Potential Grape-Derived Contributions to Volatile Ester Concentrations in Wine

Transcription

1 Moleules 215, 2, ; doi:1.339/moleules Artile OPEN ACCESS moleules ISSN Potentil Grpe-Derived Contriutions to Voltile Ester Conentrtions in Wine Pul K. Boss 1, *, Anthony D. Pere 2,, Ynji Zho 2, Emily L. Niholson 1, Eri G. Dennis 1, nd Dvid W. Jeffery 2 1 CSIRO Agriulture Flgship, PMB 2, Glen Osmond, SA 564, Austrli; E-Mils: emily.niholson@siro.u (E.L.N.); eri.dennis@u. (E.G.D.) 2 Shool of Agriulture, Food nd Wine, Wite Reserh Institute, The University of Adelide, PMB 1, Glen Osmond, SA 564, Austrli; E-Mils: _pere@post.om (A.D.P.); ynji.zho@student.delide.edu.u (Y.Z.); dvid.jeffery@delide.edu.u (D.W.J.) Present ddress: South Austrlin Reserh nd Development Institute, GPO Box 397, Adelide, SA 51, Austrli Present ddress: Deprtment of Chemistry, The University of British Columi, Okngn Cmpus, 3333 University Wy, Kelown, BC V1V 1V7, Cnd * Author to whom orrespondene should e ddressed; E-Mil: pul.oss@siro.u; Tel.: Ademi Editors: Helene Hopfer nd Susn E. Eeler Reeived: 22 Mrh 215 / Aepted: 23 April 215 / Pulished: 29 April 215 Astrt: Grpe omposition ffets wine flvour nd rom not only through vrietl ompounds, ut lso y influening the prodution of voltile ompounds y yest. C9 nd C12 ompounds tht potentilly influene ethyl ester synthesis during fermenttion were studied using model grpe juie medium. It ws shown tht the ddition of free ftty ids, their methyl esters or yl-rnitine nd yl-mino id onjugtes n inrese ethyl ester prodution in fermenttions. The stimultion of ethyl ester prodution ove tht of the ontrol ws pprent when lower onentrtions of the C9 ompounds were dded to the model musts ompred to the C12 ompounds. Four mino ids, whih re involved in CoA iosynthesis, were lso dded to model grpe juie medium in the sene of pntothente to test their ility to influene ethyl nd ette ester prodution. β-alnine ws the only one shown to inrese the prodution of ethyl esters, free ftty ids nd ette esters. The ddition of 1 mg L 1 β-lnine ws enough to stimulte prodution of these ompounds nd

2 Moleules 215, ddition of up to 1 mg L 1 β-lnine hd no greter effet. The endogenous onentrtions of β-lnine in fifty Cernet Suvignon grpe smples exeeded the 1 mg L 1 required for the stimultory effet on ethyl nd ette ester prodution oserved in this study. Keywords: grpe; wine; ester; mino id; CoA; yest; fermenttion 1. Introdution Wine is omplex solution ontining undnt voltile ompounds whih ontriute to wine rom nd flvour, nd onsequently impt wine qulity nd ppreition. Wine roms re rodly tegorised into three groups tht reflet their soure. Primry roms re grpe-derived voltiles tht pss through fermenttion often unhnged, nd re lrgely responsile for vrietl roms. Seondry roms, whih re y fr the gretest pool of voltile moleules, re produed through the winemking proess, with the gret mjority produed y yest during loholi fermenttion s metolism yproduts [1,2]. Tertiry roms develop in finished wine through storge nd mturtion, nd result from intermoleulr hemil intertions nd equilirium effets s the wine mtrix hnges. The voltile omposition of most wines, or styles of wine, is very similr, nd the vrietl differenes tht exist etween wines mde from different vrieties of grpes re lrgely due to the reltive rtios of the voltile ompounds ontined within [3]. The types of voltile ompounds re diverse, nd inlude esters, higher lohols, ldehydes, ketones, ltones, ids, phenols, N-heteroyles, isoprenoids nd sulfur ompounds [1,4]. Of these tegories, higher lohols represent the lrgest voltile pool in terms of onentrtion, ut esters hve the lrgest numer of ontriuting moleules [5]. In this work the term esters enompsses oth ethyl nd ette esters, whih hve signifint effet on wine rom y ontriuting fruity nd florl hrteristis [6,7]. These esters re minly produed y yest metolism through ftty id yl- nd etyl-coenzyme A (CoA) pthwys [7,8]. CoA is ritil oftor for lrge numer of metoli pthwys nd is used to tivte intermedites during the iosynthesis of medium hin ftty ids (MCFAs); the yl-coa intermedites formed re then esterified with ethnol y esterse nd trnsferse enzymes, forming MCFA ethyl esters. Aette esters, on the other hnd, re produed through the ondenstion of yest-derived higher lohols with etyl-coa, gin under the ontrol of ester-forming enzymes [7,8]. Ester prodution minly ours during fermenttion when yests generte mple mounts of ethnol nd higher lohols vi sugr nd mino id metolism, respetively [9], so CoA my ply signifint role in determining the ester ontent in wine [1]. For instne, it hs een suggested tht ftors tht regulte the prodution or onsumption of etyl-coa will in turn lter the mount of esters produed y yest, nd there is evidene for the ssoition of some speifi mino ids with CoA iosynthesis [11 13]. However, there re no reports diretly studying the effet of CoA-relted mino id omposition (rising from the grpes) on CoA iosynthesis, nd susequently how hnge in CoA onentrtion might influene ester levels in wine. Despite the role of yl trnsferse enzymes in ester formtion y yest, Serens et l. [14] indited tht the level of gene expression is not the limiting ftor for ester prodution, nd tht the vilility of MCFA preursors hs n importnt role. Moreover, grpe hemil omposition hs een highlighted

