: 55-78. Qiu, Y., and M. Michalak. 2009. Transcriptional control of the calreticulin gene in health and disease. International Journal of Biochemistry & Cell Biology 41: 531-538. Reddy, V.S., and A.S.N. Reddy. 2004. Proteomics of calcium-signaling components in plants. Phytochemistry 65: 1745-1776. Rex, B.L. and G. Massa. 1989. Cause, control and detection of hollow heart in potatoes: A review. American Journal of Potato Research 66: 165-183. Rodriguez-Saona, L.E. and R. Wrostald. 1997. Influence of potato composition on chip color quality. American Journal of Potato Research 74: 87-106. Rodriguez-Zas, S. L., B.R. Southey, D.W. Heyen and H.A. Lewin. 2002. Interval and composite interval mapping of somatic cell score, yield, and components of milk in dairy cattle. Journal of Dairy Science 85: 3081-3091. Sagredo, B., N. Balbyshev, A. Lafta, H. Casper, and J. Lorenzen. 2009. A QTL that confers resistance to Colorado potato beetle (Leptinotarsa decemlineata [Say]) in tetraploid potato populations segregating for leptine. Theoretical and Applied Genetics 119: 1171-1181. Saidi, Y., A. Finka, M. Muriset, Z. Bromberg, Y.G. Weiss, F.J. Maathuis, and P. Goloubinoff. 2009. The heat shock response in moss plants is regulated by specific calcium-permeable channels in the plasma membrane. The Plant Cell Online 21: 2829-2843. Saidi, Y., A. Finka, and P. Goloubinoff. 2011. Heat perception and signalling in plants: A tortuous path to thermotolerance. New Phytologist 190: 556-565. Sairam, R.K. B. Vasanthan, and A. Arora. 2011. Calcium regulates Gladiolus flower senescence by influencing antioxidative enzymes activity. Acta Physiologiae Plantarum 33: 1897-1904.
30
Sanford, J.C., and R.E. Hanneman,. 1979. Reciprocal differences in the photoperiod reaction of hybrid populations inSolanum tuberosum. American Potato Journal 56: 531-540. Sanford, J.C. and R.E. Hanneman Jr. 1982. Large yield differences between reciprocal families of Solanum tuberosum. Euphytica 31: 1-12. Simko, I, S. Costanzo, K.G. Haynes, B.J. Christ, R.W. Jones. 2004. Linkage disequilibrium mapping of a Verticillium dahliae resistance quantitative trait locus in tetraploid potato (Solanum tuberosum) through a candidate gene approach. Theoretical and Applied Genetics 108: 217–224. Slater, A.T., N.O. Cogan, and J.W. Forster. 2013. Cost analysis of the application of marker-assisted selection in potato breeding. Molecular Breeding 1: 12 Śliwka, J., H. Jakuczun, P. Kamiński, and E. Zimnoch-Guzowska. 2010. Marker-assisted selection of diploid and tetraploid potatoes carryingRpi-phu1, a major gene for resistance to Phytophthora infestans. Journal of Applied Genetics 51: 133-140. Scholte, K., and R.E. Labruybre. 1985. Netted scab: a new name for an old disease in Europe. Potato Research 28: 443-448. Sterrett, S.B. and M.R. Henninger. 1991. Influence of calcium on internal heat necrosis of Atlantic potato. American Potato Journal 68: 467-477. Sterrett, S. B. and M.R. Henninger, 1997. Internal heat necrosis in the Midatlantic region: Influence of environment and cultural management. American Journal of Potato Research 74: 233-243. Tanksley, S.D., M.W. Ganal, J.P. Prince, M.C. De Vicente, M.W. Bonierbale, P. Broun and G.B. Martin. 1992. High density molecular linkage maps of the tomato and potato genomes. Genetics 132: 1141-1160. Tawfik, A.A., M.D. Kleinhenz, and J.P. Palta. 1996. Application of calcium and nitrogen for mitigating heat stress effects on potatoes. American Journal of Potato Research 73: 261-273. Tucker, R.E.. 2013. A potato glossary. Tucker Farms Inc. Published on 6/26/13. http://www.tuckertaters.com/potato-glossary.pdf Tzeng, K.C., A. Kelman, K.E. Simmons, and K.A. Kelling. 1986. Relationship of calcium nutrition to internal brown spot of potato tubers and sub-apical necrosis of sprouts. American Journal of Potato Research 63: 87-97. Urbany, C., B. Stich, L. Schmidt, L. Simon, H. Berding, H. Junghans, K-H. Niehoff, A. Braun, E Tacke, H-R. Hofferbert, J. Lübeck, J. Strahwald, and C. Gebhardt. 2011. Association genetics in Solanum tuberosum provides new insights into potato tuber bruising and enzymatic tissue discoloration. BMC Genomics 12: 7.
31
Uitdewilligen, J.G., A.M.A. Wolters, B. Bjorn, T.J. Borm, R.G. Visser, and H.J. van Eck. 2013. A next-generation sequencing method for genotyping-by-sequencing of highly heterozygous autotetraploid potato. PloS ONE 8: e62355. Wang, C.Y. 1993. Approaches to reduce chilling injury of fruits and vegetables. Hort. Rev 15: 63-95. Waterhouse, N.J., and M.J. Pinkoski. 2007. Calreticulin: raising awareness of apoptosis. Apoptosis 12: 631-634. Webb, R.E., D.R. Wilson, J.R. Shumaker, B. Graves, M.R. Henninger, J. Watts, J.A. Frank, and H.J. Murphy. 1978. Atlantic: A new potato variety with high solids, good processing quality and resistance to pests. American Journal of Potato Research 55: 141-145. Williams, D.B. 2006. Beyond lectins: the calnexin/calreticulin chaperone system of the endoplasmic reticulum. Journal of Cell Science 119: 615–623. Wu, H.C., and T.L. Jinn. 2010. Heat shock-triggered Ca2+ mobilization accompanied by pectin methylesterase activity and cytosolic Ca2+ oscillation are crucial for plant thermotolerance. Plant Signaling & Behavior 5: 1252-1256. Wu, Q., T. Shigaki, J.S. Han, C.K. Kim, K.D. Hirschi, and S. Park. 2012. Ectopic expression of a maize calreticulin mitigates calcium deficiency-like disorders in sCAX1-expressing tobacco and tomato. Plant Molecular Biology 80: 609-619. Wyatt, S.E., P.L. Tsou and D. Robertson. 2002. Expression of the high capacity calcium-binding domain of calreticulin increases bioavailable calcium stores in plants. Transgenic Research 11: 1-10. Xu S. 2003. Estimating polygenic effects using markers of the entire genome. Genetics 163: 789−801. Yamazaki, H., and T. Hoshina. 1995 . Calcium nutrition affects resistance of tomato seedlings to bacterial wilt. HortScience 30: 91-93. Yang, T., H. Peng, B.D. Whitaker, W.M. and Jurick. 2013. Differential expression of calcium/calmodulin-regulated SlSRs in response to abiotic and biotic stresses in tomato fruit. Physiologia Plantarum 148: 445–455. Yencho, G.C., P.H. McCord, K.G. Haynes and S.R.Sterrett. 2008. Internal heat necrosis of potato: A review. American Journal of Potato Research 85: 69-76. Zeng, Z-B. 1993. Theoretical basis for separation of multiple linked gene effects in mapping of quantitative trait loci. Proceedings of the National Academy of Sciences 90: 10972–10976. Zhu, C., M. Gore, E.S. Buckler, and J. Yu. 2008. Status and prospects of association mapping in plants. The Plant Genome 1: 5-20.
32
CHAPTER 2
The Atlantic x Superior reciprocal populations segregating for yield, specific gravity,
tuber calcium, internal tuber quality, chip quality and common scab: opportunities to
study the genetics of these traits at the tetraploid level
ABSTRACT
Tuber quality traits are a major interest for breeders and the potato chip industry. This study
intended to generate populations that can be suiTable 2.for the genetic study of specific
gravity, yield, chip quality, internal quality, common scab, and tuber calcium at the tetraploid
level. Two populations were generated by reciprocally crossing Atlantic and Superior, two
cultivars contrasting for these traits. Trait segregation was assessed in both reciprocal
populations during 2009 to 2012 at Hancock, Wisconsin. A bell-shaped segregation was
observed for tuber yield, specific gravity, enzymatic browning, chip color using a visual
rating, chip color in agtron units, chip lightness, chip redness, chip yellowness, and tuber
calcium. However, the distributions were skewed towards resistance for the incidence of
hollow heart, internal brown spot, blackspot bruise, as well as pitted scab incidence and
severity. Atlantic and Superior had significantly different phenotypic performance for most
traits. In addition, the reciprocal populations differed significantly for tuber yield, internal
brown spot and tuber calcium suggesting parent-of origin effects influencing these traits. The
characteristics of these reciprocal populations offer opportunities to study the genetics of
quality traits in tetraploid potato and breed varieties that combine commercially desired traits.
33
INTRODUCTION
Commercial potato cultivars are mostly tetraploid. Potato breeding and selection of new
varieties is typically done at the tetraploid level. However, studies aimed at understanding the
genetics of various traits are conducted at the diploid level (Potato Genome Sequencing
Consortium, 2011). Therefore, taking advantage of new genetic knowledge generated in
diploid potatoes to generate breeding strategies towards new improved commercial varieties
may be a slow process. Nevertheless, genetic studies at the tetraploid level are becoming
more feasible with the development of whole genome SNP-based marker technologies for
potato (Hamilton et al. 2011), and the use of statistical approaches and software tools such as
TetraploidMap (Hackett et al. 2007) that allow quantitative genetic analyses at the tetraploid
level. Consequently, we expect that in the future of potato breeding tetraploid populations
could be used for genetic studies and selection speeding up the breeding process.
Genetic variation for agronomic traits (Freyre et al. 1994, Bradshaw et al. 2008, Haynes
2008), chip color (Pereira et al. 1995, Li et al. 2008), internal quality (Jansky and Thompson
1990, McCord et al. 2011a), common scab (McKee 1963, Bjor and Roer 1980, Haynes et al.
2010) and mineral content (Karlsson et al. 2006, Andre et al. 2007, Brown et al. 2012) has
been found among tetraploid potatoes suggesting that there is a potential for selecting
improved varieties for these traits.
Internal defects such as hollow heart, brown center, internal brown spot and black spot bruise
reduce the quality and thus the value of potatoes. Brown center is characterized as a region of
cell necrosis in the tuber pith that results in brown necrotic tissue (Bartholomew 1914).
34
Hollow heart is characterized by a cavity, usually star-shaped, in the center of the tuber that
may arise at a higher incidence under environmental or nutritional stresses (Levitt 1942; Rex
and Mazza 1989). Internal Brown Spot is an internal necrosis that first appears toward the
apical end of the tuber but in severe cases may include most of the parenchyma (Henninger et
al. 2000). Blackspot bruise is a bluish-grey to black zone beneath the skin that results from
the mechanical stress during harvesting and handling (Hughes 1980, Lærke, at al. 2002).
Internal defects such as hollow heart and brown center are more likely to appear in larger size
tubers (Nelson et al. 1979, Jansky and Thompson 1990). However, most diploid clones tend
to produce very small tubers. Thus, studies aimed at understanding the genetics of tuber
internal quality may not be feasible at the diploid level. The present study was undertaken to
develop progenies segregating for tuber quality traits using tetraploid cultivars.
Chip quality is measured using different parameters, chip color being the most important
among them. Chip color is influenced by the composition of the tubers (Rodriguez-Saona and
Wrostald 1997). This trait has been studied using visual color ratings (Work et al 1981),
three-dimensional colorimeters (Parkin and Shwobe 1990), and agtron refractance
colorimeters (Coles et al. 1993). In addition, enzymatic browning is another important trait
for chip quality. Enzymatic browning is the darkening of cut potatoes as a result of the
oxidation of various organic compounds (Makower 1964, Matheis 1983, Friedman 1996).
High enzymatic browning is considered a negative trait that reduces tuber processing quality.
Common scab is caused by several species of Streptomyces spp.; among them the
predominant species is Streptomyces scabies Thaxter (Bjor and Roer 1980, Lambert and
Loria 1989). The lesions caused by common scab range from superficial, raised, or deep
cavities (also called pits); the specific type of lesion present in a cultivar depends on the
pathogen strain and cultivar susceptibility (Loria et al. 1997, Kreuze et al. 1999). Tubers with
35
deep pitted scab cannot be used for processing (Archuleta and Easton 1981). No cultivars are
known to be immune to common scab and the difference in resistance and susceptibility is
quantitative (McKee 1958, Harrison 1962, Scholte and Labruyère 1985).
The mineral content of potato tubers is important not only for human nutrition, but also for
tuber health. Calcium has been shown to play an important role in potatoes response to biotic
(McGuire and Kelman 1986) and abiotic stresses (Palta 1996, Vega et al. 1996, Tawfik et al.
1996, Kleinhenz and Palta 2002). High levels of calcium applications have been associated to
reduced impact of soft-rot caused by Pectobacterium carotovorum (Jones) Waldee (McGuire
and Kelman 1986). Previous research has also demonstrated that application of calcium
increases tuber calcium levels (Kratzke and Palta, 1986, Clough 1994) and reduce the
incidence of internal defects (Tzeng et al. 1986, Olsen et al. 1996, Palta 1996, Ozgen et al.
2006, Karlsson et al. 2006).
Our strategy is to study the genetics of tuber quality traits and tuber calcium at the tetraploid
level using the progenies of Atlantic and Superior, two varieties with contrasting
characteristics for these traits. Atlantic is the standard variety for chipping from the field or
from very short-term storage. It has many traits that make it fit well to the chip industry needs
such as uniformity, high specific gravity, and high yield (Webb et al. 1978). Atlantic potatoes
are much less subject to enzymatic browning on cut and peeled surfaces compared to other
cultivars such as Russet Burbank (Sapers et al 1989). Conversely, Superior is a chipping
variety with low yield, low specific gravity and dark chips. Superior has higher content of
ascorbic acid and soluble proteins at harvest, and higher content of ascorbic acid, glucose,
fructose and total sugars after cold storage compared to Atlantic (Okeyo and Kushad 1995);
which might be related to its darker chip color. Superior has been reported to be resistant or
moderately resistant to internal defects and diseases for which Atlantic has been found
36
susceptible. Superior is resistant to net necrosis and relatively resistant to internal heat
necrosis (Rieman 1962). Atlantic tubers are known to be susceptible to internal heat necrosis,
particularly in sandy soils in warm, dry seasons; also, hollow heart can be serious in the
larger diameter tubers (> 10.2 cm) of Atlantic when moisture and nutrition conditions over
the season fluctuate (Webb et al. 1978). Superior is known to have less severe internal defects
and common scab. Superior has been classified as a moderately resistant and Atlantic as
moderately susceptible to common scab (Haynes et al. 2010). In addition, previous studies
have also revealed that Superior has higher tuber calcium compared to other well-known
varieties such as Russet Burbank, Snowden and Atlantic (Karlsson et al. 2006).
This research aims to assess the performance of Atlantic and Superior, and the segregation of
their progenies for tuber calcium, tuber yield, specific gravity, chip color, internal defects and
incidence of pitted scab. The potential uses of these populations for the study of potato
genetics and selection of new varieties is also discussed. Our results show not only that these
progenies produce good size tubers allowing the study of tuber internal defects, but the
progenies also provide the opportunity for selection of new varieties that could combine the
desired tuber quality with commercially important agronomic traits.
MATERIAL AND METHODS
Reciprocal populations
Reciprocal crosses of potato cultivars Atlantic and Superior were obtained in the Biotron of
the University of Wisconsin-Madison, USA in order to evaluate maternal effects on tuber
calcium and quality traits. Seeds were extracted two months after pollination, rinsed with
distilled water, air-dried, and treated with 1500 mg/Kg gibberellic acid for 24 hours at room
37
temperature to break dormancy and allow germination. The treated seeds were grown in
Murashige-Skoog media to obtain plantlets in order to multiply them clonally after
germination to perform a preliminary evaluation in greenhouse conditions reported by Vega
et al. (2006). These plantlets were then used to produce tubers that were subsequently
maintained as breeders’ seed in field conditions at the Rhinelander and Arlington Agricultural
Research Stations of the University of Wisconsin - Madison. The number of clones from each
reciprocal cross used in the experiments is indicated in Table 2.1.
Location and Experimental Design
Field trials were conducted under standard production practices for nutrient applications as
well as pest and disease control at the Hancock Agricultural Research Station located in the
commercial potato production area of Central Wisconsin, USA during 2009 to 2012 (Table
2.1). Fertilizer and pesticides were applied as needed during the season. No supplemental
calcium fertilizer was used in these trials. In the absence of rain, irrigation was scheduled
every other day. All trials were conducted under standard commercial production practices of
Central Wisconsin. In addition, a high disease pressure field, in other words a field that has
been used continuously to grow potatoes without rotation for several years and is known to
have high amounts of common scab inoculum, was used to evaluate pitted scab incidence and
severity. The standard field was used to evaluate tuber calcium, agronomic traits, internal
defects, chip quality, and pitted scab incidence; and the high disease pressure field was used
to evaluate pitted scab incidence and severity. The experimental design for the 2009 and 2010
trials in the standard field was an incomplete randomized block design (Yates, 1936) with 8-
hill plots and two or three replications for each clone depending on seed availability;
38
complete randomized block design (Fisher 1935) with 8-hill plots and three replications in
the 2011 trial in the standard field; and incomplete block design with 4-hill plots and three
replications for the 2011 and 2012 trials in the high disease pressure field. Separation
between plots was 91.4 cm and spacing between seed pieces within each plot was 30.5 cm.
The incomplete block design was structured in groups of 26 clones plus the two parents
randomized within the group, groups were as well randomized whitin replicates in one, two
or three years depending on the trait evaluated. The details of the field experiemnts, panting
and harvesting dates are presented in Table 2.1.
Tuber yield and specific gravity evaluation
Plots were harvested using a single-row digger and hand-picked to avoid mixing. All the
tubers were graded in three categories: A-grade (>4.8cm), B-grade (≤4.8cm) and culls (rotten
or green); and weighed immediately after harvest. The total tuber yield was expressed in tons
per hectare (ton/ha) and will be referred to as tuber yield. Specific gravity was determined by
the following formula: SG = Weightair/(Weightair-Weightwater), using a basket containing
approximately 2 kilograms of tubers and a scale PW-2050 (Weltech International, UK).
Tubers were stored at 12.8C until further evaluations.
Chip quality and enzymatic browning evaluation
Sixteen to twenty potato slices 1mm wide were sampled from 8 to 10 tubers, and fried in
cotton seed oil at 360ºF for 2 minutes and 20 seconds. Chip quality measurements included:
visual ratings of chip color, chip reflectance in agtron units, lightness, redness and
yellowness. Visual chip color ratings (CC) used a color scale from 1 to 5, where 1 is very
light and 5 is very dark. The chips were crushed to provide an even distribution of the sample
39
in the cup positioned on the viewer. Chips were read in triplicate and means calculated for
each plot. Agtron values (AG) were measured in the 2009 trial were performed in the USDA-
ARS Potato Worksite in East Grand Forks using an agtron M-300 reflectance
Enzymatic browning (EB) was measured as the intensity of darkening of the fresh tuber pith
tissue one hour after chopping in a visual scale from 1 to 5, 1 being light and 5 being very
dark, in 2010 and 2011.
Internal quality evaluation
The internal quality traits were evaluated immediately after harvest. All A-grade tubers were
cut in longitudinal sections to record the number of tubers with hollow heart (HH), brown
center (BC), internal brown spot (IBS) and bruise (BB). The number of tubers with internal
defects and the total number of tubers evaluated were recorded. For the data analysis, the
incidence of defects was expressed as proportions or percentages of defective tubers over the
total number of tubers depending on the analysis. The number of tubers cut depended on the
yield of A-grade tubers from a given clone; the minimum number of tubers cut for internal
evaluation was 8 tubers per plot.
Pitted scab evaluation
Deep scab lesions or pits were evaluated under standard and high disease pressure conditions.
The number of tubers with pits and the total number of tubers evaluated were recorded per
plot in the standard field (PS) and under high disease pressure (PS-E). Severity of pitted scab
40
(SPS-E) was measured as the average number of pits in the set of tubers evaluated per plot in
a high disease pressure field. Pitted scab incidence is treated as a binomial variable and
expressed as percentages or proportions of tubers having pitted scab depending on the
analysis. As noted above, the total number of tubers evaluated varied depending upon the
yield of a given clone; however, a minimum number of 8 tubers were evaluated for pitted
scab.
Estimation of calcium concentration in tuber tissue
Tuber calcium (TC) was quantified using the method described by Kratzke and Palta (1986).
For this purpose, two 1-mm-thick longitudinal slices were collected from the center of ten A-
grade tubers. The medullary tissue of the tuber was removed and dried in an oven at 60ºC,
ground to pass a 20-mesh screen, and ashed at 550 ºC. The ash was dissolved in 5 ml of 2 N
HCl. This solution was diluted with a lanthanum chloride solution to obtain a resulting
solution in 0.2 N HCl and LaCl3 at 2000 mg liter-1. The calcium concentration was
determined using an atomic absorption spectrophotometer (Varian SpectrAA 55B). Tuber
calcium was expressed in micrograms of calcium per gram of dry weight of tuber (µg/g).
Analysis of variance (ANOVA) and analysis of deviance (ANODE)
Data normality was tested using Q-Q plots, homogeneity of variances was assessed by
residuals versus fitted plots, and statistical independence was assumed. Datasets were
transformed to obtain a normal distribution when needed. Differences between parents and
reciprocal populations were assessed using multivariate models that reflect the experimental
design.
41
The models used in this study were the following:
Model I: y = μ + G + R + ε (randomized block design model)
Model II: y = μ + G + B(R) + ε (incomplete block design model)
Where, y is the observed trait measurement, μ is the overall mean, G is the genotypes, B is the
group, R is the replicate, B(R) is the group nested in replicates, and ε is the residual error.
Differences between parents were evaluated using Models I and II. The differences between
reciprocal populations were tested using a constrast between the mean of the genotypes that
belong to Atlantic x Superior versus the mean of the genotypes that belong to Superior x
Atlantic. For binomial data, which includes all incidence measurements (pitted scab and
internal defects) expressed as percentages, a Mann-Whitney test for pair-wise comparison
(Mann and Whitney 1947). This test is a non-parametric analysis and can be performed
without the assumption of normality (Holander and Wolfe 1973).
For normally distributed data, a linear model and the F-test were used for the analysis of
variance (ANOVA). For binomial data, a generalized linear model and a Chi-square test were
used for the analysis of deviance (ANODE). Visual scales were treated as numeric variables
instead of categorical variables to facilitate the analysis. Mean performances were estimated
for each clone and density plots were plotted to depict segregation for each reciprocal
population. The statistical analysis and plots were obtained using the stats and ggplot2
packages of R version 3.0.0, respectively (R Development Core Team, 2013).
42
RESULTS AND DISCUSSION
Performance of the parents: Atlantic and Superior
The mean performance of each parent was evaluated and compared in all trials (Tables 2.2
and 2.3). The two parents of the reciprocal populations used in this study, Atlantic and
Superior, were chosen due to their contrasting characteristics for several traits. One of those
traits is tuber calcium. Atlantic had tuber calcium concentrations of 130, 173 and 227 µg/g;
meanwhile, Superior had values of 182, 257 and 289 µg/g during the 2009, 2010 and 2011
trials, respectively (Table 2.2). Karlsson et al. (2006) reported concentrations of 119, 131 and
144 µg/g for Atlantic and 212, 188 and 295 µg/g for Superior in three years of evaluation,
respectively. Compared to previous reports, our estimations of tuber calcium are slightly
higher for Atlantic but it is still consistently lower than Superior by a difference of up to 80
µg/g in 2010.
The parents also differed in tuber yield (Table 2.2). Yield was constantly higher for Atlantic
with means of 73, 49, and 60 ton/ha; while, Superior had tuber yield of 60, 37 and 47 ton/ha
in 2009, 2010 and 2011, respectively. These yield values are in general higher than previous
reports that indicate mean yield of 44.8 and 32.4 ton/ha for Atlantic and Superior,
respectively (Love et al. 1998).
Atlantic and Superior also differed in specific gravity (Table 2.2). Atlantic had specific
gravities of 1.081, 1.077 and 1.079; in contrast, Superior had specific gravities of 1.073,
1.063 and 1.061 in 2009, 2010 and 2011, respectively. Specific gravities in our trials were in
the low range for both parents as compared to previous reports; for example, specific gravity
for Superior has been reported to be 1.056-1.089 (Rieman 1962); and for Atlantic 1.079-
43
1.100 (Sinha et al. 1992, Baritelle and Hyde 2002). However, as expected in our trials
Atlantic had consistently higher specific gravity than Superior.
In general, Atlantic had consistently lighter chips than Superior (Table 2.2). This difference
may be due to the fact that Atlantic is a descendant of Lenape (Webb et al. 1978); and Lenape
has been used to breed for good chipping quality in most of the current chipping varieties
(Love et al, 1998). Superior on the other hand is an old chipping variety which is now
primarily used for fresh market because of its low specific gravity and poor chip quality. The
lighter chip color of Atlantic is also shown by the higher agtron values, chip lightness and
chip yellowness but lower chip redness and visual rating compared to Superior. For chip
lightness, Atlantic had values of 48.1 and 56.6 versus 41.8 and 53.7 for Superior in 2010 and
2011, respectively. For chip yellowness, Atlantic had 20.4 and 22.9 compared to 17.8 and
21.9 for Superior. For chip redness, Atlantic had means of 8.6 and 2.1 versus 10.0 and 3.7 for
Superior (Table 2.2). The visual ratings in a scale from 1 to 5 were 3.1, 2.9 and 1.5 for
Atlantic and 3.5, 3.8 and 3.0 in a scale of 1 to 5 (light to dark) for Superior, respectively for
three years of trials. Chip color as measured in agtron units was 46.7 and 44.8 for Atlantic
and Superior, respectively (Table 2.2). These values are similar to previous reports; for
example, for Atlantic, chip lightness was 40 in the North Central Regional Trial (Navarro et
al. 2012), chip color in agtron units was reported to be between 54 - 62 (Orr and Sacks 1992,
Thompson et al. 2008), and average chip color rating of 3.0 - 3.3 on a 1 to 10 scale (light to
dark) have been reported (Webb et al. 1978, Hutchinson et al. 2003). Similarly, chip color for
Superior was reported to be 3.0 - 5.1 in a 1 to 10 scale, 1 being light and 10 being dark
(Bryan and Durre 1968, Webb et al. 1978).
44
Visual color ratings of enzymatic browning measured on a 1 – 5 scale on raw cut-potatoes
after one hour incubation at room temperature were also different for the parents (Table 2.2).
Scores of 1.04 and 1.5 were obtained for Atlantic compared to 2.14 and 3.0 for Superior in
the 2010 and 2011 trials, respectively. Sapers et al. (1989) measured enzymatic browning in
terms of the change in chip lightness; these authors reported that Atlantic had less enzymatic
browning than Russet Burbank. In another study, Atlantic showed the lowest value in
browning rate, total phenol content, and PPO activity in pre-peeled potatoes compared to
other varieties including Shepody (Jeong et al. 2005). All previous reports support our result
that Atlantic has low enzymatic browning. However, Superior was not tested for this trait by
these authors.
Atlantic and Superior also differed in the incidence of internal defects (Table 2.3). For
example, incidence of hollow heart was lower in superior compared to Atlantic. For hollow
heart, Atlantic had values of 12.0, 19.3 and 1.6% whereas Superior had values of 0.7, 0.6 and
0.6% in the 2009, 2010 and 2011 trials, respectively. Incidences varied among seasons but
were always higher for Atlantic. The Wisconsin Potato Breeding Program reported 5%
incidence of hollow heart for Atlantic in 2012 (Navarro et al. 2012).
For internal brown spot, Atlantic had incidences of 4.8, 12.8 and 0.8%; meanwhile, Superior
had mean incidences of 5.4, 13.5 and 0.0% evaluated in 2009, 2010 and 2011. Incidences
were slightly lower in Atlantic compared to Superior in 2009 and 2010, but higher in 2011. In
general, our results show similar incidences for both parents, but variable from year to year as
reported previously by Henninger et al. (2000) for internal heat necrosis, a defect related to
internal brown spot. In a study by McCord et al. (2011a), internal heat necrosis showed
incidences of 59, 15 and 20% in 2006, 2007 and 2008, respectively; however, these are much
45
higher than the incidences observed in our trials. Previous reports have shown that internal
brown spot is not very common in Wisconsin (Karlsson et al. 2006). It might be possible that
the environment in our trials was not harsh enough to observe clear differences between the
parents.
For blackspot bruise, Atlantic showed consistently higher incidence compared to Superior
with values of 12.0, 3.4, and 3.8% compared to 2.9, 2.2 and 0.6% in 2009, 2010 and 2011,
respectively (Table 2.3). Previous reports indicate that the incidence of blackspot bruise in
Atlantic was 32.6, 45.2 and 44.7% versus 9.1, 11.5 and 6.1% for Superior in 1999, 2000 and
2001, respectively (Karlsson et al. 2006). Our results also indicate higher incidences in
Atlantic than Superior, but the incidence were generally lower in our evaluations. This
difference is likely due to hand harvesting in our trials that reduced the amount of impact
damage. The earlier studies reporting higher incidences of blackspot bruise in Atlantic
compared to Superior (Karlsson et al. 2006) were machine harvested and collected in crates
where tubers fall from about 50 cm distance creating more impact damage. According to
previous research (Karlsson et al. 2006) relating high tuber calcium content with lower
internal defects, we expected that Superior would be more resistant to all internal defects
compared to Atlantic. Our observed differences in mean incidences of internal defects agree
with our expectations for hollow heart and back spot bruise, but not for internal brown spot.
As expected pitted scab incidence and severity were consistently higher in Atlantic than
Superior, with a greater difference among these cultivars when evaluated in a high disease
pressure field (Table 2.3). Mean pitted scab incidences in the standard field were 7.3, 28.0
and 10.0% for Atlantic versus 2.2, 4.0 and 3.3% for Superior for the 2009, 2010 and 2011
trials, respectively. Pitted scab incidence in the high disease field was 80.0 and 71.4% for
Atlantic compared to 3.3 and 1.5% for Superior in the 2011 and 2012 trials, respectively.
46
Pitted scab severity in the high disease pressure field was 4.6 and 2.8% for Atlantic and only
0.1 and 0% for Superior in the 2011 and 2012 evaluations, respectively. From several on-
farm trials, Pavlista (2005) reported a high incidence of scab, 24 - 35%. In addition, Haynes
et al. (2010) found an incidence of 75% for Atlantic and 55% for Superior across six years
and three locations. These previous reports agree with our results in the sense that Atlantic
and Superior differ in pitted scab tolerance and Atlantic is more susceptible.
In summary, we found that the parents were significantly different for most traits in the 2009
and 2010 evaluations (Tables 2 and 3). However, Atlantic and Superior were significantly
different only for specific gravity and pitted scab in the 2011 evaluations. It should be noted
that 2011 was an exceptionally good year for Wisconsin potato production with a very low
incidence of internal defects state-wide. Also, the trial in 2011 in the standard field included
few clones. We speculate that the lack of significant differences between the parents for
several traits in our 2011 standard field trial might be a combination of the environmental
conditions that favored low overall internal defects, the complete block design used only that
year that did not capture the spatial variability in the field as well as the incomplete block
design, and the small sample size. The difference between the complete block design and the
incomplete block design is that the latter takes into account the differences between smaller
areas of the field where the incomplete blocks or groups are grown (26 clones plus the
parents for this study); while, the complete block design takes into account only the
differences between larger areas where the complete blocks or replicates are grown (49
clones plus the parents for this study).
47
Performance of the reciprocal populations and influence of the maternal parent
For each trait, the mean values per reciprocal population were estimated and the significance
of their differences compared. The analysis of variance (ANOVA) and the analysis of
deviance (ANODE) were used to find significant differences between reciprocal crosses and
the significance level was set at p<0.05. The results indicate that there were significant
differences between reciprocal crosses for several traits including tuber calcium, tuber yield,
hollow heart incidence, internal brown spot incidence, enzymatic browning, and pitted scab
(Tables 2.4 and 2.5).
For tuber calcium, the differences between reciprocal populations were significant in three
years of evaluation and AxS had consistently lower tuber calcium compared to SxA
indicating that in average the progenies of Atlantic and Superior had tuber calcium
concentrations more similar to their female parent (Table 2.4). To our knowledge, there are
no previous reports about significant differences between reciprocal populations in mineral
content of potatoes. This relationship should be tested in several crosses using reciprocal
populations generated by crossing the high calcium cultivar Superior and other low calcium
cultivars.
For tuber yield, the reciprocal populations differed significantly only in 2009 but yields were
higher for AxS than SxA in all trials (Table 2.4). Large yield differences between reciprocal
populations in populations of Solanum tuberosum when intergroup hybrids were reported by
Sanford and Hanneman (1982). This study determined that the differences were mostly
associated to the maturity of the female parent. The maturity of Superior is medium (Rieman
1962) and Atlantic is medium-late (Webb et al. 1978); therefore, the differences in maturity
may not be the explanation for the difference in yield of our progenies.
48
The reciprocal populations also differed significantly for enzymatic browning, the visual
scale of chip color, chip color in agtron units and chip lightness only in one year of
evaluation, where population means were similar to the maternal parent (Table 2.4). Previous
studies have reported differences in chip color between reciprocal populations in diploid
populations (Lauer and Shaw 1970, Jakuczun and Zimnoch-Guzowska 2004).
On the other hand, the incidence of hollow heart was significantly different between
reciprocal populations only in one year of evaluation (Table 2.5). Similarly, significant
differences for internal brown spot between AxS and SxA was observed in 2010, a year when
internal brown spot was higher (Table 2.5). For pitted scab incidence, the differences
between reciprocal populations were significant in one year of evaluation in both standard
and high disease pressure fields (Table 2.5).
Significant differences between reciprocal populations were identified for tuber calcium,
tuber yield, enzymatic browning, visual rating of chip color, chip color in agtron units, chip
lightness, hollow heart, internal brown spot, and pitted scab incidence in the standard and the
high disease pressure field. However, a consistent relationship between reciprocal
populations was observed only for tuber calcium. This trait showed significant differences in
all years of evaluation and means closer to the maternal parent.
These results suggest that there might be some kind of parent-of-origin effects acting on tuber
calcium and quality traits of potato. In seed plants, cytoplasmic genes, chloroplast and
mitochondrial, are primarily maternally transmitted (Mogensen 1996). Reciprocal crosses
that show phenotypic differences have been used to study maternal effects (Roach and Wulff
1987). These effects have been observed for several traits in plant species such as maize
(Kollipara et al. 2002, Mach et al. 2011, Waters et al. 2011) and Arabidopsis (Duszynska et
al. 2013). In potato, differences between reciprocal populations have been previously
49
observed for several traits including male sterility, photoperiod, tuber initiation, tuber set,
vine senescence, tuber yield, flowering, male fertility and chip color (Lauer and Shaw 1970,
De la Puente and Peloquin 1968, Sanford and Hanneman 1979, 1982, Jakuczun and
Zimnoch-Guzowska 2004).
From a breeding standpoint, the identification of traits influenced by parent-of-origin effects
is not only an interesting phenomenon to study, but may also have practical applications
enabling breeders to predict which progenitors are more likely to produce offspring with the
desired phenotype when used as male or female parent.
From this study, we can conclude that for crosses between Atlantic and Superior, when
Atlantic is used as the maternal parent, the progenies will have higher tuber yields in average
than the reciprocal. On the other hand, if Superior is the maternal parent the progenies will
have less internal brown spot and higher tuber calcium. The reciprocal populations of
Atlantic x Superior could be of great value to understanding the genetic and epigenetic causes
of phenotypic differences in reciprocal populations. Future studies in reciprocal populations
will contribute to the better understanding of parent-of-origin effects, and the biological
reasons behind the differences between reciprocal populations. For this particular population,
the significant differences between reciprocal populations for tuber calcium, tuber yield and
internal brown spot incidence suggest that further analysis and conclusions generated by the
quantitative genetic analysis of these reciprocal populations should be made independently
for each population.
Segregation for tuber yield, specific gravity and chip quality
Tuber yield and specific gravity of both reciprocal populations had distributions that are
approximately normal (Figure 2.1). Most clones had intermediate or lower yield than the
50
parents and only few clones had higher yields than the parents. Atlantic was consistently
among the highest yielding clones. A similar distribution for yield was previously observed in
tetraploid populations by Bradshaw et al. (2008) in the 12601ab1 x Stirling population and
also by McCord et al. (2011a) in the Atlantic x B1829-5 population.
In the case of specific gravity, the distribution of the progeny was close to normal with
contrasting parental performances (Figure 2. 1). This distribution is similar to the distribution
presented by Haynes (2008) in a long-day adapted S. phureja x S. stenotomum diploid
population, and comparable to the distribution presented by McCord et al. (2011b) for a
tetraploid population even though their parental clones had similar values of specific gravity.
Performances for visual ratings of chip color, chip color measured as agtron units, chip
lightness (L values), chip redness (A values), and chip yellowness (B values) resemble
normal distributions with some clones with lighter chips than Atlantic (Figure 2. 1). The
normal distribution observed for the visual scale and agtron values is similar to reports from
other authors including Douches and Freyre (1994) that evaluated a diploid population of S.
tuberosum x S. chacoense; Haynes (2008) that used the 1 to 10 scale of the National Potato
Chip Institute Color Chart and evaluated a diploid S. phureja x S. stenotomum population;
and Bradshaw et al. (2008) that evaluated a tetraploid (12601ab1 x Stirling) population.
Segregation of tuber calcium
For tuber calcium, most progenies were intermediate between both parents. The performance
of both parents was contrasting for this trait (Figure 2. 1). Atlantic was consistently among
the clones with lowest tuber calcium and Superior was the highest. Tuber calcium has only
been investigated in commercial cultivars (Karlsson et al. 2006; Brown et al. 2012) and wild
51
potato germplasm (Bamberg et al. 1993), but not in segregating bi-parental populations. The
close to normal segregation for tuber calcium content in both reciprocal populations is
indicating the quantitative nature of this trait. These results are consistent with preliminary
evaluations of this population (Vega et al. 2006). This bell-shaped distribution is consistent
with a quantitative nature of the trait.suggesting that calcium concentration in the tuber may
be controlled by several genes as it has been found in soybean seeds (Zhang et al. 2009).
Segregation of internal defects
The distributions of the incidences of internal defects including hollow heart, internal brown
spot, and black spot bruise were skewed towards resistance in both reciprocal populations of
Atlantic and Superior (Figure 2. 2). These skewed distributions indicate that most progenies
have low incidence and few progenies have very high incidence of defects. These results
suggest that there might be some major resistance genes in the resistant parent Superior that
are segregating in these reciprocal populations. These skewed phenotypic segregations of
internal defects have been previously reported for the incidence of internal heat necrosis
(McCord et al. 2011a) and as a combined internal condition score (Bradshaw et al. 2008).
McCord et al. (2011a) found a skewed segregation towards lower incidence values in the
Atlantic x ‘B1829-5’ tetraploid population for internal heat necrosis (IHN). This defect has
been reported in eastern US where plants are subjected to heat stress during late season
(Sterret and Henninger 1997). On the other hand, Bradshaw et al. (2008) evaluated the
internal condition (IC), a visual score in a 1 to 9 scale, in the 12601ab1 x Stirling population.
This IC score was skewed towards higher scores (fewer defects); these results were also
similar to our results even though they performed a simultaneous evaluation of several
52
internal defects in one single score. The parents evaluated for internal heat necrosis by
McCord et al. (2011a), Atlantic and ‘B1829-5’, had contrasting values of incidence. On the
other hand, the parents evaluated for IC by Bradshaw et al. (2008), 12601ab1 and Stirling,
had similarly high internal scores which means few internal defects. However, both
populations studied by these authors showed skewed distributions. These results suggest that
the skewness of the segregation for internal defects does not depend on the difference
between parents and might be related to the genetics of these traits. The phenotypic
distribution is the result of the action of major and minor genes (Bernardo 2010); in other
words, the amount of variation explained by each gene varies and could be high or very low.
