MOLECULAR MARKER DEVELOPMENT, QTL PYRAMIDING, AND COMPARATIVE ANALYSIS OF PHENOTYPIC AND MARKER-ASSISTED SELECTION IN CUCUMBER by Matthew D. Robbins 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 2006
213
Embed
MOLECULAR MARKER DEVELOPMENT, QTL PYRAMIDING, AND … › staublab › Matt › MDRobbins_PhD... · 2006-06-16 · MOLECULAR MARKER DEVELOPMENT, QTL PYRAMIDING, AND COMPARATIVE ANALYSIS
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
MOLECULAR MARKER DEVELOPMENT, QTL PYRAMIDING, AND
COMPARATIVE ANALYSIS OF PHENOTYPIC AND MARKER-ASSISTED
SELECTION IN CUCUMBER
by
Matthew D. Robbins
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
2006
i
Dedication
To my family: Cody, Hayden, Camila, and most especially, Heidi
ii
Acknowledgements There have been many who have helped make this dissertation possible. I am thankful for several undergraduates, Danya Hooker, Caroline Gatheca, Brian Holzman, Katrina Pfaff, Sarah Yokobowski, John Alaniz, Julie Weidner, and Julianna Whan, who directly worked with me on this project. I also want to thank the graduate students, Gennaro Fazio, Sang-Min Chung, Anabel López-Sesé, Zhanyong Sun, Jaun Zalapa, Isabelle Delannay, Vanessa Gordon, Shanna Mason, Miriam Paris, and Hugo Cuevas who worked with me and gave me encouragement. I am thankful for Linda Crubaugh and her patience with me and the other graduate students. Thanks goes to my committee members, Drs. Phil Simon, Mike Casler, Mike Havey, and Jim Coors for their patience and guidance. A special thanks to my advisor, Dr. Jack Staub for his tutorage and mentorship. I am grateful for the support of my parents, Val and Judy Robbins, and my in-laws Mikel and JoLynn Stevens. I am most indebted to my children who have sacrificed time with their dad, and my wife, Heidi, without whom this work would not be possible. Finally, but most importantly, I thank my Heavenly Father for his guidance and loving care.
iii
Table of Contents
Dedication ............................................................................................................................ i Acknowledgements............................................................................................................. ii Table of Contents............................................................................................................... iii List of Tables .................................................................................................................... vii List of Figures .................................................................................................................. viii Abstract ............................................................................................................................... x Introduction......................................................................................................................... 1
Chapter 1. Comparative analysis of marker-assisted and phenotypic selection for yield components in cucumber .................................................................................................. 45
Abstract ......................................................................................................................... 45 Introduction................................................................................................................... 46 Materials and Methods.................................................................................................. 48
Considerations for MAS ........................................................................................... 58 Selection effectiveness.............................................................................................. 59 Selection methods in breeding programs.................................................................. 63
Literature Cited ............................................................................................................. 65 Chapter 2. The development of molecular markers with increased efficacy for genetic analysis in cucumber......................................................................................................... 79
iv
Abstract ......................................................................................................................... 79 Introduction................................................................................................................... 80 Materials and Methods.................................................................................................. 82
Identification of SNPs by SCAR sequencing ......................................................... 85 Identification of SNPs by BAC end sequencing .................................................... 86 SNP marker creation............................................................................................. 87
SNP marker evaluation and verification ................................................................... 90 Results........................................................................................................................... 91
Identification of SNPs by SCAR sequencing ......................................................... 92 Identification of SNPs by BAC end sequencing .................................................... 93 SNP marker creation............................................................................................. 94
SNP marker evaluation ............................................................................................. 94 Summary of RAPD conversion ................................................................................ 95
Discussion..................................................................................................................... 96 RAPD to SCAR conversion...................................................................................... 97 SCAR multiplexing................................................................................................... 99 Identification of SNPs............................................................................................. 100 SNP marker creation and evaluation....................................................................... 102
Literature Cited ........................................................................................................... 105 Chapter 3. Pyramiding QTL for multiple lateral branching in cucumber using nearly isogenic lines................................................................................................................... 121
Abstract ....................................................................................................................... 121 Introduction................................................................................................................. 122 Materials and Methods................................................................................................ 124
NIL creation ............................................................................................................ 124 Molecular marker analysis...................................................................................... 125 Open-field evaluation of NIL for MLB .................................................................. 126 Statistical analysis................................................................................................... 126
Results......................................................................................................................... 127 Discussion................................................................................................................... 129 Literature Cited ........................................................................................................... 134
Conclusions and Future Work ........................................................................................ 143 Literature Cited ........................................................................................................... 147
Appendices...................................................................................................................... 149 Appendix A. Means and linear response at two planting dates (June 23, 2004 and July 7, 2004) of five traits in four base cucumber populations (C0) of cucumber which
v
underwent three cycles of recurrent mass selection (C1-C3) using three breeding methods (see Chapter 1).............................................................................................. 149
Appendix B. The linear response of selection for five traits in four cucumber populations by marker (MAS), phenotype (PHE), and random mating (no selection; RAN) over three cycles. The five traits are earliness (measured as the number of fruits per plant in first harvest), gynoecy (measured as the percent female flowers in the first ten nodes), fruit length to diameter ratio (measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests), multiple lateral branching (measured as the number of lateral branches of at least three internodes long on the mainstem in the first 10 nodes), and yield (measured as the number of fruits per plant averaged over four harvests; see Chapter 1). ..................................................... 152
Appendix C. Sequences of RAPD bands used to make SCAR primers. Sequences are presented in FASTA format with the name of the RAPD marker, and the parental line (Gy-7 or H-19) from which the band was produced in parentheses. Incomplete sequences are indicated by “FORWARD ONLY” or “REVERSE ONLY”. Sequences that correspond to a band other than the polymorphic RAPD band (as determined by the segregation pattern of the SCAR created from the sequence and the original RAPD marker) are indicated by “DOES NOT MATCH RAPD” (Table 2.1; see Chapter 2)...................................................................................................................................... 157
Appendix D. SCAR marker database. An html (web-based) database of the SCAR markers created in Chapter 2, as well as additional SCAR markers, is available that contains primer sequences, annealing temperature gradient PCR (ATG-PCR) profiles, and spreadsheets with information on each of the SCAR markers. The structure and contents of this database are explained in the following screenshots and can be accessed at: http://www.vcru.wisc.edu/staublab/Matt/SCAR%20web%20page2/Scar%20database.htm ................................................................................................................................ 169
Appendix E. BAC clone end sequences used to create SCAR markers in Table 2.2 (Chapter 2). Sequences are presented in FASTA format with the sequence name followed by the number of bases in parentheses. Sequences are named with a one or two letter designation for the marker that hybridized to the BAC clone (AJ = AJ6SCAR, B = BC523SCAR, C = CSWCTT14, L = L19SCAR, M = M8SCAR, and W = OP-W7-1), the number of the clone that was sequenced, BE for “BAC end”, and L or R to signify the left or right end of the BAC clone, respectively (L2 or R2 indicates the second attempt to obtain the sequence). AJ-1-BE-L, for example, is the sequence of the left end of the first clone sequenced that hybridized to AJ6SCAR. . 171
Appendix F. SCAR primers from BAC clone end sequences as developed in cucumber (see Chapter 2). ........................................................................................................... 183
vi
Appendix G. SNPs identified between Gy-7 and H-19 in markers from two sequencing sources [sequencing of SCAR fragments (SCAR) and utilizing markers to identify BAC clones for sequencing (BAC end); see Chapter 2]. Asterisks (*) in either the Gy-7 or H-19 sequence (cucumber parental lines used in map construction) indicate an insertion/deletion (indel). ....................................................................................... 186
Appendix H. Characteristics of single nucleotide polymorphism (SNP) markers (comprising an allele-specific primer based on a SNP, and a non-specific primer) and their allele-specific primers as developed in cucumber (see Chapter 2). ................... 190
Appendix I. Primer names and sequences of SNP markers as developed in cucumber (see Chapter 2). ........................................................................................................... 194
Appendix J. Sample alignment of sequences from the RAPD marker OP-C1 generated in cucumber by Horesji et al. (1999; silver staining sequencing) and Chapter 2. Asterisks highlight differences in sequences and undetermined bases (N) are indicated by ^.............................................................................................................................. 199
Appendix K. The effect of increasing the number of quantitative trait loci (QTL) on the number of lateral branches in two leaf types of cucumber in two locations as determined by near-isogenic lines (NIL; Table 3.2). Lines indicate the incremental addition of QTL, and the comparisons in the legends refer to means comparisons performed (Table 3.4; Chapter 3). .............................................................................. 200
Appendix L. Analysis of variance (ANOVA) table of a test for main effects and interactions on the number of lateral branches in cucumber (MLB; Chapter 3). Effects examined were location (Hancock, Wisc. and Arlignton, Wisc.), replication (reps), with-in row spacings (10, 15, and 20 cm between plants), leaf type (standard and little leaf) and genotype [near-isogenic lines (NIL) that vary in the number of QTL (Table 3.2) for MLB]. The coefficient of variation (CV) was 19.6%................................... 201
vii
List of Tables Table 1. Sex phenotypes from combinations of the three major genes of sex expression in cucumber (Cucumis sativus L.). Adapted from Mibus and Tatlioglu (2004)…………………………………………13 Table 1.1. Mean values of yield component traits of commercial checks and parental inbred lines used to create four cucumber populations for comparison of response to selection by phenotype and marker. Means are from the replicated trial described herein…………………………………………………………….…70 Table 1.2. Cumulative selection differential over three cycles between Stage 1 and Stage 2 of phenotypic selection (PHE) for four traits in four populations of cucumber……………………. …………………..…71 Table 1.3. Characteristics of molecular markers defined in a genetic map of cucumber constructed by Fazio et al. (2003b) and used in marker-assisted selection for population improvement…………………………72 Table 1.4. Means and linear response of five traits in four base populations (C0) of cucumber which underwent three cycles of recurrent mass selection (C1-C3) using three methods………………………….73 Table 1.5. Phenotypic correlations (r) among traits in cucumber over three cycles of selection by markers (MAS) and phenotype (PHE)……………………………………………………………………………….75 Table 2.1. RAPD markers converted to SCAR markers in cucumber……………………………………110 Table 2.2. SCAR markers created from BAC end sequences in cucumber………………………………112 Table 2.3. Allele-specific markers created by four design approaches based on SNPs between two cucumber lines (Gy-7 and H-19) identified from two sequence sources………………………………….113 Table 2.4 Primer names and sequences of the polymorphic markers converted from RAPDs and verified by segregation in cucumber………………………………………………………………………………..114 Table 2.5. Results of RAPD to SCAR conversion in cucumber of RAPDs common to Horejsi et al. (1999) and this study………………………………………………………………………………………………117 Table 3.1. Characteristics of previously identified cucumber quantitative trait loci (QTL) associated with multiple lateral branching (MLB) that were introgressed to create nearly isogenic lines (NIL)………….137 Table 3.2. QTL composition and mean number of branches in near isogenic lines (NIL) of cucumber...138 Table 3.3. Marker-QTL associations used to introgress quantitative trait loci (QTL) for multiple lateral branching (MLB) into nearly isogenic lines (NIL) in cucumber…………………………………………..139 Table 3.4. Means comparisons to determine specific quantitative trait loci (QTL) effects in nearly isogenic lines (NIL) of cucumber…………………………………………………………………………………...140
viii
List of Figures
Figure 1.1. Schematic of the selection and evaluation scheme in cucumber where PHE is phenotypic selection, MAS is selection by marker, and RAN is random mating……………………………………….76
Figure 1.2. Response to selection as measured by the slope of linear regression (Y axes) over three cycles of MAS (selection by marker), PHE (phenotypic selection) or RAN (random mating) for five traits in four cucumber populations. *, **, and *** denote slopes are significant at P ≤ 0.05, P ≤ 0.01, and P ≤ 0.001, respectively………………………………………………………………………………………………….77
Figure 1.3. Time required to complete MAS (selection by marker) and PHE (phenotypic selection) for three cycles in four cucumber populations. Gray areas indicate evaluation of the populations and black areas represent recombination among selections……………………………………………………………78 Figure 2.1. SCAR multiplex reactions in cucumber. Panel C: banding patterns of individual SCAR primer pairs, including molecular weight in base pairs (bp) across a temperature gradient. Vertical numbers denote PCR annealing temperatures (oC) for each lane. Panels A, B, D, and E contain two, three, three, and four primer pairs, respectively, added to the same PCR reaction. Molecular weight of the EcoRI+HindIII digested lambda marker are to the right of panels B and E………………………………………………..118 Figure 2.2. Allele specific primer design used to create a codominant marker in cucumber from SNPs within a locus employing the optimal approach. Panel A: SNPs between H19 (bottom sequence) and Gy-7 (top sequence) in a portion of the AD14SCAR sequence are indicated by an asterisk (*). Allele specific primers match the SNP of one parent at the 3’ end with an additional mismatch (^) to both alleles within 4 bases of the 3’ end. Universal non-specific primers have no mismatch to either allele. Primer orientation and direction of extension by a polymerase (horizontal arrows) during PCR are indicated under each primer name. Panel B: Photograph after agarose gel electrophoresis of PCR reactions using Gy-7 and H-19 as template with the dominant Gy-7 allele specific marker (Gy-7 allele specific primer and Universal non-specific primer) labeled G, the dominant H-19 allele specific marker (H-19 allele specific primer and Universal non-specific primer) labeled H, and the G-y7 and H-19 allele specific markers combined (Gy-7 allele specific primer, H-19 allele specific primer, and Universal non-specific primer) in a codominant assay labeled “C”. The codominant assay was also tested on F1 and F2 individuals from a cross between Gy-7 and H-19. A 100 bp ladder (MW ladder) flanks PCR products……………….……………………119 Figure 2.3. Graphical representation of four design approaches developed herein to create Gy-7 and H-19 allele-specific markers in cucumber from a single locus depending on location and number of single nucleotide polymorphisms (SNPs). Solid horizontal lines represent a genomic fragment containing a SNP (asterisk) between Gy-7 and H-19. Arrows represent the direction of primer extension by a polymerase in PCR. Primers are designated as allele-specific (GS) and non allele-specific (GN) to amplify the Gy-7 allele, allele-specific (HS) and non allele-specific (HN) to amplify the H-19 allele, and non-allele-specific universal (UN) to amplify both alleles. The dotted line on the GS primer of the tail approach represents additional base pairs that do not anneal to the template during PCR, but are designed to add length to the PCR product. The horizontal dotted lines above and below the genomic fragment represent PCR products of Gy-7 and H-19 template, respectively. The panels on the far right represent the gel banding patterns of both approaches in each row after agarose gel electrophoresis with Gy-7, H-19, and an F1 hybrid as templates…………………………………………………………………………………………………...120
ix Figure 3.1. The effect of plant density on the number of lateral branches in cucumber near-isogenic lines (NIL) of two leaf types (little-leaf and standard-leaf). The linear and residual P-values, as well as the R2 of each leaf type are presented to the right of each leaf type. Within row spacings of 10, 15 and 20 cm correspond to plant densities of 66,700, 44,400, and 33,300 plants/hectare, respectively………………...141 Figure 3.2. The effect of increasing the number of quantitative trait loci (QTL) on the number of lateral branches in two leaf types of cucumber as determined by near-isogenic lines (NIL; Table 3.2). Lines indicate the incremental addition of QTL, and the comparisons in the legends refer to means comparisons performed (Table 3.4). Asterisks (*) indicate significant means comparisons (P < 0.05)………………..142
x
Abstract Several theoretically-based simulation studies suggest that the effectiveness of marker-
assisted selection (MAS) for polygenic traits can be greater than phenotypic selection
(PHE), but empirical comparisons are scarce and often conflicting. Therefore, existing
molecular tools [i.e., genetic linkage maps with defined quantitative trait loci (QTL)-
marker associations] were leveraged to compare the effectiveness of MAS to PHE for
several quantitative (conditioned by 2-6 QTL), yield-related traits [multiple lateral
branching (MLB), gynoecious sex expression (GYN), earliness (EAR), and fruit length to
diameter ratio (L:D)] in cucumber. Four complementary inbred lines were intermated to
produce four populations which underwent MAS, PHE, and random mating (no
selection) for three cycles of recurrent selection. Although both MAS and PHE improved
all traits, except for EAR by MAS, their effectiveness depended upon the traits and
populations under selection. PHE was most effective for GYN, EAR, and L:D, while
MAS was most effective for MLB and yield (fruit per plant). To increase the efficiency
of existing molecular markers, 43 random amplified polymorphic DNA (RAPD) markers
were sequenced to produce 22 polymorphic sequence characterized amplified region
(SCAR) markers. Sequences from bacterial artificial chromosome (BAC) clones and
monomorphic SCARs were obtained to identify single nucleotide polymorphisms (SNPs),
and four novel marker design approaches were utilized to create allele-specific markers
based on SNPs with an 80% success rate (20 markers created from 25 loci containing
xi SNPs). A total of 32 RAPDs were converted into SCAR or SNP markers that will
increase the efficiency of MAS in cucumber. Although QTL-marker associations have
provided for improvement of MLB by MAS, the epistatic effects of individual QTL have
not been characterized. Sets of nearly-isogenic lines (NIL) were created in two genetic
backgrounds (standard- and little-leaf type) with varying numbers of QTL associated
with MLB. Comparative analysis of specific QTL combinations among NIL
characterized epistatic interactions which were detected among QTL in the little leaf
background, but not in standard-leaf NIL. Genotype and QTL by environment
interactions were also identified, indicating that lateral branch production is determined
by environmental effects, interactions among other cucumber traits, and interactions
among QTL conditioning MLB.
1
Introduction
Cucumber
Commercial cucumber (Cucumis sativus L.; 2n=2x=14) has been of culinary
importance to humans for millennia. Although the cucumber is thought to originate in
India or Southern Asia, evidence from Northern Thailand suggests the earliest use of
cucumber by humans was approximately 9,750 B.C. (cucumber history reviewed by
Lower and Edwards 1986, Wehner 1989, Tatlioglu 1993, Meglic and Staub 1996a, and
Staub and Bacher 1997). The initial domestication of cucumber, however, is thought to
have occurred in India circa 3,000 years ago (Lower and Edwards 1986), which makes it
one of the oldest cultivated vegetable crops (Shetty and Wehner 2002). The
domestication of cucumber spread east from India to Western Asia, then west to Asia
Minor, North Africa, and Southern Europe before written history (Tatlioglu 1993).
Cucumber was cultivated by the Chinese (second century B.C.), Sumerians (2,500 B.C.),
ancient Greeks and Romans (300 B.C.), ancient Egyptians, French (9th century), and
English (15th century) before being carried to Haiti and New England by Christopher
Columbus at the end of the 15th century (Lower and Edwards 1986; Wehner 1989;
Tatlioglu 1993; Meglic and Staub 1996a). After its introduction into the Americas,
cucumber was grown in colonial gardens and by several North American Indian tribes
(Meglic and Staub 1996a; Staub and Bacher 1997). Cucumber is now grown in nearly all
countries in temperate zones (Tatlioglu 1993).
2 Cucumber belongs to the Cucurbitaceae or vine-crop family, which includes
only), andromonoecious (staminate and perfect flowers) or hermaphroditic (perfect
flowers only). The genetic control of sex expression can generally be explained by a
13three-gene model involving the F (Female), m (andromonoecious), and a
(androecious) genes (Table 1).
Table 1. Sex phenotypes from combinations of the three major genes of sex expression in cucumber (Cucumis sativus L.). Adapted from Mibus and Tatlioglu (2004).
fingerprinting, genetic map construction, gene tagging, and marker-assisted selection
(MAS; Collard et al. 2005). Several reviews have been published on markers regarding
their use in gene (or QTL) mapping and gene tagging, and their deployment in MAS
(Tanksley 1993; Staub et al. 1996; Jones et al. 1997; Gupta et al. 1999; Collard et al.
2005; Francia et al. 2005). Therefore, these subjects will not be reviewed here, but a
summary of important points relevant to the research presented herein will be provided.
The ideal markers for use in plant breeding programs are codominant (able to
distinguish heterozygotes), easy to develop and use, robust (repeatable and tolerable to
slight changes in detection), abundant, amenable to high-throughput systems, and low
cost (Staub et al. 1996; Jones et al. 1997; Gupta et al. 1999; Collard et al. 2005; Francia et
al. 2005). Although there are several types of markers, each has advantages and
disadvantages for their deployment in plant breeding.
Molecular marker types The three main types of genetic markers are morphological, biochemical (protein),
and DNA-based (i.e., molecular markers; Staub et al. 1996; Collard et al. 2005).
Although morphological (visualized as a phenotype, such as flower color) and
biochemical markers (allelic variants of functional enzymes, also referred to as isozymes)
were historically valuable, their paucity and variability due to environmental conditions
and developmental stages limit their effectiveness in plant genetics and breeding. The
22large majority of currently utilized markers are DNA-based because they are relatively
abundant, not influenced by the environment, and do not effect phenotype (Staub et al.
1996; Gupta et al. 1999; Collard et al. 2005).
The first widely used DNA-based markers were restriction fragment length
polymorphisms (RFLP; Botstein et al. 1980; Tanksley 1993). Although RFLPs are
codominant, fairly robust, and more prevalent than isozymes, they are costly, time-
consuming, laborious (not high-throughput), and not as abundant as other marker systems.
They also require large amounts of DNA, as well as the use of radiolabeled isotopes, and
cloning is a necessary part of marker development.
To overcome the time and labor requirements of RFLP markers, random
amplified polymorphic DNA (RAPD) markers were developed (Williams et al. 1990).
As their name implies, RAPDs are much quicker and easier to develop and utilize than
RFLPs, and they are comparatively more abundant, much less expensive, require less
DNA, and, in many cases, provide multiple markers per assay. RAPDs, however, are
typically dominant, not robust, and often methodologically problematic (Staub et al.
1996; Paran and Michelmore 1993).
Sequence characterized amplified region (SCAR) markers were initially designed
by Paran and Michelmore (1993) to convert a polymorphic RAPD marker into a robust,
single-copy marker. SCAR markers are produced by sequencing the RAPD band and
using the sequence at both ends of the fragment to extend the 10 bp RAPD primer an
additional 14 base pairs to produce a specific pair of primers. Since a SCAR marker is
defined as a fragment from genomic DNA generated from specific primers through PCR
23(Paran and Michelmore 1993), SCARs can also be derived from markers other than
RAPDs (e.g., RFLPs). The only requirement is that cloning and sequencing are needed
to design primers to specifically amplify a single product. Once developed, however,
SCARs are much more robust and repeatable than RAPDs, and are as easy and
inexpensive to use (Polashock and Vorsa 2002; Randig et al. 2002). Although SCAR
markers are usually dominant, codominant SCARs are not uncommon (Staub et al. 1996).
Because most SCAR markers produce a single band, they are amenable to multiplexing
(including two or more markers simultaneously in the same PCR reaction), which further
increases their efficiency during genotyping (Polashock and Vorsa 2002; Randig et al.
2002), and makes them amenable to high-throughput systems.
Amplified fragment length polymorphism (AFLP) markers are dominant, more
robust than RAPDs, and can provide several markers per assay (Vos et al. 1995).
Although the AFLP methodology is more technologically complicated than RAPDs, no
cloning or prior sequence knowledge is required. Initially, AFLPs required
polyacrylamide gel electrophoresis and labeling with radiolabeled isotopes, but they have
been adapted for automated sequencing platforms with fluorescent labeling (fAFLP;
Desai et al. 1998). AFLP markers are more expensive than RAPDs, and, except for
RFLPs, require only slightly more DNA than other marker systems for utilization.
There are several types of markers that require sequence information for
development in addition to SCARs. Simple sequence repeat (SSR or microsatellite)
markers take advantage of the fact that small (usually di-, tri-, tetra-, or penta-nucleotide),
tandemly repeated sequences tend to vary in length among haplotypes in a population
24(Gupta et al. 1999). These repeats are relatively abundant and highly polymorphic in
plants (Staub et al. 1996). SSRs are usually developed by creating a library enriched with
genomic fragments containing repeats, sequencing the fragments, then designing primers
flanking the repeats, which is expensive and time consuming. SSRs are codominant by
nature, and can have multiple alleles per locus because the tandem repeats vary in length
in genetically diverse populations. Once developed, SSRs are robust, but small
differences in molecular weight among band morphotypes often necessitate their
visualization by polyacrylamide gel electrophoresis. Like fAFLPs, SSRs can be
visualized in automated sequencing platforms, but unlike AFLPs or RAPDs, they can be
multiplexed in high-throughput systems.
Sequenced tag site (STS) markers were originally proposed as a standard for
simple, PCR-based markers created from RFLP probes in humans (Olson et al. 1989).
An STS is a short, single-copy marker that is associated with a specific locus and can be
amplified by PCR. Although STS and SCAR have been used synonymously in the
literature at times, STS is conventionally reserved for PCR markers made from RFLPs
(Gupta et al. 1999). STS markers are robust, relatively inexpensive, easy to use, and
amenable to high-throughput systems through multiplexing. STSs are usually dominant,
but can be codominant depending on their design and use.
Another marker type based on previously identified markers is cleaved amplified
polymorphic sequences or CAPS (Konieczny and Ausubel 1993). These markers are
based on sequence polymorphisms at restriction enzyme sites. To detect a CAPS marker,
a PCR amplified product is digested by a restriction enzyme to create a codominant
25length polymorphism that can be resolved by agarose gel electrophoresis. Sequence
information is necessary for the development of CAPS markers, and not all PCR products
contain restriction enzyme sites, which limits their development. The restriction enzyme
digestion of PCR fragments adds an extra step and expense when genotyping with CAPS
markers. In addition, the incomplete digestion of PCR products can reduce the reliability
of CAPS markers and complicate their scoring in some cases (Zheng et al. 1999; Burger
et al. 2003).
Markers based on single nucleotide polymorphisms (SNP) are gaining popularity
and are the current marker of choice for several species including crop plants (Gupta et al.
2001). This popularity is based on the idea that as more genomic resources are being
made available, SNPs are best able to fit the ideal marker for use in plant breeding. SNPs
are usually codominant and robust markers. The number of SNPs in any given genome is
much higher than any other marker type (estimated at 1 in 100 to 1 in 1000 base pairs),
including an order of magnitude higher than SSRs (Gupta et al. 2001). The rise in SNP
popularity has lead to several different methods of discovery and genotyping (Gupta et al.
2001). Some of these methods, such as pyrosequencing for SNP detection, are focused
on high-throughput systems. These and other non-gel based assays such as TaqMan,
Molecular Beacons, and array-based assays, are usually supported by proprietary
technologies which may be cost prohibitive to many plant breeding programs. SNP
genotyping, however, can be adapted to low cost methods using basic laboratory
equipment such as PCR followed by agarose gel electrophoresis in allele-specific PCR
(AS-PCR) or single-nucleotide amplified polymorphism (SNAP) assays (Drenkard et al.
262000; Moreno-Vazquez et al. 2003). The major disadvantages to the development of
SNPs markers are that sequence information is necessary for their design, and SNPs are
bi-allelic, unlike SSRs, which usually have multiple alleles per locus. The abundance of
SNPs, however, compensates for the limited number of alleles, making their development
cost-effective.
The selection of marker types for use in plant breeding depends on several factors
including project objectives, population and mating structure, genomic complexity, the
intended use of the markers, and the resources available (Staub et al. 1996; Gupta et al.
1999). For example, RAPD and AFLP are useful technologies for new marker
identification and molecular map construction because multiple markers can be identified
in each sample and no a priori sequence knowledge is needed (Paran and Michelmore
1993; Brugmans et al. 2003). Once established, however, SCAR, SNP, STS, and SSR
markers are much more useful in genotyping populations because of their robustness and
potential ability to be mutliplexed. The continued increase in sequence availability and
EST databases, allows for the creation of SNP, SSR, CAPS, and SCAR type markers
without having to generate sequence date. Furthermore, markers created from EST
databases are based upon transcribed loci, and may, therefore, be more suited to gene
tagging.
