1 CHARACTERIZATION OF THE TOMATO BRANCHED-CHAIN AMINO ACID AMINOTRANSFERASE ENZYME FAMILY MEMBERS By GREGORY S. MALONEY A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2009
92
Embed
CHARACTERIZATION OF THE TOMATO BRANCHED ...ufdcimages.uflib.ufl.edu/UF/E0/04/11/40/00001/maloney_g.pdfKID Keto acid decarboxylase KIV α-Ketoisovalerate K m Michealis-Menten constant
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
1
CHARACTERIZATION OF THE TOMATO BRANCHED-CHAIN AMINO ACID AMINOTRANSFERASE ENZYME FAMILY MEMBERS
By
GREGORY S. MALONEY
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
LITERATURE REVIEW ..............................................................................................................13
Tomato as a Research Tool for Flavor ...................................................................................13 Tomato Flavor ........................................................................................................................16 Branched-Chain VOC Metabolism .........................................................................................18 BCAA Metabolism in Plants ..................................................................................................20 Branched-Chain Aminotransferases in Plants ........................................................................23
CHARACTERIZATION OF TOMATO BCATs ..........................................................................31
Cloning of SlBCAT cDNAs ....................................................................................................31 Expression Analysis of SlBCATs ............................................................................................32
Subcellular Localization of SlBCATs ....................................................................................33 Functional Verification of SlBCATs by Bacterial Complementation ....................................34 BCAT Enzyme Assays ...........................................................................................................35 Analysis of SlBCAT1 and SlBCAT3 Transgenic Fruit ............................................................37 Conclusion of Results .............................................................................................................38
BRANCHED-CHAIN VOLATILES IN TOMATO .....................................................................49
Rationale and Background ......................................................................................................49
Results of Substrate Feeding ..................................................................................................50 BCAA and Branched-Chain Volatile Loci .............................................................................56
DISCUSSION OF RESULTS .......................................................................................................63
Diversity of SlBCAT Family ...................................................................................................63 Significance of Substrate Feeding ..........................................................................................66 Concluding Remarks ..............................................................................................................71
MATERIALS AND METHODS ...................................................................................................74
Cloning of SlBCATs...............................................................................................................74
Protein Production and Purification........................................................................................75 Enzyme Assays .......................................................................................................................76 Volatile Collection and Analysis ............................................................................................77 Microscopy and Subcellular Localization ..............................................................................78 Metabolite Feeding .................................................................................................................79 GC-MS Analyses of Nonvolatile Plant Metabolites ...............................................................79 Analysis of [U-13C6]Leucine-Labeled Samples .....................................................................79 Expression Analysis ................................................................................................................80 E. coli Complementation ........................................................................................................81 Amino Acid Analysis of Tomato Fruit by GC-MS ................................................................81 Statistical analysis ...................................................................................................................82
LIST OF REFERENCES ...............................................................................................................84
Table page 1-1 Flavor volatile compounds impacting the perception of ripe tomato fruit flavor ...............30
2-1 Measurement of E. coli cell culture growth rate .................................................................45
2-2 Kinetic parameters of SlBCATs ..........................................................................................46
2-4 Levels of free amino acids in red ripe fruit of M82 and SlBCAT over-expression lines. ...................................................................................................................................48
3-1 Label accumulation in metabolite pools following incubation with [U-13C]leucine ...........61
3-2 Occurrences of loci containing BCAA and branched-chain volatile phenotypes ...............62
4-1 Primer sequences used in this study. ...................................................................................83
1-2 Reported pathway of BCAA catabolism in plants leading to TCA cycle intermediates. ....28
1-3 Schematic of S. lycopersicum and S. pennellii introgression line population and corresponding branched-chain volatile loci. ......................................................................29
2-1 Evolutionary relationships of mature SlBCAT proteins......................................................40
2-2 Evolutionary relationships of mature SlBCAT and AtBCAT proteins ...............................41
2-3 Quantification of SlBCATs RNA in different tissue types. .................................................42
2-5 Growth complementation of E. coli ΔilvE/ΔtyrB mutant cells expressing SlBCAT3 and 4..........................................................................................................................................45
2-6 Analysis of transgenic fruit SlBCAT transcript levels .........................................................47
3-1 Leucine and KIC feeding .....................................................................................................58
3-2 Isoleucine and KMV feeding ...............................................................................................59
3-3 Valine and KIV feeding. ......................................................................................................60
4-1 Hypothesized pathways forming branched-chain volatiles from BCKAs ...........................73
Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy
CHARACTERIZATION OF THE TOMATO BRANCHED-CHAIN AMINO ACID
AMINOTRANSFERASE ENZYME FAMILY MEMBERS
By
Gregory S. Maloney
May 2010
Chair: Harry J. Klee Major: Plant Molecular and Cellular Biology
Branched-chain amino acids (BCAAs) are essential to animals and are synthesized in
plants from branched-chain keto acids (BCKAs), but their metabolism in plants is not completely
understood. The interface of BCAA anabolism and catabolism lies with branched-chain
aminotransferases (BCAT). In this study six BCAT genes from the cultivated tomato species
Solanum lycopersicum were identified and characterized. Quantitative RT-PCR showed that
SlBCAT1, 2, 3, and 4 were expressed in multiple plant tissues, while SlBCAT5 and 6 transcripts
were undetectable. SlBCAT2 and 3 were expressed nearly equally in all tissues, while SlBCAT1
was expressed most highly in ripening fruit and SlBCAT4 was expressed primarily in
inflorescences. SlBCAT1 and 2 are located in the mitochondria, SlBCAT3 and 4 in chloroplasts,
while SlBCAT5 and 6 are located in the cytosol and vacuole, respectively. Expression of
SlBCAT3 and 4 were able to restore growth of Escherichia coli BCAA auxotrophic cells, while
expression of SlBCAT1 and 2 were less effective. All SlBCAT enzymes were active in the
forward (BCAA synthesis) and reverse (BCKA synthesis) reactions. SlBCAT3 and SlBCAT4
exhibited a preference for the forward reactions while SlBCAT1 and SlBCAT2 were more active
in the reverse reactions. Over-expression of either SlBCAT1 or SlBCAT3 in tomato fruit did not
12
significantly alter amino acid levels or branched-chain volatile emissions. BCAAs and BCKAs
were applied to fruit samples which were then analyzed for branched-chain volatiles. The results
support a model in which these volatile compounds are synthesized primarily from BCKAs and
not BCAAs in tomato fruit, unlike their synthesis from BCAAs in yeast.
13
CHAPTER 1 LITERATURE REVIEW
Tomato as a Research Tool for Flavor
The cultivated tomato, Solanum lycopersicum, is the most valuable vegetable crop in
Florida, the leading US producer of fresh market tomatoes (Lucier, 2009). Consumers’ opinions
of the average grocery store tomatoes reflect poor flavor compared to the more flavorful
heirloom tomatoes. This is due in part to the fact that tomato breeding has focused principally on
economically important traits such as yield, appearance, maturity, and disease resistance. It is
difficult and impractical for breeders to select for better flavor in tomatoes because it is a highly
multigenic trait that is greatly influenced by environmental conditions and the composition of a
tasty fruit is not completely understood. Commercial tomatoes are often picked green, stored and
shipped at low temperatures, and bruised during handling, all of which have been shown to
decrease positive flavor compounds (Maul et al., 2000). Tomato varieties targeted towards home
gardeners tend to have superior flavor but are not grown commercially because they yield less
fruit, are softer, ripen too quickly, and do not ship or store well. These issues concerning fresh
tomato flavor quality may have the best chance of being resolved following research that focuses
on the genetics and biochemistry of flavor.
Tomato flavor comes from a blend of three chemical classes: acids, sugars, and volatile
organic compounds (volatiles). Flavor is also determined by the interactions of these chemicals
with alcohols, glycosidic bonds, and other compounds produced during fruit ripening (Tandon et
al., 2000; Ortiz-Serrano and Gil, 2007). Volatiles, which outnumber by far all other classes of
flavor compounds, are sensed by the olfactory system and are important in the diversity and
complexity of fresh uncooked tomato flavor.
14
Volatiles are low molecular weight hydrophobic compounds derived from fatty acids,
carotenoids, and amino acids, which diffuse readily through fruit tissue to be released into the
atmosphere. The difficulties in studying ripe tomato flavor include the multitude of metabolic
pathways leading to volatile synthesis, their interconnectivity, and the lack of genes identified in
these pathways. However, some information concerning these pathways is known and serves as a
starting point for research. For instance, the C6-alcohol and aldehyde flavor volatiles are thought
to be derived from fatty acids (Chen et al., 2004), the ketone volatiles from carotenoids (Simkin
et al., 2004), and the amino acid-derived volatiles from alanine, phenylalanine, and branched-
chain amino acids (Tressl and Drawert, 1973; Tieman et al., 2006b).
The study of tomato flavor was previously undertaken (Tieman et al., 2006) by utilizing
the S. pennellii exotic introgression line population created by Dani Zamir in Israel. This
population is a series of 76 near isogenic introgression lines (ILs) that span the whole tomato
genome and overlap to form over 100 bins. It was created from a cross of S. pennellii (accession
code LA716), a wild species, and S. lycopersicum cv. M82, a common processing tomato (Eshed
and Zamir, 1995)
The formation of an introgression population takes about 10 generations of backcrossing.
In brief, the two parent species, in this case S. pennellii and M82, are crossed. The sites of
recombination are determined using existing RFLP markers. Individuals with the best
distribution are backcrossed into the M82 parent, and this is repeated for ten or more generations.
Once lines are found with single small segments of S. pennellii genome in the M82 near-isogenic
background, those plants are self-pollinated until homozygous, creating pure lines. The last step
is to create a map showing the alignment of all of the different introgression lines such that the
whole S. pennellii genome is covered, and finally assignment of bins.
15
The importance of these lines is two-fold. First, a large amount of genetic diversity can be
exploited for new traits, given the wide range of evolutionary divergence between the two
species. Second, one can narrow down the genetic location of a phenotype to a particular
chromosomal segment via map-based cloning, eventually leading to gene discovery. This is
possible because S. pennellii has a high enough frequency of single nucleotide polymorphisms
(SNPs) that a dense map of genetic markers has been created and is available for public use
(Mueller et al., 2005).
One argument for the use of introgression populations is that the current cultivated
varieties represent a very narrow genetic base. Many wild species of the Solanaceae family and a
few closely related to tomato exist throughout the world. The genetic backgrounds of these
species are useful for introduction into the tomato varieties that are sold commercially. It is from
these wild species that tomato researchers hope to find the answers to problems with disease and
pest resistance, adaptation tolerance, yield, flavor, postharvest and other tomato qualities.
The method of improving tomato quality by QTL identification has already been used for
traits such as increased sugars and soluble solids (Fridman et al., 2002). By analyzing each
introgression line and comparing its phenotype to that of the M82 parent, the genetic location of
the trait can be immediately localized to a relatively small chromosomal segment. It has already
been found that many of these ILs contain important flavor volatile QTL (Tieman et al., 2006).