Skewed distributions observed in the distribution of a segregating population may indicate
the action of a major dominant gene; however, skewness can also exist due to the type of data
and its natural boundaries (Jansen 2007). In this study, the incidences of internal defects
expressed as proportions only can have values between 0 and 1; these values are a summary
of the presence or absence of defects in a certain number of tubers; therefore, the incidences
of internal defects follow a binomial distribution not a normal distribution. Therefore, the
skewness observed in the segregation of internal defects in this study are due to the type of
data used and may also indicate the presence of dominant major genes in the population.
Segregation of pitted scab
Pitted scab was evaluated in place of a general common scab evaluation because the
occurrence of pits reduces the value of potatoes while shallow lesions are still accepTable
2.for processing (Archuleta and Easton 1981). The evaluation of pitted scab was performed
by determining the proportion of tubers with pits evaluated as a measure of incidence and the
53
average number of pits per tuber in a plot as an indicator of severity. Pitted scab incidence
was expressed as percentages and plotted in a density plot (Figure 2.2). Incidence was
skewed towards low incidence resembling the distributions observed for incidence of internal
defects. The severity of pitted scab measured as the average number of pits per tuber had a
skewed distribution where low-values were more frequent (Figure 2.2). These values had to
be transformed by taking the square root of the values for normalization (Figure 2.2). In the
reciprocal populations of Atlantic (moderately susceptible) x Superior (moderately resistant)
most clones had lower incidences of pitted scab than Atlantic and some clones performed
similarly to Superior indicating that there is some potential for selecting clones with
improved resistance to the incidence and severity of common scab comparable to Atlantic.
In previous studies common scab severity was measured as an index from 0 to 5 (low to high)
in the Jacqueline Lee (susceptible) x MSG227-2 (tolerant) tetraploid population (Driscoll et
al. 2009). Using this severity index the population distribution was found to be skewed
towards susceptibility. In both our measurements, the proportion of tubers with disease and
the number of pits per tuber, the segregation was also skewed but towards resistance.
Therefore, we hypothesize that the skewed distribution observed for the incidence and
severity of pitted scab might be related to the mode of inheritance of scab tolerance;
specifically, the presence of major dominant genes. It is also important to note that the
incidence of pitted scab follows a binomial distribution because it is the summary of the
presence and absence of pits in a certain number of tubers.
54
CONCLUSIONS
By crossing Atlantic and Superior, we have produced reciprocal populations with clones that
have variable phenotypes for tuber calcium with values between the parents and some with
extremely high and extremely low values. Atlantic and Superior were different for most tuber
quality and pitted scab traits except for brown center or internal brown spot. Also, internal
brown spot was not different between the parents probably because this defect was generally
low; thus, this data will not be further used.
In brief, Atlantic and Superior contrast for several tuber quality traits such as yield, specific
gravity, chip quality, internal quality and pitted scab tolerance; therefore, the genetic
variation of the reciprocal populations generated by the cross of these two cultivars could be
used for the genetic study these traits.
This study demonstrates that Atlantic and Superior have contrasting phenotypes for tuber
yield, specific gravity, enzymatic browning, chip color using visual ratings, chip color as
measured in agtron units, colorimetric measurements of chip color, tuber calcium, incidence
of hollow heart and blackspot bruise, as well as incidence and severity of pitted scab. The
populations generated by reciprocal crosses between Atlantic and Superior show phenotypic
variation for all these traits. Most traits followed distributions that resemble a normal
distribution including the segregation for tuber yield, specific gravity, enzymatic browning,
chip color using visual ratings, chip color in agtron units, colorimetric measurements of chip
color and tuber calcium. However, skewed distributions were found for the incidence of
hollow heart, blackspot bruise, and pitted scab incidence and severity. The characteristics of
the reciprocal populations of Atlantic and Superior can be used to: (i) perform quantitative
genetic analyses such as broad-sense heritabilities and between-trait correlations for yield,
55
specific gravity, tuber calcium, tuber quality, chip quality, and internal quality, but also for
other traits these cultivars may differ; (ii) identify QTL for traits of commercial interest; (iii)
and identify desired clones that combine the desired traits of Atlantic and Superior to develop
new improved Atlantic-type chipping varieties. Further investigation of this population has
the potential to provide genetic and biological explanation for the variation of important
commercial desired traits of potato. This information could greatly enhance breeding efforts
to improve traits such as scab resistance, internal quality, and processing quality of potatoes.
BIBLIOGRAPHY
Andre, C.M., M. Ghislain, P. Bertin, M. Oufir, M. del R. Herrera, L. Hoffmann, J.F. Hausman, Y. Larondelle, and D. Evers. 2007. Andean potato cultivars (Solanum tuberosum l.) as a source of antioxidant and mineral micronutrients. Journal of Agricultural and Food Chemistry 55: 366-378. Archuleta, J.G. and G.D. Easton. 1981. The cause of deep- pitted scab of potatoes. American Potato Journal 58: 385-392. Bamberg, J.B., J.E Palta, L.A. Peterson, M. Martin, and A.R. Krueger. 1993. Screening tuber-bearing Solanum (potato) germplasm for efficient accumulation of tuber calcium. American Potato Journal 70: 219-226. Baritelle, A.L. and G.M. Hyde. 2003. Specific gravity and cultivar effects on potato tuber impact sensitivity. Postharvest Biology and Technology 29: 279-286. Bartholomew, E.T. 1914. A pathological and physiological study of the black heart of potato tubers. Thesis (Ph.D.) University of Wisconsin - Madison.
Bernardo, R. 2010. Breeding for Quantitative Traits in Plants. Minnnesota, US: Stemma Press. Bjor, T. and L. Roer. 1980. Testing the resistance of potato varieties to common scab. Potato Research 23: 33-47.
56
Bradshaw, J.E., C.A. Hackett; B. Pande; R. Waugh, and G.J. Bryan. 2008. QTL mapping of yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Theoretical and Applied Genetics 116: 193-211. Brown, C.R., K.G. Haynes, M. Moore, M.J. Pavek, D.C. Hane, S.L. Love, R.G. Novy, and J.C. Miller Jr. 2012. Stability and broad-sense heritability of mineral content in potato: Calcium and magnesium. American Journal of Potato Research 89: 255-261. Bryan, H.H. and N.L. Durre. 1968. White potatoes for marl soils in Dade County. In Proc. Fla. State Hort. Soc. 81: 159-163. Clough, G.H. 1994. Potato tuber yield, mineral concentration, and quality after calcium fertilization. Journal of the American Society of Horticultural Science 119: 175-179. Coles, G.D., J.P. Lammerink and A.R. Wallace. 1993. Estimating potato crisp colour variability using image analysis and a quick visual method. Potato Research 36: 127-134. De la Puente, F., and S.J. Peloquin. 1968. Male fertility of selected 24 chromosome S. tuberosum hybrids. American Potato Journal 45: 436-437. Douches, D.S. and R. Freyre. 1994. Identification of genetic factors influencing chip color in diploid potato (Solanum spp.). American Potato Journal 71: 581-590. Driscoll, J., J. Coombs, R. Hammerschmidt, W. Kirk, L. Wanner, and D. Douches. 2009. Greenhouse and field nursery evaluation for potato common scab tolerance in a tetraploid population. American Journal of Potato Research 86: 96-101. Duszynska, D., P.C. McKeown, T.E. Juenger, A. Pietraszewska‐Bogiel, D. Geelen, and C. Spillane. 2013. Gamete fertility and ovule number variation in selfed reciprocal F1 hybrid triploid plants are heritable and display epigenetic parent‐of‐origin effects. New Phytologist 198: 71-81. doi: 10.1111/nph.12147 Fisher, R.A. 1935. The design of experiments. First Edition. Oxford, England: Oliver and Boyd. Freyre, R., S. Warnke, B. Sosinski, D.S. Douches. 1994. Quantitative trait locus analysis of tuber dormancy in diploid potato (Solanum spp.). Theoretical and Applied Genetics 89: 474-480 Friedman, M. 1996. Food browning and its prevention: An overview. Journal of Agricultural and Food Chemistry 44: 631-653. Hackett, C.A., I. Milne, J.E. Bradshaw, Z. Luo. 2007. TetraploidMap for Windows: linkage map construction and QTL mapping in autotetraploid species. Journal of Heredity 98: 727-729. Hamilton, J.P., C.N. Hansey, B. R. Whitty, K. Stoffel, A.N. Massa, A. Van Deynze, W.S. De Jong, D.S. Douches, and C.R. Buell. 2011. Single nucleotide polymorphism discovery in elite North American potato germplasm. BMC Genomics 12: 302-312.
57
Harrison, M.D. 1962. Potato russet scab, its cause and factors affecting its development. American Journal of Potato Research 39: 368-387. Haynes, K.G. 2008. Heritability of chip color and specific gravity in a long-day adapted Solanum phureja x S. stenotomum population. American Journal of Potato Research 85: 361-366. Haynes, K.G., L.A. Wanner, C.A. Thill, J.M. Bradeen, J. Miller, R.G. Novy, J.L. Whitworth, D. L. Corsini, and B.T. Vinyard. 2010. Common scab trials of potato varieties and advanced selections at three U.S. locations. American Journal of Potato Research 87: 261-276. Henninger, M.R., S.B. Sterret, and K.G. Haynes. 2000. Broad-sense heritability and stability of internal heat necrosis and specific gravity in tetraploid potatoes. Crop Science 40: 977-984. Hughes, J.C. 1980. Potatoes I: Factors affecting susceptibility of tubers to damage. Span 23: 65-75. Hutchinson, C.M., J.M. White, D.M. Gergela, P.A. Solano, K.G. Haynes, R. Wenrich, and C.S. Lippi. 2003. Performance of Chip Processing Potato Varieties in Northeastern Florida. HortTechnology 13: 706-711. Jakuczun, H., and E. Zimnoch-Guzowska. 2004. Inheritance of glucose content in tubers of diploid potato families. American Journal of Potato Research 81: 359-370. Jansen, R.C. 2007. Quantitative Trait Loci in Inbred Lines. In Handbook of Statistical Genetics Vol. 1. eds. D.J. Balding, M. Bishop, C. Cannings, 595-596. Chichester, UK: John Wiley and Sons Ltd. Jansky, S.H. and Thompson D.M. 1990. Expression of Hollow Heart in segregating tetraploid potato families. American Potato Journal 67: 695-793. Jeong, J.C., H.C. Ok, O.S. Hur, C.G. Kim, C.S. Park,S.Y. Kim. 2005. Raw material characteristics affect enzymatic browning of pre-peeled potato tubers. Journal of the Korean Society for Horticultural Science 46: 246-249. Karlsson, B., J.P. Palta and P.M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. HortScience 41: 1213-1221. Kleinhenz, M.D. and J.P. Palta. 2002. Root calcium modulates the response of potato plants to heat stress. Physiologia Plantarum 115: 111-118. Kratzke, M.G. and Palta, J.P. 1986. Calcium accumulation in potato tubers: Role of the basal roots. HortScience 21:1022-1024. Kreuze, J.F., S. Suomalainen, L. Paulin, and J.P. Valkonen. 1999. Phylogenetic analysis of 16S rRNA genes and PCR analysis of the nec1 gene from Streptomyces spp. causing common scab, pitted scab, and netted scab in Finland. Phytopathology 89: 462-469.
58
Kollipara, K.P., I.N. Saab, R.D. Wych, M.J. Lauer, and G.W. Singletary. 2002. Expression profiling of reciprocal maize hybrids divergent for cold germination and desiccation tolerance. Plant Physiology 129: 974-992. Lærke, P.E., J. Christiansen, M.N. Andersen, and B. Veierskov. 2002. Blackspot bruise susceptibility of potato tubers during growth and storage determined by two different test methods. Potato Research 45: 187-202. Lambert, D.H. and R. Loria. 1989. Streptomyces scabies sp. nov., nom. rev. International Journal of Systematic Bacteriology 39: 387-392. Lauer, F., and R. Shaw. 1970. A possible genetic source for chipping potatoes from 40 F storage. American Potato Journal 47: 275-278. Levitt, J. 1942. A historical study of hollow heart potatoes. American Journal of Potato Research 19: 134-143. Li, L., M.J. Paulo, J. Strahwald, J. Lübeck, H.R. Hofferbert, E. Tacke, H. Junghans, J. Wunder, A. Draffehn, F. van Eeuwijk, and C. Gebhardt. 2008. Natural DNA variation at candidate loci is associated with potato chip color, tuber starch content, yield and starch yield. Theorethical and Applied Genetics 116: 1167-1181. Loria, R., R.A. Bukhalid, B.A. Fry, and R.R. King. 1997. Plant pathogenicity in the genus Streptomyces. Plant Disease 81: 836-846. Love, S.L., J.J. Pavek, A. Thompson-Johns, and W. Bohl. 1998. Breeding progress for potato chip quality in North American cultivars. American Journal of Potato Research 75: 27-36. Mach, J. 2011. Large-scale rna sequencing to identify maize genes with parent-of-origin expression effects. The Plant Cell Online 23(12): 4166-4166. Makower, R.U. 1964. Effect of nucleotides on enzymic browning in potato slices. Effect of nucleotides on enzymic browning in potato slices. Plant Physiology 39: 956-959. Matheis, G. 1983. Enzymatic browning of foods. Zeitschrift für Lebensmittel-Untersuchung und Forschung 176: 454-462. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011a. QTL mapping of internal heat necrosis in tetraploid potato. Theoretical and Applied Genetics 122: 129-142. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011b. Linkage Mapping and QTL Analysis of Agronomic Traits in Tetraploid Potato (Solanum tuberosum subsp. tuberosum). Crop Science 51:771-785. McGuire, R. and A. Kelman. 1986. Calcium in potato tuber cell walls in relation to tissue maceration by Erwinia carotovora pv. atroseptica. Physiology and Biochemistry 7: 401-406.
59
McKee, R.K. 1958. Assessment of the resistance of potato varieties to common scab. Potato Research 1: 65-80. McKee, R.K. 1963. Scab resistance of potato varieties. Plant Pathology 12: 106-109. Mogensen, H. L. The hows and whys of cytoplasmic inheritance in seed plants. American Journal of Botany 83:383-404. Navarro, F., B. Bowen, and J. Palta. 2012. Potato breeding program research update field year 2012. Proceedings of the Wisconsin Potato and Vegetable Growers Association (WPVGA). 36 pages. Nelson, D.C., M.C. Thoreson, D.A. Jones. 1979. The relationship between tuber size and hollow heart in Norgold Russet potatoes. American Journal of Potato Research 56: 329-337. Okeyo, J.A. and M.M. Kushad. 1985. Composition of four potato cultivars in relation to cold storage and reconditioning. HortTechnology 5: 250-253. Olsen, N.L., L.K. Hiller, and L.J. Mikitzel. 1996. The dependence of internal brown spot development upon calcium fertility in potato tubers. Potato Research 39: 165-178. Orr, P.H., and J.M. Sacks. 1992. Chipping responses of stored potatoes after handling. Transactions of the ASAE 35. Ozgen, S., B.H. Karlsson and J.P. Palta. 2006. Response of potatoes (cv. ‘Russet Burbank’) to supplemental calcium applications under field conditions. Tuber calcium, yield, and incidence of brown spot. American Journal of Potato Research 83: 195-204. Palta, J.P. 1996. Role of calcium in plant responses to stresses: Linking basic research to the solution of practical problems. HortScience 31: 51-57. Pavlista, A.D. 2005. Early-season applications of sulfur fertilizers increase potato yield and reduce tuber defects. Agronomy Journal, 97: 599-603. Parkin, K.L., and M.A. Schwobe. 1990. Effects of low temperature and modified atmosphere on sugar accumulation and chip color in potatoes (Solanum tuberosum). Journal of Food Science 55: 1341-1344. Pereira, A., G.C.C. Tai, R.Y. Yada, R.H. Coffin and V. Souza-Machado. Genetic advance for chip colour of potatoes. Euphytica 84: 133-138. Potato Genome Sequencing Consortium. 2011. Genome sequence and analysis of the tuber crop potato. Nature 475: 189-195. R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria, http://www.R-project.org.
60
Rex, B. L. and G. Mazza. 1989. Cause, control and detection of hollow heart in potatoes: A review. American Potato Journal 66: 165-183. Rieman, G.H. 1962. Superior: A new white, medium-maturing, sab resistant potato variety with high chipping quality. American Journal of Potato Research 39: 19-28. Roach D.A. and R. Wulff. 1987. Maternal effects in plants. Annual Review of Ecology and Systematics 18: 209-235. Rodriguez-Saona, L.E. and R. Wrostald. 1997. Influence of potato composition on chip color quality. American Journal of Potato Research 74: 87-106. Sanford, J.C., and R.E. Hanneman Jr. 1982. Large yield differences between reciprocal families of Solanum tuberosum. Euphytica 31: 1-12. Sapers, G.M., F.W. Douglas, A. Bilyk, A.F. Hsu, H.W. Dower, L. Garzarella, and M. Kozempel. 1989. Enzymatic browning in Atlantic potatoes and related cultivars. Journal of Food Science 54: 362-365. Scholte, K. and R.E. Labruyère. 1985. Netted scab: a new name for an old disease in Europe. Potato Research 28: 443-448. Sinha, N.K., J.N. Cash and R.W. Chase. 1992. Differences in sugars, chip color, specific gravity and yield of selected potato cultivars grown in Michigan. American Journal of Potato Research 69: 385-389. Sterret, S.B. and M.R. Henninger 1997. Internal heat necrosis in the midatlantic region-influence of environment and cultural management. American Journal of Potato Research 74: 233-243. Tawfik, A.A., M.D. Kleinhenz, and J.P. Palta. 1996. Application of calcium and nitrogen for mitigating heat stress effects on potatoes. American Journal of Potato Research 73: 261-273. Thompson, A. L., B.L. Farnsworth, N.C. Gudmestad, G.A. Secor, D.A. Preston, J.R. Sowokinos, and H. Hatterman-Valenti. 2008. Dakota Diamond: An exceptionally high yielding, cold chipping potato cultivar with long-term storage potential. American Journal of Potato Research 85: 171-182. Tzeng, K.C., A. Kelman, K.E. Simmons, and K.A. Kelling. 1986. Relationship of calcium nutrition to internal brown spot of potato tubers and sub-apical necrosis of sprouts. American Journal of Potato Research 63: 87-97. Vega, S.E., Bamberg, J.B. and Palta, J.P. 1996. Potential for improving freezing stress tolerance of wild potato germplasm by supplemental calcium fertilization. American Journal of Potato Research 73: 397-409.
61
Vega, S.E., J.P. Palta, and J.B. Bamberg. 2006. Exploiting cultivated germplasm to breed for enhanced tuber calcium accumulation ability [Abstract]. American Journal of Potato Research 83:136. Waters, A.J., I. Makarevitch, S.R. Eichten, R.A. Swanson-Wagner, C.T. Yeh, W. Xu, S.P., Schnablec, M.W. Vaughne, M. Gehringf, and N.M. Springer. 2011. Parent-of-origin effects on gene expression and DNA methylation in the maize endosperm. The Plant Cell Online 23: 4221-4233. Webb, R.E., D.R. Wilson, J.R. Shumaker, B. Graves, M.R. Henninger, J. Watts, J.A. Frank, and H.J. Murphy. 1978. Atlantic: A new potato variety with high solids, good processing quality and resistance to pests. American Journal of Potato Research 55: 141-145. Work, T.H., A.S. Kesis, and R.H. True. 1981. Factors determining potato chipping quality. Potato Chip/Snack Food Association standard color manual. Life sciences and agriculture experiment station University of Maine at Orono. Technical Bulletin 103, September 1981. Yates, F. 1936. Incomplete randomized blocks. Annals of Eugenics 7: 121-140. Zhang, B., P. Chen, A. Shi, A. Hou, T. Ishibashi, and D. Wang. 2009. Putative quantitative trait loci associated with calcium content in soybean seed. Journal of Heredity 100: 263-269.
62
TABLES Table 2.2.1. Information on the clones evaluated, experimental design and trial conditions in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations
Trials 2009 2010 2011 2011‡ 2012‡
Number of clones evaluated by reciprocal population AxS (189 clones)† 121 158 40 128 87
SxA (123 clones)† 64 107 9 86 78
Experimental design Design IBD IBD CBD IBD IBD
Plot size 8-hill 8-hill 8-hill 4-hill 4-hill
Replicates 3 2 or 3 3 3 3
Field conditions Field location C-field C-field C-field E-field E-field
Disease pressure standard standard standard high high
K application ( kg/ha) 170.1 172.4 181.4 181.4 181.4
N, P, K starter fertilizer at planting impregnated with platinum (249.5 kg/ha)
Season information Planting 05/01 04/29 04/28 05/10 05/18
Harvest (vines killed 10 days before harvest)
10/06 08/30 08/29 09/09 09/10
IBD=randomized incomplete block design, CBD=randomized complete block design, 4-hill= 4 tuber-seed pieces separated by 30.5 cm, 8-hill=8 tuber-seed pieces separated by 30.5 cm †In parenthesis are the total number of clones evaluated across all trials.The number of clones evaluated varies from season to season because not all the genotypes had enough seedto be used in the experiments. ‡Trials in the high disease pressure field.
63 Table 2. 2.2. Means by parent and population during 2009-2011 in the standard field for tuber calcium (TC), tuber yield (TY), specific gravity (SG), visual rating of chip color (CC), chip color as measured in agtron units (AG), chip lightness (L), chip redness (A), and chip yellowness (B). Evaluations made at Hancock Agricultural Research Station.
† Enzymatic browning and chip color were measured using a visual rating from 1 to 5 (light to dark).
‡Significant differences between parents in a year of evaluation (rows) at p< 0.05 as indicated by the F-test of the ANOVA are indicated in bold.
64 Table 2.2.3. Means by parent and population during 2009-2012 for the incidences of hollow heart (HH), internal brown spot (IBS), black spot bruise (BB), pitted scab in the standard field (PS), pitted scab in the high disease pressure field (PS-E), and severity of pitted scab in the high disease pressure field (SPS-E). Evaluations made at Hancock Agricultural Research Station.
†A square-root transformation was used for pitted scab severity data to normalize the data before the statistical test.
‡Significant differences between parents in a year of evaluation (rows) at p< 0.05 as indicated by the Chi-square test of the ANODE for incidence data, and the F-test of the ANOVA for pitted scab severity are indicated in bold.
§Estimated mean difference between reciprocal populations in ayear of evaluation (rows). Significance at p< 0.05 is indicated in bold.
65
Table 2. 2.4. Means by parent and population during 2009-2011 in the standard field for tuber calcium (TC), tuber yield (TY), specific gravity (SG), visual rating of chip color (CC), chip color as measured in agtron units (AG), chip lightness (L), chip redness (A), and chip yellowness (B). Evaluations made at Hancock Agricultural
Research Station.
AxS SxA
Trait units Year mean SD mean SD Contrasts‡
AxS vs. SxA p-value (T test)
TC µg/g 2009 153.6 50.6 159.1 52.0 -11.3 0.03
µg/g 2010 197.1 50.8 201.5 43.7 -12.3 0.03
µg/g 2011 268.9 65.1 291.8 62.2 -24.0 0.04
TY tons/ha 2009 51.5 15.7 48.5 14.2 5.01 0.0004
tons/ha 2010 37.2 13.5 36.4 12.3 0.55 0.36
tons/ha 2011 39.5 10.9 39.4 9.5 0.093 0.59
SG g/g 2009 1.074 0.009 1.073 0.01 -0.0009 0.4
g/g 2010 1.068 0.007 1.068 0.008 0.0007 0.1
g/g 2011 1.072 0.01 1.069 0.008 0.003 0.054
EB† 1 to 5 2010 2.09 0.99 1.92 0.85 0.17 0.002
1 to 5 2011 1.92 1.01 2.00 0.76 -0.07 0.65
AG agtron 2009 42.8 4.7 42.4 4.8 0.97 0.028
CC† 1 to 5 2009 3.5 0.6 3.5 0.6 -0.15 0.055
1 to 5 2010 3.3 0.9 3.4 0.8 -0.12 0.036
1 to 5 2011 2.1 0.9 2.0 0.8 0.07 0.54
L L values 2010 45.4 5.9 44.8 5.9 0.86 0.034
2011 55.8 4.4 55.8 3.8 -0.067 0.9
A A values 2010 9.4 1.5 9.4 1.3 -0.013 0.93
2011 3.3 0.8 3.5 1.0 -0.12 0.68
B B values 2010 18.9 2.8 18.8 2.5 -0.08 0.68
2011 23.0 1.2 23.2 1.1 -0.19 0.25 AxS =Atlantic x Superior, SxA=Superior x Atlantic, SD=standard deviation
†Enzymatic browning and chip color were measured using a visual rating from 1 to 5 (light to dark).
‡Estimated mean difference between reciprocal populations in a year of evaluation (rows). Significance at p< 0.05 is indicated in bold.
66 Table 2.2.5. Means by parent and population during 2009-2012 for the incidences of hollow heart (HH), internal brown spot (IBS), black spot bruise (BB), pitted scab in the standard field (PS), pitted scab in the high disease pressure field (PS-E), and severity of pitted scab in the high disease pressure field (SPS-E). Evaluations made at Hancock Agricultural Research Station.
pits/tuber 2012 0.54 1.29 0.58 0.05 0.034 0.67 AxS =Atlantic x Superior, SxA=Superior x Atlantic, SD=standard deviation
†A square-root transformation was used for pitted scab severity data to normalize the data before the statistical test.
‡Estimated mean difference between reciprocal populations in ayear of evaluation (rows). Significance at p< 0.05 is indicated in bold.
§Significance at p< 0.05 for the Chi-square test of a mean comparison between reciprocal populations using the Kruskal Wallis test.
67
FIGURES
Figure 2.2.1. Segregation for normally distributed traits including tuber yield, specific gravity, enzymatic browning, chip color as measured in agtron units, visual rating of chip color, and chip lightness, for chip redness, and chip yellowness, and tuber calcium in the reciprocal populations of Atlantic and Superior. Figures show density plots for one year of evaluation in Hancock, Wisconsin.
68 Figure 2.2.2. Segregation for traits with skewed distributions including the incidence of hollow heart, internal brown spot, black spot bruise, pitted scab in the standard field, pitted scab in the high disease pressure field, and severity of pitted scab in the high disease pressure field. Figures show density plots for one year of evaluation in Hancock, Wisconsin.
69
CHAPTER 3
Correlations and heritabilities of tuber quality, pitted scab and tuber calcium: Implications for selection of potatoes with improved tuber quality
ABSTRACT
A better understanding of the genetic basis of commercial traits of potato (Solanum
tuberosum) is a priority for potato breeders in order to supply growers with improved
varieties for the potato industry. Previous studies have related improved quality such as less
internal defects and reduced soft rot with field applications of calcium. However, little is
known about the genetic basis of tuber calcium content and its relationship to tuber quality. In
the present study, we estimated the correlations between tuber calcium, yield, chip quality,
internal quality and pitted scab in reciprocal populations derived from the cross of two
commercial varieties, Atlantic and Superior. Our results show significant genotype effects for
all traits evaluated indicating that the phenotypic differences observed between genotypes
have a genetic component. The broad-sense heritabilities of most traits differed among years
of evaluation and ranged between 0.19 to 0.85 for Atlantic x Superior and from 0.22 to 0.92
for Superior x Atlantic. Tuber calcium was negatively correlated to hollow heart, black spot
bruise and pitted scab. Nevertheless, tuber calcium was also negatively correlated to specific
gravity, yield, and chip yellowness; and positively correlated to chip color. Understanding the
correlations between tuber calcium and tuber quality traits as well as their heritabilities can
help potato breeders develop strategies for selection of new potato varieties with desired
phenotypes.
70
INTRODUCTION
Quantitative genetic analysis of phenotypic data helps bridge the gap between genes and
phenotypic traits. Some of the most important parameters measured in quantitative genetics
are the analysis of variances, heritabilities and correlations. The amount of variation is
measured and expressed as the variance. The analysis of variance (ANOVA) allows the
breeder to analyze data that depend on several kinds of effects, and these effects operate
simultaneously. This analysis helps breeders to decide which kinds of effects are important,
and to estimate these effects (Acquaah 2007). The assumptions of the data distribution to
apply ANOVA are the normality of the data, homogeneity of variances and independence of
observations (Falconer and Mackay 1996). Similarly, the analysis of deviance (ANODE)
measures the significance of the different effects influencing traits that follow a binomial
distribution (Jorgensen 1997). By partitioning phenotypic variance into its components one
can estimate the relative importance of the various determinants of the phenotype, in
particular the role of heredity versus environment (Falconer and Mackay 1996). Broad-sense
heritability can be defined as the ratio of the genotypic variance over the phenotypic variance
(Fehr 1987). The phenotypic variance is the sum of genotypic variance plus the residual
variance. The genotypic variance results from the differences among individuals and the
residuals variance results from the differences among genotypes caused by the failure to treat
each genotype exactly alike (Fehr 1987). The variance components can be estimated using a
multivariate analysis of variance (MANOVA), maximum likelihood (ML), restricted
maximum likelihood (REML), or Bayesian methods (Fisher 1918, Corbeil and Searle 1976,
Searle et al. 1992, Abney et al. 2000, Silva et al. 2013). The REML procedure is especially
important to deal with unbalanced datasets because by this procedure one can calculate the
71
estimators and their variances free of fixed effects (Corbeil and Searle 1976). In addition, the
estimation of genotypic correlations using REML has shown higher power of detection than a
MANOVA when there are missing data (Holland 2006).
Correlations are relationships between two or more variables or sets of variables and they
have three fundamental dimensions: significance, direction, and magnitude (Cohen and
Cohen 1983). Understanding the correlations between traits is very important information to
make decisions about selection methods in plant breeding. Correlations between characters
seriously complicate the prediction of response to phenotypic selection, because selection on
a particular trait produces not only a direct effect on the distribution of that trait in a
population, but also produces indirect effects on the distribution of correlated traits (Lande
and Arnold 1983). Correlations are due to pleiotropy, genes that affect two characters
simultaneously, but also due to linkage in crosses derived from divergent strains (Falconer
and Mackay 1996). In genetic studies, it is necessary to distinguish two causes of correlation
between characters: genetic and environmental (Falconer and Mackay 1996). Genetically
correlated traits respond to indirect selection pressures resulting from selection on other traits.
Indirect selection can be advantageous if the indirect character can be measured with more
accuracy than the primary trait (Wricke and Weber 1986).
Correlations among traits in potato have been reported for dormancy, emergence and dry
matter (Rashid and Carpena 1997); yield components (Ruiz de Galarreta et al. 2006); yield,
dry matter and plant characteristics (Felenji et al. 2011); tuber shape and weight (Bisognin et
al. 2012); yield, taste, tuber characteristics and mineral content (Flis et al. 2012). However,
most studies evaluating correlations only perform simple correlations using the phenotypic
data and evaluate one or few traits in a given population.
72
Previous research has demonstrated that in-season application of calcium increases tuber
calcium levels (Clough 1994, Kratzke and Palta 1986) and reduces the incidence of internal
defects (Tzeng et al. 1986, Olsen et al. 1996, Palta 1996, Kleinhenz et al. 1999, Ozgen et al.
2006, and Karlsson et al. 2006). For example, data from several cultivars collected over three
seasons showed that once the tuber calcium reaches more than 200ppm the incidence of
blackspot bruise is reduced dramatically (Karlsson et al. 2006). Furthermore, the incidence of
internal brown spot and sub-apical necrosis of sprouts was negatively correlated with tuber
peel calcium levels (Tzeng et al. 1986). These studies relating tuber calcium and internal
defects have been performed in a single variety or few varieties and not in segregating
populations. Therefore, the genetic basis of the relationship between calcium and internal
defects has not been explored yet.
In the present study we evaluated two reciprocal populations generated by crossing Atlantic,
a potato cultivar that has high specific gravity, high yield, light chip color, but high incidence
of internal defects and low calcium; with Superior, a potato cultivar that has low specific
gravity, low yield, dark chip color, but low internal defects and high calcium. The population
derived from these crosses contain clones that include all possible phenotypes in between the
parents for these traits. Our aims are to identify the main sources of phenotypic variance;
estimate the traits broad-sense heritabilities for tuber calcium, agronomic traits, internal
defects, chip quality and pitted scab; and estimate the degree of correlation between tuber
calcium and tuber quality traits. In addition, we discuss a strategy to select for reduced
internal defects and pitted scab by selecting for higher tuber calcium concentration, and also
discuss the effects of the correlated response to selection in case of highly correlated traits.
73
MATERIAL AND METHODS
Populations, location and experimental design
Two reciprocal bi-parental populations were used in this study, the Atlantic x Superior (AxS)
and the Superior x Atlantic (SxA) populations that include 189 and 123 clones, respectively.
The trials were grown in a standard field and in a high disease pressure field in the Hancock
Agricultural Research Station, Central Wisconsin, USA during 2009 to 2012 seasons. A
standard field was used to evaluate tuber calcium, agronomic traits, internal defects, chip
quality as well as pitted scab incidence. In addition, a high disease pressure field was also
used to evaluate pitted scab incidence and severity. Herbicide, fungicide and insecticide were
used as needed during the season and irrigation was scheduled every other day in the absence
of rain. The experimental design was a complete randomized block design in the standard
field in 2011. All other trials used an incomplete randomized block design with 3
replications. The clones evaluated were randomly sampled from the populations based on
seed availability. Groups of 28 clones that included the two parents and 26 randomly chosen
clones were formed and randomized within replicates for the incomplete randomized block
design.
Phenotypic evaluations
Tuber calcium was evaluated using the method described by Kratzke and Palta (1986) using
two 1mm-thick slices per tuber from 8 to 10 tubers, only the medullary tissue was removed,
oven-dried, ground and ashed. Tuber calcium was measured in µg/g dry weight using an
74
atomic absorption spectrophotometer (Varian SpectrAA 55B). The total tuber yield was
evaluated immediately after harvest and expressed in tons per hectare (ton/ha). Specific
gravity was determined by the following formula: SG=Weightair/(Weightair-Weightwater),
using a basket containing approximately 2 kilograms of tubers on a potato weigher PW-2050
(Weltech International, UK). Chip color was evaluated immediately after harvest and
included several measurements such as visual ratings of chip color (CC) in a scale from 1 to 5
(light to dark), agtron values (AG), chip lightness (L), chip redness (A), and chip yellowness
(B). The AG measurements were obtained with an M-300 reflectance spectrophotometer
(Fillper Magnuson, Rent, Nevada, US). The L, A and B measurements were obtained using a
scab in the standard field, pitted scab in the high disease field, and severity of pitted scab in
the high disease pressure field. The Spearman’s rank correlation analysis between years of
evaluation indicates that the correlations between ranks were positive and significant for the
visual rating of enzymatic browning, pitted scab incidence and severity evaluated in the high
disease pressure field. These traits showed correlations of 0.7, 0.69 and 0.67, for the two
years they were evaluated, respectively (Table 3.1). However, only hollow heart had
significant correlations between all years of evaluations; most traits showed significant
correlations between two out of three years of evaluation and for other traits such as
blackspot bruise all years were uncorrelated. This lack of correlation for blackspot bruise
incidence is an indication that performance of the populations were significantly different
between years and therefore further analysis and conclusions drawn from the evaluation of
this trait have to be made for each year independently. This analysis of correlations between
years of evaluation was also used to identify correlated years that could be pooled for further
global analysis.
79
The sources of variation were modeled according to the experimental design including:
Genotype, Replicate and Group for each year of evaluation and adding Year and Genotype x
Year interactions for pooled data. The ANOVA and ANODE indicate that all traits have
significant genotype effects (Tables 3.2 and 3.3). These results indicate that tuber calcium
and tuber quality traits have an important genetic component. In addition, significant replicate
and group effects were observed in most traits indicating that the variation within the field
had some influence on the phenotypic variation. Also, when data from multiple years was
analyzed, significant Genotype x Year interactions (GxY) were identified for all traits in at
least one of the reciprocal populations indicating that most quality traits of potato are
influenced by Genotype x Environment (GxE) interactions. Therefore, we can say that the
genotypes performance and their relative ranking varied from year to year. These results are
in agreement with previous reports; for example, Brown et al. (2012) also detected significant
GxE interaction for tuber calcium.
Previous research has also found significant genotypic variation in tetraploid segregating
populations for yield and specific gravity (Bradshaw et al. 2008; McCord et al. 2011a), chip
quality (Bradshaw et al. 2008), internal defects (Jansky and Thompson 1990; Henninger et al.
2000, McCord et al. 2011b) and common scab (Driscoll et al. 2009). Ours is the first report of
genetic variability for tuber calcium in a tetraploid bi-parental population. We have found
significant genotypic variation for tuber calcium and tuber quality traits that can be exploited
to select for cultivars with improved yield and specific gravity, improved chip quality and
internal quality, and tolerant to pitted scab.
80
Broad-sense heritabilities for tuber calcium, tuber quality traits, and pitted scab
The broad-sense heritabilities for tuber calcium and tuber quality traits were evaluated in
each reciprocal population per year of evaluation and using pooled data (Tables 3.2 and 3.3).
For the sake of discussion the values of heritabilities were classified as low (0.00-0.20),
moderately low (0.21-0.40), intermediate (0.41-0.60), moderately high (0.61-0.80), and high
(0.81-1.00). In general, the evaluated traits had heritabilities from moderately low to high
except for blackspot bruise that had a low heritability in 2011 (Tables 3.2 and 3.3). This
result indicates that most of the phenotypic variation for tuber calcium and tuber quality in
this population can be explained by differences between genotypes. Unfortunately, we cannot
separate the additive variance component from the total genotypic variance using bi-parental
populations but we know that the broad-sense heritability indicates the maximum value the
additive variance can reach. Therefore, the higher the broad-sense heritability, the most
reliable the conclusions we can draw about these traits.
The broad-sense heritability of tuber calcium ranged between moderately low to moderately
high with values between 0.44-0.61 for AxS and between 0.39-0.68 for SxA population
(Tables 3.2 and 3.3). Brown et al. (2012), estimated the broad-sense heritability of tuber
calcium concentration in 10 to 13 cultivars in a Tri-State, Western Regional, and Western
Regional Red/Specialty Trials obtaining values of 0.65, 0.37 and 0, respectively. Compared
to our results, the results reported were similar but we did not find values lower than 0.39 for
tuber calcium. The reason for obtaining not so low heritabilities in our case is that we are
using segregating populations of more than a hundred and fifty individuals whereas this other
study was performed in a limited number of unrelated clones. In addition, our reciprocal
populations are segregating for this trait showing a wide range of phenotypic variation
(Chapter 2). Furthermore, the experimental design used for our populations was an
81
incomplete randomized block design which is a powerful design to detect variation,
especially for traits like calcium that are influenced by soil conditions.
For tuber yield, the broad-sense heritability ranged between intermediate to moderately high
values between 0.66-0.85 for AxS and 0.51-0.82 for SxA (Tables 3.2 and 3.3). Haynes
(2001) reported a narrow-sense heritability for yield of 0.06-0.6 and a broad-sense heritability
of 0.56-0.76 (estimated using the reported variance components) in a diploid hybrid
population of S. phureja x S. stenotomum. These estimated broad-sense heritabilities were
comparable to our results.
For specific gravity, the broad-sense heritability was between low to moderately high with
values between 0.54-0.78 for AxS and 0.29-0.92 for SxA (Tables 3.2 and 3.3). A previous
report by Haynes et al. (2008) estimated the broad-sense heritability of specific gravity in a
diploid S. phureja x S. stenotomum population as 0.78, which is within the range of values for
broad-sense heritabilities that we found in our tetraploid populations. These results indicate
that our trials have been able to detect an important proportion of the genetic variation.
