Conversion of Essential Amino Acids Into Aroma Volatiles in Developing Melon Fruit

Conversion of Essential Amino Acids into Aroma Volatiles in Developing Melon Fruit

Thesis submitted in partial fulfillment of the requirements for the degree of “DOCTOR OF PHILOSOPHY”

Itay Gonda

Submitted to the Senate of Ben-Gurion University of the Negev

July 2014

Beer-Sheva

Conversion of Essential Amino Acids into Aroma Volatiles in Developing Melon Fruit

Thesis submitted in partial fulfillment of the requirements for the degree of “DOCTOR OF PHILOSOPHY”

Itay Gonda

Submitted to the Senate of Ben-Gurion University of the Negev

Approved by the advisors: Aaron Fait ______Date:______19/01/15 ____

Efraim Lewinsohn ______Date:_____21/01/15______

Approved by the Dean of the Kreitman School of Advanced Graduate Studies

July 2014

Beer-Sheva

This work was carried out under the supervision of Dr. Aaron Fait In The Jacob Blaustein Institutes for Desert Research The French Associates Institute for Agriculture and Biotechnology of Drylands and Dr. Efraim Lewinsohn Agricultural Research Organization Institute of Plant Sciences, Newe Ya’ar Research Center

The following publications resulted from this thesis:

Gonda, I., Lev, S., Bar, E., Sikron, N., Portnoy, V., Davidovich-Rikanati, R., Burger, J., Schaffer, A., Tadmor, Y., Giovannonni, J.J., Huang, M., Fei, Z., Katzir, N., Fait, A. and Lewinsoh E. (2013) Catabolism of L-methionine in the formation of sulfur and other volatiles in melon (Cucumis melo L.) fruit. Plant J. 61, 458-472.

Gonda, I., Burger, Y., Schaffer, A.A., Ibdah, M., Tadmor, Y., Katzir, N., Fait, A. and Lewinsohn, E. (2015) Biosynthesis and perception of melon aroma. In Havkin- Frenkel, D., Dudai, N. and Belanger, F.C. Eds. Biotechnology in Flavor Production. Blackwell Publishing Ltd., Oxford, UK. In press.

Thanks

First of all, I want to thank Dr. Efraim Lewinsohn, who was my main mentor in this work. He always encouraged to try and test new ideas and never stopped me from preforming new experiments, no matter what. He supported my work and enthusiasm by all means, and always was ready to hear my ideas as well as their criticisms. There are not enough words to describe my gratitude and thankful feelings for him. Many thanks to Dr. Aaron Fait who was my mentor from Ben-Gurion University. Aaron always helped me when I needed him, both scientifically and academically. He knew when to push and when to let go and gave me many good advises and ideas, and I am thankfulness for that. Also, he solved any issues I had towards the university quickly and satisfactory. Special thanks to Dr. Nurit Katzir from Newe-Ya’ar that many of the work described here was carried out in tight collaboration with her and with the Genetics lab and lab members, Vitaly Portnoy, Galil Tzuri and Rotem Harel-Beja. A big thanks to Shery Lev and Dr. Navot Galpaz also at the Genetics lab that helped me both physically and mentally with much of the work I did, especially with the RIL population that was also part of their own projects. I would like also to thank all the members of Efraims’ lab at present and in the past. Especially to Einat Bar and Rachel Davidovich-Rikanati who thought me everything I know about lab work, and also to Raz Krizevski, Karin Rand, Ilan Butnick and Asaf Levy that were always ready to help and gave good ideas and advices. Thanks to all the cucurbits department people who were always pleased to help me with the plants in the field and gave good suggestions when I need them most. In particular to Dr. Yossi Burger and Dr. Kobi Tadmor, as well as to Ayala Meir, Fabian Baumkualler and Uzi Sa’ar. Final thanks to all my friends in Newe-Ya’ar, Zohar Ben-Simhon, Hanoch Glessner, Ari Feder, Noam Chayut and many others that were already mentioned above who made my time there pleasant and valuable.

Research-Student's Affidavit when Submitting the Doctoral Thesis for Judgment

I_____Itay gonda______, whose signature appears below, hereby declare that (Please mark the appropriate statements):

_√_ I have written this Thesis by myself, except for the help and guidance offered by my Thesis Advisors.

_√_ The scientific materials included in this Thesis are products of my own research, culled from the period during which I was a research student.

___ This Thesis incorporates research materials produced in cooperation with others, excluding the technical help commonly received during experimental work. Therefore, I am attaching another affidavit stating the contributions made by myself and the other participants in this research, which has been approved by them and submitted with their approval.

Date: ______22-01-2015______Student's name: ______Itay Gonda ______

Signature:______itay______g. ___

Contents Summary ...... 8 Symbols and abbreviations ...... 10 Units ...... 13 List of figures and tables ...... 14 Literature survey ...... 17 Flavor and aroma ...... 17 The melon ...... 18 Melon aroma ...... 19 Specific characteristics of the melon aroma ...... 20 Melon aroma – biosynthesis and genetic aspects ...... 21 Amino-acid-derived aroma compounds – general aspects ...... 22 The biosynthesis of amino-acid-derived aroma compounds in plants ...... 23 Specific aspects of L-methionine degradation ...... 25 Thioesters – occurrence and biosynthesis ...... 26 The degradation of L-phenylalanine into (E)-cinnamaldehyde and other aroma volatiles ...... 27 Working hypotheses & research objectives ...... 30 Experimental procedures ...... 31 Chemicals ...... 31 Plant material...... 31 Incubation experiments and volatile analyses ...... 31 Gas-chromatography mass-spectrometry volatile analyses ...... 32 Gas-chromatography mass-spectrometry analysis of amino acids ...... 35 Liquid-chromatography mass-spectrometry analysis of amino acids ...... 36 Preparation of cell-free extracts from melon flesh tissues ...... 36 MGL enzymatic assays ...... 37 L-methionine aminotransferase radiolabeled assay ...... 38 L-methionine aminotransferase GC-MS assay ...... 38 L-methionine aminotransferase LC-MS assay ...... 39 Methanethiol acetyl transferase (ThAT) enzymatic assay ...... 39 (E)-cinnamaldehyde coupled enzymatic assays ...... 39 Cloning of CmMGL, CmMetAT and CmThAT1, and their heterologous expression in E. coli ...... 40 Sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) analyses of the recombinant proteins ...... 40 Co-expression of CmThAT1 and CmMGL and their heterologous expression in E. coli ...... 41

Cloning of CmCNL and heterologous expression in E. coli ...... 41 (E)-cinnamic acid:coenzyme A ligase enzymatic assay ...... 42 cDNA preparation and quantitative real-time PCR analysis ...... 43 RNA-seq gene expression analysis ...... 43 Bioinformatic and statistical analyses ...... 44 Results ...... 45 L-methionine-derived aroma volatiles in melon fruits ...... 45 Melon cultivars contained different levels of sulfur aroma volatiles ...... 45 Feeding melon cubes with stable-isotope-labeled L-methionine resulted in enriched volatile aroma compounds ...... 46 L-Isoleucine-derived aroma volatiles in melon cultivars ...... 57 The role of L-methionine-γ-lyase in melon aroma production ...... 59 Soluble protein extracts from mature melon fruits displayed L-methionine- γ-lyase enzymatic activity ...... 59 CmMGL encodes a methionine-γ-lyase enzyme ...... 59 Validation of CmMGL’s role in sulfur volatile formation ...... 64 The role of L-methionine aminotransferase in melon aroma production...... 65 Incubation of melon cubes with KMBA enhanced the levels of +4 enriched compounds but not of +1 enriched compounds ...... 65 Soluble protein extracts from ripe melon fruits displayed L-methionine aminotransferase enzymatic activity ...... 66 CmMetAT encodes a methionine aminotransferase (MetAT) enzyme...... 67 Thioesters production in melon fruits ...... 71 Soluble protein extracts from mature melon fruits displayed methanethiol acetyl transferase enzymatic activity ...... 71 CmThAT1 encodes a methanethiol acyl-transferase enzyme responsible for the accumulation of S-methyl thioacetate in melon fruit flesh ...... 72 The degradation of L-phenylalanine en route to aroma volatiles in melon fruits . 76 Soluble protein extracts from ripe melon fruits displayed (E)-cinnamaldehyde synthesis activities ...... 82 CmCNL encodes a an (E)-cinnamic acid:coenzyme A ligase enzyme...... 84 Discussion ...... 88 L-methionine and L-isoleucine degradation into volatile aroma compounds ...... 88 Incorporation of L-methionine into sulfur-containing aroma compounds ...... 88 Incorporation of L-methionine into L-isoleucine and L-isoleucine-derived aroma compounds ...... 91 Differential incorporation of L-methionine, L-isoleucine and L-leucine into non-sulfur volatiles ...... 92 The role of L-methionine aminotransferase in melon aroma production...... 94

KMBA is a true intermediate in the production of C3-thio-ethers in melon fruit . 94 The role of CmMetAT in volatiles’ production in melon fruit ...... 94 The production of thioesters in melon fruits is mediated by a novel thiol acyl- transferase gene product ...... 96 The catabolism of L-phenylalanine en route to aroma volatiles in melon fruits ... 97 Biosynthesis of (E)-cinnamaldehyde from (E)-cinnamic acid in melon fruit ...... 99 Conclusions ...... 101 Literature cited ...... 103 118 ...... תקציר

Summary The melon (Cucumis melo L., Cucurbitaceae) is an important crop with a worldwide annual production of more than a million tons (2001-2013). Good flavor is an essential characteristic of melon fruit quality. However, in the recent years consumers complain about a loss of fruit tastes that at least partially can be attributed to decreases in aroma production. Melon aroma is composed of a large number of diverse volatile compounds. Some of the most important volatiles in melons and other fruit are derived from essential amino acids. However, the genes and enzymes involved in their formation are only partially characterized. In particular, knowledge is lacking about the formation of sulfur- containing aroma volatiles that are postulated to originate from L-methionine. The aim of my work was to identify and characterize genes and enzymes accountable for the production of these volatiles in the melon fruit, which might be good targets for molecular breeding programs. I show here that L-methionine catabolism en route to aroma volatiles in melon fruits is performed in two parallel pathways. One involves the activity of L-methionine-γ-lyase (MGL) enzyme that cleaves L-methionine to methanethiol, ammonia and α-ketobutyrate. The coding gene for this enzyme, CmMGL, was cloned, functionally characterized and shown to effect the production of methanethiol-derived volatiles in a recombinant inbred line (RIL) population. L-methionine was shown to be a bona fide precursor of L- isoleucine (via α-ketobutyrate), that in turn serves as a precursor for both branched- and straight-chain volatiles. The second pathway involves the activity of an L-methionine aminotransferase enzyme (MetAT). The gene CmMetAT was functionally characterized and its expression was found to be ripening dependent, suggesting it has a role in sulfur volatiles production. The final step in the production of thioesters (methanethiol-derived volatiles), was shown to be catalyzed by a thiol acyl-transferase enzyme (ThAT). The coding gene, CmThAT1, was functionally characterized and the recombinant enzyme was shown to produce thioesters in vitro. CmThAT1 expression is positively correlated with the levels of the thioester S-methyl thioacetate in the aforementioned RIL population. L-phenylalanine was incorporated into 29 different aroma volatiles in the melon fruit, including (E)-cinnamaldehyde, the major aroma compound of cinnamon spice. Melon cell-free extracts converted (E)-cinnamic acid into (E)-cinnamaldehyde, depicting (E)- cinnamic acid:coenzyme A (CoA) ligase (CNL) activity following by cinnamoyl CoA reductase activity (CCR). The CmCNL gene, which expression in the RIL population is

8 positively correlated with the levels of (E)-cinnamyl acetate was functionally characterized. This work sheds light on the catabolic and anabolic processes of amino acids in ripe melon fruit, and will help generating molecular markers to produce melons with superior aroma qualities without compromising other agronomic traits of the crop.

Keywords: melon, aroma volatiles, L-methionine-γ-lyase, L-methionine aminotransferase, thioesters, (E)-cinnamaldehyde, 13C-enrichments, amino-acid catabolism

Symbols and abbreviations AA – amino acids AAT – alcohol acetyltransferase AS – authentic standard ATP – adenosine triphosphate Arg – L-arginine Asn – L-aspargine Asp – L-aspartate BCAA – branched chain amino acid bis-Tris propane – 1,3-bis(tris(hydroxymethyl)methylamino)propane BLAST – basic local alignment search tool BSTFA – N,O-bis(Trimethylsilyl)trifluoroacetamide C-terminus – carboxyl terminal CCR – cinnamoyl coenzyme A reductase CNL – (E)-cinnamic acid:coenzyme A ligase Co-A – coenzyme A cDNA – complementary deoxyribonucleic acid Da – Dalton DAA – days after anthesis DDW – double distilled water DMDS – dimethyl disulfide DMTS – dimethyl trisulfide DNA – deoxyribonucleic acid DTT – dithiothretiol E. coli – Escherichia coli EBI – European Bioinformatics Institute EST – expressed sequence tag et. – ethyl Expasy – expert protein analysis system F.W. – fresh weight GC-MS – gas chromatography mass spectrometry GC-O – gas chromatography olfactometry Gly – glycine

His – L-histidine HMM – hidden Markov model Ile – L-isoleucine IPTG – isopropylthio-β-galactoside α-KB – α-ketobutyrate KI – Kovats index KMBA – α-Keto-(methylthio) butyric acid LB – Luria Bertani LC-TOF-MS – liquid chromatography–time of flight–mass spectrometry Lys – L-lysine [M+H]+ – monoisotopic mass + proton me. – methyl Met – L-methionine MetAT – L-methionine aminotransferase MGL – L-methionine-γ-lyase MTBE – methyl tert-butyl ether NADPH – nicotinamide adenine dinucleotide phosphate (reduced) NCBI – National Center for Biotechnology Information Ni-NTA – nickle-nitrilotriacetic acid PCR – polymerase chain reaction PhyML – phylogenetic maximum likelihood PLP – pyridoxal 5'-phosphate prop. – propanoate PVP – polyvinylpyrrolidone PVPP – polyvinylpolypyrrolidone qRT-PCR – quantitative real time polymerase chain reaction RIL – recombinant inbred line RPKM – kilobase of exon model per million mapped reads RT – room temperature S-compounds – sulfur containing aroma volatiles SEM – standard error mean SD – standard deviation Ser – L-serine SDS-PAGE – sodium dodecyl sulphate polyacrylamide gel electrophoresis 11

SPME – solid phase microextraction ThAT – thiol acyl transferase Thr – L-threonine TMCS – trimethylchlorosilane TMS – trimethylsilyl Tris-HCl – 2-amino-2-(hydroxymethyl)-1,3-propanediol, hydrochloride Tyr – L-tyrosine

Units °C – degree Celsius eV – electron volt h – hour g – gram m/z – mass/charge ratio m – meter μg – microgram μl – microliter mCi – milliCurie mg – milligram ml – milliliter mM – millimolar mmole – millimole min – minute ng – nanogram r – correlation coefficient RPM – rounds per minute V – volt v/v – volume/volume w/v – weight/volume

List of figures and tables Figure 1. Taste and smell sensors in humans. Figure 2. Ehrlich pathway for volatiles’ formation from amino acids in yeasts. Figure 3. Biosynthetic routes for amino acid degradation into volatiles in plants and microorganisms. Figure 4. Degradation of L-methionine via the MGL enzyme. Figure 5. Proposed biosynthetic pathway for S-methyl thioesters from L-methionine. Figure 6. Proposed biosynthetic pathway for (E)-cinnamaldehyde from L- phenylalanine. 13 Figure 7. Incorporation levels of C5-L-methionine into sulfur aroma volatiles and methyl benzoate in ripe melon fruit flesh. 13 Figure 8.Mass spectra of volatiles isolated from melon cubes fed with 5 mM C5-L- methionine. 13 13 Figure 9. Incorporation levels of C5-L-methionine and C6-L-isoleucine into aroma volatiles in ripe melon fruit flesh. 13 13 Figure 10. Incorporation of C6-L-leucine and C6-L-isoleucine into volatile benzyl acetate in ripe melon fruit cubes. 13 13 Figure 11. Differential incorporation of C6-L-leucine and C6-L-isoleucine into the acetate fraction of various acetate esters in ripe melon fruit cubes. 2 Figure 12. Incorporation of H3-[methyl]-L-methionine into aroma volatiles in ripe melon fruit cubes. 2 Figure 13. Incorporation levels of H3-[methyl]-L-methionine into aroma volatiles in ripe melon fruit flesh. Figure 14. 2-Way hierarchical clustering of sulfur-containing aroma volatiles in the fruit. Figure 15. CmMGL gene expression in developing melon fruit flesh as analyzed by qRT-PCR. Figure 16. L-Methionine-γ-lyase activity in melon cell-free extracts and CmMGL. Figure 17. Multiple sequence alignment of amino acids from CmMGL and other functionally characterized MGL enzymes. Figure 18. Purification and molecular mass identification of the recombinant CmMGL. Figure 19. Co-segregation analysis of CmMGL gene expression levels with the levels of S-compounds enriched with +1 m/z unit in PI 414723 × 'Dulce' recombinant inbred

14 lines. Figure 20. Incubation of melon cubes with non-enriched precursors. Figure 21. Methionine aminotransferase activities in soluble protein extracts derived from ripe melon fruit. Figure 22. Relative expression of CmMetAT. Figure 23. Analyses of the recombinant CmMetAT. Figure 24. L-methionine aminotransferase enzymatic activity. Figure 25. CmMetAT substrate preferences. Figure 26. Chemical conversion of methanethiol and acetyl Coenzyme A (Co-A) to S- methyl thioacetate. Figure 27. Methanethiol acetyl transferase enzymatic activities in soluble protein extracts derived from ripe melon fruit. Figure 28. Production of thioesters by bacteria harboring CmThAT1 and CmMGL genes. Figure 29. Methanethiol acetyl transferase enzymatic activities of the recombinant CmThAT1 enzymes. Figure 30. Multivariate correlation analysis of L-phenylalanine-derived aroma volatiles in the fruit flesh throughout the PI 414723 × 'Dulce' RIL population. 13 Figure 31. Incorporation of C9-L-phenylalanine into various aroma volatiles in melon fruit rinds. 13 Figure 32. Incorporation of C9-L-phenylalanine into (E)-cinnamaldehyde in melon rinds. Figure 33. Enzymatic synthesis of (E)-cinnamaldehyde from (E)-cinnamic acid in a ripe melon cell-free extract. Figure 34. Phylogenetic tree of plant CNLs and 4CLs. Figure 35. In vitro enzymatic production of cinnamoyl coenzyme A by the recombinant enzyme CmCNL analyzed by LC-MS. 13 Figure 36. Schematic pathway illustrating the different enrichment patterns of C5-L- methionine during its incorporation into aroma volatiles. Figure 37. Schematic illustration of the ethylene (Yang) cycle.

Table 1. Ions used for enrichment and quantification calculations of L-methionine- and L-isoleucine-derived volatiles. Table 2. Ions used for the enrichment and quantification calculations of L- phenylalanine-derived volatiles. Table 3. Ratios between the levels of sulfur containing aroma volatiles in ripe PI 414723 fruit flesh and their levels in ripe ‘Dulce’ fruit flesh. Volatiles were detected by SPME-GC-MS (n=5). 13 Table 4. Enrichment levels after incubation of melon cubes with C6-L-metthionine. Table 5. Levels of 13C-enrichment in free L-isoleucine upon incubation of melon flesh 13 cubes with 5 mM C5-L-methionine as detected by GC-MS (as TMS derivatives). Table 6. Ratios between the levels of propanoates, 2-methylbutanoates and 2- methylbutyl derivatives aroma volatiles in ripe PI 414723 fruit flesh and their levels in ripe ‘Dulce’ fruit flesh. Table 7. Pairwise correlation analysis of L-phenylalanine-derived volatiles in PI 414723 × ‘Dulce’ RIL population (only significant correlations are shown). 15 13 Table 8. Enrichment levels after incubation of melon cubes with N, C9-L- phenylalanine. Table 9. Candidate genes encoding melon CNLs. Table 10. Pearson correlations coefficient (r) of CNL candidate genes expression with (E)-cinnamoyl CoA-derived volatiles in the RIL population.

Literature survey

Flavor and aroma Flavor is defined as the combination of taste and odor. However, flavor is also influenced by other sensations such as pain, heat, cold and tactile sensations, often referred to as the ‘texture’ of food (Thomson, 1987; Rolls, 2005). Humans can detect five tastes: sweet, salty, sour, bitter and umami (Temussi, 2006). In contrast, humans can distinguish between thousands of odor molecules due to a multigene family encoding 500–1,000 olfactory receptors (Buck and Axel, 1991; Rouquier et al., 2000; Yeshurun and Sobel 2010). The ability of humans to differentiate between so many molecules can be reflected in more than one trillion different odors distinguished by humans (Bushdid et al., 2014). The olfactory receptors are located in the olfactory epithelium in the nasal cavity and detect odorants reaching this area through either the nasal (orthonasal) or retronasal route (Figure 1) (Thomson, 1987; Shepherd, 2006; Huart et al., 2013).

Figure 1. Taste and smell sensors in humans.

Taste and olfactory sensory stimulation are integrated with a variety of sensory inputs, including visual, tactile, and nutrient-sensing, from the gastrointestinal tract to generate the overall flavor perception of specific foods (from Thomson, 1987). In the orthonasal route, the air that contains the volatile molecules is directly taken in through the nose to reach the olfactory receptors, while in the retronasal route, the volatile molecules are taken in through the oral cavity and reach the olfactory receptors via cavities in the throat (Figure 1) (Thomson, 1987; Shepherd, 2006). The odorant must possess certain molecular properties in order to produce an olfactory impression: a certain degree of lipophilicity and sufficiently high vapor pressure so it can be transported to the olfactory system. Also, some water solubility is required to permeate the thin layer of mucus. Furthermore, it must be present at a sufficiently high concentration to be able to interact with one or more of the olfactory receptors (Schwab et al., 2008). The aromas of fruits and vegetables are determined by unique complex combinations of volatile compounds (Thomson, 1987; Kuentzel and Bahri, 1990; Schwab et al., 2008; Dudareva et al., 2013). Although different fruits often share many aroma characteristics, each fruit has a distinctive aroma that is determined by the proportions of key volatiles and the presence or absence of unique components (Kuentzel and Bahri, 1990; Tucker, 1993). There is a general perception by consumers that fruits and vegetables have lost their full flavor as compared to their taste in the past (Baldwin et al., 2000; Klee, 2010). At least in part, that decrease in quality is attributed to alteration in aroma intensity due to intensive breeding for various agronomic traits including extension of shelf-life and pest resistant (Klee, 2010; Klee and Tieman, 2013). It is a current challenge to characterize new genes controlling aroma production that will facilitate the breeding of new cultivars with improved aroma and taste.

The melon The melon (Cucumis melo L.) is a member of the Cucurbitaceae family, which includes many species and genera, such as cucumber (Cucumis sativus L.), watermelon (Citrullus lanatus (Thunb.) Matsum. & Nakai), pumpkin and zucchini (Cucurbita pepo L., Cucurbita maxima Dush., Cucurbita moschata Dush.) (Whitaker and Davis, 1962). The melon is an old crop thought to have originated in the tropical Africa/Middle-East region, and to have spread to the Mediterranean, the Indian sub-continent and throughout Asia (Seymour and McGlasson, 1993; Schaefer et al., 2009). Melons

18 cultivated in Europe before the 15th century tasted much like cucumber, and were harvested unripe and eaten raw, cooked or as pickles (Paris, 2012). Sweet melons as we know them today reached Europe about the late 1400 AD (Paris et al., 2012). Today, melon is one of the most common agricultural crops in Israel and worldwide. In 2012, the total fruit crop of melons in Israel was 37.6 thousand tons, creating an income of more than 121 million NIS. The annual production of melons worldwide in the years 2001-2009 was more than 24 million tons, with about half produced in China. Melon is a highly polymorphic species that comprises a broad array of wild and cultivated genotypes that can be subgrouped according to different traits (Mallick and Masui, 1968; Pitrat et al., 2000). Smith and Welch (1964) divided the various cultivars by the fruit rind characters: reticulatus, cultivars with a netted rind, and inodorus, cultivars with a smooth rind. Another common division is based on the ripening physiology of the fruit, comprising climacteric and non-climacteric cultivars. The ripening of climacteric fruits is characterized by a burst of respiration, usually associated with an increase in ethylene production, followed by upregulation of specific ripening- associated genes and downregulation of other genes. As a result, changes in the color, firmness, texture, taste and aroma of the fruit take place in order to complete ripening process (Dangl et al., 2000). The climacteric cultivars are usually more aromatic than the non-climacteric (Shailt et al., 2001), though the latter have a longer shelf-life. Melons are also classified based on other fruit traits, such as sugar and acid content, rind color, flesh color, shape and size, seed characteristics and the level of total aroma (Seymour and McGlasson, 1993; Pitrat et al., 2000; Burger et al., 2006, 2009).

Melon aroma Acceptable aroma is one of the most important characteristics of melon quality (Seymour and McGlasson, 1993). Unlike sugar accumulation that stops when the fruit is harvested, the formation of aroma molecules actively continues as long as the melons are harvested when ripe (Wyllie et al., 1995; Beaulieu and Grimm, 2001). In aromatic melon varieties, volatile esters are prominent, together with sulfur-containing aroma compounds, sesquiterpenes, norisoprenes, short-chain alcohols and aldehydes (Wyllie and Leach, 1992; Seymour and McGlasson, 1993; Shalit et al., 2001; Jordán et al., 2001; Beaulieu and Grimm, 2001; Shalit et al., 2001; Aubert and Bourger, 2004; Portnoy et al., 2008; Vallone, et al., 2013). Non-aromatic varieties often have much lower levels of total volatiles and lack volatile esters (Shalit et al., 2001; Aubert and

Bourger, 2004; Burger et al., 2006; Perry et al., 2009). Young melon fruits display low levels of aroma volatiles, and only when the fruit ripens does it gain its final and unique aroma profile (Horvat and Senter, 1987; Beaulieu and Grimm, 2001; Shalit et al., 2001; Vallone, et al., 2013; Lignou et al., 2014). Numerous works described the volatile compounds constituting the “aroma of melons” (Yabumoto and Jennings 1977; Buttery et al., 1982; Horvat and Senter 1987; Schieberle et al. 1990; Wyllie and Leach 1992; Wyllie et al., 1994; Beaulieu and Grimm 2001; Jordán et al. 2001; Shalit et al., 2001; Hayata et al., 2003; Aubert and Bourger 2004; Senesi et al., 2005; Aubert and Pitrat 2006; Ibdah et al., 2006; Kourkoutas et al., 2006; Beaulieu and Lancaster 2007; Portnoy et al., 2008; Perry et al., 2009; Escribano et al., 2010; Pang et al., 2012; Vallone, et al., 2013). Volatile esters were found to be the dominant chemical group among melon volatiles, together with alcohols, aldehydes, sulfur compounds and terpenes (Vallone, et al., 2013; Gonda et al., 2014).

Specific characteristics of the melon aroma The particular odor of melons cannot be attributed to a single component. More than 150 different volatiles have been identified in a single sample of melon fruit (Gonda et al., 2014). These volatiles originated from various biosynthetic pathways. Some of these compounds depict classic melon-like, cucumber-like or fruity notes, such as ethyl butanoate, ethyl 2-methylbutyrate, (Z,Z)-3,6-nonadien-1-ol, (E,Z)-2,6-nonadienal, ethyl 2-methylpropanoate, methyl 2-methylbutyrate and ethyl 3- (methylthio)propanoate (Gonda et al., 2014). Other compounds, such as dimethyl trisulfide and 2-(methylthio) ethyl acetate, have unpleasant or sulfuric notes (Wyllie et al., 1994; Lignou et al., 2014). Using GC-olfactometry (GC-O), Schieberle et al. (1990) described 12 odorants that significantly contribute to the melon aroma, including ethyl- 2-methylpropanoate, methyl-2-methylbutanoate and ethyl butanoate, which account for fruity and sweet aroma notes. Wyllie and Leach (1992) measured the amounts of five different thio-ether esters in 26 different melon cultivars. They found that in almost all cases, at least one of the thio-ether esters was significantly abundant at a level, regarding its odor threshold, to make a marked contribution to the fruit flavor. Wyllie et al., (1994) used GC-O to show that some sulfur-containing volatiles cause unpleasant aromas in melons. More recently, also using GC-O, Jordán et al. (2001) revealed 25 odorants with significant contributions to melon aroma. They included ethyl 2- methylpropanoate, 2-methylpropyl acetate, ethyl-2-methylbutanoate, benzyl alcohol,

20 ethyl-3-(methylthio) propanoate and benzyl butanoate that confer sweet, fruity and cantaloupe-like aroma notes. Thus, each cultivar has a unique aroma profile that is composed by the absolute quantities of the volatiles and the relative proportions between them (Allwood et al., 2014). Then, differences in these profiles are crucial in forming the unique flavor and aroma of each cultivar.

Melon aroma – biosynthesis and genetic aspects In the past few years, research dealing with the biosynthesis of the melon’s aroma compounds has been published (Shalit et al., 2001; Yahyaoui et al., 2002; El-Sharkawy et al., 2005; Manríquez et al., 2006; Ibdah et al., 2006; Portnoy et al., 2008; Luchetta et al., 2008; Gonda et al., 2010). Manríquez et al. (2006) identified and characterized two distinctive alcohol dehydrogenase genes (ADH) (Cm-ADH1 and Cm- ADH2), accounting for the formation of various alcohols from aldehyde substrates in the melon fruit (see Figure 2). Shalit et al. (2001) found alcohol acetyltransferase (AAT) activity in melon cell-free extracts involved in acetate esters formation. Yahyaoui et al. (2002) characterized two genes encoding alcohol acyl transferase (Cm- AAT1 and Cm-AAT2), responsible for the biosynthesis of different volatile esters, and El-Sharkawy et al. (2005) characterized two additional melon AAT's (Cm-AAT3 and Cm-AAT4). Luchetta et al. (2008) characterized the enzymatic activity of CmAAT1-4 with thio-ether alcohols to form esters containing a thiol group. The expression of these genes is associated with the ethylene production of the fruit and are inhibited in transgenic melons unable to produce ethylene (Yahyaoui et al., 2002; El-Sharkawy et al., 2005; Manríquez et al., 2006). Ibdah et al. (2006) identified a carotenoid cleavage dioxygenase gene (CmCCD1) able to generate geranylacetone, β-ionone, α-ionone and pseudoionone from various carotenoid substrates. Portnoy et al. (2008) discovered two sesquiterpene synthase genes (CmTpsNY and CmTpsDul) from two different cultivars (‘NoyYizre'el’ and ‘Dulce’), each of them utilized the same substrate (farnesyl diphosphate) to generate different sesquiterpene products. CmTpsNY, isolated from ‘NoyYizre'el’, generated δ-cadinene, γ-cadinene and α-copaene in vitro, which are the major sesquiterpenes accumulated in 'NoyYizre'el' fruit rinds. CmTpsDul, isolated from the ‘Dulce’ cultivar, generated only α-farnesene in vitro, the sole sesquiterpene accumulated in the rind tissue of ‘Dulce’ fruits. Gonda et al. (2010) characterized a branched-chain amino-acid aminotransferase (CmBCAT1) and an aromatic amino-acid aminotransferase (CmArAT1) involved in the degradation of these amino acids en route

21 to aroma volatiles. All of these eleven genes were found to be upregulated in the last stages of fruit development (Yahyaoui et al., 2002; El-Sharkawy et al., 2005; Manríquez et al., 2006; Ibdah et al., 2006; Portnoy et al., 2008; Gonda et al., 2010), in a coordinated manner with the accumulation of the melon volatiles (Seymour and McGlasson, 1993; Beaulieu and Grimm, 2001; Shalit et al., 2001; Vallone, et al., 2013; Lignou et al., 2014).

Amino-acid-derived aroma compounds – general aspects Free amino acid content increases during melon ripening (Wyllie et al., 1995), and aroma volatiles (esters in particular) derived from amino acids are major contributors to the aroma of melons (Tressl and Drawert 1973; Yabumoyo and Jenings 1977; Buttery et al., 1982; Schieberle et al., 1990; Wyllie et al., 1994; Wyllie et al., 1995; Jordán et al., 2001; Beaulieu and Grimm, 2001; Hayata et al., 2003; Aubert and Bourger, 2004; Kourkoutas et al., 2006; Pang et al., 2012; Vallone, et al., 2013; Lignou et al., 2014) and other fruits (Buttery et al., 1987; Wyllie and Fellman, 2000; Baldwin et al., 2000; Matich and Rowan, 2007), as well of cheese and milk products (Yvon and Rijnen, 2001; Liu et al., 2008) and yeast-fermented products, such as beer and bread (Dickinson et al., 2003; Perpète et al., 2006). The amino acids that are further catabolized to aroma volatiles are usually essential amino acids and include the aromatic amino acids (L- phenylalanine, L-tyrosine and L-tryptophan), the branched-chain amino acids (BCAAs: L-leucine, L-isoleucine and L-valine) and sulfur-containing amino acids (L-methionine and L-cysteine) (Yvon and Rijnen, 2001; Goff and Klee, 2006; Liu et al., 2008). Reports regarding the formation of amino-acid-derived volatiles in cheese-dwelling microorganisms are available (see Liu et al., 2008). The first step of the pathway is relatively well understood, and in the majority of cases, it initially involves transamination (Figure 2), followed by decarboxylation (Yvon et al., 2000; Yvon and Rijnen, 2001). Further observations have indicated that this initial step leads to the formation of aroma compounds (Yvon and Rijnen, 2001; Liu et al., 2008). It had also been found that yeasts utilize the same pathway (Dickinson et al., 2003; Perpète et al., 2006), as was suggested at the beginning of the twentieth century by Ehrlich (Ehrlich, 1907) (Figure 2).

Figure 2. Ehrlich pathway for volatiles’ formation from amino acids in yeasts. The pathway includes transamination followed by decarboxylation and reduction to generate volatile alcohols (Ehrlich 1907; Dickinson et al., 2003).

The biosynthesis of amino-acid-derived aroma compounds in plants The last two steps in the formation of volatile esters (reduction to alcohol by ADH and ester formation by AAT) in plants have been extensively investigated, both in melons (Shalit et al., 2001, Yahyaoui et al., 2002; El-Sharkawy et al., 2005, Manríquez et al., 2006; Luchetta et al., 2008) and in other plants (Pérez et al., 1993; 1996; Beekwilder et al., 2004; Boatright et al., 2004; Larkov et al., 2008; Zaks et al., 2008). However, knowledge about the initial step of the pathway that involves the catabolism of amino acids for the formation of aldehydes was quite limited until recently. Studies performed in transgenic tomato (Solanum lycopersicum L.) fruit showed that the catabolism of L-phenylalanine into aroma volatiles is initially mediated by aromatic amino acid decarboxylases, releasing an amine intermediate (Tieman et al., 2006) (Figure 3, red route, solid arrows). It has been proposed that deamination of the reaction product, 2-phenethylamine, releases phenylacetaldehyde, but this step has not been confirmed (Figure 3, red route, dashed arrows). In contrast, it seems that a different biosynthetic route takes place in the petals of petunia (Petunia hybrida) and rose (Rosa hybrida), in which one enzyme is able both to decarboxylate and deaminate L- phenylalanine to release phenylacetaldehyde (Kaminaga et al., 2006; Farhi et al., 2010) (Figure 3, purple route). Flowers of transgenic petunia plants with the suppressed expression of the phenylacetaldehyde synthase gene (PhPAAS) did not emit phenylacetaldehyde and 2-phenylethanol as compared with non-silenced controls (Kaminaga et al., 2006). Gonda et al. (2010) showed the contribution of the BCAAs, L-phenylalanine and L-methionine, as well as their respective α-keto-acids, to melon

Figure 3. Biosynthetic routes for amino acid degradation to volatiles in plants and microorganisms. Red route: first decarboxylation, followed by deamination with an amine intermediate, catalyzed by amino acid decarboxylases and deaminases, as partially demonstrated in tomato fruit (Tieman et al., 2006). Purple route: aldehyde synthesis in a single enzymatic step with no free intermediates as was demonstrated in petunia and rose petals (Kaminaga et al., 2006; Farhi et al., 2010). Green route (The Erlich pathway): transamination followed by decarboxylation with an α-keto acid intermediate, catalyzed by two separate enzymes as partially demonstrated in melon fruit (Gonda et al., 2010), and in some cheese-dwelling microorganisms (Rijnen et al., 1999), as well as in several strains of Saccharomyces cerevisiae (Dickinson et al., 2003; Vuralhan et al., 2003). Solid arrows indicate that the enzyme activity and genes have been identified. Dashed arrows indicate that the proposed plant enzymes and genes are yet to be identified (Schwab et al., 2008), although the α-keto acid decarboxylases have been identified in microorganisms (Yvon and Rijnen 2001; Vuralhan et al., 2003).

aroma, and they characterized two new genes, a branched-chain amino-acid aminotransferase (CmBCAT1) and an aromatic amino-acid aminotransferase (CmArAT1) whose expressions increase in the ripe fruit. It was previously demonstrated that exogenous L-methionine causes increases in many sulfur-containing

24 volatiles in melon fruit slices (Gonda et al., 2010). Additionally, L-methionine aminotransferase activity was detected in crude extracts from ripe melon fruit (Gonda et al., 2010), but the gene responsible for this activity was unknown.