Many ripe tomato flavor volatile QTL have also been identified in the IL population of S.
habrochaites, another wild relative of cultivated tomato (Mathieu et al., 2009).
Among the many ILs that show alterations in flavor volatile profiles, twelve of them are
relevant to the research presented here. The volatile profiles of those lines were studied
previously (Tieman et al., 2006). Gas chromatography analysis of fruit from these lines revealed
16
elevated levels of the volatiles 2-methylbutanal, 2-methylbutanol, 3-methylbutanal, 3-
methylbutanol, isovaleronitrile, isobutylthiazole, and isobutyl acetate compared to the M82
control. These volatiles are collectively called branched-chain volatiles because of their
structural similarities to BCAAs. The aldehyde and alcohol branched-chain volatiles are known
to attribute unpleasant flavor to tomatoes (Tandon et al., 2000; Baldwin et al., 2008). The
metabolic route leading to these compounds in plants is unknown and presents a challenging
study. Using evidence from microorganisms and predicting putative pathways based on chemical
structures provides a platform to begin studying the synthesis of these compounds and the
enzymatic steps involved.
Tomato Flavor
Much of tomato fruit flavor composition has been worked out qualitatively and
quantitatively (Buttery et al., 1986; Buttery et al., 1987; Buttery et al., 1987; Buttery et al., 1988;
Buttery et al., 1989). The significance of those studies was in the identification of many of the
most important tomato fruit volatiles and determining their odor thresholds, gas chromatography
retention times, and mass spectra. The odor threshold of a volatile compound is the smallest
amount of the compound that can be sensed by human olfaction, and is determined by trained
panelists. If the concentration of a compound exceeds its odor threshold, it is a positive
contributor to flavor. If its concentration is below its odor threshold, it does not contribute to
perceived flavor. Odor units determine the impact of a volatile flavor compound and are
determined by dividing the log of its concentration by its odor threshold. Determination of odor
units is important in choosing compounds to study. For example, in the study by Buttery et. al
(1988), the volatile compound isovaleronitrile was discovered and was present from 10-200 ppb,
depending on the cultivar. They also found that its odor threshold was relatively high, at 150
ppb. Given both figures, this compound probably is not important to tomato aroma. However,
17
this compound is structurally related to the volatiles isobutyl cyanide and 2-isobutylthiazole, the
second of which has positive odor units in tomato fruit. On the other hand, the volatile β-
damascenone has an odor threshold of 0.002 ppb and a fresh fruit concentration of 1-3 ppb,
making it a positive contributor to fruit flavor. Table I shows the odor thresholds and
concentrations typically found in ripe fruits for most of the volatiles important in tomato flavor,
adapted from Buttery et al. (1971,1986, 1987,and 1988).
There are many thousands of aroma compounds known today, and over 400 have been
identified in tomatoes. However, only about 16 of these show evidence of being important in
flavor, given their odor units (Buttery and Ling, 1993). Interestingly, some of these compounds
have a pleasant odor in small amounts, but can be perceived as unpleasant when present at higher
concentrations. Examples of these are some of the branched-chain volatiles, which have been
described as pungent and stale, and phenylacetaldehyde, which is responsible for the malodorous
fruit aroma phenotype of IL8-2 (Tadmor et al., 2002). The branched-chain volatiles only seem to
give food a desirable quality when present at low concentration or in certain fermented foods
such as breads, cheeses, beers, and wines, in which they are crucial flavor components. Some
volatiles, however, make a fruit more pleasant when in concentration is increased, such as cis-3-
hexenal, one of the most abundant tomato volatiles, which gives a green aroma (Tandon et al.,
2000).
As mentioned previously, a study was performed in our lab on several of the S. pennellii
introgression lines to find loci important to fruit flavor (Tieman et al., 2006). Twenty-five
different quantitative trait loci (QTL) affecting 23 volatile flavor compounds were identified.
Four QTL that were significantly different in citric acid content were also identified. Other
results in the study are highly relevant to the experiments reported here. First, it was found that
18
12 loci had alterations of multiple branched-chain volatile compounds. The fact that this class of
volatiles is altered in so many QTLs suggests that there are many factors or regulatory elements
influencing their metabolism. Second, it was found that isovaleronitrile, whose synthetic
pathway has never been determined, was altered in several QTLs along with other branched-
chain volatiles, suggesting it forms from the same precursors. The authors speculate that the
changes in branched-chain volatile content observed in the IL population could be due to
transcriptional regulators or rate-limiting enzymes affecting the metabolism of any of the
intermediate compounds in their synthesis.
Branched-Chain VOC Metabolism
Largely based on structural considerations and substrate feeding studies, the BCAAs have
been proposed to be precursors for the important flavor volatiles 2-methylbutanol, 3-
methylbutanol, 2-methylbutanal, and 3-methylbutanal in plants (Tressl and Drawert, 1973;
Gonda et al., 2010). These branched-chain volatiles are produced in bacteria, fungi, and plants
(Tressl and Drawert, 1973; Vergnais et al., 1998; Perez et al., 2002; Dickinson et al., 2003;
Nimitkeatkaia et al., 2005) and contribute to flavor in many foods including tomato (Guadagni et
al., 1972; Buttery et al., 1987; Mayer et al., 2003). In microorganisms these volatiles are formed
by the catabolism of BCAAs first by a branched-chain aminotransferase (BCAT, 2.6.1.42) to
form BCKAs. Those acids are subsequently acted upon by pyruvate decarboxylase-like (PDC-
like, 4.1.1.1) enzymes to form branched-chain aldehydes followed by alcohol dehydrogenases
(ADH, 1.1.1.1) to form branched-chain alcohols (Hazelwood et al., 2008). These reactions are
collectively known as the Ehrlich pathway (Ehrlich, 1904) and are illustrated in Figure 1-1A.
BCAAs can also be catabolized by another, more established, primary metabolic pathway
that does not produce volatile compounds. In this pathway, BCKAs are produced from
deamination of BCAAs by BCATs, but are substrates for the branched-chain alpha keto acid
19
dehydrogenase (BCKADH) enzyme complex (Fujiki et al., 2000). This step and those that follow
produce acyl-CoA-conjugated compounds, which then yield the intermediate compounds of the
TCA cycle, acetyl-CoA and succinyl-CoA (Taylor et al., 2004). This catabolic pathway is
illustrated in Figure 1-2. It has been shown that high BCAA and BCKA levels induce the
expression of enzymes in this complex as a mechanism for these compounds to regulate their
own concentrations (Fujiki et al., 2001). This pathway has never been shown to produce volatile
compounds. However, when this primary pathway is highly active in yeast, very little branched-
chain alcohols and aldehydes are synthesized, presumably because this pathway uses up
intermediates that would otherwise be used in the volatile-producing pathway.
The volatile synthesis pathways of BCAA catabolism have been best studied in the
budding yeast, Saccharomyces cerevisiae, and the information gained from those studies may be
useful for gene discovery and pathway elucidation in plants. S. cerevisiae makes many long
chain alcohols including 3-methylbutanol from leucine, isobutanol from valine, 2-methylbutanol
from isoleucine, 2-phenylethanol from phenylalanine, and tryptophol from tryptophan.
Isobutanol, 3-methylbutanol, and 2-methylbutanol were shown to be synthesized from valine,
leucine, and isoleucine, respectively, by experiments utilizing 13C-valine and NMR analysis
(Dickinson et al., 1997; Dickinson et al., 1998; Dickinson et al., 2000).
Studies in yeast also demonstrated the enzymatic steps by which the branched-chain
volatiles are synthesized from BCAA catabolism. They also suggest that in some cases the
catabolic pathways of all three BCAAs require different enzymes at each particular step. While
studying leucine catabolism, for example, knocking out one decarboxylase showed that it
contributes to 94% of 3-methylbutanol production, while another decarboxylase contributes 6%.
When both genes are knocked out, no 3-methylbutanol is produced, suggesting the only pathway
20
involved in its production in yeast must involve these enzymes. Six of thirteen ADH enzymes in
yeast have been shown through mutant analysis to produce 3-methylbutanol, illustrating the
redundancy of this enzyme group (Dickinson et al., 2003). Yeast mutants lacking BCKADH still
produced wild-type levels of 3-methylbutanol, evidence that the acyl-CoA-mediated pathway of
BCAA catabolism does not contribute to branched-chain volatiles (Dickinson et al., 1997). In
studies of valine catabolism, yeast mutants lacking BCKADH produced wild-type levels of
isobutanol from valine, while PDC triple mutants produced no isobutanol and the PDC-like
mutation had no effect. The evidence together suggests that specificity between valine and
leucine branched-chain alcohol production lies with the carboxylase step. Unlike with leucine
and valine catabolism, in yeast, 2-methylbutanol from isoleucine can be produced from any of
the three PDCs, the PDC-like YDL080, or YDR380. BCKADH does not contribute to 2-
methylbutanol production (Dickinson et al., 2000).
Though it seems the specificity of branched-chain volatile metabolism lies at the
decarboxylase steps, the initiation of these pathways lies with BCATs. Yeast has two
characterized BCAT genes, the products of which are localized in the cytosol or in mitochondria
(Liepman and Olsen, 2004). Mutagenesis or overexpression of these BCATs both resulted in
dramatically decreased and increased values, respectively, of isobutanol emission (Lilly et al.,
2006). A different study found that both yeast BCATs are involved in 3-methylbutanol
production (Schoondermark-Stolk et al., 2006). The function of yeast BCATs in branched-chain
volatile metabolism has encouraged the study of this enzyme family in tomato, the results of
which are described in this work.
BCAA and Volatile Metabolism in Plants
The BCAAs leucine, isoleucine, and valine are primary plant metabolites involved in
many processes. They are synthesized from threonine or pyruvate in plastids of mostly young
21
tissues (Schulze-Siebert et al., 1984; Hagelstein et al., 1997). Threonine feeds into the isoleucine
pathway and pyruvate into the valine pathway, and hydroxyethyl-TPP into both, after which the
same four enzymatic steps are shared to form these two amino acids. Leucine is synthesized by a
branch of the valine pathway starting with α-ketoisovalerate in four enzymatic steps (Holmberg
and Petersen, 1988; Kohlhaw, 2003). These steps and their corresponding enzymes are illustrated
in Figure 1-1B.