Marker conversion Although AFLPs and RAPDs are well suited for relatively rapid identification of
new markers, the low reproducibility of RAPDs and the technical methodology of AFLPs
has prompted the conversion of these markers into other, more robust and simple marker
27types better suited for tracking alleles in MAS (Brugmans et al. 2003; Collard et al.
2005). For instance, Shirasawa et al. (2004) converted 46 AFLP markers to 8 dominant
SCAR markers, six codominant SCAR markers, 13 CAPS markers, and 13 PCR-
restriction fragment single strand conformation polymorphism (PCR-RF-SSCP or PRS)
markers. Brugmans et al. (2003) report a systematic approach to AFLP marker
conversion based on novel polymorphisms within the AFLP band or the polymorphism
that produced the original polymorphic AFLP fragment. Through this approach all ten of
the randomly selected AFLP markers were converted into a single-copy, robust marker.
The term SCAR was originally applied to single-copy, reliable markers converted
from RAPDs (Paran and Michelmore 1993). In this initial study, nine RAPDs were
converted into five dominant, three codominant, and one monomorphic SCAR, and eight
out of the nine SCAR primer pairs produced a single band. Because the majority of
RAPD polymorphisms are in the priming site (Williams et al. 1990), SCAR primer pairs
included the original RAPD primer sequence plus an additional 14 base pairs in order to
transfer the RAPD polymorphism to the SCAR. Only three of the dominant markers,
however, retained the original polymorphism. The extension of the RAPD primers to
create a SCAR primer pair allowed amplification of contrasting genotypes from the other
six SCAR markers. One of these was monomorphic, but three length polymorphisms
resulted in the three codominant markers. A fifth SCAR marker that amplified both
genotypes at the molecular weight of the RAPD, also amplified a second, dominant band
that cosegregated with the original RAPD. The sixth marker amplified both genotypes at
60o C, but raising the PCR annealing temperature to 67o C recovered a dominant
28polymorphism. One of the codominant markers was detectable by agarose
electrophoresis only after digestion with a restriction enzyme. Another codominant
marker was not polymorphic until tested on genotypes other than those that produced the
original RAPD polymorphism. Thus, the conversion from RAPD to SCAR was highly
successful (89%), but required marker optimization.
The conversion rate was not as high in a RAPD to SCAR conversion study in
cucumber (Horejsi et al. 1999). Of the 75 RAPDs attempted, 48 (64%) were successfully
sequenced by silver staining of polyacrylamide gels, from which 48 primer (18 to 22 bp)
pairs (96 primers) were designed. Only 11 of the 48 (15%) primer pairs resulted in a
polymorphism and 20 (42%) produced more than one band per DNA template. This low
RAPD to SCAR conversion rate may be partially explained by the observation that the
majority of the SCAR primers did not contain the original RAPD primer. Three of the 96
SCAR primers contained all 10 bp of the original RAPD primer, while 13 contained at
least one bp, and 80 did not contain any part of the RAPD primer. Of the 48 SCAR
markers from Horejsi et al. (1999), 20 (42%) produced more than one band per DNA
template, a percentage which is greater than the 11% obtained by Paran and Michelmore
(1993). This difference may be the result of reduced primer specificity in the cucumber
SCARs because of the comparatively shorter primers (18 bp vs. 24 bp), which allowed
additional bands to be amplified in each template. Similar to Paran and Michelmore
(1993), an increase in PCR annealing temperature by Horejsi et al. (1999) identified an
additional dominant SCAR marker, illustrating the need to empirically determine optimal
annealing temperatures of the SCAR primers in order to recover polymorphisms.
29The results of both Paran and Michelmore (1993) and Horejsi et al. (1999)
illustrate that not all SCAR markers are polymorphic, even after marker optimization. In
such cases, both alleles of the SCAR marker may be sequenced to identify additional
polymorphisms that might be exploited to genotype the locus by PCR. Sequencing the
monomorphic SCAR, for example, in the original SCAR to RAPD conversion study
revealed two SNPs (Paran and Michelmore 1993). Such polymorphisms can be utilized
to create SNP markers through a variety of methods. Three such methods are based on
CAPS (Konieczny and Ausubel 1993), AS-PCR (Newton et al. 1989; Sarkar et al. 1990)
and SNAP assays (Drenkard et al. 2000). SNP markers that utilize the CAPS assay are
codominant, based on a SNP in a restriction site, and are detected by PCR, a restriction
enzyme digest, and then agarose gel electrophoresis. The AS-PCR and SNAP assays rely
on primers designed to specifically amplify one allele at a SNP and can be codominant if
a marker is created for each allele. The advantage of the AS-PCR or SNAP assay is that
they do not require a restriction enzyme step, as PCR is followed directly by agarose gel
electrophoresis.
Genetic mapping in cucumber The utility of molecular markers is greatly enhanced when they are placed on
genetic linkage maps, which allows the identification of simply inherited genes and
genomic regions involved in agronomically important traits through QTL mapping
(Collard et al. 2005). The first genetic linkage maps in cucumber were reported almost
20 years ago and were based solely on phenotypic markers (Fanourakis and Simon 1987;
Vakalounakis 1992; Pierce and Wehner 2000). The first biochemical markers mapped in
30cucumber were isozymes (Knerr and Staub 1992), and were later combined with
phenotypic markers to produce maps of low saturation (Meglic and Staub 1996b). As
DNA-based molecular markers were developed (RFLP and RAPD), they were combined
with existing marker types in linkage maps (Kennard et al. 1994). More recently
developed maps include phenotypic and several types of DNA based markers (RAPD,
RFLP, AFLP, SCAR, SSR, and SNP; Serquen et al. 1997a; Park et al. 2000; Fazio et al.
2003b). As genetic maps continued to be refined and molecular markers were included,
the total map distance generally expanded to match the estimated range of 750 to 1000
cM (Staub and Meglic 1993). Genetic distances in such maps were reported as 166
(Fanourakis and Simon 1987), 95 (Vakalounakis 1992), 168 (Knerr and Staub 1992), 766
(narrow-based), 480 (wide-based; Kennard et al. 1994), 584 (Meglic and Staub 1996b),
600 (Serquen et al. 1997a), 816 (Park et al. 2000), and 706 cM (Fazio et al. 2003b).
The incorporation of molecular markers has also increased the saturation of
cucumber genetic linkage maps. Park et al. (2000) employed 347 RAPD, RFLP, AFLP,
and loci conditioning virus resistances to construct a map with 12 linkage groups (LOD ≤
3.5) and a mean marker interval of 4.2 cM. A map constructed by Serquen et al. (1997a)
defined nine linkage groups and spanned ca. 600 cM with an average distance between
RAPD markers of 8.4 cM. Information from this map was recently merged with other
maps (Fanourakis and Simon 1987; Knerr and Staub 1992; Kennard et al. 1994; Meglic
and Staub 1996b; Horejsi et al. 2000) to synthesize a consensus map containing 255
markers, including morphological traits, disease resistance loci, isozymes, RFLPs,
RAPDs, and AFLPs spanning 10 linkage groups (Bradeen et al. 2001). The mean marker
31interval in this consensus map was 2.1 cM spanning a total length of 538 cM. More
recently, Fazio et al. (2003b) constructed a map containing 14 SSR, 24 SCAR, 27 AFLP,
62 RAPD, one SNP, and three morphological markers (131 total markers) spanning seven
linkage groups (the theoretical number based on the haploid chromosome number) using
RIL. This map spanned 706 cM with a mean marker interval of 5.6 cM.
The development of genetic linkage maps have provided tools for the molecular
analysis of important characteristics in cucumber including fruit quality (Wenzel et al.
1995), disease resistance, (Park et al. 2000) and yield components (Serquen et al. 1997a;
Fazio et al. 2003b). The marker-QTL associations identified in these studies form the
foundation for crop improvement through marker-assisted selection.
Marker-assisted Selection
One of the primary purposes of creating genetic linkage maps coupled with QTL
analysis, is to utilize marker-QTL associations in MAS. Several simulation studies
suggest that the effectiveness of MAS for polygenic traits can be greater than traditional
breeding (Lande and Thompson 1990; Zhang and Smith 1992; Edwards and Page 1994;
Gimelfarb and Lande 1994a; Gimelfarb and Lande 1994b). In general, these studies
agree that MAS efficiency is enhanced when markers are tightly linked (< 5.0 cM) to
quantitative trait loci (QTL) and selection is performed in early generations prior to
recombination between markers and QTL, on large sample sizes, and on traits with low
heritability. In practice, MAS has been effective for the introgression of simple traits or a
small number of genes in several crop species [e.g., disease resistance in common bean
32(Phaseolus vulgaris L.; de Oliveira et al. 2005), grain protein concentration in durum
wheat (Triticum turgidum L. var. durum; Chee et al. 2001), and root depth in rice (Oryza
sativa L.; Shen et al. 2001)], but less effective for complex traits [e.g., yield in barley
(Hordeum vulgare L.; Kandemir et al. 2000), and soybean (Glycine max (L.) Merrill;
Reyna and Sneller 2001)]. Studies reporting the empirical comparison of MAS to
phenotypic selection (PHE), however, are scarce and often conflicting (Yousef and Juvik
2001; Willcox et al. 2002; Hoeck et al. 2003).
MAS has been found to be more (Yousef and Juvik 2001; Fazio et al. 2003a;
Zhang et al. 2006), equivalent (Stromberg et al. 1994; Romagosa et al. 1999; Van Berloo
and Stam 1999; Willcox et al. 2002; Moreau et al. 2004), or less (Hoeck et al. 2003)
efficient and/or effective for increasing gain from selection when compared to PHE in
various plant species. Additional comparisons of MAS and PHE have provided mixed
results within the same study (Schneider et al. 1997; Flint-Garcia et al. 2003).
Cucumber possesses several characteristics that are favorable for MAS, including
a small genome size (genetic map length of 750 to 1,000 cM, 882 Megabases; Staub and
Meglic 1993) low chromosome number (n = 7), and rapid life cycle (four cycles per year).
In addition, moderately saturated genetic linkage maps have been developed, and QTL
analyses have identified several genomic locations associated with important traits in
cucumber. Two experiments have been reported using the marker-QTL associations
identified by Serquen et al. (1997a) and Fazio et al. (2003b) for MAS of yield
components.
33Fazio et al. (2003a) compared the response of the number of lateral branches
(MLB) to phenotypic selection under open-field conditions (PHE), random intermating
without selection (RAN), and MAS employing five markers (two SSRs, two RAPDs and
one SNP) in two backcross generations. No significant differences (p < 0.001) were
detected in either backcross generation between the mean values of MLB from PHE and
MAS, which were both significantly higher than RAN (control). Since two cycles of
MAS required one year compared to three for PHE, MAS increased overall breeding
efficiency.
The effect of MAS for four yield components (MLB, gynoecy, fruit L:D, and
earliness) was evaluated in two backcross populations (line extraction) after two cycles of
phenotypic recurrent selection (population improvement). Even after PHE provided
gains in MLB and L:D, MAS continued to improve both these traits in one backcross
population and L:D in the other. MAS also provided an increase in gynoecy in both
populations (Fan et al. 2006). Thus, MAS operated to fix favorable alleles that were not
exploited by phenotypic selection.
The results of these two studies demonstrate the utility of MAS for several
quantitative traits in cucumber. Furthermore, a response to selection from MAS for MLB,
L:D, and gynoecy confirms the marker-QTL associations for these traits. Relatively little
is known, however, about the individual QTL involved these traits, such as their
architecture (i.e., single genes or gene families), specific functions, interactions with the
environment, or epistatic effects (i.e., interactions among QTL of the same trait or with
QTL for other traits).
34Epistasis
Knowledge of epistasis is critical to comprehensive genetic analysis (Kinghorn
1987; Yano 2001) and breeding (Schnell and Cockerham 1992). Although epistasis is
considered in classical quantitative genetic theory (Falconer and Mackay 1996; Bernardo
2002), it is generally ignored in QTL mapping studies (Carlborg and Haley 2004). The
term ‘epistasis’ was originally applied to simply inherited traits where the actions of one
locus mask the effects of another locus (Carlborg and Haley 2004). Epistasis can be
more broadly defined, however, as a difference in phenotype from the same genotype
when in different genetic backgrounds. Epistatic interactions can occur among QTL
affecting the same trait, or between loci involved in several traits (Carlborg and Haley
2004).
Epistatic interactions have historically been detected by classical quantitative
genetic methods at the whole genome level, but the use of molecular markers and QTL
mapping studies has provided the ability to study epistasis between individual loci
(Tanksley 1993). The small population size, and interference with other QTL make the
detection of epistasis difficult in the primary populations [F2, F2-derived F3 (F2:3), or RIL]
utilized for QTL analyses. The analysis of near-isogenic lines (NIL) that differ in
specific QTL, however, provides a more powerful examination of epistasis, including
interactions between specific QTL (Lin et al. 2000). Molecular markers linked to QTL
for heading data in rice (Oryza sativa L.) were employed during backcrossing to create
NIL that contained one, two, or all three QTL under investigation (Lin et al. 2000). An
analysis of these NIL under different daylength conditions revealed epistatic interactions
35between specific QTL. One of the QTL did not affect photoperiod sensitivity alone,
but enhanced the expression of another QTL for heading date. This approach was
utilized to create other NIL with different QTL involved in heading date, from which,
additional epistatic interactions were identified (Yamamoto et al. 2000).
The development of markers (Horejsi et al. 1999; Fazio et al. 2002) and the
subsequent construction of genetic linkage maps and QTL analyses of Serquen et al.
(1997a) and Fazio et al. (2003b) have provided tools for the molecular improvement of
yield components in cucumber. Improvements have been made for several yield
components during inbred line development by MAS in backcross breeding (Fazio et al.
2003a; Fan et al. 2006). However, there is a need for additional, efficient markers, a
more complete understanding of gene interactions, and comparative analyses of DNA-
based multi-trait selection methods. Therefore, the effectiveness of MAS was evaluated
for three cycles of population improvement by recurrent selection for MLB, earliness,
L:D, and gynoecy (Chapter 1). Four inbred lines were intermated to create four base
populations that each underwent MAS, phenotypic selection and random mating under
the same selection scheme to determine whether responses to selection from markers and
phenotype for several quantitative traits were equally effective (greater than random
mating). To increase the efficiency and effectiveness of future applications of MAS in
cucumber (Chapter 2), a number of RAPD markers from the map of Fazio et al. (2003b)
were cloned and sequenced to create polymorphic SCAR markers, and the Gy-7 and H-
19 alleles of several monomorphic and dominant SCARs were sequenced. Sequences
were also obtained from BAC clones that hybridized to markers linked to important yield
36component QTL, and SNPs between Gy-7 and H-19 sequences from both sources
were identified and utilized to create codominant markers. Lastly, several NIL with
various numbers of QTL involved in MLB were developed and evaluated to characterize
the role of epistasis among individual QTL and genetic background on the expression of
MLB (Chapter 3).
Literature Cited
Baird JR, Thieret JW (1988) The bur gherkin (Cucumis anguria var. Anguria, Cucurbitaceae). Econ Bot 42:447-451
Bernardo R (2002) Breeding for Quantitative Traits in Plants. 1st edn. Woodbury, Stemma Press
Botstein D, White RL, Skolnick M, Davis RW (1980) Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am J Hum Genet 32:314-331
Bradeen JM, Staub JE, Wye C, Antonise R, Peleman J (2001) Towards an expanded and integrated linkage map of cucumber (Cucumis sativus L.). Genome 44:111-119
Brugmans B, van der Hulst RGM, Visser RGF, Lindhout P, van Eck HJ (2003) A new and versatile method for the successful conversion of AFLP markers into simple single locus markers. Nucleic Acids Res 31:e55
Burger Y, Katzir N, Tzuri G, Portnoy V, Saar U, Shriber S, Perl-Treves R, Cohen R (2003) Variation in the response of melon genotypes to Fusarium oxysporum f.sp melonis race 1 determined by inoculation tests and molecular markers. Plant Pathol 52:204-211
Carlborg O, Haley CS (2004) Epistasis: too often neglected in complex trait studies? Nat Rev Genet 5:618-U4
Chee PW, Elias EM, Anderson JA, Kianian SF (2001) Evaluation of a high grain protein QTL from Triticum turgidum L. var. dicoccoides in an adapted durum wheat background. Crop Sci 41:295-301
Chen JF, Staub J, Adelberg J, Lewis S, Kunkle B (2002) Synthesis and preliminary characterization of a new species (amphidiploid) in Cucumis. Euphytica 123:315-322
37Chen JF, Kirkbride JH (2000) A new synthetic species of Cucumis (Cucurbitaceae) from interspecific hybridization and chromosome doubling. Brittonia 52:315-319
Chen JF, Staub JE, Tashiro Y, Isshiki S, Miyazaki S (1997) Successful interspecific hybridization between Cucumis sativus L. and C. hystrix Chakr. Euphytica 96:413-419
Collard BCY, Jahufer MZZ, Brouwer JB, Pang ECK (2005) An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142:169-196
Cramer CS, Wehner TC (2000a) Path analysis of the correlation between fruit number and plant traits of cucumber populations. HortScience 35:708-711
Cramer CS, Wehner TC (2000b) Fruit yield and yield component correlations of four pickling cucumber populations. Cucurbit Genet Coop Rpt 23:12-15
Cramer CS, Wehner TC (1999) Little heterosis for yield and yield components in hybrids of six cucumber inbreds. Euphytica 110:99-108
Cramer CS, Wehner TC (1998a) Fruit yield and yield component means and correlations of four slicing cucumber populations improved through six to ten cycles of recurrent selection. J Am Soc Hort Sci 123:388-395
Cramer CS, Wehner TC (1998b) Fruit yield and yield components of cucumber populations grown at low plant density. In: McCreight JD (eds) Cucurbitaceae '98: Evaluation and enhancement of cucurbit germplasm. 1st ed. Alexandria, ASHS Press, pp 277-285
de Oliveira EJ, Alzate-Marin AL, Borem A, Fagundes SD, de Barros EG, Moreira MA (2005) Molecular marker-assisted selection for development of common bean lines resistant to angular leaf spot. Plant Breed 124:572-575
Desai M, Tanna A, Wall R, Efstratiou A, George R, Stanley J (1998) Fluorescent amplified-fragment length polymorphism analysis of an outbreak of group A streptococcal invasive disease. J Clin Microbiol 36:3133-3137
Dijkhuizen A, Kennard WC, Havey MJ, Staub JE (1996) RFLP variation and genetic relationships in cultivated cucumber. Euphytica 90:79-87
Drenkard E, Richter BG, Rozen S, Stutius LM, Angell NA, Mindrinos M, Cho RJ, Oefner PJ, Davis RW, Ausubel FM (2000) A simple procedure for the analysis of single nucleotide polymorphisms facilitates map-based cloning in Arabidopsis. Plant Physiol 124:1483-1492
38Edwards MD, Page NJ (1994) Evaluation of marker-assisted selection through computer simulation. Theor Appl Genet 88:376-382
Fan Z, Robbins MD, Staub JE (2006) Population development by phenotypic selection with subsequent marker-assisted selection for line extraction in cucumber (Cucumis sativus L.). Theor Appl Genet 112:843-855
Fanourakis NE, Simon PW (1987) Analysis of genetic linkage in the cucumber. J Hered 78:238-242
FAOSTAT (2005) Agriculture. FAO Statistical Databases (http://faostat.fao.org, last accessed December 2005)
Fazio G, Chung SM, Staub JE (2003a) Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.). Theor Appl Genet 107:875-883
Fazio G, Staub JE, Chung SM (2002) Development and characterization of PCR markers in cucumber. J Am Soc Hort Sci 127:545-557
Fazio G, Staub JE, Stevens MR (2003b) Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor Appl Genet 107:864-874
Fazio G (2001) Comparative study of marker-assisted and phenotypic selecton and genetic analysis of yield components in cucumber. PhD dissertation University of Wisconsin Madison
Flint-Garcia SA, Darrah LL, McMullen MD, Hibbard BE (2003) Phenotypic versus marker-assisted selection for stalk strength and second-generation European corn borer resistance in maize. Theor Appl Genet 107:1331-1336
Francia E, Tacconi G, Crosatti C, Barabaschi D, Bulgarelli D, Dall'Aglio E, Vale G (2005) Marker assisted selection in crop plants. Plant Cell Tissue Organ Cult 82:317-342
Fredrick LR, Staub JE (1989) Combining ability analyses of fruit yield and quality in near-homozygous lines derived from cucumber. J Am Soc Hort Sci 114:332-338
Gimelfarb A, Lande R (1994a) Simulation of marker assisted selection for nonadditive traits. Genet Res 64:127-136
Gimelfarb A, Lande R (1994b) Simulation of marker assisted selection in hybrid populations. Genet Res 63:39-47
39Goode MJ, Bowers JL, Bassi Jr. A (1980) Little leaf, a new kind of pickling cucumber plant. Arkansas Farm Res 29:4
Gupta PK, Roy JK, Prasad M (2001) Single nucleotide polymorphisms: A new paradigm for molecular marker technology and DNA polymorphism detection with emphasis on their use in plants. Curr Sci 80:524-535
Gupta PK, Varshney RK, Sharma PC, Ramesh B (1999) Molecular markers and their applications in wheat breeding. Plant Breed 118:369-390
Hoeck JA, Fehr WR, Shoemaker RC, Welke GA, Johnson SL, Cianzio SR (2003) Molecular marker analysis of seed size in soybean. Crop Sci 43:68-74
Horejsi T, Staub JE, Thomas C (2000) Linkage of random amplified polymorphic DNA markers to downy mildew resistance in cucumber (Cucumis sativus L.). Euphytica 115:105-113
Horejsi T, Box JM, Staub JE (1999) Efficiency of randomly amplified polymorphic DNA to sequence characterized amplified region marker conversion and their comparative polymerase chain reaction sensitivity in cucumber. J Am Soc Hort Sci 124:128-135
Horejsi T, Staub JE (1999) Genetic variation in cucumber (Cucumis sativus L.) as assessed by random amplified polymorphic DNA. Genet Resour Crop Evol 46:337-350
Horst EK, Lower RL (1978) Cucumis hardwickii: A source of germplasm for the cucumber breeder. Cucurbit Genet Coop Rpt 1:5
Jones N, Ougham H, Thomas H (1997) Markers and mapping: we are all geneticists now. New Phytol 137:165-177
Kandemir N, Jones BL, Wesenberg DM, Ullrich SE, Kleinhofs A (2000) Marker-assisted analysis of three grain yield QTL in barley (Hordeum vulgare L.) using near isogenic lines. Mol Breed 6:157-167
Kennard WC, Poetter K, Dijkhuizen A, Meglic V, Staub JE, Havey MJ (1994) Linkages among RFLP, RAPD, isozyme, disease resistance, and morphological markers in narrow and wide crosses of cucumber. Theor Appl Genet 89:42-48
Kinghorn BP (1987) The nature of 2-locus epistatic interactions in animals: evidence from Sewall Wright's guinea pig data. Theor Appl Genet 73:595-604
Kirkbride JH (1993) Biosystematic monograph of the genus Cucumis (Cucurbitaceae). Boone, N.C., Parkway Publishers, pp 159
Knerr LD, Staub JE (1992) Inheritance and linkage relationships of isozyme loci in cucumber (Cucumis sativus L.). Theor Appl Genet 84:217-224
40Knerr LD, Staub JE, Holder DJ, May BP (1989) Genetic diversity in Cucumis sativus L. assessed by variation at 18 allozyme coding loci. Theor Appl Genet 78:119-128
Konieczny A, Ausubel FM (1993) A procedure for mapping Arabidopsis mutations using codominant ecotype-specific PCR-based markers. Plant J 4:403-410
Lande R, Thompson R (1990) Efficiency of marker-assisted selection in the improvement of quantitative traits. Genetics 124:743-756
Lin HX, Yamamoto T, Sasaki T, Yano M (2000) Characterization and detection of epistatic interactions of 3 QTLs, Hd1, Hd2, and Hd3, controlling heading date in rice using nearly isogenic lines. Theor Appl Genet 101:1021-1028
López-Sesé AI, Staub J (2002) Combining ability analysis of yield components in cucumber. J Am Soc Hort Sci 127:931-937
Meglic V, Staub JE (1996a) Genetic diversity in cucumber (Cucumis sativus L.): II. An evaluation of selected cultivars released between 1846 and 1978. Genet Resour Crop Evol 43:547-558
Meglic V, Staub JE (1996b) Inheritance and linkage relationships of isozyme and morphological loci in cucumber (Cucumis sativus L). Theor Appl Genet 92:865-872
Mibus H, Tatlioglu T (2004) Molecular characterization and isolation of the F/f gene for femaleness in cucumber (Cucumis sativus L.). Theor Appl Genet 109:1669-1676
Moreau L, Charcosset A, Gallais A (2004) Experimental evaluation of several cycles of marker-assisted selection in maize. Euphytica 137:111-118
Moreno-Vazquez S, Ochoa OE, Faber N, Chao SM, Jacobs JME, Malsonneuve B, Kesseli RV, Michelmore RW (2003) SNP-based codominant markers for a recessive gene conferring resistance to corky root rot (Rhizomonas suberifaciens) in lettuce (Lactuca sativa). Genome 46:1059-1069
Morton JF (1987) The horned cucumber, alias Kiwano (Cucumis metuliferus, Cucurbitaceae). Econ Bot 41:325-327
Newton CR, Graham A, Heptinstall LE, Powell SJ, Summers C, Kalsheker N, Smith JC, Markham AF (1989) Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res 17:2503-2516
Nienhuis J, Lower RL (1988) Comparison of two recurrent selection procedures for yield in two pickling cucumber populations. J Am Soc Hort Sci 113:272-276
41Olson M, Hood L, Cantor C, Botstein D (1989) A common language for physical mapping of the human genome. Science 245:1434-1435
Paran I, Michelmore RW (1993) Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor Appl Genet 85:985-993
Park YH, Sensoy S, Wye C, Antonise R, Peleman J, Havey MJ (2000) A genetic map of cucumber composed of RAPDs, RFLPs, AFLPs, and loci conditioning resistance to papaya ringspot and zucchini yellow mosaic viruses. Genome 43:1003-1010
Pierce LK, Wehner TC (2000) Review of genes and linkage groups in cucumber. HortScience 25:605-615
Polashock JJ, Vorsa N (2002) Development of SCAR markers for DNA fingerprinting and germplasm analysis of American cranberry. J Am Soc Hort Sci 127:677-684
Randig O, Bongiovanni M, Carneiro RMDG, Castagnone-Sereno P (2002) Genetic diversity of root-knot nematodes from Brazil and development of SCAR markers specific for the coffee-damaging species. Genome 45:862-870
Reyna N, Sneller CH (2001) Evaluation of marker-assisted introgression of yield QTL alleles into adapted soybean. Crop Sci 41:1317-1321
Romagosa I, Han F, Ullrich SE, Hayes PM, Wesenberg DM (1999) Verification of yield QTL through realized molecular marker-assisted selection responses in a barley cross. Mol Breed 5:143-152
Sarkar G, Cassady J, Bottema CDK, Sommer SS (1990) Characterization of polymerase chain reaction amplification of specific alleles. Anal Biochem 186:64-68
Schneider KA, Brothers ME, Kelly JD (1997) Marker-assisted selection to improve drought resistance in common bean. Crop Sci 37:51-60
Schnell FW, Cockerham CC (1992) Multiplicative vs. arbitrary gene action in heterosis. Genetics 131:461-469
Serquen FC, Bacher J, Staub JE (1997a) Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random-amplified polymorphic DNA markers. Mol Breed 3:257-268
Serquen FC, Bacher J, Staub JE (1997b) Genetic analysis of yield components in cucumber at low plant density. J Am Soc Hort Sci 122:522-528
Shen L, Courtois B, McNally KL, Robin S, Li Z (2001) Evaluation of near-isogenic lines of rice introgressed with QTLs for root depth through marker-aided selection. Theor Appl Genet 103:75-83
42Shetty NV, Wehner TC (2002) Screening the cucumber germplasm collection for fruit yield and quality. Crop Sci 42:2174-2183
Shirasawa K, Kishitani S, Nishio T (2004) Conversion of AFLP markers to sequence-specific markers for closely related lines in rice by use of the rice genome sequence. Mol Breed 14:283-292
Staub JE, Bacher J (1997) Cucumber as a Processed Vegetable. In: Smith DS, Cash JN, Nip W, Hui YH (eds) Processing Vegetables: Science and Technology. Lancaster, Technomic Publishing Co., Inc., pp 129-193
Staub JE, Serquen FC, Gupta M (1996) Genetic markers, map construction, and their application in plant breeding. HortScience 31:729-741
Staub JE, Meglic V (1993) Molecular genetic markers and their legal relevance for cultigen discrimination: A case study in cucumber. HortTechnology 3:291-300
Stromberg LD, Dudley JW, Rufener GK (1994) Comparing conventional early generation selection with molecular marker assisted selection in maize. Crop Sci 34:1221-1225
Tanksley SD (1993) Mapping polygenes. Annu Rev Genet 27:205-233
Tatlioglu T (1993) Cucumber Cucumis sativus L. In: Kalloo G, Bergh BO (eds) Genetic Improvement of Vegetable Crops. Tarrytown, Pergamon Press Ltd., pp 197-234
Trebitsh T, Staub JE, ONeill SD (1997) Identification of a 1-aminocyclopropane-1-carboxylic acid synthase gene linked to the female (F) locus that enhances female sex expression in cucumber. Plant Physiol 113:987-995
USDA NASS (2006) Vegetables: 2005 summary. United States Department of Agriculture, National Agricultural Statistics Service Vg 1-2 (06)
Vakalounakis DJ (1992) Heart leaf, a recessive leaf shape marker in cucumber: Linkage with disease resistance and other traits. J Hered 83:217-221
Van Berloo R, Stam P (1999) Comparison between marker-assisted selection and phenotypical selection in a set of Arabidopsis thaliana recombinant inbred lines. Theor Appl Genet 98:113-118
Vos P, Hoger R, Bleeker M, Reijans M, van de Lee T, Hornes M, Frijters A, Pot J, Peleman J, Kuiper M, Zabeau M (1995) AFLP: a new technique for DNA fingerprinting. Nucleic Acids Res 23: 4407-4414
Wehner TC, Cramer CS (1996a) Gain for pickling cucumber yield and fruit shape using recurrent selection. Crop Sci 36:1538-1544
43Wehner TC, Cramer CS (1996b) Ten cycles of recurrent selection for fruit yield, earliness, and quality in three slicing cucumber populations. J Am Soc Hort Sci 121:362-366
Wehner TC (1989) Breeding for improved yield in cucumber. Plant Breed Rev 6:323-359
Wehner TC, Lower RL, Staub JE, Tolla GE (1989) Convergent-divergent selection for cucumber fruit yield. HortScience 24:667-669
Wehner TC, Miller CH (1985) Effect of gynoecious expression on yield and earliness of a fresh-market cucumber hybrid. J Am Soc Hort Sci 110:464-466
Wehner TC, Staub JE, Peterson CE (1978) Inheritance of littleleaf and multi-branched plant type in cucumber. Cucurbit Genet Coop Rpt 10:33-34
Wenzel G, Kennard WC, Havey MJ (1995) Quantitative trait analysis of fruit quality in cucumber: QTL detection, confirmation, and comparison with mating-design variation. Theor Appl Genet 91:53-61
Willcox MC, Khairallah MM, Bergvinson D, Crossa J, Deutsch JA, Edmeades GO, Gonalez-de-Leon D, Jiang C, Jewell DC, Mihm JA, Williams WP, Hoisington D (2002) Selection for resistance to southwestern corn borer using marker-assisted and conventional backcrossing. Crop Sci 42:1516-1528
Williams JGK, Kubelik AR, Livak KJ, Rafalski JA, Tingey SV (1990) DNA polymorphisms amplified by arbitrary primers are useful as genetic markers. Nucleic Acids Res 18:6531-6535
Xie J, Wehner TC (2001) Gene list 2001 for cucumber. Cucurbit Genet Coop Rpt 24:110-136
Yamamoto T, Lin HX, Sasaki T, Yano M (2000) Identification of heading date quantitative trait locus Hd6 and characterization of its epistatic interactions with Hd2 in rice using advanced backcross progeny. Genetics 154:885-891
Yano M (2001) Genetic and molecular dissection of naturally occurring variation. Curr Opin Plant Biol 4:130-135
Yousef GG, Juvik JA (2001) Comparison of phenotypic and marker-assisted selection for quantitative traits in sweet corn. Crop Sci 41:645-655
Zhang C, Tar'an B, Warkentin T, Tullu A, Bett KE, Vandenberg B, Somers DJ (2006) Selection for lodging resistance in early generations of field pea by molecular markers. Crop Sci 46:321-329
44Zhang W, Smith C (1992) Computer simulation of marker-assisted selection utilizing linkage disequilibrium. Theor Appl Genet 83:813-820
Zheng XY, Wolff DW, Baudracco-Arnas S, Pitrat M (1999) Development and utility of cleaved amplified polymorphic sequences (CAPS) and restriction fragment length polymorphisms (RFLPs) linked to the Fom-2 fusarium wilt resistance gene in melon (Cucumis melo L.). Theor Appl Genet 99:453-463
45
Chapter 1. Comparative analysis of marker-assisted and phenotypic selection for yield components in cucumber
Abstract
Comparative analysis of marker-assisted (MAS) and phenotypic (PHE) selection
efficacy is important for deployment of MAS in plant breeding. For direct comparison of
the effectiveness of MAS and PHE, four cucumber (Cucumis sativus L.; 2n=2x=14)
inbred lines were intermated then bulked maternally to create four base populations for
recurrent mass selection. Each of these populations underwent three cycles of PHE
(open-field evaluations) and MAS (genotyping at 18 marker loci), as well as random
mating (RAN; no selection). The four yield-related traits that underwent selection
[multiple lateral branching (MLB), gynoecious sex expression (GYN), earliness (EAR),
and fruit length to diameter ratio (L:D)] are quantitatively inherited (2-6 QTL per trait).