Specific aspects of L-methionine degradation

Sulfur containing aroma volatiles are important contributors to melon aroma (Wyllie and Leach, 1992; Wyllie et al., 1994; Jordán et al. 2001; Kourkoutas et al., 2006; Vallone, et al., 2013; Lignou et al., 2014) and to the aroma of other fruits such as tomato (Tieman et al., 2012), potato (Buttery et al., 1973; Jansky 2010; McGorrin 2011), passion fruit (Mussinan and Kellan, 1994; Werkhoff et al., 1998), kiwi (Günther et al., 2010), and pineapple (Preston et al., 2003; Zheng et al., 2012), and to agricultural products such as meat, bread and cheese (Mussinan and Kellan, 1994; Liu et al., 2004; Landaud et al., 2008). Sulfur volatiles can originate from both L-cysteine and L- methionine, but the latter is more common to plants volatiles (see Liu et al., 2008). These volatiles can be the result of an L-methionine aminotransferase enzyme generating α-keto-γ-methylthio butyric acid (KMBA) as an intermediate (Gonda et al., 2010). Aromatic and branched-chain amino-acid aminotransferase enzymes and their coding genes have been characterized both in melons (Gonda et al., 2010) and in other plants (Diebold et al., 2002; Schuster and Binder, 2005; Maeda et al., 2011; Yoo et al., 2013) and fruits (Tieman et al., 2006; Maloney et al.,2010; Dal Cin et al., 2011; Kochevenko et al. 2012). However, only a few reports addressing specific L- methionine aminotransferases are available. A methionine:glyoxylate aminotransferase was purified and characterized from leaf extracts of Brassica carinata and Brassica napus (Chapple et al., 1990), but the coding gene was not identified. Schuster et al. (2006) found that the Arabidopsis thaliana branched-chain amino-acids aminotransferase4 (AtBCAT4) encodes an L-methionine aminotransferase enzyme that is not active towards BCAAs. However, in melons it seems that this is not the case since the closest orthologue of AtBCAT4, CmBCAT1, shows significant activity in vitro towards all BCAAs after expression in E. coli (Gonda et al., 2010), but not towards methionine (unpublished results). In addition, it is accepted that the last step of methionine recycling in the ethylene (Yang) cycle is mediated by an L-methionine aminotransferase (Miyazaki and Yang, 1987; Pommerrenig et al., 2011), but the coding gene have not been characterized. A similar enzyme, namely glutamine transaminase

K (GTK), was characterized in mammals and bacteria as the last enzyme in the methionine salvage pathway (Albers, 2009; Jaisson et al., 2009). That enzyme converts L-glutamine and KMBA substrates to α-ketoglutaramate (keto-glutamine) and L- methionine, respectively. However, since aminotransferases are fully reversible enzymes, GTK might also catalyze the reaction from L-methionine to KMBA as was demonstrated with other aminotransferases (Schuster et al., 2006). Another possible pathway for volatiles formation from L-methionine is demethiolation, which can occur chemically in some conditions or enzymatically via an L-methionine- γ-lyase enzyme (MGL) (Liu et al., 2008). An Arabidopsis gene encoding an L- methionine-γ-lyase enzyme (AtMGL) that cleaves L-methionine into methanethiol, α- ketobutyrate and ammonia has been characterized (Figure 4; Rébeillé et al., 2006). Figure 4. Catabolism of L-methionine and its conversion to L-isoleucine. The involvement of an L-methionine-γ-lyase enzyme as was demonstrated in Arabidopsis (Rébeillé et al., 2006).

However, the relation for aroma volatiles formation was hardly examined if at all. In addition, no such research was conducted in fruits. One of the products of MGL reaction, methanethiol, is known to be a precursor of many important sulfur-containing volatiles in microorganisms (Yvon and Rijnen, 2001; Dickinson et al., 2003). Another product of the MGL reaction, α-ketobutyrate, might serve as a metabolic intermediate in L-isoleucine biosynthesis as demonstrated in Arabidopsis (Rébeillé et al., 2006; Joshi and Jander, 2009). Still, although L-isoleucine contributes to the formation of aroma volatiles in melon fruit (Gonda et al., 2010), the contribution of L-methionine to L-isoleucine biosynthesis en route to the formation of aroma compounds in melons or other fruits has remained unexplored.

Thioesters – occurrence and biosynthesis Thioesters are important compounds in many food products (Liu et al., 2004; Landaud et al., 2008: Tapp et al., 2008; Vandendriessche et al., 2013; Wang and Lin 2014). Methanethiol, released from L-methionine by the activity of MGL (see above), is a

26 proposed substrate for the production of S-methyl thioesters (Figure 5). Thiols might replace the alcohol substrates in the reaction of the alcohol acyl-transferase enzyme to produce S-methyl thioesters (Figure 5, Bamforth and Kanauchi, 2003). Although considered as a somewhat spontaneous reaction, the in vitro enzymatic production of thioesters from a thiol (mostly methanethiol) and acyl coenzyme A (Co-A) substrate was demonstrated, but only in cell-free extracts of the fungus Geotrichum candidum (Helinck et al., 2000) and strawberry fruits (Noichinda et al., 1999). Tapp et al. (2008) isolated and characterized a thiosterase enzyme from passion fruit that cleaves thioester substrates and release free thiols present in the ripe fruit. However, genes encoding specific acyl-transferase enzymes that utilize thiol substrates, instead of alcohols (see Figure 5), have not yet been characterized, either from plants or from microorganisms.

Figure 5. Proposed biosynthetic pathway for S-methyl thioesters from L-methionine. The pathway includes demethiolation catalyzed by an MGL enzyme followed by S-acylation catalyzed by a ThAT enzyme. The biosynthesis of O-esters mediated by an AAT enzyme is shown at the bottom.

The degradation of L-phenylalanine into (E)-cinnamaldehyde and other aroma volatiles Many important plant volatiles are derived from L-phenylalanine-and catabolic pathways involved in their formation are known to occur in various plants (Boatright et al., 2004; Kaminaga et al., 2006; Tieman et al., 2006; Schwab et al., 2008, Gonda et al., 2010; Dudareva et al., 2013) and microorganisms (Yvon et al., 2000; Yvon and Rijnen, 2001; Liu et al., 2008; Dickinson et al., 2003; Perpète et al., 2006). L- phenylalanine derived aroma volatiles can be classified according to the length and type

27 of the carbon chain attached to the aromatic ring. C6―C1 and C6―C2 volatiles have one or two carbons in that chain, respectively, which their biosynthesis require a decarboxylation or β-oxidation steps (Boatright et al., 2004). Their production in plants and microorganisms was discussed in detail above. Melons also contain C6―C3

(phenylpropanoids) and C6═C3 (phenylpropenes) volatiles with three carbons in the chain coupled by a single or double bond (respectively) between the β and γ carbons

(Gonda et al., 2014). The production of hydroxylated and methoxylated C6═C3 and

C6―C3 (and some C6―C1) compounds was documented in many plants in various organs (reviewed in detail in Vogt, 2010 and Dudareva et al., 2013). Their biosynthesis is initiated by L-phenylalanine ammonia-lyase (PAL) enzyme generating (E)-cinnamic acid, followed by (E)-cinnamic acid 4-hydroxylase (C4H) enzyme generating p- coumaric acid. Subsequently, more volatile specific enzymes are involved, depending on the compound and the species. For example, it was shown that by the activity of NADPH-dependent reductases, leaves of sweet basil produce eugenol from coniferyl acetate and petunia hybrida petals produce isoeugenol from the same substrate (Koeduka et al., 2006). More recently, Muhlemann et al. (2014) showed that down regulation of the petunia cinnamoyl-CoA reductase (PhCCR1) resulted in enhanced levels of vanillin. However, the biosynthesis of non-substituted C6═C3 as well as

C6―C3 volatiles is much less understood. (E)-Cinnamaldehyde, the main aroma compound of the spice cinnamon (Cinnamomum spp.) (Senanayake et al., 1978; Paranagama et al., 2001), presents in low levels in various plants, including melon fruits (Gonda et al., 2014). However, its full biosynthetic pathway is not well understood. Although some melon cultivars accumulate (E)-cinnamaldehyde (as well as its reduction and further acetylation products (E)-cinnamyl alcohol and (E)-cinnamyl acetate), little is known about the catabolic pathways from L-phenylalanine that bring about the formation of (E)-cinnamaldehyde and its derivatives. It is suggested that (E)- cinnamaldehyde is biosynthesized from L-phenylalanine, initially via (E)-cinnamic acid (Figure 6) produced by the activity of L-phenylalanine ammonia-lyase (PAL). (E)- cinnamic acid is then converted to cinnamoyl coenzyme A by the activity of an (E)- cinnamic acid:CoA ligase (4CL) enzyme (Figure 6). Later, the cinnamoyl coenzyme A is reduced to generate (E)-cinnamaldehyde by a putative cinnamoyl coenzyme A reductase enzyme (CCR) (Figure 6). Similar reactions have been characterized in numerous plants (Goffner et al., 1994; Beuerle and Pichersky, 2002); however, most 4CL enzymes readily utilize p-coumaric, ferulic and caffeic acids, but cannot efficiently

O use (E)-cinnamic acid as a substrate. Recently,

OH Klempien et al. (2012) characterized a new

NH2 gene from Petunia hybrida, (E)-cinnamic L-phenylalanine acid:CoA ligase (PhCNL), that efficiently uses (E)-cinnamic acid as a substrate. A similar

phenylalanine enzyme has also been reported in Hypericum ammonia lyase calycinum (Gaid et al., 2012 Earlier, the Arabidopsis thaliana BZO1 enzyme was NH3 O shown to possesses benzoic acid:CoA ligase enzyme activity, and homozygous T-DNA OH insertion bzo1 mutants accumulated altered levels of benzoyl-glucosinolates (Kliebenstein (E)-cinnamic acid et al., 2007). In view of the characterized coenzyme A, ATP PhCNL, Lee et al. (2012) showed that the ++ (E)-cinnamic acid: Mg Arabidopsis BZO1 utilizes (E)-cinnamic acid

CoA ligase 5 times more efficiently than benzoic acid as AMP, PPi substrate. Moreover, bzo1 mutants fed with O exogenous benzoic acid produced benzoyl-

CoA glucosinolates, indicating that in vivo, BZO1 acts as a CNL enzyme upstream to benzoic (E)-cinnamoyl coenzyme A acid production (Lee et al., 2012).

+ NADPH + H

cinnamoyl CoA reductase

+ NADP O

(E)-cinnamaldehyde Figure 6. Proposed biosynthetic pathway for (E)-cinnamaldehyde from L-phenylalanine.

Working hypotheses & research objectives

Working hypotheses The degradation of L-methionine into sulfur and other aroma volatiles is initiated by a methionine aminotransferase (MetAT) enzyme, in a similar pattern to the newly discovered pathway for aromatic and branched-chain amino acid catabolism into volatiles. That enzyme produces KMBA as non-volatile intermediate en route to volatile compounds. Due to the unique nature of L-methionine being a sulfur amino-acid, in addition to the aminotransferase route, L-methionine catabolism involves other catabolic enzymes such as L-methionine-γ-lyase (MGL). Volatile thioesters (S-esters), which are sulfur-containing aroma compounds, originate from L-methionine, and the last step in their biosynthesis is mediated by a thiol acyl- transferase (ThAT) enzyme using methanethiol and acyl CoA's as substrates, in a parallel manner to the production of O-esters by alcohol acyl-transferase (AAT) enzymes. The degradation of L-phenylalanine into (E)-cinnamaldehyde in melon fruits is mediated, in part, by the enzymatic activity of (E)-cinnamic acid:coenzyme A ligase (CNL) enzyme that converts (E)-cinnamic acid to (E)-cinnamoyl CoA. Subsequently cinnamoyl CoA reductase (CCR) enzyme converts the (E)-cinnamoyl CoA into (E)- cinnamaldehyde.

Research objectives 1. To elucidate the metabolic pathways of L-methionine and other essential amino acids into aroma volatiles in ripening melon fruit using stable-isotope-labeled precursor feeding experiments. The incorporation of the labeled precursors into the aroma volatiles will be tracked using mass-spectrometry techniques in gas-chromatography and liquid-chromatography platforms. The pattern of the labeling will allow us to distinguish between different active catabolic pathways in the ripe fruit. 2. To isolate the biosynthetic genes involved in the process of the conversion of L-methionine, L-phenylalanine and other essential amino acids into melon aroma volatiles. This will be done by exploiting the large infrastructure available in melon research that includes a published genome, large EST libraries and various genetic populations. 3. To functionally characterize enzymes encoding genes that participate in the formation of melon aroma volatiles derived from amino acids. This will be performed in heterologous systems (mainly in E. coli) using both in vitro and in vivo analyses.

Experimental procedures Chemicals All chemicals were purchased from Sigma-Aldrich Co. (http://www.sigmaaldrich.com/israel.html) unless otherwise indicated. 2 13 15 [ H3-methyl]-L-Methionine, [U- C5, 97-99%; N, 97-99%]-L-methionine, 13 13 15 [U- C6, 95%+]-L-isoleucine and [U- C6, 97-99%; N, 97-99%]-L-leucine were obtained from Cambridge Isotope Laboratories, Inc. (http://www.isotope.com/cil/products/searchproducts.cfm). The radiolabeled 3 [ H3-methyl]-L-methionine was obtained from Perkin Elmer (http://www.perkinelmer.com/).

Plant material Melons (Cucumis melo L. var. ‘Dulce’, accession PI 414723 and the recombinant inbred line population of PI 414723 × ‘Dulce’, Danin-Poleg et al., 2000) were grown in an open field with drip irrigation and fertilization under commercial conditions in the summers of 2010, 2011 and 2012 (the RILs were grown only in 2011) at the Newe- Ya’ar Research Center in northern Israel. Fruit samples were collected 10, 20 and 30 days after anthesis (DAA) and at the ripe stage (after color change and development of the abscission zone).

Incubation experiments and volatile analyses Melon cubes (about 0.5 g each) cut from the flesh of mature fruit were put in sterile Petri dish plates, and 50 μl of a solution of either 5 mM labeled or non-labeled amino acids was gently applied onto the top of each cube. The plate was covered and incubated for the time specified at room temperature. Then, each cube was frozen in liquid nitrogen and ground into a uniform powder using a grinding machine (IKA, Sigma- Aldrich). The powder (1 g) was placed in a 10 ml glass vial containing 0.7 g solid NaCl. To each vial, 2 ml of a 20 % (w/v) NaCl solution and 0.2 μg of 2-heptanone, which was used as an internal standard, were added. The vial was then sealed and stored at 4°C, for no longer than one week until analysis. Solid-Phase Micro- Extraction (SPME) sampling was conducted according to Davidovich-Rikanati et al. (2007) with slight modifications. The sample was preheated to 30°C and agitated for 5 min at 500 RPM, and then a 65 μm fused silica fiber coated with

31 polydimethylsiloxane/divinylbenzene/ carboxen (PDMS/DVB/CAR) SPME fiber (SupelcoInc., http://www.sigmaaldrich.com/analytical-chromatography.html) was inserted into the vial and exposed to the sample headspace. After 25 min, the SPME fiber was introduced into the injector port of the GC-MS apparatus for further analysis (see below).

Gas-chromatography mass-spectrometry volatile analyses Volatile compounds were analyzed on a GC-MS apparatus (Agilent Technologies, http://www.home.agilent.com/agilent/home.jspx?cc=US&lc=eng) equipped with an RXI-5SIL MS (30 m × 0.25 mm, 0.25 µm) fused-silica capillary column (Restek Co., http://www.restek.com/). Helium (0.8 ml × min-1) was used as a carrier gas. The injector temperature was 250°C, set for splitless injection. The oven was set to 50°C for 1 min, and then the temperature was increased to 180°C at a rate of 5°C × min-1, then to 260°C at 20°C × min-1. Thermal desorption was allowed for 10 min. The detector temperature was 280°C. The mass range was recorded from 41 to 250 m/z, with an electron energy of 70 eV. A mixture of straight-chain alkanes (C7–C23) was injected into the column under the aforementioned conditions for the determination of retention indices. The identification and quantification of the volatiles were done according to Davidovich- Rikanati et al. (2008) and Gonda et al. (2010). The ratio between the values obtained from PI 414723 and ‘Dulce’ was determined based on the calculated areas. 2- Heptanone was used as an internal standard. The % of label enrichment was calculated as described below: [area of enriched ion] { 표푓 𝑖푛푐푢푏푎푡𝑖표푛 푤𝑖푡ℎ 푙푎푏푒푙푒푑 푝푟푒푐푢푟푠표푟} − (area of enriched + non−enriched ions) [area of enriched ion] { 표푓 𝑖푛푐푢푏푎푡𝑖표푛 푤𝑖푡ℎ 푛표푛 − 푙푎푏푒푙푒푑 푝푟푒푐푢푟푠표푟} (area of enriched + non−enriched ions) For a complete list of the ions used for quantification and enrichment levels, see Tables 1 and 2.

Table 1. Ions used for enrichment and quantification calculations of L-methionine- and L-isoleucine- derived volatiles. E Ions E Ions E Ions 13 2 13 Q Compound name +[m/z] C5-met H3-met C6-ile Ion incubations incubations incubations 2-(methylthio) ethanol - - - - 61 methyl benzoate +1 136/137* - - - (47+48)/ methanethiol +1 47/50* - 47* (48+49)* S-methyl propanethioate +1 75/76 104/107* - 104* [methylthio moiety] S-methyl 2-methyl butanethioate +1 75/76 75/78 - 85 [methylthio moiety] S-methyl butanethioate +1 118/119* 118/121 - 118* S-methyl 2-methylpropanethioate +1 118/119* 118/121 - 118* S-methyl thioacetate +1 90/91* 90/93 - 90* dimethyl trisulfide +1 +1 94/95* 94/97* - 94* dimethyl disulfide +1 +1 126/127* 126/129* - 126* dimethyl disulfide +2 +2 94/96* 94/100* - - dimethyl trisulfide +2 +2 126/128* 126/132* - - ethyl2-(methylthio) acetate +1 134/135* 134/137* - 134* methyl2-(methylthio) acetate +1 120/121* 120/123* - 61 2-(methylthio) ethyl acetate +1 74/75 74/77 - - 3-(methylthio) propanal +4 104/108* 104/107* - - ethyl 3-(methylthio) propanoate +4 148/152* 148/151* - 74 methyl 3-methylthiopropanoate +4 134/138* 134/137* - 134* 3-(methylthio) propyl acetate +4 148/152* 148/151* - 88 3-(methylthio)propanol +4 106/110* 106/109* - - propyl propanoate +3 57/60 - 57/60 57 S-methyl propanethioate +3 57/60 - 104/107* - [propanoate moiety] ethyl propanoate +3 57/60 - 57/60 57 2-methylpropyl propanoate +3 57/60 - 57/60 57 methyl propanoate +3 88/91* - 88/91* 57 2-methyl butyl propanoate +3 57/60 - 57/60 57 [propanoate moiety] 2-methylbutyl propanoate +3/+5 70/73 - 70/75 - [isoleucine moiety] ethyl-2-methylbutanoate +3/+5 57/59 - 102/105 102 butyl-2-methybutanoate +3/+5 103/106 - 103/108 57 S-methyl 2-methylbutanethioate +3/+5 85/88 - 85/90 - [isoleucine moiety] 2-methyl-2-butenal +3/+5 84/87* - 84/89* 84* 2-methylpropyl 2- +3/+5 103/106 - 103/108 57 methylbutanoate 2-methylbutyl-2-methylbutanoate +3/+5 85/88 - 85/90 57 propyl-2-methylbutanoate +3/+5 103/106 - 103/108 103 2-methylbutyl acetate +3/+5 70/73 - 70/75 43 2-methylbutanol +3/+5 70/73 - 70/75 57 2-methylbutanal +5 81/86 - (E)-ethyl 2-methylbutenoate +5 100/105 - methyl-2-methylbutanoate +3/+5 101/104 101/105 88 E Ions - ions used for the enrichment calculations; Q Ion – ion used for quantifications;*-molecular ion.

Table 2. Ions used for enrichment and quantification calculations of L-phenylalanine-derived volatiles. E Ions Enrichment 13 15 Compound name Q Ion C9, N1-L-phe pattern [m/z] incubations benzaldehyde 105 +7 106 / 113* benzyl alcohol 79 +7 108 / 115* 2-phenylacetaldehyde 91 +8 120 / 128* methyl benzoate 77 +7 136 / 143* 2-phenetyl alcohol 91 +8 122 / 130* 2-phenylacetonitrile +8 117 +8 125 / 126* 2-phenylacetonitrile +9 -- +9 125 / 127* benzyl acetate 108 +7 150 / 157* benzoic acid 122 -- -- 3-phenyl propanol 117 +9 136 / 145* 2-phenethyl acetate 104 +8 104 / 112 benzyl propanoate 108 +7 164 / 171* (E)-cinnamaldehyde 131* +9 131 / 140* (E)-cinnamyl alcohol 134* +9 134 / 143* eugenol 164* -- -- 3-phenylpropyl acetate 118 +9 117 / 126 methyl (E)-cinnamate 131 +9 162 / 171* (E)-cinnamic acid 147 -- -- (E)-cinnamyl acetate 115 +9 176 / 185* ethyl benzoate 105 +7 122 / 129* ethyl phenyl acetate 91 +8 164 / 172* ethyl (E)-cinnamate 131 -- -- 1-phenethyl alcohol 107 +8 122 / 130* 1-phenylethanone 105 +8 120 / 128* 1-phenethyl acetate 122 +8 164 / 172* benzyl 2-methyl propanoate 91 +7 178 / 185* 2-phenyl nitro ethane 104 -- -- benzyl butanoate 91 +7 178 / 185* vanillin 151* -- -- methyl vanillate 151 -- -- benzyl 2/3-methyl butanoate -- +7 192 / 199* E Ions - ions used for the enrichment calculations; Q Ion – ion used for quantifications;*-molecular ion.

- 34 -

Gas-chromatography mass-spectrometry analysis of amino acids For amino acid analyses, 250 mg of finely ground melon was added to 1 ml of a pre- chilled methanol/water/chloroform mix (25/10/10, v/v/v), vortexed for 10 min at 4°C, sonicated for 10 min at room temperature and centrifuged 10 min at 20,000 g. Then, 1 ml of the supernatant was transferred to a new vial, and 300 μl of water and 300 μl of chloroform were added, vortexed, and centrifuged. Aliquots of 50 μl from the upper phase were dried with a Speedvac (Eppendorf, http://www.eppendorf.com/int/?l=1) and derivatized according to Lisec et al. (2006) with slight modifications. Residues were redissolved and derivatized for 120 min at 37°C in 40 µl of 20 mg/ml methoxyamine hydrochloride (dissolved in pyridine), followed by a 30 min treatment with 70 µl N-methyl-N- (trimethylsilyl)trifluoroacetamide at 37°C. Seven µl of a retention time standard mixture (consisting of 0.029% v/v n-dodecane, n-pentadecane, n-nonadecane, n-docosane, n-octacosane, n-dotracontane, and n-hexatriacontane dissolved in pyridine) was added prior to trimethylsilylation. A sample volume of 1 µl was then injected into the GC column with a splitless/split mode ratio of 32:1 (split/splitless liner with Wool, Restek, USA, http://www.restek.com/). The GC-MS system consisted of an AS 3000 autosampler, a TRACE GC ULTRA gas chromatograph, and a DSQII quadrupole mass spectrometer (Thermo-Fisher ltd, http://www.thermofisher.com/global/en/home.asp). GC was performed on a VF-5 MS column (30 m × 0.25 mm × 0.25 μm) +10 m EZ-Guard (Agilent). A gradient of injection temperature (PTV) was from 60°C to 300°C in 14.5°C × sec-1.The transfer line was set to 300°C, and the ion source adjusted to 250°C. Helium was used as the carrier gas at a constant flow rate of 1 ml × min-1. The oven temperature program was 1 min isothermal at 70°C, followed by a 1°C × min-1 temperature ramp to 76°C, followed by a 6°C × min-1 temperature ramp to 350°C, and a final 5 min at 350°C. Mass spectra were recorded at 8 scans per second with a mass-to-charge ratio 70 to 700 scanning range, with an electron energy of 70 eV. Spectral searching utilized the National Institute of Standards and Technology (NIST, Gaithersburg, USA) algorithm incorporated in the Xcalibur® data system (version 2.0.7) against the retention indices libraries downloadable from the Max-Planck Institute for Plant Physiology in Golm, Germany, (http://gmd.mpimp-golm.mpg.de/) and finally normalized by the internal standard ribitol and tissue mass. Enrichment ratios were calculated as above.

Liquid-chromatography mass-spectrometry analysis of amino acids Samples were prepared according to Hacham et al. (2002) with some modifications.

Fruit tissues (250 mg) were ground in liquid N2, and extracted with 1000 μl of a water:chloroform:methanol mix (3:5:12, v/v), vigorously shaken for 10 min at 4°C, sonicated at RT and centrifuged at 20000 g, and the supernatant was collected. Then, 250 μl chloroform and 375 μl of water were added to each supernatant, and the samples were vortexed and centrifuged at 20000 g for 2 min. Dry aliquots of 200 μl were prepared using speedvac (Eppendorf). Then, the samples were resuspended in 200 μl of 0.02 mM acetic acid. From each sample, 3 μl was injected into the liquid- chromatography (LC) apparatus equipped with a time-of-flight (TOF) mass detector (Agilent Technologies). The analytical separation was performed on a Zorbax Extend

C18 column Rapid Resolution HT (2.1 x 50 mm, 1.8 μm; Agilent Technologies) using an isocratic mobile phase of 95% eluent A (0.01mM acetic acid in 0.2% (v/v) aqueous solution of formic acid) and 5% of eluent B (0.1% (v/v) formic acid in methanol) at a flow rate of 0.2 ml/min in a total run time of 6 min. Eluting compounds were subjected to a dual-sprayer orthogonal ESI source, with one sprayer for analytical flow and one for the reference compound (Agilent Technologies, Santa Clara, USA). The ESI source was operated in positive mode with the following settings: gas temp of 325 ºC with a flow of 8 L/min and nebulizer set to 50 psig. VCap set to 3200 V, the fragmentor to 110 V and the skimmer to 55 V. The scan mode of the mass detector was applied (50– 1050 m/z) with a rate of 3 scans per minute. Amino acids as their [M+H]+ ion were detected using the “find compound by formula” function. The 13C-enriched compounds were detected using the “extract ion chromatogram” function for the calculated [M+H]+ ion.

Preparation of cell-free extracts from melon flesh tissues Fruit flesh was cut into small pieces (about 2 cm3) and frozen at -20°C for no longer than six months until use. The frozen pieces were placed in a chilled mortar and ground with a pestle in the presence of seasand (−50+70 mesh) and 0.5 g polyvinylpolypyrrolidone (PVPP) until a uniform powder was obtained. Ice-cold extraction buffer [50 mM bis-tris propane (pH 7.2 for MGL, pH 8.5 for aminotransferase, 50 mM bis-tris pH 7.0 for thiol-acyl transferase and 50 mM bis-tris propane pH 7.0 for coupled (E)-cinnamaldehyde synthesis), 5% (w/v) D-sorbitol (10% (w/v) glycerol for coupled (E)-cinnamaldehyde synthesis), 10 mM DTT,

5 mM Na2S2O5, 25 µM pyridoxal 5′-phosphate (PLP) (for MGL and aminotransferase assays only), and 0.1% (w/v) polyvinylpyrrolidone (PVP-40)] was added (4:1, v/w), and the suspension was further extracted for an additional 30 s. The slurry was centrifuged at 26,000 g for 20 min at 4°C. Ammonium sulfate was added to the supernatant (crude extract) to a final concentration of 30% saturation, stirred for 20 minutes, and centrifuged at 26,000 g for 20 min at 4°C, and the pellet was discarded. Then, ammonium sulfate was added to the supernatant to a final concentration of 60% saturation, stirred for 20 minutes and then centrifuged at 26,000 g for 20 min at 4°C. The pellet was resuspended in ice-cold buffer A [50 mM bis-tris propane (pH 7.2 for MGL, pH 8.5 for aminotransferase assays, 50 mM bis-tris pH 7.0 for thiol-acyl transferase and 50 mM bis-tris propane pH 7.0 for coupled (E)-cinnamaldehyde synthesis), 5% (w/v) D-sorbitol (10% (v/v) glycerol for the coupled (E)-cinnamaldehyde synthesis), 25 µM pyridoxal 5′-phosphate (PLP)] and desalted on a P-6 DG column (BioRad Labs. Inc., http://www.bio-rad.com/). The protein-containing fractions were merged and used for enzymatic activity assays.

MGL enzymatic assays Enzymatic assays were performed by mixing 150 μl of the desalted ammonium sulfate fractions (about 25 µg protein) with buffer A containing 10 mM L-methionine, to a total volume of 300 μl in a GC-MS 2 ml glass vials. The reactions were incubated for two hours at 30°C and then stopped by an injection of 100 μl of 25% (w/v) NaCl. Then, the vials were analyzed by SPME GC-MS as described above. The methanethiol produced was identified by its retention index and mass spectrum as compared to an authentic standard. At times, peaks corresponding to dimethyl disulfide (DMDS) were also prominent (not shown). Methanethiol was quantified using a standard calibration curve, based on the sum of the areas of both methanethiol and DMDS. For the determination of the α-ketobutyrate (α-KB), the enzyme reactions were performed as above except that a final volume of 1 ml (400 μl for the His-tagged enzyme) and an overnight incubation were used. The reactions were stopped by acidification with 20 μl 10 N HCl (8 μl for the His-tagged enzyme), which was incubated for 30 min at 30°C, extracted with 7 ml ethyl acetate and washed once with

H2O. The organic phase was then evaporated to dryness under a N2 stream, dissolved in 100 μl pyridine and incubated for 90 min. Samples were derivatized with 100 μl N,O-bis(trimethylsilyl)trifluoroacetamide (BSTFA) (Supelco Inc.,

37 http://www.sigmaaldrich.com/analytical-chromatography.html) for 90 min, transferred to a GC-MS vial and injected into the GC-MS (Gonda et al., 2010). The α-KB product as its 2-TMS derivative was identified by comparing the spectral data and retention time with the authentic α-KB standard treated in an identical way as the reaction products.

L-methionine aminotransferase radiolabeled assay Enzymatic assays were performed by mixing 20 μl of desalted concentrated crude extract (about 25 µg protein) with 50 mM bis-tris propane pH 8.5 buffer containing 3 10 mM α-keto acid, 0.1 mM L-methionine {L-[methyl- H3] methionine (Perkin Elmer, http://www.perkinelmer.com/), specific activities: 0.01mCi/mmol}, 1 mM DTT, 225 µM PLP, 10% (w/v) D-sorbitol, in a total volume of 100 μl. The reactions were incubated for two hours (unless otherwise indicated) at 30°C. After incubation, each sample was acidified with 20 μl 1 N HCl, and incubated for 30 min at 30°C. The samples were extracted with 750 μl ethyl acetate, vigorously shaken and centrifuged at 20,000 g to allow phase separation. The upper organic phase (550 μl) was transferred into 5 ml scintillation vials containing 3 ml of Ultima Gold™ scintillation liquid (Perkin Elmer, http://www.perkinelmer.com/). Radioactivity was quantified using a liquid scintillation analyzer (Tri-Carb 2800TR, Perkin Elmer, http://www.perkinelmer.com/). Product amounts were calculated on the basis of the specific activity of the substrate and the counting efficiency of the machine.

L-methionine aminotransferase GC-MS assay Assays were incubated as described for the radiolabeled assays except that only a non-radiolabeled substrate was used, and the volume was increased to 2 ml for fruit- derived cell-free extracts and 0.4 ml for CmMetAT. Samples were then acidified with 40 μl 10 N HCl (8 μl for CmMetAT), and incubated for 30 min at 30°C. The samples were extracted with methyl tert-butyl ether (MTBE), the MTBE evaporated to dryness and the residue dissolved in 100 μl pyridine and incubated for 90 min at 37°C. For tri-methyl silyl derivatization, 100 μl N,O-bis(Trimethylsilyl)trifluoroacetamide (BSTFA) (Supelco Inc., http://www.sigmaaldrich.com/analytical- chromatography.html) was added. The samples were incubated for another 90 min at 37°C, transferred into glass test tubes and stored at 4°C until GC-MS analysis. A 1 μl aliquot of the derivatized extract was injected into a GC-MS system in procedure

38 similar to the one used for the α-KB in the MGL assay. The KMBA product as its 2- TMS derivative was identified by comparing the spectral data and retention time with these of an authentic KMBA standard treated in an identical way as the reaction products.

L-methionine aminotransferase LC-MS assay Enzymatic assays were performed by mixing 70 μl of desalted CmMetAT (about 3 µg protein) with 20 mM sodium phosphate buffer pH 8.5 containing 2 mM KMBA, 2 mM L-amino acid, 200 µM PLP, in a total volume of 100 μl. The reactions were 2 incubated for the time indicated at 30°C. After incubation, [ H3-methyl]-L-methionine was added to each sample as an internal standard, and the samples were inactivated by heating to 100°C for 5 min and immediately chilled on ice. Then, the samples were diluted with 0.2 mM acetic acid in an aqueous solution by a factor of 104. Finally, the samples were injected into the LC-MS as described for amino-acid analysis.

Methanethiol acetyl transferase (ThAT) enzymatic assay Enzymatic assays were performed by mixing 200 μl of either desalted crude extract (about 25 µg protein) or desalted CmThAT1 (about 3 µg protein) with 50 mM bis-tris pH 7.0 buffer containing 0.2 mM acetyl coenzyme A, 1 mM sodium methanethiolate (or propanol in the control assays) and 1 mM DTT in a total volume of 400 μl in a GC- MS 2-ml glass vial. The reactions were incubated overnight at 30°C. Then, the vials were analyzed by SPME GC-MS as described above.

(E)-cinnamaldehyde coupled enzymatic assays Enzymatic assays were performed by mixing 250 μl of desalted crude extract (about 25 µg protein) with 50 mM bis-tris propane pH 7.0 (unless otherwise indicated) buffer containing 1 mM DTT, 7.5 mM (E)-cinnamic acid, 0.1 mM coenzyme A,

0.1 mM NADPH, 5 mM MgSO4 and 5 mM ATP in a total volume of 400 μl in a GC-MS 2-ml glass vial. The reactions were incubated overnight at 30°C. Then, the vials were analyzed by SPME GC-MS as described above.

Cloning of CmMGL, CmMetAT and CmThAT1, and their heterologous expression in E. coli Specific primers (forward CmMGL primer: 5'-AGCATATGGCTGAATTGAAG-3'; reverse CmMGL primer: 5'-TCCTCGAGATTGTTGCAAAA-3'; forward CmMetAT primer: 5'-CGCATATGAAACCAAGCTCTAT-3'; reverse CmMetAT primer: 5'-TTCTCGAGTGACTTCCTAATTA-3'; forward CmThAT1 primer: 5'-CGCATATGATGGCTCCAGCA-3'; reverse CmThAT1 primer: 5'-TTCTCGAGCACAAGCTCCA-3') were designed to insert each gene into a pET21a vector (Novagen, http://www.merckmillipore.com/life-science- research/novagen/c_YTKb.s1OFbwAAAEjSGVXhFCX) with a C-terminal 6×His-tag. A pBK-CMV expression plasmid carrying the respective gene was used as a template. The recombinant plasmid was transformed into BL21 (DE3) pLys Escherichia coli. The cultures were incubated overnight with shaking in 3 ml LB medium containing 5 mM ampicillin at 37°C, and then transferred to 50 ml LB medium and incubated for another six hours at 37°C with shaking. Then, isopropylthio-β- galactoside was added to a final concentration of 0.3 mM, and the bacteria were incubated for another night at room temperature with shaking. The bacteria were centrifuged, and the pellet suspended in 1 ml of lysis buffer containing 50 mM sodium phosphate pH 8.0, 300 mM NaCl, 10 mM imidazole, 25 µM PLP, 100 µg×ml-1 lysozyme (Sigma grade VI from chicken egg, 60,000 units×mg-1 protein), and 200 U of Benzonase® Nuclease (Sigma). The samples were incubated on ice for 20 min, frozen with liquid nitrogen and thawed twice. After the cells had lysed, the suspensions were centrifuged (20,000 g for 20 min at 4°C), and the supernatants were applied to a 1.25 ml Ni-NTA column (Sigma). The tagged-enzymes were eluted with a buffer containing 50 mM sodium phosphate pH 8.0, 300 mM NaCl and 250 mM imidazole. Fractions of 1.5 ml were collected, analyzed by SDS-PAGE and tested for enzymatic activity.