Compared to previous studies, we can say that a similar proportion of the phenotypic
variation for tuber yield and specific gravity is due to genotypic differences in diploid and
tetraploid potato.
The broad-sense heritability of enzymatic browning was moderately high or high with values
between 0.73-0.74 for AxS and 0.65-0.81 for SxA (Tables 3.2 and 3.3). A previous study
reported the broad-sense heritability value for enzymatic discoloration (ED), a similar
measurement to our enzymatic browning, was 0.84 for the CxD diploid population (Werij et
al. 2007). This value is close to the values estimated from our research. This result confirms
that using a visual rating scale to evaluate enzymatic browning was a good approach to detect
genetic variation and that the proportion of this genetic variation over the phenotypic
82
variation is similar in diploid and tetraploid potatoes. It is also interesting to note that
enzymatic browning had higher heritabilities than all other traits.
For chip color measured using a visual rating, the broad-sense heritability was between
intermediate and high with values between 0.42-0.83 for AxS and 0.54-0.83 for SxA (Tables
3.2 and 3.3). These results are similar to those reported by Haynes et al. (2008) who
estimated a narrow-sense heritability of 0.68 for this trait using a visual rating in a diploid
population. These results indicate that the evaluation of chip color using a visual rating has a
very important genetic component.
For the other measurements of chip color in the agtron and colorimetric scales, the values of
broad-sense heritability were similar as the visual rating for chip color. However, these
values were slightly lower for chip redness and chip yellowness indicating that these
parameters are much more influenced by the environment. The broad-sense heritability of
chip color in agtron values evaluated only the 2009 season was high, 0.81 for AxS and 0.82
for SxA (Tables 3.2 and 3.3). The broad-sense heritability of chip lightness was between
moderately high to high with values between 0.63-0.85 for AxS and 0.63-0.77 for SxA
(Tables 3.2 and 3.3). The broad-sense heritability of chip redness was between moderately
low to moderately high with values between 0.40-0.73 for AxS and 0.45-0.68 for SxA
(Tables 3.2 and 3.3). Chip yellowness broad-sense heritability was between moderately low
and high with values ranging between 0.29-0.71 for AxS and values between 0.62-0.77 for
SxA (Tables 3.2 and 3.3).
In addition, we estimated the estimated the broad-sense heritability of internal defects
including hollow heart and blackspot bruise. The broad-sense heritabilities for the incidence
of hollow heart were intermediate to moderately high. These values were between 0.59-0.74
for AxS and between 0.50-0.73 for SxA (Tables 3.2 and 3.3) which are slightly lower
83
compared to those reported by Henninger et al. (2000) for internal heat necrosis another type
of internal defect that had broad-sense heritabilities in the range of 0.83-0.88. The values of
broad-sense heritability observed for hollow heart indicate that an important proportion of the
phenotypic variance is due the genotype variation.
The broad-sense heritability of blackspot bruise incidence was between low and moderately
low within seasons with values between 0.19-0.30 for AxS and 0.22-0.41 for SxA (Tables
3.2 and 3.3). These values were low compared to previous reports; for example, Pavek et al.
(1993) found a narrow-sense heritability of 0.85 using a five-parent half diallel. Another
study reported broad-sense heritability of bruise as 0.73 and in this study bruise was
evaluated by subjecting tubers to impact damage using a drum in diploid hybrids (Hara-
Skrzypiec and Jakuczun 2013). In our study, we found low broad-sense heritability for bruise,
which is caused by the generally low incidence of this defect in our reciprocal populations.
The reason for the low defect may be that plots were hand-picked and that the incidence of
bruising was evaluated after regular handling of the plots without an external source of
impact damage. Thus, our harvest and handling method was not harsh enough to reveal the
genotypic variation between clones.
The broad-sense heritability of pitted scab incidence evaluated under standard conditions was
between moderately low to moderately high with values between 0.34-0.39 for AxS and 0.31-
0.65 for SxA (Tables 3.2 and 3.3). However, the broad-sense heritabilities for the incidence
and severity of pitted scab incidence evaluated under high disease pressure were higher.
These values were between 0.57-0.69 for AxS and 0.56-0.78 for SxA for the incidence of
pitted scab and values between 0.74-0.82 for AxS and 0.80-0.86 for SxA for the severity of
pitted scab (Tables 3.2 and 3.3). The broad-sense heritabilities estimated in the high disease
pressure field were in agreement with those reported previously by Haynes et al. (2010) for
84
the proportion of scabby tubers, 0.3 - 0.8; and for the severity of common scab measured as
and an area index, 0 - 0.78, and a lesion index, 0.49 - 0.90, for a set of 17 to 23 tetraploid
cultivars and advanced selections that represented a wide variety of genotypes. Our study
instead used segregating populations with all degrees of scab tolerance and obtained a similar
proportion of variance due to genotypic variation in the high disease pressure field. A
possible explanation of the low heritability for the pitted scab evaluation in the standard field
is that the conditions were variable and usually mild; therefore, the differences between the
progenies for scab tolerance were not evidenced or at least were not as high as the
environmental and residual variance.
In summary, most traits showed broad-sense heritabilities between moderately low and high,
except for black spot bruise that showed low broad-sense heritability in the 2011 evaluation.
Therefore, the phenotypic variation for all traits evaluated in this study has an important
genetic component.
Broad-sense heritability of internal quality traits and pitted scab in mild and harsh
environments
Based on the observed broad-sense heritabilities, we can say that for most traits the
proportions of the observed phenotypic variances that can be explained by the genotypic
differences is important with the exception of black spot bruise in 2011. The low broad-sense
heritability of black spot bruise is explained by the low incidence caused by hand-harvesting
and a gentle post-harvest manipulation that does not reveal the differences in impact
resistance between genotypes. Therefore, the variability in the amount of impact each plot
received may have obscured genotypic differences when damage is generally low. Results of
85
our study suggest that future studies of black spot bruise should be made subjecting each plot
to a uniform external source of impact that can reveal differences to withstand impact damage
among cultivars. Similarly, the higher heritabilities of the pitted scab evaluations in the high
disease pressure field indicate that clones have to be exposed to a harsh environment in order
to reveal the genotypic variation for tolerance. These results again suggest that future studies
trying to find variability for common scab tolerance should be made in fields where the
disease is known to occur or by inoculating the pathogen in the soil. Previous research has
used greenhouse evaluations where Streptomyces spp. were inoculated (Wiersema 1970,
Driscoll et al. 2009). These greenhouse evaluations were correlated to field grown
evaluations (Driscoll et al. 2009).
Correlations between tuber calcium and internal quality
In order to capture most genetic variation, we are using the best linear unbiased predictions to
estimate genetic correlations between traits. Spearman’s rank correlations were estimated for
all traits evaluated in this population in order to predict correlations regardless of the type of
trait distribution (Table 3.4). Tuber calcium was negatively correlated to the incidence of
hollow heart and black spot bruise in both reciprocal populations. The incidence of hollow
heart was correlated to tuber calcium with rs = -0.38 and rs = -0.35 in AxS and SxA
populations, respectively. This correlation was significant only in AxS population. This
negative genetic correlation between tuber calcium and hollow heart supports our hypothesis
that the incidence of hollow heart is reduced in cultivars with higher tuber calcium. The
development of hollow heart is caused by rapid growth after a variety of abiotic stresses such
as water and nutritional stress (Rex and Mazza 1989, Mc Cann and Stark 1989). Under rapid
growth situations, cell wall extension is correlated to growth rate (Taiz 1984). Cell wall
86
extension requires calcium as a component of cell walls. Therefore, if an adequate amount of
calcium is available, cell expansion and growth can adjust to the faster rate; otherwise, cell
necrosis occurs generating hollow heart.
Also, tuber calcium was negatively correlated to the incidence of black spot bruise with rs = -
0.79 in AxS and rs = -0.63 in SxA. This negative correlation between tuber calcium and black
spot bruise supports our hypothesis that impact damage is low in cultivars with high tuber
calcium. The color of blackspots is a result of phenols oxidation to the black pigment melanin
mediated by polyphenol oxidase (PPO) (Matheis 1987). Lærke et al. (2002a, 2002b)
explained the development of blackspots by the disruption of intracellular membranes as an
immediate effect of the impact; and the consequent contact between the PPO located in the
amyloplasts and its substrates located in the vacuole. This means that the structural properties
of the tuber cells are crucial for its resistance to blackspot formation caused by impact.
Calcium contributes to cellular structural properties such as stiffening of cell walls and cell
wall strength (Taiz 1984); and therefore, to the resistance to impact damage. Karlsson et al.
(2006) also found a negative relationship between tuber calcium content and the incidence of
black spot bruise when tubers of the same variety were subjected to calcium treatment. Our
results support the notion that there is a negative correlation between tuber calcium content
and the incidence of black spot bruise in a segregating population where supplemental
calcium has not been applied.
Correlations between tuber calcium and pitted scab
An early study by Horsfall et al. (1954) reported calcium in tuber peelings to be positively
correlated to common scab severity. Also, Davis et al. (1974) found that after gypsum
application to potatoes, calcium levels in tuber peel were positively correlated with scab
87
susceptibility. Later, Lambert and Manzer (1991) concluded that high calcium in the
periderm was a consequence rather than a cause of increased scab. Our study relates common
scab, specifically in terms of pitted scab lesions, incidence and severity with the natural
variation for tuber calcium concentration without external calcium application. Our results
indicate a negative correlation between tuber calcium and pitted scab incidence and severity
under high disease pressure (Table 3.4). Tuber calcium was negatively correlated to pitted
scab incidence and severity in the high disease pressure field with rs = -0.43 for both traits in
both reciprocal populations but significant only in SxA. However, tuber calcium was
positively correlated to pitted scab incidence in the standard field with rs = 0.21 and rs = 0.22
in AxS and SxA, correspondingly. However, these correlations were not significant.
Consequently, selecting for high tuber calcium under high disease pressure may be a good
approach to select varieties with resistance to pitted scab. As stated above, reports of higher
periderm calcium as a consequence of increased scab infection had been reported (Lambert
and Manzer 1991). We tested calcium in the medullary tissue of the tuber instead of the
periderm to avoid any change in the composition after contact with the pathogen. From our
experience evaluating calcium concentration in tubers, in general we expect that cultivars
with higher medullar calcium also have higher periderm calcium (Kleinhenz et al. 1999).
Higher tuber calcium contributes to strengthen cell wall structure and cell health (Palta 1996)
which could explain increased tolerance to Streptomyces spp. infection. In support of this
idea, higher tuber calcium was reported to provide higher tolerance to pathogens such as
Pectobacterium spp. that cause soft rot of potato tubers (McGuire and Kelman 1986).
88
Correlations of tuber calcium with agronomic traits, chip color measurements, and
enzymatic browning
Tuber calcium was negatively correlated to yield and specific gravity in both reciprocal
populations (Table 3.4). The correlation between tuber calcium and tuber yield was rs = -0.71
and rs = -0.60 in the AxS and SxA populations, respectively; and the correlation between
tuber calcium and specific gravity was rs = -0.66 and rs = -0.61 in the AxS and SxA
populations, correspondingly. These correlations were significant except for tuber calcium
and tuber yield in AxS. These results indicate that selecting cultivars for higher tuber calcium
may reduce yield and specific gravity in these populations if their correlation is due to
linkage. We have not found other evidence that indicates a cause-effect relationship.
In addition, the correlation between tuber calcium with the various measurements of chip
color and enzymatic browning varied between reciprocal populations. First, tuber calcium
was negatively correlated to agtron values (rs = -0.26 in AxS and SxA), chip lightness (rs = -
0.48 and rs = -0.53 in AxS and SxA), chip yellowness (rs = -0.22 and -0.37 in AxS and SxA)
but positively correlated to the visual rating of chip color (rs = 0.52 and rs = 0.55 in AxS and
SxA), and chip redness (rs = 0.43 and rs = 0.50 in AxS and SxA), and enzymatic browning (rs
= 0.55 and rs = 0.42 in AxS and SxA) in both reciprocal populations (Table 3.4). These
results indicate that selecting for high calcium, we may be indirectly selecting for darker
chips, reddish color, and higher enzymatic browning.
These correlations between tuber calcium and tuber quality could be explained by pleiotropy,
some genes controlling tuber calcium also control tuber quality, or linkage, some genes
controlling these traits are located closely in the genome. Another explanation could be a
cause-effect relationship supported by the previous reports on tuber calcium improving
internal quality of potato.
89
Correlations between different measurements of chip color and pitted scab
Chip color is the trait of most interest in a chipping variety. Therefore, it is important to know
what type of chip color measurement may work better for selection. We have evaluated
several types of chip color measurements including a visual rating of chip color, chip color in
agtron units, chip lightness, chip redness, and chip yellowness. The visual rating of chip color
was significantly correlated to most measurements of chip quality indicating that the visual
rating summarizes all the chip quality properties evaluated (Table 3.4). The visual rating of
chip color was negatively correlated to chip color in agtron units (rs = -0.08), chip lightness
(rs = -0.79) and chip yellowness (rs = -0.65); and positively correlated to chip redness (rs =
0.76), all significant at p <0.05. The negative correlation between the visual rating with chip
color in agtron units, chip lightness and chip yellowness is expected because higher visual
scores indicate darker chip color while higher agtron and L values indicate lighter chip color.
Furthermore, higher chip redness indicates the presence of dark spots in the chips and
therefore higher visual scores.
Chip color was also positively correlated to enzymatic browning, rs = 0.26 in AxS and rs =
0.25 in SxA, significant only for AxS population (Table 3.4). The visual color scale of
enzymatic browning is not a chip trait but enzymatic browning may influence the color of the
slice before it enters to the fryer and thus indirectly influence chip color.
The recommended measurement of chip color for selection is the evaluation of chip lightness,
because it is an objective measurement estimated by a colorimeter and this parameter was
significantly correlated to the other chip color measurements. A second option, when the
colorimeter is not available is the visual rating of chip color that is also significantly
correlated to all other chip color measurements. This rating does not require specialized
90
equipment and it is not as time consuming. Nevertheless, this visual rating is challenging due
to its subjective nature requiring training and experience of the evaluator for consistency.
Different types of measurements have been used to study incidence and severity of common
scab. Scab incidence has been studied by classifying tubers as diseased or healthy (Hiltunen
et al. 2005). Scab severity was evaluated by the type of lesion: superficial, raised or deep; and
percentage of scab lesion (Hiltunen et al. 2005). In addition, a visual rating or scab index has
also been used (Driscoll et al. 2009). A thorough review of common scab of potato was
presented by Dees and Wanner (2012). In our study pitted scab incidence was measured as
the proportion of tubers with pits and severity measured as the average number of pits per
tuber. All measurements of pitted scab were highly correlated in both reciprocal populations
(Table 3.4). The incidence of pitted scab in the standard field was significantly correlated to
the incidence of pitted scab in the high disease pressure field with rs = 0.74 in AxS and rs =
0.70 in SxA; and the severity of pitted scab in the high disease pressure field with rs = 0.64 in
AxS and rs = 0.62 in SxA. However, the highest correlations were observed between pitted
scab incidence and severity under high disease pressure, rs =0.93 in AxS and rs =0.95 in SxA,
suggesting that the mechanisms of resistance for incidence and severity of pitted scab might
be controlled by the same or related genetic mechanisms of plant-pathogen interaction.
Correlations between tuber quality traits and pitted scab with tuber calcium:
implications for selection
The relationship between tuber calcium and internal quality reported by previous studies
(Tzeng et al. 1986, Olsen et al. 1996, Palta 1996, Ozgen et al. 2006, Karlsson et al. 2006)
suggested that tuber calcium may be a good candidate for indirect selection for internal
91
quality without having to cut large numbers of tubers. The results presented in our study can
be used to answer the question that if there is a genetic correlation between tuber calcium and
internal defects and pitted scab, we can use this relationship to select for higher tuber calcium
and indirectly select for internal quality. This indirect selection would reduce the number of
tubers that have to be used in the evaluation and eliminate the need to cut several tubers. We
have found that high tuber calcium is negatively correlated to internal defects incidence such
as hollow heart and blackspot bruise, as well as yield, specific gravity, and chip lightness. A
consequence of these correlations would be that selecting for high tuber calcium may
generate cultivars with low incidence of internal defects but with low yield, low specific
gravity and dark chip color. One of the causes for correlated traits is linkage between the
genes controlling them or same gene controlling both traits (pleiotropy) and the other one is
linkage (Falconer and Mackay 1996). The negative correlation between tuber calcium and
hollow heart, blackspot bruise and incidence and severity of pitted scab are consistent with
previous studies that relate high tuber calcium concentration as a consequence of
supplemental calcium application with improved tuber quality. Our results suggest that
genotypes with naturally high tuber calcium concentrations are less likely to suffer from
internal defects. Tuber calcium appears to have a very important role in the prevention of
hollow heart, blackspot bruise, and pitted scab. In addition, the application of supplemental
calcium may reduce internal defects by increasing tuber calcium content. Future studies in the
reciprocal populations of Atlantic and Superior should determine whether there is genetic
variation for the increase in tuber calcium as an effect of calcium supplementation.
Light chip color, high specific gravity and high yield are desirable traits in a chipping variety;
however, these must be accompanied by satisfactory internal quality and disease resistance.
92
We can avoid obtaining good chipping varieties that will not be adopted due to internal
quality and scab issues by selecting for all these traits simultaneously at early stages of the
breeding process in large populations.
Relationship of tuber quality traits and pitted scab with tuber calcium and tuber size
The relationship between internal defects and tuber calcium was evaluated in both reciprocal
populations using a logistic regression that related tuber calcium concentration and tuber size
classified in small, medium and large categories with the incidence of internal defects
expressed as proportions (Figure 3.1). The datasets used in this analysis were the same used
to study the genetic correlations between traits. The probability of hollow heart was similar
for small and medium tubers but significantly higher in genotypes with large tubers. Also, the
probability of hollow heart decreased at higher calcium concentrations in all tuber size
categories (Figure 3.1) following the equations indicated in Table 3.5. In addition, the
probability of blackspot bruise was also similar for small and medium tubers but significantly
higher in genotypes with large tubers. This relationship may be explained because large
tubers have larger surface area that can be affected by impact; however, Skrobacki et al.
(1989) demonstrated that larger mass was not correlated to blackspot bruise. As well, the
probability of blackspot bruise decreased at higher calcium concentrations in all tuber size
categories (Figure 3.1) following the equations indicated in Table 3.5. These results are in
agreement with previous reports that found the incidences of internal defects can be reduced
by tuber calcium concentration (Karlsson et al. 2006) and increased with tuber size (Jansky
and Thompson 1990). The plots for the relationship of calcium and pitted scab incidence and
severity reveal differences between reciprocal populations for these traits, a higher positive
effect of calcium on tolerance to pitted scab is observed in Superior x Atlantic (Figure 3.2).
93
In our study, we are calculating this relationship in terms of a numerical estimation of the
effects of each variable in these reciprocal populations. Only internal defects were evaluated
by their relationship with both tuber calcium and tuber size, all other traits were tested for
their relationship with tuber calcium. A logistic regression was also applied to determine the
probability of pitted scab in relation to tuber calcium. The results of these analyses indicated
that higher tuber calcium decreased the probability of getting tubers with pitted scab under
the high disease field conditions (Figure 3.2); however, the probability of having tubers with
pitted scab in the standard field was positively but not significantly correlated with tuber
calcium (Table 3.4). Linear regressions were used to evaluate the relationship between tuber
calcium with severity of pitted scab (square-root transformed). Pitted scab severity decreased
at higher calcium concentrations in the tuber (Figure 3.2). In addition, total yield, specific
gravity, chip lightness and chip yellowness showed a decrease at higher calcium
concentrations in the tubers (Figures 3.3 and 3.4). These results indicate that, in this
population, clones with higher tuber calcium tend to have low tuber yield and specific gravity
as well as bad chip quality.
Selection of promising clones from the reciprocal populations of Atlantic x Superior
Promising clones were selected based on their performance using a culling method that
selected first for chip color, then for incidence of hollow heart and other internal defects, next
for common scab, and finally for total yield. Clones were ranked for each trait sequentially.
Only clones with lighter chip color compared to Atlantic, less hollow heart than Atlantic; and
less than 10% incidence of pitted scab in the high disease pressure field were selected as
promising clones. Finally, clones with yields less than 35 tons/ha were excluded. Four
promising clones were identified using the pooled data with the highest heritability, the same
94
used for the analysis of correlations (Table 3.6). Two promising clones were identified in the
Atlantic x Superior population (B-032 and B-167), and two more from the Superior x
Atlantic population (C-062 and C205). All clones selected as promising have L values higher
than 50; values lower than 40 are considered unacceptable (Parkin and Shwobe 1981); and
higher tuber calcium compared to Atlantic (Table 3.6). In addition, their external appearance
and size was comparable to the parents (Figure 3.5). The identification of these promising
clones supports our initial hypothesis that we could improve internal quality by crossing
Atlantic with the high calcium cultivar Superior and obtain Atlantic-like cultivars. However,
these clones are not as high yielding as Atlantic due to the high correlation between tuber
yield and tuber calcium. Higher number of progenies of the reciprocal populations of Atlantic
and Superior might be used to find recombinant genotypes that combine the good tuber
quality of Superior and the high yield and good chipping quality of Atlantic.
CONCLUSIONS
From the evaluation of the reciprocal populations of Atlantic x Superior for broad-sense
heritability, genetic correlations and the estimation of relationships between tuber calcium,
tuber quality and pitted scab tolerance, we can draw the following conclusions:
1. In this study, we demonstrated that the phenotypic variation for tuber calcium, tuber
quality traits, and pitted scab observed in the reciprocal populations of Atlantic and
Superior has an important genetic component due to the significant genotypic effects
for all traits. These genotypic variations can be exploited to select for cultivars with
95
improved yield and specific gravity, improved chip quality and internal quality, and
tolerance to pitted scab.
2. The data also indicate that tuber quality traits, pitted scab tolerance and tuber calcium
are influenced by environmental effects, including the year of evaluation and the
spatial distribution within the field as well as significant Genotype x Environment
(year) interactions for all traits in at least one the reciprocal populations.
3. The broad-sense heritabilities of most traits varied from year to year and ranged
between 0.19 to 0.85 for Atlantic x Superior and from 0.22 to 0.92 for Superior x
Atlantic.
4. The highest broad-sense heritabilities were observed for tuber yield, followed by
pitted scab severity and incidence and enzymatic browning.
5. Black spot bruise had the lowest broad-sense heritabilities. This can be explained by
the low overall blackspot bruise incidence caused by hand-harvesting and a gentle
post-harvest manipulation that did not reveal the differences in impact resistance
between genotypes.
6. Pitted scab evaluations of incidence and severity in the high disease pressure field had
higher heritabilities compared to the pitted scab incidence evaluation in the standard
field indicating that clones have to be exposed to a harsh pathogen pressure in order to
reveal the genotypic variation for tolerance.
7. Most traits showed significant correlations between two out of three years of
evaluation and other traits such as blackspot bruise were uncorrelated for all years.
Only for hollow heart, the rankings of all years of evaluations were correlated.
Correlated years were pooled for further global analyses.
96
8. Negative Spearman’s rank correlations between the best linear unbiased predictions
(BLUP) for hollow heart, blackspot bruise as well as pitted scab incidence and
severity with tuber calcium were identified. These genetic correlations suggest that
some genes that control these traits might be the same (pleiotropy), closely linked
(linkage), or a cause-effect relationship.
9. Tuber calcium was also negatively correlated to yield, specific gravity and chip
quality. The use of tuber calcium for indirect selection of improved tuber quality may
be performed with caution by evaluating large populations in order to identify the
desired phenotypes.
10. Most chip color measurements were significantly correlated. Our results show that
chip lightness and the visual rating of chip color are correlated to all other
measurements of chip color suggesting that these traits can be used to select for chip
quality.
11. High correlations were observed between pitted scab incidence and severity under
high disease pressure suggesting that the mechanisms of tolerance to both these traits
might be controlled by the related genetic mechanisms of plant-pathogen interaction.
12. In addition, four promising clones that have good chipping quality, good internal
quality, reduced pitted scab incidence and severity as well as acceptable yield
compared to Atlantic were selected from the reciprocal populations.
BIBLIOGRAPHY
Abney, M., M.S. McPeek and C. Ober. 2000. Estimation of variance components of quantitative traits in inbred populations. The American Journal of Human Genetics 66: 629-650.
97
Acquaah, 2007. Principles of plant genetics and breeding. Oxford, UK: Blackwell Publishing. Bates, D., M. Maechler and B. Bolker. 2011. Linear mixed-effects models using S4 classes. R version 0.999375-42. http://lme4.r‐forge.r‐project.org/. Accessed on November 2012. Bates, D. 2012. Linear mixed model implementation in lme4. University of Wisconsin – Madison. http://cran.r‐project.org/web/packages/lme4/vignettes/Implementation.pdf. Accessed on November 2012. Bernardo, R. 2010. Chapter 10: Best Linear Unbiased Prediction. In Breeding for Quantitative Traits in Plants. Second Edition, 211–238. Minnnesota, US: Stemma Press. Bisognin, D.A, M.H. Rigão, S.J. Lopes, and L. Storck. 2012. Heritability and correlation among potato tuber traits. Crop Breeding and Applied Biotechnology 12: 215-219. Bradshaw, J.E., C.A. Hackett, B. Pande, R. Waugh, G.J. Bryan. 2008. QTL mapping of yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Theor Appl Genet 116: 193–211. Brown, C.R., K.G. Haynes, M. Moore, M.J. Pavek, D.C. Hane, S.L. Love, R.G. Novy, and J.C. Miller Jr. 2012. Stability and broad-sense heritability of mineral content in potato: Calcium and magnesium. American Journal of Potato Research 89: 255-261. Cohen, J. and P. Cohen. 1983. Applied Multiple Regression/Correlation Analysis for the Behavioral Sciences (Second Edition). Hillsdale, NJ: Erlbaum. Clough, G.H. 1994. Potato tuber yield, mineral concentration, and quality after calcium fertilization. Journal of the American Society of Horticultural Science 119: 175-179. Corbeil, R.R., and S.R. Searle. 1976. Restricted maximum likelihood (REML) estimation of variance components in the mixed model. Technometrics 18: 31-38. Davis, J.R., J.E. Garner, and R.H. Callihan. 1974. Effects of gypsum, sulfur, terraclor, and terraclor super-X for potato scab control. American Journal of Potato Research 51: 35-43. Dees, M.W., and L.A. Wanner. 2012. In Search of Better Management of Potato Common Scab. Potato Research 55: 249-268. Driscoll, J., J. Coombs, R. Hammerschmidt, W. Kirk, L. Wanner, and D. Douches. 2009. Greenhouse and field nursery evaluation for potato common scab tolerance in a tetraploid population. American Journal of Potato Research 86: 96–101. Fahmeir L., and G. Tutz, 1994. Multivariate Statistical Modelling Based on Generalized Linear Models. München, Germany: Springer Series in Statistics. Falconer, D.S., and T.F.C. Mackay. 1996. Introduction to Quantitative Genetics, Fourth Edition. Essex, UK: Longman, Harlow.
98
Fehr, W. 1987. Principles of Cultivar Development, vol1: Theory and Techniques. Macmillan Publishing Co., New York, USA. Felenji, H., S. Aharizad, G.R. Afsharmanesh, and M. Ahmadizadeh. 2011. Evaluating correlation and factor analysis of morphological traits in potato cultivars in fall cultivation of jiroft area. American-Eurasian Journal of Agricultural & Environmental Sciences 11: 679-684. Fisher R.A. 1918. The correlation between relatives on the supposition of Mendelian inheritance. Transactions of the Royal Society of Edinburgh 52: 399–433 Flis, B., E. Zimnoch-Guzowska, and D. Mankowski. 2012. Correlations among yield, taste, tuber characteristics and mineral contents of potato cultivars grown at different growing conditions. Journal of Agricultural Science 4: 197-207. Haynes, K. G. 2001. Variance components for yield and specific gravity in a diploid potato population after two cycles of recurrent selection. Potato Research 78: 69-75. Haynes, K. G. 2008. Heritability of chip color and specific gravity in a long-day adapted Solanum phureja–S. stenotomum population. American Journal of Potato Research 85: 361-366. Haynes, K. G., L.A. Wanner, C.A. Thill, J.M. Bradeen, J. Miller, R.G. Novy, J.L. Whitworth, D.L. Corsini and B.T. Vinyard. 2010. Common scab trials of potato varieties and advanced selections at three US locations. American Journal of Potato Research 87: 261-276. Hara-Skrzypiec, A., and H. Jakuczun. 2013. Diploid Potato Hybrids as Sources of Resistance to Blackspot Bruising. American Journal of Potato Research, 1-9. DOI 10.1007/s12230-013-9319-y. Henderson, C.R. 1974. General flexibility of linear model techniques for sire evaluation. Journal of Dairy Science 57: 963–972. Henninger, M.R., S.B. Sterrett, and K.G. Haynes. 2000. Broad-sense heritability and stability of internal heat necrosis and specific gravity in tetraploid potatoes. Crop science. 40: 977–984.
Hiltunen, L.H., A. Weckman, A. Ylhäinen, H. Rita, E. Richter, and J.P.T. Valkonen. 2005. Responses of potato cultivars to the common scab pathogens, Streptomyces scabies and S. turgidiscabies. Annals of Applied Biology 146: 395–403. Holland, J.B. 2006. Estimating genotypic correlations and their standard errors using multivariate restricted maximum likelihood estimation with SAS Proc MIXED. Crop Science 46: 642-654. Horsfall, J.G., J.P. Hollis, and H.G.M. Jacobson. 1954. Calcium and potato scab. Phytopathology 44: 19-24.
99
Jansky, S.H. and Thompson D.M. 1990. Expression of Hollow Heart in segregating tetraploid potato families. American Potato Journal 67: 695–793. Jorgensen, B. 1997. The theory of dispersion models.Monographs on Statistics and applied probability 76. London, UK: Chapman and Hall. Karlsson, B.H., J.P. Palta and P.M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. Hort Science 41: 1213-1221. Kleinhenz, M.D., J.P. Palta, C.C. Gunter, and K.A. Kelling. 1999. Impact of source and timing of calcium and nitrogen applications on Atlantic potato tuber calcium concentrations and internal quality. Journal of the American Society for Horticultural Science 124: 498–506. Kratzke, M.G. and J.P Palta. 1986. Calcium uptake by potato tubers: Role of the basal roots. Hort Science 21: 1022-1024. Lærke, P.E., J. Christiansen, M N. Andersen, and B. Veierskov. 2002a. Blackspot bruise susceptibility of potato tubers during growth and storage determined by two different test methods. Potato Research 45: 187-202. Lærke, P.E., J. Christiansen, and B. Veierskov 2002b. Colour of blackspot bruises in potato tubers during growth and storage compared to their discolouration potential. Postharvest Biology and Technology 26: 99-101. Lambert, D.H., and F.E. Manzer. 1991. Relationship of calcium to potato scab. Phytopathology 81: 632-636. Lande, R., and S.J. Arnold. 1983. The measurement of selection on correlated characters. Evolution 37: 1210-1226. Matheis, G. 1987. Polyphenol oxidase and enzymatic browning of potatoes (Solanum tuberosum). II. Enzymatic browning and potato constituents. Chem. Mikrobiol. Technol. Lebensm. 11: 33–41. Mc Cann, L.R., and J.C. Stark, 1989. Irrigation and nitrogen management effects on Potato Brown Center and Hollow Heart. HortScience 24: 950-952. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011a. Linkage Mapping and QTL Analysis of Agronomic Traits in Tetraploid Potato (Solanum tuberosum subsp. tuberosum). Crop Science 51: 771-785. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011b. QTL mapping of internal heat necrosis in tetraploid potato. Theoretical and Applied Genetics 122: 129–142. McGuire, R. and A. Kelman. 1986. Calcium in potato tuber cell walls in relation to tissue maceration by Erwinia carotovora pv. atroseptica. Physiology and Biochemistry 7: 401-406.
100
Nakagawa, S. and H. Schielzeth. 2010. Repeatability for Gaussian and non-Gaussian data: A practical guide for biologists. Biological Reviews 85: 935–956. Olsen, N.L., L.K. Hiller, and L.J. Mikitzel. 1996. The dependence of internal brown spot development upon calcium fertility in potato tubers. Potato Research 39: 165-178. Ozgen, S., B.H. Karlsson and J.P. Palta. 2006. Response of potatoes (cv. ‘Russet Burbank’) to supplemental calcium applications under field conditions. Tuber calcium, yield, and incidence of brown spot. American Journal of Potato Research 83: 195-204. Palta, J.P. 1996. Role of calcium in plant responses to stresses: Linking basic research to the solution of practical problems. HortScience 31: 51-57.
Parkin, K.L., and M.A. Schwobe. 1990. Effects of low temperature and modified atmosphere on sugar accumulation and chip color in potatoes (Solanum tuberosum). Journal of Food Science 55: 1341-1344. Paterson, H.D., and R. Thompson, 1971. Recovery of inter-block information when block sizes are unequal. Biometrika 58: 545–554. Pavek, J.J., C.R. Brown, M.W. Martin, and D.L. Corsisni. 1993. Inheritance of blackspot bruise resistance in potato. American Journal of Potato Research 70: 43-48. R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R‐project.org. Rashid, M.H.O., and A.L. Carpena. 1997. Genetic variability for tuber dormancy and associated traits in potato (Solanum tuberosum L.). Philippine Journal of Crop Science 22: 125-130. Rex, B.L., and G. Mazza. 1989. Cause, control and detection of hollow heart in potatoes: A review. American Journal of Potato Research 66: 165-183. Ruiz de Galarreta, J.I., B. Ezpeleta, J. Pascualena, and E. Ritter. 2006. Combining ability for yield components in early generations of plant breeding. Plant Breeding 125: 183-186. Searle, S.R., G. Casella, and C.E. McCulloch. 1992. Variance components. Wiley Series in Probability and Statistics Vol. 391. New Jersey, USA: Wiley-Interscience. 528 pages Silva, E., J.M.S. Viana, V.R., Faria, and M.D.V. de Resende. 2013. Bayesian inference of mixed models in quantitative genetics of crop species. Theoretical and Applied Genetics 126: 1749-1761. Spearman C. 1904. The proof and measurement of association between two things. The American Journal of Psychology 15: 72–101.
101
Taiz, L. 1984. Plant cell expansion: Regulation of cell wall chemical properties. Annual Reviews of Plant Physiology 35: 585-657. Tzeng, K.C., A. Kelman, K.E. Simmons, and K.A. Kelling. 1986. Relationship of calcium nutrition to internal brown spot of potato tubers and sub-apical necrosis of sprouts. American Journal of Potato Research 63: 87-97. Werij, J.S., B. Kloosterman, C. Celis-Gamboa, C.R. de Vos, T. America, R.G. Visser, and C.W. Bachem. 2007. Unravelling enzymatic discoloration in potato through a combined approach of candidate genes, QTL, and expression analysis. Theor Appl Genet, 115:245–252. Wiersema, H.T. 1970. A reliable method for testing scab resistance in the green house. In Proceedings of the 4th Triennial Conference of the European Association for Potato Research. Brest, France. 8-13 Sept. 1969. 210–212. Paris: Inst. Natl. de la Recherché Agronomique Publ. 70-5. Wricke, G., and E. Weber. 1986. Quantitative Genetics and Selection in Plant Breeding. Walter de Gruyter, Berlin, Germany. Page 167.
102
TABLES
Table 3.1. Spearman’s rank correlation between years of evaluation for total yield (TY), specific gravity (SG), visual rating of chip color (CC), chip color in agtron units (AG), chip lightness (L), chip redness (A), chip yellowness, (B) for tuber calcium (TC), incidences of hollow heart (HH), black spot bruise (BB), pitted scab in the standard field (PS), pitted scab in the high disease pressure field (PS-E), and severity of pitted scab in the high disease pressure field (SPS-E) in the reciprocal populations of Atlantic x Superior and Superior x Atlantic.
Trait Field Spearman’s rank correlation Years to pool 2009-2010 2010-2011 2009-2011 2011-2012
TC C 0.25** 0.22 0.01 2009, 2010 TY C 0.40 *** 0.2 0.07 2009, 2010 SG C 0.38*** 0.35* 0.23 2009, 2010 and 2010, 2011 EB C 0.70*** 2010, 2011 CC C 0.29*** 0.47** 0.18 2009, 2010 and 2010, 2011 L C 0.41* 2010, 2011 A C 0.26 - B C 0.47* 2010, 2011
HH C 0.59*** 0.30* 0.32* 2009, 2010, 2011 BB C 0.11 0.13 0.19 - PS C 0.32*** 0.01 0.03 2009, 2010
PS-E E 0.69*** 2011, 2012 SPS-E E 0.67*** 2011, 2012
103 Table 3.2. Sources of phenotypic variation, best linear unbiased predictions and heritabilities for total yield (TY), specific gravity (SG), enzymatic browning (EB), chip color in agtron units (AG), chip color visual rating (CC), chip lightness (L), redness (A), yellowness (B), tuber calcium (TC), incidences of hollow heart (HH), blackspot bruise (BB), pitted scab in the standard field (PS), pitted scab in the high disease pressure field (PS-E), and pitted scab severity in the high disease pressure field (SPS-E) in the Atlantic x Superior population.
Trait Years G Y GxY R B Mean Min Max H2 TC 2009 *** *** *** 156.24 123.77 208.95 0.61
2011, 2012 *** *** * *** NS 0.53 0.03 2.81 0.82 G=Genotype, Y=year, GxY=Genotype x Year, R=Replicate, B= Group. Mean, Min, Max are the mean, minimum and maximum BLUP. H2 =broad-sense heritability.
104 Table 3.3. Sources of variance and deviance, best linear unbiased predictions and heritabilities for total yield (TY), specific gravity (SG), enzymatic browning (EB), chip color in agtron units (AG), visual rating of chip color (CC), chip lightness (L), redness (A), yellowness (B), tuber calcium (TC), incidences of hollow heart (HH), black spot bruise (BB), pitted scab in the standard field (PS), pitted scab in the high disease pressure field (PS-E), and pitted scab severity in the high disease pressure field (SPS-E) in the Superior x Atlantic population.
Trait Years G Y GxY R B Mean Min Max H2 TC 2009 ** *** ** 159.36 136.00 179.48 0.45
2011, 2012 *** *** * *** NS 0.51 0.02 3.04 0.86 G=Genotype, Y=year, GxY=Genotype x Year, R=Replicate, B= Group. Mean, Min, Max are the mean, minimum and maximum BLUP. H2 =broad-sense heritability.
105 Table 3.4. Pair-wise Spearman’s rank correlations between best linear unbiased predictions in the Atlantic x Superior reciprocal populations for tuber calcium (TC), tuber yield (TY), specific gravity (SG), enzymatic browning (EB), chip color in agtron units (AG), visual rating of chip color (CC), chip lightness (L), chip redness (A), and chip yellowness (B), incidences of hollow heart (HH), blackspot bruise (BB), pitted scab in the standard field (PS), pitted scab incidence in the high disease pressure field (PS-E), and pitted scab severity in the high disease pressure field (SPS-E) using the datasets with the highest heritabilities from the 2009-2012 evaluations in Hancock, Wisconsin, USA.
Superior
x Atla
ntic200
9, 20
10200
9, 20
10200
9, 20
10, 2
011
2009,
2010
, 201
1
2009
2010,
2011
2011
2010,
2011
2010,
2011
2009,
2010
, 201
1
2010
2010
2011,
2012
2011,
2012
Atlantic x Superior TC TY SG EB AG L A B CC HH BB PS PS-E SPS-E
Significant Spearman’s rank correlations at p < 0.05 are indicated in bold. Correlations below the diagonal belong to the Atlantic x Superior population and correlation above the diagonal belong to the Superior x Atlantic population
106
Table 3.5. Relationship between tuber calcium concentration and its significantly correlated traits at p<0.05 including the incidences of hollow heart (HH), blackspot bruise (BB), pitted scab incidence in the high disease pressure field (PS-E), severity of pitted scab in the high disease pressure field (SPS-E), tuber yield (TY), specific gravity (SG), chip lightness (L), and chip yellowness (B) in the Atlantic x Superior reciprocal populations.