3 Moleules 215, s eing highly influentil to the level of prodution of suite of rom ompounds, nd mny grpe omponents re known to e depleted nd onverted to lohols nd esters through fermenttion [15]. It hs een widely demonstrted tht yest n lierte voltile moleules from vrious grpe-derived onjugtes, suh s glyosidilly ound nd ysteine- nd glutthione-onjugted voltile ompounds [16,17]. More reent investigtions hve sought to demonstrte the reltionship etween grpe-derived preursors nd wine voltile roms, suh s study involving ferments in model grpe juie medium spiked with vrying mounts of nturl grpe juie to test the impt on wine roms [18]. This experiment reveled n rry of omplex intertions ross the grpe juie proportions dded; however, severl fermenttion esters were identified s hving positive, lrgely liner orreltions with grpe juie ontent in the ferments. Following this work, grpe-derived liphti lohols nd ldehydes were identified s preursors to ette esters in wine [19]. In prtiulr, the C6 ompounds (E)-2-hexenl, hexnl, (E)-2-hexen-1-ol, nd hexn-1-ol were shown to e preursors to hexyl ette, wheres otn-1-ol nd enzyl lohol were identified s preursors to otyl ette nd enzyl ette, respetively [19]. These exmples my lso imply the presene of grpe-derived MCFAs, or their preursors, ontriuting to the pool of MFCA ethyl esters lierted during fermenttion, whih ould provide nother importnt link etween grpe omposition nd wine voltile profile. Ultimtely, deeper understnding of the soures nd iosynthesis pthwys of rom ompounds involves linking vitiulturl prtises, grpe vriety, grpe hemil omposition, yest strin seletion, winemking tehniques nd wine mtrix intertions etween rom onstituents with defined sensory outomes in the finished wine [2,21]. The im of this study ws to ddress one omponent of this overll gol, nd tht ws to identify links etween potentil grpe omponents nd wine ester onentrtions. A series of model fermenttions were onduted with MCFAs nd their puttive grpe-derived preursors, s well s mino ids relted to the iosynthesis of CoA, nd esters rising in these fermenttions were quntified with GC-MS nlysis. 2. Results nd Disussion 2.1. Potentil Medium Chin Ftty Aid (MCFA)-Derived Preursors of Ethyl Esters The first prt of this study investigted potentil grpe-derived preursors of the ethyl esters of C9 nd C12 ftty ids (synthesised or ommerilly ville). These two ron lengths were hosen for two resons. First, odd-numered ftty ids nd their ethyl esters re found in low undne in wine, mening tht it is esier to follow the effet of the ddition of odd-numered potentil preursors on ethyl ester prodution s the endogenous onentrtions re low. Seond, ethyl dodenote is extrted t moderte mounts y the SPME method employed in this study ompred to shorter even-numered ethyl esters nd hene does not sturte the detetor [19]. This llowed the effets of C12 preursor ddition to e redily deteted nd llowed the omprison of odd- nd even-numered preursors to e tested in se they re metolised differently during fermenttion. The potentil grpe-derived preursors tested inluded the C9 nd C12 free ftty ids, s well s ftty ids in vrious onjugted forms. Methyl esters were inluded euse they hve een identified in grpes nd represent possile ontriutors to the preursor pool [22]. Crnitine-ftty id onjugtes re