Although synthesis of BCAAs is well characterized in plants, the catabolic pathways are
not completely understood. Their catabolism is believed to be initiated in mitochondria, where
the BCKDH complex is located (Taylor et al., 2004). The primary fates of BCAAs in plant cells
are peptide elongation, glutamate recycling, glucose and sucrose linked branched-chain ester
synthesis, branched-chain fatty acid synthesis, and respiration through synthesis of TCA cycle
intermediates (Kandra et al., 1990; Walters and Steffens, 1990; Kroumova et al., 1994; Daschner
et al., 1999; Li et al., 2003; Beck et al., 2004; Taylor et al., 2004). BCAA catabolism may be
more central to plant metabolism than previously appreciated. For example, Gu et al. (2010)
showed that a mutation in isovaleryl-CoA dehydrogenase, an enzyme in the BCAA catabolic
pathway, influences the metabolism of many unrelated compounds in Arabidopsis seeds,
including twelve amino acids.
In a banana study, it was shown that leucine and valine levels accumulate during ripening,
while isoleucine levels remain constant (Tressl and Drawert, 1973). Labeling studies in banana
showed that a small but significant amount of [14C]leucine was converted into 3-methylbutanol,
3-methylbutyl esters, 3-methylbutyric acid, and KIC. The largest amount of radioactive label was
found in 3-methylbutanol. Similar results were found for [14C]valine. Banana fruit have a volatile
profile consisting mostly of esters, with small amounts of alcohols, ketones, aldehydes, and
22
phenolic ethers (Tressl and Drawert, 1973). Alcohol acetyl transferase (AAT) in banana makes
esters using 2-methylbutanol and 3-methylbutanol as precursors, and the amount of volatile
esters produced is limited by alcohol substrate. Banana also produces 2-methypropyl esters from
valine (Wyllie and Fellman, 2000). In a similar study in strawberry, incubation with isoleucine
resulted in increased 2-methylbutanol and 2-methylbutanal, but levels started to drop off after 48
h, while esters derived from these compounds increased (Perez et al., 2002). A study in
Gypsophila showed that in pre-anthesis flowers, 3-methylbutanol and 3-methylbutanal are
produced from leucine via PDC and ADH enzymes, while leucine is made into 3-methylbutyric
acid via isovaleryl-CoA from the same enzymes in open flowers (Nimitkeatkaia et al., 2005). A
recent study in cucumber fruit provided evidence that some fruit volatiles are derived directly
from amino acids by the action of aminotransferases. Although they showed direct evidence of
labeled phenylalanine incorporation into volatiles, the authors were not able to show this for
volatile formation from BCAAs via BCATs (Gonda et al., 2010), These studies all suggest that
the Ehrlich pathway probably occurs in some plants and tissues, but is not the only route to
branched-chain volatiles. The results in chapter three suggest an alternate route to the synthesis
of these volatiles in tomato which bypasses the BCAT step.
As mentioned previously, there are twelve QTLs in the S. pennellii introgression
population affecting emission of branched-chain volatiles in tomato (Figure 1-3). QTLs from
these lines were generally were altered in multiple branched-chain volatiles, suggesting common
pathways for synthesis. In a study by Schauer et al. (2006), approximately 26 QTLs for BCAA
content of ripe tomato fruits were identified in the S. pennellii introgression population. Of these
QTLs, seven were altered in the levels of all three BCAAs, suggesting that they encode major
BCAA pathway regulatory elements. However, there is no consistent correlation between QTLs
23
with increased BCAAs and QTLs with increased branched-chain volatiles. These observations
shed doubt on the likelihood that BCAAs and branched-chain volatiles are metabolically related
in tomato, a thought that is strongly supported by the results in chapter three.
BCAA metabolism is under strict control by multiple regulatory mechanisms, including
gene expression, enzymatic substrate specificity, and feedback inhibition. Most of what is known
about regulation of BCAA metabolism in plants, however, pertains to the BCAA synthesis
Figure 1-3. Schematic of S. lycopersicum and S. pennellii introgression line population and corresponding branched-chain volatile loci. S. lycopersicum chromosomes are represented by long bars and introgressed genomic fragments of S. pennellii by small bars. Introgression fragments containing loci for branched-chain volatile phenotypes are designated by letters. The letters indicate which volatile phenotypes correspond to a particularintrogression line. Figure was adapted from Tieman et al. (2006).
30
Table 1-1. Flavor volatile compounds impacting the perception of ripe tomato fruit flavor.
Volatile
Conc.
(ppb)
Odor
units* Precursor
Odor
Characteristic
cis-3-Hexenal 12,000 3.7 lipid tomato/green ß-ionone 4 2.8 carotenoid fruity/floral Hexanal 3,100 2.8 lipid green/grassy ß-Damascenone 1 2.7 carotenoid fruity 1-Penten-3-one 520 2.7 lipid fruity floral/green 2-Methylbutanal 27 2.1 BCAAs/BCKAs musty/foot odor 3-Methylbutanal 27 2.1 BCAAs/BCKAs musty/foot odor trans-2-Hexenal 270 1.2 lipid green 2-Isobutylthiazole 36 1 BCAAs/BCKAs tomato vine 1-nitro-2-Phenylethane 17 0.9 phenylalanine musty, earthy trans-2-Heptenal 60 0.7 lipid green Phenylacetaldehyde 15 0.6 phenylalanine floral/alcohol 6-Methyl-5-hepten-2-one 130 0.4 carotenoid fruity, floral cis-3-Hexenol 150 0.3 lipid green 2-Phenylethanol 1,900 0.3 phenylalanine nutty 3-Methylbutanol 380 0.2 BCAAs/BCKAs earthy, musty 2-Methylbutanol 100 0.2 BCAAs/BCKAs earthy, musty Methyl salicylate 48 0.08 phenylalanine wintergreen *Odor unit is defined as the log of the concentration of a volatile divided by its odor threshold Data are from Buttery and Ling (1993).
31
CHAPTER 2 CHARACTERIZATION OF TOMATO BCATS
Cloning of SlBCAT cDNAs
In order to better understand the dynamics of branched-chain amino acid metabolism, six
unique tomato sequences potentially encoding BCAT enzymes were identified by searching the
Sol Genomics Network tomato expressed sequence tag (EST) database
(http://solgenomics.net/index.pl) (Mueller et al., 2005) for unigene sequences similar to those of
BCATs from other species. Unigenes are collections of transcripts that appear to originate from
the same locus. Full-length cDNAs from each unigene were cloned and sequenced, and proteins
were deduced from their open reading frames. The unigene SGN-U569828 (SlBCAT1, 45
members) has the most ESTs of all putative SlBCATs, while the unigene SGN-U569952
(SlBCAT3, 27 members) has the second highest, both far surpassing the numbers of ESTs of the
other putative SlBCATs. Phylogenetic analysis of all putative SlBCATs and comparisons with
Arabidopsis BCATs (Diebold et al., 2002) revealed that SlBCAT1 is most similar to the AtBCAT2
and AtBCAT1 genes from Arabidopsis, respectively. The unigene SGN-U569830 (SlBCAT2,
seven members) is most similar to AtBCAT3. SlBCAT3 and the unigene SGN-U569953
(SlBCAT4, seven members) are highly similar to each other and most similar to AtBCAT5. The
unigenes SGN-U569831 (SlBCAT5, five members) and SGN-U569829 (SlBCAT6, two
members) are most similar to AtBCAT2 and most similar to each other within the putative
SlBCATs (Figures 2-1 and 2-2).
A seventh unigene, SGN-U566152, a putative branched-chain aminotransferase-like
protein, was identified and considered to be a possible member of the tomato BCAT family.
However, after comparing the phylogeny of this gene with the rest of the family, it was decided
that this gene is most likely not a BCAT based on its 29% homology with its closest SlBCAT
Figure 2-1. Evolutionary relationships of mature SlBCAT proteins. Mature peptides were predicted by omitting highly divergent N-terminal residues thought to be transit peptides. A, Alignment of amino acid sequences using ClustalW (Larkin et al., 2007). B, The evolutionary history was inferred using the Neighbor-Joining method (Saitou and Nei, 1987). The optimal tree with the sum of branch length = 1.21366717 is shown. The percentage of replicate trees in which the proteins clustered together in the bootstrap test (500 replicates) are shown next to the branches (Felsenstein, 1985). The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. There were a total of 332 amino acids in the final dataset. Phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007).
Figure 2-2. Evolutionary relationships of mature SlBCAT and AtBCAT proteins. Mature peptides were predicted by omitting highly divergent N-terminal residues thought to be transit peptides. A, Alignment of amino acid sequences using ClustalW (Larkin et al., 2007). B, The evolutionary history was inferred using the Neighbor-Joining method (Saitou and Nei, 1987). The optimal tree with the sum of branch length = 2.55921922 is shown. The percentage of replicate trees in which the associated taxa clustered together in the bootstrap test (500 replicates) are shown next to the branches (Felsenstein, 1985). The tree is drawn to scale, with branch lengths in the same units as those of the evolutionary distances used to infer the phylogenetic tree. There were a total of 331 positions in the final dataset. Phylogenetic analyses were conducted in MEGA4 (Tamura et al., 2007).
SlBCAT5
SlBCAT6
SlBCAT1
AtBCAT2
AtBCAT1
AtBCAT5
SlBCAT4
SlBCAT3
AtBCAT4
AtBCAT6
AtBCAT7
AtBCAT3
SlBCAT2100
100
100
100
100
100
100
100
99
39
0.1
A B
42
Figure 2-3. Quantification of SlBCATs RNA in different tissue types. A, Comparison of
expression of genes relative to each other in each tissue. B, Comparison of expression in tissues within each gene. Analysis was performed on three biological and three technical replicates of tissues for each sample. Values represent percentage of total mRNA per sample ± SD, calculated from a standard curve for each gene. Note differences in y-axes. Expression of SlBCAT5 and SlBCAT6 was below the limit of detection. L = young leaves, F = inflorescences 1dpa, G = mature green fruit, B = breaker fruit, T = turning fruit, R = red ripe fruit.
A
43
Figure 2-3. Continued.
B
44
Figure 2-4. Subcellular localization of SlBCATs. Each gene was fused to E-GFP at the C-
terminal and expressed in N. benthamiana leaf protoplasts. The left column shows GFP fluorescence, the middle column shows marker fluorescence, and the right column shows merging of GFP and marker. Chlorophyll autofluorescence was used to show presence of chloroplasts for SlBCAT-3, 4, 5, and 6. Mitotracker Orange dye was used to show mitochondria for SlBCAT-1 and 2. Scale bars are each 10µm.
45
Figure 2-5. Growth complementation of E. coli ΔilvE/ΔtyrB mutant cells expressing SlBCAT3
and 4. Cells are streaked on media lacking amino acids and SlBCAT cDNAs are expressed under control of the Pbad promoter in mutant cells. The top image shows growth restoration by SlBCAT4, while the bottom image shows restoration by SlBCAT3. Cells expressing the other SlBCATs are not shown, as they did not show visible growth on plates.
Table 2-1. Measurement of E. coli cell culture growth rate Strain BW25113 ΔilvE/ΔtyrB
Cultures were grown for 10 h after which OD600 was measured. BW25113 and ΔilvE/ΔtyrB carried an empty pBAD24 vector. SlBCAT cDNAs are under control of the Pbad promoter in the pBAD24 vector (Guzman et al., 1995). Values are averages of two replicates ± SD.