Trait Size Relationship with tuber calcium Probabilities in percentages and reponses at specific tuber calcium concentrationsǂ
† Relationship estimated by logistic regression. Other traits estimated using linear regressions.
ǂ Estimated probability or response for the indicated tuber calcium concentrations estimated using the relationship equation. Values were transformed to the regular scale for the severity of pitted scab.
P=probability, y=response, μg/g=micrograms per gram of dry weight.
107 Table 3.6. Best linear unbiased predictions of four promising clones selected from the reciprocal populations of Atlantic x Superior for tuber calcium (TC), yield (TY), specific gravity (SG), enzymatic browning (EB), chip color using a visual rating (CC), chip color in agtron units (AG), chip lightness (L), chip redness (A), chip yellowness (B), incidence of hollow heart (HH), incidence of black spot bruise (BB), incidence of pitted scab in the standard field (PS), incidence of pitted scab in the high disease pressure field (PS-E), and severity of pitted scab in the high disease pressure field (SPS-E). The parents were included for comparison.
ATL=Atlantic, SUP=Superior. Favorable values in green and non-favorable values in red.
108
FIGURES
Figure 3.1. Predicted probabilities of hollow heart and blackspot bruise at different tuber calcium concentrations in relation to tuber size in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations
100 150 200 250 300
SmallMediumLarge
tuber calcium (µg/g)
pro
bab
ility
0.00
0.05
0.10
0.15
0.20
Probability of Hollow HeartAxS
100 150 200 250 300
SmallMediumLarge
tuber calcium (μg/g)
pro
bab
ility
0.00
0.05
0.10
0.15
Probability of Blackspot BruiseAxS
A
C
100 150 200 250 300
SmallMediumLarge
tuber calcium (μg/g)
pro
bab
ility
0.00
0.04
0.08
Probability of Hollow HeartSxA
100 150 200 250 300
SmallMediumLarge
tuber calcium (μg/g)
pro
bab
ility
0.00
0.02
0.04
Probability of Blackspot BruiseSxA
B
D
A, B. Probability of hollow heart estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations. C, D. Probability of blackspot bruise estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations.
109 Figure 3.2. Predicted probabilities of pitted scab incidence and predicted severity in the high disease pressure field at different tuber calcium concentrations in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations
in the high disease pressure field
100 150 200 250 300
tuber calcium (µg/g)
pro
bab
ility
0.0
0.2
0.4
0.6
Probability of Pitted Scabin the high disease
pressure field
AxS
100 150 200 250 300
tuber calcium (µg/g)
pro
bab
ility
0.0
0.2
0.4
0.6
Probability of Pitted Scab
SxA
100 150 200 250 300
tuber calcium (µg/g)
pit
s p
er t
ub
er
0.0
0.5
1.0
1.5
2.0
2.5
Severity of Pitted Scabin the high disease
pressure field
AxS
100 150 200 250 300
tuber calcium (µg/g)
pit
s p
er
tub
er
0.0
0.5
1.0
1.5
2.0
2.5
Severity of Pitted Scabin the high disease
pressure field
SxA
A
C
B
D
A, B. Probability of pitted scab estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations. C, D. Severity of pitted scab estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations.
110 Figure 3.3. Predicted predicted yield and specific gravity at different tuber calcium concentrations in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations
100 150 200 250 300
tuber calcium (μg/g)
ton
s/h
a
010
2030
4050
Total YieldAxS
100 150 200 250 300
tuber calcium (μg/g)
ton
s/h
a
010
2030
4050
Total YieldSxA
100 150 200 250 300
tuber calcium (μg/g)
g/g
1.04
1.05
1.06
1.07
1.08
Specific GravityAxS
100 150 200 250 300
tuber calcium (μg/g)
g/g
1.04
1.05
1.06
1.07
1.08
Specific GravitySxA
A
C
B
D
A, B. Total yield estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations. C, D. Specific gravity estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations.
111 Figure 3.4. Predicted chip lightness and chip yellowness at different tuber calcium concentrations in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations
100 150 200 250 300
tuber calcium (μg/g)
L v
alu
es
010
2030
4050
60
Chip LightnessAxS
100 150 200 250 300
tuber calcium (μg/g)
Lva
lues
010
2030
4050
60
Chip LightnessSxA
100 150 200 250 300
tuber calcium (μg/g)
B v
alu
es
05
1015
20
Chip YellownessAxS
100 150 200 250 300
tuber calcium (μg/g)
B v
alu
es
05
1015
20
Chip YellownessSxA
A
C
B
D
A, B. Chip lightness estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations. C, D. Chip yellowness estimated at 100, 150, 200, 250 and 300 µg/g in the Atlantic x Superior (AxS) and Superior x Atlantic (SxA) populations.
112 Figure 3.5. External appearance and chips of the promising clones identified in the Atlantic x Superior reciprocal populations compared to the parents
113
CHAPTER 4
Mapping QTL for Tuber Calcium, Tuber Quality and Pitted Scab in a Tetraploid
Population of Potato (Solanum tuberosum) derived from Atlantic x Superior
ABSTRACT
The population generated by a cross of potato cultivars Atlantic x Superior was used to develop a
map for internal quality, tuber calcium; specific gravity, total yield, chipping quality, internal
defects incidence as well as incidence and severity of pitted scab. This population was genotyped
using the SolCAP 8300 Infinium Chip. After data quality assessment, 600 single nucleotide
polymorphisms markers with simplex x nulliplex and nulliplex x simplex dosages in the parents
from 151 genotypes were used to build a tetraploid linkage map that covered 1254 cM for
Atlantic and 939 cM for Superior. Phenotypic evaluations were performed during 2009-2012
seasons at the Hancock Agricultural Research Station of the University of Wisconsin-Madison.
Using an interval mapping approach, we identified a total of 75 QTL from both parents including
8 for tuber calcium, 10 for specific gravity, 6 for yield, 5 for enzymatic browning, 5 for chip
color in agtron scale, 8 for chip color using a visual scale, 2 for chip lightness, 2 for chip redness,
2 for chip yellowness, 7 for hollow heart, 2 for blackspot bruise, 4 for pitted scab incidence in a
standard field, 7 for pitted scab incidence in a high disease pressure field, and 7 for pitted scab
severity in a high disease pressure field. Some correlated traits had QTL located on the same
chromosome and at close positions. Differences in QTL detection and position between years
were observed.
114
INTRODUCTION
Tuber quality is the most important characteristic in processing potatoes. For chipping varieties,
good tuber quality means high specific gravity, light chip color, and low incidence of internal
and external defects. As well, low incidence and severity of diseases that cause deformities in
tubers such as pitted scab are desired. Tuber calcium has previously found to be associated with
reduced internal defects (Tzeng et al., 1986; Olsen et al. 1996, Palta 1996, Kleinhenz et al. 1999,
Karlsson et al. 2006, Ozgen et al. 2006), and resistance to soft-rot during storage (McGuire and
Kelman 1984). Tuber calcium concentration is a heritable trait as demonstrated in Chapter 3 and
genetic variation for this trait has been observed in both cultivated and wild potatoes (Bamberg
et al. 1993, Karlsson et al. 2006, Vega et al. 2006). Therefore, higher tuber calcium is a desirable
trait. A better understanding of the genetics of tuber quality traits and tuber calcium by the
identification of quantitative trait loci (QTL) associated with these traits could contribute to the
development of breeding strategies for improved chipping varieties.
As indicated in Chapters 2 and 3, the potato cultivars Atlantic and Superior differ significantly
for several tuber quality characteristics. The cross of Atlantic and Superior generated reciprocal
populations that are segregating for total tuber calcium, tuber yield, specific gravity, enzymatic
browning, a visual rate of chip color, chip color as measured in agtrons, chip lightness, chip
redness, chip yellowness, hollow heart and blackspot bruise as well as incidence and severity of
pitted scab. In addition, an important proportion of the observed phenotypic variation is due to
the differences among the genotypes. Therefore, these populations are appropiate to identify the
genetic regions that control tuber calcium, tuber quality and pitted scab tolerance. In this study
115
we used one of the reciprocal populations, the Atlantic x Superior population, to identify QTL
for these traits.
Commercial potato varieties mostly belong to the autotetraploid species Solanum tuberosum.
Polyploids have complex inheritance modes due to preferential and multiple pairing during
cross-over, and the possibility of double reduction (Bradshaw 1994, Wu et al. 2001, Leach et al.
2010). Double reduction is the formation of gametes with segments of two sister chromatids in
the same gamete in polyploids as a result of crossover between a locus and the centromere; and
thus, varies from locus to locus (Mather 1935, 1936, Haynes and Douches 1993). If
quadrivalents are always formed in a tetraploid and an effective cross-over occurs between the
locus and its centromere, the coefficient of double reduction (α) is its maximum of 1/6 (Mather
1936. If there is chromatid segregation, the coefficient of double reduction is equal to 1/7
(Haldane, 1930). The probability of double reduction for each homologous chromosome is 1/4α
(Bradshaw 1994). Theoretically, the frequency of duplex genotypes produced by double
reduction in a simplex genotype is 0 (0%) according to the random chromosome segregation
model (Muller, 1914), 1/28 (3.57%) according to the random chromatid segregation model
(Haldane, 1930), 1/24 (4.17%) according to the maximum equational segregation model
(Mather, 1935, 1936), and 1/10 (10%) according to the general polyploid model (Wu et al.,
2001). The general poplyploid model described by Wu et al. (2001) predicts a 1/10 frequency for
each possible type of gamete including the products of double reduction following the formation
of a quadrivalent assuming that the three types of segregation: no-crossover, cross-over with
equational division, and cross-over with reductional division occur in the same frequency in the
first meiotic division of an autotetraploid. These characteristics of tetraploid potato have made it
116
less preferred for genetic studies.
Potato tetraploid linkage maps have been scarce compared to diploid maps due to its complex
inheritance (Luo et al. 2001). Previous QTL maps in tetraploid populations include those
developed for late blight (Phytophtora infestans (Mont.) de Bary) by Meyer et al. (1998),
Verticillium dahliae by Simko et al. (2004), for Colorado potato beetle (Leptinotarsa
decemlineata [Say]) by Sagredo et al. (2009); for yield, agronomic traits and quality traits by
Bradshaw et al. (2008); and for agronomic traits and internal heat necrosis by McCord et al.
(2011a, 2011b). So far, QTL interval mapping in tetraploid populations has been implemented
only in TetraploidMap (Hackett et al. 2007). This software was developed to deal with simple
sequence repeats (SSR) and amplified fragment length polymorphisms (AFLP). However, the
datasets can be adapted to use single nucleotide polymorphisms (SNP) data. The SolCAP project
developed an Illumina Infinium Bead Chip for potato that evaluates simultaneously 8303 SNP
(Hamilton et al. 2011). The advantages of using SNP markers for tetraploid mapping are that
polymorphisms can be detected ideally in every position of the genome and the dosage can be
determined. To determine the dosage for each SNP locus, a clustering analysis was performed by
the SolCAP using the Illumina GenomeStudio software. A set of 354 diverse lines, two F1
tetraploid mapping populations and one F1 diploid mapping population were included in the
determination of cluster positions for the diploid and tetraploid models (Hirsch et al. 2013). With
the availability of whole-genome genotyping data and the development of software that can deal
with it, linkage mapping of tetraploid populations generated from already accepted commercial
varieties would become an increasingly used tool.
117
The present study aims to identify genomic regions that can explain the phenotypic variation
observed in the Atlantic x Superior tetraploid population (detailed data given in Chapters 2 and
3). This research is the first genetic study of tuber calcium using molecular data and an attempt
to explain some of the observed correlations between tuber quality traits and tuber calcium. In
this study we also attempted to identify regions associated with tolerance to pitted scab at the
tetraploid level.
MATERIAL AND METHODS
Population and genotyping
The population generated by the cross of the potato cultivars Atlantic (female) and Superior
(male), and was named the Atlantic x Superior population. This population consists of 184
clones obtained in the Biotron of the University of Wisconsin-Madison, USA. The Atlantic x
Superior population and their parental genotypes were genotyped using the Illumina high-
throughput SNP assay. This assay evaluated 8303 SNP markers identified by Hamilton et al.
(2011) that combine information from cDNA sequencing of popular modern varieties, Atlantic,
Premier Russet and Snowden, and data mining of existing ESTs from a set of old potato
varieties, Bintje, Kennebec and Shepody. For the SNP genotyping, leaf samples were collected
in the maintenance field in 2010 and sent to the Seed Biotechnology Center of the University of
California for DNA isolation and genotyping as part of the SolCAP Project. The advantages of
using SNP markers are their abundance and that they give us information about allele dosage. In
general, it has been observed that SNP are mostly bi-allelic with two alleles segregating in the
populations (Krawczak 1999). Therefore, in a single locus with two alleles of a tetraploid potato,
118
we can get up to 5 different genotypes including: nulliplex (BBBB), simplex (ABBB), duplex
(AABB), triplex (AAAB), and quadriplex (AAAA).
Analysis of genetic structure
The genetic structure was evaluated using the package adegenet (Jombart 2008) in R version
2.13.2 (R Development Core Team 2011). Principal components analysis (PCA) was performed
to confirm that all clones genotyped belong to the Atlantic x Superior population. The two
principal components that explain the most variance were plotted in a two-dimensional graph.
Allele frequencies were estimated and missing data were replaced by the mean allele frequency
for the genetic relatedness analysis. A dendogram was constructed as a graphical representation
of genetic relatedness using the Nei’s distance (Nei 1972) and the complete linkage method of
hierarchical clustering (Sorensen 1948). Duplicated genotypes were identified by a pairwise
comparison of genotypes and estimation of the proportion of matches using the qtl package
(Broman et al. 2003) in R version 2.13.2. Pairs of genotypes with more than 0.99 identities were
identified and only one representative was kept for further analyses.
Sorting SNP markers suitable for tetraploid mapping
Each SNP marker in the potato Infinium SNP chip was assigned a quality indicator by SolCAP.
The markers with quality indicators BAD and QUESTIONABLE, were removed, keeping only
the GOOD and SEGREGATING markers. Further, any marker missing in all clones or in one or
both parents was removed. Then, markers that could not be located (no hits) in the pseudo-
molecule of the Solanum tuberosum Group Phureja double monoploid reference genome
sequence (Potato Genome Sequencing Consortium 2011), and those located in more than two
119
positions (two hits) were also removed. Also, markers that mapped in more than two positions in
the DRH and D84 reference diploid maps (Felcher et al 2012) were removed. Due to lack of
software that allows handling of tetraploid genotypic data with known dosage, only cross types
with expected segregation ratio of 1:1 in tetraploid populations, such as simplex x nulliplex and
nulliplex x simplex (simplex markers) were used for mapping. Simplex markers were more
abundant compared to other types of markers such as duplex (duplex x nulliplex or nulliplex x
duplex) or triplex markers (triplex x nulliplex or nulliplex x triplex) (Data not shown).
Distorted markers, markers missing in most genotypes, genotypes with lots of missing markers,
duplicated markers and duplicated genotypes were analyzed using the tools for data cleaning in
the qtl package (Broman et al. 2003) in R version 2.13.2. Segregation distortion was tested for
each locus using a Chi-square test. A correction for multiple comparisons was performed using a
Bonferroni procedure (Bonferroni 1935) with an experiment wise error rate of 0.05. Therefore,
SNP markers with p-values smaller than 0.05/total number of markers were excluded from the
dataset. Duplicated markers, markers that show identical information, were identified and one
representative marker was kept and the rest were eliminated from the dataset because duplicated
markers will invariably map to the same genomic position (Broman 2010). Duplicated genotypes
are not common but also they are not rare; and may indicate sample duplications (Broman 2010).
Duplicated genotypes may also indicate genotyping errors (Pompanon et al. 2005). Therefore,
duplicated genotypes were also removed from the dataset for mapping analysis.
Evaluation of double reduction
Segregations between a locus with simplex dosage (ABBB) and a locus with nulliplex dosage
(BBBB) are called simplex markers. In a simplex marker, duplex genotypes (AABB) can only be
120
generated by double reduction. The frequency of individuals produced by double reduction in
each SNP marker was estimated as the proportion of duplex genotypes found in simplex
markers. SNP markers with frequency of duplex genotypes 16.67% (1/6) or higher were
removed from the dataset. This threshold was set because 1/6 is the maximum expected value of
double reduction as indicated by Mather (1936). Higher numbers of duplex genotypes are most
likely generated by genotyping errors. Duplex genotypes were relabeled as missing data prior to
linkage analysis in all remaining SNP locus because markers with double reduction cannot be
used in the mapping software TetraploidMap (Hackett et al. 2007).
Construction of a tetraploid linkage map and QTL interval mapping
A tetraploid linkage map was constructed using TetraploidMap (Hackett et al. 2007; available at:
http://www.bioss.ac.uk/knowledge/tetraploidmap/). This software was built to deal with SSR and
AFLP markers but can be accommodated to work with SNP (Christine Hacket from The James
Hutton Institute and Joseph Coombs from Michigan State University, personal communication).
Simplex markers were coded using a binary nomenclature, where simplex genotypes were coded
as 1 and nulliplex genotypes as 0 trying to emulate the coding of a dominant AFLP marker.
Markers with significant segregation distortion, p <0.001 for the Chi-square test, were excluded
from the linkage map analysis. Markers that belong to the same chromosome according to their
physical position in the pseudo-molecule of the potato genome sequence were used for the
clustering analysis, this cluster analysis is the first step to generate a linkage group. The cluster
analysis was performed using a distance coefficient d=1-10-2s, where s is the significance level of
the test for independent segregation (Luo et al. 2001). Recombination frequency and linkage
phase was estimated for each pair of marker loci using maximum likelihood and the Expectation-
121
Maximization (EM) algorithm as described by Luo et al. (2001). The markers were ordered in
each chromosome using a two-point linkage analysis with custom marker ordering and ripple
ordering (Hackett et al. 1998). Linkage phase was assigned manually using the pair-wise
likelihood evaluation of coupling and repulsion described by Luo et al. (2001) to assign markers
to one of the four homologous chromosomes. Markers in coupling with more than one homolog
were removed from the analysis. Overall and per-homolog maps were built for each chromosome
and each parent. Trait data were related to the genotypic data to identify QTL using the interval
mapping approach for full-sib families in autotetraploid species derived from an intercross as
described by Hackett et al. (2001). This method is included in TetraploidMap in the QTL
function (Hackett et al. 2007). The QTL analysis was performed using the full model and
compared to simpler models such as simplex x nulliplex or duplex x nulliplex segregations to
determine the dosage of the QTL and its effect as described by Hackett et al. (2007). If no
simpler model explains the QTL effects, the QTL is considered additive with a complex
inheritance. Using the Permutations function, a total of 100 permutation tests were performed in
order to define 90% and 95% confidence thresholds as described by (Churchill and Doerge
1994). Only peaks with maximum LOD above the 90% threshold of permutation tests were
considered acceptable. The QTL maximum LOD, position and variance explained were
recorded.
Marker regression
The analyses of variance by means comparison (ANOVA) and by ranks comparison (Kruskal-
Wallis test) were performed to detect significant differences between the genotypes carrying
simplex markers versus nulliplex markers. These analyses are available in TetraploidMap
122
(Hackett et al. 2007). The significance threshold was set at p<0.01. Only markers within ± 20 cM
of a detected QTL are reported here.
Phenotypic evaluations
The field performances of the progenies from the Atlantic x Superior population were evaluated
during the 2009-2012 seasons at the Hancock Agricultural Research Station of the University of
Wisconsin-Madison. The traits studied were total tuber yield (ton/ha), tuber calcium (µg/g),
specific gravity (g/g), enzymatic browning (using a visual rating from 1 to 5, light to dark), chip
color (using a visual rating from 1 to 5, light to dark), chip color in agtrons, chip lightness in L
values, chip redness in A values, chip yellowness in B values, incidence of hollow heart,
incidence of black spot bruise, incidence of pitted scab, and severity of pitted scab. All trials
were conducted under standard commercial production practices of Central Wisconsin. A
standard field was used to evaluate tuber calcium, agronomic traits, internal defects, chip quality,
and pitted scab incidence. In addition, a high disease pressure field, a field that has been used
continuously to grow potatoes without rotation and is known to have high amounts of common
scab inoculum, was used to evaluate pitted scab incidence and severity. Two to three
replications/clone and four or eight plants/plot were evaluated using an incomplete block design
for most trials except for the standard field in 2011 where a complete block design was used. In
the incomplete block design, clones were randomized within groups of 28 clones, each group
included a replicate of both parents, groups were randomized within replicates, and replicates
were randomized within years. Best linear unbiased predictions (BLUP) estimated for the
Atlantic x Superior progenies were used as trait data for QTL mapping. Incidence traits were
evaluated as proportions. Clones performances were determined per year and pooling years. The
123
model used to estimate pooled BLUP values was the following: Xijkl = μ + Gi + Yj + Bk(Rl(Yj)) +
(GY)ij + εijkl; where Xijkl are the observed trait measurements, μ is the overall mean, Gi are the
genotypes, Yj are the years, Bk(Rl(Yj)) are the groups nested in replicates and replicates nested in
years, (GY)ij is the interaction between genotypes and years, and εijkl is the residual error. All
variables were considered random. The estimation of per year BLUP used the same model
without the year and genotype x year interaction variables. Linear mixed models and generalized
linear mixed models were used depending on the type of data evaluated, direct observations or
proportions, respectively. A detailed analysis of tuber quality and calcium traits in the Atlantic x
Superior population was presented in Chapter 2 and 3.
RESULTS AND DISCUSSION
Quality assessment of genotypic data
The Atlantic x Superior population studied in this research was created in 2003 followed by
several years of field maintenance. According to the genetic structure analysis, four clones were
identified as clearly genetically distant from the rest of the population (Figure 4.1). We believe
that these clones might belong to the Superior x Snowden population which was also created in
2003 and grown in randomized plots in the same field with the Atlantic x Superior population
throughout the years. The genetic structure analysis also indicates that some clones had been
misidentified (Figure 4.1). A sample of Atlantic was mislabeled as Superior, another Atlantic
sample was mislabeled as a progeny, and a sample of Superior was mislabeled as a progeny. In
addition, clusters of duplicated genotypes were also observed in the hierarchical cluster analysis
124
(data not shown). Two clusters of duplicated genotypes corresponded to genotypes sampled
twice. Other 15 clusters of duplicated genotypes included clones with different codes.
Furthermore, the pair-wise comparison between clones indicates that some clones have high
genotypic identities (identity > 0.99) (Figure 4.2). These highly identical clones were removed
from further analyses in order to avoid redundancy; only one representative from each cluster of
duplicates was used for mapping. After sorting markers and keeping those more suitable for
tetraploid mapping (Table 4.1), the number of SNP markers was reduced from 8303 to 619
simplex markers. These markers were introduced to TetraploidMap where a final sorting
occurred by excluding 5 simplex markers due to high segregation distortion (p-value < 0.001),
and 14 simplex markers due to similarly low values of LOD for coupling with more than one
homolog. A final dataset of 600 simplex markers evaluated in 151 progenies with unique
genotypes passed the quality assessment (Table 4.1) and were used to construct the tetraploid
linkage map for the Atlantic x Superior population.
Evidence of double reduction in the Atlantic x Superior population
Double reduction was evaluated in the Atlantic x Superior population using 968 simplex markers
(Table 4.1). We detected evidence of double reduction in 28.7% out of 968 simplex loci. The
frequency of duplex genotypes produced as a result of double reduction was between 0 and
33.33%. These results contrast with previous studies that reported that double reduction occurs
sporadically in potato (Haynes and Douches, 1993). In a study by Haynes and Douches (1993),
double reduction was estimated using isozymes in 4x-2x crosses and haploid families. Using the
data presented for the Pgm locus in simplex genotypes of Katahdin and Merrimack reported by
Haynes and Douches (1993), the frequency of duplex genotypes was between 5% and 9.18%
125
which is within the range we have observed in the present study. These values exceed the
maximum expected frequency according to the chromosomal and chromatid segregation models
but are within the expected values for the general polyploid model. Nevertheless, 9 SNP markers
in our population had more than 10% duplex genotypes which is the maximum expected
according to the general polyploid model. These loci with higher than expected duplex genotypes
may be caused by genotyping errors, selection acting at the level of the gametes, high
recombination frequency between the locus and the centromere, or a combination of all these
factors. The maximum frequency of double reduction expected is 1/6 or 16.67% of duplex
genotypes. Therefore, the 6 SNP with > 16.67% duplex genotypes were removed from the
dataset for mapping.
Tetraploid linkage map for the Atlantic x Superior population using SNP markers
Tetraploid linkage maps of Atlantic and Superior were constructed using 600 simplex SNP
markers distributed over 12 chromosomes (Table 4.3). The number of SNP mapped per
chromosome ranged from 18 (Chr.12) to 49 (Chr. 2) in Atlantic and from 5 (Chr.11) to 24
(Chr.7) in Superior. Chromosome lengths ranged from 84.8 cM (Chr.12) to 123.7 cM (Chr.1) in
Atlantic and from 34.8 cM (Chr.11) to 116 cM (Chr.4) in Superior. The overall size of the map
was 1238.4 cM for Atlantic and 889 cM for Superior. More than twice (414) markers were
mapped in Atlantic compared to Superior (186). A possible explanation for this difference is that
Atlantic was one of the cultivars used in the development of the SNP markers by SolCAP
whereas Superior was not used. Therefore, one hypothesis coul be that the polymorphisms
detected by the SNP chip might be biased towards Atlantic. Another hypothesis is that Superior
126
is more imbred than Atlantic and thus may have a higher number of homozygous loci compared
to Atlantic. By looking at the pedigrees of both parents it seems that the second hypothesis is the
most plausible (Figure 4.4). On average our map has 1 SNP marker every 2.33 cM for Atlantic
and 1 SNP marker every 3.8 cM for Superior. Areas near the centromeric regions, where
increments in physical position have little effect on map distance, were identified using these
SNP markers as reported previously by Felcher et al. (2012). The physical versus linkage map
correlation plots indicate a misorientation of the cM position with respect to the physical position
suggesting an inversion at the distal end of Chr.10 (Figure not shown). These results indicate that
some chromosomal rearrangements may exist between S. tuberosum and S. phureja because we
have constructed a map using two S. tuberosum cultivars and the physical positions assigned to
the SNP markers in the SolCAP 8300 Infinium Chip were based on the genome sequence of the
double monoploid of S. phureja (DM1-3 516 R44) (Potato Genome Sequencing Consortium
2011). Felcher et al. (2012) also observed regions with chromosomal rearrangements in Chr. 5
and 12 comparing a S. phureja x S. tuberosum-based and S. phureja x S. chacoense-based
diploid maps using the same SNP markers. As a consequence of these chromosomal
rearrangements these authors suggest to exercise caution when extrapolating sequence data
between species. Additional studies and future linkage maps constructed with the SolCAP 8300
Infinium Chip will shed more light on these chromosomal rearrangements.
Phenotypic data
Performance in terms of the best linear unbiased predictions (BLUP) of 151 progenies, selected
for mapping by the criteria described before, was used as trait data to associate with the
genotypic data. BLUP values and broad-sense heritabilities for tuber calcium, tuber yield,
127
specific gravity, chip color, hollow heart, blackspot bruise and pitted scab were estimated
(Chapter 3). Broad-sense heritabilities ranged from 0.19 to 0.85 for all traits evaluated in the
Atlantic x Superior population and the correlation analyses of trait data across years using the
Spearman’s rank correlation indicated a positive correlation between ranks for different years for
most traits; however, all traits were affected by genotype x environment effects (Chapter 3).
Effect of population size and heritability on QTL detection in the Atlantic x Superior
population
Several QTL were identified for all traits evaluated across Atlantic and Superior genomes
(Figure 4.5-4.16 and Table 4.3). The population sizes used for QTL mapping varied from
season to season and ranged from 32 for the 2011 trial in the standard field to 128 in 2010 in the
standard field (Table 4.4). The effects of the QTL were usually around 10% for most years of
evaluation and pooled analysis (Table 4.5). However, higher values of variance explained were
observed in QTL identified only in the 2011 trial in the standard field (Table 4.5). For example
some QTL were identified in the standard field only in 2011, a QTL for specific gravity located
on Chr. 1 of Atlantic at 101.7 cM and explains 39.5% of the variance, another QTL for the visual
rating of chip color located on Chr.11 at 14.2 cM that explains 41.4%, and two other QTL for
hollow heart located on Chr. 1 and Chr. 5 that explain 57 and 49.6% of the variance,
respectively. These high values of variance explained are most likely due to the over-estimation
of the QTL effect when population size is small as explained by Beavis (1994, 1998). In this case
population size was 32 for the 2011 trial in the standard field (Table 4.4). The number of QTL
identified for each trait ranged from 2 to 9 in the Atlantic x Superior populations (Table 4.3).
Fewer and more unstable QTL were found for traits with low heritabilities such as blackspot
128
bruise that had broad-sense heritabilities of 0.3, 0.3 and 0.19 in 2009, 2010 and 2011,
respectively. Two QTL were detected for blackspot bruise; each of them present only in a single
year of evaluation (Table 4.3 and 4.5). This lower power of QTL detection in traits with low
heritability has also been explained by Beavis (1994, 1998).
QTL for tuber calcium
Eight QTL were identified for tuber calcium (TC), seven on Chr. 1, 3, 5, 7, 8, 9, and 11 of
Atlantic and one on Chr. 12 of Superior. The effects of these QTL ranged from 7.7 to 30.3%, two
of these QTL were detected in one year, and the rest were detected in either two years or one
year and the pooled data (Table 4.5). Only one of these QTL increases the concentration of
calcium when it is present but it also has the lowest effect among all the QTL for calcium.
However, the smaller number of markers in Superior may have reduced our power to detect QTL
coming from this parent that has the highest tuber calcium content. Our results confirm that tuber
calcium is controlled by several genes and therefore a trait of a quantitative nature with genes
distributed throughout the potato genome.
QTL for yield and specific gravity
Six QTL were identified for tuber yield (TY), 2 in Atlantic and 4 in Superior. In Atlantic, a QTL
was located on Chr. 1 at 73.7 cM that explained 10.7% of the variance. Another was located on
Chr. 3 of Atlantic at 112 cM that explained 21.8% of the variance. In Superior, two QTL for
tuber yield were located on Chr. 1 at 10 cM and 102 cM with 9.06% and 13.09% variance
explained, respectively. Another QTL detected was found on Chr. 2 of Superior located at 22 cM
and variance explained by this QTL was 11.3%. Another QTL was detected on Chr. 4 of
129
Superior at 20 cM that explained 8.9% of the variance (Table 4.5). Half of the QTL for tuber
yield were additive. The QTL with the largest effect was located in Atlantic and had a duplex
dosage where the dominant allele would increase total yield. In general, three QTL for yield
were located on Chr. 1 suggesting that this chromosome may have several loci that control this
trait. Bradshaw et al. (2008), identified a QTL for yield with dominance for low yield which was
located at 118 cM on Chr. 1 in the 12601ab1x Stirling population. Also, McCord et al. (2011a)
identified QTL on Chr. 2, 3, 5, 10, and 12 of Atlantic as well as on Chr. 5, 6, 7, and 8 of B1829-
5. In our Atlantic x Superior population, we also identified a QTL on Chr.3 of Atlantic but we
did not detect the other reported QTL but we detected another QTL on Chr. 1 of Atlantic that had
not been reported. This difference is expected because QTL detection in bi-parental populations
depends on the segregation in the specific population under study (Würschum 2012).
Ten QTL were identified for specific gravity (SG), 8 in Atlantic and 2 in Superior (Table 4.5).
These QTL explained variances between 7.7% and 39.5% and half of them were detected in a
single year (Table 4.5). For Atlantic, two QTL were found on Chr.1 located at 83.7 and 101.7
cM that explained 12.4% and 39.5% of the variance, respectively. Other QTL were identified on
Chr.2 at 7.3 cM that explained 17.6% of the variance, on Chr. 7 at 22 cM that explained 11.2%
of the variance, on Chr. 8 at 30.9 cM that explains 18.7% of the variance, on Chr. 11 at 70.2 cM
that explains 15.6% of the variance, two on Chr. 12 at 20 and 60 cM that explain 11.8 and 7.7%
of the variance, respectively. In Superior, a QTL on Chr.1 located at 104 cM that explained
11.04 of the variance and another QTL on Chr. 9 at 30.8 cM that explained 9% of the variance.
The QTL for specific gravity with the largest effect was duplex with the dominant allele
increasing the specific gravity and was found in Atlantic (Table 4.5). Our results are in general
130
in agreement with reports by Freyre et al. (1994) that detected 10 QTL in a (S. tuberosum x S.
chacoense) x S. phureja diploid population. Bradshaw et al. (2008) that found two QTL for
specific gravity on Chr.5 at 74 and 76 cM that explained 9.1 and 9.7% of the variance in the
Stirling x 12601 tetraploid population. McCord et al. (2011a) identified QTL in the Atlantic x
B1829-5 on Chr. 2, 3, 5 and 8 together with significant markers in group 7, Chr.9 and 12. We
also found QTL for specific gravity on Chr. 2, 7, 8, 9 and 12 in addition to other QTL on Chr. 1
and 11.
QTL for chip quality
Chip quality was studied using several measurements of chip color and enzymatic browning.
Enzymatic browning (EB) affects chip quality by determining the color of the potato slices
before they enter to the fryer (Krokida et al. 2001). Five QTL were identified for enzymatic
browning. Two of these QTL were located on Chr. 1 and 8 of Atlantic and three on Chr. 5, 6, and
7 of Superior. These QTL explained between 7.4 and 28.5% of the variances and most of them
were detected in two years, or one year and the pooled analysis (Table 4.5). The QTL with the
largest effect was found on Chr. 7 of Superior, followed by the QTL on Chr. 8 of Atlantic.
Enzymatic browning in raw cut potatoes has been related with the polyphenol oxidase (PPO)
enzymatic activity (Jeong et al. 2005). The PPO enzyme, which causes darkening of cut tissues,
has been previously mapped on Chr. 8 of Solanum (Newman et al. 1993). A study by Werij et al.
(2007) identified QTL for enzymatic discoloration (ED) on Chr. 1, 3 and 8 in a diploid
population using interval mapping with AFLP markers and marker regression with CAPS
markers targeting the PPO gene. The QTL on Chr. 8 found by Werij et al. (2007) was located in
a region around 50 cM. In our study, we found a QTL for enzymatic browning located on Chr. 8
131
of Atlantic using interval mapping with SNP markers in our tetraploid population. This QTL was
located at 50.9 cM, almost the same position reported by Werij et al. (2007). This QTL has a
simplex dosage that decreases the enzymatic browning when present. It appears that this QTL is
the PPO gene. These results confirm that our phenotyping method using visual ratings from 1 to
5 was adequate to detect the PPO gene and other QTL regions that might be related to the
substrates of this enzyme and therefore causing enzymatic browning.
For the visual scoring of chip color (CC), we found two QTL on Chr.1 of Superior and six QTL
on Chr. 4, 6, 7, 8, 9, and 11 of Atlantic. In Superior, two QTL on Chr. 1 were located at 68 and
80 cM that explain 10.3 and 8.8% of the total variance, respectively. In Atlantic, the QTL on
Chr. 4 was located at 34 cM that explained 10.1% of the variance. The QTL on Chr. 6 was
located at 36 cM and explained 9.1% of the variance. The QTL on Chr. 7 of Atlantic is located at
70 cM and explained 9.9% of the variance. The QTL on Chr. 8 of Atlantic was located at 84.9
cM and explained 13.8% of the variance. The QTL on Chr. 9 of Atlantic was located at 89.2 cM
and explains 14.9% of the variance. The QTL on Chr. 11 of Atlantic was located at 14.2 cM and
explains 41.4% of the variance but was detected only in 2011. All other QTL were detected at
least in two years. The QTL with largest effect was found in Atlantic on Chr.11 only during the
2011 evaluation and had a complex inheritance. Most previous studies have tried to understand
the genetics of chip color after several months in storage at low temperatures (Douches and
Freyre 1994, Menéndez et al. 2002, Bradshaw et al. 2008). However, in our study we tried to
understand the genetics of chip color shorter after harvest because neither Atlantic nor Superior
are a cold storage chip variety. Nevertheless, we have found eight QTL for chip color which is a
comparable number to what has been reported previously. For example, Douches and Freyre
132
(1994) found 6 QTL in a diploid population on Chr. 2, 4, 5 and 10. Also, Bradshaw et al. (2008)
found 4 QTL for chip color at two storage temperatures 4ºC and 10ºC on Chr. 1, 6 and 11. In this
research, we also found QTL on Chr. 1, 6 and 11 but they were located in different positions as
compared to previous reports. Our results confirm that using a visual rating for chip color is a
good approach to detect gene-based phenotypic variation. And also that chip color is a trait
controlled by several genes distributed throughout the genome.
In our study, six QTL were identified for the three components of chip color, L, A, and B, as
measured by the Hunter colorimeter (Table 4.5). For chip lightness measured as L values (L),
two QTL were detected on Atlantic which explained 11 and 11.8% variance, respectively. One
of these QTL was located on Chr. 7 at 76 cM detected in 2010 and the pooled analysis and the
other on Chr. 9 at 89.2 cM detected in 2010, 2011 and pooled analysis (Table 4.5). For chip
redness measured as A values (A), two QTL were identified in Atlantic. One of these QTL was
on Chr. 8 at 86.9 cM and the other on Chr. 9 at 101.3 cM. These QTL explained 11.5 % of the
variances, respectively; and both QTL were detected in 2010, 2011 and the pooled analysis
(Table 4.5). For chip yellowness measured as B values (B), two QTL were identified in Atlantic.
One of the QTL was located on Chr. 5 at 38 cM and the other one on Chr. 8 at 86.9 cM and
explained 8.4 % and 7.8% of the variances. Both of these QTL were detected in 2010 and the
pooled analysis (Table 4.5). Previous research has used chip lightness measurements to detect
phenotypic variation between varieties and/or under different physiological conditions (Sapers
1989, Rodriguez-Saona and Wrolstad 1997) and differences among transgenic clones (Bhaskar et
al. 2010). However, these measurements have not been previously used for QTL mapping.
133
For chip color measured by the agtron scale (AG), five QTL located on Chr. 1, 2, 4 and 5 of
Superior and one QTL on Chr. 7 of Atlantic were identified (Table 4.5). These QTL explained
between 9.3% and 18.3% of the variances and were detected only in 2009 because data for chip
color as measured in agtrons was collected only one season. The QTL with the largest effect was
found in Atlantic on Chr.7 at 74 cM and with a simplex dosage where the presence of the QTL
increased the agtron values. Previous research has studied chip color using agtron values but
these values have not been used for mapping. It is interesting to note that the QTL found for
agtron readings and L values were located in the same position or close to the QTL identified for
chip color (Table 4.5).
Genetic diversity for chip color has been demonstrated for cultivated and wild potato (Li et al,
2005, 2008; McCann, 2010). Furthermore, QTL for chip color have been identified by previous
studies using diploid populations (Douches and Freyre 1994) and tetraploid populations
(Bradshaw et al. 2008). Douches and Freyre (1994) mapped six QTL for chip color on Chr. 2, 4,
5 and 10 using a marker regression approach. Bradshaw et al. (2008) found QTL for chip color
after storage at 4°C and 10°C on Chr. 1, 6 and 11 in a tetraploid population using interval
mapping. In addition, because chip color is related to sugar content; Menéndez et al. (2002)
reported QTL located on all chromosomes of potatoes for sugar content after cold storage, a trait
that influences chip color. The QTL we found are also distributed in most of the potato
chromosomes in concordance with previous reports and these QTL were confirmed by different
types of measurements used for evaluating chip color and quality.