Both MAS and PHE were useful for multi-trait improvement in cucumber, but their
effectiveness depended upon the traits and populations under selection. The populations
with maternal parents that were inferior for a trait responded favorably to selection, while
those with maternal parents of superior trait values either did not change or decreased
during selection. Both MAS and PHE provided improvements in all traits under selection
in at least one population, except for EAR by MAS. Generally, PHE was most effective
for GYN, EAR, and L:D, while MAS was most effective for MLB and provided the only
increase in yield (fruit per plant).
46Introduction
Yield has been a focus of cucumber (Cucumis sativus L.; 2n=2x=14) breeding for
over 50 years (Lower and Edwards 1986; Wehner 1989; Wehner et al. 1989). Although
the yield of U.S. processing cucumber increased steadily from 1950 to 1980, it has
reached a plateau since the early 1980’s (Shetty and Wehner 2002). Selecting directly for
yield is difficult (Wehner 1989) which is partially due to its relatively low narrow-sense
heritability (0.07 to 0.25) and the dramatic influence of the environment on trait
expression. The most effective breeding approach for yield improvement in cucumber
may be selection for traits directly related to yield (Wehner 1989; Cramer and Wehner
1998; Cramer and Wehner 2000b).
Several theoretically-based simulation studies suggest that the effectiveness of
marker-assisted selection (MAS) for polygenic traits can be greater than traditional
breeding (Lande and Thompson 1990; Zhang and Smith 1992; Edwards and Page 1994;
Gimelfarb and Lande 1994a; Gimelfarb and Lande 1994b). In general, these studies
agree that MAS efficiency is enhanced when markers are tightly linked (< 5.0 cM) to
quantitative trait loci (QTL), and selection is performed in early generations prior to
recombination between markers and QTL, in relatively large populations, and on traits
with low heritability. In practice, MAS has been effective for the introgression of simple
traits or a small number of genes in several crop species [e.g., disease resistance in
common bean (Phaseolus vulgaris L.; de Oliveira et al. 2005), grain protein
concentration in durum wheat (Triticum turgidum L. var. durum; Chee et al. 2001), and
root depth in rice (Oryza sativa L.; Shen et al. 2001)], but less effective for complex traits
47[e.g., yield in barley (Hordeum vulgare L.; Kandemir et al. 2000), and soybean (Glycine
max (L.) Merrill; Reyna and Sneller 2001)]. Empirical comparisons of MAS to
phenotypic selection (PHE), however, are scarce and often conflicting (Yousef and Juvik
2001; Willcox et al. 2002; Hoeck et al. 2003).
Four important yield components in cucumber are earliness (EAR), gynoecious
sex expression (GYN), fruit length to diameter ratio (L:D), and multiple lateral branching
(MLB; Cramer and Wehner 2000a; Fazio et al. 2003a). Each of these traits is under the
control of two to six major genes with relatively large effects, where narrow-sense
heritabilities (h2) range from 0.14 to 0.48 depending on trait and environment (Serquen et
al. 1997a; Fazio et al. 2003b). EAR, GYN, and MLB have been shown to be positively
correlated with the number of fruit per plant (Cramer and Wehner 2000a; Cramer and
Wehner 2000b; Fazio 2001), and L:D is an important determinant of marketable fruit
yield (Serquen et al. 1997b). Negative correlations exist however, between these yield
components (e.g., GYN with MLB and EAR with L:D) making the simultaneous
improvement of these traits a challenge.
The use of MAS in plant breeding has potential for increasing the efficiency and
effectiveness of plant improvement (Francia et al. 2005). Moderately saturated linkage
maps have been developed for cucumber and genomic regions have been identified that
have proven useful for selection of yield components by MAS (Fazio et al. 2003a; Fan et
al. 2006). These studies utilized yield-associated QTL identified initially by Serquen et
al. (1997a) and then by Fazio et al. (2003b) in a mapping population derived from a cross
between lines Gy-7 (synom. G421) and H-19. Selectable markers included those linked
48to QTL for EAR (LOD ≥ 4.1), GYN (LOD ≥ 3.0), L:D (LOD ≥ 4.2), and MLB (LOD ≥
3.0).
MAS in cucumber has proven to be effective in line extraction during backcross
breeding (Fazio et al. 2003a; Fan et al. 2006), but has not been evaluated for its efficacy
in population improvement. Given the potential utility of MAS, a study was designed to
increase cucumber yield by simultaneous selection of multiple yield components by MAS
and PHE, and to compare these methods for response to selection. In order to test their
efficacy, both methods used the same selection scheme which was designed to overcome
known negative correlations between yield components. Four populations were created
by intermating four inbred lines, and each population underwent three cycles of recurrent
selection by PHE and MAS, as well as random mating (RAN; no selection). Results will
allow for managerial decisions regarding the value and use of MAS in population
improvement of cucumber.
Materials and Methods
Germplasm Four genetically diverse but complementary inbred lines were used as parents for
the development of four populations that subsequently underwent selection. These
contrasting lines were drawn from the U.S. Department of Agriculture (USDA) cucumber
breeding program, Madison, Wisc., because the combination of their complementary
phenotypes (Table 1.1) provided the basis for selection of desirable aspects of EAR,
GYN, MLB, and L:D. Lines 6996A and 6995C were drawn from a recombinant inbred
49line (RIL; Gy-7 × H-19) population (F9; Staub et al. 2002). Line 6823B originated from
a cross between a parent (H-19) of the RIL population and a USDA elite processing line
whose progeny were then selected for H-19 attributes. Line 6632E is morphologically
similar to the other parent (Gy-7) of the RIL population, but does not have either parent
in its pedigree (Staub and Crubaugh 2001).
Selection scheme The four parental inbreds were intermated in a greenhouse in Madison, Wisc., in
2000 by pollinating female flowers of each inbred with bulked male flowers from the
other three lines (Figure 1.1). The resulting seeds were bulked by maternal parent to
create four populations (i.e., Population 1-4). Since these populations did not undergo
any selection, they are designated as cycle 0 (e.g., Population 1 C0). Each of these
populations subsequently underwent PHE, MAS, and RAN for three cycles (C1-C3). All
selection and mating was performed within each of the four populations, independent of
the other three populations (i.e., intrapopulation improvement only). Each population
was independently subjected to each of the two selection methods. PHE was performed
based on phenotype alone (i.e., without marker information), and MAS was applied
without regards to phenotypic information (i.e., marker information only).
Phenotypic selection was practiced on 400 (2001) C0, 600 (2002) C1, or 600
(2003) C2 plants in each population for EAR, GYN, MLB, L:D, and standard leaf type
under open-field conditions at the University of Wisconsin Experiment Station, Hancock,
Data were taken on individual plants, where leaf type was classified as standard (LL) or
50little leaf (ll = 30-40 cm2; Staub et al. 1992). Although EAR is usually measured as the
number of fruit per plant in the first harvest on a per plot basis, EAR was assessed on
individual plants as the number of days from planting to anthesis of the first female
flower, because the fruit number of single plants is a poor predictor of fruit number in
replicated trials (Wehner 1989). Sex expression was measured as the percentage of the
first 10 flowering nodes bearing female flowers (nodes with male and female flowers
were classified as male) where 100% was designated gynoecious, 50% to 90% was
considered predominantly female (PF), and less than 50% was classified as monoecious.
Fruit L:D was estimated by visual inspection of at least four immature fruit (USDA grade
size 3A-3B; 3.0 to 5.0 cm in diameter). MLB was recorded at or after anthesis as the
number of lateral branches (at least three internodes in length) in the first ten nodes of the
mainstem. PHE was accomplished in two stages within each cycle of selection using
minimum trait thresholds for the first stage, and index selection for the second stage. For
Stage 1, individual plants were first evaluated for leaf type, EAR, GYN, and MLB, since
these are the first traits to be expressed developmentally in cucumber. Only individuals
that met pre-established thresholds (i.e., standard-leaf, EAR < 48 days to the first female
flower, GYN > 50% female flowers, and MLB > 3 branches) were evaluated for L:D.
Those plants with an L:D above the threshold (> 2.8) were designated selections of Stage
1. An informal index was employed for Stage 2 of PHE where MLB and EAR were
weighted approximately 2 (MLB) and 1.5 (EAR) times that of GYN and L:D, which were
weighted equally. Individual plants were ranked by their values of MLB, then EAR, and
the values of GYN and L:D were used to make Stage 2 selections among the highest
51ranked individuals. The relative weights among the traits are illustrated by the selection
differentials (difference in trait means of the selections from Stage 2 and the selections
from Stage 1) for each trait in each population (Table 1.2). Twenty plants were selected
from Stage 2 in each cycle (C1-C3) of PHE within each population, representing a
standardized selection intensity (i) of 2.063, 2.219, and 2.219 for C1, C2, and C3,
respectively. Meristems of each Stage 2 selection were taken for cloning as rooted
cuttings. Once these cuttings were established, the apical meristems and surrounding
leaves were treated with two applications (7 days apart) of 3 mM silver thiosulfate
[Ag(S2O3)2]3- as a foliar spray to induce male flower production (Nijs and Visser 1980).
Selections were then randomly mated by pollination of each female flower with five
random male flowers.
For MAS, individual plants were genotyped using 18 markers linked to F
(femaleness), de (determinate), ll and previously identified QTL (Serquen et al. 1997a;
Fazio et al. 2003b) for EAR, GYN, MLB, and L:D (Table 1.3), and then selections were
made based on marker genotype. All markers employed were drawn from Fazio et al.
(2003b), except AJ6SCAR, and M8SCAR which were SCARs converted from previously
mapped RAPDs (Nam et al. 2005; Chapter 2). Marker type, genetic distance from QTL,
and number of QTL in proximity to the marker were considered when identifying
markers for use in MAS. Generally, dominant markers flanking QTL or codominant
markers tightly linked to QTL were selected (Robbins et al. 2002; Robbins and Staub
2004). All four parental lines carried only Gy-7 or H-19 alleles at each marker locus.
Thus, the desired genotype, or ideotype, was created based on parental allele constitution
52at each marker locus and the relationship of the QTL surrounding the marker locus (Table
1.3).
Tissue from all individuals and parental lines was harvested and DNA extraction,
polymerase chain reaction (PCR) amplification, and agarose gel electrophoresis was
conducted according to Fazio et al. (2003b). To increase marker efficiency, markers
were multiplexed in empirically determined groups (Table 1.3) according to Staub et al.
(2004) and Chapter 2. All individuals within a population were genotyped at each
marker locus, and individuals were selected whose genotype most closely matched the
ideotype at the greatest number of marker loci. The number of individuals tested, the
selection intensity, and crossing scheme for each cycle of MAS within each of the four
populations were identical to that of PHE.
Random mating was accomplished by sowing 20 randomly selected seeds from
each of the four C0 populations, and then chemically inducing male flowers in
gynoecious plants prior to intermating using the same scheme as that for MAS and PHE
to create C1. The resulting seeds were equally bulked, and 20 seeds were randomly
selected in the same manner to create C2 and C3.
Open-field evaluation of selection Response to selection was evaluated in the open-field trial at UWESH in the
summer of 2004 at two planting dates. Seeds were sown in a greenhouse in Madison,
Wisc. on June 4, 2004 and June 16, 2004, then transplanted on June 23, 2004 and July 7,
2004, respectively. Each planting date was arranged in a split-plot design with eight
replications of each population (whole plot factor) in randomized complete blocks, with a
53combination of cycle (i.e., C0,-C3) and method of selection (i.e., MAS, PHE, and RAN)
completely randomized as subplots with 10 plants per subplot. Plots were arranged in
single rows with 18 cm between plants and 1.5 m between rows (~37,000 plants/ha).
This plant density was chosen because it optimized potential yield in MLB genotypes in
multiple harvest operations (Fredrick and Staub 1989; Staub et al. 1992). The four inbred
lines that served as parents, as well as Gy-7, H-19, and the commercial cultivar ‘Vlasset’
(Seminis Vegetable Seeds, Inc, Oxnard, Calif.) were included as controls for comparison.
The traits evaluated were MLB, GYN, EAR, L:D, and total yield. The most
efficient measurement of yield in cucumber research is the total (marketable and
oversize) number of fruits per plant, since it has a higher heritability, is more stable over
time, and is easier to measure than other yield measurements (i.e., volume, mass, or
dollar value). Furthermore, fruit number is highly correlated (genetic correlation = 0.87)
with fruit weight (Wehner et al. 1989). The number of fruit per plot was counted at each
of four harvests [59, 66, 76, and 96 (first planting date) and 54, 64, 75, and 91 (second
planting date) days after planting] to calculate four-harvest means adjusted for plant stand.
Each of the four harvests occurred as two to three oversized fruit (>51 mm in diameter)
were observed within a plot (Wehner et al. 1989), where all immature fruits >20 mm in
diameter and >10 cm in length were included in total fruit number. Both MLB and GYN
were evaluated on each plant exactly as during PHE. Mean fruit L:D was obtained per
plot by measuring the length and diameter of 5–10 randomly selected fruits (USDA 2B-
3A grade; 2.5-3.0 cm in diameter), and then averaging over three harvests. EAR was
defined as the average number of fruits per plant in the first harvest.
54Statistical analysis
All response variables were initially analyzed by analysis of variance (ANOVA)
using PROC GLM of SAS (2003) to determine treatment effects. Treatments of planting
date, populations, cycles, and methods were considered fixed effects, while blocks were
considered random. Specific single-degree of freedom contrasts within analyses of
variance were employed to determine general response to selection for biologically
important comparisons (e.g., PHE and MAS). Selection responses (linear and quadratic
effects) were computed by regression of trait means on selection cycles within each
population for each selection method by employing single-degree of freedom contrasts
within ANOVA (Steele et al. 1996). To determine the relationship between the traits
under selection, phenotypic correlations among traits were calculated by Pearson
correlation using PROC CORR of SAS (2003).
Results
All main effects (planting date, populations, and combinations of cycles and
methods) were highly significant (P < 0.001) for all traits. In general, planting date
affected the magnitude of the mean value of a trait and not the entry ranking in response
to selection over cycles (Appendix A). Generally, the means of all traits were higher for
all populations in the first planting than the second, except for MLB, which was lower.
Although the planting date by population interaction was significant for L:D (P = 0.01)
and EAR (P = 0.001), general trends over cycles were the same for both plantings for all
traits. Selection was performed in each population independent of the other populations,
55and response to selection varied by population. Therefore, results are presented by
population with both plantings combined (Table 1.4; Figure 1.2; Appendix B).
Population 1 showed the widest range of response to selection over cycles when
compared to the other populations examined (Figure 1.2). All trait values decreased
significantly from RAN in this population, except L:D, which did not change. Both
MAS and PHE were effective at increasing MLB and L:D means, but EAR was
increased only by PHE. The only trait value to decrease through PHE was that of GYN,
which was diminutive when compared to MAS. The means of both EAR and GYN
decreased dramatically by MAS when compared to RAN and PHE. Yield values
remained fairly constant during PHE, but were reduced during MAS and RAN.
The effect of RAN was the most dramatic in Population 2 (Figure 1.2). The value
of two traits, L:D and MLB, increased slightly after three cycles of RAN, while the other
three trait means were reduced. The same directional trends were also apparent after
MAS. In contrast, PHE increased the mean of EAR and GYN, while no significant
change was detected in the other three traits.
In Population 3 after three cycles of RAN, the values of all traits except L:D were
unchanged (yield and MLB) or decreased (EAR and GYN; Figure 1.2). The increase in
the mean of L:D by PHE was similar to that of RAN. The mean of MLB also increased
after PHE, while EAR, GYN and yield values decreased. The decrease in the mean of
GYN after PHE, although significant, was dramatically less than the reduction detected
after MAS. The only trait value reduced by MAS was GYN. The means of EAR and
L:D were unchanged by MAS, while MLB and yield values increased.
56The mean values of three traits (EAR, L:D, and yield) in Population 4 were
decreased after RAN, while GYN and MLB remained unchanged over cycles (Figure 1.2).
The responses to selection from MAS and PHE were similar for all traits, except EAR. A
reduction in the means of L:D, MLB, and yield was detected during MAS, but such
reductions were not as large as those resulting from PHE. The gain from PHE for GYN
was, likewise, larger than that from MAS. The mean of EAR increased by PHE, but was
reduced by MAS and RAN. When compared to the other populations examined,
Population 4 showed the least improvement as trait values were increased in only three
instances (EAR by PHE, GYN by MAS, and GYN by PHE).
The four inbred lines (Table 1.1; Figure 1.1) used as parents in this study were
specifically chosen because high values for some of the traits under selection
complimented low values found in other lines (e.g., 6632E is high for GYN and EAR, but
low for MLB and L:D). Although this disparity among trait values was predictably
minimized in the C0 populations (Table 1.1), response to MAS and PHE varied by trait
and population. In general, the populations with maternal parents (i.e., inbred lines) that
were inferior for a trait responded favorably to selection while those with maternal
parents of superior trait values either did not change or decreased for certain traits. This
was most clearly observed in PHE for all traits except for EAR, which increased in
Population 1 even though the maternal parent (6632E) was superior for EAR. Trait
values for EAR remained comparatively low after MAS in Populations 2 and 4 (both
parents inferior for EAR), and results of MAS for GYN in Population 2 were similar. In
contrast, trait values after RAN generally decreased or remained unchanged. Trait values
57increased after RAN in only three cases: L:D and MLB in Population 2, and L:D in
Population 3. In several cases, trait values decreased after RAN in populations with
inferior maternal parents (e.g., MLB in Population 1).
The correlated response to selection [phenotypic correlation values (r)] after MAS
and PHE were somewhat different, and are presented by population in Table 1.5.
Consistent, positive correlations were detected for EAR with GYN and yield (r = 0.25 to
0.70), but EAR was always negatively correlated with MLB (r = -0.14 to -0.14).
Generally, EAR and L:D were not correlated, but a negative (Population 1 after MAS; r =
-0.26) and positive (Population 3 after MAS; r = 0.30) correlation was detected between
the two traits. Negative correlations were generally detected for GYN with L:D and
MLB (r = -0.07 to -0.64), but GYN was not correlated with yield, except in Population 1
after MAS (r = 0.47). Generally, L:D was positively correlated with MLB (r = 0.06 to
0.38), although more correlations were significant after PHE than MAS. In only two
populations [Populations 2 and 4 after both MAS (r = 0.34) and PHE (r = 0.32)] were
positive correlations detected between L:D and yield, while MLB and yield were not
correlated, except for a positive correlation in Population 4 after MAS (r = 0.27).
Discussion
Marker-assisted selection has been found to be more (Yousef and Juvik 2001;
Fazio et al. 2003a; Zhang et al. 2006), equivalent (Stromberg et al. 1994; Romagosa et al.
1999; Van Berloo and Stam 1999; Willcox et al. 2002; Moreau et al. 2004), or less
(Hoeck et al. 2003) efficient and/or effective for increasing gain from selection when
58compared to PHE in various plant species. Additional comparisons of MAS and PHE
have provided mixed results within the same study (Schneider et al. 1997; Flint-Garcia et
al. 2003). Moreover, these studies did not evaluate these selection methods for their
efficacy in the improvement of multiple, quantitatively inherited traits over multiple
cycles of recurrent selection. Data presented herein provide the first comprehensive,
comparative evaluation of MAS and PHE for such traits in a vegetable crop species.
Considerations for MAS The QTL for EAR, GYN, L:D, and MLB identified for selection in this study had
a relatively large effect (cumulative R2 > 37%-85% depending on trait and environment),
high LOD scores (>3.0; Table 1.3), and were consistent over several environments
(Serquen et al. 1997a; Fazio et al. 2003b). However, in several instances, QTL were so
tightly clustered that multiple QTL for different traits were located between adjacent
marker loci (e.g., QTL for all traits were linked to CSWCT28 and L18-SNP-H19 as well
as OP-AD12-1; Table 1.3). As the desired QTL allele came from different parental lines
for separate traits (e.g., EAR and GYN from Gy-7; MLB and L:D from H-19 at
CSWCT28), strategic decisions were made based on QTL effects and neighboring genes
to determine the most appropriate parental type for each marker locus. For example, the
Gy-7 allele was selected at OP-AD12-1, the marker linked to the little leaf gene (ll) from
H-19, in order to avoid the deleterious effects of the little leaf type on GYN and EAR
(Fazio et al. 2003b). Little leaf types, however, typically have more branches than
standard leaf types, and the QTL (from H-19) with the greatest effect on MLB (LOD =
32.9, R2 = 32%) is tightly linked (0.7 cM) to ll (Fazio et al. 2003b). Selection of the H-19
59allele at OP-AD12-1, therefore, may have resulted in greater gains in MLB from MAS,
but may, in turn, have negatively affected EAR, GYN, and L:D, which are associated
with the Gy-7 allele.
Genetic distance between QTL and marker, marker inheritance, and marker type
were also considered when choosing markers for MAS. The majority of marker-QTL
associations in this study were < 5.0 cM (Fazio et al. 2003b), and codominant markers
were utilized when available. In regions where marker-QTL associations were wide,
markers flanking the QTL of interest (Edwards and Page 1994) were employed (e.g.,
AK5SCAR and M8SCAR for MLB; Table 1.3). Certain marker types (SCAR, SNP, and
SSR) were chosen over others (RAPD and AFLP) because of their inherent robustness,
ease of use, and ability to be multiplexed (Polashock and Vorsa 2002; Tang et al. 2003;
Mohring et al. 2004; Staub et al. 2004). The majority of the RAPD markers used in MAS
were repeated several times to provide certainty during genotyping. In contrast, all but
one (M8SCAR) of the SNP and SCAR markers could be multiplexed, allowing for
increased genotyping efficiency (Table 1.3). The low repeatability of RAPDs and the
advantage of multiplexing for high-throughput genotyping demonstrate the need for SNP,
SCAR, and SSR markers for MAS in cucumber.