Sodium dodecyl sulphate–polyacrylamide gel electrophoresis (SDS-PAGE) analyses of the recombinant proteins SDS-PAGE was prepared using the following concentrations. For the running gel: 375 mM Tris buffer pH 8.8, 12% (v/v) Acrylamide / Bis-Acrylamide 29:1 (BioLab, http://www.biolab-chemicals.com/page.php?content=homepage), 0.1% (w/v) SDS, 1% (v/v) glycerol, 0.15% (v/v) N,N,N,N'-tetramethylenediamine (TEMED) and 0.1 % (w/v) ammonium persulfate (APS). For the stacking gel: 125 mM Tris buffer pH

6.8, 4 % Acrylamide / Bis-Acrylamide 29:1 (BioLab), 0.1% (w/v) SDS, 0.11% (v/v) N,N,N,N'-tetramethylenediamine (TEMED) and 0.06 % (w/v) ammonium persulfate (APS). The protein samples (15 μl) were mixed in a 1:1 ratio with 2× concentrated sample buffer (Sigma), boiled for 5 min and loaded onto the staking gel. Electrophoresis was performed with a Tris-Glycine buffer containing 1% (w/v) SDS at 70 mA. Then, the gel was stained with Coomassie brilliant blue R-250 and destained by an aqueous solution containing 20% (v/v) methanol and 10% (v/v) glacial acetic acid. The molecular weight of the recombinant proteins were determined using a calibration curve based on biomarkers

(BioRad) with the following equation: Mr = Log10 (Rf/Rmax).

Co-expression of CmThAT1 and CmMGL and their heterologous expression in E. coli Specific primers (forward CmThAT1 primer: 5'-ATGGCTCCAGCAGCAGCA-3'; reverse CmThAT1 primer: 5'-TCACACAAGCTCCAACACA-3') were designed to insert CmThAT1 into a pET21a vector (Novagen) without any tag at its terminal domains. cDNA from ripe fruits were used as a template (see below). The recombinant plasmid was transformed into BL21 (DE3) Escherichia coli using ampicillin selection. New electrocompetent cells were prepared from bacteria carrying the plasmid with the CmThAT1 insert. The competent cells were transformed again with CmMGL in pBK- CMV (extracted from the cDNA library clone no. DV633932) using kanamycin selection. Bacteria harboring the two clones were incubated overnight with shaking in 3 ml LB medium containing 30 μg / ml ampicillin and 50 μg / ml kanamycin at 37°C, and then 500 μl was transferred to 2.5 ml fresh LB medium in 20 ml SPME vials and incubated for another one hour at 37°C with shaking. Then, isopropylthio-β-galactoside was added to a final concentration of 0.5 mM, and bacteria were incubated for another night at room temperature with shaking (225 RPM). Finally, the headspace of the bacteria was analyzed by SPME-GC- MS as described above.

Cloning of CmCNL and heterologous expression in E. coli Specific primers (forward primer: 5'- AACATATGGAAGATCTGAAGCCA -3'; reverse primer: 5'- AACTCGAGCAATCTACTTTTTGGT -3') were designed to insert CmMGL into a pET21a vector (Novagen, http://www.merckmillipore.com/life- science-research/novagen/c_YTKb.s1OFbwAAAEjSGVXhFCX) with a C-terminal

6×His-tag. cDNA derived from mRNA extracted from ripe melon fruit ('Dulce' cultivar) was used as a template. The recombinant plasmid was transformed into BL21 (DE3) pLys Escherichia coli. The cultures were incubated overnight with shaking in 3 ml LB medium containing 5 mM ampicillin at 37°C, and then transferred to 50 ml LB medium and incubated for another 4 hours at 37°C with shaking. Then, isopropylthio- β-galactoside was added to a final concentration of 0.5 mM, and the bacteria were incubated for another night at room temperature with shaking. Bacteria were centrifuged and the pellet suspended in 1 ml of lysis buffer containing 50 mM sodium phosphate pH 8.0, 300 mM NaCl, 10 mM imidazole, 25 µM PLP, 100 µg×ml-1 lysozyme (Sigma grade VI from chicken egg, 60,000 units×mg-1 protein), and 200 U of Benzonase® Nuclease (Sigma). The samples were incubated on ice for 20 min, frozen with liquid nitrogen and thawed twice. After the cells have lysed, the suspensions were centrifuged (20,000 g for 20 min at 4 °C) and the supernatants were applied to a 0.4 ml Ni-NTA column (Qiagen, http://www.qiagen.com/). The tagged-enzymes were eluted with 50 mM sodium phosphate buffer pH 8.0, 300 mM NaCl and 250 mM imidazole. Fractions of 1.5 ml were collected and the enzyme potassium pH 8.0.

(E)-cinnamic acid:coenzyme A ligase enzymatic assay Enzymatic assays were performed by mixing 1000 μl of desalted recombinant enzyme (about 3 µg protein) with 20 mM potassium phosphate buffer pH 8.0 and 7.5 mM (E)- cinnamic acid, 0.1 mM coenzyme A, 5 mM MgSO4 and 5 mM ATP in a total volume of 200 μl. The reactions were incubated overnight at 30°C. From each sample 1 μl was injected into the liquid-chromatography (LC) apparatus equipped with a time-of-flight (TOF) mass detector (Agilent) and analyzed according to Klempien et al (2012) with slight modifications. The analytical separation was performed on a Zorbax Extend C18 column Rapid Resolution HT (2.1x50 mm, 1.8 micron; Agilent Technologies) using a linear gradient of from 15% eluent A (0.1 % (v/v) mM ammonium acetate aqueous solution) and 85% (v/v) of eluent B (0.1 % (v/v) ammonium acetate in methanol) at a flow rate of 0.3 ml/min in a total run time of 13 min. Eluting compounds were subjected to a dual-sprayer orthogonal ESI source, with one sprayer for analytical flow and one for the reference compound (Agilent Technologies, Santa Clara, USA). The ESI source was operated in negative mode with the following settings: gas temp of 300 ºC with a flow of 8 L/min and nebulizer set to 30 psig. VCap set to 3000 V, the fragmentor to 140 V and the skimmer to 65 V. Scan mode of the mass detector was applied (100–

1700 m/z) with a rate of 2 scans per minute. Acyl coenzyme A derivatives (as their [M−H]- ion) where detected using the 'find compound by formula' function. cDNA preparation and quantitative real-time PCR analysis cDNA synthesis for a quantitative real-time PCR (qRT-PCR) analysis was performed according to Gonda et al. (2010) with slight modifications. A qRT-PCR was performed on an ABI prism7000 Sequence Detection System (Applied Biosystems, Foster, CA) using ABsoluteTM QPCR SYBR® Green Mixes (ABgene®’s Inc., Epsom, UK). All samples were run in triplicates for each primer combination. A 1-µl aliquot of cDNA was used for each RT-qPCR reaction. Thermal cycling was initiated by 15 min incubation at 95°C, followed by 40 cycles of 90°C for 15 s and 60°C for 1 min. A melting curve analysis was performed for each reaction to confirm the specificity of the amplification. The relative quantification of gene expression was performed using the melon housekeeping gene cyclophilin (MU47687 or DV632242) as a reference according to Portnoy et al. (2008, 2011). Three biological samples of each developmental stage were tested. Ct values were determined by the ABI Prism 7000 SDS software and exported into MS Excel workbook (Microsoft Inc., Redmond, WA) for statistical analysis. Real-time efficiencies (E) were calculated from the slopes of standard curves for each gene (E =10[-1/slope]) (Ramakers et al., 2003). The relative expression ratio (R) was calculated according to Pfaffl (2001) and compared to the sample of the respective young fruit (10 DAA). Primers for CmMGL were designed using the Gene Runner 3.0 software (Hastings Software, Inc.). The sequences that were used for the primer design are all available in the melon database (http://www.icugi.org). The following primers (0.2 µM final concentration) were used: CmMGL: forward primer 5'-CGACGTAAAAACCGATTCA-3' and reverse primer 5'- GACGCATGGTTTCAGGTTC-3'. Amplicon length was 169 bp. Cyclophilin (DV632242): forward primer 5'-GATGGAGCTCTACGCCGATGTC-3' and reverse primer 5'-CCTCCCTGGCACATGAAATTAG-3'. Amplicon length was 153 bp.

RNA-seq gene expression analysis Total RNAs were extracted from the flesh tissue of five pooled mature fruits of each of the RI lines. Strand-specific RNA-seq libraries were constructed from the total RNAs as previously described (Zhong et al., 2011) and sequenced on the Illumina HiSeq platform. RNA-Seq reads were first aligned to ribosomal RNA sequences using Bowtie

(Langmead et al., 2009) to remove possible rRNA contaminations. The resulting filtered reads were aligned to the melon genome (Garcia-Mas et al., 2012) using TopHat (Trapnell et al., 2009). Following alignments, raw counts for the CmMGL gene were obtained and normalized to reads per kilobase of exon model per million mapped reads (RPKM).

Bioinformatic and statistical analyses Sequence homologies were based on the EBI align tool (http://www.ebi.ac.uk/Tools/emboss/align/), as performed by the Needle program in the EBLOSUM62 matrix. Multiple sequence alignment was done with the EBI MUSCLE tool (http://www.ebi.ac.uk/Tools/muscle/), and shaded with Boxshade version 3.21 (http://www.ch.embnet.org/software/BOX_form.html). Blast analyses were prefomed using the NCBI Blastp algorithm (http://blast.ncbi.nlm.nih.gov/Blast.cgi), or the ICUGI tblastn (http://www.icugi.org/cgi-bin/ICuGI/tool/blast.cgi), or the melon genome tblastn (https://melonomics.net/tools/blast/run/). Family and subfamily divisions were determined either by the “HMM sequence scoring” tool of the Panther classification system (http://www.pantherdb.org/) or by Pfam (http://pfam.sanger.ac.uk/). N- terminus targeting predictions were done using either Predotar (http://urgi.versailles.inra.fr/predotar/predotar.html), Mitoprot (http://ihg.gsf.de/ihg/mitoprot.html), TergetP 1.1 (http://www.cbs.dtu.dk/services/TargetP/), iPSORT (http://ipsort.hgc.jp/) or WoLF PSORT (http://wolfpsort.org/). All statistical analyses were carried out using JMP® software according to the developer’s instructions.

Results L-methionine-derived aroma volatiles in melon fruits

Melon cultivars contained different levels of sulfur aroma volatiles Melon genotypes display a vast polymorphism in their aromas, also reflected in the content of sulfur and other aroma volatiles. I have characterized the sulfur-containing aroma volatiles (S-compounds) of two distant accessions differing in their organoleptic properties: the commercial ‘Dulce’ cultivar that has a very pleasant aroma, and the non- commercial PI 414723 accession that has an unpleasant sulfurous aroma (Table 3). Table 3. Ratios between the levels of sulfur-containing aroma volatiles in ripe PI 414723 fruit flesh and their levels in ripe ‘Dulce’ fruit flesh. Volatiles were detected by SPME-GC-MS (n=5). Compound PI 414723 / Methods of Compound name type ‘Dulce’ identification ratio 2-(methylthio) ethanol thio-ether alcohol 10.89 MS, AS methanethiol thiol 2.58 MS,AS dimethyl disulfide sulfide 2.82 MS,KI dimethyl trisulfide sulfide 2.54 MS,KI S-methylthio acetate thioester 0.86 MS,KI S-methyl propanethioate thioester 7.22 MS S-methyl 2-methylpropanethioate thioester 4.31 MS S-methyl thiobutanoate thioester 2.78 MS, AS S-methyl 2-methylbutanethioate thioester 3.09 MS methyl (methylthio) acetate thioether ester 17.61 MS,KI ethyl (methylthio) acetate thioether ester 6.65 MS,KI,AS 2-(methylthio) ethyl acetate thioether ester 2.42 MS,AS methyl 3-(methylthio) propanoate thioether ester 7.94 MS,KI ethyl 3-(methylthio) propanoate thioether ester 14.90 MS,KI,AS 3-(methylthio) propyl acetate thioether ester 0.25 MS,KI, AS All S-compounds - 2.53 MS – mass spectrum, KI – Kovats index, AS – comparison with an authentic standard.

Fifteen different S-compounds were present in both cultivars, including five thioesters (S-methyl thioacetate, S-methyl propanethioate, S-methyl 2-methylpropanethioate, S-methyl thiobutanoate and S-methyl 2-methylbutanethioate), six thio-ether esters (methyl (methylthio) acetate, ethyl (methylthio) acetate, 2-(methylthio) ethyl acetate,

45 methyl 3-(methylthio) propanoate, ethyl 3-(methylthio) propanoate and 3-(methylthio) propyl acetate), two sulfides (dimethyl disulfide and dimethyl trisulfide), one alcohol (2-(methylthio) ethanol) and one thiol (methanethiol) (Table 3). Trace amounts of two other S-compounds, namely 3-(methylthio) propanal and 3-(methylthio) propanol, were also detected. PI 414723 accumulated about twofold higher amounts of total sulfur- containing volatiles as compared to ‘Dulce’ (Table 3). Most of the S-compounds were present in PI 414723 at higher levels than in ‘Dulce’ (two to tenfold). Still, some compounds seemed to be of comparable levels in both genotypes, while 3-(methylthio) propyl acetate was present in higher levels in ‘Dulce’ (Table 3).

Feeding melon cubes with stable-isotope-labeled L-methionine resulted in enriched volatile aroma compounds I have previously shown that melon flesh cubes accumulate high levels of sulfur- containing aroma volatiles upon incubation with exogenous L-methionine (Gonda et al., 2010). In order to explore possible biosynthetic pathways for these conversions, I 13 incubated melon flesh cubes with C5-L-methionine. I observed four different major enrichment patterns in the mass spectra of the aroma volatiles detected, as documented by additional ions in the spectra of the co-eluting enriched and non-enriched compounds (Table 4). 1. Compounds enriched with +4 m/z units, such as ethyl 3- (methylthio) propanoate and 3-(methylthio) propyl acetate (Figure 7 red bars, Figure 8a). 2. Compounds enriched with only +1 m/z unit, such as S-methyl 2-methylbutanethioate and ethyl-2-(methylthio) acetate (Figure 7 green bars, Figure 8be). In some cases, this enrichment was associated with a +2 m/z pattern, e.g., dimethyl disulfide and dimethyl trisulfide (Figure 7 green bars, Figure 8c). 3. Compounds enriched with +3 m/z units and characterized by the L-isoleucine side- chain, such as 2-methyl butanol and propyl-2-methyl butanoate (Figure 8de, Figure 9 orange bars). 4. Esters enriched with +3 m/z units and characterized by a propanoate substitution, such as methyl and ethyl propanoate (Figure 8f, Figure 9 purple bars). No enrichment was detected in propyl esters, such as propyl acetate and propyl butanoate. Some volatiles, such as S-methyl 2-methylbutanethioate and S-methyl propanethioate, showed enrichment in the methylthio moiety (group 2) and the carbon moiety (groups 3 and 4, respectively) (Figure 8e). The compound 2-methyl butyl propanoate was enriched in both moieties (groups 3 and 4) (Figure 8g).

13 Table 4. Enrichment levels after incubation of melon cubes with C6-L-metthionine. Year 2011 2012 Cultivar PI 414723 ‘Dulce’ PI 414723 ‘Dulce’ Incubation time 6 h ON 6 h ON 6 h 6 h Compound name +m/z* Enrichment level ± SEM [%] 3-(methylthio) propanal 45.4±3.2 36.5±7.8 30.8±4.2 11.8±0.7 42±3.1 23±0.2 et. 3-(methylyhio) propanoate 44.7±6.6 35.5±8.5 32.9±2.4 29.0±4.3 52.7±1.4 25.7±4.4 me. 3-(methylthio) propanoate +4 36.0±5.6 40.9±6.8 28.2±2.7 27.9±5.0 36.2±0.9 16.4±4.9 3-(methylthio) propyl acetate 34.7±7.7 45.9±6.6 22.9±1.8 13.5±4.7 33.9±2.7 14.3±9.8 3-(methylthio) propanol 32.6±7.8 48.5±7.0 28.5±1.1 13.5±4.7 14.7±6.0 n.d.±n.d. S-me. propanethioate 50.8±3.6 33.7±9.7 n.d. n.d. 58.7±5.3 n.d.±n.d. [methylthio moiety] S-me. 2-me. butanethioate 49.8±4.0 39.6±7.3 34±2.1 n.d. 57.3±3.7 25.8±2.3 [methylthio moiety] S-me.butanethioate 46.8±3.9 47.6 n.d. n.d. 53.4±4.0 n.d. S-methyl 2- 42.6±3.1 26.6±7.0 26.6±3.1 n.d. 50.7±4.6 n.d. methylpropanethioate S-me. Thioacetate 42.4±4.1 35.8±7.6 32.3±3.7 33.2±2.6 52.1±3.7 n.d. dimethyl trisulfide +1 36±2.2 25.6±4.9 28.1±5.2 22.9±4.3 22.6±4.5 5.6±1.6 +1 dimethyl disulfide +1 35.3±2.9 27.0±4.0 27.9±5.4 23.2±2 25.1±5.0 6.2±2.5 dimethyl disulfide +2 17.3±5.2 5.1±2.5 4.3±1.4 1.6±0.7 5.3±2.7 0.3±0.2 dimethyl trisulfide +2 10.6±4.9 2.1±1.0 2.9±0.8 n.l. 1.3±0.7 0.6±0.3 methanethiol 15.0±2.9 7.6±2.2 7.2±2.9 0.6±0.6 8.3±2.0 5.5±1.8 et. (methylthio) acetate 6.8±2.7 29.6±7.1 1.7±0.6 8.3±3.6 16.1±3.3 1.8±0.5 me. 2-methylthio acetate 2.7±0.6 21.9±5.2 0.6±0.2 4.0±2.2 6.5±1.7 0.5±0.5 2-(methylthio) et. Acetate 0.9±0.2 9.6±3.0 1.2±0.3 11.2±6.3 1.0 1.2±0.3 methyl benzoate 19.3±3.5 16.8±13.1 1.3±0.6 7.0±7.0 17.6±5.1 n.l. 2-me. butanal n.l n.l n.l n.l 14.8±1.9 0.6±0.5 2-me. butanol 12.3±1.8 17.6±1.6 n.l n.l 11.6±4.1 n.l 2-me.-2-butenal 14.8±1.8 17.9±2.3 n.l n.d. 11.8±1.6 3.1±0.6 me.-2-me.butanoate 7.±0.9 9.6±1.9 0.2±0.1 n.l 7.8±0.8 n.l et.-2-me.butanoate 17.6±2.8 14.9±2.3 n.l n.l 13.0±1.4 0.2±0.1 propyl-2-me.butanoate 13.5±2.0 16.1±2.2 0.4±0.2 n.l 14.5±2 n.l +3 butyl-2-methybutanoate 17.2±3.6 15.4±3.1 n.d. n.d. 16.3±2.7 n.d. (ile) 2-me.butyl-2-me. butanoate 13.9±1.9 17.4±4.0 n.d. n.d. 14.2±2.4 n.d. 2-me. propyl 2-me. butanoate 14.2±1.7 20.9±1.4 n.d. n.d. 13.6±1.7 n.d. 2-me. butyl propanoate [ile 10.4±1.7 14.7±2.4 n.l n.l 9±1.7 n.d. moiety] S-me. 2-me.butanethioate [ile 14.9±1.8 17.4±3 1.5±0.2 n.d. 16.8±2.1 4.4±1.8 moiety] 2-me.butyl acetate 12.8±2.3 20.6±3.2 0.9±0.3 1.9±0.2 6.5±0.4 n.l propyl propanoate 17.9±3.6 21.4±2.8 10.3±2.4 n.d. 31.4±5.2 n.d. S-me. propanethioate [prop. 17.6±2.8 14.4±6.7 n.d. n.d. 22.8±5.7 n.d. moiety] et. propanoate +3 17.1±3.0 20.5±2.9 7.5±2.1 6.5±0.2 32.3±3.3 2.4±0.8 2-me. propyl propanoate (prop.) 16.6±2.7 16.2±2.9 9.0±5.7 2.1±1.7 29.7±3.9 n.d. me. propanoate 14.5±2.5 22.1±3 n.l n.l 22.4±4.3 n.d. 2-me.butyl propanoate [prop. 18.4±3.2 16.7±3.2 n.l n.l 33.4±3.5 n.d. moiety] Levels of 360 min incubations are means of six (three for overnight incubations) biological repeats from four (three for overnight incubations) independent experiments ± SE. Values are: 13C / (13C + 12C); * - m/z enrichment pattern, n.d. – compound not detected, n.l. – compound not labeled, et. – ethyl, me. – methyl, prop. – propanoate, ile – L-isoleucine. Green numerals: n=5; red numerals: n=2; blue numerals: n=1.

13 Figure 7. Incorporation levels of C5-L-methionine into sulfur aroma volatiles and methyl benzoate in ripe melon fruit flesh. (a) PI 414723. (b) ‘Dulce.’ Red bars represent compounds enriched with +4 m/z units. Green bars represent compounds enriched with either +1 or +2 m/z unit. Ripe melon flesh cubes were

13 incubated with either 5 mM C5-L-methionine or with 5 mM non-labeled L-methionine as a control for six hours at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS. The incorporation level for methanethiol is only an approximation since its dense spectra makes it difficult to accurately evaluate. Values are the percentage of the enriched compound out of the total levels (enriched and non-enriched) after the subtraction of the enriched levels in the controls (incubation with non-labeled L-methionine), and they represent the means of three (• - n=2) biological repeats ± SEM. nd: not detected.

13 Figure 8. Mass-spectra of volatiles isolated from melon cubes fed with 5 mM C5-L- methionine.

13 Figure 8. Mass-spectra of volatiles isolated from melon cubes fed with 5 mM C5-L- methionine. 13 The upper spectrum in each panel represents C5-L-methionine enrichment; the lower spectrum in each panel depicts a control sample incubated with non-labeled L-methionine. (a) 3-(methylthio) propyl acetate, (b) ethyl (methylthio) acetate, (c) dimethyl disulfide, (d) propyl-2-methyl butanoate, (e) S-methyl 2-methyl butanethioate, (f) ethyl propanoate, (g) 2-methylbutyl propanoate. (h) Mass-spectra of 2-methylbutyl propanoate isolated from

13 melon cubes fed with either 5 mM C6-L-isoleucine (upper panel) or with non-labeled 5 mM L-isoleucine (lower panel). Each spectrum shown is typical to at least five biological

13 repeats. Ripe melon flesh cubes were incubated with either 5 mM C5-labeled amino acid or with 5 mM non-labeled amino acid as controls for six hours at room temperature and the volatiles were subsequently analyzed by SPME-GC-MS.

13 13 Figure 9. Incorporation levels of C5-L-methionine and C6-L-isoleucine into aroma volatiles in ripe melon fruit flesh. Panels (a) and (b) represent 13C-enrichment levels of aroma volatiles after incubation of melon

13 13 cubes with C5-L-methionine. Panels (c) and (d) represent C-enrichment levels of aroma

13 volatiles after incubation of melon cubes with C6-L-isoleucine. (a) and (c): PI 414723; (b) and (d): ‘Dulce.’ Orange bars represent compounds bearing an isoleucine side-chain enriched with either +3 m/z units, in the case of L-methionine, or 5 m/z units, in the case of L-isoleucine incubations. Purple bars represent +3 m/z enriched propanoate derivatives.

13 (legend Figure 9 – continue) Ripe melon flesh cubes were incubated with either 5 mM C5- labeled amino acid or with 5 mM non-labeled amino acid as controls for six hours at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS. Values are the percentage of the enriched compound out of the total levels (enriched and non-enriched) after the subtraction of enriched levels in the controls (incubation with non-labeled amino acids), and represent means of three (• - n=2) biological repeats ± SEM. nd: not detected.

In general, the level of isotopic enrichment was higher in PI 414723 than in ‘Dulce’, and the +3 m/z enrichment pattern (both groups 3 and 4) was mainly detected in PI 414723 (Figure 9ab). No enrichment was detected for 2-(methylthio) ethanol. The

C2-thio-ether esters gained significant enrichment only after overnight incubation 13 (Table 4). Examination of the non-volatile fraction of the C5-L-methionine incubated melon cubes by GC-MS clearly indicated the presence of enriched L-isoleucine in PI 414723 (Table 5).

Table 5. Levels of 13C-enrichment in free L-isoleucine upon incubation of melon flesh cubes with 5 13 mM C5-L-methionine as detected by GC-MS (as TMS derivatives). Genotype PI 414723 Dulce

% of 13C-enrichment in free L-isoleucine 4.97±0.65 0.84±0.41

Values are means of four biological repeats ± SEM.

Moreover, examination of the enrichment in a liquid-chromatography, time-of-flight mass-spectrometer (LC-TOF-MS) revealed that the enrichment pattern of L-isoleucine in these incubations was of +4 m/z units, when the monoisotopic mass, + 12 13 ([M+H] ) = 136.1158 m/z, depicts a molecular formula of C2 C4H13NO2. To further explore if the +3 m/z enrichment pattern in branched-chain-derived volatiles and propanoate esters could be due to L-isoleucine catabolism, incubation experiments 13 with C6-L-isoleucine were conducted. As expected, I found that volatiles bearing the L-isoleucine side-chain displayed additional masses enriched with +5 m/z units (Figure 8h, blue ions). Propanoate derivatives, such as ethyl and propyl propanoate, were 13 enriched with +3 m/z units (Figure 8h, purple ions) when administering C6-L- 13 isoleucine in patterns identical to the enrichment observed when C5-L-methionine was administrated (Figure 8g, purple ions). This experiment also helped to distinguish between the L-isoleucine side-chain moiety (that was enriched with +5 mass units at

52 ion 70) and the propanoate moiety (that was enriched with +3 mass units at ion 57) in the volatile 2-methylbutyl propanoate (Figure 8h). No enrichment was detected in acetate esters (derived from acetyl CoA) either when 13C-L-isoleucine or 13C-L-methionine was administered. Moreover, propanoate esters 13 were not enriched when melon cubes were incubated with C6-L-leucine. Interestingly, 13 13 in contrast to incubation with C6-L-isoleucine, when incubated with C6-L-leucine, melon fruit cubes produced acetate esters enriched with +2 m/z units (Figures 10 and 11).

13 13 Figure 10. Incorporation of C6-L-leucine and C6-L-isoleucine into volatile benzyl acetate in ripe melon fruit cubes. Overlaid chromatograms of ion (m/z) 150 (black lines) depict non-enriched benzyl acetate, and

13 ion (m/z) 152 (red lines) depict C2-enriched benzyl acetate. Upper left panel – incubation with

13 non-enriched L-leucine, lower left panel – incubation with C6-L-leucine, upper right panel –

13 incubation with non-enriched L-isoleucine, lower right panel – incubation with C6-L-

13 isoleucine. Ripe melon flesh cubes were incubated overnight with either 5 mM C5-labeled amino acid or with 5 mM non-labeled amino acid as controls at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS. Each chromatogram represents at least six biological repeats from at least two independent experiments in two different years.

13 13 Figure 11. Differential incorporation of C6-L-leucine and C6-L-isoleucine into the acetate fraction of various acetate esters in ripe melon fruit cubes. 13 Overlaid chromatograms of ion (m/z) 45 depict C2-enriched acetate fraction of various acetate

13 esters. The lower panel is a continuation of the upper panel. Red line – incubation with C6-L-

13 leucine, black line – incubation with non-enriched L-leucine, green line – incubation with C6- L-isoleucine, and blue line – incubation with non-enriched L-isoleucine. Numbers depict the following compounds: 1: ethyl acetate, 2: propyl acetate, 3: 2-methylpropyl acetate, 4: butyl acetate, 5: 3-methylbutyl acetate, 6: 2-methylbutyl acetate, 7: pentyl acetate, 8: 2-(methylthio) ethyl acetate, 9: hexyl acetate, 10: 2,3-butanedioldiacetate, 11: heptyl acetate, 12: 3- (methylthio) propyl acetate, 13: benzyl acetate, 14: 2,4-diacetoxypentane, 15: (Z)-3-octen-1-ol, acetate, 16: octyl acetate, and 17: 2-phenethyl acetate. Ripe melon flesh cubes were incubated

13 overnight with either 5 mM C5-labeled amino acid or with 5 mM non-labeled amino acid as controls at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS. Each chromatogram represents at least four biological repeats from at least two independent experiments in two different years.

To further ascertain if the +1 m/z enrichment in group 2 originated from the carbon of 2 the methylthio moiety, feeding experiments were conducted with H3-[methyl]-L- methionine. The substantial enrichment of +3 m/z units was documented as two separate peaks consisting of the enriched and non-enriched compounds (in that order) of the volatile compounds and further corroborated by their mass spectra (Figure 12).

2 Figure 12. Incorporation of H3- [methyl]-L-methionine into aroma volatiles in ripe melon fruit cubes. (a) Single ion chromatograms (overlaid) of ion (m/z) 94 (typical to non-enriched dimethyl disulfide, DMDS), ion 97 (typical

2 to DMDS enriched with one H3-[methyl] group) and ion 100 (typical to DMDS

2 enriched with two H3-methyl groups). (b) Total ion chromatograms of volatiles from melon cubes incubated either with

2 H3-[methyl]-L-methionine (top) or with non-labeled L-methionine (bottom). The peak in the lower chromatogram and the right peak in the top chromatogram depict non-enriched S-methyl-2-methyl butanethioate. The left peak in the top

2 chromatogram depicts H3-S-methyl-2- methylbutanethioate. (c) Mass spectrum of

2 H3-enriched (top) and non-enriched (bottom) S-methyl-2-methyl

2 butanethioate). Ions in grey depict H3- enriched fragments. Ripe melon flesh cubes were incubated overnight with either 5 mM enriched or non-enriched amino acid at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS.

The double enrichment of +6 m/z units (in addition to the +3 enrichment) was detected in dimethyl disulfide (Figure 12, Figure 13) and dimethyl trisulfide (Figure 13). No enrichment was apparent in compounds from groups 3 or 4 of the 13C experiment. Substantial levels of enrichment were detected in methyl benzoate in PI 414723, both in 13C (Figure 7) and 2H experiments (not shown). Interestingly, other abundant methyl esters, such as methyl butanoate and methyl hexanoate, were not enriched regardless of the isotope administered.

2 Figure 13. Incorporation levels of H3-[methyl]-L-methionine into aroma volatiles in ripe melon fruit flesh. Bars represent enrichment values obtained in ‘Dulce’ (red) and ‘PI 414725’ (blue). Ripe melon

2 flesh cubes were incubated with 5 mM H3-[methyl]-L-methionine for six hours, and the volatiles were analyzed by GC-MS. Levels are averages of three biological repeats ± SEM.

In the analysis of S-compounds in 94 different recombinant inbred lines, generated from a cross between PI 414723 and ‘Dulce’ genotypes (Harel-Beja et al., 2010), I found two additional S-compounds: 3-(methylthio) propanal and 3-(methylthio) propanol (also detected in the incubation experiments, see Figure 7). These compounds probably were present in quantities below the detection levels in the parent accessions. To analyze the distribution of S-compounds in the RIL population, a two-way hierarchical clustering analysis was carried out (Figure 14). In a broad-spectrum view, two major clusters were detected: the first one contains methanethiol, sulfides and thioesters, and the second one contains thio-ether compounds that include aldehyde, alcohols and C2- and C3- thio-ether esters. The first cluster comprises of all the compounds that were significantly enriched with +1 m/z after six hours (Table 4), when all five thioesters are clustered together as well as the two sulfides.

Figure 14. Two-way hierarchical clustering of sulfur-containing aroma volatiles in the fruit flesh throughout the PI 414723 × ‘Dulce’ RIL population. For each RI line, three fruits of different plants were analyzed by SPME-GC-MS. A two-way hierarchical clustering analysis was carried out using the JMP program with the averages of each line (n=3).

In the second cluster that comprises the thio-ethers, the alcohols were clustered together, and the thio-ether esters were clustered according to their alcohol moiety (i.e., methyl, ethyl or acetate esters) and not according to the length of the thio-ether carbon chain (Figure 14).

L-Isoleucine-derived aroma volatiles in melon cultivars I further monitored the levels of the L-isoleucine-derived volatiles in the two genotypes. I identified five propanoates, four 2-methylbutyl derivatives (one of them is also a propanoate), and six 2-methylbutanoates (Table 6). This excludes S-methyl 2-methylbutanethioate and S-methyl propanethioate.

Table 6. Ratios between the levels of propanoates, 2-methylbutanoates and 2-methylbutyl derivatives aroma volatiles in ripe PI 414723 fruit flesh and their levels in ripe ‘Dulce’ fruit flesh. Volatiles were detected by SPME-GC-MS (n=5).

PI / Dul Methods of Compound name Compound type ratio identification methyl propanoate propanoate 3.08 MS,KI ethyl propanoate propanoate 1.83 MS,KI propyl propanoate propanoate 0.50 MS,KI

2-methylpropyl propanoate propanoate 0.39 MS,KI

2-methylbutyl propanoate 2-methylbutyl 0.59 MS,KI

2-methyl butanol derivate/propanoate2-methylbutyl derivate 0.40 MS,KI,AS 2-methyl-2-butenal 2-methylbutyl derivate 1.23 MS methyl 2-methylbutanoate 2-methylbutanoate derivative 2.29 MS,KI,AS ethyl 2-methylbutanoate 2-methylbutanoate derivative 1.16 MS,KI

2-methylbutyl acetate 2-methylbutyl derivate 0.16 MS,KI propyl 2-methylbutanoate 2-methylbutanoate derivative 2.12 MS,KI

2-methylpropyl 2-methylbutanoate 2-methylbutanoate derivative 2.43 MS,KI butyl 2-methylbutanoate 2-methylbutanoate derivative 5.96 MS,KI

2-methylbutyl 2-methylbutanoate 2-methylbutanoate derivative 19.27 MS,AS total isoleucine derived compounds% - 0.71 - total propanoates# - 1.10 - total isoleucine derived + - 0.75 - propanoatesPI – PI 414723,$ Dul – ‘Dulce,’ MS – mass spectrum, KI – Kovats index, AS – comparison with an authentic standard. % – including S-methyl 2-methylbutanethioate, # – including S-methyl propanethioate. $– including S-methyl 2-methylbutanethioate and S-methyl propanethioate.

Unlike the S-compounds generally more abundant in PI 414723 than in ‘Dulce’, the L- isoleucine-derived 2-methylbutyl acetate and 2-methylpropyl propanoate were measured at higher levels in ‘Dulce’ as compared to PI 414723, whereas methyl propanoate and methyl-2-methylbutanoate were present in higher levels in PI 414723 than in ‘Dulce’.

The role of L-methionine-γ-lyase in melon aroma production

Soluble protein extracts from mature melon fruits displayed L-methionine-γ- lyase enzymatic activity The different enrichment patterns observed in the experiments suggested that more than one L-methionine catabolic pathway is active in ripe melon fruit flesh. The incorporation of L-methionine into volatiles from group 1 (+4 m/z units enrichment) is probably mediated via L-methionine aminotransferase activity (Gonda et al., 2010 and see also below). The incorporation of L-methionine into other volatiles from groups 2, 3 and 4 (+1 or +3 m/z units enrichment) suggested that L-methionine is additionally broken down by an L-methionine-γ-lyase (MGL) enzyme into methanethiol, α- ketobutyrate and ammonia (Figure 16a), and that these products are further metabolized into aroma compounds through several metabolic steps. Hence, I tested melon crude cell-free extracts for their ability to support MGL enzymatic activity in the presence of L-methionine as a substrate and pyridoxal 5’-phosphate as a cofactor. I was able to detect MGL enzymatic activity in melon crude extracts under such conditions (Figure 16). The enzyme activity produced detectable levels of methanethiol and α-ketobutyrate. Methanethiol was monitored by GC-MS using SPME (Figure 16bc). Dimethyl disulfide was sometimes observed as an additional product of the reaction, probably due to the chemical conjugation of two methanethiol molecules. MGL activity was three times higher in genotype PI 414723 as compared to ‘Dulce’ (Figure 16b). The α-ketobutyrate product was monitored by GC-MS after acid extraction with ethyl acetate and derivatization with N,O-bis(trimethylsilyl) trifluoroacetamide (BSTFA) and trimethylchlorosilane (TMCS) (Figure 16d). MGL activity was detected in both PI 414723 and ‘Dulce’ cultivars in ripe fruit flesh (Figure 16b-d).