134
QTL for internal tuber quality
Internal tuber quality was evaluated in terms of the incidence of hollow heart (HH) and the
incidence of blackspot bruise (BB). For hollow heart, two QTL were identified in Superior and
four in Atlantic (Table 4.5). In Atlantic, a QTL detected on Chr. 1 at 45.7 cM explained 57% of
the variance but was detected only in 2011. Two other QTL were found on Chr. 3 at 0 and 40 cM
that explained 17% and 19.7% of the variance, respectively. Another QTL were found on Chr. 6
at 70 cM that explained 19% of the variance and a QTL on Chr. 9 at 101.2 cM that explained
71.7% of the variance. In Superior, a QTL for hollow heart was found on Chr. 3 at 44 cM that
explained 7.5% of the variance; another QTL was found on Chr.5 at 22 cM that explained 49.6%
of the total variance. The latter was detected only in 2011. The QTL on Chr.3 were the most
stable because they were present in two years or three years of evaluations and they were also
present in the pooled data.
For black spot bruise, two QTL were identified on Chr.10 and 11 of Atlantic at 21 and 50.2 cM
with explained variances of 20.9% and 7.1% and detected in 2009 and 2010, respectively. No
QTL for blackspot bruise were found in Superior. The PPO enzyme has been proposed to be
involved in the bruising reaction by previous studies (Matheis 1987, Urbany et al. 2011);
however, we did not detect the QTL on Chr.8 at 50 cM where the PPO enzyme has been located
in our blackspot bruise evaluations. QTL for tuber bruising were detected by Urbany et al.
(2011) using an association mapping approach. They found several markers associated to bruise
susceptibility that were located throughout the genome. In our study, we were not able to detect a
large number of QTL for blackspot bruise due to the low heritability of this trait under the
conditions of our evaluations.
135
Previous genetic studies for internal quality investigated internal heat necrosis (IHN), a defect
that causes necrotic patches in the tuber flesh when plants are subjected to heat stress (Sterret and
Henninger 1997); QTL for IHN have been detected in the Atlantic x B1829-5 tetraploid
population on Chr. 4, 5, 7, 10 of Atlantic, and Chr. 7 of B1829-5 (McCord et al., 2011b). In
addition, Bradshaw et al (2008) evaluated internal quality in terms of an index of internal
condition (IC) that combined all defects in a single score where higher values meant better
internal quality. A QTL for the IC index was found on Chr.5 at 116 cM. In our study, we found a
QTL for incidence of hollow heart on Chr.5 with large effects, 49.6% of variance explained but
it was detected only in the 2011 evaluation.
QTL for pitted scab
Pitted scab was evaluated by recording the incidence of this disease under standard field
conditions, and incidence and severity in a field with high disease pressure. For pitted scab
incidence under standard field conditions, four QTL were identified, two on Chr. 3 and 12 of
Atlantic and two on Chr. 4 and 5 of Superior (Table 4.5). These QTL explained between 7.8%
and 13.7% of the variances. Only the QTL on Chr. 4 of Superior and the QTL on Chr. 12 of
Atlantic were present in two years of evaluation and in the pooled analysis; the other QTL were
present only in one year or only in the pooled analysis. The other QTL with the largest effect was
located on Chr. 4 and explained 13.7% of the variance.
For pitted scab incidence under high disease pressure, seven QTL were identified, three on Chr.
3, 7 and 11 of Atlantic and four on Chr. 2 ,5, 6 and 10 of Superior. These QTL explained
between 7.8% and 23 % of the variances and most were detected in two years of evaluation and
136
the pooled data. The QTL with the largest effect was found in Atlantic and had a simplex dosage
and the incidence of pitted scab increased when the QTL was present (Table 4.5).
For pitted scab severity under high disease pressure, seven QTL were identified, three on Chr. 3,
9 and 11 of Atlantic and four on Chr. 2 ,5, 6 and 10 of Superior. These QTL explained between
9.5% and 49.6 % of the variances and most were detected in one or two years of evaluation and
pooled, and one detected only in the pooled evaluation. The QTL with the largest effect was
found on Chr.9 of Atlantic and had a complex inheritance.
In previous studies, two QTL for scab were found on Chr. 2 and 6 at positions 80 and 86 cM in
the Stirling x 12601ab1 tetraploid population (Bradshaw et al. 2008). We also found QTL in
these chromosomes but located in different positions in the genome. Braun (2013) identified a
QTL located at 10.1 cM on Chr.11 that explained 17 and 24.3% of the variance for lesion type
(LT) and percent surface area (PSA), in a diploid population generated by a cross between the
susceptible S. tuberosum clone US-W4 and the resistant S. chacoense clone 524-8. The
inheritance of pitted scab resistance was earlier studied in diploid potatoes by Alam (1972) who
proposed a simple model of inheritance that involves two loci, one where resistance is conferred
by a dominant allele and another one where resistance is conferred by a recessive allele. Several
studies have concluded that common scab resistance in haploid or diploid potato is controlled by
one or few genes. However, it does not seem to be the case in tetraploid potatoes (Dees and
Wanner 2012). In our study, we have found four QTL for pitted scab incidence under standard
disease conditions and seven QTL each for incidence and severity under high disease pressure
that suggest a quantitative nature for pitted scab tolerance. Our search of QTL for pitted scab
tolerance was successful in part because we focused on the evaluation of pitted lesions which are
137
more prominent and easier to detect and therefore the heritability of this trait is high. Also, the
evaluation in a field with high disease pressure allowed us to have enough inoculum to make
more evident the variation in pitted scab tolerance between genotypes.
Linked QTL and correlations between traits
Tuber calcium (TC) has been found to be negatively correlated to the incidence of hollow heart
(HH), brown center (BC), internal brown spot (IBS) and blackspot bruise (BB) in the Atlantic x
Superior population (see Chapter 3 for details). These correlations might be partially explained
by the linkage of QTL for these traits. In several cases correlated traits had QTL located at close
positions on the same chromosome (Table 4.5). For example on Chr.1, a QTL for TC and a QTL
for HH were located at 101.7 and 45.7 cM of Atlantic. On Chr.3, a QTL for TC was located at 50
cM, another QTL for HH located at 40 cM of Atlantic and another QTL for HH located at 44 cM
of Superior. Also, QTL for TC and HH were located at 87.2 and 101.2 cM of Chr. 9 of Atlantic.
On Chr. 11 of Atlantic, QTL for TC and BB were located in close proximity at 54.2 and 50.2 cM
of Atlantic. This result also indicates that the correlation observed between internal defects and
tuber calcium (Chapter 3) can be explained at least in part by this linkage observed among QTL.
However this hypothesis was not tested in this study.
Tuber calcium (TC) was also found to be positively correlated to chip color (CC), chip redness
(A) and enzymatic browning (EB); and negatively correlated to total yield (TY), chip lightness
(B) and specific gravity(SG) in the Atlantic x Superior population (Chapter 3). These
correlations can be partially explained by the close location of QTL for tuber calcium and yield,
specific gravity, and chip quality traits (Table 4.5). On Chr.1 of Atlantic, there is one QTL for
138
TY at 73.7 cM, two for SG at 83.7 and 101.7 cM, and one at 81.7 cM. These QTL are closely
linked to a QTL for TC located at101.7 cM (Table 4.5). On Chr. 5, there is a QTL for TC at 56
cM of Atlantic which is closely linked to a QTL for chip yellowness at 38 cM and for EB located
at 40 cM. Also on Chr.7, QTL for TC, L, AG and CC were very closely located at 76, 76, 74 and
70 cM, correspondingly. These QTL for the different measurements of chip color might actually
be the same QTL. Similarly on Chr. 8, there were QTL for TC, SG and EB in close positions at
46.9, 30.9 and 50.9 cM, respectively. On Chr. 9 of Atlantic, TC, CC, L, and A have QTL in close
positions located at 87.2, 89.2, 89.2 and 89.2 cM, respectively. On Chr. 11 of Atlantic, TC and
SG were located at 54.2 and 70.2 cM. Taken together these results indicate that the correlation of
tuber calcium with total yield, specific gravity, and the several measurements of chip color can
partly be explained by linkage of QTL for these traits.
In addition, the different types of chip color measurements including visual scores of chip color
(CC), chip lightness (L), chip redness (A), chip yellowness (B), chip color as measured by the
agtron scale (AG) and enzymatic browning (EB) evaluated in this population are highly
correlated among themselves (See Chapter 3). These traits also had some QTL located on the
same chromosomes in close positions indicating that the same genomic regions might be
explaining the phenotypic variation for chip color regardless of the type of measurement. For
instance on Chr.1, we found three QTL for chip quality in Superior. Two of these QTL were for
the visual rating of chip color at 66 cM and 80 cM, and another QTL for chip color as measured
in agtrons at 68 cM. On this Chr. a QTL for EB at 81.7 cM of Atlantic was also detected.
Furthermore on Chr.7, there were three QTL for chip quality in Atlantic when measured as
visual rating of chip color, chip color in agtrons, and chip lightness located at 70, 74 and 76 cM,
139
respectively; the small distance between these QTL suggests that they might be the same QTL.
On Chr. 8 of Atlantic, QTL for CC, A and B were found at 82.9, 86.9 and 86.9 which also could
be considered the same QTL. Furthermore on Chr.9 of Atlantic, QTL for visual rating of chip
color, chip lightness, and chip redness were identified at the same position which is 89.2 cM.
Correlated measurements of chip quality detected QTL in similar positions.
Similarly, measurements of pitted scab incidence under standard field (PS) and high disease field
conditions (PS-E) and severity under high disease field conditions (SPS-E) were highly
correlated (Chapter 3). This correlation was reflected in several QTL located in the same
chromosome and even the same position for these traits (Table 4.5). QTL for pitted scab
incidence and severity in the high disease pressure field were located in the same position or less
than 5 cM apart on Chr. 2, 3, 5, 6, 10 and 11. The QTL for PS were also located on Chr. 3 and 5
but their location was shifted by more than 20 cM. These results suggest that incidence and
severity of pitted scab evaluated in a field with high disease pressure seem to be controlled by
the same genes.
In summary, we have found that some of the QTL for correlated traits are closely linked causing
the simultaneous inheritance of the parental alleles for these traits unless there is recombination.
Therefore, large populations should be evaluated to increase chances of finding recombinants
between correlated traits. Evaluation of large numbers of progenies should help in finding clones
that could be selected as cultivars that combine the various desired traits including high tuber
calcium and good internal quality but also have good chipping quality as well as acceptable
specific gravity and yield.
140
Year to Year variability in QTL detection
Variable detection of QTL among years for all traits studied. Was observed in this study and one
of the reasons could be the significant genotype x environment (GxE) interactions observed in all
the traits evaluated. GxE interactions have a clear influence in the detection and localization of
QTL, as it has previously been observed for several traits in potato (Freyre and Douches 2004),
rice (Zhuang et al. 1997); cotton (Paterson et al. 2003), Arabidopsis and others plants (Ungerer et
al. 2003, Tétard-Jones 2011). Some important sources of variation are the differences in weather
and soil conditions from season to season that can change the relative performance of clones and
therefore the detection and location of QTL. Another important source of error in the QTL
detection and estimation of effects is population size that varied among years and was very small
in the standard field in 2011. In our study, 33 QTL were identified in two or three years of
evaluation plus the pooled QTL analysis. These QTL can be called “stable QTL”. We also
detected 22 QTL in either two years of evaluation or one year plus the pooled analysis. These
QTL can be called “semi-stable QTL”. In addition, 15 QTL were detected in one year of
evaluation or only in the pooled analysis and are called “unstable QTL” (Table 4.5). The last
classification does not include the 5 QTL detected for AG which was evaluated only in 2009. An
example of a stable QTL is the QTL for tuber calcium in Chr1. of Atlantic and this is presented
in Figure 4.17.
The QTL detected in multiple environments are more stable and may be of more importance for
plant breeding (Liu et al. 2006). QTL detected in several environments are most likely QTL with
larger effects (Broman and Sen 2009). Therefore, trait data from several trials, either several
141
years or several locations should be considered when performing QTL analysis in order to have a
stronger support for the detected QTL. This would be an additional confidence criterion to
develop markers for marker assisted selection. Our approach to evaluate traits by year and
pooled years was successful in detecting stable QTL and semi-stable QTL.
Regression analysis of SNP markers and opportunities for marker assisted selection
The best model that explains each QTL effect is indicated in Table 4.5. All QTL were initially
detected using the full or additive model and then compared to simpler models. If no significant
differences were observed with a simpler model, a new QTL analysis and permutations test were
performed with the simpler model. The simpler model was assumed to be the best model if the
peaks were above the LOD threshold; otherwise, the additive model was selected as the best.
Significant differences were identified in several SNP markers for all traits. Only SNP with p-
values < 0.01 for the ANOVA or the Kruskal-Wallis test and located ±20 cM from a detected
QTL are reported. The markers and their type of effect were also given for the SNP with
significant effects (Table 4.5). A positive effect of a marker (+) means that the mean is higher in
the simplex genotypes compared to the nulliplex genotypes; and a negative effect (-) indicates
that means are lower in the simplex genotypes. Combining these results with the availability of
the potato genome sequence; the markers with significant effects located nearby QTL regions
could be good candidates for further analysis towards the development of markers for marker
assisted selection. This is especially true for those with larger effects, present in most years of
evaluation and in simplex dosage where the effect of the marker is the desired effect on the trait.
The presence or absence of a significantly associated marker even if it is closely linked to the
QTL can be used successfully to select for the best phenotypes only if the QTL is in simplex
142
dosage and in coupling with the marker. For example, a marker with potential to be converted in
a presence/absence marker used in marker assisted selection for pitted scab tolerance was found
on Chr. 2 of Superior (Figure 4.18). This c2_ 13213414 marker is nearby a QTL in simplex
dosage that increases the incidence and severity of pitted scab when present. This marker was
located in the same homolog as the QTL and had the same effect. This marker could be used to
select against this QTL.
Because of the complex genetic nature of tetraploid potatoes, the direct application of most of the
associated markers identified in this study is going to be challenging. Several QTL were additive
meaning that their segregation is complex and cannot be explained by a simple inheritance
model. Additional fine mapping analyses within the identified QTL regions need to be performed
in order to identify candidate genes that explain the phenotypic variation. Specific markers for
those regions should be developed using the available potato genome sequence information.
Marker assisted selection (MAS) in potato offers great opportunities to use the currently
available genetic data (Barone 2004) that have not been extensively exploited. MAS in potato
has been applied mostly for pathogen resistance (Pineda et al. 1993, Hämäläinen et al. 1997,
Oberhagemann et al. 1999, Colton et al. 2006, Gebhardt 2006, Śliwka et al. 2010, Lopez-Pardo
et al. 2013) and in a few cases related to tuber quality (Freyre and Douches 1994, Li et al. 2013).
We expect to contribute with new target regions that can be exploited to use MAS for improving
tuber quality traits and scab tolerance. Future analyses of the reciprocal population Superior x
Atlantic would be performed to confirm the QTL regions, and the associations of some SNP
markers with the traits of interest. Lastly, the information provided by our study is a starting
point to locate the actual genes that control tuber calcium, tuber quality and pitted scab tolerance.
143
CONCLUSIONS
The results of this study lead us to conclude the following:
1. A tetraploid map was successfully constructed using 600 simplex SNP markers. The map
for Atlantic had more than twice the number of markers, 414, compared to Superior, 186,
suggesting that the SolCAP 8300 Infinium.
2. A large number of SNP loci with duplex genotypes were identified. Most of these are
probably products of genotyping errors.
3. Using an interval mapping approach, several quantitative trait loci were identified for
tuber calcium, tuber quality traits and pitted scab tolerance in the Atlantic x Superior
population.
4. Few and unstable QTL were detected for blackspot bruise, a trait with low heritability in
this study. Also, the small population size of the 2011 standard field trial caused over-
estimation of QTL effects. Both these results suggest that the Beavis effect can also occur
in tetraploid populations.
5. Tuber quality traits correlated to tuber calcium showed QTL in close positions on the
same chromosome indicating that their correlation can be explained at least partially due
to linkage.
6. The detection and sometimes location of QTL varied between years of evaluation. The
sources of error included environmental variation and population size.
7. Markers with significant effects were identified for several traits in the marker regression
analysis. We are reporting only those close to QTL positions.
8. Approximately half of the QTL were additive. It means that these QTL have a complex
144
inheritance. Nevertheless, some QTL had simpler inheritance such as dominant alleles in
simplex and duplex dosage.
9. Finally, combining the information of this QTL map and the potato genome sequence
offers opportunities to develop markers suitable for MAS and the identification of the
genes that are responsible for controlling tuber quality and tuber calcium in the future.
BIBLIOGRAPHY
Alam, Z. 1972. Inheritance of scab resistance in 24-chromosome potatoes. Ph.D. dissertation University of Wisconsin. Dissertation Abstract. B 32:6764-6765. Bamberg, J.B., J.E Palta, L.A. Peterson, M. Martin, and A.R. Krueger. 1993. Screening tuber-bearing Solanum (potato) germplasm for efficient accumulation of tuber calcium. American Journal of Potato Research 70: 219-226. Barone, A. 2004. Molecular Marker-assisted selection for potato breeding. American Journal of Potato Research 81: 111-117. Beavis, W.D. 1994. The power and deceit of QTL experiments: lessons from comparative QTL studies. In Proceedings of the Forty-Ninth Annual Corn and Sorghum Industry Research Conference, 250-266. Washington, DC: American Seed Trade Assoc. Beavis, W.D. 1998. QTL analyses: power, precision, and accuracy in molecular dissection of complex traits, ed A.H. Paterson, 145-162. Boca Raton, FL: CRC. van Berloo, R., R.C.B. Hutten, H.J. van Eck and R.G.F. Visser. 2007. An online potato pedigree database resource. Potato Research 50: 45-57. Bhaskar, P.B., L. Wu, J.S. Busse, B.R. Whitty, A.J. Hamernik, S.H. Jansky and J. Jiang. 2010. Suppression of the vacuolar invertase gene prevents cold-induced sweetening in potato. Plant Physiology 154: 939-948. Bonferroni, C.E. 1935. Il calcolo delle assicurazioni su gruppi di teste. In Studi in Onore del Professore Salvatore Ortu Carboni, 13-60. Rome, Italy. Bradshaw, J. E. 1994. Quantitative Genetics Theory for Tetrasomic Inheritance. In Potato Genetics, eds. J.E. Bradshaw and G.R. Mackay, 76-77. Wallingford, UK: CAB International. Bradshaw, J.E., C.A. Hackett, B. Pande, R. Waugh, and G.J. Bryan. 2008. QTL mapping of
145
yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Theoretical and Applied Genetics 116: 193-211. Braun, S.R. 2013. Quantitative trait loci analysis and breeding for resistance to common scab in potato. Ph.D. dissertation Plant Breeding and Plant Genetics. University of Wisconsin. Broman, K.W., H. Wu, S. Sen and G.A. Churchill. 2003. R/qtl: QTL mapping in experimental crosses. Bioinformatics 19: 889-890. Broman, K.W., and Sen S. 2009. A guide to QTL mapping with R/qtl. Springer, New York. Broman, K.W. 2010. Technical Report # 214. University of Wisconsin - Madison. Accessed on August 2013. Colton, L. M., H.I. Groza, S.M. Wielgus and J. Jiang. 2006. Marker-assisted selection for the broad-spectrum potato late blight resistance conferred by gene RB derived from a wild potato species. Crop Science 46: 589-594. Churchill, G.A., and R. W. Doerge. 1994. Empirical threshold values for quantitative trait mapping. Genetics 138: 963-971. Dees, M.W., and L.A Wanner. 2012. In search of better management of potato common scab. Potato Research 55: 249-268. Douches, D., and R. Freyre, 1994. Identification of Genetic Factors Influencing Chip Color in Diploid Potato (Solanum spp.). American Journal of Potato Research 71: 581-590. Felcher K.J., J.J. Coombs, A.N. Massa, C.N. Hansey, J.P. Hamilton, R.E. Veilleux, C.R. Buell, and D.S. Douches. 2012. Integration of two diploid potato linkage maps with the potato genome sequence. PLoS ONE 7: e36347. doi:10.1371/journal.pone.0036347 Freyre, R., and D.S Douches.1994. Development of a model for marker-assisted selection of specific gravity in diploid potato across environments. Crop Science 34: 1361-1368. Gebhardt, C., D. Bellin, H. Henselewski, W. Lehmann, J. Schwarzfischer, and J.P.T. Valkonen. 2006. Marker-assisted combination of major genes for pathogen resistance in potato. Theoretical and Applied Genetics 112: 1458-1464. Hackett, C.A, J.E. Bradshaw, R.C. Meyer, J.W. McNicol, D. Milbourne, and R. Waugh. 1998. Linkage analysis in tetraploid species: A simulation study. Genetical Research 71: 143-153. Hackett, C.A., J.E. Bradshaw, and J.W. McNicol, 2001. Interval mapping of QTLs in autotetraploid species. Genetics 159: 1819-1832. Hackett, C.A., I. Milne, J.E. Bradshaw, and Z. Luo. 2007. TetraploidMap for Windows: Linkage
146
map construction and QTL mapping in autotetraploid species. Journal of Heredity 98: 727-729. Haldane, J. B. S. 1930. Theoretical genetics of autopolyploids. Journal of Genetics 22: 349-372. Hamilton, J.P., C. N. Hansey, B. R. Whitty, K. Stoffel, A.N. Massa, A. Van Deynze, W.S. De Jong, D.S. Douches, and C.R. Buell. 2011. Single nucleotide polymorphism discovery in elite North American potato germplasm. BMC Genomics 12: 302-312. Hämäläinen, J. H, K.N. Watanabe, J.P.T. Valkonen, A. Arihara, R.L. Plaisted, E. Pehu, L. Miller, and S.A. Slack. 1997. Mapping and marker-assisted selection for a gene for extreme resistance to potato virus Y. Theoretical and Applied Genetics 94: 192-197. Hirsch, C. N., C. D. Hirsch, K. Felcher, J. Coombs, D. Zarka, A. Van Deynze, W. De Jong, R. E. Veilleux, S. Jansky, P. Bethke, D.S. Douches and C.R. Buell. 2013. Retrospective View of North American Potato (Solanum tuberosum L.) Breeding in the 20th and 21st Centuries. G3: Genes| Genomes| Genetics 3: 1003–1013. Haynes, K.G. and D.S. Douches. 1993. Estimation of the coefficient of double reduction in the cultivated tetraploid potato. Theoretical and Applied Genetics 85: 857-862. Jeong, J.C., H.C. Ok, O.S. Hur, C.G. Kim, C.S. Park, and S.Y. Kim. 2005. Raw material characteristics affect enzymatic browning of pre-peeled potato tubers. Journal of the Korean Society for Horticultural Science 46: 246-249. Jombart, T. 2008. Adegenet: a R package for the multivariate analysis of genetic markers. Bioinformatics Applications Note 24: 1403-1405. Karlsson, B.H., J.P. Palta and P. M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. Hort Science 41: 1213-1221. Kleinhenz, M.D., J.P. Palta, C.C. Gunter, and K.A. Kelling. 1999. Impact of source and timing of calcium and nitrogen applications on Atlantic potato tuber calcium concentrations and internal quality. Journal of the American Society for Horticultural Science 124: 498-506. Krawczak, M. 1999. Informativity assessment for biallelic single nucleotide polymorphisms. Electrophoresis 20: 1676-1681. Krokida, M.K., V. Oreopoulou, Z.B. Maroulis and D. Marinos-Kouris. 2001. Colour changes during deep fat frying. Journal of Food Engineering 48: 219-225. Leach, L. J., L. Wang, M. J. Kearsey, and Z. Luo. 2010. Multilocus tetrasomic linkage analysis using hidden Markov chain model. PNAS 107: 4270-4274. Li, L., J. Strahwald, H.R. Hofferbert, J. Lübeck, E. Tacke, H. Junghans, and C. Gebhardt. 2005. DNA variation at the invertase locus invGE/GF is associated with tuber quality traits in
147
populations of potato breeding clones. Genetics 170: 813-821. Li, L., M. J. Paulo, J. Strahwald, J. Lübeck, H. R. Hofferbert, E. Tacke, H. Junghans, J. Wunder, A. Draffehn, F. van Eeuwijk, and C. Gebhardt. 2008. Natural DNA variation at candidate loci is associated with potato chip color, tuber starch content, yield and starch yield. Theoretical and Applied Genetics 116: 1167-1181. Li, L.,E. Tacke, H. R. Hofferbert, J. Lübeck, J. Strahwald, A. M. Draffehn, B. Walkemeier, and C. Gebhardt. 2013. Validation of candidate gene markers for marker-assisted selection of potato cultivars with improved tuber quality. Theoretical and Applied Genetics 126: 1039-1052. Liu, P. Y., Z. Jun, and Y. Lu. 2006. Impacts of QTL x environment interactions on genetic response to marker-assisted selection. Acta Genetica Sinica 33: 63-71. Lopez‐Pardo, R., L. Barandalla, E. Ritter and J.I. Ruiz de Galarreta. 2013. Validation of molecular markers for pathogen resistance in potato. Plant Breeding 132: 246-251. Luo, Z. W., C. A. Hackett, J. E. Bradshaw, J. W. McNicol, and D. Milbourne, 2001. Construction of a genetic linkage map in tetraploid species using molecular markers. Genetics 157: 1369-1385. Matheis, G. 1987. Polyphenol oxidase and enzymatic browning of potatoes (Solanum tuberosum). II. Enzymatic browning and potato constituents. Chemie Mikrobiologie Technologie der Lebensmittel 11: 67-70. Mather, K. 1935. Reductional and equational separation of the chromosomes in bivalents and multivalents. Journal of Genetics 30: 53-78. Mather, K. 1936. Segregation and linkage in autotetraploids. Journal of Genetics 32: 287-314. McCann, L. C., P. C. Bethke, and P. W. Simon. 2010. Extensive variation in fried chip color and tuber composition in cold stored tubers of wild potato (Solanum) germplasm. Journal of Agricultural and Food Chemistry 58: 2368-2376. McCord, P. H., B. R. Sosinski, K. G. Haynes, M. E. Clough, and G. C. Yencho. 2011a. Linkage mapping and QTL analysis of agronomic traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Crop Science 51: 771-785. McCord, P.H., B. R. Sosinski, K.G. Haynes, M. E. Clough, and G. C. Yencho. 2011b. QTL mapping of internal heat necrosis in tetraploid potato. Theoretical and Applied Genetics 122: 129-142. McGuire, R. G., and A. Kelman. 1984. Reduced severity of Erwinia soft rot in potato tubers with increased calcium content. Phytopathology 74: 1250-1256.
148
Menéndez, C. M., E. Ritter, R. Schäfer-Pregl, B. Walkemeier, A. Kalde, F. Salamini, and C. Gebhardt. 2002. Cold sweetening in diploid potato: mapping quantitative trait loci and candidate genes. Genetics 162: 1423-1434. Meyer, R. C., D. Milbourne, C. A. Hackett, J. E. Bradshaw, J. W. McNichol, and R. Waugh. 1998. Linkage analysis in tetraploid potato and association of markers with quantitative resistance to late blight (Phytophthora infestans). Molecular and General Genetics 259: 150-160. Muller, H.J. 1914. A new mode of segregation in Gregory’s tetraploid Primulas. The American Naturalist 48: 508-512. Nei, M. 1972. Genetic distance between populations. The American Naturalist 106: 283-292. Newman, S.M., N.T. Eannetta, H.Yu, J.P. Prince, M.C. de Vicente, S.D. Tanksley, and J.C. Steffens. 1993. Organisation of the tomato polyphenol oxidase gene family. Plant Molecular Biology 21: 1035-1051. Oberhagemann, P., C. Chatot-Balandras, R. Schäfer-Pregl, D. Wegener, C. Palomino, F. Salamini, and C. Gebhardt. 1999. A genetic analysis of quantitative resistance to late blight in potato: towards marker-assisted selection. Molecular Breeding 5: 399-415. Olsen, N. L., L. K. Hiller, and L. J. Mikitzel. 1996. The dependence of internal brown spot development upon calcium fertility in potato tubers. Potato Research 39: 165-178. Ozgen, S., B. H. Karlsson, and J. P. Palta. 2006. Response of potatoes (cv. ‘Russet Burbank’) to supplemental calcium applications under field conditions. Tuber calcium, yield, and incidence of brown spot. American Journal of Potato Research 83: 195-204. Palta, J.P. 1996. Role of calcium in plant responses to stresses: Linking basic research to the solution of practical problems. HortScience 31: 51-57. Paterson, A. H, Y. Saranga, M. Menz, C. X. Jiang, and R. J. Wright. 2003. QTL analysis of genotype × environment interactions affecting cotton fiber quality. Theoretical and Applied Genetics 106: 384-396. Pineda, O., M.W. Bonierbale, R.L. Plaisted, B.B. Brodie, and S.D. Tanksley. 1993. Identification of RFLP markers linked to the H1 gene conferring resistance to the potato cyst nematode Globodera rostochiensis. Genome 36: 152-156. Pompanon, F., A. Bonin, E. Bellemain, and P. Taberlet. 2005. Genotyping errors: causes, consequences and solutions. Nature Reviews Genetics 6: 847-846. Potato Genome Sequencing Consortium. 2011. Genome sequence and analysis of the tuber crop potato. Nature 475: 189-195.
149
R Development, Core. Team. 2011: R: A Language and Environment for Statistical Computing. R Foundation for Statistical Computing, Vienna, Austria. Rodriguez-Saona, L.E., and R. Wrostald. 1997. Influence of potato composition on chip color quality. American Journal of Potato Research 74: 87-106. Sagredo, B., N. Balbyshev, A. Lafta, H. Casper, and J. Lorenzen. 2009. A QTL that confers resistance to Colorado potato beetle (Leptinotarsa decemlineata [Say]) in tetraploid potato populations segregating for leptine. Theoretical and Applied Genetics 119: 1171-1181. Sapers, G.M., F.W. Douglas, A. Bilyk, A.F. Hsu, H.W. Dower, L. Garzarella, and M. Kozempel. 1989. Enzymatic browning in Atlantic potatoes and related cultivars. Journal of Food Science 54: 362-365. Simko, I, S. Costanzo, K. G. Haynes, B. J. Christ, R. W. Jones. 2004. Linkage disequilibrium mapping of a Verticillium dahliae resistance quantitative trait locus in tetraploid potato (Solanum tuberosum) through a candidate gene approach. Theoretical and Applied Genetics 108: 217-224. Śliwka, J., H. Jakuczun, P. Kamiński, and E. Zimnoch-Guzowska. 2010. Marker-assisted selection of diploid and tetraploid potatoes carrying Rpi-phu1, a major gene for resistance to Phytophthora infestans. Journal of Applied Genetics 51: 133-140. Sorensen, T. 1948. A method of establishing groups of equal amplitude in plant sociology based on similarity of species content and its application to analysis of vegetation on Danish commons. Biologiske Skrifter 5: 1-34. Sterret, S.B. and M.R. Henninger 1997. Internal heat necrosis in the midatlantic region-influence of environment and cultural management. American Journal of Potato Research 74: 233-243. Tétard-Jones, C., M.A. Kertesz, and R.F. Preziosi. 2011. Quantitative trait loci mapping of phenotypic plasticity and genotype-environment interactions in plant and insect performance. Philosophical Transactions of the Royal Society B: Biological Sciences 366: 1368-1379. Tzeng, K. C., A. Kelman, K. E. Simmons, and K. A. Kelling. 1986. Relationship of calcium nutrition to internal brown spot of potato tubers and sub-apical necrosis of sprouts. American Journal of Potato Research 63: 87-97. Ungerer, M.C., S.S. Halldorsdottir, M.D. Purugganan, and T.F.C. Mackay. 2003. Genotype-environment interactions at quantitative trait loci affecting inflorescence development in Arabidopsis thaliana. Genetics 165: 353-365. Urbany, C., B. Stich, L. Schmidt, L. Simon, H. Berding, H. Junghans, K.H. Niehoff, A. Braun, E. Tacke, H.R. Hofferbert, J. Lübeck, J. Strahwald, and C. Gebhardt. 2011. Association genetics in Solanum tuberosum provides new insights into potato tuber bruising and enzymatic tissue discoloration. BMC genomics 12: 7.
150
Vega, S.E., Palta, J.P., Bamberg, J.B. 2006. Exploiting cultivated germplasm to breed for enhanced tuber calcium accumulation ability [abstract]. American Journal of Potato Research 83: 136. Werij, J.S., B. Kloosterman, C. Celis-Gamboa, C.R. de Vos, T. America, R.G. Visser, and C.W. Bachem. 2007. Unravelling enzymatic discoloration in potato through a combined approach of candidate genes, QTL, and expression analysis. Theoretical and Applied Genetics 115: 245-252. Wu, R., M. Gallo-Meagher, R. C. Littell, and Z. B. Zeng. 2001. General polyploid model for analyzing gene segregation in outcrossing tetraploid species. Genetics 159: 869-882. Würschum, T. 2012. Mapping QTL for agronomic traits in breeding populations. Theoretical and Applied Genetics 125: 201-210. Zhuang, J. Y., H.X. Lin, J. Lu, H.R. Qian, S. Hittalmani, N. Huang, and K.L. Zheng. 1997. Analysis of QTL× environment interaction for yield components and plant height in rice. Theoretical and Applied Genetics 95: 799-808.
151
TABLES Table 4.1. Selecting SNP markers from the 8303 SNP genotyped in the Atlantic x Superior population to be used in tetraploid linkage mapping. Numbers indicate remaining SNP markers and genotypes.
Criteria to remove from the analysis Remaining
SNP Remaining Genotypes
Initial SNP and genotypes 8303 184 Questionable and bad SNPs 7666 Missing data in all genotypes 5024 No hit or more than two hits 4605† Mixed and twice-sampled genotypes 176 Located in two positions in the genome 4498 Missing in one or both parents 4088 Keep only simplex markers 968‡ SNP with >16.67% double reduction 962 Markers with high segregation distortion (<0.05/#markers)
913
Duplicated markers 628 Markers with <130 individuals typed 619 Individuals with <580 markers typed 166 Duplicated genotypes (>0.99 identity) 151 Markers with high segregation distortion (<0.001) 614 Markers in coupling with more than two homologs 600 Final SNP and genotypes 600 151 † Used in the genetic structure analysis
‡ Data used in the double reduction analysis
152 Table 4.2. The Atlantic x Superior linkage map using SNP markers and QTL identified by parent and chromosome
153 Table 4.3. Broad-sense heritabilities per year and pooled data, and QTL present in Atlantic (ATL) and Superior (SUP) for tuber calcium (TC), tuber yield (TY), specific gravity (SG), chip color (CC), browning (EB), agtron (AG), chip lightness (L), chip redness (a), and chip yellowness (b), incidence of hollow heart (HH), black spot bruise (BB), pitted scab (PS), incidence (PS-E) and severity (SPS-E) of pitted scab in a high disease pressure field.
75 47 28 †Heritabilities presented in Chapter 3 ‡ Visual scale from 1 to 5 used as a numerical variable for genetic analysis, where 1 is very light and 5 is very dark P-value symbols: (**) p <0.01, (*) p < 0.05, (.) p < 0.1, (NS) p ≥0.1
154 Table 4.4. Total and mapped population size from the Atlantic x Superior population Year Field Population
size Mapped population size
Single year evaluations 2009 Standard 121 89 2010 Standard 158 128 2011 Standard 40 32 2011 High disease pressure 128 106 2012 High disease pressure 87 70 Pooled evaluations 2009-2010 Standard 168 128 2010-2011 Standard 160 128 2011-2012 High disease pressure 130 106 2009-2011 Standard 189 128
155 Table 4.5. QTL and SNP markers with significant effects (p<0.01) identified in the Atlantic x Superior population phenotyped during the 2009-2012 seasons for tuber calcium, tuber quality traits and pitted scab
Trait Chr. Pa cM CI LOD %V BM Year detected† Sig. SNP (effect) TC I ATL 101.7 98.7-113.7 2.7 10.7 dup(-) Pool1, 09, 10 TC III ATL 50.0 40-62 3.1 12.2 addit. Pool1, 09, 11 c3_26996294(-) TC V ATL 56.0 42-68 2.5 7.7 simp(+) Pool1, Pool2, 09,
c12_53697469(-) †Years with peaks, at least one above the permutations threshold. Data on the table corresponds to the year in bold and the best model for the dosage and effects. Chr.=chromosome, Pa=parents, cM=centimorgan, CI=confidence interval, %V=%variance explained, BM=best model, addit.=addititve, complex inheritance, dup(+)=duplex, the dominant allele increases the mean, dup(-)=duplex, the dominant allele decreases the mean, simp(+)=simplex, the mean increases if present, simp(-)=simplex, the mean decreases if present, (+)=mean increase, and (-)=mean decrease, TC = tuber calcium, TY = total yield, SG = specific gravity and EB = enzymatic browning. Population size varied from year to year as indicated in Table 4.4. BLUP values were estimated including the mixed and duplicated genotypes.
Table 4.5 continues on the next page
156 Continue
Trait Chr. Pa cM CI LOD %V BM Year detected† Sig. SNP (effect) EB I ATL 81.7 70.2-92.7 3.0 7.4 addit. Pool, 10, 11 c1_66696384(-)
c1_67731638(-) c1_67731638(-)
EB V SUP 40.0 32-51 3.2 7.9 addit. Pool, 10 EB VI SUP 53.5 12.5-53.5 2.7 8.6 simp(+) Pool, 10 c6_52436008(+) EB VII SUP 6.0 0-9 2.4 28.5 simp(+) Pool, 11 EB VIII ATL 50.9 27-62 3.2 13.0 simp(-) Pool, 10 AG I SUP 68.0 53-101 3.5 20.7 dup(-) 09 c1_68432076(-)
c1_80006794(+) c1_80510838(+)
AG II SUP 86.0 58-88.7 2.0 9.6 simp(-) 09 c2_44147246(-) AG IV SUP 4.0 0-18 3.8 15.8 addit. 09 AG V SUP 6.0 0-17 2.5 11.8 dup,dom+ 09 c5_4797133(-) AG VII ATL 74.0 62.5-77.5 4.1 18.3 simp(+) 09 c7_44184334(+)
CC VIII ATL 84.9 78.9-86.9 3.9 13.8 addit. Pool1, Pool2, 09, 10
CC IX ATL 89.2 87.2-94.2 2.7 14.9 dup(+) Pool1, Pool2, 10, 11
CC XI ATL 14.2 10.2-35.2 4.8 41.4 addit. 11 L VII ATL 76.0 68-88.2 3.3 11.0 dup(-) Pool, 10 c7_45450767(-) L IX ATL 89.2 83.2-99.7 2.2 11.8 dup(-) Pool, 10, 11 c9_47031950(-)
† Years with peaks, at least one above the permutations threshold. Data on the table corresponds to the year in bold and the best model for the dosage and effects. Chr.=chromosome, Pa=parents, cM=centimorgan, CI=confidence interval, %V=%variance explained, BM=best model, addit.=addititve, complex inheritance, dup(+)=duplex, the dominant allele increases the mean, dup(-)=duplex, the dominant allele decreases the mean, simp(+)=simplex, the mean increases if present, simp(-)=simplex, the mean decreases if present, (+)=mean increase, and (-)=mean decrease, EB=enzymatic browning, AG=chip color in agtrons, CC=visual rate of chip color, L=chip lightness, A=chip redness, and B=chip yellowness. Population size varied from year to year as indicated in Table 4.4. BLUP values were estimated including the mixed and duplicated genotypes.