Selection effectiveness Each of the four base populations underwent random mating (RAN) following the
same mating scheme as MAS and PHE to provide four estimates of genetic drift. When
considered over all five traits in each of the four populations, 15 of the 20 slopes were
significant after RAN (Table 1.4; Figure 1.2). The significant changes in trait values of
60L:D and MLB after RAN are most likely due to genetic drift, since regression slopes
were positive, negative, or not significant, depending on the population, and the R2 values
were generally low. In the three instance where trait values increased after RAN (L:D in
Populations 2 and 3 and MLB in Population 3), the similar increase from MAS or PHE
cannot be attributed to selection. In contrast, the regression slopes for EAR, GYN, and
yield were significantly negative in almost every population after RAN, and the R2 values
are relatively high, indicating a small, but steady reduction in these traits inconsistent
with the random nature of genetic drift. A probable explanation for the reduction in GYN
is that gynoecious plants needed to be chemically induced to produce male flowers for
pollination during RAN. The production of male flowers varies in time and quantity
among gynoecious individuals, which introduces flowering time and fecundity
differences, possibly leading to a reproductive disadvantage for gynoecious plants.
Chemical induction of male flowers may also have also affected EAR and yield indirectly
[e.g., correlated responses similar to GYN and EAR after MAS and PHE (Table 1.5)], but
other physiological factors (i.e., source-sink relationships, reduced fitness of higher
yielding individuals) may have also affected these traits. Although trait values were
generally not static in the absence of selection, the general reduction in trait values after
RAN indicates that increases after MAS or PHE can be attributed to a response from
selection.
Both MAS and PHE provided improvements in all traits under selection in at least
one population, except EAR by MAS (Table 1.4, Figure 1.2). Generally, PHE was most
effective for GYN, EAR, and L:D, while MAS was slightly more effective for MLB.
61The similar response to MAS and PHE for MLB confirms the results from Fazio et al.
(2003a) where MAS and PHE equally improved MLB in two cycles of backcross
selection in cucumber. Both PHE and MAS were generally effective at improving
populations with inferior traits, but not as effective at maintaining traits with high values.
Based on trait value changes in response to selection, PHE was more effective than MAS
in Populations 1, 2, and 4, but MAS was slightly more effective than PHE in Population 3.
Thus, the choice of selection methods for cucumber improvement through plant
architectural manipulation (i.e., yield components) will depend upon the populations and
traits under selection.
Yield was not under direct selection in this study, but was evaluated to test the
efficacy of indirect yield improvement by selection for yield components. Yield was
higher in every C0 population than the maternal parent that produced it, except the
highest yielding parent (6823B; Population 2), suggesting the possibility of a heterotic
yield effect in these populations. Cucumber is considered a cross pollinated crop, and
although it exhibits little inbreeding depression, heterosis for yield has been observed in a
number of cases (Wehner 1989). Using the mean of the four parents (1.81) as the mid-
parent value, the mid-parent percent heterosis for yield is 22%, 12%, 2.6%, and 27% for
Populations 1-4, respectively. These values are similar to those reported for fruit number
in previous studies (Wehner 1989). Given this heterotic yield effect, and the difficulty of
simultaneously increasing several yield components, inbred lines with high values for
specific yield components (e.g., GYN with EAR, or MLB with L:D) could be developed
in parallel, and then crossed to create high yielding hybrids. This approach would
62involve the extensive combining ability or test cross evaluation of inbred lines in multiple
environments, and would likely be population specific.
Indirect selection by MAS or PHE was generally not effective at increasing yield
in this study. Nevertheless, the hypothesis that yield increases with the improvement of
all four yield components cannot be rejected, since in no instance did improvement of all
four traits occur. This contention is supported by the results of Fan et al. (2006), who
independently evaluated the effectiveness of MAS for MLB, GYN, and L:D in a crossing
scheme using selections from C2 of PHE in Population 1 of this study as recurrent parents
to produce two backcross populations. In addition to the gains made by PHE for MLB
and L:D after two cycles of selection, MAS continued to improve MLB and L:D in one
backcross population, and L:D in the other, while GYN was improved in both
populations after MAS. These results, coupled with the observation that MAS increased
GYN, L:D, and MLB in this study, confirms the potential value of the marker-QTL
associations for these three traits in these cucumber populations. The challenge to
improving yield in cucumber will likely be the simultaneous improvement of yield
components using both MAS and PHE.
The simultaneous increase in all four traits under selection in this study will be
predictably difficult given the negative correlations among some yield component traits.
The strength and direction of these correlations have been documented in a wide range of
genetic backgrounds (Kupper and Staub 1988; Serquen et al. 1997b; Cramer and Wehner
1998; Cramer and Wehner 1999; Cramer and Wehner 2000b; Fazio et al. 2003b).
Recombination of four inbred lines, recurrent selection, and four different populations
63were used herein to mitigate negative correlations among yield components. These
strategies were generally ineffective, however, because GYN and EAR were positively
correlated as were MLB and L:D in all four populations (Table 1.5). The correlations
among these yield components are most likely due to a combination of pleitropy with the
F, de, and ll genes (Fazio et al. 2003b), and linkage among individual QTL (Robbins and
Staub 2004). Selection by MAS or PHE will not overcome pleitropic effects, but may
identify recombinants between QTL that may help to diminish negative correlations
among yield components. Fine mapping in regions with clustered QTL would assist in
determining the extent of linkage between QTL and identify molecular markers that
could be useful for selecting recombinants between tightly linked QTL (Nam et al. 2005).
Selection methods in breeding programs For MAS to be employed in plant improvement programs, it must provide
resource (cost/benefit) and/or technical (improved effectiveness or efficiency) advantages
over PHE. In this study, the cumulative time required to complete three cycles of MAS
in all four populations, was 19 months as compared to one cycle per year for PHE (Figure
1.3). Populations were evaluated for PHE simultaneously, while they were offset for
MAS such that genotyping usually occurred in one population while other populations
were intermated. The increased efficiency of MAS may, in some cases, be an advantage
over PHE under Wisconsin conditions. For example, the improvement of GYN per year
in Population 4 was fairly similar between MAS (4.9%/cycle × 3 cycles/yr = 14.7%/yr)
and PHE (12.6%/cycle × 1 cycle/yr = 12.6%/yr). The efficiency of MAS could be further
improved by the use of codominant, single-copy markers that can be multiplexed, such as
64SCARs, SNPs, and SSRs in combination with automation technologies (e.g., robotics,
gel-less PCR assays, microarrays, etc.; Gupta et al. 2001; Collard et al. 2005).
Substantial investments required for automation technologies are currently cost
prohibitive for minor crops such as cucumber, but may be mitigated as genomic tools
become more available and affordable.
Recurrent selection is the method of choice for traits with low heritability and has
been used extensively for yield improvement in cucumber (Lower and Edwards 1986;
Wehner 1989; Cramer and Wehner 1998). Two important considerations for recurrent
selection are selection intensity and genetic drift. Selection intensity must be stringent
enough to increase desired alleles (make gain from selection), but modest enough to
allow diversity to continue improvement in subsequent cycles of selection (Casler 1999;
Bernardo 2002). These two factors must be balanced since increasing selection intensity
by decreasing the number of individuals selected increases the effect of drift. The results
from RAN indicate that selecting 20 out of 600 individuals to obtain fairly high selection
intensities results in genetic drift for some traits. Genetic drift may be minimized by
relaxing the selection intensity so that more individuals are intermated (e.g., select 50-75
individuals), which also increases the probability of recombination among tightly linked
QTL. Using this approach, gains from selection per cycle would be expected to be lower,
and thus, increased cycles of selection should be employed. In addition, the finding that
the values of GYN, EAR, and yield were reduced in the absensce of selection indicates
the importance of continual selection for all three traits. An alternative to relaxing the
selection intensity is to increase the number of individuals evaluated and selected
65proportionately, thereby maintaining the selection intensity. The evaluation of 600
individuals in each population was the maximum allowable for each method with the
resources available in this study. However, selecting 600 individuals by both MAS and
PHE in same cycle and intermating 40 selections is possible. Using this approach, high
selection intensities are maintained, but evaluating a greater number of individuals may
allow for recombination among tightly linked QTL, while intermating more individuals
may overcome genetic drift. Thus, selection for improved yield in cucumber may be
most effective by combining both MAS and PHE.
Literature Cited
Bernardo R (2002) Breeding for Quantitative Traits in Plants. 1st edn. Woodbury, Stemma Press
Casler MD (1999) Phenotypic recurrent selection methodology for reducing fiber concentration in smooth bromegrass. Crop Sci 39:381-390
Chee PW, Elias EM, Anderson JA, Kianian SF (2001) Evaluation of a high grain protein QTL from Triticum turgidum L. var. dicoccoides in an adapted durum wheat background. Crop Sci 41:295-301
Collard BCY, Jahufer MZZ, Brouwer JB, Pang ECK (2005) An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142:169-196
Cramer CS, Wehner TC (2000a) Path analysis of the correlation between fruit number and plant traits of cucumber populations. HortScience 35:708-711
Cramer CS, Wehner TC (2000b) Fruit yield and yield component correlations of four pickling cucumber populations. Cucurbit Genet Coop Rpt 23:12-15
Cramer CS, Wehner TC (1999) Little heterosis for yield and yield components in hybrids of six cucumber inbreds. Euphytica 110:99-108
Cramer CS, Wehner TC (1998) Fruit yield and yield component means and correlations of four slicing cucumber populations improved through six to ten cycles of recurrent selection. J Am Soc Hort Sci 123:388-395
66de Oliveira EJ, Alzate-Marin AL, Borem A, Fagundes SD, de Barros EG, Moreira MA (2005) Molecular marker-assisted selection for development of common bean lines resistant to angular leaf spot. Plant Breed 124:572-575
Edwards MD, Page NJ (1994) Evaluation of marker-assisted selection through computer simulation. Theor Appl Genet 88:376-382
Fan Z, Robbins MD, Staub JE (2006) Population development by phenotypic selection with subsequent marker-assisted selection for line extraction in cucumber (Cucumis sativus L.). Theor Appl Genet 112:843-855
Fazio G, Chung SM, Staub JE (2003a) Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.). Theor Appl Genet 107:875-883
Fazio G, Staub JE, Stevens MR (2003b) Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor Appl Genet 107:864-874
Fazio G (2001) Comparative study of marker-assisted and phenotypic selecton and genetic analysis of yield components in cucumber. PhD dissertation University of Wisconsin Madison
Flint-Garcia SA, Darrah LL, McMullen MD, Hibbard BE (2003) Phenotypic versus marker-assisted selection for stalk strength and second-generation European corn borer resistance in maize. Theor Appl Genet 107:1331-1336
Francia E, Tacconi G, Crosatti C, Barabaschi D, Bulgarelli D, Dall'Aglio E, Vale G (2005) Marker assisted selection in crop plants. Plant Cell Tissue Organ Cult 82:317-342
Fredrick LR, Staub JE (1989) Combining ability analyses of fruit yield and quality in near-homozygous lines derived from cucumber. J Am Soc Hort Sci 114:332-338
Gimelfarb A, Lande R (1994a) Simulation of marker assisted selection for nonadditive traits. Genet Res 64:127-136
Gimelfarb A, Lande R (1994b) Simulation of marker assisted selection in hybrid populations. Genet Res 63:39-47
Gupta PK, Roy JK, Prasad M (2001) Single nucleotide polymorphisms: A new paradigm for molecular marker technology and DNA polymorphism detection with emphasis on their use in plants. Curr Sci 80:524-535
Hoeck JA, Fehr WR, Shoemaker RC, Welke GA, Johnson SL, Cianzio SR (2003) Molecular marker analysis of seed size in soybean. Crop Sci 43:68-74
67Kandemir N, Jones BL, Wesenberg DM, Ullrich SE, Kleinhofs A (2000) Marker-assisted analysis of three grain yield QTL in barley (Hordeum vulgare L.) using near isogenic lines. Mol Breed 6:157-167
Kupper RS, Staub JE (1988) Combining ability between lines of Cucumis sativus L. and Cucumis sativus var. hardwickii (R.) Alef. Euphytica 38:197-210
Lande R, Thompson R (1990) Efficiency of marker-assisted selection in the improvement of quantitative traits. Genetics 124:743-756
Mohring S, Salamini F, Schneider K (2004) Multiplexed, linkage group-specific SNP marker sets for rapid genetic mapping and fingerprinting of sugar beet (Beta vulgaris L.). Mol Breed 14:475-488
Moreau L, Charcosset A, Gallais A (2004) Experimental evaluation of several cycles of marker-assisted selection in maize. Euphytica 137:111-118
Nam YW, Lee JR, Song KH, Lee MK, Robbins MD, Chung SM, Staub JE, Zhang HB (2005) Construction of two BAC libraries from cucumber (Cucumis sativus L.) and identification of clones linked to yield component quantitative trait loci. Theor Appl Genet 111:150-161
Nijs APMD, Visser DL (1980) Induction of male flowering in gynoecious cucumbers (Cucumis sativus L.) by silver ions. Euphytica 29:273-280
Polashock JJ, Vorsa N (2002) Development of SCAR markers for DNA fingerprinting and germplasm analysis of American cranberry. J Am Soc Hort Sci 127:677-684
Reyna N, Sneller CH (2001) Evaluation of marker-assisted introgression of yield QTL alleles into adapted soybean. Crop Sci 41:1317-1321
Robbins MD, Staub JE (2004) Strategies for selection of multiple, quantitatively inherited yield components in cucumber. In: Lebeda A, Paris HS (eds) Progress in Cucurbit Genetics and Breeding Research. Proceedings of Cucurbitaceae 2004, the 8th EUCARPIA Meeting on Cucurbit Genetics and Breeding. Olomouc, Czech Republic, Palacky University, pp 401-408
Robbins MD, Staub JE, Fazio G (2002) Deployment of molecular markers for multi-trait selection in cucumber. In: Cucurbitaceae 2002. Alexandria, Amer. Soc. Hort. Sci. Press, pp 41-47
68Romagosa I, Han F, Ullrich SE, Hayes PM, Wesenberg DM (1999) Verification of yield QTL through realized molecular marker-assisted selection responses in a barley cross. Mol Breed 5:143-152
Schneider KA, Brothers ME, Kelly JD (1997) Marker-assisted selection to improve drought resistance in common bean. Crop Sci 37:51-60
Serquen FC, Bacher J, Staub JE (1997a) Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random-amplified polymorphic DNA markers. Mol Breed 3:257-268
Serquen FC, Bacher J, Staub JE (1997b) Genetic analysis of yield components in cucumber at low plant density. J Am Soc Hort Sci 122:522-528
Shen L, Courtois B, McNally KL, Robin S, Li Z (2001) Evaluation of near-isogenic lines of rice introgressed with QTLs for root depth through marker-aided selection. Theor Appl Genet 103:75-83
Shetty NV, Wehner TC (2002) Screening the cucumber germplasm collection for fruit yield and quality. Crop Sci 42:2174-2183
Staub JE, Robbins MD, Chung S, López-Sesé AI (2004) Molecular methodologies for improved genetic diversity assessment in cucumber and melon. Acta Hort 637:41-47
Staub JE, Crubaugh LK, Fazio G (2002) Cucumber recombinant inbred lines. Cucurbit Genet Coop Rpt 25:1-2
Staub JE, Crubaugh LK (2001) Cucumber inbred line USDA 6632E. Cucurbit Genet Coop Rpt 24:6-7
Staub JE, Knerr LD, Hopen HJ (1992) Plant density and herbicides affect cucumber productivity. J Am Soc Hort Sci 117:48-53
Steele RGD, Torrie JH, Dickey DA (1996) Principles and Procedure in Statistics. 3rd edn. New York, McGraw-Hill
Stromberg LD, Dudley JW, Rufener GK (1994) Comparing conventional early generation selection with molecular marker assisted selection in maize. Crop Sci 34:1221-1225
Tang SX, Kishore VK, Knapp SJ (2003) PCR-multiplexes for a genome-wide framework of simple sequence repeat marker loci in cultivated sunflower. Theor Appl Genet 107:6-19
69Van Berloo R, Stam P (1999) Comparison between marker-assisted selection and phenotypical selection in a set of Arabidopsis thaliana recombinant inbred lines. Theor Appl Genet 98:113-118
Wehner TC (1989) Breeding for improved yield in cucumber. Plant Breed Rev 6:323-359
Wehner TC, Lower RL, Staub JE, Tolla GE (1989) Convergent-divergent selection for cucumber fruit yield. HortScience 24:667-669
Willcox MC, Khairallah MM, Bergvinson D, Crossa J, Deutsch JA, Edmeades GO, Gonalez-de-Leon D, Jiang C, Jewell DC, Mihm JA, Williams WP, Hoisington D (2002) Selection for resistance to southwestern corn borer using marker-assisted and conventional backcrossing. Crop Sci 42:1516-1528
Yousef GG, Juvik JA (2001) Comparison of phenotypic and marker-assisted selection for quantitative traits in sweet corn. Crop Sci 41:645-655
Zhang C, Tar'an B, Warkentin T, Tullu A, Bett KE, Vandenberg B, Somers DJ (2006) Selection for lodging resistance in early generations of field pea by molecular markers. Crop Sci 46:321-329
Zhang W, Smith C (1992) Computer simulation of marker-assisted selection utilizing linkage disequilibrium. Theor Appl Genet 83:813-820
70 Table 1.1. Mean values of yield component traits of commercial checks and parental inbred lines used to create four cucumber populations for comparison of response to selection by phenotype and marker. Means are from the replicated trial described herein.
Inbred line/Check Population EARa GYNb L:Dc MLBd Yielde Leaf typef
6632E 1 2.47 99.1 2.57 2.1 2.04 Standard 6823B 2 0.76 5.7 3.51 4.5 2.11 Little 6996A 3 1.84 99.2 2.71 0.9 1.50 Standard 6995C 4 0.43 10.6 3.01 3.1 1.60 Standard Gy-7 (check) line 2.13 99.8 2.74 1.0 1.70 Standard H-19 (check) line 0.42 6.1 3.03 5.7 1.84 Little ‘Vlasset’g (check) hybrid 1.98 77.9 2.70 3.0 2.28 Standard a Earliness measured as the number of fruits per plant in first harvest b Gynoecy measured as the percent female flowers in the first ten nodes c Fruit length to diameter ratio measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests d Multiple lateral branching measured as the number of lateral branches (at least three internodes long) on the mainstem in the first 10 nodes e Yield measured as the noumber of fruits per plant averaged over four harvests f Leaf type classified as Standard (> 40 cm2) or Little leaf (30-40 cm2; Staub et al. 1992) g Commercial cultivar from Seminis Vegetable Seeds, Inc, Oxnard, Calif.
71Table 1.2. Cumulative selection differential over three cycles between Stage 1 and Stage 2 of phenotypic selection (PHE) for four traits in four populations of cucumber.
Traita Population Selection differentialb P-valuec Percentd
a Traits are EAR = earliness measured as the number of days to anthesis of the first female flower, GYN = gynoecy measured as the percentage of plants classified as gynoecious (100% female flowers in the first 10 nodes), MLB = multiple lateral branching measured as the number of lateral branches (at least three internodes long) on the mainstem in the first 10 nodes, and L:D = fruit length to diameter ratio measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests, b Sum over three cycles of the difference between the mean of Stage 2 selections and the mean of the selections from Stage 1 at each cycle. The harmonic mean number of individuals selected at Stage 1 over three cycles was 125.2, 128.9, 114.2, and 130.1 for Populations 1, 2, 3, and 4, respectively. c P-value of a t-test for the significance ot the selection differential d Selection differential expressed as the percent of the mean of Stage 1 selections
72
Table 1.3. Characteristics of molecular markers defined in a genetic map of cucumber constructed by Fazio et al. (2003b) and used in marker-assisted selection for population improvement.
Marker TypeaLinkage group
Map position (cM) Parentb
Multiplex groupc Ideotype QTL (mapping parent and LOD score) and gene associations d
CSWCT28 SSR 1 5.0 G&H G&H EAR(G, 7.1), MLB(H, 10.4), GYN(G, 13.0), L:D(H, 5.7), F L18-SNP-H19 SNP 1 7.4 H 1 H EAR(G, 7.1), MLB(H, 10.4), GYN(G, 13.0), L:D(H, 5.7) OP-AG1-1 RAPD 1 31.8 G H EAR(G, 6.4), MLB(H, 11.6), GYN(G, 7.3), de AJ6SCAR SCAR 1 61.4 G 3 H MLB(H, 3.3) BC523SCAR SCAR 1 66.5 G 2 H MLB(H, 3.3) OP-AD12-1 RAPD 1 70.2 H G EAR(G, 4.1), MLB(H, 32.9), GYN(G, 3.7), L:D(G, 8.6), ll AW14SCAR SCAR 3 3.9 G&H 1 G GYN(G, 5.1) CSWTAAA01 SSR 4 34.1 G&H 2 H MLB(H, 4.6) OP-AI4 RAPD 5 101.0 G G GYN(G, 3.0) OP-AO12 RAPD 5 117.3 G G GYN(G, 3.0) OP-AI10 RAPD 6 22.5 H G L:D(G, 7.3) AK5SCAR SCAR 6 33.0 G 2 H MLB(H, 3.0) M8SCAR SCAR 6 39.1 H H MLB(H, 3.0) OP-W7-1 RAPD 6 83.4 H G GYN(G, 4.1) L19-2-SCAR SCAR 6 115.0 H 1 G MLB(G, 4.2), GYN(G, 4.1) NR60 SSR 6 137.4 G&H G MLB(G, 4.2) BC515 RAPD 7 0.0 H H L:D(H, 4.2) L19-1-SCAR SCAR 7 9.9 H 3 H L:D(H, 4.2) a SSR simple sequence repeat, SNP single nucleotide polymorphism, RAPD random amplified polymorphic DNA, and SCAR sequence characterized amplified region
b Allelic constitution based on mapping parents H-19 and Gy-7 (synom. G421) (Fazio et al. 2003b), where G = present in Gy-7, H = present in H-19, G&H = present in Gy-7 and H-19 (codominant marker) c Markers used in multiplex were placed in multiplexing groups (1, 2, or 3) d Markers associated with QTL for DTF = earliness, MLB = multiple lateral branching, GYN = gynoecious, and L:D = length to diameter ratio. The parentheses contain the parent contributing the QTL (G = Gy-7, H = H-19) followed by the highest LOD score for each QTL obtained from multiple field trials (Serquen et al. 1997a; Fazio et al. 2003b). Genes are F = femaleness, de = determinate, and ll = little leaf
73Table 1.4. Means and linear response of five traits in four base populations (C0) of cucumber which underwent three cycles of recurrent mass selection (C1-C3) using three methods.
a Traits are EAR = earliness measured as the number of fruits per plant in first harvest, GYN = gynoecy measured as the percent female flowers in the first ten nodes, L:D = fruit length to diameter ratio measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests, MLB = multiple lateral branching measured as the number of lateral branches (at least three internodes long) on the mainstem in the first 10 nodes, and Yield measured as the number of fruits per plant averaged over four harvests b Methods are MAS = selection by marker, PHE = phenotypic selection, and RAN = random mating (no selection) c Slope of linear regression of means over cycles d P-values from F-tests of linear response to selection e Populations were created by intermating four inbred lines, and then bulking by the maternal parent (Figure 1.1)
75
Table 1.5. Phenotypic correlations (r) among traits in cucumber over three cycles of selection by markers (MAS) and phenotype (PHE). Selection by markers (MAS) Selection by phenotype (PHE) Traita EAR GYN L:D MLB Traita EAR GYN L:D MLB Population 1b Population 1 GYN 0.70*** GYN 0.35** L:D -0.26* -0.57*** L:D 0.19 -0.33** MLB -0.53*** -0.53*** 0.13 MLB -0.29* -0.31* 0.36** Yield 0.56*** 0.47*** -0.18 -0.04 Yield 0.25* 0.04 0.17 -0.02 Population 2 Population 2 GYN 0.37** GYN 0.30* L:D 0.21 -0.44*** L:D 0.01 -0.20 MLB -0.54*** -0.43*** 0.14 MLB -0.35** -0.32** 0.38** Yield 0.54*** 0.17 0.34** -0.07 Yield 0.56*** -0.01 0.28* 0.12 Population 3 Population 3 GYN 0.30* GYN 0.45*** L:D 0.30* -0.26* L:D -0.11 -0.07 MLB -0.38** -0.37** 0.06 MLB -0.26* -0.14 0.31* Yield 0.54*** 0.12 -0.01 -0.17 Yield 0.55*** 0.16 -0.03 -0.08 Population 4 Population 4 GYN 0.37** GYN 0.42*** L:D 0.03 -0.53*** L:D 0.07 -0.64*** MLB -0.46*** -0.41*** 0.32** MLB -0.43*** -0.42*** 0.23 Yield 0.39** 0.08 0.32** 0.27* Yield 0.49*** -0.12 0.32** 0.13
a Traits are EAR = earliness measured as the number of fruits per plant in first harvest, GYN = gynoecy measured as the percent female flowers in the first ten nodes, L:D = fruit length to diameter ratio measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests, MLB = multiple lateral branching measured as the number of lateral branches (at least three internodes long) on the mainstem in the first 10 nodes, and Yield measured as the number of fruits per plant averaged over four harvests b Populations were created by intermating four inbred lines, and then bulking by the maternal parent (Figure 1.1) * = P ≤ 0.05, ** = P ≤ 0.01, *** = P ≤ 0.001
76
6632E(Inbred 1)
Cross with2,3,4
Pop. 1 C0
6823B(Inbred 2)
6632E(Inbred 1)
Cross with1,3,4
Pop. 2 C0
6996A(Inbred 3)
Cross with1,2,4
Pop. 3 C0
6995C(Inbred 4)
Cross with1,2,3
Pop. 4 C0
Cycle 1
Replicated Trial
MASPHE RAN MASPHE RAN MASPHE RAN MASPHE RAN
MASPHE RANCycle 2
MASPHE RANCycle 3
MASPHE RAN
MASPHE RAN
MASPHE RAN
MASPHE RAN
MASPHE RAN
MASPHE RAN
Cross with2,3,4
Pop. 1 C0
6823B(Inbred 2)
Cross with1,3,4
Pop. 2 C0
6996A(Inbred 3)
Cross with1,2,4
Pop. 3 C0
6995C(Inbred 4)
Cross with1,2,3
Pop. 4 C0
Cycle 1
Replicated Trial
MASPHE RAN MASPHE RAN MASPHE RAN MASPHE RAN
MASPHE RAN MASPHE RAN MASPHE RAN MASPHE RANCycle 2
MASPHE RANCycle 3
MASPHE RAN MASPHE RAN MASPHE RAN
Figure 1.1. Schematic of the selection and evaluation scheme in cucumber where PHE is phenotypic selection, MAS is selection by marker, and RAN is random mating.
Figure 1.2. Response to selection as measured by the slope of linear regression (Y axes) over three cycles of MAS (selection by marker), PHE (phenotypic selection) or RAN (random mating) for five traits in four cucumber populations. *, **, and *** denote slopes are significant at P ≤ 0.05, P ≤ 0.01, and P ≤ 0.001, respectively.
78
Figure 1.3. Time required to complete MAS (selection by marker) and PHE (phenotypic selection) for three cycles in four cucumber populations. Gray areas indicate evaluation of the populations and black areas represent recombination among selections.
79
Chapter 2. The development of molecular markers with increased efficacy for genetic analysis in cucumber
Abstract
The genetic base of cucumber (Cucumis sativus L.; 2n=2x=14) is narrow, thus the
recovery of useful markers from genetic analyses is typically low. Although the largely
RAPD-based genetic linkage maps of cucumber have proven effective for marker-
assisted selection (MAS), marker genotyping is inefficient. Conversion of RAPD to
SCAR markers has not been extremely successful in cucumber and, thus, methods were
developed to increase conversion of these RAPDs into more efficient and effective SCAR
and SNP markers. Forty-three dominant RAPD bands were sequenced to create 22
polymorphic (17 dominant, five codominant) SCAR markers. Evaluation of three
multiplexing sets revealed that the optimal number of SCARs that could be combined
during PCR was four to five. Twenty-three monomorphic or dominant SCARs were
sequenced to identify SNPs. Additional sequences were obtained using the amplicons of
six dominant markers (five SCAR and one RAPD) linked to important yield traits to
probe a cucumber BAC library, that led to the sequencing of a subset of positive clones.
Four different approaches were utilized to create allele-specific markers based on SNPs
for 20 (19 codominant, one dominant) of the 25 loci that contained SNPs. Of the 43
initial RAPDs, 32 were converted into SCAR or SNP markers, six of which have proven
effective in MAS, while others provide a starting point for molecular investigation of
several QTL associated with yield-related traits in cucumber.