CmMGL encodes a methionine-γ-lyase enzyme In order to identify a gene encoding for MGL enzyme, I searched for melon sequences homologous to the functionally characterized AtMGL (Rébeillé et al. 2006) in the cucurbits genomic database (http://www.icugi.org/) using tBlastn algorithm. This database has proven to be a good source for identifying many genes affecting melon quality characteristics, including aroma qualities (Ibdah et al., 2006; Portnoy et al., 2008; Gonda et al., 2010). The best hit was the unigene MU49895 (GeneBank accession no. JX673982, termed CmMGL) composed of nine clones, six of them from fruits. It

59 shares 75% AA identities with AtMGL, while the second best hit (MU44961) shares only 29% AA identities with AtMGL. A partial sequence of CmMGL is tentatively annotated as a cystathionine gamma synthase in the NCBI database (González et al., 2010). CmMGL is a full-length clone coding for a 449 AA protein with an estimated molecular weight of 48.7 KDa, as predicted by Expasy. Both ‘Dulce’ and PI 414723 genotypes displayed identical CmMGL nucleotide sequences. CmMGL was aligned with other functionally characterized MGL enzymes from plant and bacterial origins (Figure 17). Bioinformatics analyses of CmMGL reveals no subcellular localization to any known organelle by all tools used (predotar, MitoProt II, iPSORT, TargetP, and WoLF PSORT). The expression of CmMGL increased in the ripe fruit flesh of the genotype PI 414723 (Figure 15). However, CmMGL expression levels in ‘Dulce’ fruit flesh remained relatively constant during fruit development with a slight decrease in the ripe stage (Figure 15). Figure 15. CmMGL gene expression in developing melon fruit flesh as analyzed by qRT-PCR. RNA was extracted from the fruit flesh at various stages of fruit development. Expression levels were normalized against the internal control gene cyclophilin. Values represent the means of at least two biological triplicates ± SD.

Next, with the melon genome that was recently published (Garcia-Mas et al., 2012) further analyses were conducted using the tblastn tool provided in melon genome website (https://melonomics.net/). That reveal another putative MGL enzyme (MELO3C005786T1) with 74% AA identities to AtMGL. However, it didn’t expressed in ‘Duce’, or in PI 414723. The next two tblastn hits (MELO3C026534T1 and MELO3C009845T1) show much lower AA identities to AtMGL (33% and 29% respectively). Hence, no further analyses was conducted with these three genes.

Figure 16. L-Methionine-γ-lyase activity in melon cell-free extracts and CmMGL. (a) Scheme of the MGL reaction. (b) MGL activities in cell-free extracts derived from PI 414723 and ‘Dulce’ measured by quantification of the methanethiol product generated. (c) GC-MS chromatograms indicating enzymatic formation of methanethiol from L-methionine. Upper chromatograms in each subpanel depict full reactions. Lower chromatograms depict reactions with heat inactivated enzymes. Assay mixtures, containing cell-free extracts the recombinant CmMGL protein were incubated at 30°C for two hours and analyzed by SPME- GC-MS. (d) GC-MS chromatograms indicating the enzymatic formation of α-KB from L-methionine. Subpanel are as in (c). Assay mixtures were incubated overnight at 30°C, stopped by acidification, extracted with ethyl acetate, dried and derivatized with BSTFA+TMCS, and analyzed by GC-MS. The mass spectra and RI of the biosynthetic α-KB were identical to those of an authentic α-KB (after derivatization). Each chromatogram represents at least four assays from at least two independent experiments.

CmMGL 1 MAELKNHN....FRGTKRASPDESDDVKTDSKKPTAMISSLLEDPAAAIANTRHEFGEHGGVNMSIEASATFTVMEPETM CmMGLAtMGL 1 MAELKNHN....FRGTMAHFLETQEPLVFSGKKRARNDRDDEDGDALVAKKSALAVCD..ADPAAASPDESDDVKTDSKKPTAMISSLLEDPAAAIANTRANIRHEFGEHGGVNMSIEASATFTVMEPEDTM AtMGLBlMGL 1 MAHFLETQEPLVFSGK...... MSKRNDRDDEDGDALVAKKSALAVCD..ADPAAAITQNGIS...... TRSIVANIRHSGANPHEFGEHESHTGGVSVVNMSAPIIEAFQTSTFSATFTVMEPMMDTP...DTM CmMGLBlMGLPpMGL 11 MAELKNHN....FRGT...... MS...... MRDSKHRANNISTPDESDDVKTDSKKPTAMISSLLEDPAAAQNGIS...... GFS...... TRSRAIVIHANTRHSGANPHGYDPLSHGHEFGEHESHTGGGAGVSVVLVNMSAPPIPIVYEAFQTSTFQTSAATFTYTVMEPMAMFDPTVTP...EETMYG AtMGLEhMGL1 11 MAHFLETQEPLVFSGK...... MTAQDKRNDRDDEDGDALVAKKSALAVCD..ADPAAAIT...... TTLILANIRH..PHKEFGEHGDHVLHSHAYGGVNMSPIIEAFQTSTFSATFTVMEPCFDSTQQDTMG CmMGLPpMGL 1 MAELKNHN....FRGT...... MRDSKHRANNSTPDESDDVKTDSKKPTAMISSLLEDPAAAGFS...... TRAIIHANTRHGYDPLSHGHEFGEHGGAGVLVNMSPPIVYEAQTSATFTYTVMEPAFPTVETMYG BlMGLEhMGL1EhMGL2 11 ...... MS...... MTAQD...... MSQLKDIILTTQ...... QNGIS...... TTRSTLRVVLHHSGANP..PTPSKGWGEPLHEDSHTHVLHSHAYGSAVVHTFAPIPIFFQTSTFQTSTFQTSTYLMCMFDDTP...STQQDTQMG PpMGLTvMGL1AtMGL 11 ...... MRDS...... MSHERMAHFLETQEPLVFSGKHKNNRNDRDDEDGDALVAKKSALAVCD..ADPAAAMTTGFS...... PA...... TTRAACIHIHIANIRHGYDPLSHGANPQKHEFGEHD.QFGAGAGGVLVAINMSPPPPIVYIYEAQTQTSTFSAATTFYAVTVMEPFFDPTVNCQQEDYTMGG BlMGLEhMGL2CmMGL 1 ...... MS...... MSQLMAELKNHN....FRGTKDRAILTQ...... SQNGIS...... PDESDDVKTDSKKPTAMISSLLEDPAAATRSRVVLIHANTRSGANPTPSKHWGEPLHEFGEHESHTGSAGVVVHTFNMSAPIIFEAQTSTFQTSTSATFYLMTVMEPMFDDTP...DTQMETMG EhMGL1TvMGL1AtMGLTvMGL2 11 ...... MTAQD...... MSHERMAHFLETQEPLVFSGK...... MSGKHRNDRDDEDGDALVAKKSALAVCD..ADPAAAAIMITTDPTHTD...... PA...... TTTLACLSLIHIHANIR..PANPQKKHGEFGEHDDHVLHSHAY.QFGAGGVAIVINMSPAPIPIIFYEAQTSTFQTSTFSATFCVTVMEPLFDFDSTQQNCQQNCDTMQGG CmMGLPpMGL 1 MAELKNHN....FRGT...... MRDSKHRANNSTPDESDDVKTDSKKPTAMISSLLEDPAAAGFS...... TRAIIHANTRHGYDPLSHGHEFGEHGGAGVLVNMSPPIVYEAQTSATFTYTVMEPAFPTVETMYG EhMGL2TvMGL2BlMGL 11 ...... MSQL...... MSG...... MSKHDALIQ...... DPTHTD...... TQNGIS...... TTRVLSRSLIHVHHTPSANPQKSGANPKWGEPLHDE.QFSHTGAGASHTFIVVVAPIPIFYFQTSTQTSTFYLLMFDFDMDDTQMNCTP...DQGG AtMGLEhMGL1 1 MAHFLETQEPLVFSGK...... MTAQDKRNDRDDEDGDALVAKKSALAVCD..ADPAAAIT...... TTLILANIRH..PHKEFGEHGDHVLHSHAYGGVNMSPIIEAFQTSTFSATFTVMEPCFDSTQQDTMG TvMGL1 PpMGLCmMGL 11 ...... MSHER...... MRDSMAELKNHN....FRGTHKNNRAMTTSPA...... GFS...... PDESDDVKTDSKKPTAMISSLLEDPAAATTACRAIHIHIANTRANPQKHGYDPLSHGHEFGEHD.QFGAGAGGVALVINMSPPPIPVYIYEAQTSTFQTSATTFYVATVMEPFDFPTVNCQQEYTMGG BlMGLEhMGL2 1 ...... MS...... MSQLKDILTQ...... QNGIS...... TRSRVVLHSGANPTPSKWGEPLHESHTGSAVVHTFAPIFQTSTFQTSTYLMMFDDTP...DTQMG TvMGL2 AtMGLEhMGL1CmMGL 7711 ...... MSGMAHFLETQEPLVFSGK...... MTAQDRRMFAGELGPDRDFFIHKYARNDRDDEDGDALVAKKSALAVCD..ADPAAASIIRDPTHTD...... THF...... NPTVLNLSRQMAALEGTAAAYCTSTTSLSTLGMIHILSANIRHAIANPQK..PAAVHKEFGEHGLDDL.QFHVLHSHAYQLLGAGASGVIVGDHNMSAPIPIVVIYEAFASRQTSTFQTSTFSATLTFYLTVMEPCGFDFDGTHALNCSTQQDDQTMMAGG PpMGLTvMGL1 1 ...... MRDS...... MSHERHNNMTGFS...... PA...... TRAACIHHGYDPLSHGANPQKD.QFGALVAIPPPIVYYQTQTSTFATYAVFFDPTVNCQQEYG EhMGL2CmMGLBlMGLAtMGL 79771 ...... MSQL...... MSRRMFATGELGPDRELGPDNDFFIYVYKDSRLIQ...... HFTQNGIS...... NPTVLNLSRQMAALEGTATQAAYCTSTSGMRVRSLSVHAITPSSGANPAAVSSVKWGEPLHLMLEQSHTLLLCSSASGASHTFGDHVVGGAHPIVVFASRAQTSTQTSTFASTLYYLGMGFDMTHALDDTQMTP...MALSG EhMGL1TvMGL2 1 ...... MTAQD...... MSGHAITDPTHTD...... TTLLSLIHH..PANPQKKGDHVLHSHAY.QFGAIVAPIFYQTSTFCLFDSTQQNCDQG AtMGLPpMGLTvMGL1BlMGL 79 421 RR...... MRDS...... MSHER....MFTGGQT....RAELGPDNDFGYVFDYHSNNARMHFTGFS...... PA...... GNPTVTPNRSDLNLLESRQDVMLACEALEGLENTQASFAAYCAATVNSSTGMRAACGTSSIHAIAHGYDPLSHGANPQKEVAVFSASSVMLDQ.QFLLLCSSGAGPGLVAGDGIHPEVVPPIIVYIAYIPRDAQTQTSTFSTLATIYGAVGFFDTHALPTVTNCQQYRELLSYLKG EhMGL2 1 ...... MSQLKDLQ...... TRVLHTPSKWGEPLHAHTFPIFQTSTYLFDDTQMG CmMGLBlMGLTvMGL2EhMGL1PpMGL 7742 471 RR...... MSG...... MTAQDAACMFFAGAGQT....RAGELGPDRE....EPDFGFIFHFDYYHSASRRIHFTDPTHTD...... IG...... SNPTVTNPTPNRSDLLALNLLESRQDVQRMLMACEAASLEGLELEGNTAAGSFEAAAAGYCAALATVNLSASSTGMGLSTLGMTSSIHLGAIAIAHEVAVFSAANPQK..PAAVTSKTILGLDWTQ.QFHVLHSHAYLLLLASGPGARPGDHGDIVEAVVIPILIIASRIPRDYFVQTSTFGRTLTLIYYGGLCGGFDCTHALTYRNCSTQQFSFLHLDMALKQG TvMGL1 1 ...... MSHERMTPA...... TACIHANPQKD.QFGAAIPPIYQTSTFVFDNCQQG AtMGLPpMGLEhMGL2 EhMGL1 7947 411 RR...... MSQLAACDMLFFTAMGGELGPDNE....EPK....GDEFGYVHFHIYYKSSDRRLHFIQ...... LSGNPTVNPTVNPTLLALENQLLEFSRQEQREMMMVAACSASLEGLEGIEGTQGAEAAAGAGSYCLAAATLFSAGSSTGMGMRVSGAILAIHTPSSSVTSSSTIKSMWGEPLHTLWTLQAFLLLCSSLRQKPAGGDHTFGDHLIGHEVVLIPIAVFAAGQTSTGDSRTLTLYYYLGGGCFDTHALTFSFLHDTQMVSLFLSGT BlMGLEhMGL2TvMGL2CmMGL 4244 771 ...... MSGRRADLMFGGAQT....RAGK....RELGPDREDGGFFHFIDMYYHASARRITMDPTHTD...... HFGGNPTVTNPTVPNRSDEYFLNLELESRQDVELLVMCENSAALELEGNAGTASFVGAATAAAAAYCVNFTGSSSTGGMLSTSGAIIHSAAIEVAVFSAANPQKHSATAAVLMDLG.QFQLLLLGPKGAASAGDGDHLIIVGDHEAIPIVVIIPRDSYASRGDQTSTFTLIYYGGLGCFDGTTTHALYRVENCLLFDLKQMAG AtMGLEhMGL1 TvMGL1 7941 1 RR...... MSHERADMLFTMGELGPDNK....GDEFGYVHIYSRMHFLTGNPTVPA...... NPTVLENQLFSRQEEMMVACSALEGIEGTQAAAGASYCAATFSGSTGMACSGAIIHAIANPQKSSVSSSMTLDQAF.QFLCSSLQKGAGGDHLIAGIHPVVPIAYAGDQTSTFSTLYGVGCFDTHALTVSNCQQLFLSTG PpMGLCmMGLTvMGL1 474477 ARRGACNRMFFAAGGE....EPELGPDRQ....ESDGFGHFFIYIYYSSTRRIHFLSGNPTVNPTNPTVLALLSNLELLESRQQRGKMMIAASAFLEGLEGLEKTEGTAEAAGACVLYCAATLTASSSGMGMGAISGAIAITSAAVAATITVLWTLQTLLLLILRASKPAGDGDHGDHLIELIVVVASRSDGRETLTLCLYYGGCCTTHALTHALFFSFLHMAE BlMGLEhMGL2TvMGL2 4244 1 ...... MSGADLFGQT....RAGK....REGFHDMYHASRITMDPTHTD...... GGNPTVTPNRSDEYFLEEDVELLVCENSLELEGNAGSFVGATAAVNFGSTGGMLSTSGAIIHAEVAVFSAANPQKHSATMDG.QFLLGPKGAAGDGDHLIIVEAIPIIIPRDSYGDQTSTFTLIYGLGCFDTYRVENCLLFDLKQG EhMGL1AtMGLTvMGL2 417947 ARRGDARLMFFMTGGGK....GELGPDNK....EAEDGFGHYVYMIYYSSTRRLHFIGNPTVGNPTNPTVNSAELQNFLLEESRQEMGKVMICSAAKILEGLEEGHATQAAEGASACAAYCAATFTGSASSGMGMGAISGAIAISSSSVAASSVTMLLWTFAFQLLCSSLQKKAGDHLIGGDHLIGHVVAASDGDASDTLTLCLYYGGCCTTHALTHALFVSLFLSTE PpMGLTvMGL1 4744 AGACNRFAGE....EPQ....ESGHFYIYSTRILSGNPTVNPTLALSNLEQRGKMIASFLEGLEKTEGEAGCVLAATLASSGMGAITSAATITVWTLTLLILRKPAGDGDHLIELIVSDGRETLCLYGCGTTHALFFSFLHE EhMGL2BlMGL CmMGL 444277 A....RRDLMFFGGAGQT....RAGK....RELGPDREDGGFHFFIMDYYSASRRMTHFGNPTVGTNPTVPNRSDEYFLNLELESRQELDVVLMNSCEAALEGLELEGNGATAVSFGATAAAAAYCFVNTGSSSGMGMTSGAISAAIEVAVFSAHSATAAVLMLGQLLLLKGPASAGDHLIGDGDHEIVVISIPRDASRGDTLTLIYYGGCGTTTHALVEYRLFLLKMAG EhMGL1TvMGL2 4147 AGDARLFMGGK....GK....EAEGHYMIYSTRLIGNPTVGNPTNSAEQFLEEEMGKVICSAKILEEGHAAEGASCAAAATFGASGMG GAISSAASSVTLWTFAFLQKKAGDHLIASDGDDTLCLYGCGTTHALFVSLFTE TvMGL1CmMGLPpMGL AtMGL 44477977 GRRANRACMFFAATGGQ....ESELGPDRE....EPELGPDNDGFGYIFIHFYVYYTSRRLHFIGNPTVSNPTVNPTLSLALNNLELLESRQGKQRIMAAFASLELEGKTETAGTQEAACVAGYCLATATLSSASSGMGMGAISGAIAIAAAAVTSSSVTVTILMLLWTTQILLLLCSSKASRAPGDHLIGDHGDGGEHVVLISDASRVAGAERSCTLLYYGGCCGTHALFTHALTFSFLHMALSE EhMGL2 44 ADLFGGK....REGHMYSRMGNPTVEYFEELVNSLEGGVGTAAFGSGMGAIHSATMGLLKAGDHLISGDTLYGCGTVELFG TvMGL2AtMGLEhMGL1CmMGLBlMGL 15747794142 GRR....AHARDFFPMLFFGTMRGGGTSNQT....RAK....EAELGPDNK....GITTTDEFVDGFGYMYVHFIIGDYYTSDARRLKEIHFLTGNPTGNPTVGNPTVVTEPKAINRSDNSALENQVEGKLELFLESRQEGKEMDVTKIMVLAACSVLCEKALELEGIYFELEEGHNATQASEAVSSFAAGCASNPTLAATYCAATFVNASGASSVGMGMANGTSGAISGAIIAIPAELEVAVFSAAASSVSSCSVSRIGMTWTFLQAFHL.LLLCSSLEKGVKQKGPAGDHLIGGDHLIGDGTHVVVDNEVVIISDAIPRDAGDDSTCTLFALIYY.GGPCCMIGTHALFTHALTVSLYRSLFPLLSARLKET TvMGL1 44 GNRFAGQ....ESGYIYTRLGNPTVSNLEGKIAFLEKTEACVATSSGMGAIAATVLTILKAGDHLISDECLYGCGTHALFE BlMGLEhMGL2AtMGLPpMGL 159424447 ....AHDFACLFPGRAQT....RAGTCNK....RE....EPITTSEFVDGFHHFDMIYTASDRHGTMIGGNPTVAVSTNPTPANNRSDAILEYFALVEGRLEEDVELQRTLVQMCENSVLASLELEGYFENAGSSFVEGANPTLAATGAALAVNFLGTASVGGMATSDIGAIAPEVAVFSAELHSATTSSRMTIMAHWTGLL.EKGVGPKRAPGDGDHLITEVVVDNILIIIPRDSVGDGTRTLFAIY.GPGCMVTYRVELFSFLHSLLFPLKAKG TvMGL2CmMGL 15747 GHARFFPFGRGTSNK....EAITTTFVDGYMIGYTDRLKEIGNPTVEKAINSAVEGKLEGKTKIAVLKLEYFEHASEVSACNPTLAATASVGMANGAIIPELAACSVRIGWTFH.LEKGVKAGDHLITVVVDNSDDTCFALY.GPCMITHALFLSPARE PpMGLTvMGL1BlMGLEhMGL1 114474441 NEYEAGACNRDLFARMGWE....EPQ....ES.K....GGISIRTEGVDHFYIHLIYTSTDRTEILSGNPTVANPTLAALAIALSEN.SAKQLEFEQRGKEMTMIAVAICSSFVLEGLEWIVEGKTEEGTAEPAASGGCVNPSLAAAATGLFASDGSIVGMDIGAIAETSAATSSAKTITVSLTWTLAHTAFLLI.AANALRKQKPAGDGDHLIILELIAVDVSDAGSGDRETTLCFALYTGPCILTTHALFFSFLHQRVSPLFIET AtMGL 159 HFLPRTCNITTSFVDITDHGAVANAIVEGRTQVLYFESVANPTLTVADIPELSRMAH.EKGVTVVVDNTFA.PMVLSPAK CmMGLEhMGL1TvMGL2PpMGLEhMGL2 157 123414744 HAGHFFPDARGLIFGERMGTSNGFK....GK....EA.K....RGIVTTTKIRHFVDEGVDHYMIGILMYNDSTDLKERAKLIMGNPTVGNPTAVLEKAKAINSAAIEEYFVEGKQ.NSKFLEEEMGKELTKTVIRMIVLCSANSKYFEILEYFELEGEGHSATGVSAEPANPVGANPTLSCTAAAATNFMAGAQVSLVANGMDIIGAIPAAELSSAAVHSATCARIGSSVEAVR.GHDTLWTFMHAFG.LLEKGVLQKKAGDHLIVTHVVVDNVVVDNASDSGDTDTFATLCYLC.YTPGPMICYTTHALFLLVSQRVESPLFPLEARTEG BlMGL 114 NEYERW.GISIRTVDLTDTEALAAAI.SAKTAIVWVETPSNPGLDIVDIAETAKLAH.AANAILAVDSTFATPILQRPIE AtMGL EhMGL2EhMGL1TvMGL1 159 11744 HAHGFDWNRLLPFPRGRFATCNGK....R.Q....ESGIITTSEVDLFVDEGIHDYIIMTSTYDSDTHGRVEKMLAVGNPTVVANKAAIAEYFW.SVEGRNKPNTKLEEELGKTQVIVLNSMAVYFYFELEGLELEKTESGSVVPANPTANPTLGATCVAAATFCTGKVSVSAGMSDIDIGAIPKGELHSATIAASARMVTVVAHMCLGHT.LL.IEKGVELKRAGGDHLIARTVVVDNLVVDSSDGDATTEFATLFCTL.YSPGPMVCCFTTHALFLVELSKPLFPLEAKGE CmMGLPpMGL 157123 HFFPGIGERTSNF.GIVTTTKIRHFVDVDIGLNDLKEAKAVLEKAKAIAIVEGK.NSKTKTRMIVLYFESTVSPANPNPTLNMAQVLVANDIIPAAELVCARIGEAVR.GHDH.EKGVVTHVVVDNTFAYC.TPMIYLLQRSPPLEAR BlMGL TvMGL1EhMGL2TvMGL2 114 1204447 NEYENRFKDM.GNRARFRAGWG.Q....ESK....EAGIGVSDIVRTSFVVDGYINTYMLTAYDTSDETERLIAGNPTVAVGNPTLAAVKAAINSASW..SAKNLEKPNTKGKTAIIAMVFVYKWLEVLEKTEEHTSAPPANPTESANPCVCAATGATLCDSKVAIVSSGMDIDIGAIAKGETAAIAKAKTVSVLIAHLCWTFTH.AANAI.LEKGKAGDHLIAILILAVVDVDSDSCTETDFACFLTTTYPGPILVCTHALFFQRLPAPILEEE AtMGLEhMGL1 159117 HFWLPRRFTCN.GIITTSEVDLFVDIDITSTYDHGVEKR AVVANKAAIAW.VEGRKPNTKTQVLMVYYFELESVPANPTANPTLCTKVVASGMDIPKGELISARMVVAHCH.EKGVERGARTVVVDNLVVDATFAFT.SYPMVCFC LSKPPLEAK PpMGLTvMGL2CmMGLTvMGL1 123 15712047 HGHGARFFPAILGEFTGRKFGTSNF.K....EA.GGIVIKTTTQIVRHDFVDFVDGIYMNTLIGNYADTDIPGEAKRLKEIAGNPTLVKAEKKAIKAINSAHM.NSKVEGK.LEKPNTKGKTTKRMIIAVLIKVYFEYFELEYFEHTASTPANPEVSPANPTLKACNPTLAATNMQALVSVIIGMANDIDGAIIMAAPERELVAAVACEAVR.GHDSVRIGKDWTFAHH.SQELEKGVKAGVVGDHLIHTLVVVDNVVVDNVIASDDNTTDYCFAFCLCTY.SPGPYCMILTHALFQRLTNSPLEPARVDE BlMGLEhMGL2 114120 NEYENRFKDM.RW.GIGVSDIVRTSFVVDNTLTADSDETEAAVLAAVKAAIW..SAKKPNTKTAIMVVYWVLETSPPANPTSNPGLCDKVIVSDIAKGETIAKLIAHCH.AANA.EKGAILAVVDVDSCTFAFTTPILVFQRLPAPILEE EhMGL1AtMGLTvMGL2 117159123 HHWFQLLPPRRFRKTCNF..GIGIVETTSEVVDLDFVDFIIDDTSIMATDDVPGNVEKHGAVVIKAANEKAAIHW.LVEGR.KPNTKKPNTTQRIMVLVYVYFEYFELESSTPANPTVPANPTLKVANPTLCKVTVSAIDIDIKGPEELDIAVASVRMKVQCAHAHRKQ..EEKGVRKGDARILTLVVVDNVVVDIVDNATTFFATS.SPPCFMVILLLTNKSPLEPPLAKD PpMGLTvMGL1 123120 HGAILGETKF.GGIVKQIVRHDFVDINTLNADIPGEAKALVKAKKAIHM.NSK.KPNTKTRMIIVYFEYFETPANPPANPTLKNMQLVIIDIDMAAERVACEAVR.GHDKDAHSQEGVVHLVVVDNVIADNTYFCTSPYMILQRTNPLEPVD EhMGL2 BlMGLCmMGL 120114157 NRFKDM.NEYEHFFPRWTSN.GGIVIDSTTTVISRTFVFVDVDNTLIGATSDEDTELKEAVALVVAAEKAKAIAIW..SAKVEGKKPNTKTTKAMIVLVYVWYFELVEESTSPANPTPVSSNPNPTLGCLKVDAIVVSANDIDIIKGAPEELITAKAKCRIGILCAHHH..AANA.EKGEKGVAILILTVVVDNVVDAVDCSTTFFATTT.PPVILMIFLQRLPASPLEIARE EhMGL1TvMGL2 117123 HWQLPRRFKF.GIGVEVDLDFIDTSMADVPGNVEKVIKAEKAHW.L.KPNTKKPNTRIMVYVYFELESTPANPTPANPTLKVCKVSIDIKGEDIAVAVKVQCAHRKQ.ERKGDARILLVVVDIVDNATFFATSPCFILLTNKPLEPLD TvMGL1CmMGLPpMGL AtMGL 120157123159 HHAFFPGFLILTGEPKRFTSNFTCN..GIGIVQTTTKTTSVIDRHFFVDIVDNTIGLIANTIPGEDLKEAKHGAAVVVLKEKAANKKAIHAIM.VEGK.NSKVEGRKPNTKTKTRMIQIVLVYFEYFETSTPANPTLKVSPANPVANPTLNPTLNMAQTIIVLVANADDIMIERPAAELVVCCASKRIGEAVR.GHDRMDAHAHHSQE.EKGVGVVLTHVVVVDNIADNTTFFAYCCS.TPPMIMIYMVLTNLQRSPPPLEVDARAK EhMGL2 120 NRFKDM.GVDVSFVNTASDEAVVKAW.KPNTKMVYLESPANPTCKVSDIKGIAKICH.EKGAILVVDCTFTTPVFLPALE TvMGL2AtMGLEhMGL1CmMGLBlMGL 123159117235114 HNEYEHLGADVVVHQFWLLRPKRRFFTCNW..GGIVISETTSEISSVVIDDLKRTFFVDFIIVDDDSMAITSGLTGVPGNDAHGVEKDTEIAVIAGAIVLEANKAAAKAHAIAVLW...... VEGR.SAKKPNTKPNTKTRIQAVLMCGIVYFEVYVYFEP.TKWLVETSTPANPTLKVVPANPTLVPANPTLSNSNPMGCMDLRQLTKVDVIVIASDIDIGEPKGSLMADELEAVITSALAKKRMVLQVGLAAHCPTRKQH.M.AANAEKGVENAKRDGKILARVTAFILVVVVDNLIVVDEAVDNLSVDAERSTTFAFAFITPHS.STPPLILMVCFSILLTNLRMKQRSKPLPPLEAKEIDHE TvMGL1 120 HALTKF.GIQVDFINTAIPGEVKKHM.KPNTKIVYFETPANPTLKIIDMERVCKDAHSQEGVLVIADNTFCSPMITNPVD BlMGLEhMGL2AtMGLPpMGL 114120237123 NEYENRFKDM.LGADVVVHHGIGERWF.GIGVSSDISKIVRTSKRHFVFVDINTSLGTAGNDSDEATEDAKIAAVIAGLAAVKAKAAAIVW...... SAK.NSKKPNTKTARMIIMCGVVYWS.ENYFEVLETSPPANPTLVPANPSNPKEGMNLCMDLRGMDKVQIVLVSDIGAKGSLMAAETIVAKLALEAVR.GHDLIGAHCPTH.AANA.MEKGNAKAVILAFHVVVDNAVVDEVDLSSCERTFAFIYTPHCTPLILVGYFLQRLRMPAPRPLEILEEEH TvMGL2CmMGL 123235 HLGADVVVHQLRKF.GVSEISVDKFIDSMAGGVPGNADIIAGIEKAHVL...... KPNTRICGVYFEP.TKTPANPTLKVLVNSMMDLRQIDIGESLMDAVLKLQGAPTRKQMNAKDKILVAFVIEVDNLSERTFAIPHSPLILSLTNRMKPLEDH PpMGLTvMGL1BlMGLEhMGL1 123120191117 HLGADVVGAWILGETPKRFF.IGGIHVSKQTEIVTRHDKDLFVDIINGHNTDLTSNADIPGESAKDVIVEKALVGKAKGKAAIHVAMLAGDGSTW..NSK.KPNTKTRMIICMVYFEPVYYFERALAETKSPANPPANPTLKVPANPTVERLNME..S..YLASVCQKVLVIISDIDMAAERKGVIACEAVR.GHDKVDGVAHLGCHSQEI.AEPRGVFGVDAARHLVVVDNVWLILVVDATRDNRGIKTLATYFCTTSPYMIPCFLVQRTNRMLKPLEPAKVDH AtMGL 237 LGADVVVHSISKFISGGADIIAGAV...... CGS.ENLVKEMMDLRGGSLMLLGPTMNAKVAFELSERIPHLGLRMREH CmMGLEhMGL1TvMGL2PpMGLEhMGL2 235117123200120 LGADVVVHNRFKDM.HLGADWQLPRRFKLFVVH.GIGSVSISEADVTKDLDKSFFYFVIILSDSNTGTSMAGHGAADVPGNGSDEDVEKDIIIAGTAVVIAGKAEVAKLKAVAHV...... W.L...... V.KPNTKKPNTRICGMVYVYFEGP.TKRL.KAESTLVPANPTPANPTLKVLVNSDRMIMDLRQCRKV..LESIDIGGSLMKGELDKIAVDLAMAKLVKTGGVQIPTCAHRKQVLM.NAESEKGRKPKGDHVARILDAAAFILLVSEVVDILLSVDNLMERARGIKTLCTIFFAPHTSTLPSCFILAVLLFRMKLTNRMKPAPLEPLDRELEHDH BlMGL 191 LGADVVIHSTTKFINGHSDVIGGAVLAGDGSTCPRAAKVVERLE..S..YLASVGLGIAPFDAWLTRRGIKTLPVRMAKH AtMGLEhMGL2 EhMGL1TvMGL1 237120194 LGADVVVHNRFKDM.LGADHALTIKAFL.GHGISVSISDVSQVKSKDFFVYINGHFIISNTGGAASDEGIPGEDDVIIIAGAVGVVGKAKAVKVS...... SH...... W.M.KPNTKCGMIVYAKVYFES.ENLTEASETLVPANPTDPANPTLKIKEATMIMDLRGCKKV..F..YIISDIDGMSLMKGERRIKDAVLAKCLKGIGDPTCSAHHLM.MSQENAEKGAPKMGVVADAFLAFILLVEVVDILSCAADNERCRGTIFMPHTKTLCTSLPGVPMILFIRMLRMTNPARQIPELEVDHH CmMGLPpMGL 235200 LGADVVVHLGADLVVHSISATKFYILSGGHGAGDIIAGTAGALV...... VCGGP.TKR.KALVNSDRMIMDLRQR..LEGSLMLKDLMLTGGPTAVLMNASPKHVDAAFSELLSLMERRGIKTLIPHLSALRMKRMDREH BlMGLTvMGL1 EhMGL2TvMGL2 191120197123 LGADVVHLGADVVAQLTRKFI.LHGIHGSSVTQIETVTKDKFFYINGHINGHINTDMAASIPGEGVPGNDVIDVVGGVIGKGEAKVVHVLAGDGSTM...... TTL.KPNTKKPNTRICIPVYFERKANPAKTEVPANPTLKLPANPTLKVVLEQRLKIE..S..YLASVK..A..FIIIDDIMERERDVKDAVCKTGGDQLGAHSALRKQISQEMAAPPKFGVMDDADAFLILLWVLITRACIVDNDNRGIKTLRGTFVFACKTLSPPMIPILVMRMTNRMKAKPPLVDVHHD AtMGLEhMGL1 237194 LGADVVVHLGADIALHSISVSKFYINGHISGGAGDDVIIIAGGGAVVS...... S...... CGAKS.ENTAELVDIKEATMIMDLRGK..F..YGSLMRKDALLGPTSLMNAAPKMVDAFLAFELSCAERRGIMPHKTLLGPLIRMRQIEH PpMGLTvMGL2CmMGL TvMGL1 200123235198 LGADFHLGADVVVHQGLVRDVVVHLKVVHF.GSVSAEISATVTKDKYFYINGHLISDSGHMAGGGVPGNATDDDVIITIAGVAGIAGELKALIVHV...... VL...... KPNTRICGGVYFERP.TKK.KA.ATDLVPANPTLKVLVLLDNSQQRIMIRMDLRQR..LE..MVIDIGGLESLMIKDKDAVDMLITGKLTGQGAAPTSVLRKQVIMSNASPKPHDKHDAILVDAAFSVWLIELLVDNLSITMRGIKTLERRGTFAILPHSSTLPLAILSNLLIRMTNRMKDRPLEAEHDH BlMGLEhMGL2 191197 LGADVVILHSTITKFYINGHINGHSGDVIDVVGGGAVVLAGDGST...... TTCPRKANPAKEVLVLEQRLKIE..S..YLASVK..A..FRKDTGGLGSLIMAPFMDADAFLWLTRCIRGIKTLRGVKTLPVMRMRMKAKVH EhMGL1 AtMGLTvMGL2 194237201 LGADLGADVVVHLGVDIIAVVHLHSSVSISATKKYINGHFYINGHISGGGATDVIDDVIIAGVGAGGVALS...... SV...... CGCAKSS.ENRT.AAEDIIDLVIATKEAKVKIMKMDLRG..F..Y..SQGSLMIRKKDADLILTGGGSPTALIIMMANASPPMKHDAFLVDAAFWELCALSITRGERRGMITLKTLPHTLLPGDILMRMRMRVQIRKERAHH PpMGLTvMGL1 200198 FLGADGVDVVVHLVVHSATKYYINGHLSGHGTDDVITVAGLLIV...... V...... CGGRK.KA.ADLVLLDQQRIR..LE..MVGLIKDMITGASVLVISPHDASWLLITMRGIKTLRGLSTLANLIRMRMKDRAEH EhMGL2 BlMGLCmMGL 197191235 LGADVVLGADVVLGADVVVHLIHHSSITISTTKKYINGHFINGHISGGGSADVDVIDIVGIAGGGGVAVV...... TTLAGDGST...... CCGPKRP.TKNPAAEKLVLVLVQENSKRLIMKE..S..YLASVMDLRQ..A..FGSLMRKDLTGLGSLGPTLMIMAANAPPMFKDAFLDAVAFWLECITRLSRGRGIKTLERVIKTLPHLPPSMVLRMKRMRMKAKVEHH EhMGL1TvMGL2 194201 LGADLGVDIAVVHLHSVSATKYINGHGTDVIDVVGAGGVLS...... SV...... CAKSRT.AEDIIDIATAKVKIK..F..Y..SQGIRKKDADITGGSALIIMASPMHDAFLDAWLCAITRGMTLKTLTLPDIMRMRVQIKRAH TvMGL1CmMGLPpMGL AtMGL 198235200237 FLGADVVVHLGADGVDVVVHLVVHSSAISATTKKYINGHFYILSGGHGTAGDVDIVIAGTAGAGLIALV...... VCGCGGKP.TKRS.EN..KAADLLVLQQNSDKERIMIRMDLRQRMDLRG..MV..LEGGISLMLKKDDILMTGLTGGSPTAVIVLMSNASPPHKHDAVDAAFWSLELITLSLMRGERRGIKTLLISPHTLLNSAGILRMKRMKRMDRRAEEH EhMGL2 197 LGADVVLHSITKYINGHGDVVGGVV...... TTKNPELLQKIK..A..FRKDTGSLMAPMDAFLCIRGVKTLPMRMKVH TvMGL2AtMGLEhMGL1CmMGLBlMGL 201237194307191 LGCRRLGADVVVHLGADLGADVVVDAIISVFVVHALIHAESSAISVSRTTTMKKKYINGHFYINGHKAG..LIINGHSGGTAGSDVDDVIKVIVIAGIAGGYPGLEGLAVVVS...... S...... LAGDGSTDHPQHQCCGLCSAKMPRS.ENKR.TSLANPEYAAADIIEKLVDVIVAKEATEKVKRLMIGFGMDLRGKE..S..YLASV..SQ..F..YGMLGGCVISLMKRDDKDAMILG.TGLGTAPTSELGIILEMRISNAAPPNHKMKLMFDAVDAFLDAAFWSIWLELITLSCALTRQNTRGERRGRGIKTLTLIMTPHKTLQFGFTLLDGPMLIMVRRMAVSLGVKRQIAKRAEH TvMGL1 198 FGVDVVVHSATKYINGHTDVVAGLI...... CGK.ADLLQQIR..MVGIKDITGSVISPHDAWLITRGLSTLNIRMKAE BlMGLEhMGL2AtMGLPpMGL 191197309200 SHRLGADVVLGADAQLVYVVHILHAESTIRATMRKFYINGHDLG..MYINGHLSGHSGDVIDVKVDIVGITGYPGLEAGGAVLVLAGDGST...... TT...... VTHPQHKCLPFGRKANPGM.KAAKEVVLNRDYLVVLEQDRLKRIGE..S..YLASVKYR..A..FG..LEGLLSGILRDKKDMDE.MTGGTLGSEALEVLIMKASPNFMKLMHDADAFLWAYSLTRCILQNAMRGIKTLRGVTKTLQFGFPAVMLRMRMKAVSLGAKDRVH TvMGL2CmMGL 201307 CRRLGVDAISVFVVHAESARTMKKYINGHKAG..LTDVKVVIAGYPGLELV...... DHPQHQCLSMRK.SLANPEYADIIAKVKGFG..SQGMLGCVIKDDMIG.TGTAEIIERSAPNHKLMDAWSILITLQNTRGTLTQFGFTLDMRAVSLGVKRA PpMGLTvMGL1EhMGL1BlMGL 200198194267 FCLGADGENAQVDVVVHLIVVHALVAHSQAVSWTLKESRYYINGHLSPGHEIGTAEDDVDVIIVTVYAGGYPGLGLLIVVS...... S...... V...... PSHPGHEVCGGAKAKKQMRK.KA.TADELVLDS...LIDQQATRIGFGRK..LE..MV..F..YGVVGSFLIKRRTD.DKDAMITGGTASEARVLVILMSAPLSHMDADAFLLVSWKSTLLITCAMKLRGIKTLRGILMTSKTL....TLANPLILARMRMKEDRQISLGAEH AtMGL 309 SHRAQVYAERMRDLG..MKVIYPGLETHPQHKLFKGMVNRDYGYGGLLSIDME.TEEKANKLMAYLQNATQFGFMAVSLG CmMGLEhMGL1TvMGL2PpMGLEhMGL2 307194201270197 CRRCALGADLGLGADVVVNAQADSVFIAVVHQLVAEAEHSRVSAIMKTLKARQKAG..LYINGHPQVGTELKVDVIDVIIVHVGYPGLEGAGYPGLGVLS...... SV...... P...... TTDSHPFAQYQHQELLCMAKSAKRQKSLANPEYT.RNPAQMEDIIDRLI...LPLATAQKVKKGFGIK..F..Y..SQG..A..FGGMMLICVGAIFDRKELMKDADKDG.IKTGGGITGESAELIIERMAAASGRRNPKLMMHDAFLDAFMSIWNALLCAITLCIQNTQRGLFA....RTMTLVQFGFKTLTLPDMIMAVSLGRMRAVSLGRMKVQIKRAVH BlMGL 267 CENAQAVAQWLESRPEIAEVYYPGLPSHPGHEVAKKQMS...GFGGVVSFRTD.TEARALSLVKSTKLIT....LAESLG AtMGLEhMGL2 TvMGL1EhMGL1 309197198263 SHRFMLGADVVGENVAQDVVVHGLVYKVALAEHSKRIAFMRTLKEQDLG..MYINGHHEKIGTVKVDVKVIVGVNHYPGLEAGGPGLESVLIV...... TT...... THPFPGHQHKDLCGIAKKQMFKKGMNP.AVEDNRDYLT...LQQQKGIGYKRYG..A..F..MVGGSTFLLSGSFIIDRKEMKDDMKE.ITG.TSESFELVIEKMAAASAKKLMNPKLMMHDAFLDAAYWELHLCIITLKQNARGVCTVLTQFGFKTLS....TLPNMMILAVSLGAVSLGRMKVAEH CmMGLPpMGL 307270 CRRCANAQASVFQVAEAERAMKLARQKAG..LPQVELKVIIHYPGLEYPGLPDSHPFAQYQHQELLMAKQSLANPEYRQMR...LPGFGGGMLICVAFDELMG.KGGITEERAAGRRNKLMFMSINALQNTQLFA....RTQFGFMAVSLG BlMGLTvMGL1 TvMGL2EhMGL2 267198201266 CFMLGENAQGENVDVVVHDGILAVVHKVAVAQSKWAFLTLESRKEQYINGHHPPEKIIAETKQDVVVYVNHYPGLAGPGLESLILVP...... SHPGHFPGHEVKCGCIAAKKQMSKRME.AQMDDIIS...LKL...QQAKVKGFGIGFGR..MV..SQGSTFVVSFGSFIRTD.KEDMKITG.TSEARSAFVIIIEAASAKKLMLSPHLDAVKSTWELHITVKLKLRGICG....TLTL....STLNDLAIMLAVSLGRMKREVSLGKAERA AtMGLEhMGL1 309263 SHRMENAQGLVYKVAAEKRFMRLEQDLG..MHEKIVKVINHYPGLEPGLESTHPFPGHQHKDLIAKKQMFKGMVNRDYT...GYGGSTFLLSSFIDEMMKE..TSEFEKAAAKKLMNKLMAYEHLLKQNAVCTQFGF....MLAVSLGAVSLG PpMGLTvMGL2CmMGL TvMGL1 270201307268 CACRRSLGENANAQVDAISVFMQVVHKVAEVAEAESAARYLTMKLARQKYINGHKAG..LSHPPQAVVELTEDVKVIHVIYYPGLAGYPGLEYPGLVFP...... ESDFAQYHPHEQGHHELQDCLIAKKQMASMQRKR.SLANPEYQMADIIRR...LP...MSAKVKGFGG..SQGGGSMMILITAGCVFIFELKDIDMLKIG.KGGITGSGFTAEEIIEARGGRRSAAKKLPNHKLMFDAMLDNAWSILNLITLLKLQQNTLRGFA....RITLTQFGF....TLDMLAVSLGAVSLGRAVSLGVKRA BlMGLEhMGL2 267266 CMENAQENGLAKVAVAQKWFLESREQHPPEKIAEKQVYNHYPGLPGLESPSHPGHFPGHEVKIAAKKQMMEQMS...K...GFGGSTFVVSFRTD.EMK.TSEARFEAAAKKLMLSLVKSTEHVKLICG....T....LALAVSLGESLG EhMGL1 AtMGLTvMGL2 263309271 MSHRAENENAQKVAEGAQLKVAVYAEKFRFLMRLEQHDLG..MEHHEKKAIVVKKVKVNHIYYPGLEYPGLPGLESPDTFHPHPGHPGHQHDKEIAKKQMLIAKKQMFKGMVT...NRDYK...MGGYYFGGGSTFGSLLMISFSAIFEDDVMKME.D.GL.STFEEEAKAKKLMAAKKNKLMVLDEAYHNCHLKLQNAVVCVTTS....QFGF....LAVSLGMLAVSLGAVSLG PpMGLTvMGL1 270268 CASENANAQMQKVAEVAEAYLARQKSHPPQAVELEKVIHYYPGLYPGFPESDFAQYHEGHELDIAKKQMAQRQMR...LP...MSGGSMIITAFELILKGGISGFEAGGRRAKKLFMLDNANLLKLQLFA....RIT....LAVSLGAVSLG EhMGL2 BlMGLCmMGL 266267307 MCCRRENENAQGALSVFKVAAVAAEKQFWRLLMKEQESRKAG..LHPPKEIIKQAEKVVVNHYIYPGLYPGLEPGLESPSHPGHDFHPPGHQHKEVQIALAKKQMMMEKSLANPEYQMKS...... GFGGFGSTFGVVMLSFSFCVERTD.DMKMG..STFEAREEEARAKKLMALSNKLMLVEKSTSIHVLKLKLQNTCG....IT....QFGFLAVSLGLAMAVSLGESLG EhMGL1TvMGL2 263271 MAENENAQKVAEGLKVAKFLEQHEHEKKAIVVKKVNHYYPGLPGLESPDFHPGHPGHDEIAKKQMT...K...MGYFGGSTFSMISFAFEDVMKD.GL.SFEAKAKKLMAKKVLDEHNCHLKVCVTS....LAVSLG TvMGL1CmMGLPpMGLAtMGL 268307270309 SCRRCASHRENANAQAAQMSVFKVAEQVYVAEAEYRALMKLMRKARQSKAG..LDLG..MHPPAQVVEELKVKVIYIHYPGYPGLEYPGLFEPDDSTHHPFAQYEGHQHDQELKIAKKQMLMAFKQSLANPEYRGMQMVRRNRDY...MS...LPGFGGYGGSGMMLLITLICVASFFIIDELLMKG.KE.SGFGGITEEEGRAKAKKLAGRRNKLMFLDMSINAAYNLKLLQNTQQNALIFA....RTT....QFGFLAVSLGMAVSLG EhMGL2 266 MENGLKVAKFLEQHPKIKQVNHPGLESFPGHKIAMEQMK...GFGSTFSFEMK.SFEAAKKLMEHVKLCG....LAVSLG TvMGL2AtMGLEhMGL1BlMGLCmMGL 271309263267384 ASHRMCYYENAQKVAEENENAQEAQGTLVYKVAAMVASCSAEKQFRFWGLMRLSHEQESRSEDLG..MTHHSSEKAEKPEVIMKVAESGEKVKVVYINHYEREYPGLYPGLEYPGLPGLESLAGIPDPTSHPGHHPGHHPFSPGHPGLVRQHEKDEVIAKKQMLIAKKQMFAKKQMMKSGMIGVYKNRDYT...S...M...MGTLEGGFGFGYQKWSQFEKGSGSTFMLLVVIASSFFIDVDERTD.MMKALD.GLE..AKTSEFEARVEEQDIGVPFCNN...... KKAAKKAAKKLMNLSKLMVLDLVAYEKSTNCHHLLKQNAKLVVVCIST....QFGF....LAVSLGMLAVSLGLAAVSLGESLG TvMGL1 268 SENAMKVAEYLKSHPAVEKVYYPGFEDHEGHDIAKKQMR...MSGSMITFILKSGFEGAKKLLDNLKLIT....LAVSLG BlMGLEhMGL2PpMGLAtMGL 267266270386 CMCAYYENAQENNAQEGTLAKVAQMVAVAESCSQKWFAGLSESREQARQSTHPSSEPEKQIVLAEKQELDPSQVIYNHHYPGLKPGLESEAAGIPSHPGHSFFAQYSPGHPGLVREVKELIAAKKQMAMMEQSVGRQMYS...KRV...... LPGTLEGFGQKWTQFEKGGSTFVVMISFAFRTD.EELMKAKFLR.GGITSEARFME...... AAAKKLMGRRLSLFVMKSTENAHVLKLQLICG....FA....RT....LALAVSLGAVSLGESLG TvMGL2CmMGL 271384 AYYENAQKVAEETLMSCSFGLSHSETHSSEKAVMKSGEKVYEREYPGLLAGIPDHPGHSPGLVREIAKKQMMSIGYKM...MGTLEFGQKWSQFEKSMIAFDVALD.GLAKVEQDIGVPFCNN...... KAKKVLDNCHVVS....LAVSLG PpMGLTvMGL1EhMGL1BlMGL 270268263339 CASMGVENAENNAQEGSMLLIQKVAEKVAVAEDHPAKAYFLTARQKEQMTHSHPHPEKLQAVIELEVADCKVIHYNHEYPGLYPGL...SPGLESFPESDVSFAQYHFEPGHPGHTFELDIIAKKQMRLSAQRLQMG.RT...I...LP...MSEDIGAYDIGGSSTFLMADLIITASFFDAELIELMKALKGGISGF.ASTVQTPDGVSSDAEALGAEVPIADPSSFEAGGRRAKKLAKKLMFMLDNAENHLLKLLKQLVFA....RICT....LAVSLGAVSLG AtMGL 386 YYETLMSCSGSSTSSELDPSQKEAAGISPGLVRMSVGYVGTLEQKWTQFEKAFLRM...... CmMGLEhMGL1TvMGL2EhMGL2PpMGL 384263271266343 YYMADAENENAQKVAEEETGSLLMKVAAQSCSHPASMTHKGFSLSEQHTESSEHHPEKKASSYTAKKMIVSGEVKKQKVVERENHYERYPGLPGLESAHHLAGIPDGISFHPGHSPGLVRPGHEGLVRLSVGDEKIAKKQMIAMSMEIGQMY.MT...KLGT...M...EDLEVEGGFGQKWSQFEKYFGDGLLSTFSMADISFAVFEEDVQALALMKD.GLAK.QACS...... SVFQDIGVPFCNN...... EAKAKKLMAKKVLDEHNCHLKVKLVCVCG....TS....LAVSLG BlMGL 339 GVESLIDHPATMTHLAVADCEL...SVSPTFIRLSLG.IEDIADILADLDAALASTVQTPDGVSSDAEALGAEVPIADPS AtMGLEhMGL2 TvMGL1EhMGL1 386266268335 YYMSCVENENAEDTTGLLMLIMKVAKVAESCSEHPASMTHAKGFYSLSEQKTSSSEHPKALIVVPDPSQKQEKVENIMRKQVNHYKYPGEPGLESAAGIFEGIDSFHTPGLVRPGHEPGHELVRKDIAIAKKQMMISVGMESVGQMY.VKRIGT...... MSENLEVDGFGQKWTQFEKDIIADLKQALGSTFSMITSFFEIAMKLFLRK.SGFESLW...... MF...... EAGAKKLMAKKLLDEHNVLKLKLCG....IT....LAVSLG CmMGLPpMGL 384343 YYDAETSLMAQSCSHPASMTHGSSTSSESSYTAKMSGEEREERAHHLAGIGISPGLVREGLVRLSVGMSIGY.MLGTEDLEVEQKWSQFEKDLLADVEQALALAKQACS...... VQDIGVPFCNN...... BlMGLEhMGL1TvMGL1 TvMGL2EhMGL2 339335268271338 GVCVSATLENAENAQKVAEEDTSLILIMKVAEDEHPAHPASMTHAYFTLMTHKHSEHPHLKAAAVVPVADCEKENIMRKQKVEHLLKQQEYL...SYPGYPGLFGIEPDGVSDLTTHHPGHPPERETFEGHLVRIDERLSIAKKQMISVGLG..IIRKLEE...MS...MDNNPIVDADDIDIIADLKQALFGGLSADLMITIADAFIDVALLKD.GLASTVQTPDGVSSDAEALGAEVPIADPSESGFLW...... QC...... EGKAKKLAKKVLDLDNNCHLKLVIVTS....LAVSLG TvMGL1AtMGL 341386 GCYYESTLILMQSCSHPASMTHAGSSTSSEVVPLDPSQKEEREKEAAGITSDPGLVRGMIRLSVGMSVG.YIVEGTDALEDELQKWTQFEKIADFKQGALFLRDALML...... PpMGLEhMGL2TvMGL2CmMGL 343338271384 DATLAYYENAQKVAEEDTESTLLILAQMESCSHPASMTHHPASMTHAFGLSHSETHSSESSYTAKKAAVPVMKSGEEHLLKQQKVERYEREYPGLAHHLAGIGIGPDLTSHPGHSEREPGLVRGLVRLSVGLVREIAKKQMIMSVGSIG..YLLKMEE...MGTDNPVELEDDDIIADLKQALFGQKWSQFEKLLSADMIVAEFQALDVALD.GLQACS...... EAKQC...... VEQDIGVPFCNN...... KAKKVLDNCHVVS....LAVSLG TvMGL2BlMGL 343339 GPGVESLIQDHPASMTHAHPATMTHLGAVPVADCKEEREEL...SAAGLTVSDNPTFLIRLSVGRLSLG.C.IENDVIQADIIDILDADLDLKQALDAALDASTVQTPDGVSSDAEALGAEVPIADPSLVL...... EhMGL1TvMGL1 AtMGL 335341386 CVGCYYDTESTLILILMEQSCSHPASMTHAHPASMTHAGSSTSSEAVVPVPLDPSQENIMRKQKEEREKEAAGIGITTSPDPGLVREGLVRMIRLSVGIMSVGSVG..YIIVEEGTNDAVDLEDELDIIADLKQALQKWTQFEKIADFKQGALFLREDLW...... ALML...... PpMGL 343 DAESLAQHPASMTHSSYTAKERAHHGISEGLVRLSVG.LEDVEDLLADVEQALQACS...... EhMGL2TvMGL2 BlMGLCmMGL 338343339384 TLGPGVYYDTESTLILILMEQDSCSHPASMTHAHPASMTHAHPAGTSMTHSTSSELAGAVPVPVMADCSGEEHLLKQQKEEREEL...SALAGAGIGLTLTVSSREDNPPGLVRTFLVRLIRLSVGRLSIMSVGSLIG..C.YLIMEEGTNPNDVILEDQADIIADLKQALDIIDIQKWSQFEKLDADLDLKQALDAALEDASTVQTPDGVSSDAEALGAEVPIADPSAKQC...... LVL...... QDIGVPFCNN...... EhMGL1 335 CVDTLIEHPASMTHAAVPENIMRKQGITPELVRISVG.IENVDDIIADLKQALELW...... TvMGL1 CmMGLPpMGLAtMGL 341384343386 GCYYDAEESTSLILMAQQSCSHPASMTHAHPASMTHGSSTSSESSYTAKVVPMLSGEDPSQKEEREEREERKAHHEALAAGIAGIGITSDPGLVREGGLVRLSVGMIRLSVGMSSVGIG.Y.IMLVEGTEDADLEVEDELQKWSQFEKDQKWTQFEKLLIADADFVKQEQALGALALFLRDAKQACS...... ALVML...... QDIGVPFCNN...... CmMGLEhMGL2 338 ...... TLDTLIEHPASMTHA AVPEHLLKQQGLTRELVRISVG.LENPDDIIADLKQALEQC...... TvMGL2 AtMGLEhMGL1BlMGL 343386335339 GPYYCVGVEEDTSTSLILLIMQSCSEDHPASMTHAHPASMTHAHPAGSTSMTHTSSELGAVPLVPVDPSQADCKEENIMRKQEREEKL...SEAAAGAGIGILTVSSTDNPGLVRPETFLLVRIIRLSVGRLSMISVGLG.CY.VIEGTENNDVLEVDIQADIIQKWTQFEKDIIADLKQALDILDADLDLKQALDAAALFLRDEASTVQTPDGVSSDAEALGAEVPIADPSLLW...... VML...... AtMGLTvMGL1 341 ...... GCESLIQHPASMTHA VVPKEEREAAGITDGMIRLSVG.IEDADELIADFKQGLDALL...... CmMGLBlMGLPpMGLEhMGL2 339343338 ...... GVDATLEDTSLILAQDEHPAHPASMTHHPASMTHAT MTHLSSYTAKAVVPADCEHLLKQQEERL...SAHHGIGVSLTSPERETFGLVRLSVGLVRIRLSISVGLG.ILEDNPIVEADDIDDIIADLKQALLLLADLADVDAEQALALASTVQTPDGVSSDAEALGAEVPIADPSQACS...... EQC...... BlMGLTvMGL2 415343 PSTLQHPVATVGPESLIQHPASMTHA GVPKEEREAAGLTDNLIRLSVG.CENVQDIIDDLKQALDLVL...... AtMGLPpMGLEhMGL1TvMGL1 343335341 ...... DACVGCEDTSLLIAQEQHPASMTHHPASMTHA SSYTAKAVVPENIMRKQKEEREREAHHAAGIGISTEPDGLVRLSVGEGLVRMIRLSVGISVG.LIEDNDAVEVDDELDDIIADLKQALLLIADADVFEKQQALGLQACS...... EDLW...... ALL...... PpMGL ...... CmMGLBlMGLEhMGL1EhMGL2TvMGL2 415335338343 ...... PSTLQHPVATVCVTLGPDTESLIEQHPASMTHA AGVPENIMRKQEHLLKQQKEEREAAGGIGLTTPREDNELVRLIRLSVGISVG..CILENNPVDVDQDIIADLKQALDIIDDLKQALEDLW...... QC...... LVL...... EhMGL1 ...... AtMGLPpMGLEhMGL2TvMGL1 338341 ...... TLGC...... DTESLIEQHPASMTHA AVVPEHLLKQQKEEREAAGIGLTTREDGLVRMIRLSVGISVG.LIENPDADDELDIIADLKQALIADFKQGLEDQC...... ALL...... EhMGL2CmMGL ...... BlMGLEhMGL1TvMGL1TvMGL2 415 341343 PSTLQHPVATV...... GCGPESLIQHPASMTHA VGVPKEEREAAGIAGLTTDDNGMILIRLSVG..CIEDANVDELQDIIIADDDLKQALFKQGLDALLVL...... TvMGL1AtMGL ...... PpMGLEhMGL2TvMGL2 CmMGL 343 ...... GPESLIQHPASMTHA GVPKEEREAAGLTDNLIRLSVG.CENVQDIIDDLKQALDLVL...... TvMGL2BlMGL 415 ...... PSTLQHPVATV EhMGL1TvMGL1 AtMGL ...... PpMGL ...... EhMGL2FigureTvMGL2 CmMGLBlMGL 1 415 7 ...... PSTLQHPVATV Multiple sequence alignment of amino acids from CmMGL and other TvMGL1CmMGLAtMGL PpMGLEhMGL1 ...... EhMGL2 ...... TvMGL2AtMGL BlMGLEhMGL1 415 ...... PSTLQHPVATV fTvMGL1 unctionally ...... characterized MGL enzymes. PpMGLBlMGLEhMGL2 415 PSTLQHPVATV...... TvMGL2 ...... EhMGL1PpMGLTvMGL1 ......