Table 4.5 continues on the next page
157 Continue
Trait Chr. Pa cM CI LOD %V BM Year detected† Sig. SNP (effect) A VIII ATL 86.9 82.9-86.9 3.3 10.5 simp(-) Pool, 10, 11 c8_35411921(-)
c8_39195950(-) c8_40106266(-)
A IX ATL 89.2 76.2-96.2 2.4 13.3 dup(+) Pool, 10, 11 B V ATL 38.0 28-47 3.1 8.4 addit. Pool, 10 c5_4792932(-)
c5_5117198(-) B VIII ATL 86.9 84.9-86.9 2.1 6.6 dup(+) Pool, 10
HH I ATL 45.7 31.7-48.7 5.4 57.0 addit. 11 HH III ATL 0.0 0-16 4.6 17.0 addit. Pool, 09, 10 c3_14240272(+)
c5_14027989(-) HH VI ATL 70.0 67-74 3.4 19.0 dup(-) Pool, 10 c6_47540578(-) HH IX ATL 101.2 99.2-104.2 4.8 71.7 addit. 09, 11 BB X ATL 21.0 8-31 4.3 23.9 dup(+) 09 BB XI ATL 50.2 38.2-64.2 3.0 7.1 dup(-) 10
†Years with peaks, at least one above the permutations threshold. Data on the table corresponds to the year in bold and the best model for the dosage and effects. Chr.=chromosome, Pa=parents, cM=centimorgan, CI=confidence interval, %V=%variance explained, BM=best model, addit.=addititve, complex inheritance, dup(+)=duplex, the dominant allele increases the mean, dup(-)=duplex, the dominant allele decreases the mean, simp(+)=simplex, the mean increases if present, simp(-)=simplex, the mean decreases if present, (+)=mean increase, and (-)=mean decrease, HH=incidence of hollow heart and BB=incidence of blackspot bruise. Population size varied from year to year as indicated in Table 4.4. BLUP values were estimated including the mixed and duplicated genotypes.
Table 4.5 continues on the next page
158 Continue
Trait Chr. Pa cM CI LOD %V BM Year detected† Sig. SNP (effect) PS III ATL 48.0 39-53 3.2 10.7 dup(-) Pool c3_29440562(-) PS IV SUP 24.0 10-27 4.6 13.7 addit. Pool, 09, 10, 11 c4_6259961(+)
c4_8401238(+) PS V SUP 26.0 17.5-37.5 3.1 7.8 addit. 10 c5_41648687(-) PS XII ATL 6.0 0-18.5 3.8 10.2 addit. Pool, 09, 10 c12_9644160(-)
PS-E II SUP 18.0 14-28 4.2 23.0 simp(+) Pool, 11, 12 c2_1498967(-) c2_13213414(+)
PS-E III ATL 80.0 67-93 3.2 12.8 dup(-) Pool, 12 c3_36273874(-) PS-E V SUP 6.0 0-8 2.6 9.7 addit. Pool, 11, 12 c5_3536366(-)
c5_4796997(-) c5_5845469(-) c5_6413644(-)
PS-E VI SUP 3.5 3.5-16 1.9 7.8 dup(+) Pool, 12 PS-E VII ATL 38.0 20-62 3.2 10.8 addit. Pool, 12 PS-E X SUP 24.0 12.5-28 2.2 9.8 dup(-) Pool, 11, 12 PS-E XI ATL 74.2 80.2-86.2 3.1 12.1 dup(-) Pool, 11, 12
SPS-E II SUP 18.0 0-30 2.6 14.7 simp(+) Pool, 11, 12 c2_13213414(+) SPS-E III ATL 80.0 67-86 4.4 17.6 dup(-) Pool, 11, 12 c3_36273874(-) SPS-E V SUP 0.0 0-10 3.4 9.5 addit. Pool, 11, 12 c5_3536366(-)
c5_4796997(-) c5_5845469(-) c5_6413644(-)
SPS-E VI SUP 3.5 0-14.5 3.1 10.1 addit. Pool, 12 SPS-E IX ATL 101.2 97.2-103.7 4.0 49.6 addit. 12 SPS-E X SUP 24.0 20-27.5 2.5 11.0 dup(-) Pool, 11, 12 c10_39831786(-)
c10_40518785(+) SPS-E XI ATL 80.2 74.2-86.2 4.0 17.6 dup(-) Pool, 11
†Years with peaks, at least one above the permutations threshold. Data on the table corresponds to the year in bold and the best model for the dosage and effects. Chr.=chromosome, Pa=parents, cM=centimorgan, CI=confidence interval, %V=%variance explained, BM=best model, addit.=addititve, complex inheritance, dup(+)=duplex, the dominant allele increases the mean, dup(-)=duplex, the dominant allele decreases the mean, simp(+)=simplex, the mean increases if present, simp(-)=simplex, the mean decreases if present, (+)=mean increase, and (-)=mean decrease, PS= incidence of pitted scab in the standard field, PS-E= incidence of pitted scab in the high disease pressure field, and SPS-E= severity of pitted scab in the high disease pressure field. Population size varied from year to year as indicated in Table 4.4. BLUP values were estimated including the mixed and duplicated genotypes.
159
FIGURES Figure 4.1. Genetic structure based on principal components analysis of Atlantic x Superior progenies and parents explaining 8.53% of the total genotypic variance. ATL3* had been misidentified as Superior, ATL4* had been misidentified as B-028, and SUP2* had been misidentified as B-064.
160 Figure 4.2. Pairwise comparison of genotypic identity. Identities higher than 0.99 indicating duplicated genotypes are shown with an arrow.
Pairwise Comparison of Genotype Identity
Pairwise identity
Nu
mb
er o
f p
air-
wis
e co
mp
aris
on
s
0.0 0.2 0.4 0.6 0.8 1.0
020
060
010
00
Duplicated genotypes(identity>0.99)
161 Figure 4.3. Genetic maps for Atlantic and Superior based on 600 simplex SNP markers.
162 Figure 4.4. Pedigree of the parents
Atlantic Superior
Plots obtained from the Potato pedigree database (van Berloo et al. 2007)
163
Figure 4.5. Genetic maps and QTL located on chromosome 1 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
164 Figure 4.6. Genetic maps and QTL located on chromosome 2 of Atlantic and Superior. . Error bars indicate the ± 1 LOD confidence interval for the QTL location
165 Figure 4.7. Genetic maps and QTL located on chromosome 3 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
166 Figure 4.8. Genetic maps and QTL located on chromosome 4 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
167 Figure 4.9. Genetic maps and QTL located on chromosome 5 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
168 Figure 4.10. Genetic maps and QTL located on chromosome 6 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
169 Figure 4.11. Genetic maps and QTL located on chromosome 7 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
170 Figure 4.12. Genetic maps and QTL located on chromosome 8 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
171 Figure 4.13. Genetic maps and QTL located on chromosome 9 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
172 Figure 4.14. Genetic maps and QTL located on chromosome 10 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
173 Figure 4.15. Genetic maps and QTL located on chromosome 11 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
174 Figure 4.16. Genetic maps and QTL located on chromosome 12 of Atlantic and Superior. Error bars indicate the ± 1 LOD confidence interval for the QTL location
175 Figure 4.17. LOD profiles in different years of evaluation for the QTL for tuber calcium concentration present on chromosome 1 of Atlantic
176 Figure 4.18. SNP marker with potential to be used in marker assisted selection targeting a QTL located on chromosome 2 of Superior and significantly increases pitted scab incidence and severity when present
Nulliplex(BBBB)
Simplex(ABBB)
1020
3040
5060
Genotype
% I
nci
den
ce (
2011
-20
12)
Pitted Scab IncidenceSNP c2_13213414Chr.2 of Superior
Nulliplex(BBBB)
Simplex(ABBB)
0.0
0.5
1.0
1.5
2.0
2.5
Genotype
Pit
s p
er
tub
er
(20
11-2
012)
Pitted Scab SeveritySNP c2_13213414Chr.2 of Superior
QTL: %V = 23%
SNP: p < 0.01
QTL: %V = 14.7%
SNP: p < 0.01
This QTL explained 23% of the variance for pitted scab incidence and 14.7% of the variance for pitted scab severity in the high disease pressure field. The marker c2_13213414 was associated to this QTL with p <0.01.
177
CHAPTER 5
Over-expressing the Vacuolar Antiporter CAX1 in the Potato Cultivar Atlantic: Phenotype of the Transformed Clones and Implications to Understand the Role of Calcium on Tuber
Quality and Plant Health
ABSTRACT
In vitro-grown plantlets of the potato variety Atlantic were transformed with an Agrobacterium
strain LBA4404 harboring a short version of the Cation Exchanger 1 (CAX1) gene, a tonoplast
calcium-cation transmembrane transporter of Arabidopsis thaliana, driven by the CaMV35S
promoter and the cdc2a promoter. The objective of this study was to evaluate the effect of the
increased calcium transport into the vacuole on the calcium content of in-vitro grown plantlets as
well as calcium distribution in leaves and tubers of greenhouse grown potato plants. Greenhouse
and in-vitro evaluations showed calcium deficiency symptoms in the transgenic clones when
growing under sufficient amounts of media or soil calcium content. Some clones needed up to
ten times the normal media or soil calcium content to grow without calcium deficiency
symptoms. We were able to reach several conclusions from the evaluation of these plants. First,
the increased transport of calcium into the vacuoles of the transgenic lines generates calcium
deficiency symptoms in the plant such as apical shoots damage and leaf margin necrosis. Second,
as most calcium transported to the foliage and tubers moves with water via the apoplast; growing
the transgenic plants with higher amounts of media or soil calcium mitigated calcium deficiency
symptoms in the shoots. Third, the sub-cellular localization analysis of calcium indicates that
calcium is being stored as calcium oxalate in the vacuoles of transgenic plants and therefore
becomes unavailable. Calcium oxalate crystals were not observed in the leaves of the wildtype
178
plants. Fourth, the deficiency symptoms, plant health damage and internal defects appear to be
caused by a reduction of the calcium concentration in the apoplast resulting in compromised
plant health and tuber quality. Our results show that maintaining the homeostasis of the sub-
cellular localization of calcium is very important for tuber quality and plant health.
INTRODUCTION
Calcium is essential for the integrity, maintenance and formation of cell-membrane systems and
cell walls of plants (Kirkby and Pilbeam 1984, Hirschi 2004). Calcium stabilizes cell membranes
by connecting adjacent polar head groups of membrane lipids (Legge et al. 1982, Palta 1996,
Hirschi 2004). It also protects membranes from adverse effects of stress such as salinity (Cramer
et al. 1985), freezing injury (Arora and Palta 1986), and heat stress (Tawfik et al. 1996,
Kleinhenz and Palta 2002, Saidi et al. 2009). Calcium is also a component of cell walls that
forms stiff gels through Ca+2-mediated crosslinking of its carboxyl groups through ionic and
coordinate bonds with a pectin component called homogalacturonan or polygalacturonic acid
(Cosgrove 2005).
Calcium homeostasis maintains the concentration of extracellular calcium in the milimolar range
whereas the cytoplasmic concentration of calcium is in the nanomolar to micromolar range
(Kauss 1987, Gilroy et al. 1993). Calcium antiporters and efflux pumps are important to maintain
the cytoplasmic calcium at low levels and restoring the normal Ca2+ levels after perturbation
(Tuteja and Mahajan 2007). A constant supply of Ca2+ in the range of 1-10 mM is required to
maintain normal growth and development at the whole-plant level (Epstein 1972, Clarkson and
Hanson 1980). Calcium moves with water mainly by the apoplast and its redistribution within
179
the plant is very low (Bangerth 1979, Clarkson 1984, Busse and Palta 2006). In plant tissues that
are deficient in calcium, cell walls disintegrate and tissue collapses resulting in necrosis (Kirby
and Pilbeam 1984, Marschner 1995).
The H+/ Ca2+ antiporter 1, CAX1, is a tonoplast calcium antiporter identified and cloned from
Arabidopsis thaliana by suppressing yeast mutants defective in vacuolar Ca2+ transport (Hirschi
et al. 1996). Several CAX antiporters have been identified in Arabidopsis with different ion
specificities such as CAX2 that transports heavy metals (Hirschi et al. 2000), CAX3 that
transports Ca2+ mainly in roots (Manohar et al. 2011), and CAX4 (Cheng et al. 2002) among
others. The N-terminal regulatory region of CAX1 acts as an autoinhibitory domain for H+/ Ca2+
transport activity when expressed in yeast (Pittman and Hirschi 2001, Pittman et al. 2002). This
region was removed to generate a deregulated short version, denominated sCAX1 (Cheng et al.
2005). The over-expression of sCAX1 in tobacco (Hirschi 1999), carrots (Park et al. 2004),
potato (Park et al. 2005a) and tomato (Park et al. 2005b) has shown to increase calcium content
in leaves, roots, tubers and fruits, respectively.
The Arabidopsis mutants of cax1 showed a reduction of Ca2+/ H+ antiport activity, a reduction in
tonoplast V-type H-translocating ATPase activity, an increase in tonoplast Ca2+-ATPase activity,
and the lack of CAX1 was compensated by increased expression of CAX3 and CAX4 (Cheng et
al. 2003). Also, cax1 mutants have shown characteristics of plants that grow in serpentine soils
which include greater tolerance for low Ca2+, increased tolerance to Mg2+ and higher Mg2+
requirement for maximum growth, and Mg2+ exclusion from leaves (Bradshaw 2005). In
addition, cax1/cax3 double mutants displayed a severe reduction in growth, including leaf tip and
flower necrosis and pronounced sensitivity to exogenous Ca2+ and other ions (Cheng et al. 2005).
CAX1 has been recognized as a key regulator of the apoplastic Ca2+ concentration. CAX1 keeps
180
apoplastic Ca2+ low through compartmentation into mesophyll vacuoles that is essential for
sufficient plant function and productivity (Conn et al. 2011). The unbound extracellular Ca2+
must be maintained in equilibrium across the apoplast/symplast boundary. In leaves, rather than
the cell wall, the vacuole serves as the reservoir for excess accumulation of Ca2+ (Robertson
2013).
Ca2+ deficiency-like symptoms have been reported in transgenic lines of tobacco that are over-
expressing CAX1 (Hirschi 1999). In tomatoes expressing sCAX1, a modest compromise of plant
growth had been reported by Park at al. (2005b); however, another study reported severe
deficiency symptoms and high incidence of blossom end rot in fruits of tomato plants (de Freitas
et al. 2011). A recent study has demonstrated that some of these Ca2+ deficiency symptoms can
be relieved by the expression of another gene encoding for the calcium binding protein
calreticullin (Wu et al. 2012). These deficiency symptoms have not been reported in potato. Park
et al. (2005a) reported that increased levels of sCAX1 do not appear to alter potato growth and
development, nor tuber morphology or yield. They further reported an increased calcium
concentration of various tuber tissues in transgenic plants.
The objectives of our study are to determine if there are calcium deficiency symptoms in potato
caused by the expression of sCAX1; to evaluate the effects on plant health, tuber quality and
calcium content of increased Ca2+ transport into the vacuole by the expression of sCAX1, to test
the hypothesis that because Ca2+ is sequestered in the vacuole there is less Ca2+ available in the
apoplast that weakens the cell walls causing calcium deficiency symptoms and internal defects in
the tubers, and to test if the deficiency symptoms can be relieved by supplemental calcium.
181
MATERIAL AND METHODS
Plant material and transformation constructs
In vitro-grown plantlets 5-6 weeks old of the potato variety Atlantic were transformed with the
Agrobacterium strain LBA4404 harboring the small version of the CAX1 gene from Arabidopsis
thaliana. Two gene constructs developed by Park et al (2005a), one carrying the CaMV35S
promoter, CaMV35S::sCAX1; and another with the cdc2a promoter, cdc2a::sCAX1, were used
for transformation. The CaMV and the cdc2a promoters were chosen due to their high and non-
tissue specific transcription (Odell et al. 1985, Doerner et al. 1996). The Atlantic transgenic lines
were named using AT1 for lines carrying the CaMV35S::sCAX1 construct and AT2 for lines
carrying the cdc2a::sCAX1 construct. The numbers following indicate plate number, explants
number, and shoot number separated by an underscore. In addition, four Russet Norkotah
transgenic lines, CAX1 #34, CAX1 #36 K-3-1, CAX1 #36 K-3-2, and CAX1 #49; and a
wildtype line of Russet Norkotah RN#12, generated by Park et al. (2005a) were evaluated as
controls. The Russet Norkotah sCAX1-expressing lines, the Russet Norkotah wildtype and the
gene constructs were kindly provided by Kendall Hirschi from Texas A&M University.
Potato transformation
Stem cuttings of approximately 1cm long were used for transformation. Stem cuttings and the
Agrobacterium strain containing the construct were co-cultured in a media containing Trans-
media). After 4 to 6 days of co-cultivation under low light, stem cuttings were transferred to ZIG
182
media + Cefotaxime 100 mg/L + Kanamycin 50mg/L (ZIG++); and then, transferred to ZIG ++
media every two weeks until some shoots appeared in the callus formed at both ends of the stem
cuttings. Shoots of at least 2cm long were transferred to Murashige-Skoog media + Kanamycin
50mg/L (MS+), if the explants showed any signs of Agrobacterium growth they were grown in
MS + Cefotaxime 100 mg/L + Kanamycin 50mg/L (MS++). If the explants grew and generated
roots in either MS+ or MS++ media, they were considered positively transformed; otherwise,
they were eliminated. Wildtype explants were also grown in MS+ and MS++ media as negative
controls.
Polymerase chain reaction (PCR)
Genomic DNA was isolated from tissue culture plantlets using an extraction protocol based on a
CTAB 2X buffer (Doyle and Doyle 1990); with a tissue lysis step of 20 seconds using 1/4-inch
ceramic beads performed with a Bead Beater (Biospec) or a Fast Prep-12 instrument (MP
Biomedicals); and treated with RNAse A (Invitrogen) after isolation. The presence of the
transgene was determined by polymerase chain reaction (PCR) of the neomycin
phosphotransferase (NPT II) marker gene. The sequences of the primers used for the NPT II gene
were 5’- AGC CAA CGC TAT GTC CTG AT-3’ and 5’- GAA GGG ACT GGC TGC TAT TG-
3’(GenBank accession: U09365).The presence of a fragment of 370bp indicated that the plant
had been transformed.
PCR efficiency and copy number determination
Transgene copy number determination based on real time PCR offers a fast, inexpensive and
high throughput alternative (Yuan et al. 2007). In real time PCR, the threshold cycle (Ct) is the
183
PCR cycle at which fluorescence exceeds background and a significant increase in fluorescence
is observed (Higuchi et al. 1993). Another important variable to set up a quantitative PCR
(qPCR) assay is the PCR efficiency. Efficiency close to 100% is ideal and an indicator of the
quality of the assay (Saunders 2004). The DNA concentration was determined using a nanodrop
ND-2000 (Thermo Scientific). Template amounts of 50, 100, 200 and 400 ng of the same
transgenic clone in triplicate were used for the calibration curve. The efficiency of the PCR
reactions in this study was estimated by the R2 and the slope of the linear relationship between
the Ct values obtained from a CFX96 real time detection system (BIO-RAD) versus the log2
transformed DNA amounts. The slope should be close to -1 and R2 close to 100%. Primers that
target the CAX1 gene, the NPT II gene and single copy genes including the granule-bound starch
(GBSS, also called waxy) gene and EF1- were tested to determine their PCR efficiency (Figure
5.1). The primer sequences used were 5’-GAAGAAATCGCTCCACTTGC-3’ and 5’-
CTCCCCAGCAAAAACCAAT-3’to amplify a 387bp fragment of the GBSS gene (Genebank
accession: X83220); the 5’-ATTGGAAACGGATATGCTCCA-3’ and 5’-
TCCTTACCTGAACGCCTGTCA-3’ to amplify a 101bp fragment of the EF1- gene (Nicot et
al. 2005); the 5’-AGACAATCGGCTGCTCTGAT-3’ and 5’-AGTGACAACGTCGAGCACAG-
3’ to amplify a 370bp fragments of the NPT II gene (GenBank accession: U09365); and the 5’-
GCAACAGGAGGAGGAGTTTT-3’ and 5’-AACCCACCCACAAGAAGAAT-3’ to amplify a
250bp fragment of the CAX1 gene (TAIR: AT2G38170). The PCR efficiency was high for all
primer pairs evaluated (slope < -0.95, R2 > 0.97). The PCR reaction mix contained 1X Ex-Taq
buffer (Takara), 3mM MgCl2, 300µM DNTPs, 0.8µM of each primer, 0.8X Evagreen (Biotium),
and 0.125 units of Takara Ex-Taq Hot Start version (Takara) in a final volume of 25 µL.
184
The primer pairs used for the copy number determination on most lines were those targeting the
EF1- and CAX1 genes that presented slightly higher PCR efficiency. For each transgenic line,
two replicates of the real time reactions for both genes were performed in different wells of the
same plate. One line was used in all plates as a plate variation control because not all lines could
fit in a single plate. A mixed model that accounts for all the sources of variation was used to
determine the predicted Ct values for each line and gene. The model used was: Ct~ log2dna +
gene + line + plate + line*gene, where log2dna and gene were used as fixed effects variables and
line, plate and line*year were used as random effects variables. The predicted Ct values were
used to estimate copy number using a ΔΔCt method as described by Livak et al.(1995). An
external calibration curve was used to normalize for the variation in DNA concentration. The
formula to estimate copy number is copy = 2^ ΔΔCt, where ΔΔCt = (CtEF1α-CtCAX1)unknown-(CtEF1α-
CtCAX1)reference. The lines with known copy number were CAX1#34 (1 copy), CAX1#36 k-3-1
(1copy), CAX1#36 k-3-2 (1copy), and CAX1#49 (2 copies).
Test for the expression of CAX1 in the transgenic lines by the quantification of transcripts
The tips of two tissue culture plantlets measuring approximately 1.5cm were removed, quick
frozen in liquid nitrogen and stored at -70ºC until RNA extraction. Total RNA was extracted
using the RNAeasy Plant Mini kit (Qiagen) and the contaminating DNA was removed using the
Turbo DNA-freeTM kit (Ambion) according to the manufacturer’s recommendations. The cDNA
was produced using the Superscript III First Strand cDNA Synthesis System (Promega) from
1µg of purified total RNA according to the manufacturer specifications. The RNA and cDNA
concentrations were determined using a nanodrop ND-2000 (Thermo Scientific). The PCR
185
efficiency of the qPCR reaction was tested for the housekeeping gene EF1-α and the CAX1 gene
using different amounts, 25, 50 and 100ng, of cDNA from a single clone as template. The EF1-α
was chosen as the control housekeeping gene because it was one of the most stable under biotic
and abiotic stress in potato (Nicot et al. 2005)
The PCR reaction and primers were the same as the copy number assay. The linear model used
to determine the expected Ct values was Ct ~ line + gene + treatment + log2cdna + line*gene,
where all variables were treated as fixed effects. The relative difference between the amount of
transcripts for the reference gene EF1-α and the transgene CAX1 and the relative difference of
the amount of transcripts among lines were determined by the 2ΔCt and the 2ΔΔCt methods,
respectively. An internal standard curve generated with 25, 50, 100 and 200ng of cDNA from the
same clone was included in this evaluation.
In-vitro culture maintenance
Wild type clones were maintained using a modified Murashige and Skoog (MS) basal media
(Murashige and Skoog 1962) which contains 3mM Calcium in the form of CaCl2. Transgenic
clones were maintained at different levels of calcium in the media depending on which level
reduced significantly apical shoot damage. These levels of calcium were determined by
cultivating the transgenics in a range of calcium concentrations. Calcium levels of 6mM, 10mM,
15mM and 20mM were used for propagation of transgenic plants. The calcium concentration that
gave the healthiest plants was chosen to propagate these transgenics. Plantlets of transgenic
clones grown at their optimal calcium level had adequate development to be used in the tissue
culture and greenhouse experiments.
186
Tissue Culture Experiments
Initial in-vitro experiment: Higher demand of calcium by transgenic clones
Each transgenic clone was grown in tissue culture using standard MS media containing 3mM,
10mM and 30mM Ca+2 levels in the form of CaCl2. The transgenic clones were evaluated using
an augmented design with a single evaluation for most transgenic lines and three evaluations for
the wildtype clone. Each evaluation unit consisted on a set of 4 to 8 plantlets for the transgenic
lines and the wildtype lines. Most clones were included in this experiment. From these
experiments and additional tissue culture maintenance evaluations, the optimum media calcium
concentration for each transgenic line was defined.
In-vitro experiments: sCAX1 effects on plant health and calcium concentration
Each transgenic clone was grown at sufficient calcium concentration, 3mM, and one or more
high calcium concentrations depending on the aim of the experiment. For this purpose MS media
with calcium concentrations of 3mM,10mM, 15mM and 30mM; or 3mM and 15mM have been
used in this study in the form of CaCl2. The transgenic clones were evaluated using a randomized
block design with three replications of each transgenic line and the wildtype. Each replication
contained 6 to 8 plants. A subset of clones with a single copy of sCAX1 was evaluated in these
experiments.
Greenhouse Experiments
These experiments were conducted in order to determine the changes in total leaf calcium, water
soluble calcium, acid soluble calcium, and total tuber calcium, cell wall tuber calcium and
incidence of internal defects in tubers.
187
Plants were grown in the Greenhouses located on the Madison campus of the University of
Wisconsin. Calcium treatments used CaCl2 as the source of calcium. Commercial soil mixes
were not used due to the added gypsum and/or lime they contain. We produced our own growing
mix that contained vermicultite:perlite:peat in a 1:1:1 volume ratio in order to have a better
control of the amount of nutrients applied. The mix was placed in 4L pots up to ¾ of the volume,
and washed several times with tap water, after washing the growing mix pH was 5. Plants were
fertilized once a day with 500mL of a 300 ppm solution of Peat-Lite Special (Peter’s
Professional) made with tap water that contains 70ppm of CaCO3, plus the necessary amount of a
stock solution of 1M or 2.5 M CaCl2 to get the final concentration of the treatment. Treatments
of 1mM and 10mM; or 1mM, 10mM, 15mM and 20mM were used in the experiments. The 1mM
treatment was considered the sufficient calcium treatment and the treatments with 10mM or
more were considered high calcium treatments. The calcium coming from the CaCO3 in the tap
water was not counted as part of the treatment. The transgenic clones were evaluated using a
randomized block design with three replications per clone and different calcium treatments. The
temperature was 20°C/15°C day/night, day length was 16h/8h (light/dark periods), and light
intensity was on average 600μmoles per square meter per second. The light was supplemental
with high intensity discharge lamps when needed.
Total calcium extraction and quantification from leaf tissues of greenhouse grown and in-
vitro plants
A protocol based on the method presented by Kratzke and Palta (1986) was used for extraction
and quantification of tissue calcium. For the tissue culture experiments, whole in-vitro grown
188
plantlets 5-6 week old were either freeze dried or dried in an oven at 60C. For the greenhouse
experiments, the second and third fully extended leaves were sampled and oven-dried
inmediately. Samples were ground to pass a 20-mesh screen and ashed at 550 ºC. The ashes were
dissolved in 5mL of 2N HCl, this solution was filtered using an acid-treated filter paper and
collected in a 50 mL volumetric flask. Ten mL of LaCl3 (2000 mg/L solution) was added to the
flask and the volume raised to 50 mL. The calcium concentration in micrograms per gram of dry
weight (µg/g) was determined using an atomic absorption spectrophotometer Varian SpectrAA
55B (Agilent Technologies, US).
Cell wall and total tuber calcium extraction from tubers
Cell walls were extracted from potato tubers using a protocol adapted from Hoff and Castro
(1969). An average of 20 slices of 1mm-thick was cut from 5 to 16 tubers depending on the size
of the tubers. Slices were rinsed in pre-chilled distilled-deionized water to remove the excess of
starch and dried in a mesh cloth before weighing. The slices were separated in two subsamples,
one for total calcium and the other for cell wall calcium, each with average weights of 10 grams.
The subsample used for total calcium quantification was frozen at -20ºC immediately after
sampling until calcium extraction. The subsample for cell wall calcium extraction was blended
using 100mL of cold distilled-deionized water. The blended tissue was filtered using a propylene
mesh sieve of 100 µm x 100 µm pores and washed with about 900mL of cold distilled-deionized
water to remove starch. The tissue was transferred the to a small beaker with 40mL cold
distilled-deionized water and sonicated for 10 min in sets of 15 seconds followed by 30s of
cooling at amplitude of 20 on an ice bath. The sonicated tissue was filtered again with the
189
propylene mesh sieve, rinsed with 300 mL of water and a final rinse with 30 mL of absolute
ethanol. The cell walls were frozen at -20ºC until calcium extraction.
For calcium extraction, both subsamples, tuber tissue and extracted tuber cell walls were freeze
dried for two weeks. The dried samples were ground manually with a glass rod, ashed and
processed for quantification of calcium using the same procedure as described above for leaf
tissue samples.
Water-extractable and HCl-extractable calcium fractions
The water-extractable and acid-extractable fractions were determined in order to indirectly
measure the amount of apoplastic calcium and calcium oxalate, respectively. For the water-
extractable fraction, 5 mL of distilled-deionized water was added to approximately 0.1 g of dry
tissue in case of leaves and 0.5 g in case of tubers, and let it sit overnight at 4ºC. For the HCl-
extractable fraction, 5 mL of 0.1 N HCl was added to approximately 0.1 g of dry tissue in case of
leaves and 0.5 g in case of tubers, and let it sit overnight at 4ºC. For both fractions, the mix was
filtered using an acid-treated filter paper and collected in a 50 mL volumetric flask. One mL of
H2O2, 5 mL of 2 N HCl and 10 mL of 2000 mg/L LaCl3was added and the final volume of 50
mL was made with distilled-deionized water. These samples were then read in the atomic
absorption spectrophotometer.
Polarized light and environmental scanning electron microscopy
For polarized light microscopy, a thin layer of vascular tissue was peeled off from the abaxial
surface of the leaf using a small forceps and put immediately in a drop of bi-distilled water.
190
Transverse sections of 100-200 microns were obtained from a leaflet of the third fully extended
leaf using a vibratome sectioning system (Oxford). In this procedure, the sample is submerged in
tap water and live sections of the leaf tissue are cut by a vibrating blade. The transversal sections
were immersed in pre-boiled water cooled at room temperature and subjected to vacuum for 5
minutes in order to remove air from the sample. Tap water was used throughout the sectioning to
avoid cell injury.
For the environmental scanning electron microscopy, squares of 5mm x 5mm were sampled from
leaf margins and immediately observed in a FEI Quanta environmental scanning electron
microscope (ESEMTM), the sample temperature was set at 5ºC and vacuum was applied,
accelerated voltage was 20 kV, spot size was 4, and water vapor pressure was 5.08 Torr.
Statistical Analysis
The statistical analysis and plots were obtained using the stats and graphics packages of R
version 3.0.0, respectively (R Development Core Team, 2013).
RESULTS AND DISCUSSION
Calcium deficiency symptoms in the transgenic lines expressing the sCAX1 gene are
alleviated by supplemental calcium
The presence of the transgene was verified by PCR. Most of the lines selected with the MS
media + kanamycin were transgenic; however, a few lines had grown in the media but did not
show a band for the NPT II gene (Figure 5.2). Forty-nine transformed lines were successfully
191
obtained, 48 with the CaMV35S::sCAX1 construct and 1 with the cdc2a::sCAX1 construct. In
addition, the transgene copy number was determined for several lines using qPCR. An initial in-
vitro experiment was performed to test the incidence of calcium deficiency symptoms such as
apical shoot damage, the incidence of stressed leaves and other indicators of plant health such as
plant height and biomass at different calcium treatments compared to the wildtype. There was a
wide range of phenotypic variability among lines that is a consequence of the differences in copy
number, gene expression, somaclonal variation, and the specific genomic position where the
CAX1 gene was inserted by Agrobacterium (Table 5.1). In general, the incidence of apical shoot
damage was reduced at higher media calcium treatments (Table 5.1). Most transgenic lines
showed apical shoot damage in the MS media containing 3mM calcium, the normal
concentration of calcium in the MS media except for few lines presenting multiple copies of the
transgene such as AT1_01_06_02 (6 copies), AT1_02_04_01 (2 or 3 copies), AT1_02_09_01,
AT1_04_03_01 (6 copies), and AT1_06_09_01 (1 copy) that had no apical shoot damage at the
3mM calcium treatment indicating that the multiple copies of sCAX1 may have suffered post-
transcriptional silencing as it was observed in other studies (Tang et al. 2007, Nocarova et al.
2010) (Table 5.1). In addition, at higher calcium treatments, plant height and plant biomass
increased in the transgenics but slightly decreased in the wildtype lines. Also, the total calcium
of the in-vitro cultured plantlets from all transgenic lines and the wildtype increased at higher
calcium treatments. The stress symptoms in the leaves or yellowness were less at higher calcium
treatments except for some of the CAX1 lines of Russet Norkotah. This initial experiment
covered most of our transgenic lines in order to determine general trends. The follow-up
experiments included a reduced set of lines that had a single copy of the sCAX1 gene. Also, the
CAX1 #34 and CAX1 #49 were excluded from further evaluations because of the weakness of
192
the plants and leaf malformations, respectively. Another experiment with a set of clones with a
single copy of the transgene was performed to determine the minimum concentration at which
the transgenic clones alleviate most of their calcium deficiency symptoms without showing salt
stress signs such as yellowish leaves. The single copy transgenics had their sufficient growth at
15mM calcium because at this concentration the apical shoot damage was alleviated in in-vitro
grown plants and was comparable to the wildtype at 3mM (Figure 5.3). These results indicate
that the expression of CAX1 causes the calcium deficiency symptoms observed in in-vitro
cultured plants of potato and that the symptoms are reduced at higher calcium treatments in the
growth media.
A greenhouse experiment was performed in order to study the calcium needs of transgenic plants
to grow normal under greenhouse conditions. In greenhouse grown plants, the calcium
deficiency symptoms observed were apical shoot damage and leaf margin necrosis. In general,
the apical shoot damage was reduced in the high calcium treatment of 10mM, compared to the
1mM calcium treatment. It is important that 1mM was sufficient for normal growth of wildtype
plants (Figure 5.4). The leaf margin necrosis symptom was more evident when the transgenic
lines were grown in the greenhouse because leaves were bigger compared to the in-vitro grown
plants. At higher calcium levels less leaves had this symptom and the symptoms were less severe
in the transgenic lines (Figure 5.5). Nevertheless, the calcium deficiency symptoms caused by
the expression of sCAX1 was not completely or uniformly alleviated under greenhouse
conditions for the whole transgenic plants, especially for leaf margin necrosis. We hypothesized
that the over- expression of the symptoms might be related to the position of the leaf in the plant
and the amount of evapotranspiration in each particular leaf. Comparing the leaf morphology of
normal leaves and leaves showing margin necrosis from the same plant at the 10mM calcium
193
treatment, we observed differences in the morphology of the upper leaf surface (Figure 5.6). The
epidermal cells size was smaller and their shape was more squared. Furthermore, a higher
number of stomata and trichomes were observed on transgenic plants compared to the wildtype
(Figure 5.6). For example, in a given area of upper leaf surface 12 stomata can be observed in
the transgenic leaf with margin necrosis whereas there are no stomata in the similar size area of a
transgenic leaf without deficiency symptoms as well as in the wildtype (Figure 5.6). The higher
stomatal and trichome density are consistent with previous reports of responses under water
stress (Sam et al. 2000). These morphological changes suggest that the sCAX1 lines are under
nutritional stress even though they are growing at sufficient calcium levels.
The transgenic line AT2_01_09_01, the only line that was generated using the cdc2a promoter
had stronger deficiency symptoms than the other transgenic lines at the sufficient and high
calcium treatments. These stronger deficiency symptoms could be caused by a difference in the
levels of expression between the CMV35S promoter and the cdc2a promoter. The comparison of
the relative gene expression between clones revealed that AT2_01_09_01 had almost double the
expression compared to the other lines (Table 5.2).
These results of the in-vitro and greenhouse experiments suggest that the increased transport of
Ca2+ towards the vacuole reduces the availability of calcium in other cell compartments and
therefore calcium deficiency symptoms are observed in the transgenic lines expressing sCAX1
even when they are growing using media or soil calcium concentrations that are sufficient for the
wildtype plants.
194
Calcium content in the sCAX1-expressing lines
As it was indicated before, there was a variable degree of alleviation of calcium deficiency
symptoms in the leaves of plants given the high calcium treatment under greenhouse conditions.
The calcium concentrations in leaf tissue were also variable among the transgenic lines with
some lines showing increased and others showing decreased total calcium concentration. The
results of this evaluation were not presented because of the high variation between samples of
the same clone. The variation observed in calcium from greenhouse leaf samples might be
caused by several sources of variation including leaf age, position in the plant, amount of light
and shade and slight differences in the nutrient solution and soil mix from pot to pot.
Nevertheless, comparing the calcium content of leaves with and without calcium deficiency
symptoms from the same transgenic line at the high calcium treatment, we found that leaves with
margin necrosis had less calcium than leaves without margin necrosis in the greenhouse
evaluation (Table 5.3 and Figure 5.7). These results suggest that when more calcium is
transported to a leaf of the sCAX1 transgenic lines, the calcium deficiency symptoms are
ameliorated.
In order to have a more controlled system and a more homogeneous plant environment, we
performed experiments using in-vitro cultured plants to determine the effect of the expression of
sCAX1on the plantlet calcium concentration. In the in-vitro plantlets, most CAX1 clones showed
similar total leaf calcium content to the wildtype and the amount of total calcium was found to
increase at higher media calcium levels. The concentration of total leaf calcium in dry weight
basis under greenhouse conditions was similar to the wildtype for most transgenic lines in the
3mM calcium treatment. For all transgenic lines, the total concentration of calcium in plants
grown at 15mM calcium was similar or slightly higher as compared to the wildtype. The LSD
195
analysis indicates that the differences are significant between the transgenic lines and the
wildtype at the 15mM calcium treatment (Table 5.4 and Figure 5.8).
In the case of total tuber calcium, there was variability between cultivars and treatments but not
between transgenic lines and the wildtype lines as revealed by the ANOVA analysis (Table 5.5).
The main differences were between the Atlantic and Russet Norkotah lines (Figure 5.9). These
results indicate that if the total amount of calcium in the transgenic lines is similar to the
wildtype but more is being transported into the vacuole in the transgenic plants, we can assume
that there is less in other cellular compartments and especially in the apoplast because the Ca2+
transport into the vacuole modulates apoplastic Ca2+ concentrations (Conn et al. 2011). A close
relationship between apoplastic calcium and cell wall calcium has been documented by previous
research (Demarty et al. 1984, Pechanova et al. 2010, Gilliham et al. 2011a, Wang et al. 2013).
This suggests that in our transgenic lines may be affected by a decrease in apoplastic calcium.
This reduced apoplastic calcium may be compromising the cell walls by decreasing the cell wall-
bound Ca2+ or by a reduction of the cell wall biomass. Further experiments in the following
section will determine which of these mechanisms are affecting the cell walls of the sCAX1-
expressing transgenic lines.
Effects of sCAX1 in the sub-cellular localization of calcium
The expression of sCAX1 in potato causes an increased transport of Ca2+ into the vacuole, but
what happens to that extra calcium? We know from previous reports that calcium cannot be
redistributed to other parts of the cell (Gilliham et al. 2011b). Therefore, calcium might be
forming salts with other compounds stored in the vacuole. One type of calcium salts stored in
vacuoles is calcium oxalate. Calcium oxalate is abundant in certain plants such as Oxalis, beets
196
and spinach. Calcium oxalate forms crystals that vary in bioavailability depending on their
hydration, purity and size (Libert and Franceschi 1987). In our study, we evaluated sCAX1
transgenic lines for the presence of calcium oxalate crystals using polarized light microscopy.