80Introduction
Molecular markers are becoming proven, valuable tools for the improvement of
many crop species (Collard et al. 2005; Francia et al. 2005). Markers have been
employed in plant breeding programs for genetic diversity assessment, cultivar identity,
genetic similarity estimation, fingerprinting, genetic map construction, gene tagging, and
marker-assisted selection (MAS; Collard et al. 2005). The ideal markers for use in plant
breeding programs are codominant, easily adapted, robust (repeatable), abundant, and
amenable to low cost, high-throughput systems (Staub et al. 1996; Mohan et al. 1997;
Gupta et al. 1999; Collard et al. 2005).
In cucumber (Cucumis sativus L.; 2n=2x=14), a moderately saturated genetic
linkage map has been constructed (Fazio et al. 2003b) that defined economically
important marker-trait associations which have been useful in MAS. Gain from selection
using MAS has been demonstrated for quantitative traits (multiple lateral branching,
gynoecious sex expression, and fruit length to diameter ratio) in backcross breeding
(Fazio et al. 2003a; Fan et al. 2006) and in recurrent selection (Chapter 1).
Approximately 70% of the 131 markers on the Fazio et al. (2003b) map, however, are
random amplified polymorphic DNA (RAPD) or amplified fragment length
polymorphism (AFLP) markers, and roughly 80% of all markers are dominant. Although
AFLPs and RAPDs are well suited for relatively rapid identification of new markers, the
low reproducibility of RAPDs and the technical methodology of AFLPs as well as their
dominant nature make them less efficient than more robust sequence characterized
amplified region (SCAR), and codominant simple sequence repeat (SSR) and single
81nucleotide polymorphism (SNP) based markers, which are better suited for tracking
alleles during MAS (Brugmans et al. 2003; Shirasawa et al. 2004; Collard et al. 2005).
The historically low rate of new marker identification due to the narrow genetic base of
cucumber (3-8% polymorphism among adapted cultivars for any given marker; Knerr et
al. 1989; Dijkhuizen et al. 1996; Horejsi and Staub 1999; Fazio et al. 2002) suggests that
existing resources should be used to convert mapped RAPD markers to more efficient
markers for use in MAS.
SCAR markers were initially designed to convert polymorphic RAPD markers
into robust, single-copy markers by extending the original 10 bp RAPD primer an
additional 14 bases pairs on the 3’ end based on the internal sequence of both ends of the
RAPD fragment (Paran and Michelmore 1993). Although the SCAR marker may retain
the original RAPD polymorphism in some cases, lengthening RAPD primers to increase
PCR specificity may mask the original RAPD polymorphism in others. If the latter
occurs, raising the PCR annealing temperature may reveal polymorphisms that are not
apparent under standard PCR conditions (Paran and Michelmore 1993; Horejsi et al.
1999). Additionally, both alleles of the SCAR marker may be sequenced to identify
polymorphisms for exploitation in genotypic analysis. Although initially created from
RAPDs, SCAR markers can be created from any genomic sequence by designing specific
primer pairs for PCR (Paran and Michelmore 1993).
Codominant SNP-based markers are particularly attractive for use in MAS
because they are robust, abundant, and can be detected by several methods, including low
cost methods using basic laboratory equipment (Gupta et al. 2001). Cleaved amplified
82polymorphic sequence (CAPS; Konieczny and Ausubel 1993), allele-specific PCR
(AS-PCR; Newton et al. 1989; Sarkar et al. 1990) and single-nucleotide amplified
polymorphism (SNAP) markers (Drenkard et al. 2000) are all based on SNPs and are
detected by PCR and agarose gel electrophoresis. The advantage of the AS-PCR or
SNAP markers, however, is that they do not require a restriction enzyme step after PCR,
which is essential for CAPS markers.
The limited success of a previous RAPD to SCAR conversion study in cucumber
using silver staining technology for sequencing (Horejsi et al. 1999) compared to that of
Paran and Michelmore (1993), prompted a reevaluation of the conversion of existing
RAPDs into more efficient markers using current sequencing methods (automated
fluorescent sequencing). This was accomplished by: 1) converting several RAPDs into
polymorphic SCAR markers; 2) determining if the new SCAR markers could be
multiplexed to increase their efficiency; 3) creating codominant markers based on SNPs
from sequences of monomorphic or dominant SCARs (a) and specific BAC clones (b);
and 4) evaluating allele-specific SNP markers from both sequence sources. The
development, optimization, and evaluation of these markers for PCR will allow for
increased MAS efficiency during plant improvement in cucumber.
Materials and Methods
RAPD to SCAR conversion Forty-three RAPD markers mapped by Serquen et al. (1997) and Fazio et al.
(2003b) were selected for conversion to SCAR markers (Objective 1; Table 2.1).
83Parental lines of these linkage maps, Gy-7 (synon. G421) and H-19, were used as
templates for amplification of RAPD markers employing PCR conditions used by
Serquen et al. (1997). The PCR products were resolved on agarose (1.6%) gels, and
polymorphic bands were excised and purified with the QIAquick gel extraction kit
(QIAGEN Inc., Valencia, CA) following the manufacturer’s protocol. The resulting
fragments were subsequently cloned into the pGEM®-T vector (Promega Corporation,
Madison, Wisc.) following manufacturer’s instructions, and single colonies were picked
and used to inoculate 3 mL of LB broth with ampicillin (100μg/mL) for overnight
incubation at 37o C. Plasmids were then isolated using the QIAprep spin miniprep kit
(QIAGEN Inc., Valencia, CA) according to the manufacturer’s protocol, and the inserts
were amplified in preparation for sequencing using M13 forward (5’-CGC CAG GGT
Mohring et al. 2004), common bean (Phaseolus vulgaris L.; Masi et al. 2003),
Arabidopsis (Torjek et al. 2003), rice (Oryza sativa L.; Blair et al. 2002), sunflower
(Helianthus annuus L.; Tang et al. 2003), and Brassica spp. (Mitchell et al. 1997).
Annealing temperature is one of the most important parameters in multiplexing
(Henegariu et al. 1997). Therefore, determining the annealing temperature range of
SCAR markers by ATG-PCR was necessary for identifying markers that could be
100successfully multiplexed. The multiplex evaluations performed herein were designed
to optimize the use of SCAR markers for genotyping. Although multiplexing was not
tested over all possible PCR conditions (i.e., PCR reagent concentrations and thermal
cycling conditions other than annealing temperature), the evaluations conducted provided
for critical insights, which subsequently led to the design of multiplexing primer sets. All
markers (up to five) were scoreable within all three multiplexing sets at an optimal
annealing temperature. These results are similar to those of Polashock and Vorsa (2002)
who found the optimal SCAR multiplex marker number was four to five in cranberry.
Optimization of multiplex PCR conditions proved useful for increasing the efficiency of
MAS in cucumber (Fan et al. 2006; Chapter 1), and provides guidelines for the
identification of potential multiplex groups where large-scale genotyping is demanded.
However, since the interactions of markers are typically unknown and the multiplexing
potential of primer sets is unpredictable (Polashock and Vorsa 2002), empirical testing of
potential multiplex groups is essential.
Identification of SNPs SCAR-based multiplexing provides the ability to evaluate the equivalent of four
to five RAPD markers simultaneously. Not all RAPD markers were converted to
polymorphic SCARs, however, and the majority of polymorphic SCARs were dominant.
Two sources of sequence data [SCAR markers (Objective 3a) and BAC ends (Objective
3b)] were utilized, therefore, to create more stable and efficient RAPD marker
replacements. Sequencing SCAR markers provided a marker at the same physical
location as its RAPD counterpart. Sequencing BAC ends, however, created a set of
101markers physically near the RAPD locus linked to yield component QTL that could
then be utilized for fine mapping or map-based cloning (Fazio et al. 2003b; Nam et al.
2005), and to identify SNPs where no polymorphism was detected by a SCAR and no
internal sequence polymorphism exists.
The two sources of sequences to identify SNPs (SCAR markers and BAC ends)
provided contrasting SNP frequency results. The percentage of sequences that contained
at least one SNP was higher from SCAR markers (63%) than from the BAC end
sequences (35%). This may be explained by the fact that the average number of base
pairs sequenced per locus was much higher in the SCARs (838 bp/locus) than in the BAC
end sequences (427 bp/locus), which increases the chance of identifying a SNP. The
SNP frequency was lower in the SCARs (one SNP every 210 bp) than in the BAC end
sequences (one SNP every 173 bp). However, the large number of SNPs detected in C-3-
BE-L (31 SNPs) and C6-BE-L (24 SNPs) likely inflated the average SNP frequency in
the BAC end sequences such that the true average SNP frequency is most likely about
one SNP every 210 bp. In coding regions of melon (Cucumis melo L.), SNPs and indels
were identified at an average frequency of one SNP every 441 bp and one indel every
1,666 bp (Morales et al. 2004). In 22 diverse soybean genotypes, SNPs were identified at
a frequency of one SNP every 503 bp in coding regions and one SNP every 237 bp in
non-coding regions (Zhu et al. 2003). Likewise, the SNP frequency in elite maize (Zea
mays) lines was estimated at one SNP per 31 bp (non-coding regions) or 124 bp (coding
regions; Ching et al. 2002). In contrast, SNP frequencies of tomato, have been
determined to be one SNP every 7,000-9,500 bp in coding regions (Nesbitt and Tanksley
1022002; Yang et al. 2004; Labate and Baldo 2005). The SNP frequency identified
herein (non-coding regions) is much greater than tomato, and is similar to that detected in
soybean and to what may be expected for non-coding regions of melon. Lines Gy-7 and
H-19 are two of the most genetically diverse accessions known in processing cucumber
(Horejsi and Staub 1999), and thus, SNP frequencies in other germplasm would be
predictably lower. Nevertheless, the abundance of SNPs in cucumber provides an
opportunity to create a large number of new markers utilizing the methodologies
presented herein. The fact that a SNP marker was successfully created for 20 of the 25
sequences that contained a SNP indicates that the recovery rate of SNP-based markers
will be greater than that previously reported in the same germplasm for RAPDs (4.8%;
Serquen et al. 1997) and SSRs (7.2%; Fazio et al. 2002).
SNP marker creation and evaluation SNP-based, allele-specific primers were designed such that the final 3’ nucleotide
matched only one allele (Gy-7 or H-19), but an additional mismatch to both alleles was
included within the last four nucleotides to increase allele specificity in PCR (Newton et
al. 1989; Sarkar et al. 1990; Drenkard et al. 2000; Figure 2.2). Because not all internal
mismatches provide the same level of specificity, Drenkard et al. (2000) has
recommended testing four allele-specific primers at each SNP to obtain a polymorphic
marker. Indeed, 21% of the SNP-based primers created herein were not allele-specific.
However, the fact that an allele-specific marker was recovered for the Gy-7 and H-19
allele in 87.5% and 80.0% of loci attempted (Table 2.3), respectively, indicates that three
allele-specific primers for both Gy-7 and H-19 were usually sufficient. For those
103sequences where allele-specific markers were not identified, additional allele-specific
primers could be tested, which would likely produce a suitable marker (Drenkard et al.
2000). The high recovery rate of allele-specific markers, coupled with the finding that
71% of the SNP-based markers were allele specific, demonstrates that the 3’ and
additional, internal mismatches are highly, but not universally, effective at distinguishing
alleles. Thus, when designing allele-specific, SNP-based markers by methodologies
described herein, only a few (3-4) allele-specific primers are required to detect most
SNPs, while others may require designing additional allele-specific primers.
Four different design approaches to create allele-specific markers (Figure 2.3)
were utilized since none of the individual approaches were applicable in all cases due to
the variability in location (sequence surrounding SNP for primer design, and proximity to
other SNPs) and number of SNPs in a given sequence. When considered together,
however, the four different approaches provided allele-specific markers for almost any
SNP. In only one case was an allele-specific marker not created from a SNP-containing
sequence (Gy-7 allele at W-2-BE-R) because the AT content of the sequence was too
high to design a primer flanking the SNP. Thus, the only main requirements of the four
design approaches are that sequences on at least one side of the SNP need to be amenable
to primer annealing, and sufficient sequence flanking the SNP is needed to create a non-
specific primer.
The optimal marker design approach was utilized for the majority of loci (17 of
25; Table 2.3) because they contained more than one SNP in close proximity to each
other. This approach was preferred over the other three approaches from a marker design
104standpoint since the allele-specific primers for Gy-7 and H-19 were anchored to two
different genomic DNA sites (in contrast to the tail approach) to reduce primer site
competition and both allele-specific markers were designed to utilize the same non-
specific (universal) primer. Thus, only three primers are required to detect both alleles
(i.e., a codominant marker) when both markers are combined in PCR, which reduces the
complexity of the reaction (in contrast to the sandwich and opposite approaches). The
optimal approach was effective and produced a codominant marker in 14 of 17 cases
(Table 2.3). All three of the markers produced by the sandwich approach were
codominant, but they could not be combined in a single assay. This was most likely due
to the increased PCR complexity of combining four primers in a single reaction.
Codominant markers were also produced by both the tail and opposite approaches. The
creation of 19 codominant markers from 25 sequences that contained SNPs demonstrates
the effectiveness of these four approaches, which are predicted to produce similar results
in any crop species because of their applicability to sequences with one or more SNPs in
any position.
The conversion of RAPDs to more efficient and effective SCAR and SNP
markers was highly successful. Compared to a previous RAPD to SCAR conversion
study, the results reported herein illustrate the importance of: 1) including the RAPD
primer sequence in the SCAR primers to increase the recovery of polymorphisms; 2)
creating SCAR primers of sufficient length (24 bp) and utilizing ATG-PCR to increase
marker specificity (recovering single bands); and 3) verifying the new SCAR markers in
a segregating population. The methodologies utilized to create SNP markers to detect
105two alleles at a locus in one or two assays by relatively simple and inexpensive
procedures (conventional PCR and agarose gel electrophoresis), should be applicable to a
diverse array of crop species. In addition, BAC libraries can be used as a tool to obtain
sequence physically proximal to a RAPD marker from which SNPs can be identified.
Markers derived from BAC clones also provide a potential starting point for fine
mapping and map-based cloning of yield component QTL.
Literature Cited
Blair MW, Hedetale V, McCouch SR (2002) Fluorescent-labeled microsatellite panels useful for detecting allelic diversity in cultivated rice (Oryza sativa L.). Theor Appl Genet 105:449-457
Brugmans B, van der Hulst RGM, Visser RGF, Lindhout P, van Eck HJ (2003) A new and versatile method for the successful conversion of AFLP markers into simple single locus markers. Nucleic Acids Res 31:e55
Ching A, Caldwell KS, Jung M, Dolan M, Smith OS, Tingey S, Morgante M, Rafalski AJ (2002) SNP frequency, haplotype structure and linkage disequilibrium in elite maize inbred lines. BMC Genet 3:19
Collard BCY, Jahufer MZZ, Brouwer JB, Pang ECK (2005) An introduction to markers, quantitative trait loci (QTL) mapping and marker-assisted selection for crop improvement: The basic concepts. Euphytica 142:169-196
Dijkhuizen A, Kennard WC, Havey MJ, Staub JE (1996) RFLP variation and genetic relationships in cultivated cucumber. Euphytica 90:79-87
Drenkard E, Richter BG, Rozen S, Stutius LM, Angell NA, Mindrinos M, Cho RJ, Oefner PJ, Davis RW, Ausubel FM (2000) A simple procedure for the analysis of single nucleotide polymorphisms facilitates map-based cloning in Arabidopsis. Plant Physiol 124:1483-1492
Fan Z, Robbins MD, Staub JE (2006) Population development by phenotypic selection with subsequent marker-assisted selection for line extraction in cucumber (Cucumis sativus L.). Theor Appl Genet 112:843-855
106Fazio G, Chung SM, Staub JE (2003a) Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.). Theor Appl Genet 107:875-883
Fazio G, Staub JE, Stevens MR (2003b) Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor Appl Genet 107:864-874
Fazio G, Staub JE, Chung SM (2002) Development and characterization of PCR markers in cucumber. J Am Soc Hort Sci 127:545-557
Francia E, Tacconi G, Crosatti C, Barabaschi D, Bulgarelli D, Dall'Aglio E, Vale G (2005) Marker assisted selection in crop plants. Plant Cell Tissue Organ Cult 82:317-342
Frey JE, Frey B, Sauer C, Kellerhals M (2004) Efficient low-cost DNA extraction and multiplex fluorescent PCR method for marker-assisted selection in breeding. Plant Breed 123:554-557
Gupta PK, Roy JK, Prasad M (2001) Single nucleotide polymorphisms: A new paradigm for molecular marker technology and DNA polymorphism detection with emphasis on their use in plants. Curr Sci 80:524-535
Gupta PK, Varshney RK, Sharma PC, Ramesh B (1999) Molecular markers and their applications in wheat breeding. Plant Breed 118:369-390
Henegariu O, Heerema NA, Dlouhy SR, Vance GH, Vogt PH (1997) Multiplex PCR: Critical parameters and step-by-step protocol. BioTechniques 23:504-511
Horejsi T, Box JM, Staub JE (1999) Efficiency of randomly amplified polymorphic DNA to sequence characterized amplified region marker conversion and their comparative polymerase chain reaction sensitivity in cucumber. J Am Soc Hort Sci 124:128-135
Horejsi T, Staub JE (1999) Genetic variation in cucumber (Cucumis sativus L.) as assessed by random amplified polymorphic DNA. Genet Resour Crop Evol 46:337-350
Knerr LD, Staub JE, Holder DJ, May BP (1989) Genetic diversity in Cucumis sativus L. assessed by variation at 18 allozyme coding loci. Theor Appl Genet 78:119-128
Konieczny A, Ausubel FM (1993) A procedure for mapping Arabidopsis mutations using codominant ecotype-specific PCR-based markers. Plant J 4:403-410
Kwok S, Kellogg DE, Mckinney N, Spasic D, Goda L, Levenson C, Sninsky JJ (1990) Effects of primer template mismatches on the polymerase chain reaction: Human immunodeficiency virus type 1 model studies. Nucleic Acids Res 18:999-1005
107Labate JA, Baldo AM (2005) Tomato SNP discovery by EST mining and resequencing. Mol Breed 16:343-349
Masi P, Zeuli PLS, Donini P (2003) Development and analysis of multiplex microsatellite markers sets in common bean (Phaseolus vulgaris L.). Mol Breed 11:303-313
Merdinoglu D, Butterlin G, Bevilacqua L, Chiquet V, Adam-Blondon AF, Decroocq S (2005) Development and characterization of a large set of microsatellite markers in grapevine (Vitis vinifera L.) suitable for multiplex PCR. Mol Breed 15:349-366
Mitchell SE, Kresovich S, Jester CA, Hernandez CJ, SzewcMcFadden AK (1997) Application of multiplex PCR and fluorescence-based, semi-automated allele sizing technology for genotyping plant genetic resources. Crop Sci 37:617-624
Mohan M, Nair S, Bhagwat A, Krishna TG, Yano M, Bhatia CR, Sasaki T (1997) Genome mapping, molecular markers and marker-assisted selection in crop plants. Mol Breed 3:87-103
Mohring S, Salamini F, Schneider K (2004) Multiplexed, linkage group-specific SNP marker sets for rapid genetic mapping and fingerprinting of sugar beet (Beta vulgaris L.). Mol Breed 14:475-488
Morales M, Roig E, Monforte AJ, Arus P, Garcia-Mas J (2004) Single-nucleotide polymorphisms detected in expressed sequence tags of melon (Cucumis melo L.). Genome 47:352-360
Moreno-Vazquez S, Ochoa OE, Faber N, Chao SM, Jacobs JME, Malsonneuve B, Kesseli RV, Michelmore RW (2003) SNP-based codominant markers for a recessive gene conferring resistance to corky root rot (Rhizomonas suberifaciens) in lettuce (Lactuca sativa). Genome 46:1059-1069
Nam YW, Lee JR, Song KH, Lee MK, Robbins MD, Chung SM, Staub JE, Zhang HB (2005) Construction of two BAC libraries from cucumber (Cucumis sativus L.) and identification of clones linked to yield component quantitative trait loci. Theor Appl Genet 111:150-161
Nesbitt TC, Tanksley SD (2002) Comparative sequencing in the genus Lycopersicon: Implications for the evolution of fruit size in the domestication of cultivated tomatoes. Genetics 162:365-379
Newton CR, Graham A, Heptinstall LE, Powell SJ, Summers C, Kalsheker N, Smith JC, Markham AF (1989) Analysis of any point mutation in DNA. The amplification refractory mutation system (ARMS). Nucleic Acids Res 17:2503-2516
108Papp AC, Pinsonneault JK, Cooke G, Sadee W (2003) Single nucleotide polymorphism genotyping using allele-specific PCR and fluorescence melting curves. BioTechniques 34:1068-1072
Paran I, Michelmore RW (1993) Development of reliable PCR-based markers linked to downy mildew resistance genes in lettuce. Theor Appl Genet 85:985-993
Polashock JJ, Vorsa N (2002) Development of SCAR markers for DNA fingerprinting and germplasm analysis of American cranberry. J Am Soc Hort Sci 127:677-684
Rozen S, Skaletsky HJ (2000) Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S (eds) Bioinformatics Methods and Protocols: Methods in Molecular Biology. Totowa, Humana Press, pp 365-386
Sarkar G, Cassady J, Bottema CDK, Sommer SS (1990) Characterization of polymerase chain reaction amplification of specific alleles. Anal Biochem 186:64-68
Serquen FC, Bacher J, Staub JE (1997) Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random-amplified polymorphic DNA markers. Mol Breed 3:257-268
Sheffield VC, Cox DR, Lerman LS, Myers RM (1989) Attachment of a 40-base-pair G+C-rich sequence (GC-clamp) to genomic DNA fragments by the polymerase chain reaction results in improved detection of single-base changes. Proc Natl Acad Sci U S A 86:232-236
Shirasawa K, Kishitani S, Nishio T (2004) Conversion of AFLP markers to sequence-specific markers for closely related lines in rice by use of the rice genome sequence. Mol Breed 14:283-292
Staden R (1996) The Staden sequence analysis package. Mol Biotechnol 5:233-241
Staub JE, Serquen FC, Gupta M (1996) Genetic markers, map construction, and their application in plant breeding. HortScience 31:729-741
Tang SX, Kishore VK, Knapp SJ (2003) PCR-multiplexes for a genome-wide framework of simple sequence repeat marker loci in cultivated sunflower. Theor Appl Genet 107:6-19
Torjek O, Berger D, Meyer RC, Mussig C, Schmid KJ, Sorensen TR, Weisshaar B, Mitchell-Olds T, Altmann T (2003) Establishment of a high-efficiency SNP-based framework marker set for Arabidopsis. Plant J 36:122-140
Willcox MC, Khairallah MM, Bergvinson D, Crossa J, Deutsch JA, Edmeades GO, Gonalez-de-Leon D, Jiang C, Jewell DC, Mihm JA, Williams WP, Hoisington D (2002)
109Selection for resistance to southwestern corn borer using marker-assisted and conventional backcrossing. Crop Sci 42:1516-1528
Yang WC, Bai XD, Kabelka E, Eaton C, Kamoun S, van der Knaap E, Francis D (2004) Discovery of single nucleotide polymorphisms in Lycopersicon esculentum by computer aided analysis of expressed sequence tags. Mol Breed 14:21-34
Zhu YL, Song QJ, Hyten DL, Van Tassell CP, Matukumalli LK, Grimm DR, Hyatt SM, Fickus EW, Young ND, Cregan PB (2003) Single-nucleotide polymorphisms in soybean. Genetics 163:1123-1134
Table 2.1. RAPD markers converted to SCAR markers in cucumber. Original RAPD SCAR namea
a No primers were created for OP-AT15, OP-T2, and OP-Y10 because reliable sequence information was not obtained b SCAR markers were either the same locus as the original RAPD (Yes) or another locus (No) c SCAR markers verified to match original RAPD by segregation on 20 F2 individuals either directly or from SNP markers created from the SCAR marker (Seg = Segregation), or by molecular weight determined by agarose gel electrophoresis (MW) d The molecular weight of the PCR product amplified from Gy-7 (GMW) or H-19 (HMW) e Polymorphism type: dominant for Gy-7 (Dom-G), or H-19 (Dom-H), or codominant (Codom) f The number of PCR products produced from Gy-7 or H-19 g The range of PCR annealing temperatures that produces the lowest number of bands for Gy-7 and H-19 as determined by annealing temperature gradient PCR (ATG-PCR) h The optimal PCR annealing temperature i The number of base pairs sequenced from both Gy-7 and H-19 to identify SNPs. No seq = sequencing was attempted but could not be obtained for both genotypes
111
112Table 2.2. SCAR markers created from BAC end sequences in cucumber. Marker used as probe
a The name of the BAC library clone from Nam et al. (2005) b Molecular weight c Polymorphism type: dominant for Gy-7 (Dom-G), or H-19 (Dom-H), or codominant (Codom) d The number of base pairs sequenced from both Gy-7 and H-19 to identify SNPs
Table 2.3. Allele-specific markers created by four design approaches based on SNPs between two cucumber lines (Gy-7 and H-19) identified from two sequence sources. Gy-7 allele-specific marker H-19 allele-specific marker Codominant SNP markere
a Sequences were obtained by sequencing SCARs converted from RAPDs (SCAR) or BAC ends (BAC) b Allele-specific markers based on SNPs were created by one of four design approaches (Figure 2.3). The Gy-7 and H-19 sequences of C8-BE-R were completely different, and markers for only H-19 allele were designed for W-2-BE-R c SNP markers were dominant for Gy-7 (Dom-G), or H-19 (Dom-H), codominant by a single assay (Codom), or codominant by two assays [Codom (2 assay; Figure 2.2)] d Optimal PCR annealing temperature e If the Gy-7 and H-19 allele-specific markers were successfully combined in a single assay they were considered as a single marker and their names were combined (e.g., AB14SNPG1 and AB14SNPH1 becomes AB14SNPG1H1)
113
Table 2.4 Primer names and sequences of the polymorphic markers converted from RAPDs and verified by segregation in cucumber. Original RAPD New Marker Marker type Polymorphisma Primer name Primer sequence (5’ to 3’) BC231 BC231SCAR SCAR Dom-H BC231SCARF AGGGAGTTCCAAACTTTTCAGTAC BC231SCARR AGGGAGTTCCCCTGTGATCTCTCT BC388 BC388SCAR SCAR Dom-G BC388SCARF CGGTCGCGTCCTTAGACCAACCAC BC388SCARR CGGTCGCGTCATTCTGTATGAGGC BC403 BC403SCAR SCAR Dom-G BC403SCARF GGAAGGCTGTCTTCCTTATGTCTT BC403SCARR GGAAGGCTGTGCAAGGTCGAGGGA BC469 BC469SCAR SCAR Dom-H BC469SCARF CTCCAGCAAACTAACAATTGAGGG BC469SCARR CTCCAGCAAAGATTTCAAAAGGCT BC515 BC515SCAR SCAR Dom-H BC515SCARF GGGGGCCTCATTATGAGGAATGAA BC515SCARR GGGGGCCTCAAGTGAAAACAATCA BC523 B-3-BE-R SCAR Codom B-3-BE-R_F CCAAAACATACGACCCATCC B-3-BE-R_R TTCAATCGGTTTCCATGTTC B-5-BE-L SCAR Dom-G B-5-BE-L_F CCCGAGTTTATGTGGAAATG B-5-BE-L_R AAGAGGTGCTTGGGAAAGTG BC523SCAR SCAR Dom-G BC523SCARF ACAGGCAGACCCGACGAGGGGCAG BC523SCARR ACAGGCAGACAAGAGTTTGAGGAT BC526 BC526SCAR SCAR Codom BC526SCARF AACGGGCACCCGTCTCACTGGAAA BC526SCARR AACGGGCACCCACATAGTGAAAAC BC592 BC592SCAR SCAR Dom-H BC592SCARF GGGCGAGTGCAATATCTAAAATGG BC592SCARR GGGCGAGTGCATGCGAACACAAAA OP-AB14 AB14SNPG1H1 SNP Codom AB14SCARF AAGTGCGACCGGGTCAGTAAATTA AB14SNP451G1 ACTTGGAAAGCGGACATAGA AB14SNP491H1 GTTATCATATCTATCAGTAACAGAAGGAA OP-AC17 AC17SNPG1H2 SNP Codom AC17SNPG1T GATACAGATACACTCGGTTACCTGTAGTCTTGAACTA AC17SNPH4 GGTTACCTGTAGTCTTGAACGG AC17SNPUR TTTTTCCTGTTCTGTCATCGTG OP-AC9 AC9SNPG3H3 SNP Codom AC9SCARF AGAGCGTACCACTATGAGTGAGAA AC9SNP274G6 TCCGCATATTGATCCTTCTATTA AC9SNP322H6 TCCTTGAGATTCGACCAAACTA OP-AD14 AD14SNPG1H2 SNP Codom AD14SCARR GAACGAGGGTGAATGTTGCGAAAC AD14SNP602G1 ATTTTGTAAACATCTCGAAGTGAAC AD14SNP648H3 TCTGCTCGTCCATCTCGTCA 114
116Original RAPD New Marker Marker type Polymorphisma Primer name Primer sequence (5’ to 3’) OP-L19-2 L19SCAR SCAR Dom-H L19SCARF GAGTGGTGACCATATATTAAAGTG L19SCARR GAGTGGTGACTGTAATATCACAAA L1LG3H3 SNP Codom L-1-BE-L_R ACACTAGAGCTGGAAAGTGAAGC L1L195H6 ACGATGTTGAAATGGAAACG L1L76G9 TGTTTATCATCTTTGTCTTTTATCGA OP-M8 M4LG2 SNP Dom* M4L152G2 GTTTGTAAAAGAGAGGGAAATCA M4LH2 SNP Dom* M4L152H3 CAAACTATTTCATTTCAGTCCGTA M-5-BE-L SCAR Codom M-5-BE-L_F GAATAAGCGCTCCAGCTCAG M-5-BE-L_R TAATCGGACGGTGTGAAAGG M7LG3 SNP Dom* M7L171G6 GTACCGTGATGGGAGTAAAC M7LH3 SNP Dom* M7L171H8 TTACAGCTACAGAGCTGCCTCTA M8SCAR SCAR Dom-H M8SCARF TCTGTTCCCCATACAAGAATTAAA M8SCARR TCTGTTCCCCATGATGTAGACTTC M8SNPG3H1 SNP Codom M8SCARR TCTGTTCCCCATGATGTAGACTTC M8SNP146G5 TCTAATTTAGTTGAAAATGTTAATCAATATC M8SNP234H2 GCAACTTAATTACAATTTGGTCAT OP-N8 N8SNPG2H3 SNP Codom N8SCARF ACCTCAGCTCCCAAACATTAAAAT N8SNP497H5 GCGTGTCAAACACACTCTACC N8SNP545G4 AATGTCAATAGGATTATTCACTGATG OP-P14 P14SCAR SCAR Dom-H P14SCARF CCAGCCGAACACAAGAAGGTCTGA P14SCARR CCAGCCGAACGCCTCACTCTTGAT OP-W7-1 W7aSCAR SCAR Dom-H W7aSCARF CTGGACGTCAACTAAAAGGTAATT W7aSCARR CTGGACGTCAATGTAGAGAGGGCT OP-W7-2 W7SNPG1H3 SNP Codom W7SCARF CTGGACGTCACATATCAGTAAGTA W7SNP248G1 CACCCCTCTTTAATTATTAATTGAAA W7SNP364H3 GAGCGTTGGGAGTTCCTATAT a Polymorphism type: dominant for Gy-7 (Dom-G), or H-19 (Dom-H), or codominant (Codom) * The two dominant markers sets of M4LG2/M4LH2 and M7LG3/M7LH3 are each considered a codominant marker but require genotyping with both markers to determine heterozygosity (i.e., a two assay, codominant marker)
Table 2.5. Results of RAPD to SCAR conversion in cucumber of RAPDs common to Horejsi et al. (1999) and this study. SCARs from Horejsi et al. (1999) SCARs from this study
a The number of base pairs of the original RAPD primer (10 bp) that are part of the two SCAR primers. If the entire RAPD primer is included in both SCAR primers, the value is 20 (highest possible value) b Polymorphism type: dominant for line Gy-7 (Dom-G), or H-19 (Dom-H), or codominant (Codom) c The number of bands (including polymorphic band) produced by the SCAR marker d The similarity of the sequences obtain between the two studies. NA indicates the sequences were not aligned and are completely different. An asterisk (*) indicates that differences were detected in the priming site of at least one of the SCAR primers created by Horejsi et al. (1999)
117
C1SCAR (372 bp)
N8SCAR (1300 bp)
C10SCAR (726 bp)50
.650
.651
.653
.455
.058
.561
.464
.266
.768
.769
.970
.1 oC
AF7SCAR (1700 bp)
N8SCAR
AF7SCAR
564
947831
137515841904
3530 bp
2027
564
947831
137515841904
3530 bp
2027
C1SCAR
C10SCAR
A
C1SCAR
AF7SCAR
C1SCAR
C10SCAR
AF7SCAR
C1SCAR
C10SCAR
N8SCAR
B
ED
C50
.650
.651
.653
.455
.058
.561
.464
.266
.768
.769
.970
.1 oC 50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC
50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC
C1SCAR (372 bp)
N8SCAR (1300 bp)
C10SCAR (726 bp)50
.650
.651
.653
.455
.058
.561
.464
.266
.768
.769
.970
.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC
AF7SCAR (1700 bp)
N8SCAR
AF7SCAR
564
947831
137515841904
3530 bp
2027
564
947831
137515841904
3530 bp
2027
C1SCAR
C10SCAR
A
C1SCAR
AF7SCAR
C1SCAR
C10SCAR
AF7SCAR
C1SCAR
C10SCAR
N8SCAR
B
ED
C50
.650
.651
.653
.455
.058
.561
.464
.266
.768
.769
.970
.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC 50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC
50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC50.6
50.6
51.6
53.4
55.0
58.5
61.4
64.2
66.7
68.7
69.9
70.1 oC
Figure 2.1. SCAR multiplex reactions in cucumber. Panel C: banding patterns of individual SCAR primer pairs, including molecular weight in base pairs (bp) across a temperature gradient. Vertical numbers denote PCR annealing temperatures (oC) for each lane. Panels A, B, D, and E contain two, three, three, and four primer pairs, respectively, added to the same PCR reaction. Molecular weight of the EcoRI+HindIII digested lambda marker are to the right of panels B and E.