AlignmentsEhMGL2EhMGL1TvMGL2 were...... performed using the EBI ‘MUSCLE’ tool and shaded using BoxShade version TvMGL1EhMGL2 ...... 3.21.TvMGL2 TvMGL1 Amino ...... acids conserved in four genes or more are shaded in black. Amino acid similarities TvMGL2 ...... are shaded in gray. Red background indicates the conserved Lys residues that covalently bind the PLP cofactor, and the pink background indicates conserved amino acids important for the formation of the homotetramer according to X-ray crystal analyses of PpMGL. CmMGL: Cucumis melo L-methionine-γ-lyase, GeneBank accession no. AGF70153; AtMGL: Arabidopsis thaliana L-methionine γ-lyase, GeneBank accession no. AEE34271.1; BlMGL: Brevibacterium linens L-methionine-γ-lyase, GeneBank accession no. AAV54600; PpMGL: Pseudomonas putida L-methionine-γ-lyase. GeneBank accession no. 1GC2_A; EhMGL1: Entamoeba histolytica L-methionine γ-lyase 1, GeneBank accession no. BAC75877; EhMGL2: Entamoeba histolytica L-methionine γ-lyase 2, GeneBank accession no. BAC75878; TvMGL1: Trichomonas vaginalis L-methionine γ-lyase 1, GeneBank accession no. CAA04124; TvMGL2: Trichomonas vaginalis L-methionine γ-lyase 2, GeneBank accession no. CAA04125.

CmMGL displayed all the conserved amino acids that are considered crucial for tertiary and quaternary structure and activity. They include Lys 246 that covalently binds the PLP co-factor (Figure 17, red shaded; Sato and Nozaki, 2009), as well as Tyr 93, Arg 95, Gly 123, Met 124, Tyr 148, Thr 224, Ser 243 and Arg 416 that are important in the formation of the homotetramer and the active site (Figure 17, pink shaded; Sato and Nozaki, 2009). Moreover, Gly 150, which is located in the active site, is conserved only in melon, Arabidopsis thaliana and in Brevibacterium linens (Figure 17, green shaded). In the two latter species, MGL activity was shown to be restricted to L-methionine and its close derivatives (e.g., L-ethionine, L-homocysteine and seleno-L-methionine) as substrates (Goyer et al., 2007; Kudou et al., 2008). In the other encoded enzymes, Gly 150 is replaced with Cys (Figure 17, blue shaded). These enzymes can efficiently utilize a broader range of substrates, such as L-cysteine and homocystathionine (Kudou et al., 2008). Among the functionally characterized MGL enzymes used for the alignment, CmMGL shows the highest similarity to AtMGL (72% AA identity, by ClustalW2, http://www.ebi.ac.uk/Tools/msa/clustalw2/) that is also reflected in longer N-terminal domain. That is in compare to the non-plant MGL enzymes where AA identity range from 26% to 35%. CmMGL belongs to the Cys/Met metabolism PLP-dependent enzymes according to Pfam (http://pfam.sanger.ac.uk/). According to the Panther classification system (http://www.pantherdb.org/), CmMGL is a methionine-γ-lyase enzyme and predicted to possess a catalytic lyase activity. The CmMGL full-length sequence was cloned into a pET21a vector attached to 6×His-tag at its C-terminus, and purified on a Ni-NTA column using buffer containing 250 mM imidazole for elution. The purified enzyme was resolved on SDS-PAGE (Figure 18), and its calculated molecular mass was found to be 52017 Da. This is in a good agreement with the predicted mass of the recombinant protein (by Expasy) that is 49803 Da.

Figure 18. Purification and molecular mass identification of the recombinant CmMGL.

Proteins from E. coli harboring the 6xHis- tagged CmMGL construct were cleaned by Ni- affinity chromatography and eluted with a sodium phosphate buffer containing 250 mM imidazole. The elution fractions, as well as the flow thru (FT) and wash (W) fractions, were resolved on an SDS-PAGE with Coomassie blue staining compared to the first elution fraction of proteins from E. coli harboring an empty pET21a vector.

CmMGL exhibited in vitro enzymatic γ-elimination of L-methionine, generating the products ammonia, methanethiol and α-ketobutyrate (Figure 16cd). α-Keto-γ-methylthio butyrate (KMBA) was an inefficient substrate for CmMGL, unlike other MGL enzymes isolated from Saccharomyces in which KMBA served as a substrate (Perpète et al., 2006).

Validation of CmMGL’s role in sulfur volatile formation To validate the role of CmMGL in the production of the +1 enriched volatiles, I analyzed its expression levels in the PI 414723 × ‘Dulce’ RILs population (Harel-Beja et al., 2010). RNA-sequencing was used to monitor gene expression levels of selected recombinant inbred lines. The S-compound levels of 19 recombinant inbred lines, exhibiting the highest CmMGL expression levels, were compared to those of 19 lines exhibiting the lowest CmMGL expression values. As expected, lines with high CmMGL expression showed significantly higher total amounts of S-compounds enriched with +1 m/ z unit than those with low CmMGL expression (Figure 19a). The differences are significant according to the Wilcoxon test (α<0.0001). Conversely, no significant differences were observed in the total amounts of S-compounds enriched with +4 m/z units between the high and low CmMGL-expressing groups (Figure 19).

Figure 19. Co-segregation analysis of CmMGL expression levels with the levels of S- compounds enriched with +1 m/z unit in PI 414723 × ‘Dulce’ recombinant inbred lines. The total levels of S-compounds exhibiting enrichment with +1 m/z unit were compared in the fruit flesh of 19 RI lines expressing CmMGL at the highest levels in the population and 19 RI lines expressing CmMGL at the lowest levels in the population (upper panel). Differences are significant (α<0.0001) as determined by the Wilcoxon test. No segregation was found between the total levels of S-compounds enriched with +4 m/z units in the same RI lines (lower panel). For each RI line, three fruits of different plants were analyzed by SPME-GC- MS.

The role of L-methionine aminotransferase in melon aroma production

Incubation of melon cubes with KMBA enhanced the levels of +4 enriched compounds but not of +1 enriched compounds To validate the role of KMBA as an intermediate in the production of C3-thio-ethers 13 (enriched with +4 m/z units upon incubation with C5-L-methionine, Figure 7), I carried out incubation experiments with non-labeled precursors in lower concentrations than those used by Gonda et al. (2010), measuring the levels all S- compounds. Incubation of melon cubes with 5 mM L-methionine significantly enhanced the levels of both +4 and +1enriched compounds by comparable levels in the PI 414723 genotype but not in the ‘Dulce’ genotype (Figure 20). Incubation of melon cubes with 30 mM L-methionine significantly enhanced the levels of +4 and +1enriched compounds in both of the genotypes (Figure 20). In contrast, incubation of melon cubes with 5 mM KMBA significantly enhanced the levels of +4 enriched compounds but not of +1 enriched compounds in both of the genotypes (Figure 20).

Figure 20. Incubation of melon cubes with non-enriched precursors.

Volatile levels in melon cubes incubated with either 5 mM L-methionine, 30 mM L-methionine, 5 mM KMBA or DDW (control). Bars represent the sum levels of all +1 m/z or +4 m/z enriched compounds (see Fig. 7), in either ‘Dulce’ or PI 414723 genotypes. Values represent the mean of three biological repeats ± SEM. Asterisks indicate treatments significantly different from the control incubation with DDW (purple bars) analyzed by paired t-tests (α<0.05). Acids were administered to the melons for 6 h, and volatiles were analyzed by SPME-GC-MS.

Soluble protein extracts from ripe melon fruits displayed L-methionine aminotransferase enzymatic activity I have previously shown that melon crude extracts possess L-methionine aminotransferase (MetAT) enzymatic activity that produced KMBA from an L-methionine substrate when α-keto glutarate was used as a co-substrate (Gonda et al., 2010). Here, I checked three commercially available α-keto acids as a possible co-substrate for L-methionine aminotransferase enzyme. Figure 21 shows that melon enzymes (desalted crude extracts) can utilize α-keto glutarate (derived from L-glutamate), oxaloacetic acid (derived from L-aspartate) and glyoxylate (derived from glycine) for the production of KMBA from L-methionine. However, the highest L- methionine aminotransferase activity was observed when glyoxylate was used as a co- substrate (Figure 21b). The production of KMBA in this reaction was further validated in an assay analyzed by GC-MS after derivatization with TMS (Figure 21c).

Figure 21. Methionine aminotransferase activities in soluble protein extracts derived from ripe melon fruit. (a) Scheme of L-methionine aminotransferase enzymatic activity. In (b) and (c) L-Methionine aminotransferase enzymatic activities in soluble protein extracts derived from melon fruit. (b) About 25 µg of desalted proteins were assayed with 10 mM α-keto acid and

3 0.1μM H3-[methyl]-L-methionine. The reactions were incubated overnight at 30°C. Then, samples were acidified, and the radiolabeled product was extracted with ethyl acetate. The organic phase was transferred into liquid scintillation and analyzed in a β-counter. Results are the means of two repeats ± SEM. (c) Desalted protein extracts were assayed as above but with 5 mM L-methionine and 5 mM glyoxylate. The KMBA product was analyzed by GC-MS after derivatization with BSTFA. Each chromatogram represents at least four similar replicates.

CmMetAT encodes a methionine aminotransferase (MetAT) enzyme In order to identify a gene encoding for the MetAT enzyme, I used an RNA-seq database of ripening melon fruits for ‘Dulce’ and PI 414723 genotypes that is available in the lab. I looked for aminotransferases whose expression increases in the ripe fruit. I found three aminotransferase genes whose expression was enhanced in the ripe fruit stage. One was a branched-chain amino-acid aminotransferase (CmBCAT1, Gonda et al., 2010), another one was an aromatic amino-acid aminotransferase (CmArAT1, Gonda et al., 2010), and a third one was the unigene DV633223 (later termed CmMetAT). The expression pattern of CmMetAT is shown in Figure 22.

Figure 22. Relative expression of CmMetAT. Expression of CmMetAT measured by 454 pyrosequencing in PI 414723 and ‘Dulce’ genotypes at various fruit development stages both in rind and flesh. The data are from a database available in the lab. DAA – days after anthesis.

According to Pfam, CmMetAT belongs to the aminotransferase class I and II family, and according to the Panther classification system, it belongs to subgroup I aminotransferase related family. When the amino-acid sequence was checked in the NCBI Blastp algorithm, the most homologous genes were either unknown or hypothetical proteins from plants, but a closer look at more distant revealed a bacterial methionine aminotransferase. The AA sequence contains the catalytic lysine residue (Lys293) that forms an internal aldimine bond with the PLP co-factor. The subcellular localization of the protein is predicted to be in the mitochondria according to predotar (probability = 0.57), MitoProt II (probability = 0.92), iPSORT, and TargetP (probability = 0.838), and in the chloroplast according to WoLF PSORT. The predicted presequence length by TargetP is 70 aa, and by MitoProt II is 35 aa. I decided to construct a form of the protein truncated between position 28 and 29, since two Ser residues in positions 30 and 31 are conserved in the primary structure of many aminotransferases from plants (Cañas et al., 2008). The truncated CmMetAT sequence was cloned into a pET21a vector attached to 6×His-tag at its C-terminus, and purified on a Ni-NTA column using a sodium phosphate buffer containing 250 mM imidazole for elution. The purified enzyme was resolved on SDS-PAGE (Figure 23a), and its calculated molecular mass was found to be 55760 Da. This is in agreement with the predicted mass of the recombinant protein (by Expasy) that is 47492 Da. The gene product was assayed for L-methionine aminotransferase enzymatic 3 activity in three different assays. In an assay with L-[ H3-methyl]-methionine (specific activity: 0.01 mCi/mmol), I found that the first two elution fractions that also contained the recombinant enzyme (Figure 23a) showed the highest methionine aminotransferase enzymatic activity (Figure 23b).

Figure 23. Analyses of the recombinant CmMetAT.

(a) Purification and molecular mass identification of the recombinant CmMetAT. Proteins from E. coli harboring the 6xHis-tagged CmMetAT construct were cleaned by Ni-affinity chromatography and eluted with a sodium phosphate buffer containing 250 mM imidazole. The elution fractions, as well as the flow-thru (FT) and wash (W) fractions, were resolved on an CmMetAT SDS-PAGE with Coomassie blue staining and compared to the same fractions of proteins from E. coli harboring an empty pET21a vector. (b) L-Methionine aminotransferase enzymatic activities of the recombinant CmMetAT enzyme. About 3 µg of recombinant CmMetAT 3 (elution fraction 1) were assayed with 10 μM H3-[methyl]-L- methionine and 10 mM glyoxylate. The reactions were incubated overnight at 30°C. After incubation, the samples were acidified, and the radiolabeled product was extracted with ethyl acetate. The organic phase was transferred into liquid scintillation and analyzed in a scintillation counter. Results are the means of two repeats ± SEM. (c) L-Methionine aminotransferase enzymatic activity of the recombinant CmMetAT enzyme assayed with 5 mM L-methionine and 5 mM glyoxylate as a co-substrate. KMBA was analyzed by GC-MS after derivatization with BSTFA. Each chromatogram represents at least four similar replicates.

In a manner similar to the assays with the melon cell-free extracts, the production of KMBA in this reaction was validated in an assay analyzed by GC-MS after derivatization with TMS (Figure 23c). Because some α-keto acids are not commercially available, I used the reversible nature of aminotransferase enzymes to test whether CmMetAT can utilize more substrates. For this, I performed assays with KMBA and various amino acids as substrates and looked for the amino product, L-methionine, in an LC-TOF-MS. The amino acid that was the most efficient co-substrate for L-methionine production from KMBA by CmMetAT was, by far, L-glutamine (Figures 24a and 25).

Figure 24. L-methionine aminotransferase enzymatic activity of CmMetAT analyzed by LC-TOF-MS. (a) L-methionine aminotransferase enzymatic activity of CmMetAT. Purified CmMetAT was assayed for L-methionine aminotransferase enzymatic activity in the presence of 20 mM sodium phosphate pH 8.5, 2 mM KMBA, 2 mM L-glutamine and 0.2 mM PLP at 30°C in an overnight incubation. Then, the reactions were stopped by heating to 100°C for 5 min, chilled, diluted by 1:100 with 0.2 mM acetic acid, and 1 µl was injected into the LC-TOF. L-methionine product was identified by comparison with an authentic standard and by exact mass calculation. (b) Mass spectrum of the L-methionine generated in the assay display a typical natural isotopic distribution pattern of [M+H]+.

Figure 25. CmMetAT substrate preferences. L-methionine aminotransferase enzymatic activity of CmMetAT with different amino substrates. Purified CmMetAT was assayed for L-methionine aminotransferase enzymatic activity in the presence of 20 mM sodium phosphate pH 8.5, 2 mM KMBA, 2 mM amino-acid and 0.2 mM PLP at 30°C for overnight incubation. Then, the reactions were stopped by heating to 100°C for 5 min, chilled, diluted by 1:100 with 0.2 mM acetic acid, and 1 µl was injected into the LC-TOF.

Thioesters production in melon fruits

Soluble protein extracts from mature melon fruits displayed methanethiol acetyl transferase enzymatic activity Melon fruits contain five different thioesters (Table 3). Although present in many fruits and other food products, their biosynthesis has rarely been studied. I tested the hypothesis that thioesters (S-esters) are biosynthesized in similar way to the production of O-ester (through the activity of alcohol acyl transferase (AAT) enzyme, Figure 5, bottom), but by utilizing thiol instead of alcohol as a substrate (Figure 5, top). Since melon fruits contain only S-methyl thioesters, I tested methanethiol as the thiol substrate. I initially checked only acetyl coenzyme A as the acyl-CoA substrate. Since the production of thioesters from methanethiol and acyl-CoA could be spontaneous, the chemical production of S-methyl thioacetate from methanethiol and acetyl-CoA was tested at various pH values. Figure 26 shows that S-methyl thioacetate’s spontaneous

71 production is pH dependent, with the highest value between pH 9.5 to 10.5, while at pH 7.5, the optimal pH for AAT enzyme activity (Yahyaoui et al., 2002), its production is much slower.