Under polarized light, only tri-dimensional bi-refringent objects such as crystals, that refract light
in two slightly different directions, shine under the microscope. The results of the polarized light
microscopy evaluation indicate that there are calcium oxalate crystals in the leaves of the
transgenic lines at the sufficient and high calcium treatments; meanwhile, the wildtype only
shows few or no crystals at both calcium treatments (Figure 5.10). Calcium oxalate crystals were
observed in the vascular tissue, epidermis and trichomes, and the mesophyll of the leaves of
transgenic plants (Figure 5.10). In the mesophyll, crystals were observed in the spongy cells and
palisade cells (Figure 5.11). Even though we did not test for the presence of calcium oxalates in
tubers, our observations in the leaf cells suggest that the calcium being stored by these transgenic
lines is most likely not going to be bioavailable since it is sequestered in the form of calcium
oxalate. One would expect that the tuber tissue will also contain calcium oxalates. A study of
bioavailability and absorption of sCAX1-expressing carrots showed that there was a reduced
incorporation of calcium into bonds of mice and 10% reduction in fractional (%) absorption from
sCAX1-expressing carrots whereas the total concentration of calcium absorbed in 100g of carrots
was 42% higher compared with an equal quantity of the control indicating that not all of the
calcium sequestered in the vacuole by ectopic expression of sCAX1 is bioavailable (Morris et al.
2008). Therefore, efforts to improve the nutritional quality of potatoes using the expression of
sCAX1 should evaluate calcium bioavailability and absorption to determine to what extent these
calcium oxalates could contribute to human nutrition.
197
Another experiment was performed to test the hypotheses that the increased transport of Ca2+
into the vacuole results in weaker cell walls due to a reduction of the concentration of calcium in
the cell walls or a reduction of the amount of cell walls. Cell walls were extracted from tubers
produced under greenhouse conditions. There was variability in tuber size, with smaller tubers in
the transgenics than the wildtype. A smaller amount of cell walls was observed in the transgenic
lines compared to the wildtype and was confirmed by the ANOVA analysis (Table 5.6 and
Figure 5.12). The ANOVA analysis revealed that the transgenic lines had similar cell wall
calcium concentration as the wildtype (Table 5.7 and Figure 5.13). These results support the
hypothesis that the cell walls biomass decreased as a consequence of the reduced apoplastic Ca2+
concentration in the sCAX1-expressing lines.
The water-extractable fraction of the calcium concentration was also evaluated as an indirect
measurement of apoplastic calcium. The water-extractable fraction in the in-vitro grown plants
was higher in the wildtype than the transgenic lines at the sufficient calcium treatment (3mM)
but at the higher calcium treatment (15mM) the water-extractable fraction becomes more similar
between the wildtype and the transgenic lines (Table 5.8 and Figure 5.14). This result is in
agreement to the predicted changes in apoplastic calcium that is expected to be low in the
transgenic lines at the sufficient calcium treatment (3mM) but increase and become closer to the
wildtype at the high calcium treatment (15mM). Interestingly, significant differences between
the Atlantic and Russet Norkotah lines at =0.1 were observed due to the higher concentrations
for the Russet Norkotah lines, especially at the high calcium treatment. This difference was not
expected because the plants show similar phenotypes.
198
The solubility of calcium oxalate in water is very low which approximately 9 mg/L. is However,
its solubility is much higher in HCl which is around 489.2 mg/L at 25°C (Seidell 1919).
Therefore, the HCl-extractable fraction can be used as an indicator of the amount of calcium
sequestered as calcium oxalate. Some transgenic lines had a higher amount of HCl-extractable
calcium than the wildtype at the sufficient calcium treatment but most of them were higher than
the wildtype at the high calcium treatment (Table 5.9 and Figure 5.15). The lower amount of
water-extractable calcium in the transgenic lines (Figure 5.15) suggests that the apoplastic
calcium is lower in the transgenic lines compared to the wildtype. Furthermore, higher amounts
of HCl-extractable calcium in the transgenic clones (Figure 5.15) indicate an increased
formation of calcium oxalate in the transgenic clones as compared to the wildtype.
Plant heath and tuber quality on the sCAX1 expressing lines
Our observation of the plants show that the over-expression of sCAX1 does not seem to be
beneficial and it is indeed compromising plant health (Figures 5.3 and 5.4). We measured plant
height and biomass as indicators of plant health. The results show that the health of the potato
plant is compromised as shown by lower biomass in transgenic plants (Figures 5.16, 5.17 and
5.18). Under greenhouse conditions, transgenic plants have less biomass as compared to the
wildtype at the sufficient calcium treatment of 1mM. However, their biomass becomes much
similar at higher calcium treatments (Figure 5.16). The ANOVA analysis indicates that the
differences in biomass between the wildtype and the transgenic lines are significant (Table
5.10). Similarly, biomass evaluated under in-vitro conditions was lower in the transgenic lines
compared to the wildtype at the sufficient calcium concentration of 3mM in the media, but their
biomass became more similar at the 15mM calcium treatment (Table 5.11 and 5.17). The data on
199
plant height shows similar trends as biomass. The transgenic lines were shorter than the wildtype
at the sufficient calcium treatment of 3mM, but increased at high calcium treatment and became
more similar to the wildtype (Figure 5.18). The ANOVA analysis indicates that the plant height
differences between the wildtype and the transgenic lines are significant (Table 5.12). These
results show that the over-expression of sCAX1 in potato compromises plant health as shown by
reduced growth compared to the wildtype.
A higher incidence of internal defects, specifically hollow heart, was found in the transgenic
clones as compared to the wildtype in the greenhouse studies (Figure 5.19). The data analysis
indicates that the differences between the incidence of hollow heart in the wildtype and the
transgenic lines are significant (Table 5.13). These results suggest that the increased transport of
calcium into the vacuoles in the sCAX1 transgenic lines has a negative effect on tuber quality.
Even though the total tuber calcium was similar among the transgenic and wildtype lines, the
transgenic lines showed higher tuber defects. These results suggest that an inadequate supply of
Ca2+ to the tuber cell walls compromises tuber cells health resulting in increased internal defects.
Our results also show that the tuber defects were alleviated when soil was given high calcium
treatment (Figure 5.20). These results further confirm the earlier results that tuber internal
defects can be mitigated by a supplemental soil calcium application (Tzeng et al. 1986, Olsen et
al. 1996, Palta 1996, Ozgen et al. 2006, Karlsson et al. 2006).
The relationship between root abundance and calcium uptake under in-vitro culture
conditions
In-vitro plants of Atlantic and its transgenic lines always showed lower calcium than Russet
Norkotah and its transgenic lines at all calcium treatments (Figure 5.8). We hypothesized that
200
this difference might be related to a differential calcium uptake from the media between these
two cultivars. Because calcium is taken up by roots and transported to the plant along with water
mainly in the apoplast, the most logical difference expected between the two potato varieties
would be the amount of roots produced. The roots were collected from 8 in-vitro grown plants in
triplicate, weighed and compared. The results indicate that roots biomass of Atlantic was much
lower than Russet Norkotah (Figure 5.21 and Table 5.14). Data analysis also shows that both
cultivars have significant differences in calcium concentrations (Tables 5.4 and 5.5). These
results indicate a differential response to high calcium levels in the media exists among potato
cultivars and also suggests that an increased amount of roots can contribute to increased uptake
of calcium.
SUMMARY AND CONCLUSIONS
The results presented in this study show that the over-expression of sCAX1 leads to symptoms of
calcium deficiency and compromises tissue health. Our study further illustrates this deficiency
results from transport of calcium to the vacuole where it is sequestered and made unavailable in
the form of calcium oxalate. These results are in agreement with current published reports
(Hirschi 1999, Park et al. 2005b, de Freitas et al. 2011, Wu et al. 2012).
The antiporter H+/Ca2+ CAX1 is known to regulate apoplastic calcium in Arabidopsis as reported
by Conn et al. (2011). In our study, we estimated the apoplastic calcium concentration indirectly
by measuring the water-extractable fraction and assessing the effect of the increased transport of
Ca2+ into the vacuole on cell walls. A reduction in apoplastic calcium would reduce the amount
201
of Ca2+ available to bind to the cell walls. We found that even though the cell wall calcium
concentration did not change significantly, the total amount of cell walls formed was much lower
in the sCAX1 transgenic lines. These results together with the observation that a similar total
tuber calcium concentration was found between the transgenic lines and the wildtype shows that
if the same total amount of total calcium is present in the tissue and more is transported into the
vacuole there is less calcium available in the apoplast and therefore in the cell walls causing
calcium deficiency symptoms and tuber internal defects. Our result also indirectly confirms that
the calcium that goes to the vacuole is trapped in the form of calcium oxalate and cannot be
redistributed to other parts of the cell.
Potato plants grown under very low amounts of calcium such as 30μM are known to show
deficiency symptoms such as apical shoot damage, leaf margin and tip necrosis, and leaf lamina
rolled towards the midrib (Singh and Sharma 1972, Seling et al. 2000, Busse et al. 2004). All
these symptoms were observed in the sCAX1 transgenic lines at sufficient calcium levels (1mM)
and depending on the clone even at higher calcium levels (10mM or more) in the media and/or in
the nutrient solution for plants grown in pots.
Potato tubers are normally low in calcium (Wiersum 1966, Davies and Millard 1985, Kratzke
and Palta 1986) and more than 90% of the calcium in tuber tissue is present in a physiologically
active form (Davies and Millard 1985). This is because calcium moves with water in plants and
very little water moves to the tuber as compared to the leaf tissue (Palta 1996, Busse and Palta
2006). A sub-cellular localization analysis found that the distribution of calcium differs among
tuber cells and most of the calcium is found in the vacuole and very little is present in the cell
walls (Oparka and Davies 1988). When this homeostasis is altered, internal defects appear. In
our study, the transgenic lines showed tuber internal defects in plants grown with sufficient
202
calcium levels (1mM calcium) as a consequence of the increased transport of Ca+2 into the
vacuole in the transgenic lines. Previous studies have shown that increased calcium levels in
potato tubers can improve potato tuber quality by reducing the incidence of internal defects
(Kleinhenz et al. 1999, Karlsson et al. 2006, Ozgen et al. 2006). In our study, we observed that
internal defects of transgenic plants in tubers are mitigated by supplying high calcium (10mM).
In addition, our results suggest that the strategies to respond to supplemental calcium differ
between Atlantic and Russet Norkotah since only Russet Norkotah tends to increase the amount
of roots in presence of additional calcium. The natural variation for total calcium content and
responses to additional calcium levels in the soil and/or media observed among potato clones
may result from the interplay of CAX-like apoplastic calcium regulators and other regulators of
calcium uptake and transport at the root and shoot levels.
Atlantic is a standard chipping cultivar that is susceptible to internal defects such as hollow heart,
internal brown spot, brown center and blackspot bruise. This cultivar also contains less tuber
calcium than other varieties such as Superior (Karlsson et al. 2006). Atlantic is therefore a good
cultivar to test genes that potentially could be used to produce genetically modified potato with
increased tuber calcium concentration without compromising plant health or increasing the
incidence of internal defects. Our results show that even though there was an increased transport
of Ca2+ into the vacuole in the transgenic clones, the total calcium did not increase in the tubers
of Atlantic and Norkotah lines over-expressing sCAX1. This study suggests that CAX1 is a very
important gene that regulates apoplastic calcium and cell wall strength. The natural variation for
this gene should be studied in relation to tuber calcium and tuber quality. Furthermore, the
transgenic lines generated in this study are of great importance for the understanding of calcium
regulation in crops and could be used for example to study genes that restore plant health and
203
tuber quality. This has been done in tobacco and tomato where the co-expression of calreticulin,
a chaperone found in the endoplasmic reticulum, in the sCAX1-expressing lines relieved the
calcium deficiency symptoms caused by sCAX1 (Wu et al. 2012).
BIBLIOGRAPHY
Arora, R. and J.P. Palta. 1988. In Vivo perturbation of membrane-associated calcium by freeze-thaw stress in onion bulb cells simulation of this perturbation in extracellular KCl and alleviation by calcium. Plant Physiology 87: 622-628. Bangerth, F. 1979. Calcium-related physiological disorders of plants. Annual Review of Phytopathology 17: 97-122. Bradshaw, H.D. 2005. Mutations in CAX1 produce phenotypes characteristic of plants tolerant to serpentine soils. New Phytologist 167: 81-88. Busse, J.S. and J.P. Palta. 2006. Investigating the in vivo calcium transport path to developing potato tuber using 45Ca: a new concept in potato tuber calcium nutrition. Physiologia Plantarum 128: 313-323. Cheng, N.H., J.K. Pittman, T. Shigaki, and K.D. Hirschi. 2002. Characterization of CAX4, an Arabidopsis H+/cation antiporter. Plant Physiology 128: 1245-1254. Cheng, N.H., J.K. Pittman, B.J. Barkla, T. Shigaki, and K.D. Hirschi. 2003. The Arabidopsis CAX1 mutant exhibits impaired ion homeostasis, development, and hormonal responses and reveals interplay among vacuolar transporters. The Plant Cell Online 15: 347-364. Cheng N.H., J.K. Pittman, T. Shigaki, J. Lachmansingh, S. LeClere, B. Lahner, D.E. Salt, and K.D. Hirschi. 2005. Functional association of Arabidopsis CAX1 and CAX3 is required for normal growth and ion homeostasis. Plant Physiology 138: 2048-2060. Clarkson, D.T. 1984. Calcium transport between tissues and its distribution in the plant. Plant, Cell & Environment 7: 449-456. Clarkson, D.T., and J.B. Hanson. 1980. The mineral nutrition of higher plants. Annual Review of Plant Physiology 31: 239-298. Cramer, G.R., A. Läuchli, and V.S. Polito. 1985. Displacement of Ca2+ by Na+ from the plasmalemma of root cells. A Primary Response to Salt Stress?. Plant Physiology 79: 207-211.
204
Conn, S.J., M. Gilliham, A. Athman, A.W. Schreiber, U. Baumann, I. Moller, N. Cheng, et al. 2011. Cell-specific vacuolar calcium storage mediated by CAX1 regulates apoplastic calcium concentration, gas exchange, and plant productivity in Arabidopsis. The Plant Cell Online 23: 240-257. Cosgrove, D.J. 2005. Growth of the plant cell wall. Nature Reviews Molecular Cell Biology 6: 850-861. Davies, H.V., and P. Millard. 1985. Fractionation and distribution of calcium in sprouting and non-sprouting potato tubers. Annals of Botany 56: 745-754. Demarty, M., C. Morvan, and M. Thellier. 1984. Calcium and the cell wall. Plant, Cell & Environment 7: 441-448. Doerner, P., J.E. Joergensen, J.E., R. You, J. Steppuhn., and C. Lamb. 1996. Control of root growth and development by cyclin expression. Nature 380: 520-523. Doyle, J.J., and J.L. Doyle. 1990. A rapid total DNA preparation procedure for fresh plant tissue. Focus 12: 13-15. Epstein, E. 1972. Mineral nutrition of plants. Principles and Perspectives. New York: John Wiley and Sons. de Freitas, S.T., M. Padda, Q. Wu, S. Park, and E.J. Mitcham. 2011. Dynamic alternations in cellular and molecular components during blossom-end rot development in tomatoes expressing sCAX1, a constitutively active Ca2+/H+ antiporter from Arabidopsis. Plant Physiology 156: 844-855. Gilliham M., C. Jordans, S. Conn, I. Moller, A. Schreiber, R. Burton, B. Kaiser, S. Tyerman, and R. Leigh. 2011a. Re-examining the role of apoplastic calcium in cell wall modification. 22nd International Conference on Arabidopsis Research. Madison, Wisconsin. Gilliham, M., M. Dayod, B.J. Hocking, B. Xu, S.J. Conn, B.N. Kaiser, S.D. Tyerman. 2011b. Calcium delivery and storage in plant leaves: Exploring the link with water flow. Journal of Experimental Botany 62: 2233-2250. Gilroy, S., P.C. Bethke, and R.L. Jones. 1993. Calcium homeostasis in plants. Journal of Cell Science 106: 453-461. Higuchi, R., C. Fockler, G. Dollinger, and R. Watson. 1993. Kinetic PCR: Real-time monitoring of DNA amplification reactions. Nature Biotechnol 11: 1026-1030. Hirschi, K.D., R.G. Zhen, K.W. Cunningham, P.A. Rea, and G.R. Fink. 1996. CAX1, an H+/Ca2+ antiporter from Arabidopsis. Proceedings of the National Academy of Sciences 93: 8782-8786.
205
Hirschi, K.D. 1999. Expression of Arabidopsis CAX1 in tobacco: altered calcium homeostasis and increased stress sensitivity. The Plant Cell Online 11: 2113-2122. Hirschi, K.D.,V.D. Korenkov, N.L. Wilganowski, and G.J. Wagner. 2000. Expression of Arabidopsis CAX2 in tobacco. Altered metal accumulation and increased manganese tolerance. Plant Physiology 124: 125-134. Hirschi, K.D. 2004. The calcium conundrum. Both versatile nutrient and specific signal. Plant Physiology 136: 2438-2442. Hoff, J.E., and M.D. Castro. 1969. Chemical composition of potato cell wall. Journal of Agricultural and Food Chemistry 17: 1328-1331. Karlsson, B., J.P. Palta, and P.M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. HortScience 41: 1213-1221. Kauss, H. 1987. Some aspects of calcium-dependent regulation in plant metabolism. Annual Review of Plant Physiology 38: 47-72. Kirkby, E.A., and D.J. Pilbeam. 1984. Calcium as a plant nutrient. Plant, Cell & Environment 7: 397-405. Kleinhenz, M.D., J.P. Palta, C.C. Gunter, and K.A. Kelling. 1999. Impact of source and timing of calcium and nitrogen applications on ‘Atlantic’ potato tuber calcium concentrations and internal quality. Journal of the American Society for Horticultural Science 124: 498-506. Kleinhenz, M.D. and J.P. Palta. 2002. Root calcium modulates the response of potato plants to heat stress. Physiologia Plantarum 115: 111-118. Kratzke, M.G. and J.P. Palta. 1986. Calcium accumulation in potato tubers: Role of the basal roots. HortScience 21:1022-1024. Legge, R.L., J.E. Thompson, J.E. Baker, and M. Lieberman. 1982. The effect of calcium on the fluidity and phase properties of microsomal membranes isolated from post-climacteric Golden Delicious apples. Plant and Cell Physiology 23: 161-169. Libert, B. and V.R. Franceschi. 1987. Oxalate in crop plants. Journal of Agricultural and Food Chemistry 35: 926-938. Livak, K.J., S.J. Flood, J. Marmaro, W. Giusti, and K. Deetz. 1995. Oligonucleotides with fluorescent dyes at opposite ends provide a quenched probe system useful for detecting PCR product and nucleic acid hybridization. Genome Research 4: 357-362.
206
Manohar, M.,T. Shigaki, H. Mei, S. Park, J. Marshall, J. Aguilar, and K.D. Hirschi. 2011. Characterization of Arabidopsis Ca2+/H+ exchanger CAX3. Biochemistry 50: 6189-6195. Mann, H.B. and D.R. Whitney. 1947. On a test of whether one of two random variables is stochastically larger than the other. The Annals of Mathematical Statistics 18: 50-60. Marschner, H. 1995. Mineral Nutrition of Plants. Second Edition. Academic Press, Boston. Morris, J., K.M. Hawthorne, T. Hotze, S.A. Abrams, and K.D. Hirschi. 2008. Nutritional impact of elevated calcium transport activity in carrots. Proceedings of the National Academy of Sciences 105: 1431-1435. Murashige, T., and F. Skoog. 1962. A revised medium for rapid growth and bio assays with tobacco tissue cultures. Physiologia Plantarum 15: 473-497. Nicot, N., J.F. Hausman, L. Hoffmann, and D. Evers. 2005. Housekeeping gene selection for real-time RT-PCR normalization in potato during biotic and abiotic stress. Journal of Experimental Botany 56: 2907-2914 Nocarova, E., Z. Opatrny, and L. Fischer. 2010. Successive silencing of tandem reporter genes in potato (Solanum tuberosum) over 5 years of vegetative propagation. Annals of Botany 106: 565-572. Odell, J.T., F. Nagy, and N.H. Chua. 1985. Identification of DNA sequences required for activity of the cauliflower mosaic virus 35S promoter. Nature 313: 810-812. Olsen, N.L., L.K. Hiller, and L.J. Mikitzel. 1996. The dependence of internal brown spot development upon calcium fertility in potato tubers. Potato Research 39:165-178. Oparka, K.J., and H.V. Davies. 1988. Subcellular localisation of calcium in potato tubers. Potato Research 31: 297-304. Ozgen, S., B.H. Karlsson, and J.P. Palta. 2006. Response of potatoes (cv Russet Burbank) to supplemental calcium applications under field conditions: Tuber calcium, yield, and incidence of internal brown spot. American Journal of Potato Research 83: 195-204. Palta, J.P. 1996. Role of calcium in plant responses to stresses: Linking basic research to the solution of practical problems. HortScience 31: 51-57. Park, S., C.K. Kim, L.M. Pike, R.H. Smith, and K.D. Hirschi. 2004. Increased calcium in carrots by expression of an Arabidopsis H+/Ca2+ transporter. Molecular Breeding 14: 275-282. Park, S., T.S. Kang, C.K. Kim, J.S. Han, S. Kim, R.H. Smith, and K.D. Hirschi. 2005a. Genetic manipulation for enhancing calcium content in potato tuber. Journal of Agricultural and Food Chemistry 53: 5598-5603.
207
Park, S., N.H. Cheng, J.K Pittman, K.S. Yoo, J. Park, R.H. Smith, and K.D. Hirschi. 2005b. Increased calcium levels and prolonged shelf life in tomatoes expressing Arabidopsis H+/ Ca2+ transporters. Plant Physiology 139: 1194-1206. Pechanova, O., C.Y. Hsu, J. Adams, T. Pechan, L. Vandervelde, J. Drnevich, S. Jawdy et al. 2010. Apoplast proteome reveals that extracellular matrix contributes to multistress response in poplar. BMC Genomics 11(1), 674. doi:10.1186/1471-2164-11-674. Pittman, J. K., and K.D. Hirschi. 2001. Regulation of CAX1, an Arabidopsis Ca2+/H+ antiporter. Identification of an N-terminal autoinhibitory domain. Plant Physiology 127: 1020-1029. Pittman, J.K., T. Shigaki, N.H. Cheng, and K.D. Hirschi. 2002. Mechanism of N-terminal autoinhibition in the Arabidopsis Ca2+/H+ Antiporter CAX1. Journal of Biological Chemistry 277: 26452-26459. R Development Core Team. 2013. R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org. Robertson, D.N. 2013. Modulating Plant Calcium for Better Nutrition and Stress Tolerance. ISRN Botany. doi: 10.1155/2013/952043 Saidi, Y., A. Finka, M. Muriset, Z. Bromberg, Y.G. Weiss, F.J. Maathuis, and P. Goloubinoff. 2009. The heat shock response in moss plants is regulated by specific calcium-permeable channels in the plasma membrane. The Plant Cell Online 21: 2829-2843. Sam, O., E. Jerez, J. Dell'Amico, and M.C. Ruiz-Sanchez. 2000. Water stress induced changes in anatomy of tomato leaf epidermes. Biologia Plantarum 43: 275-277. Saunders, N.A. 2004. Quantitative real-time PCR. In Real-time PCR: an essential guide. Norfolk, 103-125, United Kingdom: Horizon Bioscience. Seidell, A. 1919. Solubilities of inorganic and organic compounds: a compilation of quantitative solubility data from the periodical literature. Second Edition. New York: D. Van Nostrand Company. Seling, S., A.H. Wissemeier, P. Cambier, and P. Van Cutsem. 2000. Calcium deficiency in potato (Solanum tuberosum ssp. tuberosum) leaves and its effects on the pectic composition of the apoplastic fluid. Physiologia Plantarum 109: 44-50. Singh, O.S. and V.K. Sharma. 1972. Alterations in growth and metabolism of potato plants by calcium deficiency. Plant and Soil 36: 363-369.
208
Tang, W., R.J. Newton, and D.A. Weidner. 2007. Genetic transformation and gene silencing mediated by multiple copies of a transgene in eastern white pine. Journal of Experimental Botany 58: 545-554. Tawfik, A.A., M.D. Kleinhenz, and J.P. Palta. 1996. Application of calcium and nitrogen for mitigating heat stress effects on potatoes. American Journal of Potato Research 73: 261-273. Tuteja, N., and S. Mahajan. 2007. Calcium signaling network in plants: an overview. Plant Signaling and Behavior 2: 79-85. Tzeng, K.C., A. Kelman, K.E. Simmons, and K.A. Kelling. 1986. Relationship of calcium nutrition to internal brown spot of potato tubers and sub-apical necrosis of sprouts. American Journal of Potato Research 63: 87-97. Wang, L.,X. Lv, H. Li, M. Zhang, H. Wang, B. Jin, and T. Chen. 2013. Inhibition of apoplastic calmodulin impairs calcium homeostasis and cell wall modeling during Cedrus deodara pollen tube growth. PloS One 8: e55411. doi:10.1371/journal.pone.0055411 Wiersum, L.K. 1966. Calcium content of fruits and storage tissues in relation to the mode of water supply. Acta Botanica Neerlandica 15: 406-418. Wu, Q., T. Shigaki, J.S. Han, C.K. Kim, K.D. Hirschi, and S. Park. 2012. Ectopic expression of a maize calreticulin mitigates calcium deficiency-like disorders in sCAX1-expressing tobacco and tomato. Plant Molecular Biology 80: 609-619. Yuan, J.S., J. Burris, N.R. Stewart, A. Mentewab, and C.N. Stewart. 2007. Statistical tools for transgene copy number estimation based on real-time PCR. BMC Bioinformatics 8 (Suppl 7): S6.
209
TABLES Table 5.1. Transgenic and wildtype from Atalantic and Russet Norkotah clones evaluated under in-vitro conditions for various attributes including %multiple shoots, % stressed leaves, height, biomass, calcium and copies of the transgene at three calcium treatments, 3mM, 10mM and 30mM.
line 3mM 10mM 30mM 3mM 10mM 30mM 3mM 10mM 30mM 3mM 10mM 30mM 3mM 10mM 30mM copies MS
† Russet Norkotah lines: RN#12 is wildtype and CAX1 #34, CAX1 #36 k-3-1, CAX1 #36 k-3-2 and CAX 1 #49 are transgenic lines expressing sCAX1. High values in red, intermediate in yellow, and low values in green. MS= calcium concentration for best growth, ND= not determined
210 Table 5.2. Analysis of the expression of CAX1 in relation to the house keeping gene EF1-α and a comparison between Atlantic lines.
Atlantic Lines
Threshold cycle difference between CAX1 and EF1-
ΔCt=CtEF1-α-CtCAX1
Difference between clones for threshold cycle variations between CAX1 and EF1-
ΔΔCt
Relative expression
between genes in each line
2ΔCt
Relative expression between
lines 2ΔΔCt
ATL-WT -28.3588 -22.7613 0.000 0.0 AT1_02_01_01 -5.3756 0.2219 0.024 1.2AT1_08_02_01 -5.5975 0 0.021 1.0 AT2_01_09_01 -4.6337 0.9638 0.040 2.0 Table 5.3. Analysis of variance of the calcium concentration in leaves with and without calcium deficiency symptoms evaluated in leaves from the same line. Source DF Mean Sq F-test P (>F) Line 3 4459310 2.0218 0.144967 Symptom 1 21070071 9.5531 0.006017 ** Residual 19 2205568 Lm= Line + Symptom Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) Type indicates a categorical variable with levels transgenic and wildtype Table 5.4. Analysis of variance of the total calcium concentration in plants grown under in-vitro conditions
Treatment x Type 1 3921300 3921300 2.6695 0.106269
Residuals 79 1.16e+08 1468936
Lm = Treatment + Cultivar + Block + Type + Treatment x Type Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.5. Analysis of variance of the total calcium concentration in tubers under greenhouse conditions on a fresh weight basis
Source DF Sum Sq Mean Sq F value Pr(>F)
Cultivar 2 2984.65 1492.32 28.0758 6.00e-09 ***
Treatment 1 1230.61 1230.61 23.1519 1.37e-05 ***
Block 2 98.51 49.25 0.9266 0.4025
Type‡ 1 0.15 0.15 0.0028 0.9584
Residuals 51 2710.83 53.15
Lm = Treatment + Line + Block + Block x Line Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype
211 Table 5.6. Analysis of variance of the percent cell wall extracted from fresh tuber tissue under greenhouse conditions
Source DF Sum Sq Mean Sq F value Pr(>F)
Cultivar 2 2.5517 1.2758 2.2329 0.134597
Block 2 1.2796 0.6398 1.1197 0.346969
Type‡ 1 10.6734 10.6734 18.6796 0.000368 ***
Residuals 19 10.8564 0.5714
Lm =Cell wall% ~ Type + Cultivar +Block Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.7. Analysis of variance of the calcium concentration in the tuber cell wall under greenhouse conditions in fresh weight basis
Source DF Sum Sq Mean Sq F value Pr(>F)
Cultivar 2 10.152 5.076 2.2306 0.1348
Block 2 0.175 0.0873 0.0384 0.9624
Type‡ 1 2.506 2.5062 1.1014 0.3071
Residuals 19 43.236 2.2756
Lm =Cell wall% ~ Type + Cultivar +Block Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.8. Analysis of variance of the percentage of water-extractable fraction of calcium concentration in in-vitro grown plants expressed in dry weight basis as an indirect measurement of apoplastic calcium
Source DF Sum Sq Mean Sq F value Pr(>F)
Treatment 1 0.35662 0.35662 18.2097 5.39 e-05 ***
Cultivar 1 0.06853 0.06853 3.4994 0.06505 .
Block 2 0.07267 0.03633 1.8552 0.16308
Type‡ 1 0.06705 0.06705 3.4237 0.06796 .
Residuals 80 1.56673 0.01958
Lm = HCl soluble calcium ~ Treatment + Type + Cultivar +Block Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype
212 Table 5.9. Analysis of variance of the HCl-extractable fraction of calcium concentration in in-vitro grown plants in dry weight basis as an indirect measurement of calcium oxalate
Lm = HCl soluble calcium ~ Treatment + Type + Cultivar +Block Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.10. Analysis of variance of the plant biomass under greenhouse conditions
Source DF Sum Sq Mean Sq F value Pr(>F)
Treatment 3 531449 177150 4.5923 0.009232 **
Block 3 64491 21497 0.5573 0.647314
Type‡ 1 410243 410243 10.6349 0.002765 **
Treatment x Type 3 378327 126109 3.2692 0.034763 *
Residuals 30 1157253 38575
Lm= Biomass ~ Treatment + Type + Block + Treatment x Type Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.11. Analysis of variance of the plant biomass under in-vitro conditions
Source DF Sum Sq Mean Sq F value Pr(>F)
Treatment 1 5.0031 5.0031 134.0155 < 2.2 e-16 ***
Block 2 0.3915 0.1958 5.2438 0.005501 **
Type‡ 1 2.7917 2.7917 74.7803 < 2.2 e-16 ***
Treatment x Type 1 1.3845 1.3845 37.086 1.91 e -09 ***
Residuals 665 24.8258 0.0373
Lm= Biomass ~ Treatment + Type + Block + Treatment x Type Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype
213 Table 5.12. Analysis of variance of the plant height under in-vitro conditions
Source DF Sum Sq Mean Sq F value Pr(>F)
Treatment 1 683.3 683.29 119.9937 < 2.2e-16 ***
Block 2 78.2 39.1 6.8671 0.001117 **
Type‡ 1 576.7 576.68 101.2704 < 2.2e-16 ***
Treatment x Type 1 115.7 115.71 20.3196 7.74e-06 ***
Residuals 665 3786.8 5.69
Lm = Height ~ Treatment + Type + Block + Treatment x Type Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.13. Analysis of deviance of the incidence of hollow heart in tubers under greenhouse conditions
Source DF Deviance Resid. Dev Pr(>Chi)
null 181.345
Treatment 1 2.085 179.26 0.1487
Type‡ 1 105.777 73.482 <2e-16 ***
Cultivar 2 0.432 73.051 0.8059
Block 2 0.553 72.498 0.7586
Glm= Hollow heart ~ Treatment + Type + Cultivar +Block Pr(>Chi) indicates the p values of the Chi-square test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype Table 5.14. Analysis of variance of root weight of in-vitro grown plants
Source DF Sum Sq Mean Sq F value Pr(>F)
Treatment 1 42.28 42.28 59.3076 3.32e-11 ***
Cultivar 1 58.042 58.042 81.4173 8.77e-14 ***
Block 2 4.276 2.138 2.9993 0.05553 .
Type‡ 1 4.311 4.311 6.0467 0.01612 *
Treatment x Type 1 13.468 13.468 18.892 4.08e-05 ***
Residuals 79 56.319 0.713
Lm= Root weight ~ Treatment + Type + Cultivar + Block + Treatment x Type Pr(>F) indicates the p values of the F test. P-value symbols: p< 0.05 (*), p< 0.01 (**), and p<0.001 (***) ‡Type indicates a categorical variable with levels transgenic and wildtype
214
FIGURES Figure 5.1. PCR efficiencies of the qPCR for all primers used for copy number determination and relative amount of transcripts. The R2 values are indicated in the grey box.
Slope=-1.09 Slope=-1.02
Slope=-1.06 Slope=-0.95
Slope=-0.91 Slope=--0.96
5 6 7 8 9
2425
2627
EF1- for q-PCRUsing gDNA
log2(dna1)
Ct
0.996
5 6 7 8 9
2425
26
CAX1 for qPCRUsing gDNA
log2(dna1)
Ct
0.995
4 5 6 7 8
2021
2223
EF1- for q-PCRUsing cDNA
log2cdna1
Ct
0.976
5 6 7 8 9
2425
2627
CAX1 for q-PCRUsing gDNA
log2(dna1)
Ct
0.996
5 6 7 8 9
2425
26NPT II for qPCR
Using gDNA
log2(dna1)
Ct
0.978
4 5 6 7 8
2526
2728
CAX1 for q-PCRUsing cDNA
log2cdna
Ct
0.991
215 Figure 5.2. Verification of transgenic lines by PCR detection of the nptII gene.
CA
X1
#34
AT
L
SU
P
-+ -
100bp ladder
controls
Putative transgenic lines
++ + ++ +- + ++ ++ +
370 bp
Figure 5.3. Apical shoot damage is alleviated at the 15mM calcium treatment under in-vitro grown conditions
Apical shoot damage
Alleviated shoot
Transgenic3mM
Transgenic15mM
Atlantic ‐WT 3mM
Normal shoot
Figure 5.4. Apical shoot damage is alleviated at the 10mM calcium treatment under greenhouse conditions
Alleviated shoot
Apical shoot damage
Transgenic 1mM
Transgenic 10mM
Atlantic ‐WT1mM
216 Figure 5.5. Leaf margin calcium deficiency symptom alleviated to some degree at 10mM in greenhouse grown transgenic plants.
2 cmAtlantic‐WT Transgenic
1mM 10mM 1mM 10mM
217
Figure 5.6. Morphology of the adaxial epidermal cells in the wildtype, and sCAX1 transgenic lines in leaves with and without margin necrosis
Leaf edge
Leaf edge
Leaf edge
300X Transgenic with margin necrosis
10mM
Transgenic without margin necrosis
10mM
Atlantic‐WT1mM
300X 300X
500X 500X 500X
Pictures taken with a FEI Quanta environmental scanning electron microscope (ESEMTM) under the conditions indicated in the pictures. The green arrows indicate stomata.
218
Figure 5.7. Total calcium concentration (means ± SD) in leaf tissue from symptomatic leaves versus asymptomatic leaves grown in the greenhouse at the high calcium treatment (10mM) under greenhouse conditions. Evaluations performed in single copy lines fom Atlantic and Russet Norkotah lines.
Symptomatic versus Asymptomatic Leavesat the high calcium treatment
a
ab
a
b
a
b
a
Condition
Atlantic lines Russet Norkotah lines
The differences in calcium concentration between symptomatic versus asymptomatic leaves were significant. A protected LSD test was performed between the means of the symptomatic versus asymptomatic, LSD=1269. ANOVA analysis presented on Table 5.3.
219
Figure 5.8. Calcium concentration of in-vitro grown plants (means ± SD) at the sufficient (3mM) and high (15mM) calcium treatments. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
Ca
lciu
m (μ
g/g
)
020
0040
0060
0080
0010
000
1200
0Total calcium concentration in in-vitro grown plants
3mM 15mM 3mM 15mM
Atlantic lines Russet Norkotah lines
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
b
a
bb
c
b b
a
b
aa
a a
a
Calcium treatments
The differences in total calcium concentration in of vitro grown plants between the transgenic and wildtype lines were significant at 15mM. A protected LSD test was performed between the means for transgenic versus wildtype lines, LSD= 804.8. ANOVA analysis presented on Table 5.4.
220
Figure 5.9. Total calcium concentration (means ± SD) in tubers under greenhouse conditions in the sufficient (1mM) and high calcium treatments (10mM), values expressed on fresh weight basis. The potato cultivar Superior was included for comparison. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
Cal
ciu
m (
µg
/g f
resh
we
igh
t)
020
4060
80
Total calcium concentration in tubersat the sufficient and high calcium treatments
1mM 10mM 1mM 10mM 1mM 10mM
Superior
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Atlantic lines Russet Norkotah lines
Calcium treatments
Significant differences for total tuber calcium content were detected between cultivars and treatments but not between the transgenic versus wildtype lines by the ANOVA analysis (Table 5.5).
221
Figure 5.10. Detection of calcium oxalates in the sCAX1 transgenic lines
B
C D
A
E F
Transgenic Transgenic
Transgenic Transgenic
WT WT
A. Vascular tissue of a sCAX1 transgenic line showing calcium oxalates. B. Trichomes showing crystals of calcium oxalate. C, D. Close-up to vascular cells with calcium oxalate crystals under polarized light and brightfield in a transgenic line. E, F. Close-up to vascular cells of the Atlantic wildtype under polarized light and brightfield showing no calcium oxalate crystals.
222
Figure 5.11. Calcium oxalates in the mesophyll of the sCAX1 transgenic lines
B
pc
WT
ue
sc
D Transgenic
pc ue
sc
le
vb
C Transgenic
pc ue
sc
le
vb
Aupper epidermis (ue)
palisade cells (pc)
spongy cells (sc)
lower epidermis (le)
Transgenic
vascular bundle (vb)
A. Transversal section of the leaf of a sCAX1 transgenic line showing the differential distribution of crystals among mesophyll cells. B. Epidermis and mesophyll cells showing no crystals of calcium oxalate in the Atlantic wildtype. C, D. Epidermis and mesophyll cells of a transgenic line showing crystals under brightfield and polarized light.
223
Figure 5.12. Percent cell wall biomass extracted from fresh tuber tissue (means ± SD) at the sufficient (1mM) calcium treatment under greenhouse conditions. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines. The potato cultivar Superior was included for comparison.
The differences in the percentage of dry weight cell walls extracted per gram of fresh tuber tissue between the transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD= 0.69%. ANOVA analysis shown in Table 5.6.
Cell wall/tuber fresh weight (% g/g)
0
1
2
3
4
Percent cell wall biomass extracted from tubersat the sufficient calcium treatment
a
b b
b
a
b b
SuperiorAtlantic lines Russet Norkotah lines
224
Figure 5.13. Calcium concentrations (means ± SD) in tuber cell walls under greenhouse conditions in the sufficient (1mM) calcium treatment in fresh weight basis. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines. The potato cultivar Superior was included for comparison.
Calcium concentration in tuber cell wallsat the sufficient calcium treatment
SuperiorAtlantic lines Russet Norkotah lines
Cal
ciu
m (
µg
/g f
resh
we
igh
t)
24
68
a
ab
a
b
a
b
a
0
The differences in the cell walls calcium concentration in fresh weight basis between the transgenic and wildtype lines were not significant. The LSD test was performed between the means for transgenic versus wildtype lines. LSD= 1.38. ANOVA analysis shown in Table 5.7.