118
119
……Gy-7
H-19
Gy-7 allele specific primer5’-3’ (sense)
H-19 allele specific primer5’-3’ (sense)
Universal non-specific primer3’-5’ (antisense)
3’ mismatch*
^
** **
^
3’ mismatch
Gy-7
H-19
Gy-7
H-19
Gy-7
H-19 F
1 F2 Individuals
G H C
MW
ladder
MW
ladder
300 bp
200 bp
A
B
……Gy-7
H-19
Gy-7 allele specific primer5’-3’ (sense)
H-19 allele specific primer5’-3’ (sense)
Universal non-specific primer3’-5’ (antisense)
3’ mismatch*
^
** **
^
3’ mismatch
……Gy-7
H-19
Gy-7 allele specific primer5’-3’ (sense)
H-19 allele specific primer5’-3’ (sense)
Universal non-specific primer3’-5’ (antisense)
3’ mismatch*
^
** **
^
3’ mismatch
Gy-7
H-19
Gy-7
H-19
Gy-7
H-19 F
1 F2 Individuals
G H C
MW
ladder
MW
ladder
300 bp
200 bp
Gy-7
H-19
Gy-7
H-19
Gy-7
H-19 F
1 F2 Individuals
G H C
MW
ladder
MW
ladder
300 bp
200 bp
A
B Figure 2.2. Allele specific primer design used to create a codominant marker in cucumber from SNPs within a locus employing the optimal approach. Panel A: SNPs between H19 (bottom sequence) and Gy-7 (top sequence) in a portion of the AD14SCAR sequence are indicated by an asterisk (*). Allele specific primers match the SNP of one parent at the 3’ end with an additional mismatch (^) to both alleles within 4 bases of the 3’ end. Universal non-specific primers have no mismatch to either allele. Primer orientation and direction of extension by a polymerase (horizontal arrows) during PCR are indicated under each primer name. Panel B: Photograph after agarose gel electrophoresis of PCR reactions using Gy-7 and H-19 as template with the dominant Gy-7 allele specific marker (Gy-7 allele specific primer and Universal non-specific primer) labeled G, the dominant H-19 allele specific marker (H-19 allele specific primer and Universal non-specific primer) labeled H, and the G-y7 and H-19 allele specific markers combined (Gy-7 allele specific primer, H-19 allele specific primer, and Universal non-specific primer) in a codominant assay labeled “C”. The codominant assay was also tested on F1 and F2 individuals from a cross between Gy-7 and H-19. A 100 bp ladder (MW ladder) flanks PCR products.
*GS
HS
GN
HN
Sandwich
*GS UN
HS*
*GS GN
* HSHN
Opposite
Optimal
*GS UN
HS
Tail
Gy-7
H-19 F
1
Gy-7
H-19 F
1
*GS
HS
GN
HN
*GS
HS
GN
HN
Sandwich
*GS UN
HS**
GS UN
HS*
*GS GN
* HSHN
*GS GN
* HSHN
Opposite
Optimal
*GS UN
HS
*GS UN
HS
Tail
Gy-7
H-19 F
1
Gy-7
H-19 F
1
Figure 2.3. Graphical representation of four design approaches developed herein to create Gy-7 and H-19 allele-specific markers in cucumber from a single locus depending on location and number of single nucleotide polymorphisms (SNPs). Solid horizontal lines represent a genomic fragment containing a SNP (asterisk) between Gy-7 and H-19. Arrows represent the direction of primer extension by a polymerase in PCR. Primers are designated as allele-specific (GS) and non allele-specific (GN) to amplify the Gy-7 allele, allele-specific (HS) and non allele-specific (HN) to amplify the H-19 allele, and non-allele-specific universal (UN) to amplify both alleles. The dotted line on the GS primer of the tail approach represents additional base pairs that do not anneal to the template during PCR, but are designed to add length to the PCR product. The horizontal dotted lines above and below the genomic fragment represent PCR products of Gy-7 and H-19 template, respectively. The panels on the far right represent the gel banding patterns of both approaches in each row after agarose gel electrophoresis with Gy-7, H-19, and an F1 hybrid as templates. 120
121
Chapter 3. Pyramiding QTL for multiple lateral branching in cucumber using nearly isogenic lines
Abstract
Multiple lateral branching (MLB) is a quantitatively inherited trait associated with
yield in cucumber (Cucumis sativus L.; 2n=2x=14). Although quantitative trait loci
(QTL) have been identified for MLB and QTL-marker associations have been verified by
marker-assisted selection, the epistatic effects of these QTL have not been characterized.
Therefore, markers linked to MLB were utilized to create two sets (standard- and little-
leaf types) of nearly isogenic lines (NIL) possessing various numbers of QTL to test for
epistatic interactions among QTL. As the number of QTL increased, the number of
branches generally increased in the little-leaf NIL, but decreased in the standard-leaf NIL,
demonstrating an epistatic effect of genetic background on lateral branch development.
Comparative analysis of NIL that differed by a single QTL indicated that the effects of
two specific QTL were dependent on genetic background in the little-leaf NIL, but
similar epistatic effects were not detected among QTL in standard-leaf NIL. The
evaluation of NIL in two Wisconsin environments at three plant densities revealed a
genotype independent decrease in the number of branches at higher plant densities, as
well as genotype by environment and QTL by environment interactions involved in MLB.
Thus, the production of lateral branches is determined by growing environment,
interactions among other cucumber traits, and interactions among QTL conditioning
branching.
122Introduction
Cucumber (Cucumis sativus var. sativus L.) is the fifth most widely grown
vegetable crop worldwide (2,427,436 hectares harvested in 2004) and ranks seventh in
the US in area harvested (68,660 hectares; FAOSTAT, 2005). Although the yield of US
processing cucumber has reached a plateau in the last twenty years, evidence from
several studies indicates that selection for multiple lateral branching (MLB) types (i.e.,
plants with several lateral branches) may increase cucumber yield (i.e., fruit per plant;
Fredrick and Staub 1989; Cramer and Wehner 1998; Cramer and Wehner 1999; Cramer
and Wehner 2000b). Path analysis of yield component traits revealed that the number of
branches per plant was consistently and highly correlated (r > 0.7) with yield when tested
in several populations, cycles of selection, and environments (Cramer and Wehner
2000a). Moreover, highly branched, determinate processing cucumber genotypes are
desirable for once-over machine harvest operations (Lower and Edwards 1986; Wehner
1989; Staub et al. 1992).
A feral relative of cucumber, C. sativus var. hardwickii (R) Alef. (hereafter
referred to as C. s. var. hardwickii; Horst and Lower 1978), and a cucumber inbred line
‘Little John’ (line H-19; synom. AR 79-75; Goode et al. 1980) both possess a multiple
lateral branching habit not present in commercial cucumber. Multiple lateral branching
in both sources is quantitatively inherited (Wehner et al. 1978; Serquen et al. 1997b;
Fazio et al. 2003b) demonstrating mostly additive genetic variance with narrow-sense
heritabilities (h2) ranging between 0.00 and 0.61 (Wehner et al. 1978; Serquen et al.
1997b). Genetic evaluation of an F3 population derived from lines Gy-7 (synom. G421)
123and H-19, indicated that MLB is controlled by at least four genes (Serquen et al.
1997b), where four QTL explained 48% to 66% of the observed variation (R2) depending
upon environment (Serquen et al. 1997a). Using recombinant inbred lines (RIL) derived
from the same parents, Fazio et al. (2003b) identified five QTL with a combined R2 of
37% to 55% depending on location. In both QTL studies, one major QTL was detected
that accounted for 32% (Fazio et al. 2003b) to 40% (Serquen et al. 1997a) of the variation,
which mapped near the little-leaf locus (ll).
The QTL analyses of Gy-7 × H-19 derived populations (Serquen et al. 1997a;
Fazio et al. 2003b) provided marker-QTL relationships that have been exploited in
marker-assisted selection (MAS) of MLB (Fazio et al. 2003a; Fan et al. 2006). The
increase in the number of branches from MAS was comparable to phenotypic selection
after two generations of backcrossing to Gy-7, the low branching parent (Fazio et al.
2003a). Likewise, two generations of MAS backcrossing after two cycles of phenotypic
recurrent selection for MLB in a similar population continued to increase the number of
lateral branches, and operated to fix favorable alleles that were not exploited by
phenotypic selection (Fan et al. 2006). Moreover, MAS was slightly more effective than
phenotypic selection for the improvement of MLB during recurrent selection in four
genetically distinct populations (Chapter 1).
A knowledge of epistasis is critical to comprehensive genetic analysis (Kinghorn
1987; Yano 2001) and breeding (Schnell and Cockerham 1992) of crop species.
Relatively little is known, however, about epistatic interactions among individual QTL
involved in MLB in cucumber. Therefore, nearly isogenic lines (NIL) possessing
124differing numbers of QTL for MLB were created to test for epistasis among QTL
involved in MLB in two leaf types, little- (30-40 cm2) and standard-leaf (> 40 cm2; Staub
et al. 1992). These NIL were evaluated at two locations and three plant densities to test
the effect of genetic background (leaf type) and environment on lateral branch production.
Materials and Methods
NIL creation Six major QTL affecting MLB identified by Serquen et al. (1997a) and Fazio et al.
(2003b) were chosen for the development of NIL because of their relatively high LOD
scores, R2 percentages, and genetic effects (Table 3.1). The effect of specific QTL on
MLB was estimated by comparing NIL that differed by a single QTL. NIL with specific
sets of QTL were compared in different growing environments to define epistatic effects
and to provide for an understanding of environmental factors governing MLB.
The same parents (Gy-7 and H-19) used in cross-progeny QTL analysis (Serquen
et al. 1997a; Fazio et al. 2003b) were utilized to create NIL with varying numbers of QTL
for MLB (Table 3.2). Lines Gy-7 (standard-leaf type; LL) and H-19 (little-leaf type; ll)
were crossed to create F1 individuals that were subsequently backcrossed to both parents
resulting in BC1 populations that were standard-leaf (Gy-7 as recurrent parent) and
segregating for leaf type (H-19 as recurrent parent). Individuals from both BC1
populations were phenotypically selected for high lateral branch number and leaf type in
an open-field nursery at the University of Wisconsin Agricultural Research Station,
125backcrossed to the recurrent parents to create little- and standard-leaf BC2 and BC3
families. These families were evaluated for MLB at Hancock in 2001, and 50 highly
branched individuals (30 standard-leaf and 20 little-leaf) were selfed to create BC2S1 and
BC3S1 families of each leaf type.
Twenty-four seedlings from each of the 50 BC2S1 and BC3S1 families were
genotyped using markers linked to QTL for MLB (Table 3.3). Since none of the
individuals contained all six QTL, the marker information was used to identify
individuals with complementing QTL compositions that could be crossed to provide for
as many QTL as possible in segregating progeny. A total of 11 little-leaf and nine
standard-leaf crosses were made, and 10 to 20 seeds of each cross were evaluated at
Hancock for MLB to identify 10 (five little-leaf and five standard-leaf) crosses where the
marker genotype (parental types) was confirmed by the branch number of progeny. All
the progeny (two to 96 individuals) from each of these 10 crosses were genotyped to
select plants that possessed as many QTL as possible, and then these selections were
genotyped and selfed for four to five generations to produce BC2S5/6 or BC3S5/6 NIL
differing in the number of QTL associated with MLB in both standard- and little-leaf
backgrounds (Table 3.2).
Molecular marker analysis Nine molecular markers and the ll gene linked to QTL for MLB were employed to
track the introgression of QTL into NIL (Table 3.3). All markers employed were from
Fazio et al. (2003b), except AJ6SCAR and M8SCAR, which were SCAR markers
(Chapter 2) converted from the RAPD marker mapped by Fazio et al. (2003b). Markers
126flanking each QTL were utilized, where available, to eliminate potential marker-QTL
recombination events. Leaf tissue for all genotyped lines was harvested, and DNA was
extracted, and then subjected to polymerase chain reaction (PCR) amplification and
agarose gel electrophoresis according to Fazio et al. (2003b).
Open-field evaluation of NIL for MLB To examine the effects of individual QTL, QTL number, and growing
environment on lateral branch production, NIL were evaluated in an open-field trial in the
summer of 2005 at three spacings within two locations; the University of Wisconsin
Agricultural Research Stations at Hancock and Arlington [Plano silt loam (Typic
Argiudoll)], Wisc. Seeds were sown in a greenhouse in Madison, Wisc. on June 11 and
June 14, 2005, and then transplanted on June 27 and July 29 to Hancock and Arlington,
respectively. Each location was arranged in a split-plot design with four replications of
spacing (whole plot factor) in randomized complete blocks, with the NIL completely
randomized as subplots with 10 plants per subplot. Plots were arranged in single rows
with 1.5 m between rows and 10, 15, and 20 cm between plants, corresponding to
approximately 66,700, 44,400, and 33,300 plants/ha, respectively. Lateral branch
number (at least three internodes in length) of each plant was recorded at or after anthesis
in the first ten nodes of the mainstem.
Statistical analysis Branching data were analyzed by analysis of variance using PROC GLM of SAS
(2003) to test for the main effects and interactions of location, spacing, leaf type, and
entry (NIL) nested within leaf type. All effects, except blocks, were considered fixed
127effects. Location was considered a fixed effect because the two locations chosen did
not represent a sampling of all cucumber growing environments, and inferences were
desired for Hancock, specifically, because it was the location originally used for the
detection of QTL (Serquen et al. 1997a; Fazio et al. 2003b). The two degrees of freedom
for spacings were partitioned into single degree of freedom contrasts within analyses of
variance to test for the linear (regression) and residual effect of spacing. The regression
coefficients of MLB on spacing were then obtained using PROC REG of SAS (2003).
Least squares means (lsmeans) are presented because of missing plots [22 of 288 (7.6%)].
To test for effects of specific QTL, comparisons were made between specific NIL that
differed in a single QTL (i.e., 12 NIL allowed for 11 single degree of freedom contrasts;
Table 3.4). For example, all little-leaf NIL in Table 3.2 have MLB1 and MLB6. In
addition, NIL-146 has MLB4, NIL-1246 has MLB2 and MLB4, and NIL-12456 has MLB2,
MLB4, and MLB5. Thus, NIL-146 and NIL-1246 have all MLB QTL in common, except
for MLB2. The addition effect of MLB2, therefore, is tested by comparing the means of
NIL-146 and NIL-1246 (Comparison A; Table 3.4). Likewise, the effect of adding
MLB5 to MLB2 and MLB4 is tested by Comparison B (NIL-12456 vs. NIL-1246).
Results
All main effects (location, spacing, leaf type, and entry nested within leaf type)
were highly significant (P < 0.001; Appendix L). The interactions of location with
spacing and leaf type were not significant, but the location by entry interaction was
highly significant (P < 0.001). Therefore, location lsmeans of MLB for each NIL (Table
1283.2) were used to interpret location effects. The number of branches of little-leaf NIL
tended to decrease in Hancock, but increase in Arlington with increasing number of QTL
(Table 3.2; Appendix K). In contrast, the number of branches of the standard-leaf NIL
tended to decrease in Arlington with increasing numbers of QTL, but no trend was
evident in Hancock as the number of QTL increased. Although the spacing by entry
interaction was not significant, the spacing by leaf type interaction was marginally
significant (P = 0.049). Therefore, the effect of spacing on MLB is presented separately
for each leaf type (Figure 3.1). Both the linear and residual effects of spacing were
significant in both leaf types (Figure 3.1). The lsmeans for MLB were 6.2, 7.6, and 7.8
(little-leaf NIL) and 0.9, 1.6, and 1.8 (standard-leaf NIL) branches per plant at within row
spacings of 10, 15, and 20 cm, respectively [LSD (α = 0.05) = 0.3].
Two of the six means comparisons for little-leaf NIL (Comparisons C and E)
were significant (Table 3.4). Of the three means comparisons (A, C, and F) that
examined the addition effect of MLB2 in little-leaf NIL, only one was significant
(Comparison C). Similarly, one of the three comparisons that evaluated the addition
effect of MLB5 (Comparison E) was significant. QTL by location interactions (Table
3.4) for MLB2 were highly significant (P < 0.001; Comparison C), marginally significant
(P = 0.04; Comparison A), or not significant (Comparison F), while one such interaction
for MLB5 was highly significant (P < 0.001; Comparison E) and two were marginally
significant [P = 0.06 (Comparison D) and 0.09 (Comparison B)]. Of the standard-leaf
NIL (Comparisons G-J), the effect of MLB3, when evaluated with MLB2 (Comparison G)
or alone (Comparison J), was not significant. In contrast, the effect of MLB5 was
129significant when alone (Comparison H) or with MLB2 (Comparison I). Only one
QTL by location interaction tested in the standard-leaf NIL was significant (P = 0.05;
Comparison J).
Increasing the number of QTL in the little-leaf background from three to four
increased the number of branches in two cases (NIL-136 to NIL-1236, and NIL-136 to
NIL-1356), while no such change was detected in a third case (NIL-146 to NIL-1246;
Figure 3.2). In all instances, the number of branches did not change when the number of
QTL was increased from four to five in little-leaf NIL. The increase from zero to two
QTL either did not change (NIL-0 to NIL-23) or decreased (NIL-0 to NIL-25) the
number of branches in standard-leaf plants (Figure 3.2). Similarly, the number of lateral
branches decreased (NIL-23 to NIL-235) or did not change (NIL-25 to NIL-235) when a
third QTL was added.
Discussion
Segregating populations such as F2, F2-derived F3 (F2:3), or recombinant inbred
lines (RIL) typically used for QTL identification are inadequate for the detection of
epistatic interactions among individual QTL (Lin et al. 2000). In contrast, the use of NIL
to detect epistatic interactions among QTL for heading date in rice (Oryza sativa L.)
illustrates the power of NIL over such populations for inter-allelic analyses. Therefore,
the marker-QTL relationships for MLB, previously identified and verified by MAS (Fan
et al. 2006; Chapter 1), were utilized to create NIL for the characterization of epistatic
interactions among QTL for MLB in cucumber. Based on the effects estimated from the
130QTL analyses of Serquen et al. (1997a) and Fazio et al. (2003b), the alleles of the
highly-branched H-19 parent at all but one QTL (MLB3) should contribute to higher
lateral branch number (Table 3.1). Although the estimated effects of each QTL are
dissimilar, incrementally combining the Gy-7 allele of MLB3 with the H-19 allele at all
other QTL should predictably increase the number of lateral branches in either little- or
standard-leaf NIL under a no-epistasis inheritance model.
The number of lateral branches in H-19 or C. s. var. hardwickii derived
germplasm was not affected by environment (Georgia and Hancock, Wisc.; Serquen et al.
1997b) or planting date (early and late; Fredrick and Staub 1989). In addition, the same
four QTL were identified in three different environments (Hancock, Wisc, in 1999 and
2000 and Utah in 1999; Fazio et al. 2003b), indicating that MLB habit can be relatively
stable across environments. However, Fazio et al. (2003b) identified a QTL specific to
Hancock (LOD 2.7-3.0 in both years), and seven other QTL (LOD 2.8-6.1) unique to a
single environment. This result, coupled with an estimate of narrow-sense heritability of
0.48 for MLB (Serquen et al. 1997b), indicates that MLB is affected by the environment.
Indeed, the number of lateral branches varied across years (López-Sesé and Staub 2002)
and planting dates (Chapter 1) when grown at Hancock, Wisc. The expression of MLB
was affected by environment in this study (Table 3.2), as evidenced by the location main
effect and location by entry interaction. Not only were genotype by location effects
evident, but QTL by location interactions were detected in both leaf-type backgrounds
(Table 3.4). Thus, although selection for MLB in one environment under a given plant
131density may increase the number of lateral branches, selection for MLB should be
performed in several commercial growing environments to optimize gain from selection.
The finding that the number of lateral branches decreased with increased plant
density (Figure 3.1) is consistent with previous results in C. s. var. hardwickii germplasm
(Fredrick and Staub 1989). This spacing effect was similar among all lines and across
locations examined. The highly significant residual effect of spacing on MLB suggests
their relationship is not linear (Figure 3.1). The decrease in the number of branches when
the spacing between plants was reduced from 20 to 15 cm was smaller than when spacing
was reduced from 15 to 10 cm in both leaf types. This difference, however, was less
pronounced for the little-leaf NIL than for standard-leaf NIL (i.e., spacing by leaf type
interaction; Figure 3.1). This reduction in number of lateral branches at higher plant
densities, coupled with the positive correlation of branching with fruit per plant, suggests
that the optimal plant density for yield of highly branched genotypes may be lower than
that of unbranched genotypes in machine harvest operations (100,000 to 200,000 plants/
ha; Wehner 1989; Staub et al. 1992). Thus, the relationship of branching and yield must
be evaluated in highly branched genotypes at plant densities above 44,400 plants/ha to
determine the optimal density.
The pyramiding of QTL for MLB did not necessarily increase the number of
lateral branches (Figure 3.2). In fact, increasing the number of QTL in the standard-leaf
NIL decreased the number of branches (Figure 3.2). The same QTL when in the little-
leaf NIL, however, generally increased the number of branches, and branching increased
when the number of QTL increased from three to four. Thus, the number of lateral
132branches is enhanced by the linkage block that contains the little-leaf gene. The QTL
with the greatest effect on MLB (MLB1, R2 = 32.4%; Table 3.1) was mapped to the
genomic region of the ll gene (Fazio et al. 2003b). Because ll was used as a marker for
the introgression of MLB1 (Table 3.3), only little-leaf NIL contained this QTL. The
effect of this genomic region on MLB is illustrated by the substantial difference in branch
number between the standard-leaf NIL-36 and the little-leaf NIL-136, which differ only
in MLB1 and ll (Table 3.2).