Activities of purified recombinant SlBCAT proteins on all BCAA and BCKA substrates. Km is presented as average ± SE. Km data were obtained using GraphPad Prism5 software. Other parameters were obtained by calculations listed in Materials and Methods.
47
Table 2-3. Volatile emissions of BCAT over-expressing transgenic ripe tomato fruit. Line 3-methyl-
V Values are percentages compared to 100% of M82 control fruit volatiles, ± SD (n=6). Values in bold are statistically significant as determined by one-way ANOVA followed by Dunnet’s test, (p<0.05).
Figure 2-6. Analysis of transgenic fruit SlBCAT transcript levels. Analysis was performed on three biological and three technical replicates of ripe fruit for each sample. Values represent percentage of total mRNA per sample ± SD, calculated from a standard curve for each gene. Data were analyzed by one-way ANOVA and values of each line are significantly different compared to control as determined by Dunnet’s test (p<0.05).
48
Table 2-4. Levels of free amino acids in red ripe fruit of M82 and SlBCAT over-expression lines.
Values are expressed as means ± SD, (n=3).Values in bold are significantly different from control fruit by one-way ANOVA followed by Dunnet’s test (P<0.05).
49
CHAPTER 3 BRANCHED-CHAIN VOLATILES IN TOMATO
Rationale and Background
In order to identify the components involved in the synthesis of branched-chain flavor
compounds in fruit, tomato fruit pericarp discs were supplemented with potential precursors to
those volatiles, including the branched-chain amino acids and their α-keto acid analogs. These
experiments were performed to provide evidence that the pathway leading to branched-chain
volatiles in tomato fruit is either similar or dissimilar to the Ehrlich pathway in yeast. The
experiments may also provide evidence of the rate limiting factors and steps in the putative
pathways, whether they be substrate limited or enzyme limited.
Similar experiments to those described here were performed in yeast with positive results,
showing that such a feeding experiment can yield the downstream branched-chain volatile
compounds as long as the appropriate Ehrlich pathway enzymes are present: branched-chain
amino acid aminotransferase, pyruvate decarboxylase, and aldehydes dehydrogenase (Dickinson
et al., 2003). The yeast studies showed that the flavor volatiles 2-methylbutanal and 2-
methylbutanol are downstream products of isoleucine catabolism and 3-methylbutanal and 3-
methylbutanol are downstream products of leucine catabolism (Dickinson et al., 1997; Dickinson
et al., 2000).
The hypothesis held before performing these experiments was that the branched-chain
volatiles emitted from tomato fruit are synthesized from BCAA substrates. The justification is
that the same volatiles in yeast are products of BCAAs and these substrates share similar
chemical structure with the volatiles. Since BCKAs are synthesized from BCAAs and are
downstream in the Ehrilch pathway, it was predicted that feeding these compounds would also
increase branched-chain volatile emissions. The results from the following experiments,
50
however, were unexpected and required a revision of the current hypothesis that branched-chain
volatiles in tomato are formed via the Ehrlich pathway.
Results of Substrate Feeding
The free amino acids leucine, isoleucine, and valine were applied to M82 red ripe fruit
pericarp discs and incubated for eight hours after which volatiles were quantified by GC-MS.
Emissions of seventeen tomato volatiles having the most impact on flavor were analyzed and
compared to emissions from controls supplemented with only water. Significant changes were
not seen in many of the non-branched-chain volatiles, as was expected due to structural
dissimilarities to the provided substrates.
In tissues supplemented with leucine, small but significant (p<0.05) increases of about 1.5-
fold were observed in 3-methylbutanol and 3-methylbutanal emissions (Figure 3-1). Both of
these volatiles share side-chain structure with leucine. A significant decrease in cis-3-hexenal
and highly variable increase in trans-2-hexenal were also observed in these samples.
The volatile trans-2-hexenal, along with other green leaf volatiles (GLV), is usually
formed from linolenic and linoleic acids via lipoxygenase enzymes (Hatanaka, 1993). GLVs are
thought to exist in low concentrations in healthy plant tissues, but concentrate rapidly in stressed
or wounded tissues, and that tissue laceration provides the lipid substrates and triggers hydrolysis
to form these volatiles (Matsui et al., 2000). It is possible that the tissue injury from creating
pericarp discs and addition of high amounts of branched-chain substrates elicits a stress response
that causes trans-2-hexenal to form.
In tissues supplemented with KIC, significant increases of about nine-fold and ten-fold in
3-methylbutanal and 3-methylbutanol, respectively, were observed (Figure3-1). This result may
be explained by the fact that KIC is one step closer to these volatile compounds than leucine in
the hypothesized Ehrlich pathway. Although 3-methylbutanal is closer in this pathway to KIC
51
than 3-methylbutanol, a greater concentration of 3-methylbutanol was emitted. This could be due
to the fact that 3-methylbutanol is the end product in this pathway causing it to accumulate, and
no reverse reaction exists for conversion back into the aldehyde. Interestingly, a decrease of
almost 50% was seen in 2-methylbutanal in KIC-supplemented tissue. Three possible
explanations include: the steps converting KIC and KMV to 3- and 2-methylbutanal,
respectively, occur via the same enzyme, for which both substrates are in competition, the
unnaturally high levels of substrates applied causes competitive or feedback inhibition of the
isoleucine synthetic pathway, or a several-fold increase in 3-methylbutanal or 3-methylbutanol
signals feedback inhibition of that pathway. In either case, 2-methylbutanal is still catabolized to
cause accumulation of 2-methylbutanol, as shown by the lack of change in 2-methylbutanol
levels from control samples.
When isoleucine was fed to M82 pericarp discs, small but significant increases of about
2.5-fold and 1.5-fold were observed in the levels of 2-methylbutanal and 2-methylbutanol
emissions, respectively (Figure 3-2). This result was expected because these two volatiles share
the same side chain structure as isoleucine and thus predicted to be metabolically related. This
result shows that only a small amount of isoleucine can be converted to branched-chain
volatiles.
M82 pericarp discs supplemented with KMV showed significant increases of about 15-fold
and 9-fold in 2-methylbutanal and 2-methylbutanol emissions, respectively, from water fed
controls (Figure 3-2). Like the results from the KIC and leucine feeding experiment, the greater
conversion of KMV than isoleucine to branched-chain volatiles may be explained by a revision
of the putative pathway to branched-chain volatiles. These results suggest BCKAs are more
likely than BCAAs to be the primary precursors to these volatiles. Similar to the KIC feeding
52
results, KMV feeding resulted in decreases of about 60% and 20% in levels of 3-methylbutanal
and 3-methylbutanol emissions. As with KIC, this may be due to high substrate competitive or
feedback inhibition of the pathway leading to these volatiles by the substrate. Similar decreases
of about 20% were seen in the branched-chain volatiles isovaleronitrile and 2-isobutylthiazole,
suggesting high amounts of supplied KMV or the 2-methylbutanal or 2-methylbutanol products
may cause inhibition of the synthetic pathways of those compounds as well. There is also a
significant 1.5-fold increase of isobutyl acetate observed from KMV feeding and significant
increases in trans-2-hexenal in both isoleucine and KMV feeding. Lack of structural similarities
between these volatiles and the substrates suggest they are not related metabolically, but that the
substrates signal a change in these volatiles’ pathways. These changes are not likely due to
metabolic relatedness of volatiles to substrate but rather to regulatory mechanisms.
M82 pericarp discs supplemented with valine did not exhibit significant increases in
branched-chain flavor volatiles (Figure 3-3). In yeast, catabolism of valine results in KIV and the
volatiles isobutanal and isobutanol (Dickinson et al., 1998). Though these volatile compounds
are likely produced in tomato fruit, given that leucine and isoleucine feeding yield their
respective branched-chain volatiles, they are not considered important compounds to flavor
perception and are of less concern in this study. Also, due to their very low molecular weights,
they have a very short retention time on the gas chromatograph and are not detectable with the
methods used in this analysis. The only significant changes observed with valine feeding were a
three-fold increase in trans-2-hexenal and about 0.3-fold decreases in cis-3-hexenal and trans-2-
heptenal.
M82 pericarp discs supplemented with KIV displayed increases of about 4.5-fold and 5-
fold in 3-methylbutanal and 3-methylbutanol emissions, respectively. This result is similar to that
53
observed with leucine feeding and can be explained by the fact that KIV is a precursor to KIC,
therefore it is expected that the same branched-chain volatiles would increase in this experiment.
This explanation is reinforced by results from yeast, in which it was reported that products of
both leucine and valine catabolic pathways are derived from a shared pool of KIV (Dickinson et
al., 1998). It is interesting, however, that although the same amount of leucine and KIV were fed
in each experiment the amount of 3-methylbutanal and 3-methylbutanol increase was greater
with KIV than with leucine. It is possible that a high concentration of leucine, but not KIV, in the
cell exhibits feedback inhibition to steps downstream in the pathway to KIC and volatile
synthesis, such as isopropylmalate synthase. It is also possible that the flux from KIV to KIC is
greater than that of leucine to KIC, since application of KIC yields the greatest increase in 3-
methylbutanal and 3-methylbutanol of all branched-chain substrates. The likely explanation is
that the hypothesis of the Ehrlich pathway occurring in tomatoes is incorrect. These data suggest
that KIC is more likely the primary precursor of 3-methylbutanal and 3-methylbutanol, rather
than leucine. Leucine catabolism may not result in these volatiles at all under biological
conditions, but is observed only when unnaturally high concentrations are fed to fruit. A decrease
of about 40% in 2-methylbutanal emission was observed with KIV feeding, similar to the result
from KIC feeding. KIV feeding also resulted in an increase of about fourteen-fold of the
branched-chain ester isobutyl acetate, making it the only tomato flavor volatile so far known to
be produced predominantly from the valine catabolic pathway. This compound is likely
produced by the reaction of isobutanol with acetyl-CoA and an acetyltransferase enzyme. There
was also an observed 1.5-fold increase in isobutyl acetate in tissues supplemented with KIC,
which may be explained by feeding of KIC into the valine pathway (Figure1-1). As in the other
feeding experiments, feeding of KIV caused significant increases in trans-2-hexenal.
The analysis of the feeding experiments above gave evidence for the synthetic pathways of
the volatile end-products of branched-chain catabolism, but did not measure the non-volatile
intermediates upon feeding of BCAAs. Additional evidence is also needed to support the
possibility that the Ehrlich pathway does not occur in tomato fruit as it does in yeast. The above
feeding experiments only provide indirect evidence of a pathway leading to branched-chain
volatiles. To further test and validate the hypothesis of a pathway in tomato fruit, a similar
feeding experiment was performed in which [U-13C]leucine heavy isotope was used as a
substrate. It is likely that the catabolic pathway of leucine is parallel to the pathways of
isoleucine and valine, similar to these pathways in yeast. Leucine was used as a representative of
all three BCAAs and their potential catabolic pathways.