Figure 26. Chemical conversion of methanethiol and acetyl Coenzyme A (Co-A) to S-methyl thioacetate. Single ion chromatograms of m/z=90 depict S-methyl thioacetate. Sodium methanethiolate and acetyl Co-A were incubated for two hours at various pH conditions. The production of S-methyl thioacetate was monitored by SPME-GC-MS.

The synthesis of S-methyl thioacetate by cell-free extracts from melon flesh was found to be much faster than its spontaneous chemical production (measured without an enzyme or using a heat-inactivated enzyme), indicating methanethiol acetyl transferase enzymatic activity in the melon fruit flesh (Figure 27).

Figure 27. Methanethiol acetyl transferase enzymatic activities in soluble protein extracts derived from ripe melon fruit. Single ion chromatograms of m/z = 90 depict S-methyl thioacetate. Each enzymatic assay contained 1 mM sodium methanethiolate and 0.2 mM acetyl Co- A in 50 mM Bis- Tris buffer pH 7.0. The production of S-methyl thioacetate was monitored by SPME-GC-MS. Colored chromatograms depict the assay with their color in the legend. Each chromatogram represents at least six similar replicates from six independent experiments.

CmThAT1 encodes a methanethiol acyl-transferase enzyme responsible for the accumulation of S-methyl thioacetate in melon fruit flesh To identify the acyl transferase gene that is in charge of the production of S-methyl thioacetate, I checked in the PI 414723 × ‘Dulce’ RIL population for genes whose

72 expression levels correlate with the levels of S-methyl thioacetate. The gene whose expression profile showed the highest correlation with S-methyl thioacetate levels is MELO3C024190 (termed CmThAT1, for Thiol Acyl Transferase), annotated as acetyl- CoA acetyl transferase, cytosolic 1 [Arabidopsis thaliana] with a correlation coefficient (r) of 0.59 (Pearson correlation). CmThAT1 belongs to the Thiolase-like superfamily, according to Pfam, and to the mitochondrial acetyl-CoA acetyl transferase subfamily according to the Panther Classification System. However, it is only distantly related to the four alcohol acyltransferase characterized from melon (CmAAT1-4, El-Sharkawy et al., 2005), with AA identities of 6%, 7% 11% and 2% respectively. The resulted protein is not predicted to localized to any organelle by most programs tested (only iPSORT predicted chloroplast localization). Blastp reveals that its closest orthologous are predicted proteins from plants annotated as acetyl CoA acetyl transferase. When Blastp against human proteins, the best hit is 2IB7_A with 50% AA identity. The recombinant enzyme was kinetically characterized and its crystal structure was determined as a homotetramer (Haapalainen et al., 2007). 2IB7uses acetyl CoA as substrate that is condensed with a free thiol of a cysteine residue in the enzyme, then condensed with another acetyl CoA to be released as acetoacetyl CoA (Haapalainen et al., 2007). CmThAT1 was cloned to the expression vector pET21a (ampicillin resistance selection) with and without a 6×His tag at its C-terminus end. The construct without the tag was transformed to E. coli BL21 (DE3) to be tested for the accumulation of the thioester product in the headspace of the bacteria. The bacteria were then transformed with the vector pBK-CMV (kanamycin resistance selection) carrying the CmMGL gene for the production of the methanethiol substrate. Due to its function in all living cells, the acetyl CoA substrate was hypothesized to be present in the bacteria, and no additional enzyme was needed for its production. Then, bacteria harboring the two plasmids were grown in an LB medium, the proteins production was induced by the addition of IPTG, and the bacteria were transferred to closed vials. The headspace of the bacteria was analyzed with SPME-GC-MS in comparison to control bacteria harboring an empty pET21a plasmid and a pBK-CMV carrying CmMGL plasmid. Bacteria expressing both CmThAT1 and CmMGL produced significantly higher amounts of S-methyl thioacetate than bacteria harboring empty pET21a and CmMGL in pBK-CMV (Figure 28a). In addition, the bacteria harboring the two genes produced higher amounts of S-methyl propanethioate than the control bacteria (Figure 28b), but four times lower levels than

73 the levels of S-methyl thioacetate. Interestingly, when the bacteria were supplied with α-keto butyrate (the precursor for propanoyl Co-A, the substrate for S-methyl propanethioate), even more S-methyl propanethioate was produced (Figure 28c), but the levels of S-methyl thioacetate production were not changed (Figure 28c, inset). The production of other thioesters was not detected. Moreover, when supplied with branched-chain α-keto acids that are precursors for branched-chain thioesters present in the melon, the bacteria also did not produce these thioesters.

Figure 28. Production of thioesters by bacteria harboring CmThAT1 and CmMGL genes. (a) Single ion chromatograms of m/z=90 depict S-methyl thioacetate production in bacteria expressing either CmThAT1 or empty pET21a simultaneously with CmMGL. (b) Single ion chromatograms of m/z=104 depict S-methyl propanethioate production in the same bacteria. (c) Single ion chromatograms of m/z=104 depict S-methyl propanethioate production in the same bacteria with and without the addition of α-KB. Inset depicts the effect of α-KB addition on the production of S- methyl thioacetate. Bacteria were grown in an LB medium, enzymes production was induced by IPTG, and α-KB was added to the vials as indicated. The headspace of each vial was analyzed by SPME-GC-MS. Each chromatogram represents at least four similar replicates from two independent experiments. α-KB: α-keto butyrate.

The constructs with the 6×His tag at the C-terminus (for both sequences from ‘Dulce’ and PI 414723 genotypes) were transformed to E. coli BL21 (DE3) pLys, and the

74 purified recombinant enzymes were checked for in vitro methanethiol acetyl transferase enzymatic activities. CmThAT1 enzymes, both from the ‘Dulce’ genotype and the PI 414723 genotype, showed methanethiol acetyl transferase enzymatic activity (Figure 29) but lacked alcohol acetyl transferase enzymatic activity (tested for the production of propyl acetate).

Figure 29. Methanethiol acetyl transferase enzymatic activities of the recombinant CmThAT1 enzymes. Single ion chromatograms of m/z=90 depict S-methyl thioacetate. Each enzymatic assay contained about 3 μg enzyme, 1 mM sodium methanethiolate, and 0.2 mM acetyl Co-A in 50 mM Bis-Tris buffer pH 7.0. Production of S-methyl thioacetate was monitored by SPME-GC-MS. Each chromatogram represents at least six similar replicates from three independent experiments.

The degradation of L-phenylalanine en route to aroma volatiles in melon fruits The levels of 29 L-phenylalanine-derived aroma compounds in the PI 414723 × ‘Dulce’ RIL population were quantified (using SPME-GC-MS) and analyzed by a multivariate Pearson correlation analysis (Table 7, significant correlations only). The compounds were clustered based on the pairwise correlations (Figure 30). Two major clusters were observed, one containing aldehydes and alcohols, as well as the esters, benzyl 2- methylpropanoate and benzyl butanoate (r values vary from 0.3 to 0.72). The second cluster contains acetate esters (r values vary from 0.54 to 0.83), which are negatively correlated with the aldehydes (r values vary from −0.15 to −0.45). A high correlation (positive correlation, hereafter) was also observed for the ethyl esters, ethyl phenyl acetate and ethyl (E)-cinnamate (r = 0.64), and to lesser extent, for ethyl benzoate (r = 0.28, r = 0.16, respectively). Significant correlations were also observed for methyl (E)- cinnamate and (E)-cinnamic acid (r = 0.53), and for methyl benzoate and benzoic acid (r = 0.29). The compounds eugenol and vanillin show high correlations with r = 0.43. The three benzyl esters, other than benzyl acetate, display high correlation values (r = 0.75, 0.52, 0.47). All three 1-phenyl ethyl compounds show high correlation values (r = 0.56, 0.31, 0.31), in contrast to the 2-phenyl ethyl compounds that do not display a clear correlation picture. To further understand the degradation of L-phenylalanine into aroma volatiles in melon fruit, I incubated melon 15 13 fruit rinds with N, C9-L-phenylalanine. As shown in Figure 31, the enrichment pattern was according to the length of the carbon chain attached to the aromatic ring, as expected (see Table 8). An exception was 2-phenylacetonitrile that displayed +8 and +9 m/z units enrichment patterns (Table 8).

Table 7. Pairwise correlation analysis of L-phenylalanine-derived volatiles in PI Variable By Variable r p-value 414723 × ‘Dulce’ RIL population (only significant correlations are shown). benzyl alcohol 1-phenethyl alcohol 0.49 <.0001 Variable By Variable r p-value benzyl alcohol 3-phenyl propanol 0.72 <.0001 benzaldehyde benzyl alcohol 0.48 <.0001 benzyl alcohol vanillin 0.41 <.0001 benzaldehyde benzyl butanoate 0.50 <.0001 benzyl alcohol (E)-cinnamaldehyde 0.34 <.001 benzaldehyde benzyl 2-methylpropanoate 0.56 <.0001 benzyl alcohol 2-phenylacetaldehyde 0.33 <.001 benzaldehyde (E)-cinnamaldehyde 0.65 <.0001 benzyl alcohol eugenol 0.27 <.01 benzaldehyde (E)-cinnamyl alcohol 0.42 <.0001 benzyl alcohol 2-phenetyl alcohol 0.27 <.01 benzaldehyde 2-phenylacetaldehyde 0.71 <.0001 benzyl alcohol benzyl propanoate 0.21 <.05 benzaldehyde 2-phenethyl acetate -0.39 <.0001 benzyl alcohol (E)-cinnamic acid 0.20 <.05 benzaldehyde 3-phenyl propanol 0.70 <.0001 benzyl alcohol ethyl benzoate 0.21 <.05 benzaldehyde benzyl acetate -0.35 <.001 benzyl alcohol 1-phenethyl acetate 0.23 <.05 benzaldehyde 1-phenethyl alcohol 0.34 <.001 benzyl butanoate benzyl 2-methylpropanoate 0.75 <.0001 benzaldehyde (E)-cinnamic acid 0.27 <.01 benzyl butanoate benzyl propanoate 0.52 <.0001 benzaldehyde (E)-cinnamyl acetate -0.21 <.05 benzyl butanoate (E)-cinnamyl alcohol 0.42 <.0001 benzaldehyde ethyl (E)-cinnamate 0.22 <.05 benzyl butanoate 1-phenethyl alcohol 0.39 <.0001 benzaldehyde 2-phenetyl alcohol -0.22 <.05 benzyl butanoate 3-phenyl propanol 0.65 <.0001 benzaldehyde 3-phenylpropyl acetate -0.24 <.05 benzyl butanoate (E)-cinnamaldehyde 0.33 <.001 2-phenylacetonitrile eugenol 0.24 <.05 benzyl butanoate ethyl benzoate 0.33 <.001 2-phenylacetonitrile methyl vanillate 0.23 <.05 benzyl butanoate 2-phenylacetaldehyde 0.35 <.001 2-phenylacetonitrile 2-phenylacetaldehyde 0.23 <.05 benzyl butanoate vanillin 0.26 <.05 benzoic acid 1-phenethyl alcohol 0.48 <.0001 benzyl 2-methylpropanoate benzyl propanoate 0.47 <.0001 benzoic acid 1-phenylethanone 0.39 <.0001 benzyl 2-methylpropanoate (E)-cinnamaldehyde 0.46 <.0001 benzoic acid methyl benzoate 0.29 <.01 benzyl 2-methylpropanoate (E)-cinnamyl alcohol 0.46 <.0001 benzoic acid benzyl acetate -0.22 <.05 benzyl 2-methylpropanoate 3-phenyl propanol 0.59 <.0001 benzyl acetate benzyl propanoate 0.58 <.0001 benzyl 2-methylpropanoate 2-phenylacetaldehyde 0.37 <.0005 benzyl acetate (E)-cinnamyl acetate 0.74 <.0001 benzyl 2-methylpropanoate 1-phenethyl alcohol 0.35 <.001 benzyl acetate 2-phenylacetaldehyde -0.44 <.0001 benzyl 2-methylpropanoate ethyl benzoate 0.24 <.05 benzyl acetate 1-phenethyl acetate 0.66 <.0001 benzyl 2-methylpropanoate 2-phenyl nitro ethane 0.24 <.05 benzyl acetate 2-phenethyl acetate 0.73 <.0001 benzyl 2-methylpropanoate vanillin 0.24 <.05 benzyl acetate 1-phenylethanone 0.40 <.0001 benzyl propanoate ethyl benzoate 0.40 <.0001 benzyl acetate 3-phenylpropyl acetate 0.83 <.0001 benzyl propanoate 3-phenylpropyl acetate 0.46 <.0001 benzyl acetate ethyl benzoate 0.28 <.01 benzyl propanoate 1-phenethyl acetate 0.36 <.0005 benzyl acetate vanillin 0.24 <.05 benzyl propanoate (E)-cinnamyl acetate 0.35 <.001 benzyl alcohol benzyl butanoate 0.60 <.0001 benzyl propanoate 2-phenethyl acetate 0.26 <.01 benzyl alcohol benzyl 2-methylpropanoate 0.50 <.0001 benzyl propanoate 1-phenylethanone 0.28 <.01 benzyl alcohol (E)-cinnamyl alcohol 0.60 <.0001 benzyl propanoate (E)-cinnamyl alcohol 0.20 <.05 - 77 -

Variable By Variable r p-value Variable By Variable r p-value benzyl propanoate 2-phenyl nitro ethane 0.20 <.05 eugenol 2-phenyl nitro ethane 0.29 <.01 (E)-cinnamaldehyde (E)-cinnamyl alcohol 0.67 <.0001 eugenol 3-phenyl propanol 0.25 <.05 (E)-cinnamaldehyde 2-phenylacetaldehyde 0.57 <.0001 methyl benzoate 1-phenylethanone 0.49 <.0001 (E)-cinnamaldehyde 3-phenyl propanol 0.47 <.0001 methyl benzoate methyl (E)-cinnamate 0.32 <.01 (E)-cinnamaldehyde 1-phenethyl alcohol 0.37 <.0005 methyl benzoate methyl vanillate 0.22 <.05 (E)-cinnamaldehyde eugenol 0.21 <.05 methyl benzoate 2-phenetyl alcohol 0.22 <.05 (E)-cinnamaldehyde 2-phenethyl acetate -0.21 <.05 methyl (E)-cinnamate 1-phenylethanone 0.35 <.001 (E)-cinnamic acid methyl (E)-cinnamate 0.53 <.0001 methyl vanillate 2-phenethyl acetate 0.30 <.01 (E)-cinnamic acid 1-phenethyl alcohol 0.35 <.0005 methyl vanillate 3-phenylpropyl acetate 0.27 <.01 (E)-cinnamic acid ethyl (E)-cinnamate 0.31 <.01 methyl vanillate 2-phenyl nitro ethane 0.22 <.05 (E)-cinnamic acid 1-phenylethanone 0.27 <.01 methyl vanillate vanillin 0.25 <.05 (E)-cinnamic acid 3-phenyl propanol 0.33 <.01 1-phenethyl acetate 2-phenethyl acetate 0.54 <.0001 (E)-cinnamyl acetate 1-phenethyl acetate 0.63 <.0001 1-phenethyl acetate 1-phenylethanone 0.56 <.0001 (E)-cinnamyl acetate 2-phenethyl acetate 0.60 <.0001 1-phenethyl acetate 3-phenylpropyl acetate 0.76 <.0001 (E)-cinnamyl acetate 3-phenylpropyl acetate 0.75 <.0001 1-phenethyl acetate 1-phenethyl alcohol 0.31 <.01 (E)-cinnamyl acetate 2-phenylacetaldehyde -0.30 <.01 1-phenethyl acetate vanillin 0.29 <.01 (E)-cinnamyl acetate 1-phenylethanone 0.28 <.01 2-phenethyl acetate 1-phenylethanone 0.48 <.0001 (E)-cinnamyl acetate vanillin 0.31 <.01 2-phenethyl acetate 3-phenylpropyl acetate 0.68 <.0001 (E)-cinnamyl acetate (E)-cinnamyl alcohol 0.23 <.05 2-phenethyl acetate vanillin 0.29 <.01 (E)-cinnamyl acetate ethyl benzoate 0.22 <.05 2-phenethyl acetate 2-phenetyl alcohol 0.24 <.05 (E)-cinnamyl acetate methyl vanillate 0.22 <.05 2-phenethyl acetate 3-phenyl propanol -0.23 <.05 (E)-cinnamyl alcohol 3-phenyl propanol 0.49 <.0001 1-phenethyl alcohol 3-phenyl propanol 0.41 <.0001 (E)-cinnamyl alcohol 2-phenylacetaldehyde 0.33 <.001 1-phenethyl alcohol 1-phenylethanone 0.31 <.01 (E)-cinnamyl alcohol 1-phenethyl alcohol 0.30 <.01 1-phenethyl alcohol 2-phenetyl alcohol 0.22 <.05 (E)-cinnamyl alcohol eugenol 0.25 <.05 2-phenetyl alcohol 1-phenylethanone 0.26 <.01 ethyl benzoate 1-phenethyl acetate 0.41 <.0001 2-phenetyl alcohol vanillin 0.30 <.01 ethyl benzoate 3-phenylpropyl acetate 0.46 <.0001 2-phenylacetaldehyde 2-phenethyl acetate -0.45 <.0001 ethyl benzoate 3-phenyl propanol 0.40 <.0001 2-phenylacetaldehyde 3-phenyl propanol 0.50 <.0001 ethyl benzoate ethyl phenyl acetate 0.28 <.01 2-phenylacetaldehyde 1-phenethyl alcohol 0.34 <.001 ethyl benzoate 1-phenylethanone 0.21 <.05 2-phenylacetaldehyde 3-phenylpropyl acetate -0.34 <.001 ethyl (E)-cinnamate ethyl phenyl acetate 0.64 <.0001 2-phenylacetaldehyde 1-phenethyl acetate -0.23 <.05 ethyl (E)-cinnamate methyl (E)-cinnamate 0.28 <.01 1-phenylethanone 3-phenylpropyl acetate 0.50 <.0001 ethyl (E)-cinnamate 3-phenyl propanol 0.23 <.05 1-phenylethanone vanillin 0.41 <.0001 ethyl phenyl acetate methyl benzoate 0.22 <.05 3-phenyl propanol vanillin 0.28 <.01 eugenol vanillin 0.43 <.0001 3-phenylpropyl acetate vanillin 0.31 <.01 eugenol 2-phenylacetaldehyde 0.28 <.01 Analysis performed in JMP® software. r – correlation coefficient (Pearson). - 78 -

Figure 30. Multivariate correlation analysis of L-phenylalanine-derived aroma volatiles in the fruit flesh throughout the PI 414723 × ‘Dulce’ RIL population. A multivariate correlation analysis was carried out using the JMP® program for L-phenylalanine-derived aroma volatiles. Result are displayed in "cluster the correlations" view. Analysis was performed with the averages of each line (n=3). Red squares – positive correlation, green squares – negative correlation.

- 79 -

Figure 31. Incorporation of 13C9-L-phenylalanine into various aroma volatiles in melon fruit rinds. For each volatiles group, the enrichment pattern is shown in blue. Ripe melon rind strips were

15 13 incubated with 5 mM N, C9-L-phenylalanine for six hours at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS. PAL – phenylalanine ammonia lyase; AADC – aromatic amino acids decarboxylase; ArAT – aromatic amino acids aminotransferase.

15 13 Table 8. Enrichment levels after incubation of melon cubes with N, C9-L-phenylalanine. Method of Compound name m/z PI 414723 ‘Dulce’ identification

benzaldehyde +7 15.9 ± 3.2 43 ± 1.7 MS,KI,AS

benzyl alcohol +7 10.1 ± 1.9 41.8 ± 1.3 MS,KI,AS

benzyl acetate +7 30.5 ± 4.7 47.8 ± 2.5 MS,KI

benzyl propanoate +7 25.1 ± 2.8 44 ± 2.2 MS,KI

benzyl butanoate +7 30.3 ± 3.7 43.8 ± 2.2 MS,KI

benzyl 2/3-methyl butanoate +7 28.3 ± 4.3 37.5 ± 2.3 MS,KI

benzyl 2-methylpropanoate +7 29.1 ± 3.8 36 ± 3.8 MS,KI

methyl benzoate +7 46.5 ± 0.9 35.8 ± 3.6 MS,KI

ethyl benzoate +7 21.3 ± 3.9 16.6 ± 2.7 MS,KI

2-phenylacetaldehyde +8 65 ± 4.5 47.4 ± 2.5 MS,KI,AS

2-phenethyl alcohol +8 43.6 ± 5.6 28.5 ± 2.1 MS,KI,AS

2-phenethyl acetate +8 74.9 ± 2.9 69.1 ± 0.8 MS,KI,AS

ethyl phenyl acetate +8 71.4 ± 6.5 n.d. MS,KI

1-phenylethanone +8 47.8 ± 4.2 36.8 ± 2.5 MS,KI

1-phenylethanol +8 30 ± 2 12.1 ± 1.9 MS,KI

1-phenethyl acetate +8 30.2 ± 6 n.d. MS,KI

2-phenylacetonitrile +8 +8 50.2 ± 3.5 44.7 ± 1.6 MS,KI

2-phenylacetonitrile +9 +9 31.4 ± 2.6 31.8 ± 0.7 MS,KI

3-phenyl propanol +9 42.7 ± 3.5 34.7 ± 2.5 MS,KI

3-phenylpropyl acetate +9 40.4 ± 10.6 47.4 ± 2.7 MS,KI

(E)-cinnamaldehyde +9 n.d. 47.7 ± 3 MS,KI,AS

(E)-cinnamyl alcohol +9 n.d. 41.4 ± 4.2 MS,AS,KI

(E)-cinnamyl acetate +9 n.d. 49.8 ± 4.9 MS,AS,KI

methyl (E)-cinnamate +9 64.5 (n=1) 56.4 ± 8 (n=2) MS,AS

Values are: 13C / (13C + 12C), m/z – enrichment pattern, n.d. – compound not detected; MS – mass spectrum; AS – comparison with an authentic standard.

Soluble protein extracts from ripe melon fruits displayed (E)-cinnamaldehyde synthesis activities I focused next on the mechanisms of (E)-cinnamaldehyde production, an important constituent of cinnamon aroma, but whose biosynthesis is poorly understood. The enrichment pattern of +9 m/z units (Figure 32) strengthen the hypothesis that (E)-cinnamaldehyde is produced from (E)-cinnamic acid, mediated by the two enzymes (E)-cinnamic acid:Co-A ligase (CNL) and cinnamoyl Co-A reductase (CCR) (Figure 33a).

13 Figure 32. Incorporation of C9-L-phenylalanine into (E)-cinnamaldehyde in melon rinds. The upper spectrum represents (E)-cinnamaldehyde extracted from melon rinds incubated with

12 5 mM C9-L-phenylalanine. The lower spectrum represents (E)-cinnamaldehyde extracted

13 12 from melon rinds incubated with 5 mM C9-L-phenylalanine. Black ions depict C9-(E)-

13 cinnamaldehyde and blue ions depict C9-(E)-cinnamaldehyde. Each spectrum shown is typical to at least three biological repeats. Ripe melon rind strips were incubated with the indicted solution for six hours at room temperature, and the volatiles were subsequently analyzed by SPME-GC-MS.

For this, I performed coupled enzymatic assays of these enzymes in cell-free extracts derived from ripe melon fruits. Figure 33 shows that soluble protein extracts from ripe melon fruits displayed (E)-cinnamaldehyde synthesis activities. Since the optimum pH of the two enzymes is probably quite different based on information obtained for the enzymes of other plants (pH 8.0 for the Petunia CNL, Klempien et al., 2012; pH 6.3 for Eucalyptus CCR, Goffner et al., 1994), I tested the coupled reactions at pH levels of 6.3, 7.0 and 8.0. The intermediate level of pH 7.0 was found to be optimal for the coupled assay (Figure 33c), readily demonstrating the ability of melon tissue to

82 biosynthesize (E)-cinnamaldehyde from (E)-cinnamic acid in the presence of Co-A, ATP, Mg++ and NADPH.

Figure 33. Enzymatic synthesis of (E)-cinnamaldehyde from (E)-cinnamic acid in a ripe melon cell-free extract. (a) Scheme of coupled reaction for (E)-cinnamaldehyde biosynthesis. In (b) and (c) GC-MS chromatograms indicating the enzymatic formation of (E)-cinnamaldehyde in cell-free extracts derived from ripe fruit. (b) The upper chromatogram depicts the full reaction. The lower chromatograms depict reactions with no enzyme, with heat-inactivated enzymes and without a Co-A substrate. (c) Top to bottom: the chromatograms depict the full reaction in pH 7.0, 8.0 and 6.3. Melon enzymes were extracted using a buffer containing 50 mM Bis-Tris propane pH 7.0, 5mM DTT and 10% (v/v) glycerol, and desalted on a P-6 column. Assay mixtures were incubated overnight at 30°C and analyzed by SPME-GC-MS. Each chromatogram represents four assays in two independent experiments.

CmCNL encodes a an (E)-cinnamic acid:coenzyme A ligase enzyme In order to find genes encoding (E)-cinnamic acid:coenzyme A ligase (CNL) enzyme, I searched the melon genome for genes showing high similarity to the Petunia hybrida CNL (PhCNL) sequence using tBlastn algorithm. Nine candidate genes were found with AA identities to PhCNL that are vary from 63% to 37% (Table 9). Table 9. Candidate genes encoding melon CNLs. % of AA Melon genome identities to Annotation (from melon genome website) accession number PhCNL

MELO3C025105 36% Medium-chain-fatty-acid--CoA ligase (Pseudomonas oleovorans) (uniprot_sprot:sp|Q00594|ALKK_PSEOL) MELO3C025111 84% Putative acyl-CoA synthetase YngI (Bacillus subtilis) (uniprot_sprot:sp|O31826|YNGI_BACSU) MELO3C017017 84% 2-succinylbenzoate--CoA ligase (Geobacillus sp. (strain WCH70)) (uniprot_sprot:sp|C5D6U5|MENE_GEOSW) MELO3C025110 84% Putative acyl-CoA synthetase YngI (Bacillus subtilis) (uniprot_sprot:sp|O31826|YNGI_BACSU) MELO3C017018 84% Putative acyl-CoA synthetase YngI (Bacillus subtilis) (uniprot_sprot:sp|O31826|YNGI_BACSU) MELO3C007967 84% Medium-chain-fatty-acid--CoA ligase (Pseudomonas oleovorans) (uniprot_sprot:sp|Q00594|ALKK_PSEOL) MELO3C006703 84% Medium-chain-fatty-acid--CoA ligase (Pseudomonas oleovorans) (uniprot_sprot:sp|Q00594|ALKK_PSEOL) MELO3C007966 88% Endochitinase (Pisum sativum PE=2 SV=1) (uniprot_sprot:sp|P36907|CHIX_PEA) MELO3C013735 64% Putative acyl-CoA synthetase YngI (Bacillus subtilis) (uniprot_sprot:sp|O31826|YNGI_BACSU) Interestingly, three of these genes (MELO3C025105/10/11) are adjacently located on chromosome 9. Of these nine genes only five show significant expression levels in some of the RIL population families. Next, I looked for positive correlations between the expression levels of the five candidate genes and levels of (E)-cinnamaldehyde and its derived volatiles (E)-cinnamyl alcohol, ethyl-(E)-cinnamate and (E)-cinnamyl acetate. Table 10. Pearson correlations coefficient (r) of CNL candidate genes expression with (E)-cinnamoyl CoA-derived volatiles in the RIL population (*α<0.05, **α<0.0001) MELO3C0 17017 7967 25110 6703 25111 (E)-cinnamaldehyde ns ns ns ns ns (E)-cinnamyl alcohol ns -0.23* ns ns ns (E)-ethyl cinnamate ns ns ns ns ns (E)-cinnamyl acetate ns -0.21* 0.38** ns ns

The only gene that showed significant positive correlation with one of the compounds was MELO3C025110 (later termed CmCNL), which its expression levels correlated with (E)-cinnamyl acetate with correlation coefficient of 0.38 (α=0.0001, Pearson correlation). Phylogenetic analysis of CmCNL with other functionally characterized CNLs and 4CLs, indicated that CmCNL is likely to act as a bona fida CNL enzyme and not a 4CL one (Figure 34).

Figure 34. Phylogenetic tree of plant CNLs and 4CLs. Amino acid sequences of representative CNL and 4CL plant enzymes that have been functionally validated were compared utilizing the "One click" mode of Phylogeny.fr program (http://www.phylogeny.fr). Red numbers indicate branch support values. CmCNL: Cucumis melo cinnamic acid:CoA ligase XP_008463174.1; HcCNL: Hypericum calycinum cinnamate:CoA ligase AFS60176.1; PhCNL: Petunia x hybrida cinnamic acid:CoA ligase AEO52693.1; AtBZO1: Arabidopsis thaliana benzoate-CoA ligase 1 NP_176763.1; Ph4CL1: Petunia x hybrida 4-coumarate:CoA ligase AEO52694.1; Nt4CL2: Nicotiana tabacum 4- coumarate-CoA ligase 2 O24146.1; Nt4CL1: Nicotiana tabacum 4-coumarate-CoA ligase 1 O24145.1; Ri4CL2: Rubus idaeus 4-coumarate:coA ligase 2 AF239686_1; At4CL1-5: Arabidopsis thaliana 4-coumarate CoA ligase 1-5 AAQ86588.1; AAQ86587.1; AAQ86589.1; AAQ86590.1; AAQ86591.1 Gm4CL1-4: Glycine max 4-coumarate:coenzyme A ligase 1-4 AF279267_1; AAC97600.1; AAC97599.1; CAC36095.1

CmCNL is predicted to be localized in the peroxisomes (certainty = 0.854) according to PSORT prediction. The full-length gene was extracted from cDNA derived from ripe melon fruit ('Dulce' cultivar) and cloned into pET21a expression vector with 6×His tag at its’ C-terminus. Then, the gene was heterologously expressed in E. coli and the recombinant enzyme was purified on Ni-NTA resin. CmCNL showed (E)-cinnamic acid:coenzyme A ligase activity in vitro that was analyzed by LC-TOF-MS (Figure 35a). In the presence of ATP and Mg++ CmCNL catalyzed the production of (E)- cinnamoyl coenzyme A from (E)-cinnamic acid and coenzyme A substrates. The cinnamoyl CoA product was identified by retention time and mass-spectrum characteristics against cinnamoyl-CoA standard produced by the Nicotiana tabacum 4CL (Beuerle and Pichersky, 2002). It shows a typical [M-H]- mass-spectrum of 896.1493 m/z ratio, and display natural isotopic distribution, both confirmed by “Find by Formula” algorithm (Figure 35b). Moreover, CmCNL was also able to catalyze the enzymatic conversion of ferulic and p-coumaric and acids to their respective acyl- CoA's at similar levels. The conversion of benzoic acid to benzoyl CoA was catalyzed albeit at lower levels and caffeic acid to caffeoyl CoA were formed at much lower level (not shown).

Figure 35. In vitro enzymatic production of cinnamoyl coenzyme A by the recombinant enzyme CmCNL analyzed by LC-MS.

Analyses of (E)-cinnamic acid:coenzyme A ligase enzyme activity catalyzed by CmCNL enzyme. (a) LC-MS extracted ion chromatograms (EIC=896.1493) represent the full reaction (upper) and the following controls: no enzyme, heat inactivated enzyme, without ATP and without Mg++. Lower chromatogram depict cinnamoyl coenzyme A standard produced by Nicotiana tobacco 4CL (Beuerle and Pichersky, 2002). Purified CmCNL was assayed for (E)- cinnamic acid:coenzyme A ligase activity in the presence of 20 mM potassium phosphate pH

8.0, 5 mM (E)-cinnamic acid, 0.1 mM coenzyme A, 5 mM ATP and 5 mM MgSO4 at 30°C in an overnight incubation. Then, the reactions were filtered and 1 µl was injected to the LC-TOF. Cinnamoyl coenzyme A was identified by comparison to a standard produced by Nicotiana tobacco 4CL and by exact mass calculation. (b) “Find by formula” analysis of the mass spectrum of the cinnamoyl coenzyme A generated in the assay display a typical natural isotopic distribution pattern of [M-H]-.

Discussion L-methionine and L-isoleucine degradation into volatile aroma compounds Melon varieties differ in their aroma, and this is reflected in the content of specific volatiles. I have compared the levels of L-methionine-derived volatiles in two contrasting melon varieties that differ in their aroma properties and utilized the differences in attempts to elucidate the respective biosynthetic pathways. Generally, the genotype PI 414723 is characterized by a higher abundance of sulfur-containing aroma volatiles (S-compounds) as compared to ‘Dulce,’ and most of the volatile thioesters were more abundant in PI 414723 than in ‘Dulce’ (Table 3). The only S-compound whose level was notably higher in ‘Dulce’ was 3-(methylthio) propyl acetate. Different volatiles profiles may have a distinct influence on the aroma of foods. This is a consequence of their concentration, their specific odor thresholds and their respective odor descriptor properties. No single compound is responsible for the unique aroma of melons. Moreover, the exact impact of each compound on melon aroma is difficult to determine. However, it is clear that thio-ether esters greatly contribute to the fruity aroma of melon fruits (Wyllie and Leach, 1992; Wyllie et al., 1994; Jordán et al., 2001). Conversely, some thioesters and sulfides clearly have unpleasant stale/garlic/sulfurous aroma notes (Wyllie et al., 1994; Berger et al., 1999) and negatively impact melon aroma and flavor. Considering all of the above, differential levels of sulfur volatiles displayed in the genotypes PI 414723 and ‘Dulce’ might partially explain the noticeable difference in the aroma and commercial acceptability of the genotypes (Danin-Poleg et al., 2000).

Incorporation of L-methionine into sulfur-containing aroma compounds Using incubation experiments with stable-isotope–labeled precursors, I outlined the possible degradation pathways of L-methionine into aroma volatiles in ripe melon fruits. The biosynthetic scheme illustrated in Figure 36, summarizes the results obtained 13 when C5-L-methionine was administered to ripe melon flesh cubes.

13 Figure 36. Schematic pathway illustrating the different enrichment patterns of C5-L- methionine during its incorporation into aroma volatiles. Red route (left): compounds enriched with +4 m/z units, probably arising through the action of L-methionine aminotransferase (Gonda et al., 2010); green route (middle left): compounds enriched with +1 m/z unit, (and +2), probably arising through a reaction with methanethiol, generated by the action of MGL. Black route (middle right): compounds enriched with +3 m/z units that bear a 2-methylbutyl side-chain. Purple route (right): compounds enriched with +3 m/z units that bear a propanoate group. Compounds enriched in more than one pattern are shown by overlaid colors.

Compounds enriched with +4 m/z units included exclusively C3-thio-ethers, which are probably biosynthesized via the transamination of L-methionine, a reaction that was demonstrated to occur in melon fruit (Figure 21bc, Gonda et al., 2010). Eleven S- compounds, including three C2-thio-ether esters, a thiol, two sulfides and five thioesters, were isotopically enriched with +1 m/z units (Figure 36, green box). When analyzing the mass spectra fragment enrichment of the above compounds, I concluded that their enrichment originated from the methylthio moiety of the original exogenous 13 C5-L-methionine (Figure 8). Moreover, the isotopic enrichment of the same 2 compounds with +3 m/z units upon incubation with H3-[methyl]-L-methionine (Figures 12 and 13) further supports that conclusion. Accordingly, the isotopic enrichment of +3 and +6 m/z units was observed for dimethyl disulfide and dimethyl 2 trisulfide when utilizing H3-[methyl]-L-methionine (Figures 12 and 13). Some microorganisms produce S-compounds via the demethiolation of methionine mediated by MGL enzyme activities (Dias and Weimer, 1998; Perpète et al., 2005). The enrichment of S-compounds only in their methylthio fraction upon incubation with either 13C or 2H-labeled L-methionine, (Figures 7, 12 and 13), together with the presence of MGL enzymatic activity in cell-free extracts derived from melon flesh (Figure 16b-d), suggests that MGL activity plays a crucial role in the formation of these S-compounds. The seemingly constant expression pattern of CmMGL in ‘Dulce’ (Figure 15) is apparently enough to support the enzymatic activity detected in the cell- free extracts’ assays from the ripe fruit (Figure 16b-d), as well as the labeling of volatiles from group 2. The enhanced MGL activity measured in PI 414723, as compared to ‘Dulce’ (Figure 16b), is probably due to the differential expression of CmMGL in the two genotypes (Figure 15). These differences probably account for the different levels of S-compounds accumulated in the two genotypes (Table 3). In spite of the low expression of CmMGL in ‘Dulce’ flesh, volatiles from group 2 (+1 m/z unit) were also enriched in ‘Dulce’ fruit tissues (Figure 7), although at lower rates than in PI 414723. This indicates that ‘Dulce’ flesh tissues do possess MGL activity as corroborated by the enzymatic analyses (Figure 16). The association between the levels of S-compounds from group 2 and CmMGL expression in the recombinant inbred lines (Figure 19a) strongly suggests that CmMGL is involved in the production of S- compounds from this group. No association was apparent between CmMGL expression levels and the levels of S-compounds of group 1 (Figure 19b). Nevertheless, I cannot exclude the possibility that other genes encoding more MGL enzymes play additional

90 roles in S-compounds’ formation in melons. To conclude, CmMGL is a good target for future breeding programs as being responsible for the formation of often undesirable aroma traits (Figure 19; Wyllie et al., 1994; Perry et al., 2009; Wang and Lin, 2014). Although SNP does not exist in the coding region of the enzyme, non-coding regions might bear such differences. SNPs in the gene promoter for instance can explain its’ varied expression in the RIL population and might be used as markers in marker- assisted selection (MAS) breeding.