Fazio et al. (2003b) estimated that the effects of individual QTL are not equal
(Table 3.1), which was confirmed by this study. The means comparisons used to
estimate the effect of specific QTL (Table 3.4) revealed that the addition of some QTL
(i.e., MLB2 and MLB5 in little-leaf NIL) affected the number of lateral branches, while
others did not (i.e., MLB3 in standard-leaf NIL). In little-leaf NIL, the effect of a QTL
was dependent on the QTL with which it was combined. The differential effect of MLB2
and MLB5 when in combination with other QTL indicate that epistasis among QTL is a
major factor in the expression of MLB in little-leaf germplasm, and may be due to
duplicate gene action of MLB2 and MLB5. Epistasis was not detected, however, among
QTL in the standard-leaf NIL. The addition of MLB3, in every case (alone or with
another QTL), caused no change in the number of lateral branches (Table 3.2, Figure 3.2).
When MLB5, was added (either alone or with MLB2), however, a decrease in lateral
branches number was observed. The finding that MLB5 increased (Comparison E), did
not change (Comparisons B and D), or decreased (Comparisons H and I) the number of
lateral branches, depending on genetic background (i.e., QTL present and leaf type),
133indicates that epistasis plays a major role in the expression of MLB5. Thus, selection
for MLB will require the identification of favorable QTL combinations depending on the
genetic background employed.
The increased number of branches is desirable for once-over machine harvest and
branch number is higher in little-leaf than standard-leaf types. However, little-leaf types
tend to have poor fruit quality, and are monoecious and later flowering, resulting in low
first harvest yield, which is undesirable for machine harvest operations (Lower and
Edwards 1986; Wehner 1989). Previous studies indicate that MLB is also affected by the
determinate (de) and female (F) loci (Serquen et al. 1997a; Fazio et al. 2003a; Fazio et al.
2003b). Although determinant and/or gynoecious individuals with several branches have
been identified in little- and standard-leaf backgrounds, they generally have fewer
branches than their indeterminate and/or monoecious counterparts. Thus, even if all the
QTL that promote MLB are pyramided, the expression of MLB will depend upon plant
architecture (e.g., de and F) and genetic background.
The specific function of the QTL associated with MLB is largely unknown. It is
possible that the QTL mapped to the ll and de loci are not primarily involved with
branching, but represent pleitropic effects of ll and de on MLB (Fazio et al. 2003b).
Many of the QTL affecting MLB have been mapped near QTL for other traits, such as
sex expression, fruit length to diameter ratio, and fruits per plant, and MLB is
consistently correlated (either negatively or positively) with these traits (Kupper and
Staub 1988; Serquen et al. 1997b; Cramer and Wehner 1998; Cramer and Wehner 1999;
Cramer and Wehner 2000b; Fazio et al. 2003b). In fact, the simultaneous improvement
134of MLB and negatively correlated traits (e.g., gynoecy and earliness) using both
phenotypic selection and MAS has been largely unsuccessful (Chapter 1). Thus, some of
the factors involved in MLB may not be genes whose primary function is involved in
branching, but may be involved in the regulation of developmental processes and source
(i.e., photosynthetic base) to sink (i.e., fruit) relationships. Therefore, other traits (e.g.,
leaf type, determinate character, gynoecious sex expression, earliness, and fruit length to
diameter ratio) must be considered when developing highly branched cucumber
phenotypes to improve yield in cucumber.
Literature Cited
Cramer CS, Wehner TC (2000a) Path analysis of the correlation between fruit number and plant traits of cucumber populations. HortScience 35:708-711
Cramer CS, Wehner TC (2000b) Fruit yield and yield component correlations of four pickling cucumber populations. Cucurbit Genet Coop Rpt 23:12-15
Cramer CS, Wehner TC (1999) Little heterosis for yield and yield components in hybrids of six cucumber inbreds. Euphytica 110:99-108
Cramer CS, Wehner TC (1998) Fruit yield and yield component means and correlations of four slicing cucumber populations improved through six to ten cycles of recurrent selection. J Am Soc Hort Sci 123:388-395
Fan Z, Robbins MD, Staub JE (2006) Population development by phenotypic selection with subsequent marker-assisted selection for line extraction in cucumber (Cucumis sativus L.). Theor Appl Genet 112:843-855
FAOSTAT (2005) Agriculture. FAO Statistical Databases (http://faostat.fao.org, last accessed December 2005)
Fazio G, Chung SM, Staub JE (2003a) Comparative analysis of response to phenotypic and marker-assisted selection for multiple lateral branching in cucumber (Cucumis sativus L.). Theor Appl Genet 107:875-883
135Fazio G, Staub JE, Stevens MR (2003b) Genetic mapping and QTL analysis of horticultural traits in cucumber (Cucumis sativus L.) using recombinant inbred lines. Theor Appl Genet 107:864-874
Fredrick LR, Staub JE (1989) Combining ability analyses of fruit yield and quality in near-homozygous lines derived from cucumber. J Am Soc Hort Sci 114:332-338
Goode MJ, Bowers JL, Bassi Jr. A (1980) Little leaf, a new kind of pickling cucumber plant. Arkansas Farm Res 29:4
Horst EK, Lower RL (1978) Cucumis hardwickii: A source of germplasm for the cucumber breeder. Cucurbit Genet Coop Rpt 1:5
Kinghorn BP (1987) The nature of 2-locus epistatic interactions in animals: evidence from Sewall Wright's guinea pig data. Theor Appl Genet 73:595-604
Kupper RS, Staub JE (1988) Combining ability between lines of Cucumis sativus L. and Cucumis sativus var. hardwickii (R.) Alef. Euphytica 38:197-210
Lin HX, Yamamoto T, Sasaki T, Yano M (2000) Characterization and detection of epistatic interactions of 3 QTLs, Hd1, Hd2, and Hd3, controlling heading date in rice using nearly isogenic lines. Theor Appl Genet 101:1021-1028
López-Sesé AI, Staub J (2002) Combining ability analysis of yield components in cucumber. J Am Soc Hort Sci 127:931-937
Schnell FW, Cockerham CC (1992) Multiplicative vs. arbitrary gene action in heterosis. Genetics 131:461-469
Serquen FC, Bacher J, Staub JE (1997a) Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random-amplified polymorphic DNA markers. Mol Breed 3:257-268
Serquen FC, Bacher J, Staub JE (1997b) Genetic analysis of yield components in cucumber at low plant density. J Am Soc Hort Sci 122:522-528
136Staub JE, Knerr LD, Hopen HJ (1992) Plant density and herbicides affect cucumber productivity. J Am Soc Hort Sci 117:48-53
Wehner TC (1989) Breeding for improved yield in cucumber. Plant Breed Rev 6:323-359
Wehner TC, Staub JE, Peterson CE (1978) Inheritance of littleleaf and multi-branched plant type in cucumber. Cucurbit Genet Coop Rpt 10:33-34
Yano M (2001) Genetic and molecular dissection of naturally occurring variation. Curr Opin Plant Biol 4:130-135
137
Table 3.1. Characteristics of previously identified cucumber quantitative trait loci (QTL) associated with multiple lateral branching (MLB) that were introgressed to create nearly isogenic lines (NIL)
Recombinant inbred line analysisa F2:3 progeny analysisb
Hancock, WI 1999 Hancock, WI 2000 Utah 1999 Tifton, GA Hancock, WI QTL Namec LODd R2 (%)e Effectf LOD R2 (%) Effect LOD R2 (%) Effect LOD R2 (%) LOD R2 (%) MLB1 mlb1.4 32.9 32.4 0.63 7.0 8.1 0.23 9.8 17.2 0.42 4.6 13.6 nd nd MLB2 mlb1.1 11.6 9.1 0.36 8.2 10.6 0.24 6.8 11.5 0.33 10.4 39.6 10.1 37.0 MLB3 mlb6.2 4.2 3.7 -0.17 3.7 4.3 -0.15 2.7 4.9 -0.18 MLB4 nd nd 3.3 11.0 MLB5 mlb4.4 3.0 1.7 0.17 4.6 4.6 0.37 2.7 3.3 0.20 MLB6 mlb6.1 2.7 1.5 0.11 3.0 2.9 0.17 ndg nd nd a QTL identified by Fazio et al 2003b b QTL identified by Serquen et al 1997a c QTL name given by Fazio et al 2003b d Log of likelihood ratio e Percentage of the phenotypic variation explained f The effect of the H-19 allele on the number of lateral branches g Not detected
138
Table 3.2. QTL composition and mean number of branches in near isogenic lines (NIL) of cucumber. Quantitative trait locic Number of lateral branchesd
NILa Leaf typeb MLB1 MLB2 MLB3 MLB4 MLB5 MLB6 Generation Hancock, Wisc. Arlington, Wisc. Mean NIL-1246 little MLB1 MLB2 MLB4 MLB6 BC2S5 6.75 7.30 7.03 NIL-136 little MLB1 MLB3 MLB6 BC2S5 7.89 4.68 6.29 NIL-1236 little MLB1 MLB2 MLB3 MLB6 BC2S5 7.10 8.28 7.69 NIL-1356 little MLB1 MLB3 MLB5 MLB6 BC2S6 7.20 7.44 7.32 NIL-12456 little MLB1 MLB2 MLB4 MLB5 MLB6 BC2S5 7.13 6.77 6.95 NIL-12356 little MLB1 MLB2 MLB3 MLB5 MLB6 BC2S5 7.24 8.19 7.71 NIL-146 little MLB1 MLB4 MLB6 BC2S6 7.64 7.18 7.41 NIL-36 standard MLB3 MLB6 BC3S5 1.45 1.38 1.41 NIL-23 standard MLB2 MLB3 BC2S5 2.13 1.98 2.05 NIL-25 standard MLB2 MLB5 BC2S5 0.93 1.25 1.09 NIL-235 standard MLB2 MLB3 MLB5 BC3S5 1.30 0.59 0.94 NIL-0 standard BC2S5 1.38 1.89 1.64
a NIL names reflect their QTL composition (i.e., NIL-1246 contains MLB1, MLB2, MLB4, and MLB6) b Leaf type classified as standard- (> 40 cm2) or little-leaf (30-40 cm2; Staub et al. 1992) c QTL names are from Table 3.1 d Lsmeans for Hancock, Wisconsin, Arlington, Wisconsin, and combined locations. LSD’s (α= 0.05) are 0.96, 0.60, and 0.57, and Coefficient of variation (CV) are 22.5%, 15.1%, and 19.6%, respectively.
139Table 3.3. Marker-QTL associations used to introgress quantitative trait loci (QTL) for multiple lateral branching (MLB) into nearly isogenic lines (NIL) in cucumber Flanking markera Distance (cM)b QTL Distance (cM) b Flanking markerc
OP-AD12-1 9.1 MLB1 3.7 lld
MLB2 6.1 OP-AG1-1 L19-2-SCAR 6.3 MLB3 16.1 NR60 AJ6SCAR MLB4e BC523SCAR CSWTAAA01 2.3 MLB5 AK5SCAR 4.1 MLB6 2.0 M8SCAR a The marker above the QTL in the genetic map of Fazio et al. (2003b) b The genetic map distance between the marker and the QTL c The marker below the QTL in the genetic map of Fazio et al. (2003b) d The little-leaf gene (Pierce and Wehner 2000) e MLB4 was mapped between AJ6SCAR and BC523SCAR, which are separated by 5.1 cM (Fazio et al. 2003b)
140Table 3.4. Means comparisons to determine specific quantitative trait loci (QTL) effects in nearly isogenic lines (NIL) of cucumber
Test hypothesisb
Means comparisona Comparison QTL effect Background P-valuecQTL × loc P-valued
NIL-1246 vs. NIL-146 A MLB2 MLB4 0.1542 0.0416 NIL-12456 vs. NIL-1246 B MLB5 MLB2 & MLB4 0.7716 0.0866 NIL-1236 vs. NIL-136 C MLB2 MLB3 <0.0001 <0.0001 NIL-12356 vs. NIL-1236 D MLB5 MLB2 & MLB3 0.2173 0.0598 NIL-1356 vs. NIL-136 E MLB5 MLB3 0.0001 <0.0001 NIL-12356 vs. NIL-1356 F MLB2 MLB3 & MLB5 0.1323 0.1756 NIL-23 vs. NIL-0 G MLB2 & MLB3 0.1130 0.2044 NIL-235 vs. NIL-23 H MLB5 MLB2 & MLB3 <0.0001 0.2946 NIL-25 vs. NIL-0 I MLB2 & MLB5 0.0375 0.7117 NIL-235 vs. NIL-25 J MLB3 MLB2 & MLB5 0.5782 0.0522 a The means (over both locations) of the two lines in the means comparison column were tested by single degree of freedom contrasts with the null hypothesis as equal means b Null hypothesis, where adding the QTL in QTL effect column to the QTL already present in the background column has no effect on the number of lateral branches c The P-value of the means comparison d The P-value of the QTL by location (loc; Hancock and Arlington, Wisc.) interaction
Figure 3.1. The effect of plant density on the number of lateral branches in cucumber near-isogenic lines (NIL) of two leaf types (little-leaf and standard-leaf). The linear and residual P-values, as well as the R2 of each leaf type are presented to the right of each leaf type. Within row spacings of 10, 15 and 20 cm correspond to plant densities of 66,700, 44,400, and 33,300 plants/hectare, respectively.
P-linear < 0.0001 P-residual = 0.0022 R2 = 0.92
142
Little-leaf NIL
NIL-146
NIL-1246 NIL-12456
NIL-1236
NIL-136
NIL-1356
NIL-12356
5.5
6.0
6.5
7.0
7.5
8.0
2 3 4 5 6
Num
ber
of la
tera
l bra
nche
s
Comparisons A & B
Comparisons C & D
Comparisons E & F
**
Standard-leaf NIL
NIL-235
NIL-23
NIL-0
NIL-25
NIL-36
0.0
0.5
1.0
1.5
2.0
2.5
0 1 2 3 4
Number of QTL
Num
ber o
f lat
eral
bra
nche
s
Comparisons G & H
Comparisions I & J
NIL-36
*
*
Figure 3.2. The effect of increasing the number of quantitative trait loci (QTL) on the number of lateral branches in two leaf types of cucumber as determined by near-isogenic lines (NIL; Table 3.2). Lines indicate the incremental addition of QTL, and the comparisons in the legends refer to means comparisons performed (Table 3.4). Asterisks (*) indicate significant means comparisons (P < 0.05).
143
Conclusions and Future Work
The expression of cucumber (Cucumis sativus L.; 2n=2x=14) yield component
traits investigated herein [multiple lateral branching (MLB), gynoecious sex expression
(GYN), earliness (EAR), and fruit length to diameter ratio (L:D)] is complex and
influenced by genetic and environmental effects at several levels. Epistatic interactions
were detected among quantitative trait loci (QTL) for MLB, as well as between MLB and
leaf type. The expression of MLB was affected by the environment and QTL by
environment interactions were detected. Thus, manipulation of these traits by phenotypic
selection (PHE) or marker-assisted selection (MAS) will not be straightforward. Indeed,
both MAS and PHE provided improvements for these traits, but response to selection was
dependent upon genetic background (population). In some populations, trait means
decreased, but in others, improvements were made from both selection methods. MAS
and PHE were considered separately herein for a direct comparison of their efficacy, but
since both methods provided improvements, selection for yield component traits may be
most effective by combining both MAS and PHE.
Using the methods and resources of this study, MAS and PHE could be combined
by evaluating 600 individuals and selecting 20 by PHE during the summer. During the
establishment of cuttings from PHE, another 20 selections could be made by MAS from
600 individuals of the same population and all 40 selections could be intermated in the
greenhouse in the fall. Another option would be to genotype a large number of seedlings
with a select group of reliable sequenced characterized amplified region (SCAR), simple
144sequence repeat (SSR), or single nucleotide polymorphism (SNP) markers before
transplanting in the field for subsequent phenotypic selection. The multiplexing
guidelines and new SCAR and SNP markers developed herein would allow for efficient
genotyping of seedlings to greatly reduce the number of plants evaluated by phenotypic
selection. MAS and PHE could also be combined using and index based on both marker
and phenotypic information. Both phenotypic and genotypic data could be taken for all
traits on each plant, or plants could be phenotyped for GYN, EAR, and L:D and
genotyped for MLB, since these method/trait combinations were generally most effective
in this study. The combination of MAS and PHE using one or a combination of these
methods would allow a test of the hypothesis that MAS and PHE are more effective at
improving yield components when utilized together than separately.
Yield was generally not increased, nor was the simultaneous improvement of
MLB, GYN, EAR, and L:D accomplished during population improvement (recurrent
selection) from MAS or PHE, so the hypothesis that yield increases with the
improvement of all four traits cannot be rejected. These results suggest that obtaining
high-parent values of all four traits (i.e., > 6 branches, 100% female flowers, 2.0
fruit/plant in the first harvest, and L:D > 3.0) simultaneously in this germplasm by
population improvement will be difficult given the correlations between traits, which are
due, in large part, to pleitropic effects and/or linkage between QTL conditioning these
traits. When considering this hypothesis, the results of a parallel study are informative.
Fan et al. (2006) independently evaluated the effectiveness of two cycles of MAS for
MLB, GYN, and L:D in a backcrossing scheme using selections from cycle 2 (C2) of
145PHE in Population 1 of this study as recurrent parents to produce two backcross
populations. In addition to the gains for MLB and L:D after two cycles of recurrent
selection by PHE, MAS continued to improve MLB and L:D in one backcross population,
and L:D in the other, while GYN was improved in both populations and yield increased
in the former population after only two cycles of MAS. These results suggest that line
extraction (backcrossing) is more efficient than population improvement (recurrent
selection) for simultaneous multi-trait improvement. Indeed, individual plants with
standard leaves, > 6 branches, 100% female flowers, < 32 days to the first female flower,
and L:D > 3.0 were identified during phenotypic selection of this study. Thus, combining
all four traits is possible, and the development of inbred lines from these individuals
would allow a test of the hypothesis that combining all four traits increases yield.
The results of this study have demonstrated that MAS can be effective for the
improvement of quantitative yield components and illustrated that epistasis is an
important factor in the expression of MLB. Still, relatively little is known about the
individual QTL involved MLB and other yield components. Epistasis is known to play a
significant role among the three major genes involved in sex expression in cucumber, but
little is known about modifiers of the F locus. The characterization of interactions among
these loci and with the F locus, similar to the detection of epistatic interactions using NIL
presented herein, would provide valuable information for breeding for gynoecious sex
expression in cucumber. The clustering of QTL on Linkage Groups 1 and 6, the mapping
of QTL near known genes, and the correlation among yield component traits suggest that
some of these QTL may not be solely involved in one trait, but may be pleitropically
146involved in several yield traits. A closer inspection of these QTL by fine mapping would
help determine if relationships among traits are due to pleitropy or tight linkage. Several
BAC clones have been identified from markers linked to important yield component QTL,
and new markers have been created from several of these BAC clones that provide a
potential starting point for fine mapping and map-based cloning of yield component QTL.
In addition, the function of some of these QTL could be determined by comparative
analysis of characterized genes from other plant species. For example, the identification
and mapping of homologs of known branching genes from Arabidopsis or corn (Zea
mays), may help determine if the primary function of QTL associated with MLB is
indeed the production of branches. The characterization of the QTL associated with yield
components, including a knowledge of their function, will help define the relationships
among yield components and provide valuable information for trait manipulation by
selection.
The conversion of random amplified polymorphic DNA (RAPD) markers to
SCAR and SNP markers utilizing the methods described herein was highly successful.
Although the new markers were verified by co-segregation with the original RAPD from
which they were derived, they have not been placed on the current cucumber genetic
linkage map. In addition, these new markers have not been tested for their utility in
germplasm other than Gy-7 × H-19 crosses. The conversion of several RAPD markers
on the current map that were not included in the conversion study is recommended and
should be highly successful utilizing the methods described herein.
147The frequency of SNPs identified in this study (1 SNP every ~210 bp) and the
recovery rate of SNP-based markers in sequences that contain SNPs (80%) suggest that
the recovery rate of new markers based on SNPs in cucumber is expected to be much
greater than that previously reported for RAPDs (4.8%; Serquen et al. 1997) and SSRs
(7.2%; Fazio et al. 2002). Sequences for SNP identification can be obtained from current
markers, or by BAC clones. Sequences obtained from a randomly selected set of BAC
clones should provide SNP markers that are randomly distributed throughout the
cucumber genome, while sequences from BACs in targeted regions (i.e., linked to
important QTL), will provide physically clustered SNP markers useful for fine mapping
and QTL dissection. The creation of allele-specific markers based on SNPs utilizing the
four marker design approaches reported herein was highly successful and applicable to
almost any SNP. Although these methods are very useful for creating markers in specific
germplasm (i.e., Gy-7 × H-19 crosses), the creation of SNP markers that can distinguish
any allele would be useful for a wide range of germplasm. Several methods are available
(i.e., differential hybridization or heteroduplex analysis) to create broadly applicable SNP
markers. The new markers created herein will increase the efficiency of MAS in
cucumber, while the marker development methods will be useful to create new markers
in cucumber.
Literature Cited Fan Z, Robbins MD, Staub JE (2006) Population development by phenotypic selection with subsequent marker-assisted selection for line extraction in cucumber (Cucumis sativus L.). Theor Appl Genet 112:843-855
148Fazio G, Staub JE, Chung SM (2002) Development and characterization of PCR markers in cucumber. J Am Soc Hort Sci 127:545-557
Serquen FC, Bacher J, Staub JE (1997) Mapping and QTL analysis of horticultural traits in a narrow cross in cucumber (Cucumis sativus L.) using random-amplified polymorphic DNA markers. Mol Breed 3:257-268
149
Appendices
Appendix A. Means and linear response at two planting dates (June 23, 2004 and July 7, 2004) of five traits in four base cucumber populations (C0) of cucumber which underwent three cycles of recurrent mass selection (C1-C3) using three breeding methods (see Chapter 1).
First Planting Second planting
Traita Methodb C0 C1 C2 C3 bc R2 Pd C0 C1 C2 C3 bc R2 Pd
a Traits are EAR = earliness measured as the number of fruits per plant in first harvest, GYN = gynoecy measured as the percent female flowers in the first ten nodes, L:D = fruit length to diameter ratio measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests, MLB = multiple lateral branching measured as the number of lateral branches (at least three internodes long) on the mainstem in the first 10 nodes, and Yield measured as the number of fruits per plant averaged over four harvests b Methods are MAS =selection by marker, PHE = phenotypic selection, and RAN = random mating (no selection) c Slope of linear regression of means over cycles d P-values from F-tests of linear response to selection e Populations were created by intermating four inbred lines, and then bulking by the maternal parent (Figure 1.1)
152
Appendix B. The linear response of selection for five traits in four cucumber populations by marker (MAS), phenotype (PHE), and random mating (no selection; RAN) over three cycles. The five traits are earliness (measured as the number of fruits per plant in first harvest), gynoecy (measured as the percent female flowers in the first ten nodes), fruit length to diameter ratio (measured as the mean length to diameter ratio of 5-10 randomly selected fruit averaged over three harvests), multiple lateral branching (measured as the number of lateral branches of at least three internodes long on the mainstem in the first 10 nodes), and yield (measured as the number of fruits per plant averaged over four harvests; see Chapter 1).