The [U-13C]leucine compound was fed to red ripe fruit pericarp discs and incubated in
Petri dishes. Tissues were immediately frozen after 4 and 8 h time points. Metabolites were
extracted with methanol, then derivitized with methoxyamine hydrochloride in pyrimidines
followed by MSTFA and analyzed by GC-MS. The two time points were used to allow
measurement of 13C label incorporation over time. Analyzing the concentrations of compounds
which have incorporated label should provide evidence of the steps in the leucine catabolic
pathway in tomato fruit, and analyzing them temporally should give an indication of the
sequence of steps in the pathway. 13C-labeled substrate and GC-MS analysis was used in the
experiment to assess this pathway because it has proved useful in previous studies, such as in the
analysis of glycolysis in Arabidopsis (Giege et al., 2003). Molar enrichment of 13C label into
intermediate compounds was determined by comparison to 12C spectral fragments from
55
standards. We specifically analyzed the accumulation of label in putative BCAA catabolic
intermediates and branched-chain volatiles (Table 3-1).
A significant quantity of 13C was observed in KIC, which is the direct catabolic product of
leucine by BCAT (Table 3-1). Significant amounts of label in KIV and KMV were also
observed. The quantities of labeled KIC and KIV were about double that of labeled KMV, which
is likely due to the connection of the leucine metabolic pathway to that of valine but not
isoleucine. It is interesting that the quantity of 13C label in KIV was higher than in KIC after 4 h,
but the opposite is observed after 8 hours. This may be due to a lower rate of carbon transfer into
KIC than KIV, which eventually reaches a steady state while KIC continues to accumulate label.
Of the branched-chain volatiles the incorporation of label was very low, though 3-
methylbutanol and 2-methylbutanol accumulated the most label, while their aldehyde forms had
nearly 10-fold less accumulation. The incorporation of label into these two alcohols about
doubled from 4 h to 8 h, while most other branched-chain intermediates appeared to reach their
isotopic steady-state by 4 h. Both of these observations are evidence that the branched-chain
alcohols are likely the end-products of the pathway in tomato. It also appears from these results
that the conversion of BCKAs from BCAAs occurs rapidly before reaching a steady-state. The
very low concentration of label in the branched-chain volatiles is likely because the pool of
BCKAs with label is used in other more primary pathways such as protein synthesis, leaving
only a small amount to be used in volatile synthesis. There may also be technical difficulties in
detecting such small amounts of volatile compounds.
That there is not a much higher amount of label in the volatiles supports the hypothesis
that BCAAs are not immediate precursors to branched-chain volatiles. It is also supported by the
fact that more label was incorporated into 2-methylbutanol than 3-methylbutanol, while the
56
opposite result would be expected if 3-methylbutanol was produced by a direct route from
leucine. The results of this experiment show that carbon flux from leucine is highly dynamic,
which creates difficulty in interpreting the data. This is evident by the unexpected incorporation
of label into valine and isoleucine, which suggests that much of the carbon was first cycled into
pyruvate and threonine, the precursors to valine and isoleucine.
BCAA and Branched-Chain Volatile Loci
As was mentioned in the introductory chapter, much information can be gained about
tomato traits by identifying loci of interest using the population of S. pennellii introgression
lines. It was previously determined that there are at least 26 genetic loci with significant volatile
emission phenotypes, 12 of which are altered in branched-chain volatile emissions (Tieman et
al., 2006). This high number of branched-chain volatile loci suggests there are many factors
contributing to their synthesis and many points of regulation in their pathways in tomato.
Therefore, it is not surprising that drastic phenotypes were not observed in the transgenic
BCAT over-expression plants, or that the pathway leading to these volatiles may not be as
straightforward as they are in yeast. To further support the theory that branched-chain volatiles
are not derived directly from BCAAs, it is possible to cross-reference volatile loci to loci that are
changed in BCAA quantity in the S. pennellii population. These BCAA loci have already been
established in a previous report in which, like the volatiles, there were a surprisingly high
number of loci with BCAA phenotypes (Schauer et al., 2006). As mentioned in the introduction,
there is no consistent pattern of BCAA loci and branched-chain volatile loci occurring together,
as illustrated in Table 3-2. This observation adds additional evidence to the hypothesized
pathway which suggests branched-chain volatiles are not derived from BCAA catabolism, but
directly from catabolism of BCKAs.
57
The evidence presented in this chapter suggests that branched-chain volatiles are more
likely derived directly from BCKAs, and not from BCAAs as they are in yeast (Dickinson et al.,
2003). Feeding of BCKAs in tomato fruit tissue resulted in much greater production of branched-
chain volatiles than BCAA feeding. The isotopic feeding showed that only a small transfer of
carbon into branched-chain volatiles from leucine occurs. Along with the IL data, these
observations reinforce the possibility of this novel branched-chain catabolic pathway existing in
tomato fruit.
58
Figure 3-1. Leucine and KIC feeding. Levels of relevant flavor volatiles extracted from tomato
pericarp samples supplemented with either leucine or KIC, expressed as a percentage of water-fed control pericarp normalized to 100%. Asterisks mark compounds that were significantly different from control, as determined by one-way ANOVA followed by Dunnet’s test, (p<0.05) (n=3).
59
Figure 3-2. Isoleucine and KMV feeding. Levels of relevant flavor volatiles extracted from
tomato pericarp samples supplemented with either isoleucine or KMV, expressed as a percentage of water-fed control pericarp normalized to 100%. Asterisks mark compounds that were significantly different from control, as determined by one-way ANOVA followed by Dunnet’s test, (p<0.05) (n=3).
60
Figure 3-3. Valine and KIV feeding. Levels of relevant flavor volatiles extracted from tomato
pericarp samples supplemented with either valine or KIV, expressed as a percentage of water-fed control pericarp normalized to 100%. Asterisks mark compounds that were significantly different from control, as determined by one-way ANOVA followed by Dunnet’s test, (p<0.05) (n=3).
61
Table 3-1. Label accumulation in metabolite pools following incubation with [U-13C]leucine.
Metabolite Label accumulation
(nmol gFW-1) 4 h 8 h
BCAAs
Leu 1302 ± 129 1285 ± 106 Iso 412 ± 32.0 461 ± 42.0 Val 256 ± 25.0 267 ± 24.0
2-isobutylthiazole 0.005 ± 0.000 0.005 ± 0.001 Values are presented as means ± SE from four independent determinants.
62
Table 3-2. Occurrences of loci containing BCAA and branched-chain volatile phenotypes.
BCAA IL
locus
Fold-
change
BCV
locus BCAA
IL
locus
Fold-
change
BCV
locus
L 2-2 0.74 - I 6-2 1.83 - V 0.65 V 1.89 V 2-3 0.61 + L 6-2-2 2.07 - L 2-4 2.15 + I 3.19 I 2.6 V 2.43 V 2-5 0.36 + L 7-1 1.72 + L 2-6-5 1.41 - I 2.16 L 3-1 0.64 - V 1.76 L 3-2 2.49 - L 7-2 2.02 + I 1.96 I 2.37 V 1.69 V 7-3 2.4 - L 3-4 1.94 - I 8-1 2.59 - V 1.87 V 1.99 V 3-5 0.59 + I 8-1-1 2.35 L 4-1 1.76 - V 2.06 I 2.61 L 8-3-1 2.33 - V 1.99 L 9-3 1.78 + L 4-4 1.54 + L 10-1 2.83 + I 2.09 I 2.75 I 5-1 0.55 + V 1.64 V 0.63 L 11-2 2.11 + L 5-2 1.75 - V 1-1-3 0.68 - I 2.6 V 12-1-1 1.51 - V 2.06 L 12-3 2.08 - I 5-3 2.24 - I 2.66 V 1.97
The first column indicates which BCAA is changed. The second column indicates the introgression line containing the phenotype. The third column values indicate fold differences of BCAA concentration in red ripe fruit from each introgression line from M82 control values, as reported by Schauer et al. (2006). The fourth column indicates if an overlapping branched-chain volatile phenotype was observed in the line, according to Tieman et al. (2006). Shaded rows indicate ILs with phenotypes for all three BCAAs. BCAA = branched-chain amino acid, BCV = branched-chain volatile, L = leucine, I = isoleucine, V = valine. Data have been adapted from Schauer et al. (2006) and Tieman et al. (2006).
63
CHAPTER 4 DISCUSSION OF RESULTS
Diversity of SlBCAT Family
In this work we have discovered six BCATs that exist in S. lycopersicum, and expression of
four of these cDNAs was detected in the parts of the plant that were analyzed. The experiments
that followed their discovery revealed important information about this enzyme family and its
individual members in a plant species in which they have not yet been extensively characterized.
It is clear from BCAT studies in other plant species that this is a very important metabolic
enzyme family whose members perform functions in addition to BCAA synthesis and
degradation, making their continued study important to the overall understanding of plant
metabolism. The SlBCATs have very distinct characteristics from each other, suggesting some of
them they may also perform non-primary metabolic processes in the plant.
SlBCAT1 and SlBCAT2 were shown to be located in mitochondria in tobacco leaf cells.
Given that mitochondria are the primary location of BCAA catabolism in plant cells and the
preferences of these two enzymes is primarily catabolic and only minimally restore growth in E.
coli auxotrophs, these are likely to be the primary BCAA-degrading enzymes in tomato. Since
SlBCAT1 is highest in expression in ripening fruit, it may be the primary enzyme for recycling of
BCAAs resulting from protein degradation in senescent fruit tissue. Since SlBCAT2 is expressed
in all green tissues analyzed and very highly in floral tissues, it may be the responsible for
BCAA catabolism throughout. We expect a catabolic mitochondrial gene to exist in all tissues
since BCAAs are known precursors to the TCA cycle intermediates succinyl-CoA and acetyl-
CoA. This gene may also function in the crucial process of maintaining steady-state levels of
BCAAs in cells, which is crucial for compounds of such primary importance in cellular
metabolism.
64
The very low preference of SlBCAT1 for the forward direction was may be explained by
the multiple amino acid substitutions this enzyme has in conserved regions. It has an I101 in its
active site whereas the other SlBCATs all have F101, as do the E. coli, human, and yeast BCAT
proteins at this position. However, AtBCAT1 also contains an isoleucine at this position and is
highly active in the forward direction, making this substitution an unlikely cause (Schuster and
Binder, 2005). More likely are the residues S122, which is an asparagine in all BCATs from
Arabidopsis, human, and yeast, and N175, which is a serine or threonine at the same position in
all Arabidopsis and yeast genes. The human mitochondrial BCAT has this same amino acid
substitution, but humans do not synthesize BCAAs. Future mutagenesis experiments may
determine if forward activity can be enhanced in this tomato enzyme by changing these residues
to their more conserved amino acids. If successful it will provide crucial evidence to the amino
acid configurations of BCAT enzymes that give them reversible catalytic activity.