225
Figure 5.14. Water-extractable calcium concentration (means ± SD) in in-vitro grown plants as an indirect measurement of apoplastic calcium. The data on water-extractable fraction is presented as the percentage of total calcium. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
Water extractable calcium concentrationunder in-vitro conditions
Per
cen
tag
e
010
2030
4050
15mM 3mM 15mM3mM
a a
b b
a
b
b
aa a
a
a a
a
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
Atlantic lines Russet Norkotah lines
Calcium treatments
The differences in the percentage of water-extractable calcium concentration between the transgenic and wildtype lines were significant at =0.1. An LSD test was performed between the means for transgenic versus wildtype lines. LSD= 6.6%. ANOVA analysis shown in Table 5.8.
226
Figure 5.15. HCl-extractable calcium concentration in (means ± SD) in-vitro grown plants as an indirect measurement of calcium oxalates. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
The differences in the HCl-extractable calcium concentration between the transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD= 310.4. ANOVA analysis shown in Table 5.9.
227
Figure 5.16. Plant biomass (means ± SD) at different calcium treatments under greenhouse conditions. Evaluations performed in single copy lines and the wildtype fom Atlantic.
1mM 10mM 15mM 20mM
Bio
ma
ss
(g)
050
010
0015
0020
00
Plant biomass at different calcium treatments
under greenhouse conditions
a
b
b
a
b b
ab
a
b
a a
a
ATL-WTAT1_02_01_01AT1_08_02_01
Lines
Calcium treatments
The differences in the plant biomass under greenhouse conditions between the transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD= 251.9. ANOVA analysis shown in Table 5.10.
228
Figure 5.17. Plant biomass (means ± SD) at different calcium treatments under in-vitro conditions. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
3mM 15mM
Bio
ma
ss
(g
)
0.0
0.5
1.0
1.5
Plant biomass at different calcium treatments
under in-vitro conditions
3mM 15mM
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
a
b bb
ab
a a
a
b b
a a a
Atlantic lines Russet Norkotah lines
Calcium treatments
The differences in the plant biomass under in-vitro conditions between the transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD= 0.046. ANOVA analysis shown in Table 5.11.
229
Figure 5.18. Plant height (means ± SD) at different calcium treatments under in-vitro conditions. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
Hei
gh
t (c
m)
05
1015
20Plant height at different calcium treatments
under in-vitro conditions
3mM 15mM 3mM 15mM
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
a
b
cd
a a ab
a
b b
ab
b
Atlantic lines Russet Norkotah lines
Calcium treatments
The differences in the plant height under in-vitro conditions between the transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD= 0.57. ANOVA analysis shown in Table 5.12.
230
Figure 5.19. Internal defects in transgenic and wildtype lines of Atlantic and Russet Norkotah
Russet Norkotah
transgenic
Russet Norkotah
WT
Atlantic WT
Atlantictransgenic
hollow heart
hollow heart
brown center
231
Figure 5.20. Incidence of hollow heart (means ± SD) at the sufficient (1mM) and high calcium treatment (10mM). The potato cultivar Superior was included for comparison. Evaluations performed in single copy lines and the wildtype fom Atlantic and Russet Norkotah lines.
Pe
rcen
tin
cid
enc
e
020
4060
8010
0
Incidence of internal defects: hollow heartat the sufficient and high calcium treatment
1mM 10mM 1mM 10mM 1mM 10mM
ATL-WT
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
b
aa
a
b
aba
ab
b
aa
b
a
ab
SuperiorAtlantic lines Russet Norkotah lines
Calcium treatments
The differences in hollow heart incidence at sufficient and high calcium treatments between the transgenic and wildtype lines as demonstrated by the ANODE (Table 5.13). The Kruskal-Wallis test performed to detect differences between the transgenic versus the wildtype lines were significant at =0.05 for the Atlantic lines and at =0.1 for the Russet Norkotah lines. The non-parametric Mann-Whitney test (Mann and Whitney 1947) performed for the incidence of hollow heart at the sufficient and high calcium treatments in a pair-wise manner also demonstrated that defects were significantly reduced only for AT2_01_09_01 with a p-value=0.076.
232
Figure 5.21. Root weight of 8 plants (means ± SD) at the sufficient and high calcium treatments under in-vitro conditions. Evaluations performed in single copy lines fom Atlantic and Russet Norkotah lines, and the wildtype fom Russet Norkotah.
Ro
ot
we
igh
t (g
)
02
46
810
Root weight of 8 plants
under in-vitro conditions
3mM 15mM 3mM 15mM
ATL
RN #12-WT
AT1_02_01_01AT1_08_02_01AT2_01_09_01
CAX1 #36 K-3-1CAX1 #36 K-3-2
Lines
a
b bb
a aa a
a
b b
b
a a
Atlantic lines Russet Norkotah lines
Calcium treatments
The differences in the amount of roots between transgenic and wildtype lines were significant. A protected LSD test was performed between the means for transgenic versus wildtype lines. LSD=0.56. ANOVA analysis presented on Table 5.14.
233
CHAPTER 6
General Discussion and Conclusions
DISCUSSION
The present study utilized two populations generated by the reciprocal crosses of Atlantic and
Superior. Our evaluation demonstrated that these populations were well suited for the study of
tuber calcium and its relationship with tuber quality. In addition to tuber calcium the parents used
contrasted for tuber yield; specific gravity; enzymatic browning; visual ratings of chip color, chip
color in agtron units, colorimetric measurements of chip color, incidence of hollow heart and
blackspot bruise; as well as incidence and severity of pitted scab. Our results show that these
reciprocal populations segregated for most of these traits except for blackspot bruise that had low
overall incidences in both reciprocal populations. The evaluation of phenotypic variation for
tuber calcium, tuber quality traits, and pitted scab observed in the reciprocal populations of
Atlantic and Superior had an important genetic component due to significant genotype and
genotype x environment (GxE) effects. These traits also showed in most cases intermediate to
high broad-sense heritabilities.
The populations studied in this research are unique in the sense that these are the first tetraploid
segregating populations that have been studied to understand the genetics of tuber calcium in
relation to several commercially important traits. Previous studies have investigated the
234
phenotypic variation for tuber calcium in commercial cultivars (Karlsson et al. 2006; Brown et
al. 2012), and wild potato germplasm (Bamberg et al. 1993), but not in segregating bi-parental
populations. The progenies of these populations had large enough tubers that permitted precise
estimation of tuber internal quality. This is particularly important because larger tubers are
known to be more susceptible to internal defects such as hollow heart compared to small tubers
(Jansky and Thompson 1990).
Another important aspect of our reciprocal populations of Atlantic x Superior is that the parents
are two important commercially accepted chipping cultivars and this is one of the reasons why
SolCAP selected this population for genotyping. Previous tetraploid populations of potato
evaluating quantitative trait loci for commercial traits have used crosses between a released
cultivar and an advanced breeding line. For instance, the processing advanced line 12601ab1 was
crossed to the fresh market cultivar Stirling to study tuber yield, agronomic quality, and a
common scab index (Bradshaw et al. 2008). A tetraploid population generated by the cross of the
fresh market cultivar Jacqueline Lee (susceptible to scab) and the advanced chip processing line
MSG227-2 (tolerant to scab) was previously used to study pitted scab severity (Driscoll et al.
2009). In addition, a tetraploid population generated by the cross of the chipping cultivar Atlantic
and the advanced chipping line B1829-5 population was used to evaluate internal heat necrosis,
yield, dry matter, specific gravity, maturity, texture and flower color (McCord et al. 2011a,
2011b).
The distribution of the genotypic variation was evaluated to depict segregation. A bell-shaped
distribution was observed for tuber yield, specific gravity, enzymatic browning, visual rating of
235
chip color, chip color in agtron units, chip lightness, chip redness, chip yellowness, and tuber
calcium. However, the distributions were skewed towards resistance for the incidence of hollow
heart, internal brown spot, blackspot bruise, as well as pitted scab incidence and severity.
Phenotypic variation for quantitative traits results from the segregation of alleles at multiple
quantitative trait loci (QTL) with effects that are sensitive to the genetic, sexual, and external
environments (Mackay 2001). Bell-shaped distributions have been related to the segregation of
several loci; however, these can also be obtained if large environmental error is affecting a single
genotype (Falconer and Mackay 1996). Therefore, the phenotypic distributions by itself do not
give information about the quantitative nature of a trait. The phenotypic distributions together
with the heritability estimations, significant genotype effects and QTL analysis performed in this
study suggest that all the traits evaluated, including tuber calcium concentration, are quantitative.
The comparison between the reciprocal populations of Atlantic x Superior revealed that the
progenies had tuber calcium more similar to the maternal parent in all years of evaluation.
Previous studies trying to understand maternal inheritance in potato have been performed in
reciprocal populations using parents from different cultivar groups. These reports studied
cytoplasmic male sterility (De la Puente and Peloquin 1968); tuber initiation, tuber set, vine
senescence, tuber yield, flowering, and male fertility generated by the difference in photoperiod
(Sanford and Hanneman 1979). Further evaluation of yield by Sanford and Hanneman (1982) in
reciprocal populations indicated that the higher-yielding reciprocal always had the higher-
yielding parent as the maternal parent. These authors reported that opposite maturities between
the parents was the proposed explanation for the differences between reciprocal populations for
yield. In our study the two parents were not very different in maturity suggesting that this may
236
not be the cause of the difference between reciprocal populations we studied. Differences in chip
color performance between reciprocal populations were observed in diploid populations (Lauer
and Shaw 1970, Jakuczun and Zimnoch-Guzowska 2004) but not in tetraploid populations
(Coffin et al. 1988, Ehlenfeldt et al. 1990, and Pereira et al. 1993). Nevertheless, we observed
significant differences in performance for enzymatic browning, the visual scale of chip color,
chip color in agtron units and chip lightness but only for one year of evaluation. In our study the
population size was variable between the reciprocal populations and that factor is most likely
influencing the differences between the means and their significance. The differences between
the reciprocal populations of Atlantic x Superior for tuber calcium were significant for the three
years of evaluation suggesting that tuber calcium may be influenced by maternal effects.
A tetraploid map was successfully constructed using 600 simplex SNP markers from the SolCAP
8300 Infinium Chip using the Atlantic x Superior population. Previous mapping studies at the
tetraploid level have been performed using AFLP and SSR markers and therefore were targeting
mostly neutral regions, however the SNP chip from SolCAP was designed to target expressed
regions (Hamilton et al, 2011). In our study QTL were identified for tuber calcium, tuber quality
traits and pitted scab tolerance using an interval mapping approach. The interval mapping
approach in tetraploid populations is described by Luo et al (2001). This interval mapping
approach allowed us to detect 75 QTL in total for all traits studied. However, the ±1 LOD
confidence intervals were large for most QTL. The interval mapping approach is not as powerful
compared to other methodologies such as composite interval mapping that uses covariates to
remove the effects of other QTL allowing a more precise localization of the putative QTL (Zeng
1993, Jansen 1993). Future development of new software tools that can perform composite
237
interval mapping or multiple interval mapping in tetraploid populations and the use of all types
of segregation will allow breeders to take full advantage of the 8300 Infinium SNP Chip and to
map QTL more precisely.
Our study is the first attempt to identify the genomic regions that control tuber calcium. Calcium
is an important trait to study in potato tuber because it has important structural properties that
stabilize cell walls and membranes (Demarty et al. 1984, Hirschi 2004). In our study, eight QTL
were detected for tuber calcium concentrations. This number of QTL is in agreement with
calcium being a quantitative trait. The quantitative nature of calcium concentration in the tuber is
somewhat expected because there is a large number of calcium transporters and calcium binding
regulatory proteins that work in complex networks within cells (Boudsocq and Sheen 2010). Half
of the QTL for tuber calcium were additive. The other half were dominant in a simplex or duplex
dosage. On average, each of these eight QTL explained approximately 10% of the variance.
These results are consistent with a study in soybeans where four QTL explaining around 10% of
the variance each were identified for seed calcium in a F2 population generated by a cross of a
low seed calcium cultivar with a high seed calcium cultivar (Zhang et al. 2009). In another study
in common beans two QTL for calcium content were detected and the sum of the variance
explained by both QTL was 25% (Guzmán-Maldonado et al. 2003), indicating similar QTL
effects compared to the QTL for tuber calcium concentration detected in our study.
Hollow heart studied in the Atlantic x Superior population had an intermediate to moderately
high broad-sense heritability. Seven QTL were found for hollow heart. Four of these QTL were
additive, two were dominant in duplex dosage and one was dominant in simplex dosage. The
238
most stable QTL for hollow heart were located on Chr.3 at 0 and 40 cM respectively, and
explained between 7.5 to 19.7% of the total variance. The several QTL identified for hollow
heart suggests that this trait is controlled by several genes. Bradshaw et al. (2008) evaluated the
internal condition (IC), a simultaneous evaluation of incipient hollow heart, hollow heart,
internal necrosis and flecking in a single visual score in a 1 to 9 scale, in the 12601ab1 x Stirling
population. These authors detected only one QTL for IC in Chr. 5 with a single copy of an allele
for increased defects. It seems that these authors may have had less power to detect QTL because
they combined several defects in a new single score. Interestingly, significantly negative genetic
correlations between hollow heart and tuber calcium were identified in our study. These
correlations are in agreement with previous studies reporting that tuber calcium content can be
increased by seasonal calcium application and this increase was related to reduced internal
defects such as blackspot bruise (Karlsson et al. 2006) and internal brown spot (Ozgen et al.
2006). Calcium has important structural properties that maintain cell strength; thus clones with
an adequate amount of tuber calcium may be able to withstand cell damage by physiological
stress, which is considered a cause of hollow heart (Levitt 1942). Therefore, the negative
correlation observed between tuber calcium and hollow heart can be explained by a direct
protective effect of calcium on the tubers to tolerate physiological stress. This negative
correlation may also be partially explained by linkage between loci for tuber calcium and hollow
heart since QTL for tuber calcium concentration and hollow heart incidence were located less
than 20 cM apart on Chr. 3 and 9.
Another interesting finding in our study is related to the evaluation of pitted scab. When
incidence and severity of pitted scab are studied under high disease pressure conditions, the
239
heritabilities were higher suggesting that the differences between genotypes for tolerance to this
pathogen are better observed in a field that was used to grow potatoes for several years without
rotation and where there is enough inoculum to reveal genotypic differences. This finding is
supported by other published reports indicating that seed-borne inoculum contributes
significantly to the disease level (Cairns et al. 1936, Wilson et al. 1999).
Several studies have concluded that common scab resistance in haploid or diploid potato is
controlled by one or few genes (Alam 1972, Krantz and Eide 1941, Murphy et al. 1995) but it
does not seem to be the case in tetraploid potatoes (Dees and Wanner 2012). A segregating
tetraploid population showed continuous variation in common scab resistance, indicating
complex genetics (Driscoll et al. 2009). Four QTL for pitted scab incidence under standard
disease conditions and seven QTL each for incidence and severity under high disease pressure
were detected in our study in agreement with a quantitative nature for common scab tolerance as
it was suggested by previous research (Dionne and Lawrence 1961, Cipar and Lawrence 1972).
Bradshaw et al. (2008) found two QTL for a scab index from 1 (susceptible) to 9 (resistant) on
Chr. 2 and 6 at positions 80 and 86 cM in the Stirling x 12601ab1 tetraploid population in a
standard field. Recently, Braun (2013) identified a QTL located at 10.1 cM on Chr.11 using two
different types of common scab rating, lesion type (LT) and percent surface area (PSA), in a
diploid population generated by a cross between the susceptible S. tuberosum clone US-W4 and
a scab resistant S. chacoense clone 524-8. Similarly, we found that most QTL identified for the
two measurements of pitted scab resistance in our tetraploid populations, pitted scab incidence
and severity, were located in the same positions suggesting that both traits may be controlled by
the same genes. The number of QTL identified by our study was larger than for previous reports.
240
This may, in part, be due to the evaluation method used in our study. We focused on the
evaluation of pitted lesions which are more prominent and easier to detect and therefore the
heritability of this trait was high. In addition, our evaluations were performed in a high disease
field that had enough inoculum that was effective in revealing genotypic differences.
We also found significantly negative correlations between tuber calcium and pitted scab
incidence and severity under high disease pressure. Previous studies had shown contradictory
data, Horsfall et al. (1954) and Davis et al. (1974) reported a positive correlation between
calcium in the tuber peel and common scab severity. Lambert and Manzer (1991) concluded that
high calcium in the periderm was a consequence rather than a cause of increased scab. Our study
relates incidence and severity of pitted scab lesions with calcium concentration in the tuber
medullary tissue and therefore is not influenced by contact with the pathogen. Several studies
have demonstrated that increased calcium concentrations in the potato tuber and tomato stems
reduced the effects of pathogen infections (McGuire and Kelman 1986, Yamazaki and Hoshina
1995, Jiang et al. 2013). Also Flego et al. (1997) demonstrated that an increase in extracellular
calcium concentration in the plant repressed the expression of a pectic enzyme-encoding gene by
the pathogen. The results of our study are in agreement with these published reports and indicate
a protective action of tuber calcium to prevent pathogen attack.
Nearly half of the QTL identified in this study were additive. For instance, we detected three
additive QTL for tuber yield in Chr. 1 that explained approximately 32.9% of the variance. This
results contrast with previous reports that have found that the non-additive genetic effects are the
main component of variance for yield in diploid potato (Mendiburu and Peloquin 1971, Killick
241
1977). In addition, additive genetic variance was minimal or zero and therefore narrow-sense
heritability was low for average external lesion diameter and internal lesion depth caused by
Fusarium (Burkhart et al. 2007). These studies suggest that non-additive genetic variance is a
main component of the genetic variance for some potato traits at the diploid level. However, Hill
et al. (2008) used evidence from empirical studies of genetic variance components across a range
of traits and species to imply that most genetic variance is additive and typically accounts for
over half, and often close to 100% of the total genetic variance. These authors also presented
theoretical results, based upon the distribution of allele frequencies under neutral and other
population genetic models that show why there is mainly additive genetic variance at the level of
gene action even if there are non-additive effects. For example, at a completely dominant locus
almost all the variance contributed is additive if the recessive gene is at high frequency (Falconer
and Mackay 1996). Hill et al. (2008) postulated two primary explanations to explain that most
genetic variance appears to be additive genetic, first that there is indeed little real dominant or
epistatic gene action, or second that it is mainly because allele frequencies are distributed
towards extreme values, as for example in the neutral mutation model. In addition, bottlenecks
usually change the proportion of variance that is additive due to the dispersal of gene frequencies
and the reduction in mean heterozygosity (Cheverud and Routman 1996). These evidence
support that genetic variance is mainly additive at the level of gene action. In addition, additive
and dominance components evaluated in crosses between clones of group Phureja and haploid
clones of Tuberosum indicates that in some of their trials dominance components were higher
than additive but in other trials the additive and dominance components were similar (Rowe
1969). In support of our findings, previous studies that identified QTL for tuber yield also found
half of the QTL with additive effects. Bradshaw et al. (2008) identified two QTL for tuber yield,
242
one dominant in simplex dosage that explained 13.3% of the variance on Chr. 6 and another one
additive on Chr.1. that explained 5.3% of the variance. Interestingly, the three additive QTL
identified in our study is also located on Chr.1. and explained between 9.1 and 13.1% of the
variance. The QTL effects reported by Bradshaw were based on the analysis of 227 progenies
whereas those in our study were based on 128 progenies. This smaller population size in our
study may be causing an over-estimation of the effects of QTL and the real effects may be lower
than predicted. Another factor to consider is that two of those additive QTL in Chr.1 come from
the same parent Superior thus are linked. The presence of linked QTL can interfere with the
estimation of QTL effects especially when using interval mapping without covariates (Broman
and Sen 2009), as it is the case in TetraploidMap. It appears from the discussion above that the
three QTL identified in Chr. 1 in our study may be additive, however, the estimated effects may
not be the “real” effects.
Different measurements of chip color were found to be highly correlated and most QTL for these
traits were located in close proximity or the same location in the genome. The joint analysis of
multiple phenotypes can increase the power for QTL detection and the precision of QTL
localization and can allow one to test for pleiotropy, a single QTL affecting multiple phenotypes
(Broman and Sen 2009). In our study multiple measurements for the chip color detected QTL for
this trait in the same region indicating that the different methods employed were in agreement.
The results of our study suggest that the over-expression of CAX1 is transporting large amounts
of Ca+2 into the vacuole making it less available in other parts of the cell. This detrimental effect
of an increased transport of Ca+2 into the vacuole has shown to increase the incidence of blossom
243
end defect in tomato fruits (de Freitas et al. 2011) and caused leaves necrosis in tobacco (Hirschi
1999). In our study, the effects of the increased transport of calcium into the vacuole were
associated with apical shoot damage, leaf margin necrosis and internal tuber defects. These
symptoms are a consequence of the induced calcium deficiency in the plant but they can be
alleviated by either increasing the apoplastic calcium or the symplastic movement of calcium.
Apoplastic calcium can be increased by growing the plants under high calcium treatments. The
amelioration of tuber internal defects in the CAX1 transgenic lines resembles that one observed
when supplemental calcium is applied to potatoes in-season (Karlsson et al. 2006, Ozgen et al.
2006). However, the calcium deficiency is very high in the transgenic lines so that the alleviation
is not complete. Symplastic calcium transport has been reported to increase by over-expressing
Calreticullin (CRT) (Wu et al. 2012). CRT is a calcium binding protein that can mobilize Ca+2
through the endoplasmic reticulum and the cytoplasm (Wyatt et al. 2002). The endoplasmic
reticulum is closely associated with plasmodesmata and the desmotubule provides a potential
pathway for movement between cells (Roberts and Oparka 2003). Our results suggest that it
would be interesting to select for potato cultivars that have a high efficiency of apoplastic and
symplatic calcium transport and therefore an adequate supply of calcium where and when it is
needed. These two genes CAX1 and CRT deserve to be studied in more detail in relation to tuber
calcium and tuber quality.
Another question generated by our study is if total tuber calcium is the best predictor of internal
quality. The observed effects of the over-expression of the CAX1 gene in potato demonstrate that
not only the total amount of calcium but its sub-cellular distribution is very important to maintain
cellular health and therefore internal tuber quality. These results suggest that it might be
244
important to determine which fractions of tuber calcium (cell wall calcium, apoplastic calcium or
vacuolar calcium) besides the total tuber calcium is the best predictor of adequate internal
quality. The results of our study suggest that an adequate amount of calcium distributed where it
is needed in the tuber cells is necessary to withstand physiological and biotic stress.
CONCLUSIONS
The goal of this thesis was to generate new knowledge about the genetics of tuber calcium and
its relationship with tuber quality. The results of the research presented in this thesis lead us to
come up with several conclusions listed by chapter in the following lines.
Characteristics of the reciprocal populations of Atlantic x Superior (Chapter 2)
1. Atlantic and Superior have contrasting phenotypes for tuber yield, specific gravity,
enzymatic browning, chip color using visual ratings, chip color in agtron units,
colorimetric measurements of chip color, tuber calcium, incidence of hollow heart and
blackspot bruise, as well as incidence and severity of pitted scab. The reciprocal
populations of Atlantic and Superior are segregating for all traits these cultivars differ.
2. The performance of these populations differed significantly for several traits but only in
one year of evaluation, except for tuber calcium that was significantly different between
the reciprocal populations for the three years of evaluation.
245
3. Tuber yield, specific gravity, chip quality traits and tuber calcium had bell-shaped
distributions that resemble normal distributions. However, the incidences of hollow heart
and blackspot bruise as well as pitted scab incidence and severity had skewed
distributions that suggest the presence of major dominant genes for these traits. Bell-
shaped distributions have been related to the segregation of several loci; however, these
can also be obtained if large environmental error is affecting a single genotype.
Correlations and broad-sense heritabilities between traits in the Atlantic x Superior
reciprocal populations (Chapter 3)
1. The phenotypic variation for tuber calcium, tuber quality traits, and pitted scab observed
in the reciprocal populations of Atlantic and Superior has an important genetic
component due to the significant genotypic effects for all traits. This genotypic variation
can be exploited to select for cultivars with improved yield and specific gravity,
improved chip quality and internal quality, and tolerant to pitted scab.
2. All traits studied including tuber quality traits, pitted scab tolerance and tuber calcium are
influenced by environmental effects, including the year of evaluation as well as
significant genotype x environment (year) in at least one the reciprocal populations.
3. The performances of the reciprocal populations showed significant correlations between
two out of three years of evaluation for most traits. Hollow heart was the only trait that
had significant correlations between all years of evaluation, whereas all years were
uncorrelated for blackspot bruise.
246
4. A significantly negative correlation for hollow heart with tuber calcium was found. This
correlation is in agreement with previous reports that relate high tuber calcium and
decreased incidence of defects.
5. Most measurements of chip color were significantly correlated. We recommend chip
lightness and the visual rating of chip color for selection because they are correlated to all
other measurements of chip color.
6. High correlations were observed between pitted scab incidence and severity under high
disease pressure indicating that pitted scab incidence could be scored for quick
assessments.
7. The results of the relationship analyses indicated that higher tuber calcium decreased the
probability of getting tubers with hollow heart as well as tubers with pitted scab and the
number of pits per tuber under the high disease field conditions.
8. Four promising clones that have good chipping quality, good internal quality, reduced
pitted scab incidence and severity as well as acceptable yield based on three years of
evaluation compared to Atlantic were selected from the reciprocal populations.
Identification of QTL for tuber calcium, tuber quality and pitted scab tolerance in the
Atlantic x Superior tetraploid population (Chapter 4)
10. A tetraploid map was successfully constructed using 600 simplex SNP markers from the
SolCAP 8300 Infinium Chip. The map for Atlantic had more than twice the number of
markers, 414, compared to Superior, 186, probably due to the higher inbred pedigree of
247
Superior.
11. Several SNP loci had more than expected duplex genotypes suggesting that the method
for the estimation of dosage used by SolCAP may not be completely accurate.
12. Several quantitative trait loci were identified for tuber calcium, tuber quality traits and
pitted scab tolerance in the Atlantic x Superior tetraploid population using an interval
mapping approach.
13. The correlation between hollow heart and tuber calcium can be explained at least
partially due to linkage because two QTL for hollow heart were in close proximity to
QTL for tuber calcium concentration indicating that part of the correlation between these
traits may be explained by linkage.
14. For some correlated traits such as several methods of measurement of chip color as well
as measurements of pitted scab incidence and severity, we found several QTL in the same
chromosomes and similar position indicating that these correlated traits might be
detecting the same QTL.
15. The detection and sometimes location of QTL varied from year to year due to the effects
of different population sizes and also probably environmental effects.
16. Markers with significant effects were identified for several traits in the marker regression
analysis and many of them were located ±20 cM of a QTL.
17. Approximately half of the QTL detected were additive with a complex inheritance. The
other half consisted of QTL in a simplex and duplex dosage. These simplex and duplex
QTL have higher potential to become markers for marker assisted selection (MAS).
248
Effect of the over-expression of the CAX1 gene in potato (Chapter 5)
1. A wide range of phenotypic variability was observed among the transgenic lines
probably due to differences in copy number, gene expression and the specific genomic
position where the CAX1 gene was inserted.
2. These results of the in-vitro and greenhouse experiments suggest that the increased
transport of Ca2+ towards the vacuole reduces the availability of calcium in other cell
compartments and therefore calcium deficiency symptoms are observed in the transgenic
lines expressing sCAX1 even when they are growing using media or soil calcium
concentrations that are sufficient for the wildtype plants.
3. The appearance of calcium deficiency symptoms such as apical shoot damage and
marginal necrosis was mitigated at high calcium concentrations in the media or the soil.
4. Squared-shaped and smaller epidermal cells as well as increased number of stomata and
trichomes were observed comparing the leaf morphology leaves showing margin necrosis
and normal leaves from the same transgenic plant at the high (10mM) calcium treatment
in greenhouse grown plants. These characteristics are indicators of nutritional stress even
though they are growing at high calcium levels.
5. The total amount of calcium in in-vitro grown plants and tubers of the transgenic lines is
similar to the wildtype at the sufficient calcium treatment. However, high amount of
calcium are stored in the vacuole. Therefore, we can assume that there is less calcium
available in the apoplast because the Ca2+ transport into the vacuole modulates apoplastic
calcium concentrations.
6. Calcium oxalate crystals were detected in the vascular tissue, epidermis, trichomes as
249
well as the palisade and spongy cells of the mesophyll of leaves from the sCAX1
transgenic plants.
7. The transgenic lines had less cell walls biomass as a consequence of the reduced
apoplastic Ca2+ concentration.
8. The over-expression of sCAX1 in potato compromises plant health as shown by reduced
biomass and height in the transgenic lines compared to the wildtype. Nevertheless, plant
health is ameliorated by high tuber calcium treatments.
9. Higher incidence of internal defects, specifically hollow heart, was found in the
transgenic clones as compared to the wildtype in the greenhouse studies. This internal
defect was mitigated by increased tuber calcium in the soil.
10. Russet Norkotah had significantly higher root biomass than Atlantic at the sufficient and
high calcium treatments. Also, in the presence of high calcium Russet Norkotah increased
whereas Atlantic maintained the root biomass. The transgenic lines of Atlantic increased
their root biomass to similar values compared to the wildtype at the high calcium
treatment. The Russet Norkotah transgenic lines also increased their root biomass but to
significantly higher values compared to the wildtype.
BIBLIOGRAPHY
Alam, Z. (1972). Inheritance of scab resistance in 24-chromosome potatoes. University of Wisconsin--Madison. Bamberg, J.B., J.E Palta, L.A. Peterson, M. Martin, and A.R. Krueger. 1993. Screening tuber-bearing Solanum (potato) germplasm for efficient accumulation of tuber calcium. American Potato Journal 70: 219-226.
250
Boudsocq, M., and J. Sheen. 2010. Stress signaling II: calcium sensing and signaling. In Abiotic Stress Adaptation in Plants, 75-90. Springer Netherlands. Bradshaw, J.E., C.A. Hackett, B. Pande, R. Waugh, and G.J. Bryan. 2008. QTL mapping of yield, agronomic and quality traits in tetraploid potato (Solanum tuberosum subsp. tuberosum). Theoretical and Applied Genetics 116: 193–211. Braun, S.R. 2013. Quantitative trait loci analysis and breeding for resistance to common scab in potato. Ph.D. dissertation Plant Breeding and Plant Genetics. University of Wisconsin. Broman K. W., and S. Sen. 2009. A Guide to QTL Mapping with R/qtl. Springer, New York. Brown, C.R., K.G. Haynes, M. Moore, M.J. Pavek, D.C. Hane, S.L. Love, R.G. Novy, and J.C. Miller Jr. 2012. Stability and Broad-Sense Heritability of Mineral Content in Potato: Calcium and Magnesium. American Journal of Potato Research 89: 255-261. Burkhart, C.R., B.J, Christ, and K.G. Haynes, 2007. Non-additive genetic variance governs resistance to Fusarium dry rot in a diploid hybrid potato population. American journal of potato research 84: 199-204. Cairns, H., T.N. Greeves, and A.E. Muskett, 1936. The Control of Common Scab (Actinomyces scabies (Thaxt.) Güss.) of the Potato by Tuber Disinfection. Annals of Applied Biology 23(4): 718-74. Cipar, M.S., and C.H. Lawrence. 1972. Scab resistance of haploids from two Solanum tuberosum cultivars. American Journal of Potato Research 49: 117–119. Cheverud, J.M., E.J. Routman. 1996. Epistasis as a source of increased additive genetic variance at population bottlenecks. Evolution 50: 1042–1051. Coffin R, MK Keenan, R Yada, G Tai, M Slavnik, and G Johnston. 1988. Inheritance of genetic factors controlling chip processing quality after tuber storage at low temperatures [Abstract]. American Journal of Potato Research 65:474. Davis, J.R., G.M. McMaster, R.H. Callihan, J.G. Garner, and R.E. McDole. 1974. The relationship of irrigation timing and soil treatments to control potato scab. Phytopathology 64: 1404-1410. Dees, M.W., and L.A. Wanner. 2012. In Search of Better Management of Potato Common Scab. Potato Research 55: 249-268. De la Puente, F., and S. J. Peloquin. 1968. Male fertility of selected 24 chromosome S. tuberosum hybrids. American Potato Journal 45: 436-437. Demarty, M., C. Morvan, and M. Thellier. 1984. Calcium and the cell wall. Plant, Cell & Environment 7: 441-448.
251
Dionne, L.A. and C. H. Lawrence. 1961. Early scab resistant derivatives of Solanum chacoense x Solanum phureja. American Journal of Potato Research 38: 6-8 Driscoll, J., J. Coombs, R. Hammerschmidt, W. Kirk, L. Wanner, and D. Douches. 2009. Greenhouse and Field Nursery Evaluation for Potato Common Scab Tolerance in a Tetraploid Population. American Journal of Potato Research 86: 96–101. Ehlenfeldt M.K., D.F. Lopez-Portilla, A.A. Boe, and R.H. Johansen. 1990. Reducing sugar accumulation in progeny families of cold chipping potato clones. American Journal of Potato Research 67: 83-91. Falconer, D.S., and T.F.C. Mackay. 1996. Introduction to Quantitative Genetics, Fourth Edition. Essex, UK: Longman, Harlow. Flego, D., M. Pirhonen, H. Saarilahti, T.K. Palva, and E.T. Palva. 1997. Control of virulence gene expression by plant calcium in the phytopathogen Erwinia carotovora. Molecular Microbiology 25: 831-838. de Freitas, S.T., M. Padda, Q. Wu, S. Park, and E.J. Mitcham. 2011. Dynamic alternations in cellular and molecular components during blossom-end rot development in tomatoes expressing sCAX1, a constitutively active Ca2
+/H+ antiporter from Arabidopsis. Plant Physiology 156: 844-855. Guzmán-Maldonado, S.H., O. Martínez, J.A. Acosta-Gallegos, F. Guevara-Lara, and O. Paredes- López. 2003. Putative quantitative trait loci for physical and chemical components of common bean. Crop Science 43: 1029-1035. Hamilton, J.P., C. N. Hansey, B. R. Whitty, K. Stoffel, A. N. Massa, A. Van Deynze, W.S. De Jong, D.S. Douches, and C.R. Buell. 2011. Single nucleotide polymorphism discovery in elite North American potato germplasm. BMC Genomics 12: 302-312 Hill, W.G., M.E. Goddard and P.M. Visscher. 2008. Data and Theory Point to Mainly Additive Genetic Variance for Complex Traits. PLoS Genetics 4: e1000008. doi:10.1371/journal.pgen.1000008 Hirschi, K.D. 1999. Expression of Arabidopsis CAX1 in tobacco: altered calcium homeostasis and increased stress sensitivity. The Plant Cell Online 11: 2113-2122. Hirschi, K.D. 2004. The calcium conundrum. Both versatile nutrient and specific signal. Plant Physiology 136: 2438-2442. Horsfall, J.G., J.P. Hollis, and H.G.M. Jacobson. 1954. Calcium and potato scab. Phytopathology 44: 19-24. Jansen, R.C. 1993. Interval mapping of multiple quantitative trait loci. Genetics 135: 205-211.
252
Jansky, S.H. and Thompson D.M. 1990. Expression of Hollow Heart in segregating tetraploid potato families. American Potato Journal 67: 695-793. Jiang, J.F., J.G. Li, and Y.H Dong. 2013. Effect of calcium nutrition on resistance of tomato against bacterial wilt induced by Ralstonia solanacearum. European Journal of Plant Pathology 136: 547-555. Karlsson B.H., J.P. Palta, and P.M. Crump. 2006. Enhancing tuber calcium concentration may reduce incidence of blackspot bruise injury in potatoes. Hort Science 41: 1213-1221. Killick, R.J. 1977. Genetic analysis of several traits in potatoes by means of a diallel cross. Annals of Applied Biology 86: 279-289. Krantz, F.A. and C.J. Eide. 1941. Inheritance of reaction to common scab in the potato. Journal of Agricultural Research 63:219-231. Jakuczun, H. and E. Zimnoch-Guzowska. 2004. Inheritance of glucose content in tubers of diploid potato families. American Journal of Potato Research 81: 359-370. Lambert, D.H., and F.E. Manzer. 1991. Relationship of Calcium to Potato Scab. Phytopathology 81: 632-636. Lauer, F. and R. Shaw. 1970. A possible genetic source for chipping potatoes from 40 F storage. American Journal of Potato Research 47: 275-278. Levitt, J. 1942. A historical study of hollow heart potatoes. American Journal of Potato Research 19: 134-143. Luo, Z. W., C. A. Hackett, J. E. Bradshaw, J. W. McNicol, and D. Milbourne, 2001. Construction of a genetic linkage map in tetraploid species using molecular markers. Genetics 157: 1369-1385. Mackay, T. F. 2001. The genetic architecture of quantitative traits. Annual Review of Genetics 35: 303-339. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011a. Linkage Mapping and QTL Analysis of Agronomic Traits in Tetraploid Potato (Solanum tuberosum subsp. tuberosum). Crop Science 51: 771-785. McCord, P.H., B.R. Sosinski, K.G. Haynes, M.E. Clough, and G.C. Yencho. 2011b. QTL mapping of internal heat necrosis in tetraploid potato. Theoretical and Applied Genetics 122: 129–142. McGuire, R. and A. Kelman. 1986. Calcium in potato tuber cell walls in relation to tissue maceration by Erwinia carotovora pv. atroseptica. Physiology and Biochemistry 7: 401-406.
253
Mendiburu, A.O. and S.J. Peloquin 1971. The significance of 2N gametes in potato breeding. Theoretical and Applied Genetics 49: 53-61. Murphy, A.M., H. Jong, G.C.C. Tai. 1995. Transmission of resistance to common scab from the diploid to the tetraploid level via 4x-2x crosses in potatoes. Euphytica 82:227–233 Ozgen S., B.H. Karlsson, and J.P. Palta. 2006. Response of potatoes (cv. ‘Russet Burbank’) to supplemental calcium applications under field conditions. Tuber calcium yield and incidence of brown spot. Journal American Potato Research 83: 195-204. Pereira, A.D.S., R.H. Coffin, R.Y. Yada and V.S. Machado. 1993. Inheritance patterns of reducing sugars in potato tubers after storage at 12C and 4C followed by reconditioning. American Potato Journal 70: 71-76. Roberts, A. G. and K.J. Oparka. 2003. Plasmodesmata and the control of symplastic transport. Plant, Cell and Environment 26: 103-124. Rowe, P.R. 1969. Quantitative variation in diploid potatoes. American Journal of Potato Research 46: 14-17. Sanford, J.C., and R.E. Hanneman,. 1979. Reciprocal differences in the photoperiod reaction of hybrid populations inSolanum tuberosum. American Potato Journal 56: 531-540. Sanford, J.C. and R.E. Hanneman Jr. 1982. Large yield differences between reciprocal families of Solanum tuberosum. Euphytica 31: 1-12. Wilson, C.R., L.M. Ransom and B. M Pemberton. 1999. The Relative Importance of Seed‐borne Inoculum to Common Scab Disease of Potato and the Efficacy of Seed Tuber and Soil Treatments for Disease Control. Journal of Phytopathology 147: 13-18. Wu, Q., T. Shigaki, J.S. Han, C.K. Kim, K.D. Hirschi, and S. Park. 2012. Ectopic expression of a maize calreticulin mitigates calcium deficiency-like disorders in sCAX1-expressing tobacco and tomato. Plant Molecular Biology 80: 609-619. Wyatt, S.E., P.L. Tsou and D. Robertson. 2002. Expression of the high capacity calcium-binding domain of calreticulin increases bioavailable calcium stores in plants. Transgenic Research 11: 1-10. Yamazaki, H., and T. Hoshina. 1995. Calcium nutrition affects resistance of tomato seedlings to bacterial wilt. HortScience 30: 91-93. Zhang, B., P. Chen, A. Shi, A. Hou, T. Ishibashi, and D. Wang. 2009. Putative quantitative trait loci associated with calcium content in soybean seed. Journal of Heredity 100: 263-269.
254
Zeng, Z-B. 1993. Theoretical basis for separation of multiple linked gene effects in mapping of quantitative trait loci. Proceedings of the National Academy of Sciences 90: 10972–10976.