Incorporation of L-methionine into L-isoleucine and L-isoleucine-derived aroma compounds 13 Enriched L-isoleucine was detected upon incubation with C5-L-methionine (Table 5). The enrichment pattern suggests that the L-isoleucine pool in melon fruit can be at least partially derived from the degradation of L-methionine by the action of MGL (see Figure 36). MGL activity releases α-ketobutyrate, which can then be metabolized into L-isoleucine via the classic L-isoleucine biosynthesis pathway (Singh 1999). This observation is also supported by the presence of L-isoleucine-derived aroma volatiles 13 upon incubation with C5-L-methionine enriched with +3 m/z units (Figure 36, black box). In Arabidopsis flowers and siliques, 13C-L-methionine was incorporated into 13C- L-isoleucine displaying an enrichment pattern of +4 m/z units (Joshi and Jander, 2009). Furthermore, the authors showed that an mgl knockdown mutant accumulates much lower quantities of enriched L-isoleucine than the wild type. Recently, Huang et al. (2014) showed that knockdown of the potato StMGL1 increased the methionine:isoleucine ratio in their free form and suggest that it might give rise to methionine-derived aroma volatiles. However, the role of MGL in L-isoleucine formation, as related to volatiles biosynthesis, has never been shown before. Using a GC-MS derivatization method, the enrichment detected here for L-isoleucine was of +3 m/z units (Table 5), but it was apparent only in mass fragments (ion 89 vs. 86 for 1TMS, and ion 161 vs. 158 for 2TMS). This apparent difference is probably due to the very low intensity of the molecular ion detected, likely caused by the electrospray ionization mass spectrometry methodology used. When considering that the ions used for the analyses exclude the carboxylic carbon, I assumed that the enrichment was originally a +4 m/z pattern. This is in agreement with the LC-MS data that clearly shows enrichment of +4 m/z units for L-isoleucine. The observed enrichment of the L-isoleucine-derived volatiles was +3 m/z units as well,

91 both in fragments and in the molecular ion (when apparent in the fragmentation mass spectrum). Since in most cases, volatile biosynthesis involves decarboxylation, there is accordance between the findings in these two enrichment experiments. The enrichment patterns of free L-isoleucine (Table 5) and L-isoleucine-derived volatiles (Figures 8 and 9) suggest that the MGL pathway is active in the ripe fruit. The reason that the enrichment levels of L-isoleucine are lower than those of its derived volatiles is probably due to existence of different sub-cellular pools. However, I cannot exclude the possibility that some 2-methylbutyl volatile derivatives are biosynthesized directly from keto-isoleucine, bypassing L-isoleucine, as suggested by Kochevenko et al. (2012).

Differential incorporation of L-methionine, L-isoleucine and L-leucine into non- sulfur volatiles I found substantial isotopic enrichment in volatile propanoate esters (e.g., ethyl propanoate, 2-methylbutyl propanoate) with +3 m/z units (Figure 36, purple box), but interestingly, propanol-derived volatiles (e.g. propyl acetate) were not enriched. The propanoate derivatives were enriched with +3 m/z units, both when 13C-labeled L-methionine (Figure 9, top) and when 13C-labeled L-isoleucine (Figure 9, bottom) were administered. In contrast, when 13C-labeled L-leucine was administrated, propanoate esters were not enriched at all. This evidence strongly suggests that L-isoleucine, but not L-leucine, is a bona fida precursor of propanoate volatiles. While the conversion of fatty acids to C6 and C9 volatiles is documented in other plants (Galliard and Phillips, 1976; Phillips and Galliard, 1978; Anthon and Barrett, 2003), the origin of shorter straight-chain volatiles (i.e., C3 and C4) is less obvious. The observation that only propanoates were labeled but not propanol-derived volatile esters, suggest that the propanoate moiety in propanoate volatile esters is a result of L- isoleucine degradation into propanoyl CoA that is subsequently incorporated into the propanoate ester. The finding that in melon, L-isoleucine is a precursor for propanoate esters can offer new ideas about their biosynthetic origin and the origin of other short straight-chain aliphatic volatiles. More experiments are needed to better understand this biosynthetic pathway. Furthermore, the incorporation of L-leucine into many acetate esters (Figures 10 and 11) is probably a result of its catabolism into acetyl CoA. Interestingly, the observation that L-leucine, but not L-isoleucine, was incorporated into volatile acetates is in contrast to what have been found in apple fruits (Matich and

Rowan, 2007). These results strengthen the observation that the metabolism of these two branched-chain amino-acids is differentially regulated and varies among fruits. L-methionine is a precursor of S-adenosyl-L-methionine (SAM), a key metabolite that controls the levels of ethylene, polyamines and biotin and also serves as the primary methyl group donor in the biosynthesis of many methylated natural products such as flavonoids and phenylpropenes (Amir, 2010). Methyl benzoate was substantially enriched in PI 414723 when 13C-L-methionine was administered (Figure 7), and showed a relatively high correlation with the levels of benzoic acid within the RIL population families (Figure 30, Table 7). This might indicate that methyl benzoate is biosynthesized in melon fruits in a similar manner to its biosynthesis in Antirrhinum majus involving the activity of an S-adenosyl-L-methionine:benzoic acid carboxyl methyltransferase (BAMT) enzyme (Murfitt et al., 2000) (see below for further discussion). However, no enrichment was found in other volatile methyl esters, such as methyl hexanoate and methyl butanoate in 13C-L-methionine-fed fruit slices (methyl- (E)-cinnamate was not detected in these experiments). It could be that compartmentalization barriers or different pool sizes caused the differential enrichment of methylated compounds. Still, it could also be that other methyl donors, such as methanol (Pritchett and Metcalf, 2005) or 5-methyl-tetrahydrofolate (Stead et al., 2006), are involved in the formation of volatile aliphatic methyl esters in melon fruit. While ethanol, propanol and many other alcohol substrates were tested as substrates for melon AAT enzymes (Shalit et al., 2001; Yahyaoui et al., 2002; El-Sharkawy et al., 2005; Luchetta et al., 2008) as well as for other plant AATs (e.g. Aharoni et al. 2000), much less attention was dedicated for the production of straight- and branched-chain methyl esters. Aharoni et al. (2000) showed that AAT from strawberry can utilizes methanol for the production of methyl acetate. Suzuki et al. (2014) showed that crude enzyme extracts of snake fruit (Salacca zalacca, Arecaceae) can enzymatically produce methyl hexanoate from methanol and hexanoyl CoA. Wang and De Luca (2005) isolated an anthraniloyl-CoA:methanol acyltransferase (AMAT) from fruits of Concord grapes producing methyl anthranilate. The recombinant AMAT enzyme efficiently produced also methyl benzoate and methyl acetate in vitro (Wang and De Luca, 2005). Further in vitro analyses of the melon AATs for their ability to use methanol as a substrate, together with incubation of melon slices with labeled methanol might give a better understanding of the production of these methyl esters in melon fruits.

The role of L-methionine aminotransferase in melon aroma production

KMBA is a true intermediate in the production of C3-thio-ethers in melon fruit With the analysis of all S-compounds in the incubation experiments with non-labeled precursors, I was able to clearly distinguish between two major methionine catabolic pathways into volatiles. Incubation with L-methionine significantly enhanced the total levels of volatiles from both group 1 (enriched with +4 m/z units) and group 2 (enriched with +1 m/z unit) in both genotypes, when administrated at a 30 mM concentration, and in PI 414723 only when administrated at a 5 mM concentration (Figure 20). The fact that incubation with 5 mM KMBA significantly enhanced the total levels of C3-thio- ethers more than fivefold, as compared to incubation with 5 mM L-methionine for both genotypes (Figure 20), suggests that KMBA is located downstream in the biosynthetic pathway of these volatiles. Moreover, the fact that incubation with 5mM KMBA did not show significant enhancement in the total levels of the other S-compounds (Figure 20) (as was observed for incubation with 5 mM L-methionine) makes it unlikely that the supplemented KMBA was reverted to L-methionine to generate aroma volatiles from group 1. The enrichment pattern observed for the C3-thio-ethers (+4 m/z units) also prompted me to reject the possibility that the KMBA intermediate in their production is generated as part of the ethylene (Yang) cycle, since if this was the case, the enrichment pattern would have been +1 m/z unit (see Figure 37).

The role of CmMetAT in volatiles’ production in melon fruit The novel melon gene product of CmMetAT is shown to possess L-methionine aminotransferase activity in vitro as supported by radiolabeled assays (Figure 23b), GC- MS-derivatized assays (Figure 23c) and LC-MS-based assays (Figures 24 and 25). These function validations, together with its’ ripening-dependent expression (Figure

22), can make a link between CmMetAT and the C3-thio-ethers’ production in melon fruit. However, since aminotransferases are reversible enzymes, as was also demonstrated here, it could be that CmMetAT plays a different role, especially considering that a similar enzyme is crucial in the last step of the ethylene cycle (see Figure 37), and its levels are highly increased in ripe melons in parallel to volatiles’ production (Seymour and McGlasson, 1993; Beaulieu and Grimm 2001). Recently, Ellens et al. (2014) characterized glutamine aminotransferases K (GTK) enzymes from tomato and maize converting KMBA to methionine. However, these orthologous of CmMetAT that are suggested to catalyze the final step in the methionine salvage

94 pathway, where not tested for their role in the Yang cycle or in volatiles formation. Finally, since the ethylene cycle takes place predominantly in the plastid and the cytosol and that CmMetAT is predicted to be localized in the mitochondria might suggest that CmMetAT does not play a substantial role in this cycle. However, I cannot exclude roles other than volatiles formation for CmMetAT, including regeneration of L- methionine in the Yang cycle. Further experimentation, including the analyses of knockout mutants of CmMetAT, or its orthologous equivalent in other plants, in respect of volatiles production, might give better understanding of the function of methionine aminotransferases in planta. However, since aminotransferases are highly promiscuous enzymes and do not show strict substrate specifity, many times redundant enzymes might blur the effect or make the determinations less decisive (Maeda et al., 2011; Ellens et al., 2014). Figure 37. Schematic illustration of the ethylene (Yang) cycle. Illustration of the Yang cycle representing ethylene production and sulfur fraction regeneration in plants. Enzymes appear in blue italics: A – SAM synthase, B – ACC synthase, C – ACC oxidase, D – MTA nucleosidase, E – MTR kinase, F – spontaneous reaction. Compound names in black: KMBA – α-keto-γ-(methylthio) butyric acid, SAM – S-adenosyl-L-methionine, MTA – 5-(methylthio) adenosine, MTR – 5-(methylthio) ribose, MTR-1-P – 5-(methylthio) ribose-1- phosphate.

The production of thioesters in melon fruits is mediated by a novel thiol acyl- transferase gene product Thioesters are important compounds in many food products. However, their enzymatic production is poorly understood. The occurrence of volatile thioesters is documented in many plants, microorganisms and food products (Liu et al., 2004; Landaud et al., 2008: Nguyen et al., 2012; Vandendriessche et al., 2013; Wang and Lin, 2014). Methanethiol, one of the products of the MGL reaction, is the most likely substrate for the production of S-methyl thioesters. It may replace the alcohol moiety in the reaction of alcohol acetyltransferase enzymes generating thioesters (Bamforth and Kanauchi, 2003). While some data is available from microorganisms (Helinck et al., 2000), observations of cell-free extracts exhibiting ThAT activities from strawberry fruit (Noichinda et al., 1999) and mung bean roots (Riov and Jaffe, 1972) are the only reports of the presence of S-acylating activities in plants. The spontaneous production of S- methyl thioacetate was documented before in the fungus Geotrichum candidum (Helinck et al., 2000) displaying an optimum pH of 7.0 that is different from what I found here (pH 9.5, Figure 26). The enzymatic activity I found here, for S-methyl thioacetate production (Figure 27) in melon fruits, indicates that melon fruits enzymatically synthesize S-methyl thioacetate in similar way to what was found in cell- free extracts derived from strawberry fruits (Noichinda et al., 1999). Molecular studies aimed at genes encoding enzymes that produce volatile thioesters are not available, either in plants or in microorganisms. I found a novel gene (CmThAT1) encoding a thiol acyl-transferase enzyme. CmThAT1 catalyzes the conversion of acetyl coenzyme A (Co-A) and methanethiol to S-methyl thioacetate in vitro (Figure 29). In addition, co-expression experiments, utilizing E. coli expressing both CmThAT1 and CmMGL, produced significant levels of S-methyl thioacetate but also of S-methyl propanethioate as compared to the controls (Figure 28ab), indicating that CmThAT1 can also mediate the production of thioesters other than S-methyl thioacetate. CmThAT1 expression in the PI 414723 × ‘Dulce’ RIL population families significantly correlated with the presence of S-methyl thioacetate, but lacked a correlation with the levels of other thioesters. Still, in a 2-way hierarchical clustering all the thioesters were clustered together (Figure 14), and S-methyl thioacetate showed high correlations with the all other thioesters (r = 0.25, 0.32, 0.38, 0.49), which also correlated between themselves (r vary from 0.57 to 0.82). This suggests that thioesters undergo a common biosynthetic pathway, are regulated together or both. Since the acyl-CoA substrate for S-methyl

96 thioacetate is acetyl-CoA, a ubiquitous molecule in all living cells, it is likely that the limiting steps to S-methyl thioacetate production are methanethiol availability and ThAT enzyme presence. This makes CmThAT1 expression an important parameter determining the levels of S-methyl thioacetate that is reflected in the high correlation between these two latter parameters (r = 0.59). However, since other acyl-CoA substrates are not as common as acetyl-CoA, most likely, factors other than CmThAT1 expression determine the levels of these thioesters. Considering all the above (the co- expression experiments and the clustering analysis), although CmThAT1 expression in the PI 414723 × ‘Dulce’ RIL population correlates only with S-methyl thioacetate, it appears that CmThAT1 might also be responsible for the production of thioesters other than S-methyl thioacetate. However, the fact that the addition of α-ketobutyrate to E. coli expressing CmThAT1, together with CmMGL, increased the levels of S-methyl propanethioate, but the addition of branched-chain α-keto acids did not have the same effect (not shown) might suggest that CmThAT1 is responsible for the production of straight-chain, but not branched-chain, thioesters. Wang and Lin (2014) showed that S- methyl thioacetate significantly contributes sulfuric aroma to melons, but also mango, sulfuric-pineapple, tropical fruit aroma notes. CmThAT1 might serve as good target to MAS breeding programs trying to control these traits.

The catabolism of L-phenylalanine en route to aroma volatiles in melon fruits The correlation matrix of the 29 L-phenylalanine-derived aroma compounds in the PI 414723 × ‘Dulce’ RIL population (Table 8, Figure 30) shed light on some of the biosynthetic and regulation processes involved in aromatic volatiles’ production. Alcohols and aldehydes were clustered in one group, and the acetate esters were clustered in a separate group, negatively correlated with the aldehydes. This might indicate that the important biosynthetic step controlling the levels of these aroma volatiles is the final acetylation transferase step. Moreover, the lack of correlation (positive correlation, hereafter) between acetate esters and some of the non-acetate esters is probably due to the different biosynthetic pathways, since acetate esters (such as benzyl acetate or 2-phenethyl acetate) are produced from acetyl-CoA and the other esters (such as ethyl benzoate or ethyl phenyl acetate) are produced from phenyl- CoA derivatives. However, this absence of correlation could also be the result of a different regulation of the production of these different esters. The high correlation between (E)- cinnamic acid and benzoic acid with their derived methyl esters implies that these

97 methyl esters are biosynthesized by a benzoic / (E)-cinnamic acid O-methyl transferases. The enrichment of methyl benzoate upon incubation with 13C-L- methionine (Figure 7) strengthens this hypothesis, since methyltransferase enzymes use S-adenosyl-L-methionine (SAM) as the methyl donor (derived from L-methionine). Searching for genes similar to the AmBAMT enzyme (GeneBank accession no. Q9FYZ9, Murfitt et al., 2000) using the tBlastn algorithm on the melon-genome- derived transcripts, yielded a hit (MELO3C003803) that also showed expression in the RIL population. MELO3C003803 expression in the population families correlates with methyl (E)-cinnamate (r = 0.37) and methyl benzoate (r = 0.26) levels, but to lesser extent with the levels of the respective acids (r = 0.31 and 0.14, respectively). The high correlation between vanillin and eugenol (r = 0.46) makes sense since both of them are biosynthetically derived from ferulic acid. The relatively high correlations of the ethyl esters levels (highly correlated with L-methionine-derived ethyl esters, as well) and of the benzyl esters suggest that the availability of the alcohol substrates is more restricting for the formation of these esters than the respective CoA derivatives. More experiments are needed to verify the roles of O-methyltransferase enzymes in the biosynthesis of (E)-methyl cinnamate and methyl benzoate in melons, but a candidate belonging to the SABATH family has been preliminary identified. The enrichment patterns of the 29 L-phenylalanine-derived aroma compounds after the 15 13 incubation of melon fruit rinds with N, C9-L-phenylalanine were in agreement with the length of the carbon chain attached to the aromatic ring (Figure 31, Table 8). This shows that their biosynthesis is straightforward and does not contain shortening of the carbon chain to be elongated again later as in the case of cathinone in Ephedra sinica and Catha edulis (Krizevski et al., 2012). The enrichment pattern of C6−C3 compounds by +9 m/z units rules out any involvement of chain-shortening enzymes, such as aldehyde synthase or decarboxylase (Figure 3), in the biosynthesis of these compounds. Interestingly, 2-phenylacetonitrile showed enrichment patterns of both +8 and +9 m/z units (Table 8). Although 2-phenylacetonitrile is a C6-C2 compound, it contains a nitrogen atom proposed to originate from L-phenylalanine after initial decarboxylation in tomato (Tieman et al., 2006). Following this pathway, the expected enrichment was

+9 m/z units resulting from C8−N1. Indeed, a significant enrichment pattern of +9 m/z units (31% only) was detected, but in addition, 47% enrichment of +8 m/z units was detected as well. This might indicate an additional active biosynthetic route for 2- phenylacetonitrile. However, I cannot exclude the possibility that endogenous melon

15 13 14 13 enzymes converted the supplemented N, C9-phenylalanine to N, C9- phenylalanine. For example, aromatic amino acid aminotransferase can convert the 15 13 13 supplemented N, C9-L-phenylalanine to C9-phenyl pyruvic acid, which can be 14 13 converted back to N, C9-L-phenylalanine by the same enzyme. This nitrogen exchange would also give rise to such an enrichment pattern.

Biosynthesis of (E)-cinnamaldehyde from (E)-cinnamic acid in melon fruit The biosynthesis of (E)-cinnamaldehyde, the main compound giving cinnamon bark its unique aroma, is poorly understood. The +9 m/z units enrichment pattern of

(E)-cinnamaldehyde and of other C3=C6 compounds strengthens the assumption that they are derived in a straightforward pathway from (E)-cinnamic acid. The enzymatic production of (E)-cinnamaldehyde by melon cell-free extracts (Figure 33bc) demonstrates that the enzymes, (E)-cinnamic acid:CoA ligase (CNL) and cinnamoyl CoA reductase (CCR) (Figure 33a), can catalyze (E)-cinnamaldehyde production in melon fruits. The functional characterization of CmCNL shows that it is an (E)- cinnamic acid:coenzyme A ligase enzyme (Figure 35). That observation, together with the relatively high correlation of (E)-cinnamyl acetate with CmCNL expression levels in the RIL population (Table 10), suggest that CmCNL has an important role in (E)- cinnamaldehyde production, that is later converted to (E)-cinnamyl acetate. CmCNL was also found to be a promiscuous enzyme towards its acid substrates, able to utilize benzoic, ferulic and caffeic acids in addition to (E)-cinnamic acid. In that view I additionally found that CmCNL expression is highly correlated with benzyl acetate (r=0.38, α<0.0001). This is in accordance to the high correlation observed between (E)- cinnamyl acetate and benzyl acetate levels within the RIL population families (r=0.74, Table 7). However, while the levels of these two acetate esters are highly correlated with those of butyl acetate (r=0.54), CmCNL expression did not correlate with the levels of non-aromatic acetate esters such as butyl acetate. Taking these together, it is possible that CmCNL controls the levels of benzyl acetate via direct utilization of benzoic acid as a substrate generating benzoyl CoA to be converted to benzyl acetate thereafter. Alternatively, it would be interesting to check if CmCNL controls benzyl acetate levels via (E)-cinnamic acid in a similar manner to the control of the Arabidopsis BZO1 in benzoyl-glucosinolates (Lee et al., 2012). The β-oxidation pathway was suggested as an optional pathway to benzaldehyde formation in petunia petals (Boatright et al., 2004). Still, this possibility should be further experimentally verified by incubating of

99 melon tissues with 13C-labeld (E)-cinnamic acid. Finally, the melon genome contain 15 candidate genes encoding possible CmCCRs (similar to PhCCR1 by melon genome tblastn algorithm). Five of them have significant expression in the RIL population. However, none of them shows correlation with the relevant volatiles. More experiments, such as knockout mutations, are needed to identify the CCR gene responsible for (E)-cinnamaldehyde production in melon fruit.

100

Conclusions Essential amino acids are known precursors of aroma volatile in melons and other fruits and food products. The importance of sulfur-containing aroma volatiles and L-isoleucine-derived aroma volatiles in influencing the aroma and flavor of melon and other fruits is well documented, but this work demonstrates the interaction between these two pathways in the context of aroma volatiles’ formation. Here I show that two different and parallel pathways for L-methionine catabolism contribute to the formation of aroma volatiles in melon fruits. L-methionine degradation via methionine-γ-lyase, as well as through L-methionine aminotransferase, is vital in the formation of sulfur and other melon volatiles. I also show that the catabolic pathway of L-methionine into L- isoleucine might also provide precursors for important aroma compounds. This work demonstrates that MGL enzymes are not only an important factor in volatile production in microorganisms, but also in melon fruits. The incorporation of L-methionine into L- isoleucine in ripe fruit tissue shows that, although mature fruit is considered a sink tissue, it still has the capability to biosynthesize amino acids. I also show that propanoate esters that are straight-chain short volatiles originate from the branched- chain amino acid L-isoleucine. In contrast, L-leucine is a precursor to the acetyl moiety of various acetate esters. This shows that the metabolism of these branched-chain amino acids is differentially regulated. The discovery of the melon gene CmMetAT that uses L-glutamine and KMBA as substrates is one step towards the understanding of volatile

C3-thio-thers production in melons, as well as in other fruits and vegetables. I also found that melon fruits enzymatically produce thioesters. This production is mediated by a thio-acyl transferase enzyme encoded by a novel gene, CmThAT1. The enzymatic production of (E)-cinnamaldehyde by cell-free extracts of ripe melon fruit shows that melon biosynthesizes (E)-cinnamaldehyde and its related volatiles via (E)-cinnamic acid. This is further supported by the characterization of CmCNL and its correlation to (E)-cinnamyl acetate levels. Overall, these findings contribute to the understanding of the biology of complex plant traits, such as aroma formation in fruit. Although the work described here was performed in melons, the findings are also very relevant to understanding volatile aroma formation in other fruits, which have a great impact on general fruit quality and consumer acceptability. Specific volatiles contribute to the flavor and acceptability of melons and they include volatiles imparting a desirable melon-like aromas, as well as those imparting unpleasant off-flavour aromas.

101

Additionally, the genes identified can be used as targets for marker-assisted selection programs as well as for transgenic interfering, aimed at providing the desired aroma of melon fruits and also to repress the levels of non-desired aromas and flavors.

102

Literature cited

Aharoni, A., Keizer, L.C., Bouwmeester, H.J., Sun, Z., Alvarez-Huerta, M., Verhoeven, H.A., Blaas, J., van Houwelingen, A.M., De Vos, R.C., van der Voet, H., Jansen, R.C., Guis, M., Mol, J., Davis, R.W., Schena, M., van Tunen, A.J. and O'Connell, A.P. (2000) Identification of the SAAT gene involved in strawberry flavor biogenesis by use of DNA microarrays. Plant Cell, 12, 647–662. Albers, E. (2009). Metabolic characteristics and importance of the universal methionine salvage pathway recycling methionine from 5′‐methylthioadenosine. IUBMB life, 61, 1132-1142. Allwood, J.W., Cheung, W., Xu, Y., Mumm, R., De Vos, R.C.H., Deborde, C., Biais, B., Maucourt, M., Berger, Y., Schaffer, A. a, Rolin, D., Moing, A., Hall, R.D. and Goodacre, R. (2014) Metabolomics in melon: a new opportunity for aroma analysis. Phytochem., 99, 61–72. Amir, R. (2010) Current understanding of the factors regulating methionine content in vegetative tissues of higher plants. Amino Acids, 39, 917-931. Anthon, G.E. and Barrett, D.M. (2003) Thermal inactivation of lipoxygenase and hydroperoxytrienoic acid lyase in tomatoes. Food Chem., 81, 275–279. Aubert, C. and Pitrat, M. (2006) Volatile compounds in the skin and pulp of Queen Anne’s pocket melon. J. Agric. Food Chem., 54, 8177-8182. Aubert, C. and Bourger, N. (2004) Investigation of volatiles in Charentais cantaloupe melons (Cucumis melo var. cantalupensis) characterization of aroma constituents in some cultivars. J. Agric. Food Chem., 52, 4522-4528. Baldwin, E.A., Scott, J.W., Shewmaker, C.K. and Schuch, W. (2000) Flavor trivia and tomato aroma: biochemistry and possible mechanisms for control of important aroma components. HortSci., 35, 1013-1022. Bamforth, C. and Kanauchi, M. (2003) Use of novel assays to indicate that O-esters and S-esters are produced by the same enzyme in brewing yeast. FEMS Microbiol. Lett., 228, 111–113. Beaulieu, J.C., and Grimm, C.C. (2001) Identification of volatile compounds in cantaloupe at various developmental stages using solid phase microextraction. J. Agric Food Chem., 49, 1345-1352. Beaulieu, J.C. and Lancaster, V.A. (2007) Correlating volatile compounds, sensory attributes and quality parameters in stored fresh-cut cantaloupe. J. Agric. Food

103

Chem., 55, 9503–9513 Beekwilder, J., Alvarez-Huerta, M., Neef, E., Verstappen, F.W.A., Bouwmeester, J.H. and Aharoni, A. (2004) Functional characterization of enzymes forming volatile esters from strawberry and banana. Plant Physiol., 135, 1865-1878. Berger, C., Martin, N., Collin, S., Gijs, L., Khan, J.A., Piraprez, G., Spinnler, H.E. and Vulfson, E.N. (1999) Combinatorial approach to flavor analysis. 2. Olfactory investigation of a library of S-methyl thioesters and sensory evaluation of selected components. J. Agric. Food Chem., 47, 3274-3279. Beuerle, T. and Pichersky, E. (2002) Enzymatic synthesis and purification of aromatic coenzyme A esters. Anal. Biochem., 302, 305-312. Boatright, J., Negre, F., Chen, X., Kish, C.M,, Wood, B., Peel, G., Orlova, I., Gang, D., Rhodes, D. and Dudareva, N. (2004) Understanding in vivo benzenoid metabolism in petunia petal tissue. Plant Physiol., 135, 1993-2011. Buck, L. and Axel, R. (1991) A novel multigene family may encode odorant receptors: a molecular basis for odor recognition. Cell, 65, 175-187. Burger, Y., Sa’ar, U., Paris, H.S., Lewinsohn, E., Katzir, N., Tadmor, Y. and Schaffer, A.A. (2006) Genetic variability for valuable fruit quality traits in Cucumis melo. Isr. J. Plant Sci., 54, 233-242. Burger, Y., Paris, H.S., Cohen, R., Katzir, N., Tadmor, Y., Lewinsohn, E. and Schaffer, A.A. (2009) Genetic diversity of Cucumis melo. Hort. Rev., 36, 165-168. Bushdid, C., Magnasco, M.O., Vosshall, L.B. and Keller, A. (2014). Humans can discriminate more than 1 trillion olfactory stimuli. Science, 343, 1370-1372. Buttery, R.G., Guadagni, D.G. and Ling, L.C. (1973) Volatile components of baked potatoes. J. Sci. Food Agric., 24, 1125–1131. Buttery, R.G., Seifert, R.M., Ling, L.C., Soderstrom, E.L., Ogawa, J.M. and Turnbaugh, J.G. (1982) Additional aroma components of honeydew melon. J. Agric. Food Chem., 30, 1208-1211. Buttery, R.G., and Teranishi, R. and Ling, L.C. (1987) Fresh tomato aroma volatiles: a quantitative study. J. Agric. Food Chem., 35, 540-544. Cañas, R.A., Villalobos, D.P., Díaz-Moreno, S.M., Cánovas, F.M. and Cantón, F.R. (2008). Molecular and functional analyses support a role of ornithine-δ- aminotransferase in the provision of glutamate for glutamine biosynthesis during pine germination. Plant Physiol., 148, 77-88. Chapple, C.C., Glover, J.R. and Ellis, B.E. (1990) Purification and characterization

104

of methionine: glyoxylate aminotransferase from Brassica carinata and Brassica napus. Plant physiol., 94, 1887-1896. Dal Cin V., Tieman D.M., Tohge T., McQuinn R., de Vos R.C.H., Osorio S., Schmelz E.A., Taylor M.G., Smits-Kroon M.T., Schuurink R.C., Haring M.A., Giovannoni J., Fernie A.R. and Klee H.J. (2011) Identification of genes in the phenylalanine metabolic pathway by ectopic expression of a MYB transcription factor in tomato fruit. Plant Cell, 23, 2738-2753. Dangl, J.L., Dietrich, R.A. and Thomas, H. (2000) Senescence and programmed cell death. In Buchanan, B.B, Gruissem, W. and Jones, L.R, Eds, Biochemistry & Molecular Biology of Plants. American Society of Plant Physiologists: Rockville, Maryland, pp 1044-1100. Danin-Poleg, Y., Reis, N., Baudracco-Arnas, S., Pitrat, M., Staub, J.E., Oliver, M., Arus, P., de-Vicente, C.M. and Katzir, N. (2000) Simple sequence repeats in Cucumis mapping and map merging. Genome, 43, 963–974. Davidovich-Rikanati, R., Sitrit, Y., Tadmor, Y., Iijima, Y., Bilenko, N., Bar, E., Carmona, B., Fallik, E., Dudai, N., Simon, J.E., Pichersky, E. and Lewinsohn, E. (2007) Enrichment of tomato flavor by diversion of the early plastidial terpenoid pathway. Nature Biotechnol., 25, 899-901. Davidovich-Rikanati, R., Lewinsohn, E., Bar, E., Iijima, Y., Pichersky and E., Sitrit, Y. (2008) Overexpression of the lemon basil α-zingiberene synthase gene increases both mono- and sesquiterpene contents in tomato fruit. Plant J., 56, 228- 238. Dias, B. and Weimer, B. (1998) Purification and characterization of L-methionine gamma-lyase from Brevibacterium linens BL2. Appl. Environ. Microbiol., 64, 3327-3331. Dickinson, J.R., Salgado, L.E.J. and Hewlins, M.J.E. (2003) The catabolism of amino acids to long chain and complex alcohols in Saccharomyces cerevisiae. J. Biol. Chem., 278, 8028-8034. Diebold, R., Schuster, J., Däschner, K. and Binder, S. (2002) The branched-chain amino acid transaminase gene family in Arabidopsis encodes plastid and mitochondrial proteins. Plant Physiol., 129, 540-550. Dudareva, N., Klempien, A., Muhlemann, J.K. and Kaplan, I. (2013). Biosynthesis, function and metabolic engineering of plant volatile organic compounds. New Phytol., 198, 16-32.

105

Ehrlich, F. (1907) Über die bedingungen der fuselölbildung und über ihren zusammenhang mit dem eiweissaufbau der hefe. Ber. Dtsch. Chem. Ges., 40, 1027- 1047. El-Sharkawy, I., Manríquez, D., Flores, F.B., Regad, F., Bouzayen, M., Latché, A. and Pech, J.C. (2005) Functional characterization of a melon alcohol acyl- transferase gene family involved in the biosynthesis of ester volatiles. Identification of the crucial role of a threonine residue for enzyme activity. Plant Mol. Biol., 59, 345-362. Ellens, K.W., Richardson, L.G., Frelin, O., Collins, J., Ribeiro, C.L., Hsieh, Y.F., Mullen R.T. and Hanson, A.D. (2014) Evidence that glutamine transaminase and omega-amidase potentially act in tandem to close the methionine salvage cycle in bacteria and plants. Phytochemistry. doi: 10.1016/j.phytochem.2014.04.012 Escribano, S., Sánchez, F.J. and Lázaro, A. (2010) Establishment of a sensory characterization protocol for melon (Cucumis melo L.) and its correlation with physical–chemical attributes: indications for future genetic improvements. Eur. Food Res. Technol., 231, 611–621. Farhi, M., Lavie, O., Masci, T., Hendel-Rahmanim, K., Weiss, D., Abeliovich, H. and Vainstein A. (2010) Identification of rose phenylacetaldehyde synthase by functional complementation in yeast. Plant Mol. Biol., 72, 235-245. Gaid, M.M., Sircar, D., Müller, A., Beuerle, T., Liu, B., Ernst, L., Hänsch, R. and Beerhues, L. (2012). Cinnamate: CoA ligase initiates the biosynthesis of a benzoate-derived xanthone phytoalexin in Hypericum calycinum cell cultures. Plant Physiol., 160, 1267-1280.

Galliard, T. and Phillips, D.R. (1976) The enzymic cleavage of linoleic acid to C9 carbonyl fragments in extracts of cucumber (Cucumis sativus) fruit and the possible role of lipoxygenase. BBA - Lipid Lipid Met., 431, 278–287. Garcia-Mas, J., Benjak, A., Sanseverino, W., et al. (2012) The genome of melon (Cucumis melo L.). Proc. Natl. Acad. Sci. USA, 109, 11872-11877. Goff, S.A. and Klee, H.J. (2006) Plant volatile compounds: sensory cues for health and nutritional value? Science, 311, 815–819. Goffner, D., Campbell, M.M., Campargue, C., Clastre, M., Borderies, G., Boudet, A. and Boudet, A.M. (1994). Purification and characterization of cinnamoyl- coenzyme A: NADP oxidoreductase in Eucalyptus gunnii. Plant Physiol., 106, 625- 632.

106

Gonda, I., Bar, E., Portnoy, V., Lev, S., Burger, J., Schaffer, A.A., Tadmor, Y., Gepstein, S., Giovannoni, J.J., Katzir, N. and Lewinsohn, E. (2010) Branched- chain and aromatic amino acid catabolism into aroma volatiles in Cucumis melo L. fruit. J. Exp. Bot., 61, 1111-1123. Gonda, I., Burger, Y., Schaffer, A.A., Ibdah, M., Tadmor, Y., Katzir, N., Fait, A. and Lewinsohn, E. (2014) Biosynthesis and perception of melon aroma. In Havkin- Frenkel, D., Dudai, N. and Belanger, F.C. Eds. Biotechnology in Flavor Production. Blackwell Publishing Ltd., Oxford, UK. In press. González, V.M., Garcia-Mas, J., Arús, P. and Puigdomènech, P. (2010) Generation of a BAC-based physical map of the melon genome. BMC Genomics 11, 339. Goyer, A., Collakova, E., Shachar-Hill, Y. and Hanson, A.D. (2007) Functional characterization of a methionine -lyase in Arabidopsis and its implication in an alternative to the reverse trans-sulfuration pathway. Plant Cell Physiol., 48, 232– 242. Günther, C.S., Matich, A.J., Marsh. K.B. and Nicolau, L. (2010) (Methylsulfanyl)alkanoate ester biosynthesis in Actinidia chinensis kiwifruit and changes during cold storage. Phytochem., 71, 742–750. Haapalainen, A.M., Meriläinen, G., Pirilä, P.L., Kondo, N., Fukao, T. and Wierenga, R.K. (2007) Crystallographic and kinetic studies of human mitochondrial acetoacetyl-CoA thiolase: the importance of potassium and chloride ions for its structure and function. Biochem., 46, 4305-4321. Hacham, Y., Avraham, T. and Amir, R. (2002) The N-terminal region of Arabidopsis cystathionine γ-synthase plays an important regulatory role in methionine metabolism. Plant physiol., 128, 454-462. Harel-Beja, R., Tzuri, G., Portnoy, V. et al. (2010) A genetic map of melon highly enriched with fruit quality QTLs and EST markers, including sugar and carotenoid metabolism genes. Theor. Appl. Genet., 121, 511-533. Hayata, Y., Sakamoto, T., Maneerat, C., Li, X., Kozuka, H. and Sakamoto K., (2003) Evaluation of aroma compounds contributing to muskmelon flavor in porapak Q extracts by aroma extract dilution analysis. J. Agric. Food Chem., 51, 3415-3418. Helinck, S., Spinnler, H. E., Parayre, S., Dame‐Cahagne, M. and Bonnarme, P. (2000). Enzymatic versus spontaneous S‐methyl thioester synthesis in Geotrichum candidum. FEMS Microbial. Lett., 193, 237-241.