SlBCAT3 and SlBCAT4 were located in chloroplasts of tobacco leaf cells and were able to
restore growth of E. coli cells lacking the ability to synthesize BCAAs, suggesting that these
enzymes are primarily involved in BCAA synthesis. The substrate affinities of these two
enzymes, especially of SlBCAT3, suggest the same. It is possible, however, that under certain
conditions these two enzymes may have a role in BCAA catabolism as well, which is evident
from these two enzymes’, especially SlBCAT4, ability to use BCAAs as substrates in the kinetic
studies. SlBCAT3 may function as the BCAA synthetic enzyme throughout the plant, since it is
expressed nearly equally in all tissues analyzed, including the four fruit stages. SlBCAT4 appears
to be more specialized in function, given that its expression is by far the highest in flowers and is
relatively low in other tissues compared to the other SlBCATs. An enzyme with such a specific
expression pattern may have a specialized function, possibly in the metabolism of compounds
65
other than BCAAs as is observed with the Arabidopsis AtBCAT3 and AtBCAT4 (Schuster et al.,
2006; Knill et al., 2008). Kochevenko et al. (2010) showed that tomato inflorescences contain
nearly twice the concentration of all three BCAAs as leaves. The same study also revealed that
SlBCAT4 maps to IL3-2 and SlBCAT1 maps to IL12-3, both of which have QTLs for increased
levels of all three BCAAs. A N. benthamiana BCAT, recently found to be involved in hormonal
regulation, is also expressed highly in flowers and located in chloroplasts (Gao F, 2009). The
similarities of SlBCAT4 to this N. benthamiana gene make it an even more interesting subject for
the study of non-primary BCAT functions.
SlBCAT5 and SlBCAT6 transcripts were not detected in any of the tissues analyzed. The
only EST accessions that exist for these two unigenes were isolated from callus tissue cDNA,
therefore they may only be expressed under specific developmental, hormonal, or environmental
conditions. Despite their scarcity of expression, the distinct localization patterns and multi-
substrate kinetics of these two enzymes make them interesting candidates for study. The vacuolar
location of SlBCAT6 suggests that it may function in the recycling of BCAAs from the products
of proteolysis. This function is further supported by its lack of substrate specificity and relatively
high affinity for all branched-chain substrates. Of particular interest is this enzyme’s specificity
on KIV, which is extremely high compared to all other SlBCATs and substrates. It is unclear
why this enzyme has such a strong tendency toward valine synthesis.
The overall results of the SlBCAT kinetic experiments show that these enzymes are
diverse in function and are not entirely specific to individual branched-chain substrates, as might
have been expected from a multi-gene family. SlBCATs tend to have higher overall efficiencies
toward the BCKAs than the BCAAs, although it is assumed that BCATs are generally reversible
enzymes.
66
The lack of striking fruit volatile or BCAA phenotypes in transgenic plants was
unexpected. However, a simple explanation may exist as to why there were no clear phenotypes.
It would be expected that an increase in the expression of SlBCAT1, a catabolic enzyme, would
cause a decrease in BCAAs and thus a decrease in volatiles. As is observed in bacteria (Massey
et al., 1974), the enzymes in the pathway of BCAA synthesis are very tightly regulated by
substrate feedback, therefore if catabolism of BCAAs is increased, one might expect a
simultaneous increase in BCAA synthesis. The opposite effect may occur with overexpression of
SlBCAT3, a primarily anabolic enzyme. Increased BCAA synthesis resulting from an
overproduction of this enzyme might elicit a feedback mechanism to earlier steps in the BCAA
synthesis pathway, such as threonine deaminase or acetolactate synthase, inhibiting excess
BCAAs from being synthesized.
Significance of Substrate Feeding
The substrate feeding experiments were performed to gain insight into the catabolic
pathways of BCAAs and the formation of their respective volatiles in tomato fruit, of which little
is known. It seems the tomato branched-chain volatile synthesis pathways share at least some of
the same steps as in yeast (Dickinson et al., 2003). However, it was particularly interesting that
in the case of isoleucine/KMV and leucine/KIC, feeding of the α-keto-acid resulted in much
greater incorporations into the same branched-chain volatiles than feeding of the amino acid.
These results may be explained by the fact that the α-keto-acids are capable of a greater flux into
these compounds than amino acids. It may also be that the amino acids are quickly metabolized
by other pathways in the cell, such as in protein synthesis, while the α-keto-acids are in less
demand by enzymes other than those forming branched-chain aldehydes. It is likely that BCATs
and BCAAs do not participate in the volatile synthesis pathway, and the BCKAs are converted to
67
volatiles directly following their synthesis via dihydroxy-acid dehydratase and 3-isopropylmalate
dehydrogenase (Figure 4-1).
Tomato fruit cells are capable of disposing a small percentage of excess BCKA substrate
in the form of volatiles. The quantities of 3-methylbutanal and 3-methylbutanol emitted after
KIC feeding were 0.5 and 5.7 nmols/g FW, respectively, corresponding to 0.25 and 2.9 percent,
respectively, of substrate fed. The quantities of 2-methylbutanal and 2-methylbutanol emitted
after KMV feeding were 0.7 and 3.4 nmols/g FW, respectively, which corresponds to 0.35 and
1.7 percent, respectively, of substrate fed. The quantity of isobutyl acetate emitted after KIV
feeding was 0.45 nmols/g FW, corresponding to 0.25 percent of substrate fed. The quantity of
emissions after BCKA feeding compared to control fruit is evidence that these compounds are
limiting factors in the production of branched-chain volatile compounds. The results show that
tomato fruit cells are capable of producing more volatiles than they do under standard conditions
but are limited by the quantity of these intermediate metabolites in the cell. Conversely, it seems
that BCAAs are not limiting to the production of branched-chain volatiles. This is evident from
the lack of correlation between BCAA and volatile phenotypes in the S. pennellii IL population
and from low incorporation of BCAAs into volatiles seen in the feeding experiments.
The slight increase of isobutyl acetate observed with KIC feeding cannot be explained at
present. Since isopropylmalate synthase, the enzyme diverting KIV to the leucine synthetic
pathway, is thought to be irreversible, the conversion of KIC back to KIV and then to isobutyl
acetate is unlikely.
Decreases of 3-methylbutanal and 3-methylbutanol were observed after KMV feeding
and a decrease of 2-methylbutanal after KIC feeding. There are two possible explanations for
these occurrences: the steps in the pathways converting KIC and KMV to 3-methylbutanal and 2-
68
methylbutanal, respectively, may occur via the same enzyme and therefore both substrates would
be in competition for that enzyme, or a great increase in KIC, 3-methylbutanal, or 3-
methylbutanol may cause feedback inhibition of the pathway generating 2-methylbutanal and
vice-versa. In either case the step converting the branched-chain aldehydes to alcohols does not
seem to be affected as much, given the wild-type levels of 2-methylbutanol observed in KIC
feeding and only a 20% decrease of 3-methylbutanol compared with a 60% decrease of 3-
methylbutanal observed in KMV feeding. Also with KMV feeding, decreases of about 20% each
were seen in the branched-chain volatiles isovaleronitrile and 2-isobutylthiazole, suggesting that
KMV, 2-methylbutanal, or 2-methylbutanol may cause feedback inhibition in the pathways
synthesizing those compounds as well.
Feeding of valine and KIV yielded some expected and unexpected results. No significant
changes were seen in branched-chain volatile emissions with valine feeding. This lack of aroma
change was also observed in strawberries supplemented with valine (Perez et al., 2002).We
expected to see an increase in leucine-related volatiles since the valine and leucine pathways are
interrelated. On the other hand, valine is known to be one of the most potent feedback inhibitors
of BCAA synthesis by inhibiting ALS (Borstlap, 1972), so if that enzyme is involved in
branched-chain volatile production it would likely be repressed upon valine feeding (Massey et
al., 1974; Massey et al., 1976). We had not previously seen any evidence of a pathway leading to
isobutyl acetate synthesis, but feeding KIV has provided such evidence. The increase in isobutyl
acetate was greater than any other volatile in these feeding experiments, suggesting that KIV
follows a direct path to the synthesis of this compound. As is the case in microorganisms
(Dickinson et al., 1998), it is likely that this pathway occurs by conversion of KIV to isobutanal,
then to isobutanol, then to isobutyl acetate, this final step being catalyzed by an alcohol acetyl-
69
transferase. Dissection of this putative pathway may provide a new branch of tomato flavor-
related research in our laboratory. Also of interest in KIV feeding was that although the same
concentration of leucine and KIV were supplied in each experiment the amount of 3-
methylbutanal and 3-methylbutanol increases were greater with KIV feeding than with leucine
feeding. These volatiles were expected to increase with KIV feeding, but not to a greater degree
than with leucine, which was predicted to be closer to these volatiles. It could be that a high
concentration of leucine in the cell exhibits feedback inhibition to the downstream steps in the
volatile pathway while KIV does not. It is more likely that the flux from KIV to KIC is greater
than the flux from leucine to KIC, since KIC supplementation yields the greatest increase in
these two volatiles of all substrates supplies. This observation adds strong support to the
hypothesis that branched-chain volatiles are derived from BCKAs and do not involve the activity
of BCATs and BCAA substrates. The facts that only very small amounts of 13C label are
incorporated into branched-chain volatiles and that there is more label in 2-methylbutanol than in
3-methylbutanol also support the hypothesis that leucine is not a direct precursor of 3-
methylbutanal and 3-methylbutanol.
There is no clear indication in current literature that the involvement of BCKAs and
BCAAs in formation of trans-2-hexenal. Those compounds do not share enough structural
similarity to suggest that trans-2-hexenal may be converted from them directly. It is more likely
that the branched-chain compounds induce formation of trans-2-hexenal from long chain fatty
acids. A plausible explanation was given in Chapter 3 for this volatile’s biogenesis. It is known
that trans-2-hexenal and other C6-volatile compounds are synthesized from the degradation of
polyunsaturated fatty acids in plants (Zhuang et al., 1996). It is also known that trans-2-hexenal
and other C6 compounds are induced upon wounding and stress, and that their application to
70
plants induces defense-related genes (Kishimoto et al., 2005). The signaling pathways for these
stress responses are currently unknown. A recent study, however, has implicated γ-aminobutyric
acid (GABA) in this signaling pathway after isolating a GABA-aminotransferase mutant in
Arabidopsis. The authors showed that an increase in GABA elicits a resistance to the effects of
the trans-2-hexenal defense response on the plant (Mirabella et al., 2008). It is possible that
exogenous addition of branched-chain compounds down-regulates GABA formation, which in
turn causes an up-regulation of trans-2-hexenal production. This interaction with the GABA
pathway is feasible because BCAAs and BCKAs share similar structure with GABA, all are
precursors to TCA cycle intermediates, and all are substrates for aminotransferase enzymes and
use glutamate and α-ketoglutarate as donor and acceptor molecules, respectively. An application
of such high amounts of exogenous BCAAs and BCKAs to tomato fruit may have caused
competitive inhibition of GABA-aminotransferase. In addition, high concentrations of branched-
chain compounds may directly trigger a stress response which up-regulates the production of
trans-2-hexenal.