Earliness (EAR)
Population 1
0.40
0.90
1.40
1.90
2.40
0 1 2 3
Frui
t per
pla
nt (1
st h
arve
st)
Population 2
0.40
0.90
1.40
1.90
2.40
0 1 2 3
Population 3
0.40
0.90
1.40
1.90
2.40
0 1 2 3
Selection cycle
Frui
t/pla
nt (1
st h
arve
st)
Population 4
0.40
0.90
1.40
1.90
2.40
0 1 2 3
Selection cycle
MAS PHE RAN G421 H19Parent Vlassett Linear (MAS) Linear (PHE) Linear (RAN)
153
Gynoecy (GYN)
Population 1
0.0010.00
20.0030.0040.0050.00
60.0070.0080.00
90.00100.00
0 1 2 3
Perc
ent f
emal
e flo
wer
s
Population 2
0.0010.00
20.0030.0040.0050.00
60.0070.0080.00
90.00100.00
0 1 2 3
Population 3
0.0010.00
20.0030.0040.0050.00
60.0070.0080.00
90.00100.00
0 1 2 3
Selection cycle
Perc
ent f
emal
e flo
wer
s
Population 4
0.0010.00
20.0030.0040.0050.00
60.0070.0080.00
90.00100.00
0 1 2 3
Selection cycle
MAS PHE RAN G421 H19Parent Vlassett Linear (MAS) Linear (PHE) Linear (RAN)
154
Mean fruit length to diameter ratio (L:D)
Population 1
2.50
2.70
2.90
3.10
3.30
3.50
0 1 2 3
Mea
n L:
D
Population 2
2.50
2.70
2.90
3.10
3.30
3.50
0 1 2 3
Population 3
2.50
2.70
2.90
3.10
3.30
3.50
0 1 2 3
Selection cycle
Mea
n L:
D
Population 4
2.50
2.70
2.90
3.10
3.30
3.50
0 1 2 3
Selection cycle
MAS PHE RAN G421 H19Parent Vlassett Linear (MAS) Linear (PHE) Linear (RAN)
155
Multiple lateral branching (MLB)
Population 2
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 1 2 3
Population 3
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 1 2 3
Selection cycle
Num
ber o
f lat
eral
bra
nche
s
Population 1
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 1 2 3
Num
ber o
f lat
eral
bra
nche
s
Population 4
0.00
1.00
2.00
3.00
4.00
5.00
6.00
0 1 2 3
Selection cycle
MAS PHE RAN G421 H19Parent Vlassett Linear (MAS) Linear (PHE) Linear (RAN)
156
Yield
Population 1
1.401.50
1.601.701.801.90
2.002.102.20
2.302.40
0 1 2 3
Frui
t/pla
nt (m
ean
of fo
ur h
arve
sts)
Population 2
1.401.50
1.601.701.801.90
2.002.102.20
2.302.40
0 1 2 3
Population 3
1.401.50
1.601.70
1.801.902.00
2.102.20
2.302.40
0 1 2 3
Selection cycle
Frui
t/pla
nt (m
ean
of fo
ur h
arve
sts)
Population 4
1.401.50
1.601.70
1.801.902.00
2.102.20
2.302.40
0 1 2 3
Selection cycle
MAS PHE RAN G421 H19Parent Vlassett Linear (MAS) Linear (PHE) Linear (RAN)
157Appendix C. Sequences of RAPD bands used to make SCAR primers. Sequences are presented in FASTA format with the name of the RAPD marker, and the parental line (Gy-7 or H-19) from which the band was produced in parentheses. Incomplete sequences are indicated by “FORWARD ONLY” or “REVERSE ONLY”. Sequences that correspond to a band other than the polymorphic RAPD band (as determined by the segregation pattern of the SCAR created from the sequence and the original RAPD marker) are indicated by “DOES NOT MATCH RAPD” (Table 2.1; see Chapter 2). >BC231 (H-19) agggagttccaaacttttcagtacagtaaggggtattaaaatacaaaggaaggttgttgctgtcaaaaac atatgttctactacccacaatcatgcatacctatcatgattttttccttggaggacattcagggttccta agaacgtacaagaggatgaccggagagatttactaggaaggaatgaagaaagacataaagaaatactatg aagaatgtcttgtatgccaaaggaataaaacacttgcattatcacctgccggattgctgctgccacttga gattcctaacacagtgtggtcggatatttccatggattttatagatggacttccgaaatcagctggaaaa gaagtgattgttgtggttgtagatcgtctgagtaagtatgcacattttatagctataaatcatccttaca tgggcagcttcgtagcagaagtttttgttaaagaagtggtgagattgcacgggtatccctagtcaattat atcggatcgagacaaaatctttattagtcatttttgaaaggaaatgttcaggttagtgggaactaaaatg aatagaaggaaatgttcaggttagtgggaactaaaatgaatagaaggaaatgttcaggttagtgggaact aagcccatatgcaccaagttcgaatatagaacttgcttcttactgaaagaataatgattatgggacttgc tataagggtatccaatattcgaagtaaatgaatgattacgaaatgattgtttcctcagtgtgaagaacaa caaaaaagaaaaaaggataaaaataaaaaaacaaactttctttttgaaattctaaaagaatccaaaaaga ctttaaaaggtgcaaccaagaaaggaggtcaaagaagggctctgatagatccccctatcatacgtcaata tagttgaaagaaaaagagagagatcacaggggaactccct >BC388 (Gy-7) cggtcgcgtccttagaccaaccacccaagttgatgcgctgtgcatggcagcagacctgagtttgtatgag aggacgatgttggccaaggctgtagagaaggggccaacttcgggacataaacggaagactgagcagcagc ctatagtagcaccacaaaggaatttgagatcaatggtttgttccaacggcactgacaacaagctactcag acaggccaaaccttgaaggtactattcatgactcttaattatggaagacctcactcgggtcgttacctag caggaagtggagtatgttacaaatgcagacatagaggttgacttcgccatagagcttctagggcttccgc gctgcagagaggttgacttcgccatagagctcgaaccaggtattggtctcatatcgagagcctcatacag aatgacgcgaccg >BC403 (Gy-7) ggaaggctgtcttccttatgtcttcatctcgtacccggatttgatggtaactcgatttcaaatctctctt ggagaaaacacttgagccattcaactcatccaacagctcatcaatcattggaattgggaatttgtcaggt atcgtagctcgattcaaagccctataatcgacgcaaaatctccaaccgccatccttcttttttactagaa tcaccgggctggaaaaggggctgatgctcggtcgtatgatacctgaagtcaacatttttctgtgcgtgtg ggtaacggtatggtcttacattaatggggtcggtaccttccttcagttggattctatgatcaatctgtcc cattggaggcagctcgtttggcataacgaatacatcttcgaattcattttggagttgttcaatttcgggt tgtatttcttctattgattctgttgccatcaacatccgattggtttcgggaattcccatagctctgaagt caactagaaatccttgatcatctgattgccacgttttaaccagcatcttcaaagatatttccattctggt cagtgagggatctcttttcaagatgacgttagtgcccccgacaaagaatgtcatggttaacgctttccaa tcaacagtcatcaccccttgctttcgaagccactacattcctagtaccatatctaagttgcctaattcca gcgataagaaatcctcaatgatggtcagtactggcagcccaactatgatgtccttgcacatccctcgacc ttgcacagccttcc >BC469 (H-19) ctccagcaaactaacaattgagggaaaaatacatacctgtactaggaatacatccaaaacaagaacaggg ctattcccaggggaggcagcctggtaatatggaattgctgaggaagtaagagccttagcaactaacaatc ccaccgttgcttgcatccctagaatggcagcacccatgcctaacaggtttactactattccatttctgag
169Appendix D. SCAR marker database. An html (web-based) database of the SCAR markers created in Chapter 2, as well as additional SCAR markers, is available that contains primer sequences, annealing temperature gradient PCR (ATG-PCR) profiles, and spreadsheets with information on each of the SCAR markers. The structure and contents of this database are explained in the following screenshots and can be accessed at: http://www.vcru.wisc.edu/staublab/Matt/SCAR%20web%20page2/Scar%20database.htm Database homepage:
Clicking on the lin ll mark m S ct” on t
wink “Info on a ers created fro CAR proje he left will
open the follo g page:
170he na ar ill TG-P R gel photo er seq at mark
Clicking on t me of a SCAR m ker on the left w display an A Cand the prim uences of th er:
Information can be found in the database for the following SCAR markers: AA9SCAR AJ2SCAR AT1SCAR BC551SCAR N8SCAR AA9bSCAR AJ6SCAR AT2SCAR BC592SCAR P13SCAR AB14SCAR AJ18SCAR AT15SCAR BC600SCAR P14SCAR AC9SCAR AK5SCAR AW14SCAR C1SCAR R13SCAR AC17SCAR AK16SCAR B12-1SCAR C7SCAR T2SCAR AD12SCAR AM2SCAR BC231SCAR C10SCAR U15-2SCAR AD14SCAR AN5SCAR BC388SCAR D11SCAR W7SCAR AF7SCAR AO7SCAR BC403SCAR F4SCAR W7-2 SCAR AF15SCAR AO8SCAR BC413SCAR H13SCAR W7aSCAR AG1SCAR AO12SCAR BC450SCAR I1SCAR W7bSCAR AG1-1SCAR AO14SCAR BC469SCAR I20SCAR Y10SCAR AG17SCAR AQ18SCAR BC515SCAR J5SCAR
AI4SCAR AR13-1SCAR BC519SCAR K7SCAR
AI10SCAR AR13-2SCAR BC523SCAR L19SCAR
AJ1SCAR AS5SCAR BC526SCAR M8SCAR
171Appendix E. BAC clone end sequences used to create SCAR markers in Table 2.2 (Chapter 2). Sequences are presented in FASTA format with the sequence name followed by the number of bases in parentheses. Sequences are named with a one or two letter designation for the marker that hybridized to the BAC clone (AJ = AJ6SCAR, B = BC523SCAR, C = CSWCTT14, L = L19SCAR, M = M8SCAR, and W = OP-W7-1), the number of the clone that was sequenced, BE for “BAC end”, and L or R to signify the left or right end of the BAC clone, respectively (L2 or R2 indicates the second attempt to obtain the sequence). AJ-1-BE-L, for example, is the sequence of the left end of the first clone sequenced that hybridized to AJ6SCAR. >AJ-1-BE-L (861) CCTGTACAAGGAAGGTCAAGGCGCCAGGTTAGAAGTTAGAAGGAAGCGCAGTAAGCAGCTCCCTACTAGTTGGAGAGGCATCCCATCCCTTGAGCTTGAAAGTTCCTTGTCTATTATGATTCTGCTCTGTAGCACTTGTCCGCCAGAGCGGTCCAAGCTTCGCCTCGTGGGATTATTCCTTATTCTTTCATTTCATTCTCTCTTTCGGATCGGAATAAAGGATCGGAAGAAAGGGAAGAGAAAGAGGCAGAAAGGCGAGATGGAATTGAGTCGAGACCACCCACGCTTGATTTTGACTGTATCAACTCTGAATCACTATCCTTAATTGTAAGAAGGCAAGGGCCTTTGACGAAGTTGACTTTCGATCAGAAGCGAGACCAGAGCGGAGAGGAGTTCAACTGCGAGACTGTAACTTCAAAAAGGTGAATTTCCTCGATTCCGGAAGACACTCCCGGGCTCACTCGCGAAAACAAGGTCCTGTGAGGACTGATTCCGTAATTGTGCACTCATGGTCGATCGCTTAGCTTTACTACCTTTTTCTTGAAATAGGGAGCATCTAAAAATGCACTTGCACTCCTATTAGATTAAATCGCTTCTCCGCACTCTTCCCCTGAAAAAGTAAGTCGAAGATATGAAGAAGAAAAGATGAAAAGATATAATAAGATAATGATATTAGAACTCAAATCCCCCCTCCCCCAGTGCAAAGAATATATAGAAGGCCCATATGCGAGGCGGAAAGAGCCGTAAACTACTTTATTTATCCTTTCCGCGTAGCGACTCTGCTCTTGAAACTTCTTATTAAGCTACTTAGGACCGGAGCGACTCTATCTTTATTGAGAGCGTAGCTACTTAATCTGCCCG >AJ1-BE-R2 (822) ATTCGTCTGAGAAAGCAATTCCATTAGAATGCAGCGGAGTCCTTTCCGCAAGGGGAATATATGCGAACTGCAACACATTCATGTCCATTTTCCTCAATCGGGAGGCAGGAATTCTTCTTATATCTATTCTTGGATTTTCGTCGTTATGTAATAAATAGAGGTCGCTCTCTTGGGTTCCCTCAATGAAAAGGACGGATTTCAGTTTCCGCTTCCTAAAAAGAGGTTTTTCTATAAAAGGGAGTTCTAGTTTCGTTTTGGATTGGGCCGGCGGCCCTGTCTCCACGAATAATAGGCTGCCCTTCGGCACATGCAACCATACATGGTTGCGGCTATTCTGTCCTCTTGTAGGTCTTCGCAGTGGGGGGAGTCGAATGCTAAAGGCATAATTCTTGTACCCCTTTCTTTCTAAAGCTTTTGGGCAAGAGTTTTCCCATCAAAAGGTACGTTAGCCGTTAAAAGTCTTTTCATATTTGGCATCTAGCCTGCACTCCTGCCCATCAAGAAGAGGGAGAAGGTCCCTCGCGGAGGAGGGCTGCTCGCCTTCATCATATTCGTTTTCTTCGATGAACTCGTCGAAGACCCAAAAATCAAAGAAATCATGGGCTCCGGAGTAGTCATTCCTTCTTTTCCCGACTCCTTAAGCCAATCTTTTTAAGAAGACTTATAAGGAGGGTTTTCCGCGTATTTGATCTTCCGTAGAGAAAGAGCAGCCTTGTCTTTATTGGCCGCTCGAAGAGAAGGTTGGTTGCTATCCAATTCAGCAGGAGATACTTGTCTTGCAAGAATTCGTCAGTGTATTCTTGGGGCCGGCCTTTCCTTTCT >AJ-2-BE-L (824) ACCCTTCCTTTCAAGTCTCCCTTCCTTTCTTCCCGAAAGCAATGAAGGTATAAGATTCACCCAAAGGGCTAACTTCACTAGTTCCCTGGGATCACAGAACTAGCGTGAAAAGACCGGGAGTCCTAGTCCACCAAAGCGTGCAGATCCATCATGGCAAAGACCGGATCTGATCTAACTCTAAGATAAGAGCGGATTGGCCGACTGAAAAAGTATGAAAATCAGTTCCACAGGTCCGAAGGACTATCCATATTAACCCGTCTGTCCGAAGGACCTGTGACGCCTTGTTCATTGATTCTTTTTTCTTTCAGTGAATTAGAATGTGGCTCCTCGACCGAAAGGTTTATGACAAGTCTTCTTTCCTAAGAGTGAGTCAGTCGCTGTGGCAGGCACTAGTCAAATAGGGGAAGTCATATGCTCGGGAGAGGGACAAAGGAAAGCGGGAGAGCAAGCCTTGCTCTTTCAGTTGCCGTTGAAGAAGAAGGCAAAAACAAAGGCGCCATAAGACAGGTCCGAAGGACTATCCATTAGGGGATGAGAGGCCTATCCATTAGAGAATTTGACTTGCGACGTCCGCTTTAAGAGCCTAGAGGTCAATTTTTCTTAGGGGTGCTTTCGTGGAATCATTTCAGAGAAAAATCACTGTCTTTCTCTTTCTCACAAAGGACCGGAGAAGGGCCTGACCTCGATCCGACCCGAATGAGTTACCGGTG
a Length of the PCR product using the F (forward) and R (reverse) primer for each sequence P
b The number of base pairs from the beginning of the sequence to the start position of the primer (where the 5’ base anneals to the sequence) P
c Length of the primer in base pairs P
d Melting temperature as calculated by Primer 3 P
e GC base content of primer
186Appendix G. SNPs identified between Gy-7 and H-19 in markers from two sequencing sources [sequencing of SCAR fragments (SCAR) and utilizing markers to identify BAC clones for sequencing (BAC end); see Chapter 2]. Asterisks (*) in either the Gy-7 or H-19 sequence (cucumber parental lines used in map construction) indicate an insertion/deletion (indel). Marker Sequence source SNP SNP posa Gy-7 sequence H-19 sequence AB14SCAR SCAR 1 423 G A 2 451 T A 3 459 C T 4 491 C T 5 627 G A 6 738 G A 7 760 A T AC17SCAR SCAR 1 235 A G AC9SCAR SCAR 1 95 G A 2 135 A G 3 270 T A 4 274 A G 5 287 A G 6 322 G A 7 557 T C AD14SCAR SCAR 1 36 A C 2 108 G T 3 157 A T 4 403 G A 5 410 ** TC 6 497 G A 7 500 C T 8 578 A G 9 589 A G 10 602 C T 11 629 C T 12 648 C A 13 663 G A 14 784 A T 15 809 G A AI4SCAR SCAR 1 172 G T 2 460 G A 3 499 T C 4 655 A G 5 760 C T 6 893 A G 7 914 T C 8 985 C G 9 1067 A *
187Marker Sequence source SNP SNP posa Gy-7 sequence H-19 sequence AK16SCAR SCAR 1 153 C A 2 173 C G 3 841 C T 4 985 A T 5 1081 C G 6 1106 T C 7 1133 A G 8 1241 G C 9 1263 A G AT1SCAR SCAR 1 233 G A 2 283 G C 3 654 C G 4 658 G T 5 771 C A C10SCAR SCAR 1 255 G A 2 508 A T 3 643 T C C1SCAR SCAR 1 97 T G 2 247 A G D11SCAR (Forward) SCAR 1 270 A T 2 380 T * D11SCAR (Reverse) SCAR 1 116 C T 2 135 G A 3 263 A G M8SCAR (Reverse) SCAR 1 146 G A 2 183 * T 3 234 G A N8SCAR SCAR 1 398 * A 2 431 G A 3 497 A G 4 545 C T 5 597 T C 6 676 G A 7 685 C T 8 699 G A 9 843 A G 10 923 A G 11 926 T C W7SCAR SCAR 1 248 T G 2 296 A G 3 352 A T 4 364 TA AT 5 378 G A 6 405 C T B-1-BE-L BAC end 1 237 ACCTCTCT TTCTTCCC 2 287 AAAG CAAA
188Marker Sequence source SNP SNP posa Gy-7 sequence H-19 sequence C-1-BE-R BAC end 1 103 G A 2 131 G T 3 138 G A 4 152 C T 5 171 T A 6 180 A T C-3-BE-L BAC end 1 106 C T 2 127 G A 3 132 T C 4 154 T C 5 179 AG CN 6 187 A G 7 199 G A 8 212 A G 9 231 C T 10 249 G A 11 261 C T 12 270 CT TC 13 275 G A 14 279 TG CA 15 317 T C 16 320 AA GT 17 330 TTTT TTG 18 338 A G 19 343 G N 20 346 A G 21 360 TTG ACA 22 366 T N 23 381 TC CT 24 384 CA AG 25 410 T N 26 420 A G 27 470 G A 28 484 TTCA CTTC 29 493 A G 30 495 A G 31 498 G A C3-BE-R2 BAC end 1 124 A G C-4-BE-L BAC end 1 25 * T 2 46 TTAT AATAA 3 257 ** TT 4 286 T A 5 334 T A 6 352 T C 7 391 G A 8 439 G A 9 451 C T
189Marker Sequence source SNP SNP posa Gy-7 sequence H-19 sequence C-5-BE-L BAC end 1 174 T A 2 290 T A 3 318 T C 4 393 T A C6-BE-L BAC end 1 77 A * 2 86 A G 3 93 T C 4 97 A G 5 104 T C 6 109 G A 7 121 T C 8 128 T C 9 132 T C 10 150 G T 11 156 TA AT 12 165 A A 13 168 A G 14 178 G T 15 183 T C 16 185 G A 17 187 A G 18 209 T C 19 225 T C 20 233 T G 21 247 T C 22 262 A G 23 266 T G 24 283 A G L-1-BE-L BAC end 1 76 A G 2 195 A G 3 296 A G M-4-BE-L BAC end 1 152 * A 2 328 ** AT 1 134 T C 2 213 T C M7-BE-L BAC end 1 171 G A W-2-BE-R BAC end 1 264 TTAATTTTTTTA ************ 2 326 A C 3 343 A G 4 348 T C 5 356 T G a The number of base pairs from the beginning of the sequence to the SNP
190Appendix H. Characteristics of single nucleotide polymorphism (SNP) markers (comprising an allele-specific primer based on a SNP, and a non-specific primer) and their allele-specific primers as developed in cucumber (see Chapter 2). Allele-specific primer characteristics
Sequencea SNP Marker Alleleb Allele specificityc 3' mismatchdInternal mismatche
Mismatch positionf
AB14SCAR AB14SNPG1 Gy-7 allele specific A:A G:T 2 AB14SNPG2 Gy-7 allele specific A:A C:T 2 AB14SNPG3 Gy-7 allele specific A:A T:T 2 AB14SNPH1 H-19 allele specific A:C A:G 2 AB14SNPH2 H-19 allele specific A:C G:G 2 AB14SNPH3 H-19 allele specific A:C C:C 3 AC17SCAR AC17SNPG1 Gy-7 allele specific A:C T:T 2 AC17SNPG2 Gy-7 allele specific A:C T:G 3 AC17SNPG3 Gy-7 allele specific A:C G:G 3 AC17SNPH1 H-19 allele specific G:T T:T 2 AC17SNPH2 H-19 allele specific G:T G:T 2 AC17SNPH3 H-19 allele specific G:T C:T 4 AC9SCAR AC9SNPG1 Gy-7 allele specific A:C G:A 3 AC9SNPG2 Gy-7 allele specific A:C C:A 3 AC9SNPG3 Gy-7 allele specific A:C A:G 4 AC9SNPH1 H-19 not allele specific A:C A:A 2 AC9SNPH2 H-19 not allele specific A:C G:G 3 AC9SNPH3 H-19 allele specific A:C A:G 4 AD14SCAR AD14SNPG1 Gy-7 allele specific C:A A:A 2 AD14SNPG2 Gy-7 allele specific C:A T:T 3 AD14SNPG3 Gy-7 allele specific C:A C:T 3 AD14SNPH1 H-19 allele specific A:G C:T 3 AD14SNPH2 H-19 allele specific A:G T:T 3 AD14SNPH3 H-19 allele specific A:G C:C 4 AI4SCAR AI4SNPG1 Gy-7 allele specific A:C A:A 2 AI4SNPG2 Gy-7 allele specific A:C G:A 4 AI4SNPG3 Gy-7 allele specific A:C G:A 2 AI4SNPH1 H-19 allele specific T:G A:A 3 AI4SNPH2 H-19 not allele specific T:G G:A 3 AI4SNPH3 H-19 allele specific T:G C:A 2 AK16SCAR AK16SNPG1 Gy-7 allele specific A:A T:C 3 AK16SNPG2 Gy-7 no product A:A C:C 2 AK16SNPG3 Gy-7 allele specific A:A C:C 3 AK16SNPH1 H-19 allele specific C:A G:A 2 AK16SNPH2 H-19 allele specific C:A G:A 3 AK16SNPH3 H-19 allele specific C:A G:A 4 AT1SCAR AT1SNPG1 Gy-7 allele specific C:A A:A 2 AT1SNPG2 Gy-7 allele specific C:A G:A 2 AT1SNPG3 Gy-7 allele specific C:A C:T 3 AT1SNPH1 H-19 allele specific G:G G:A 3 AT1SNPH2 H-19 allele specific G:G A:A 3 AT1SNPH3 H-19 allele specific G:G C:A 3
191 Allele-specific primer characteristics
Sequencea SNP Marker Alleleb Allele specificityc 3' mismatchdInternal mismatche
Mismatch positionf
B-1-BE-L B1LG1 Gy-7 not allele specific T:T G:T 5 B1LH1 H-19 not allele specific A:G T:G 4 C10SCAR C10SNPG1 Gy-7 allele specific A:A G:G 2 C10SNPG2 Gy-7 allele specific A:A G:A 3 C10SNPG3 Gy-7 allele specific A:A A:A 3 C10SNPH1 H-19 allele specific C:A G:A 2 C10SNPH2 H-19 no product C:A G:A 3 C10SNPH3 H-19 no product C:A A:A 3 C-1-BE-R C1RG1 Gy-7 allele specific A:A G:T 2 C1RG2 Gy-7 allele specific A:A C:T 2 C1RG3 Gy-7 allele specific A:A T:T 2 C1RH1 H-19 not allele specific A:G C:T 4 C1RH2 H-19 not allele specific A:G A:A 3 C1RH3 H-19 not allele specific A:G T:T 4 C-3-BE-L C3LG1 Gy-7 not allele specific A:C G:T 3 C3LG2 Gy-7 no product A:A A:C 2 C3LG3 Gy-7 allele specific A:A A:C 2 C3LH1 H-19 no product *:T C:G 4 C3LH2 H-19 not allele specific G:T A:C 2 C3LH3 H-19 allele specific A:C C:A 2 C3-BE-R2 C3R2G1 Gy-7 allele specific A:C C:C 3 C3R2G2 Gy-7 not allele specific A:C A:C 2 C3R2G3 Gy-7 allele specific A:C C:C 2 C3R2H1 H-19 allele specific C:A C:C 2 C3R2H2 H-19 not allele specific C:A T:C 2 C3R2H3 H-19 allele specific C:A A:C 2 C-4-BE-L C4LG1 Gy-7 allele specific G:T G:G 2 C4LG2 Gy-7 allele specific G:T G:A 3 C4LG3 Gy-7 allele specific G:T A:A 3 C4LH1 H-19 allele specific A:A C:A 2 C4LH2 H-19 allele specific A:A A:A 2 C4LH3 H-19 allele specific A:A G:A 2 C-5-BE-L C5LG1 Gy-7 allele specific A:A G:T 2 C5LG2 Gy-7 no product A:A C:T 2 C5LG3 Gy-7 allele specific A:A T:T 2 C5LH1 H-19 allele specific G:T C:A 2 C5LH2 H-19 allele specific G:T A:A 2 C5LH3 H-19 allele specific G:T A:C 3 C-6-BE-L C6LG1 Gy-7 not allele specific A:C G:T 4 C6LG2 Gy-7 not allele specific T:G G:G 2 C6LG3 Gy-7 not allele specific T:G A:G 2 C6LH1 H-19 not allele specific A:C C:A 6 C6LH2 H-19 not allele specific C:A C:A 5 C6LH3 H-19 not allele specific C:A A:G 3
192 Allele-specific primer characteristics
Sequencea SNP Marker Alleleb Allele specificityc 3' mismatchdInternal mismatche
Mismatch positionf
D11SCAR D11SNPG1 Gy-7 allele specific C:A T:T 2 D11SNPG2 Gy-7 allele specific C:A C:T 3 D11SNPG3 Gy-7 allele specific C:A C:T 2 D11SNPH1 H-19 allele specific A:A C:A 2 D11SNPH2 H-19 allele specific A:A G:A 2 D11SNPH3 H-19 allele specific A:A A:A 2 L-1-BE-L L1LG1 Gy-7 allele specific A:C A:A 2 L1LG2 Gy-7 not allele specific A:C C:A 4 L1LG3 Gy-7 allele specific A:C G:A 2 L1LH1 H-19 not allele specific G:T A:G 2 L1LH2 H-19 allele specific G:T G:G 2 L1LH3 H-19 allele specific G:T A:C 3 M-4-BE-L M4LG1 Gy-7 allele specific A:A G:G 2 M4LG2 Gy-7 allele specific A:A T:T 3 M4LG3 Gy-7 allele specific A:A A:G 2 M4LH1 H-19 allele specific A:C A:A 2 M4LH2 H-19 allele specific A:C A:G 3 M4LH3 H-19 allele specific A:C A:G 2 M-4-BE-R M4RG2 Gy-7 not allele specific A:G A:A 3 M4RG3 Gy-7 not allele specific A:G G:A 3 M4RG8 Gy-7 not allele specific A:G C:C 2 M4RH1 H-19 not allele specific C:A A:G 2 M4RH2 H-19 not allele specific C:A A:G 3 M4RH5 H-19 not allele specific C:A A:G 4 M7-BE-L M7LG1 Gy-7 allele specific A:C C:C 2 M7LG2 Gy-7 no product A:C T:C 2 M7LG3 Gy-7 allele specific A:C A:C 2 M7LH1 H-19 allele specific A:C A:A 2 M7LH2 H-19 allele specific A:C G:A 2 M7LH3 H-19 allele specific A:C C:T 3 M8SCAR M8SNPG1 Gy-7 allele specific C:A G:A 3 M8SNPG2 Gy-7 allele specific C:A G:A 4 M8SNPG3 Gy-7 allele specific C:A A:A 3 M8SNPH1 H-19 allele specific T:G A:A 2 M8SNPH2 H-19 not allele specific T:G G:G 3 M8SNPH3 H-19 allele specific T:G G:A 4 N8SCAR N8SNPG1 Gy-7 allele specific G:T C:A 2 N8SNPG2 Gy-7 allele specific G:T A:G 3 N8SNPG3 Gy-7 allele specific G:T T:C 4 N8SNPH1 H-19 allele specific C:A T:T 2 N8SNPH2 H-19 allele specific C:A G:A 4 N8SNPH3 H-19 allele specific C:A C:T 2 W-2-BE-R W2RH1 H-19 no product C:A G:A 3 W2RH2 H-19 no product C:A G:T 4 W2RH3 H-19 no product C:A C:T 2 W2RH4 H-19 no product C:A C:T 4
193 Allele-specific primer characteristics
Sequencea SNP Marker Alleleb Allele specificityc 3' mismatchdInternal mismatche
Mismatch positionf
W7SCAR W7SNPG1 Gy-7 allele specific A:G G:G 4 W7SNPG2 Gy-7 not allele specific A:G C:T 3 W7SNPG3 Gy-7 not allele specific A:G T:T 3 W7SNPH1 H-19 allele specific TA:TA C:A 3 W7SNPH2 H-19 allele specific TA:TA G:G 4 W7SNPH3 H-19 allele specific TA:TA A:G 4 a Original sequence from which the SNP was identified b Each marker was designed to amplify either the Gy-7 or H-19 allele during PCR c Allele specific = the SNP marker amplified the expected allele, not allele specific = the SNP marker amplified both alleles, no product = no product was detected after PCR d The final base at the 3’ end of the allele specific primers matches only one of the alleles (line Gy-7 or H-19). The mismatch between the non-target allele is reported as the sequence of primer:template e The allele specific primer contains a mismatch to both alleles near the 3’ end of the primer. The mismatch is reported as the sequence of primer:template f The position of the internal mismatch as the number of base pairs from the 3’ end of the primer
Appendix I. Prim
SNP Ma
194er names and sequences of SNP markers as developed in cucumber (see Chapter 2).
rker Allele-specific primer name Allele-specific primer sequence (5’ to 3’)
Non-specific primer name Non-specific primer sequence (5’ to 3’)
200Appendix K. The effect of increasing the number of quantitative trait loci (QTL) on the number of lateral branches in two leaf types of cucumber in two locations as determined by near-isogenic lines (NIL; Table 3.2). Lines indicate the incremental addition of QTL, and the comparisons in the legends refer to means comparisons performed (Table 3.4; Chapter 3).
Little leaf NILHancock, Wisc.
NIL-146
NIL-1246
NIL-12456NIL-1236
NIL-136
NIL-1356 NIL-12356
4.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
2 3 4 5 6
Num
ber
of la
tera
l bra
nche
s
Standard leaf NILHancock, Wisc.
NIL-0
NIL-23
NIL-235
NIL-25
NIL-36
0.0
0.5
1.0
1.5
2.0
2.5
0 1 2 3 4
Little leaf NILArlington, Wisc.
NIL-12456
NIL-1246NIL-146
NIL-1236 NIL-12356
NIL-1356
NIL-1364.5
5.0
5.5
6.0
6.5
7.0
7.5
8.0
8.5
2 3 4 5 6
Number of QTL
Num
ber o
f lat
eral
bra
nche
s
Standard leaf NILArlington, Wisc.
NIL-235
NIL-23
NIL-0
NIL-25
NIL-36
0.0
0.5
1.0
1.5
2.0
2.5
0 1 2 3 4
Number of QTL
Comparisons A & B Comparisons C & D
Comparisons E & F
Comparisons G & H Comparisons I & J NIL-36
201
Source of variation DFa Mean square F value P value
Appendix L. Analysis of variance (ANOVA) table of a test for main effects and interactions on the number of lateral branches in cucumber (MLB; Chapter 3). Effects examined were location (Hancock, Wisc. and Arlignton, Wisc.), replication (reps), with-in row spacings (10, 15, and 20 cm between plants), leaf type (standard and little leaf) and genotype [near-isogenic lines (NIL) that vary in the number of QTL (Table 3.2) for MLB]. The coefficient of variation (CV) was 19.6%.