107

Horvat, R.J and Senter, S.D. (1987) Identification of additional volatile compounds from cantaloupe. J. food sci., 52, 1097-1098. Huart, C., Rombaux, P. and Hummel, T. (2013) Plasticity of the human olfactory system: the olfactory bulb. Molecules, 18, 11586-11600. Huang, T., Joshi, V. and Jander, G. (2014) The catabolic enzyme methionine gamma- lyase limits methionine accumulation in potato tubers. Plant Biotechnol. J., 12, 883– 893. Ibdah, M., Azulay, Y., Portnoy, V., Wasserman, B., Bar, E., Meir, A., Burger, Y., Hirschberg, J., Schaffer, A.A., Katzir, N., Tadmor, Y. and Lewinsohn, E. (2006) Functional characterization of CmCCD1, a carotenoid cleavage dioxygenase from melon. Phytochem., 67, 1579-1589. Jaisson, S., Veiga-da-Cunha, M. and Van Schaftingen, E. (2009). Molecular identification of ω-amidase, the enzyme that is functionally coupled with glutamine transaminases, as the putative tumor suppressor Nit2. Biochimie., 91, 1066-1071. Jansky, S.H. (2010) Potato flavor. Am. J. Pot. Res., 87, 209–217 Jordán, M.J., Shaw, P.E. and Goodner, K.L. (2001) Volatile components in aqueous essence and fresh fruit of Cucumis melo cv. athena (muskmelon) by GC-MS and GC-O. J. Agric. Food Chem., 49, 5929-5933. Joshi, V. and Jander, G. (2009) Arabidopsis methionine γ-lyase is regulated according to isoleucine biosynthesis needs but plays a subordinate role to threonine deaminase. Plant Physiol., 151, 367–378. Kaminaga, Y., Schnepp, J., Peel, G., Kish, C.M., Ben-Nissan, G., Weiss, D., Orlova, I., Lavie, O., Rhodes, D., Wood, K., Porterfield, D.M., Cooper, A.J.L., Schloss, J.V., Pichersky, E., Vainstein, A. and Dudareva, N. (2006) Plant phenylacetaldehyde synthase Is a bifunctional homotetrameric enzyme that catalyzes phenylalanine decarboxylation and oxidation. J. Biol, Chem., 281, 23357- 23366. Klee, H.J. (2010) Improving the flavor of fresh fruits: genomics, biochemistry, and biotechnology. New Phytol., 187, 44–56. Klee, H.J. and Tieman, D.M. (2013) Genetic challenges of flavor improvement in tomato. Trends in Genet., 29, 257-262. Klempien, A., Kaminaga, Y., Qualley, A., Nagegowda, D.A., Widhalm, J.R., Orlova, I., Shasany, A.K., Taguchi, G., Kish, C.M., Cooper, B.R., D’Auria, J.C., Rhodes, D., Pichersky, E. and Dudareva, N. (2012) Contribution of CoA

108

ligases to benzenoid biosynthesis in petunia flowers. The Plant Cell, 24, 2015–2030. Kliebenstein, D.J., D’Auria, J.C., Behere, A.S., Kim, J.H., Gunderson, K.L., Breen, J.N., Lee, G., Gershenzon, J., Last, R.L. and Jander, G. (2007) Characterization of seed-specific benzoyloxyglucosinolate mutations in Arabidopsis thaliana. The Plant J., 51, 1062–1076. Kochevenko, A., Araújo W.L., Maloney, G.S., Tieman, D.M., Thi Do, P., Taylor, M.G., Klee, H.J. and Fernie, A.R. (2012) Catabolism of branched chain amino acids supports respiration but not volatile synthesis in tomato fruits. Mol. Plant, 5, 366–375.

Kourkoutas., D, Elmore, J.S. and Mottram, D.S. (2006) Comparison of the volatile compositions and flavour properties of cantaloupe, Galia and Honeydew muskmelons. Food Chem., 97, 95-102. Koeduka, T., Fridman, E., Gang, D.R., Vassão, D.G., Jackson, B.L., Kish, C.M., Orlova, I., Spassova, S.M., Lewis, N.G., Noel, J.P., Baiga, T.J., Dudareva, N. and Pichersky, E. (2006) Eugenol and isoeugenol, characteristic aromatic constituents of spices, are biosynthesized via reduction of a coniferyl alcohol ester. Proc. Natl Acad. Sci. USA, 26, 10128-10133. Krizevski, R., Bar, E., Shalit, O., Levy, A., Hagel, J. M., Kilpatrick, K., Marsolaise, F., Facchini, P. J., Ben-Shabat, S., Sitrit,Y. and Lewinsohn, E. (2012) Benzaldehyde is a precursor of phenylpropylamino alkaloids as revealed by targeted metabolic profiling and comparative biochemical analyses in Ephedra spp. Phytochem., 81, 71-79. Kudou, D., Misaki, S., Yamashita, M., Tamura, T., Esaki, N. and Inagaki, K. (2008) The role of cysteine 116 in the active site of the antitumor enzyme L- methionine γ-lyase from Pseudomonas putida. Biosci. Biotechnol. Biochem., 72, 1722–1730. Kuentzel, H. and Bahri, D. (1990) Synthetic ingredients of food flavourings. In Ashurts, P.R. Ed, Food Flavorings; Blackie Academic & Professional, Glasgow and London, pp. 115-157. Landaud, S., Helinck, S. and Bonnarme, P. (2008) Formation of volatile sulfur compounds and metabolism of methionine and other sulfur compounds in fermented food. Appl. Microbiol. Biotechnol., 77, 1191-1205.

109

Langmead, B., Trapnell, C., Pop, M., Salzberg, S.L. (2009) Ultrafast and memory- efficient alignment of short DNA sequences to the human genome. Genome Biol., 10:R25. Larkov, O., Zaks, A., Bar, E., Lewinsohn, E., Dudai, N., Mayer, A.M. and Ravid, U. (2008) Enantioselective monoterpene alcohol acetylation in Origanum, Mentha and Salvia species. Phytochem., 69, 2565-2571. Lee, S., Kaminaga, Y., Cooper, B., Pichersky, E., Dudareva, N. and Chapple, C. (2012) Benzoylation and sinapoylation of glucosinolate R-groups in Arabidopsis. The Plant J., 72, 411–422. Lignou, S., Parker, J.K., Baxter, C. and Mottram, D.S. (2014) Sensory and instrumental analysis of medium and long shelf-life Charentais cantaloupe melons (Cucumis melo L.) harvested at different maturities. Food Chem., 148, 218–29. Lisec, J., Schauer, N., Kopka, J., Willmitzer, L. and Fernie, A.R. (2006). Gas chromatography mass spectrometry–based metabolite profiling in plants. Nat. protoc., 1, 387-396. Liu, S.Q., Holland, R. and Crow,V.L. (2004) Esters and their biosynthesis in fermented dairy products: a review. Int. Dairy J., 14, 923–945. Liu, M., Nauta, A., Francke, C. and Siezen, R.J. (2008) A comparative genomics of enzymes in flavor-forming pathways from amino acids in lactic acid bacteria. Appl. Environ. Microbiol., 74, 4590–4600. Luchetta, L., Manríquez, D., El-Sharkawy, I., Flores, F.B., Sanchez-Bel, P., Zouine, M., Ginies, C., Bouzayen, M., Rombaldi, C., Pech, J.C. and Latché, A. (2007) Biochemical and catalytic properties of three recombinant alcohol acyltransferases of melon. Sulfur-containing ester formation, regulatory role of CoA-SH in activity, and sequence elements conferring substrate preference. J. Agric. Food Chem., 55, 5213–5220. Maeda H., Yoo H., and Dudareva N. (2011) Prephenate aminotransferase directs plant phenylalanine biosynthesis via arogenate. Nat. Chem. Biol. 7, 19-21. Mallick, M.F.R. and Masui, M. (1968) Origin, distribution and taxonomy of melons. Scientia. Hortic., 28, 251-261. Maloney, G.S., Kochevenko, A., Tieman, D.M., Tohge, T., Krieger, U., Zamir, D., Taylor, M.G., Fernie, A.R. and Klee, H.J. (2010) Characterization of the branched-chain amino acid aminotransferase enzyme family in tomato. Plant physiology, 153, 925-936.

110

Manríquez, D., El-Sharkawy, I., Flores, F.B., Yahyaoui, F., Regad, F., Bouzayen, M., Latché, A, and Pech, J.C. (2006) Two highly divergent alcohol dehydrogenases of melon exhibit fruit ripening-specific expression and distinct biochemical characteristics. Plant Mol. Biol., 61, 675-685. Matich, A. and Rowan, D. (2007) Pathway analysis of branched-chain ester biosynthesis in apple using deuterium labeling and enantioselective gas chromatography-mass spectrometry. J. Agric. Food Chem., 55, 2727-2735. McGorrin, R.J. (2011) The significance of volatile sulfur compounds in food flavors. In Volatile Sulfur Compounds in Food. Qian, M.C., Fan, X. and Mahattanatawee K. eds. ACS Symposium Series, 1068, pp. 3–31. Miyazaki, J.H. and Yang, S.F. (1987) Metabolism of 5-methylthioribose to methionine. Plant physiol., 84, 277-281. Muhlemann, J. K., Woodworth, B. D., Morgan, J. A. and Dudareva, N. (2014) The monolignol pathway contributes to the biosynthesis of volatile phenylpropenes in flowers. New Phytol., 204, 661–670. Murfitt, L.M., Kolosova, N., Mann, C.J. and Dudareva, N. (2000) Purification and characterization of S-adenosyl-L-methionine: benzoic acid carboxyl methyltransferase, the enzyme responsible for biosynthesis of the volatile ester methyl benzoate in flowers of Antirrhinum majus. Arch. Biochem. Biophys., 382, 145-151. Mussinan, C.J. and Kellan, M.E. (1994) Sulfur compound in food. An overview. In Mussinan, C.J. and Keelan, M.E. eds. Sulfur Compounds in Foods. Washington DC: ACS Symposium Series 564, American Chemical Society, pp. 1-6. Nguyen, D.D., Nicolau, L. and Kilmartin, P.A. (2012) Application of an automated headspace solid phase micro-extraction for the GC-MS detection and quantification of reductive sulfur compounds in wines. In Salih, B. and Çelikbıçak, Ö. Eds. Gas Chromatography in Plant Science, Wine Technology, Toxicology and Some Specific Applications. Intech, 171-194. Noichinda, S., Ueda, Y., Imahori, Y. and Chachin, K. (1999) Thioester production and thioalcohol specificity of alcohol acetyltransferase in strawberry fruit. Food Sci. Technol. Res., 5, 99-103. Pang, X., Guo, X., Qin, Z., Yao, Y., Hu, X. and Wu, J. (2012) Identification of aroma- active compounds in Jiashi muskmelon juice by GC-O-MS and OAV Calculation J. Agric. Food Chem., 60, 4179−4185.

111

Paranagama, P.A., Wimalasena, S., Jayatilake, G.S., Jayawardena, A.L., Senanayake, U.M. and Mubarak, A.M. (2001) A comparison of essential oil constituents of bark, leaf, root and fruit of cinnamon (Cinnamomum zeylanicum Blum) grown in Sri Lanka. J. Natl. Sci. Found. Sri., 29, 147-153. Paris H.S. (2012) Semitic-language records of snake melons (Cucumis melo, Cucurbitaceae) in the medieval period, the ‘piqqus’ of the ‘faqqous’. Genet. Resour. Crop Ev. 59, 31-38. Paris, H.S., Amar, Z. and Lev, E. (2012) Medieval emergence of sweet melons, Cucumis melo (Cucurbitaceae). Ann. Bot. 110, 23-33. Pérez, A.G., Sanz, C. and Olias, J.M. (1993) Partial purification and some properties of alcohol acetyltransferase from strawberry fruits. J. Agric. Food Chem., 41, 1462- 1466. Pérez, A.G., Sanz, C., Olias, R., Rios, J.J. and Olias, J.M. (1996) Evolution of strawberry alcohol acetyltransferase activity during fruit development and storage. J. Agric. Food Chem., 44, 3286-3290. Perpète, P., Duthoit, O., De-Maeyer, S., Imray, L., Lawton, A.I., Stavropoulos, K.E., Gitonga, V.W., Hewlins, M.J.E. and Dickinson, J.R. (2006) Methionine catabolism in Saccharomyces cerevisiae. FEMS Yeast Res., 6, 48–56. Perry, P.L., Wang, Y. and Lin, J. (2009) Analysis of honeydew melon (Cucumis melo var. inodorus) flavour and GC–MS/MS identification of (E,Z)-2,6-nonadienyl acetate. Flavour Fragr. J., 24, 341–347. Pfaffl, M.W. (2001) A new mathematical model for relative quantification in real-time RT-PCR. Nucleic Acids Res., 29, 2002-2007. Phillips, D.R. and Galliarda, T. (1978) Flavour biogenesis. Partial purification and properties of a fatty acid hydroperoxide cleaving enzyme from fruits of cucumber. Phytochem., 17, 355–358. Pitrat, M., Hanelt, P. and Hammer, K. (2000) Some comments on interspecific classification of cultivars of melon. In Katzir, N., Paris, H.S. eds. Proceeding of cucurbitaceae 2000, 7th Eucarpia Meeting on Cucurbit Genetics and Breeding. Acta Hort., 510, 29-36. Pommerrenig, B., Feussner, K., Zierer, W., Rabinovych, V., Klebl, F., Feussner, I. and Sauer, N. (2011) Phloem-specific expression of Yang cycle genes and identification of novel Yang cycle enzymes in Plantago and Arabidopsis. The Plant Cell, 23, 1904-1919.

112

Portnoy, V., Benyamini, Y., Bar, E., Harel-Beja, R., Gepstein, S., Giovannoni, J.J., Schaffer, A.A., Burger, J., Tadmor, Y., Lewinsohn, E. and Katzir, N. (2008) The molecular and biochemical basis for varietal variation in sesquiterpene content in melon (Cucumis melo L.) rinds. Plant Mol. Biol., 66, 647-661. Portnoy, V., Diber, A., Pollock , S., Karchi, H., Lev, S., Tzuri, G., Harel-Beja, R., Forer, R., Portnoy, V.H., Lewinsohn, E., Tadmor, Y., Burger, J., Schaffer, A. and Katzir, N. (2011) Use of non-normalized, non-amplified cDNA for 454-based RNA sequencing of fleshy melon fruit. Plant Genome, 4, 36-46. Preston, C., Richling, E., Elss, S., Appel, M., Heckel, F., Hartlieb, A. and Schreier, P. (2003) On-line gas chromatography combustion/pyrolysis isotope ratio mass spectrometry (HRGC-C/P-IRMS) of pineapple (Ananas comosus L. Merr.) volatiles. J. Agric. Food Chem., 51, 8027-8031. Pritchett, M.A. and Metcalf, W.W. (2005) Genetic, physiological and biochemical characterization of multiple methanol methyltransferase isozymes in Methanosarcina acetivorans C2A. Mol. Microbiol. 56, 1183–1194. Ramakers, C., Ruijter, J.M., Deprez, R.H. and Moorman, A.F. (2003) Assumption- free analysis of quantitative real-time polymerase chain reaction (PCR) data. Neurosci. Lett., 339, 62-66. Rébeillé, F., Jabrin, S., Bligny, R., Loizeau, K., Gambonnet, B., van-Wilder, V., Douce, R. and Ravanel, S. (2006) Methionine catabolism in Arabidopsis cells is initiated by a γ-cleavage process and leads to S-methylcysteine and isoleucine syntheses. Proc. Natl Acad. Sci. USA, 103, 15687–15692. Rijnen, L., Bonneau, S. and Yvon, M. (1999) Genetic characterization of the major lactococcal aromatic aminotransferase and its involvement in conversion of amino acids to aroma compounds. Appl. Environ. Microbiol., 65, 4873-4880. Riov, J. and Jaffe, M.J. (1972) Evidence for enzymatic thioltransacetylation in plant extracts. Phytochem., 11, 2437-2438. Roessner-Tunali, U., Liu, J., Leisse, A., Balbo, I., Perez-Melis, A., Willmitzer, L. and Fernie, A.R. (2004) Kinetics of labelling of organic and amino acids in potato tubers by gas chromatography-mass spectrometry following incubation in 13C labelled isotopes. The Plant J., 39, 668–679. Rolls, E.T. (2005) Taste, olfactory, and food texture processing in the brain, and the control of food intake. Physiol. Behav., 85, 45–56.

113

Rouquier, S., Blancher, A. and Giorgi, D. (2000) The olfactory receptor gene repertoire in primates and mouse: evidence for reduction of the functional fraction in primates. Proc. Natl Acad. Sci. USA, 97, 2870–2874. Sato, D. and Nozaki, T. (2009) Methionine gamma-lyase: the unique reaction mechanism, physiological roles, and therapeutic applications against infectious diseases and cancers. IUBMB Life, 61, 1019–1028. Schaefer, H., Heibl, C. and Renner, S.S. (2009) Gourds afloat: a dated phylogeny reveals an Asian origin of the gourd family (Cucurbitaceae) and numerous oversea dispersal events. Proceedings of the Royal Society B – Biological Sciences, 276, 843-851. Schieberle, P., Ofner, S. and Grosch, W. (1990) Evaluation of potent odorants in cucumbers (Cucumis sativus) and muskmelons (Cucumis melo) by aroma extract dilution analysis. J. Food Sci., 55, 193-195. Schuster, J. and 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. and Binder, S. (2006) Branched- chain aminotransferase4 is part of the chain elongation pathway in the biosynthesis of methioninehionine-derived glucosinolates in Arabidopsis. Plant Cell, 18, 2664- 2679. Schwab, W., Davidovich-Rikanati, R. and Lewinsohn, E. (2008) Biosynthesis of plant-derived flavor compounds. The Plant J., 54, 712-732. Senanayake, U.M., Lee, T.H. and Wills, R.B. (1978). Volatile constituents of cinnamon (Cinnamomum zeylanicum) oils. J. Agric. Food Chem., 26, 822-824. Senesi. E., Di Cesare, L.F., Prinzivalli, C. and Lo Scalzo, R. (2005) Influence of ripening stage on volatiles composition, physicochemical indexes and sensory evaluation in two varieties of muskmelon (Cucumis melo L. var reticulates Naud). J. Sci. Food Agric., 85, 1241–1251. Seymour, G.B. and McGlasson, W.B. (1993). Melons. In Seymour, G.B., Taylor, J.E and Tucker, G.A. eds. Biochemistry of Fruit Ripening. London: Chapman & Hall, pp. 273-290. Shalit, M., Katzir, N., Tadmor, Y., Larkov, O., Burger, Y., Shalechet, F., Lastochkin, E., Ravid, U., Amar, O., Edelstein, M., Karchi, Z. and Lewinsohn,

114

E. (2001) Acetyl CoA: alcohol acetyl transferase activity and aroma formation in ripening melon fruits. J. Agric. Food Chem., 49, 794-799. Shepherd, G.M. (2006) Smell images and the flavour system in the human brain. Nature, 444, 316-321. Singh, B.K. (1999) Biosynthesis of valine, leucine and isoleucine. In Singh, B.K. ed. Plant Amino Acids. Biochemistry and Biotechnology. Princeton, New Jersey: American Cyanamid company, pp. 227-249. Smith, P.G. and Welch, J.E. (1964) Nomenclature of vegetables and condiment herbs grown in United States. Proc. Am. Soc. Hort. Sci., 84, 534-548. Stead, L.M., Brosnan, J.T., Brosnan, M.E., Vance, D.E. and Jacobs, R.L. (2006) Is it time to reevaluate methyl balance in humans? Am. J. Clin. Nutr., 83, 5-10. Suzuki, M., Wu, S., Tomita, N., Fujita, A. and Watanabe, N. (2014). Biogenesis of volatile methyl esters in snake fruit (Salacca edulis, Reinw) cv. Pondoh. Biosci. Biotechnol. Biochem., 67, 1267–1271. Tapp, E.J., Cummins, I., Brassington, D. and Edwards, R. (2008) Determination and isolation of a thioesterase from passion fruit (Passiflora edulis Sims) that hydrolyzes volatile thioesters. J. agric. food chem., 56, 6623-6630. Temussi, P. (2006). The history of sweet taste: not exactly a piece of cake. J. Mol. Recognit., 19, 188–99. Thomson, D.M.H. (1987). The meaning of flavor. In Birch, G.G. and Lindley, M.G. eds. Developments in Food Flavors. Elsevier, London, pp 1-21. Tieman, D., Taylor, M., Schauer, N., Fernie, A.R., Hanson, A.D. and Klee, H.J. (2006) 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, D., Bliss, P., McIntyre, L.M., Blandon-Ubeda, A., Bies, D., Odabasi, A.Z., Rodríguez, G.R., van der Knaap, E., Taylor, M.G., Goulet, C., Mageroy, M.H., Snyder, D.J., Colquhoun, T., Moskowitz, H., Clark, D.G., Sims, C., Bartoshuk, L. and Klee, H.J. (2012) The chemical interactions underlying tomato flavor preferences. Current Biol., 22, 1035-1039. Trapnell, C., Pachter, L. and Salzberg S.L. (2009) TopHat: discovering splice junctions with RNA-Seq. Bioinformatics 25, 1105-1111. Tressl, R. and Drawert, F. (1973) Biogenesis of banana volatiles. J. Agric. Food Chem., 21, 560-565.

115

Tucker, G.A. (1993). Introduction. In Seymour, G.B., Taylor, J.E. and Tucker, G.A. eds. Biochemistry of Fruit Ripening. Chapman & Hall, London pp. 1-51. Vallone, S., Sivertsen, H., Anthon, G.E., Barrett, D.M., Mitcham, E.J., Ebeler, S.E. and Zakharov, F. (2013). An integrated approach for flavour quality evaluation in muskmelon Cucumis melo L. reticulatus group) during ripening. Food chem. 139, 171-183. Vandendriessche, T., Vermeir, S., Mayayo Martinez, C., Hendrickx, Y., Lammertyn, J., Nicolaï, B.M. and Hertog, M.L.A.T.M. (2013). Effect of ripening and inter-cultivar differences on strawberry quality. LWT-Food Sci. Technol., 52, 62-70. Vogt, T. (2010) Phenylpropanoid biosynthesis. Mol. Plant, 3, 2-20. Wang, J. and De Luca, V. (2005) The biosynthesis and regulation of biosynthesis of Concord grape fruit esters, including ‘foxy’ methylanthranilate. The Plant J., 44, 606–619. Wang, Y. and Lin, J. (2014) Gas chromatography–olfactometry and gas chromatography–tandem mass spectrometry analysis of fresh cantaloupe (Cucumis melo L. var. cantalupensis Naudin) aroma. Flavour Fragr. J., 29, 87–94. Werkhoff, P., Güntert, M., Krammer, G., Sommer, H. and Kaulen, J. (1998) Vacuum headspace method in aroma research: flavor chemistry of yellow passion fruits. J. Agric. Food Chem., 46, 1076-1093. Whitaker, T.W. and Davis, G.N. (1962) Cucurbits: botany, cultivation and utilization. Interscience publishers Inc. New York. Wyllie, S.G. and Leach, D.N. (1992) Sulfur-containing compounds in the aroma volatiles of melons (Cucumis melo). J. Agric. Food Chem., 40, 253-256. Wyllie, S.G., Leach, D.N., Wang, Y. and Shewfelt, R.L. (1994) Sulfur volatiles in Cucumis melo cv. Makdimon (muskmelon) aroma. Sensory evaluation by gas chromatography-olfactometry. In Mussinan, C.J. and Keelan, M.E. eds. Sulfur Compounds in Foods. Washington DC: ACS Symposium Series 564, American Chemical Society, pp 36−48. Wyllie, S.G., Leach, D.N., Wang, Y. and Shewfelt, R.L. (1995) Key aroma compounds in melons. Their development and cultivar dependence. In Roussef, R.L. and Leahy, M.M. eds. Fruit Flavors. ACS Symposium Series 596, American Chemical Society: Washington, DC, pp 248-257.

116

Wyllie, S.G. and Fellman, J.K. (2000) Formation of volatile branched chain esters in bananas (Musa sapientum L.). J. Agric. Food Chem., 48, 3493-3496. Yabumoyo, K. and Jennings, W.G. (1977) Volatile constituents of cantaloupe, Cucumis melo, and their biogenesis. J. Food Sci., 42, 32-37. Yahyaoui, F., Wongs-Aree, C., Latché, A., Hackett, R., Grierson, D. and Pech, J.C. (2002) Molecular and biochemical characteristics of a gene encoding an alcohol acyl-transferase involved in the generation of aroma volatile esters during melon ripening. Eur. J. Biochem., 269, 2359-2366. Yeshurun, Y. and Sobel, N. (2010) An odor is not worth a thousand words: from multidimensional odors to unidimensional odor objects. Annu. Rev. Psychol., 61, 219-241. Yoo H., Widhalm J.R., Qian Y., Maeda H., Cooper B.R., Jannasch A.S., Gonda I., Lewinsohn E., Rhodes D. and Dudareva N. (2013) An alternative pathway contributes to phenylalanine biosynthesis in plants via a cytosolic tyrosine: phenylpyruvate aminotransferase. Nat. Commun. 4, doi: 10.1038/ncomms3833. Yvon, M., Chambellon, E., Bolotin, A. and Roudot-Algaron, F. (2000) Characterization and role of the branched-chain aminotransferase (BcaT) isolated from Lactococcus lactis subsp. Cremoris NCDO 763. Appl. Environ. Microbiol., 66, 571-577. Yvon, M. and Rijnen, L. (2001) Cheese flavour formation by amino acid catabolism. Inter. Dairy J., 11,185-201. Zaks, A., Davidovich-Rikanati, R., Bar, E., Inbar, M. and Lewinsohn, E. (2008) Biosynthesis of linalyl acetate and other terpenes in Mentha aquatica var. citrata. Israel J. Plant Sci., 56, 233-244. Zheng, L.Y., Sun, G.M., Liu, Y.G., Lv, L.L., Yang, W.X., Zhao, W.F. and Wei, C.B. (2012). Aroma volatile compounds from two fresh pineapple varieties in China. Int. J. Mol. Sci., 13, 7383-7392. Zhong, S., Joung, J.G., Zheng, Y., Chen, Y.R., Liu, B., Shao, Y., Xiang, J.Z,, Fei, Z. and Giovannoni, J.J. (2011) High-throughput Illumina strand-specific RNA sequencing library preparation. Cold Spring Harb. Protoc., 2011, 940-949.

117

תקציר

המלון ).Cucumis melo L( הינו מין השייך למשפחת הדלועיים הכוללת בין היתר את המלפפון, האבטיח, הקישוא והדלעת. המלון הוא מין הכולל מגוון רחב של זני בר ותרבות, ומשמש כגידול חקלאי ברחבי העולם. בשנת 2102 היבול הכולל של מלונים בישראל עמד על 3..6 אלף טון, ועל יותר ממיליון טון בארה"ב. בעולם כולו היצור השנתי של מלונים עומד על יותר מ22- מיליון טונות )2110- 2112( כאשר כמעט ממחציתו מגיע מסין. טעם טוב הינו מאפיין חיוני באיכותו של פרי המלון כשהארומה היא מרכיב מרכזי בקביעת הטעם של כל פרי וזן. ארומת המלון מורכבת ממספר רב של חומרי ארומה נדיפים אשר הינם נגזרות של חמורי מוצא שונים כגון, חומצות אמיניות, חומצות שומן, קרוטנואידים וטרפנים אחרים. ידוע כי חומצות אמיניות חיוניות מהוות חומרי מוצא לחומרי ארומה רבים המשפיעים על טעמו של המלון כמו גם על טעמם של פירות ומזונות שונים. חומצות אמיניות אלו כוללות בין היתר, מתיוניון, פנילאלאנין והחומצות מסועפות השרשרת: איזולאוצין, לאוצין ו-ואלין. חומרי ארומה נדיפים המכילים אטום גופרית הינם מרכיב חשוב בארומה של פרי המלון ושל פירות ומאכלים נוספים. זנים וגנוטיפים שונים של מלון מכילים כמויות שונות של נדיפים המכילים גופרית. החומצה האמינית מתיונין הוצעה כחומר המוצא של חומרים נדיפים אלו. הדגרה של קוביות מלון עם מתיונין מסומנת באיזוטופים יציבים )13C או 2H( הראתה כי בפרי המלון פעילים שני מסלולי פירוק מקבילים של חומצה אמינית זו לחומרי ארומה: הראשון מייצר אסטרים תיו-

אתריים )C3( באמצעות האנזים מתיונין אמינוטרנספראז )MetAT(. השני, באמצעות האנזים מתיונין-גמא-ליאז )MGL( המשחרר את התוצר מתאנתיול אשר הינו חומר מוצא לחומרי ארומה

המכילים גופרית שונים הכוללים תיול, סולפידים, תיו-אסטרים ואסטרים תיו-אתריים )C2(. מתיונין גם נמצא כחומר מוצא לחומרי ארומה אשר אינם מכילים גופרית באמצעות מטבוליזם של אלפא-קטובוטיראט, תוצר נוסף של האנזים MGL. מאלפא-קטובוטיראט מיוצרת החומצה האמינית איזולאוצין וחומרי ארומה נוספים הכוללים אסטרים מסועפי ויישרי שרשרת. שימוש באיזולאוצין מסומנת באיזוטופים יציבים )13C( הראה כי חומצה זו משמשת כחומר מוצא לאסטרים פרופנואטים )יישרי שרשרת( ולא רק לחומרי ארומה הנושאים את השרשרת הצדדית שלה )-2מתילבוטיל(, אך לא לאסטרים אצטטים. מנגד, החומצה האמינית לאוצין כן נמצאה כחומר מוצא לאסטרים אצטטיים ולא רק לחומרי ארומה הנושאים את השרשרת הצדדית שלה )-6 מתילבוטיל(, אך לא לאסטרים פרופנואטים. מיצויי חלבונים אל-תאיים מפרי מלון בשל נמצאו כבעלי פעילות אנזימטית של MGL. גן חדש ממלון )CmMGL( המקודד לאנזים MGL אותר באמצעות חיפוש בספריית EST ובוטא באופן

118

הטרולוגי בחיידקי אשריכיה קולי. נמצא כי החלבון הרקומביננטי הינו בעל פעילות אנזימטית של MGL בריאקציות אל-תאיות המייצר גם מתאנתיול וגם אלפא-קטובוטיראט. ביטויו של CmMGL הוא נמוך בשלבי ההתפתחות המוקדמים של הפרי, אך עולה בפרי הבשל כתלות בזן הנבדק. בנוסף, נבדקו רמות הביטוי של גן זה באוכלוסייה של קווים רקומביננטיים )RIL( שנוצרה מהכלאה בין הגנוטיפ PI 414723 בעל ארומה גופריתית ולא נעימה, לבין הגנוטיפ המסחרי 'דולצ'ה' אשר הינו בעל ארומת מלון טובה. נמצא כי בקווי האוכלוסייה ביטוי הגן מתכנס יחד עם הכמות הכוללת של חומרי הארומה המכילים הגופרית אשר ההנחה היא כי הינם מיוצרים במסלול המתחיל באנזים MGL. גן נוסף, CmMetAT, המקודד לאנזים מתיונין אמינוטראנספראז בודד ואופיין במערכות הטרולוגיות. ביטויו נמצא נמוך ברקמות הפרי הצעיר וגבוה ברקמות הפרי הבשל בשני הגנוטיפים PI 414723 ו'דולצ'ה'. לאנזים זה תפקיד משוער ביצור של חומרי גופרית מהמסלול הראשון. בניסויי הדגרה נמצא כי תוצר הריאקציה האנזימטית של חלבון זה, גמא-)מתילתיו( בוטיראט,

משמש כחומר מוצא לאסטרים תיו-אתריים )C3( מהמסלול הראשון אך לא לחומרי גופרית מהמסלול השני. מתיל-תיואסטרים )S-methyl thioesters( הינם חומרים נדיפים חשובים התורמים לארומת פרי המלון, אך ברמות גבוהות יכולים לגרום לריחות לוואי גופריתיים לא רצויים. הם מיוצרים ממתאנתיול, אחד מהתוצרים האנזימטיים של האנזים MGL. הביוסינתיזה שלהם נחקרה מעט מאוד עד היום בצמחים ובמיקרואוגניזמים כאחד. ההשערה הינה כי ייצורם מתבצע ע"י אנזים הדומה לאלכוהול אציל-טראנספראז )AAT( המתפקד בייצורם של אסטרים חמצניים )O-esters( והינו תיול-אציל טראנספראז )ThAT(. פעילות אנזימטית של ייצור מתיל-תיואצטט מהסובסטראטים מתאנתיול ואצטיל קו-אנזים A נמצאה במיצויי חלבונים אל-תאיים שהופקו מרקמת פרי מלון בשל. בהסתכלות על חמישה תיואסטרים מתוך .0 נדיפים המכילים גופרית באוכלוסיית הקווים הרקומביננטיים הנ"ל נמצא כי התיואסטרים מתקבצים יחדיו בנפרד מחומרי גופרית אחרים. הרמות של מתיל-תיואצטט הראו קורלציה גבוהה עם רמות הביטוי של גן המקודד לאנזים המשוער לתפקד כאלכוהול אצטיל טרנספראז )r=0.59, α<0.0001, קורלציית פירסון( בקווי אוכלוסייה זו. גן זה )CmThAT1( נמצא כמקודד לאנזים תיול-אציל טראנספראז המייצר מתיל-תיואצטט מהסובסטראטים מתאנתיול ואצטיל קו-אנזים A במבחנה. כמו-כן בביטוי הטרולוגי משותף בחיידקי קולי של CmThAT1 ושל CmMGL נמצא ייצור של שני תיואסטרים: מתיל-תיואצטט ומתיל-פרופאנתיואט. בנוסף למתיונין ולחומצות האמיניות מסועפות השרשרת, החומצה האמינית פנילאלאנין משמשת גם היא כחומר מוצא לחומרי ארומה נדיפים בפרי המלון. בניסויי הדגרה של קוביות מלון עם פנילאלאנין מסומנת באטומים יציבים )13C ו – 15N( נמצאו

119

22 נדיפים שונים שמקורם בחומצה אמינית זו. ביניהם נמצא גם החומר צינמאלדהיד, המרכיב העיקרי בארומת הקינמון. המסלול הביוסינטתי של ייצור חומר זה אינו מתועד היטב בספרות. מיצוי חלבונים אל-תאי מפרי מלון בשל מכיל פעילויות אנזימטיות מצומדות של חומצה צינמית:קו-אנזים A ליגאז )CNL( וצינאמויל קו-אנזים A רדוקטאז )CCR(. בריאקציות אלו מיוצר צינמאלדהיד מהסובסטרט חומצה צינמית הנגזרת מפנילאלאנין. אנליזה של 22 החומרים שמקורם בפנילאלאנין באוכלוסיית הקווים הרקומביננטיים חשפה מספר גנים קנדידטיים לאנזימים האחראים לייצורם של חלק מהחומרים הנ"ל. עבודה זו מקדמת אותנו מספר צעדים קדימה בהבנת הייצור של מכלול חומרי הארומה הנדיפים בפרי המלון ובפירות נוספים. היא שופכת אור על מספר תהליכים קטבוליים כמו גם אנאבוליים של חומצות אמיניות בפרי המלון, ותתרום ליצירה של כלים לטיפוח סלקטיבי ומדויק של מלונים עם איכות פרי גבוהה מבלי לפגוע בתכונות אחרות של הפרי.

120

העבודה נעשתה בהדרכת

אהרון פאיט המכונים לחקר המדבר ע"ש יעקב בלאושטיין המכון לביוטכנולוגיה וחקלאות של איזורים צחיחים

ובהדרכת אפרים לוינסון מנהל המחקר החקלאי המכון למדעי הצמח, מרכז מחקר נווה-יער

121

ייצור חומרי ארומה נדיפים מחומצות אמיניות חיוניות בפרי המלון

מחקר לשם מילוי חלקי של הדרישות לקבלת תואר "דוקטור לפילוסופיה"

מאת

איתי גונדה

הוגש לסינאט אוניברסיטת בן גוריון בנגב

תמוז התשע"ד יולי 4102

באר שבע

122