It is important to bear in mind that these feeding experiments are partly artificial in nature,
and the results may not entirely reflect the in vivo characteristics of fruit metabolism, due to the
higher-than-natural concentration of substrate supplementation. In addition, it is not currently
known which subcellular compartments volatile synthesis occurs in, and most of the substrate
presumably pools in the cytosol after application. Nevertheless, the results of these feeding
experiments show convincing evidence that the pathways to branched-chain flavor volatiles in
tomato fruit may share the same steps as in yeast and the Ehrlich pathway (Figure 1-1B),
excluding the first BCAT step. The results also show that the BCKAs are converted into a much
greater amount of their corresponding volatiles than are BCAAs. This may be due to the fact that
71
the BCKAs are only one enzymatic step from their corresponding volatiles while the amino acids
are two steps away, according to the Ehrlich pathway. This could also indicate that the BCAT
step converting BCAAs to BCKAs is much more limiting than the decarboxylase step converting
the BCKAs to aldehydes. Alternatively, and what is now our current hypothesis, BCATs and
BCAAs have only minimal roles in the branched-chain volatile synthetic pathways, while
BCKAs are the primary volatile precursors. BCKAs may be converted to volatiles directly from
their synthesis via dihydroxy-acid dehydratase and 3-isopropylmalate dehydrogenase in plastids
(Figure 4-1).
Concluding Remarks
Taken together, the results of the experiments presented here give much new detail about
BCAA metabolism and to the characteristics of the BCAT family, adding support to the
information that already exists about them in plants. Future work on this family will provide a
full characterization of its members and how they coordinate BCAA anabolism and catabolism
in plant organs. Alternative functions are also expected to be found in more detailed studies of
individual BCATs, particularly in SlBCATs 4, 5, and 6. Transgenic plants over-expressing and
silencing the four SlBCATs not described in this work are in the process of construction and
should soon expand what we know about these enzymes in vivo.
The complexity of tomato flavor, as stated previously, is an interaction of sugars, acids,
and volatile organic compounds. There are only a handful of volatiles that are thought to
contribute to tomato flavor, and of these there are seven believed to be derivatives of the BCAAs
and/or the BCKAs: 3-methylbutanol, 3-methylbutanal, 2-methylbutanol, 2-methylbutanal,
isobutylthiazole, isovaleronitrile, and isobutyl acetate. These compounds were thought to be
directly derived from their most structurally related BCAAs, but the tomato feeding experiments
shown here with valine and KIV suggest that isobutyl acetate derives directly from KIV and not
72
valine, since KIV precedes valine in the pathway of its synthesis. Similar results were found with
the addition of KIC and KMV and labeled metabolite feeding did not indicate a clear pathway
from BCAAs to volatiles, all of which support our new hypothesis that branched-chain volatiles
are the products of BCKAs and not BCAAs, which is a deviation from the Ehrlich pathway.
Physiol Plant 129: 68-78 Borstlap AC (1972) Changes in free amino-acids of Spirodela polyrhiza (L) Schleiden during
growth inhibition by L-valine, L-isoleucine, or L-leucine - Gas-Chromatographic Study. Acta Botanica Neerlandica 21: 404
Bradford MM (1976) A rapid and sensitive method for the quantitation of microgram quantities
of protein utilizing the principle of protein-dye binding. Anal Biochem 72: 248-254 Buttery RG, Ling LC (1993) Volatile components of tomato fruit and plant-parts - relationship
J Agric Food Chem 35: 540-544 Buttery RG, Teranishi R, Ling LC, Flath RA (1988) Tomato volatiles - composition and
sensory studies. Abstracts of Papers of the American Chemical Society 195: 38 Buttery RG, Teranishi R, Ling LC, Turnbaugh JG (1986) Tomato aroma and volatile
composition. Abstracts of Papers of the American Chemical Society 192: 21 Chen G, Hackett R, Walker D, Taylor A, Lin Z, Grierson D (2004) Identification of a
specific isoform of tomato lipoxygenase (TomloxC) involved in the generation of fatty acid-derived flavor compounds. Plant Physiol 136: 2641-2651
85
Chen WP, Yang XY, Gray W, Cohen JD (2010) Microscale analysis of amino acids using gas chromatography-mass spectrometry after methyl chloroformate derivatization. J Chromatography B, Submitted
Cherepanov PP, Wackernagel W (1995) Gene disruption in Escherichia coli: TcR and KmR cassettes with the option of Flp-catalyzed excision of the antibiotic-resistance determinant. Gene 158: 9-14
Daschner K, Thalheim C, Guha C, Brennicke A, Binder S (1999) In plants a putative
isovaleryl-CoA-dehydrogenase is located in mitochondria. Plant Mol Biol 39: 1275-1282 Dickinson JR, Eshantha L, Salgado J, Hewlins MJE (2003) The catabolism of amino acids to
long chain and complex alcohols in Saccharomyces cerevisiae. J Biol Chem 278: 8028-8034
Dickinson JR, Harrison SJ, Dickinson JA, Hewlins MJE (2000) An investigation of the
metabolism of isoleucine to active amyl alcohol in Saccharomyces cerevisiae. J Biol Chem 275: 10937-10942
Dickinson JR, Harrison SJ, Hewlins MJ (1998) An investigation of the metabolism of valine
to isobutyl alcohol in Saccharomyces cerevisiae. J Biol Chem 273: 25751-25756 Dickinson JR, Lanterman MM, Danner DJ, Pearson BM, Sanz P, Harrison SJ, Hewlins
MJE (1997) A C-13 nuclear magnetic resonance investigation of the metabolism of leucine to isoamyl alcohol in Saccharomyces cerevisiae. J Biol Chem 272: 26871-26878
Diebold R, Schuster J, Daschner K, Binder S (2002) The branched-chain amino acid
transaminase gene family in Arabidopsis encodes plastid and mitochondrial proteins. Plant Physiol 129: 540-550
Ehrlich F (1904) The natural isomerism of leucine [I Announcement ]. Berichte Der Deutschen
Chemischen Gesellschaft 37: 1809-1840 Emanuelsson O, Brunak S, von Heijne G, Nielsen H (2007) Locating proteins in the cell using
TargetP, SignalP and related tools. Nat Protoc 2: 953-971 Eshed Y, Zamir D (1995) An introgression line population of Lycopersicon pennellii in the
cultivated tomato enables the identification and fine mapping of yield-associated QTL. Genetics 141: 1147-1162
Felsenstein J (1985) Confidence-limits on phylogenies - an approach using the bootstrap.
Evolution 39: 783-791 Fridman E, Liu YS, Carmel-Goren L, Gur A, Shoresh M, Pleban T, Eshed Y, Zamir D
(2002) Two tightly linked QTLs modify tomato sugar content via different physiological pathways. Mol Genet Genomics 266: 821-826
86
Fujiki Y, Ito M, Nishida I, Watanabe A (2001) Leucine and its keto acid enhance the
coordinated expression of genes for branched-chain amino acid catabolism in Arabidopsis under sugar starvation. FEBS Lett 499: 161-165
Fujiki Y, Sato T, Ito M, Watanabe A (2000) Isolation and characterization of cDNA clones for
the e1beta and E2 subunits of the branched-chain alpha-ketoacid dehydrogenase complex in Arabidopsis. J Biol Chem 275: 6007-6013
Gao F, Wang C, Wei C, Li Y (2009) A branched-chain aminotransferase may regulate hormone
levels by affecting KNOX genes in plants. Planta 230: 611-623 Gelfand DH, Steinberg RA (1977) Escherichia coli mutants deficient in the aspartate and
LJ (2003) Enzymes of glycolysis are functionally associated with the mitochondrion in Arabidopsis cells. Plant Cell 15: 2140-2151
Gonda I, Bar E, Portnoy V, Lev S, Burger J, Schaffer AA, Tadmor Y, Gepstein S,
Giovannoni JJ, Katzir N, Lewinsohn E (2010) Branched-chain and aromatic amino acid catabolism into aroma volatiles in Cucumis melo L. fruit. J Exp Bot, Epub ahead of print PMID: 20065117
Gu L, Jones AD, Last RL (2010) Broad connections in the Arabidopsis seed metabolic network revealed by metabolite profiling of an amino acid catabolism mutant. Plant J, Epub ahead of print PMID: 19929878
Guadagni DG, Buttery RG, Venstrom DW (1972) Contribution of some volatile tomato components and their mixtures to fresh tomato aroma. J Sci Food Agric 23: 1445-1450
level expression by vectors containing the arabinose PBAD promoter. J Bacteriol 177: 4121-4130
Hagelstein P, Schultz G (1993) Leucine synthesis in spinach chloroplasts: partial
characterization of 2-isopropylmalate synthase. Biol Chem Hoppe Seyler 374: 1105-1108 Hagelstein P, Sieve B, Klein M, Jans H, Schultz G (1997) Leucine synthesis in chloroplasts:
Leucine/isoleucine aminotransferase and valine aminotransferase are different enzymes in spinach chloroplasts. J Plant Physiol 150: 23-30
Hatanaka A (1993) The biogeneration of green odour by green leaves. Phytochemistry 34:
1201-1218
87
Hazelwood LA, Daran JM, van Maris AJ, Pronk JT, Dickinson JR (2008) The Ehrlich pathway for fusel alcohol production: a century of research on Saccharomyces cerevisiae metabolism. Appl Environ Microbiol 74: 2259-2266
Holmberg S, Petersen JG (1988) Regulation of isoleucine-valine biosynthesis in
ocimene activate defense genes and induce resistance against Botrytis cinerea in Arabidopsis thaliana. Plant Cell Physiol 46: 1093-1102
Knill T, Schuster J, Reichelt M, Gershenzon J, Binder S (2008) Arabidopsis branched-chain
aminotransferase 3 functions in both amino acid and glucosinolate biosynthesis. Plant Physiol 146: 1028-1039
Kochevenko A, Maloney GS, Tohge T, Zamir D, Klee HJ, Fernie AR (2010) Characterisation
of two distinct branched chain amino acid transferases with important roles in the synthesis of leucine, isoleucine and valine in tomato. Plant Physiol, Submitted
Kohlhaw GB (2003) Leucine biosynthesis in fungi: Entering metabolism through the back door. Microbiol Mol Biol Rev 67: 1-15
Koncz C, Schell J (1986) The promoter of TL-DNA gene 5 controls the tissue-specific
expression of chimeric genes carried by novel type of Agrobacterium binary vector. Mol Gen Genet 204: 383-396
Kopka J, Schauer N, Krueger S, Birkemeyer C, Usadel B, Bergmuller E, Dormann P,
Weckwerth W, Gibon Y, Stitt M, Willmitzer L, Fernie AR, Steinhauser D (2005) [email protected]: the Golm Metabolome Database. Bioinformatics 21: 1635-1638
Kroumova AB, Xie Z, Wagner GJ (1994) A pathway for the biosynthesis of straight and
branched, odd- and even-length, medium-chain fatty acids in plants. Proc Natl Acad Sci USA 91: 11437-11441
Valentin F, Wallace IM, Wilm A, Lopez R, Thompson JD, Gibson TJ, Higgins DG (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947-2948
Li L, Thipyapong P, Breeden DC, Steffens JC (2003) Overexpression of a bacterial branched-
chain alpha-keto acid dehydrogenase complex in Arabidopsis results in accumulation of branched-chain acyl-CoAs and alteration of free amino acid composition in seeds. Plant Sci 165: 1213-1219
Liepman AH, Olsen LI (2004) Genomic analysis of aminotransferases in Arabidopsis thaliana.
CRC Crit Rev Plant Sci 23: 73-89 Lilly M, Bauer FF, Styger G, Lambrechts MG, Pretorius IS (2006) The effect of increased
branched-chain amino acid transaminase activity in yeast on the production of higher alcohols and on the flavour profiles of wine and distillates. Fems Yeast Res 6: 726-743
Lisec J, Schauer N, Kopka J, Willmitzer L, Fernie AR (2006) Gas chromatography mass
spectrometry-based metabolite profiling in plants. Nat Protoc 1: 387-396 Lucier G (2009) Vegetable and Melons: Tomatoes. In USDA Briefing Rooms, October 5, 2009 Malatrasi M, Corradi M, Svensson JT, Close TJ, Gulli M, Marmiroli N (2006) A branched-
chain amino acid aminotransferase gene isolated from Hordeum vulgare is differentially regulated by drought stress. Theor App Gen 113: 965-976
Massey LK, Conrad RS, Sokatch JR (1974) Regulation of leucine catabolism in
compounds in tomato fruits: identification of loci and potential pathways affecting volatile composition. J Exp Bot 60: 325-337
Matsui K, Kurishita S, Hisamitsu A, Kajiwara T (2000) A lipid-hydrolysing activity involved
in hexenal formation. Biochem Soc Trans 28: 857-860 Maul F, Sargent SA, Sims CA, Baldwin EA, Balaban MO, Huber DJ (2000) Tomato flavor
and aroma quality as affected by storage temperature. J Food Sci 65: 1228-1237 Mayer F, Takeoka G, Buttery R, Nam Y, Naim M, Bezman Y, Rabinowitch H (2003)
Aroma of fresh field tomatoes. In KR Cadwallader, H Weenen, eds. Amer Chemical Soc, pp 144-161
89
McCormick S, Niedermeyer J, Fry J, Barnason A, Horsch R, Fraley R (1986) Leaf disc transformation of cultivated tomato (L. esculentum) using Agrobacterium tumefaciens. Plant Cell Rep 5: 81-84
Schuurink RC (2008) The Arabidopsis her1 mutant implicates GABA in E-2-hexenal responsiveness. Plant J 53: 197-213
Mueller LA, Solow TH, Taylor N, Skwarecki B, Buels R, Binns J, Lin CW, Wright MH,
Ahrens R, Wang Y, Herbst EV, Keyder ER, Menda N, Zamir D, Tanksley SD (2005) The SOL Genomics Network. A comparative resource for Solanaceae biology and beyond. Plant Physiol 138: 1310-1317
Nimitkeatkaia H, Uedaa Y, Furukawaa H, Inamotoa K, Doib M (2005) Emission of
methylbutyric acid from Gypsophila paniculata L. during bud opening: Changes in amino acid catabolism. Sci Hortic 106: 370-380
Ortiz-Serrano P, Gil JV (2007) Quantitation of free and glycosidically bound volatiles in and
effect of glycosidase addition on three tomato varieties (Solanum lycopersicum L.). J Agric Food Chem 55: 9170-9176
Perez AG, Olias R, Luaces P, Sanz C (2002) Biosynthesis of strawberry aroma compounds
through amino acid metabolism. J Agric Food Chem 50: 4037-4042 Powell JT, Morrison JF (1978) Role of the Escherichia coli aromatic amino acid
aminotransferase in leucine biosynthesis. J Bacteriol 136: 1-4 Prohl C, Kispal G, Lill R (2000) Branched-chain-amino-acid transaminases of yeast
Saccharomyces cerevisiae. Methods Enzymol 324: 365-375 Richins RD, Scholthof HB, Shepherd RJ (1987) Sequence of figwort mosaic virus DNA
(caulimovirus group). Nucleic Acids Res 15: 8451-8466 Roessner-Tunali U, Liu J, Leisse A, Balbo I, Perez-Melis A, Willmitzer L, Fernie AR (2004)
Kinetics of labelling of organic and amino acids in potato tubers by gas chromatography-mass spectrometry following incubation in (13)C labeled isotopes. Plant J 39: 668-679
Saitou N, Nei M (1987) The neighbor-joining method: a new method for reconstructing
Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
90
Schauer N, Semel Y, Roessner U, Gur A, Balbo I, Carrari F, Pleban T, Perez-Melis A,
Bruedigam C, Kopka J, Willmitzer L, Zamir D, Fernie AR (2006) Comprehensive metabolic profiling and phenotyping of interspecific introgression lines for tomato improvement. Nat Biotechnol 24: 447-454
Schauer N, Steinhauser D, Strelkov S, Schomburg D, Allison G, Moritz T, Lundgren K,
Roessner-Tunali U, Forbes MG, Willmitzer L, Fernie AR, Kopka J (2005b) GC-MS libraries for the rapid identification of metabolites in complex biological samples. FEBS Lett 579: 1332-1337
Schauer N, Zamir D, Fernie AR (2005a) Metabolic profiling of leaves and fruit of wild species
tomato: a survey of the Solanum lycopersicum complex. J Exp Bot 56: 297-307 Schoondermark-Stolk SA, Jansen M, Veurink JH, Verkleij AJ, Verrips CT, Euverink
GJW, Boonstra J, Dijkhuizen L (2006) Rapid identification of target genes for 3-methyl-1-butanol production in Saccharomyces cerevisiae. Appl Microbiol Biotechnol 70: 237-246
Schulze-Siebert D, Heineke D, Scharf H, Schultz G (1984) Pyruvate-derived amino acids in
spinach chloroplasts: Synthesis and regulation during photosynthetic carbon metabolism. Plant Physiol 76: 465-471
Schuster J, Binder S (2005) The mitochondrial branched-chain aminotransferase (AtBCAT-1) is
capable to initiate degradation of leucine, isoleucine and valine in almost all tissues in Arabidopsis thaliana. Plant Mol Biol 57: 241-254
Schuster J, Knill T, Reichelt M, Gershenzon J, Binder S (2006) BRANCHED-CHAIN
AMINOTRANSFERASE4 is part of the chain elongation pathway in the biosynthesis of methionine-derived glucosinolates in Arabidopsis. Plant Cell 18: 2664-2679
Simkin AJ, Schwartz SH, Auldridge M, Taylor MG, Klee HJ (2004) The tomato carotenoid
cleavage dioxygenase 1 genes contribute to the formation of the flavor volatiles beta-ionone, pseudoionone, and geranylacetone. Plant J 40: 882-892
Singh BK, Shaner DL (1995) Biosynthesis of branched chain amino acids: From test tube to
field. Plant Cell 7: 935-944 Tadmor Y, Fridman E, Gur A, Larkov O, Lastochkin E, Ravid U, Zamir D, Lewinsohn E
(2002) Identification of malodorous, a wild species allele affecting tomato aroma that was selected against during domestication. J Agric Food Chem 50: 2005-2009
Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary genetics
analysis (MEGA) software version 4.0. Mol Biol Evol 24: 1596-1599
91
Tandon KS, Baldwin EA, Shewfelt RL (2000) Aroma perception of individual volatile compounds in fresh tomatoes (Lycopersicon esculentum, Mill.) as affected by the medium of evaluation. Postharvest Biol Technol 20: 261-268
Taylor NL, Heazlewood JL, Day DA, Millar AH (2004) Lipoic acid-dependent oxidative
catabolism of alpha-keto acids in mitochondria provides evidence for branched-chain amino acid catabolism in Arabidopsis. Plant Physiol 134: 838-848
Tieman DM, Taylor M, Schauer N, Fernie AR, Hanson AD, Klee HJ (2006b) Tomato
aromatic amino acid decarboxylases participate in synthesis of the flavor volatiles 2-phenylethanol and 2-phenylacetaldehyde. Proc Natl Acad Sci USA 103: 8287-8292
Tieman DM, Zeigler M, Schmelz EA, Taylor MG, Bliss P, Kirst M, Klee HJ (2006)
Identification of loci affecting flavour volatile emissions in tomato fruits. J Exp Bot 57: 887-896
Tressl R, Drawert F (1973) Biogenesis of banana volatiles. J Agric Food Chem 21: 560-565 Vartak NB, Liu L, Wang BM, Berg CM (1991) A functional leuABCD operon is required for
leucine synthesis by the tyrosine-repressible transaminase in Escherichia coli K-12. J Bacteriol 173: 3864-3871
Vergnais L, Masson F, Montel MC, Berdague JL, Talon R (1998) Evaluation of solid-phase
microextraction for analysis of volatile metabolites produced by Staphylococci. J Agric Food Chem 46: 228-234
Walters DS, Steffens JC (1990) Branched chain amino acid metabolism in the biosynthesis of
Lycopersicon pennellii glucose esters. Plant Physiol 93: 1544-1551 Wyllie SG, Fellman JK (2000) Formation of volatile branched chain esters in bananas (Musa
sapientum L.). J Agric Food Chem 48: 3493-3496 Yoo SD, Cho YH, Sheen J (2007) Arabidopsis mesophyll protoplasts: a versatile cell system for
transient gene expression analysis. Nat Protoc 2: 1565-1572 Zhuang H, Hamilton-Kemp TR, Andersen RA, Hildebrand DF (1996) The impact of
alteration of polyunsaturated fatty acid levels on C6-aldehyde formation of Arabidopsis
thaliana leaves. Plant Physiol 111: 805-812
92
BIOGRAPHICAL SKETCH
Gregory Stephen Maloney was born in Wilmington, Delaware and graduated in 2005from
the University of Delaware with a Bachelors of Science degree in Plant Biology and Landscape
Horticulture. After working at Pioneer Hi-Bred for half a year after college, he started graduate
school at University of Florida, where he earned a Doctor of Philosophy in 2010 in Plant
Molecular and Cellular Biology. From there he went on to a post-doctorate position at Wake