4 Moleules 215, potentil ethyl ester preursors, s they re the intrellulr trnsport moleule involved in the movement of ftty ids ross memrnes s prt of yest metoli tivities [23]. Amino id-ftty id onjugtes hve een reported in mmmlin systems nd in insets (e.g., [24,25]), ut s yet hve not een desried in plnts. Nevertheless, ftty id-mino id onjugtes n t s induers of plnt defene responses when they exist in herivorous inset sliv (e.g., [26]) nd so mino id onjugtes were inluded s puttive preursors Free Ftty Aids s Diret Preursors C9 nd C12 ftty ids were dded to model grpe juie medi (MGJM) fermenttions ross rnge intended to pture the onentrtions reported in the literture for omprle moleules in grpes [27,28]. No signifint differene in the prodution of ethyl nonnote ws seen etween the ontrol nd the fermenttions supplemented with 1 μm nonnoi id (Figure 1A). However, there were signifint inreses when 1 μm or 1 μm nonnoi id ws dded to the MGJM, with ester onentrtions rehing 112 nmol L 1. The ddition of 1 mm nonnoi id ppered to e toxi to the yest, s negligile mss ws lost throughout the fermenttion period. It hs een shown in previous studies tht free ftty ids n e toxi to yest t high onentrtions [29,3]. No signifint inrese in the prodution of ethyl dodenote ws oserved fter dditions of up to 1 μm dodenoi id to the fermenttions (Figure 1A). However, there ws lrge response when 1 mm dodenoi id ws dded, leding to men ethyl dodenote onentrtion of 8155 nm (i.e., 1.86 mg L 1, sizele quntity) ompred to men of 118 nm in the ontrol wine. An inresed lg phse ws seen in the 1 mm smples, otherwise the fermenttion kinetis ppered not to hve een impted in the dodenoi id series of dditions (dt not shown). The esterifition of the ftty id my e detoxifying tion tht ours ove prtiulr toxiity threshold. The findings of Stevens nd Hofmeyr [31] support this, s esterifition of C8 nd C1 ftty ids redued their toxiity to the yest. The results suggest tht onentrtion threshold for this toxiity exists, nd tht grpe-derived MCFAs ould ontriute to these onentrtions. It would pper tht the threshold is muh lower for nonnoi id thn for dodenoi id nd, sine most previous reserh hs foused on the effets of even hin MCFAs [8,32], the results of this study suggest potentil odd/even ron numer effet tht is worthy of further investigtion. Alterntively, the stimultion of ethyl nonnote prodution y lower onentrtion of nonnoi id thn tht oserved for dodenoi id nd its orresponding ethyl ester my reflet the size of the endogenous pool of sustrte. As odd-numered MCFAs re less ommon thn the even-numered ones, lower onentrtion would e required to inrese the ftty id sustrte onentrtion to point where the prodution of the ethyl ester is signifintly greter thn tht in the ontrol wine. This my lso e the se if the ftty ids hve differentil ilities to diffuse or e tively trnsported into the yest where esterifition ours.

5 Moleules 215, (A) Free ftty ids Ethyl ester (nm) ontrol 1µM 1µM 1µM 1mM 1 Nonnoi id Dodenoi id (B) Methyl esters 1 ontrol 1µM 1µM 1µM 1mM Ethyl ester (nm) d d Ethyl ester (nm) Methyl nonnote Methyl dodenote (C) Ayl-rnitine onjugtes ontrol 1µM 1µM 1µM 1mM 1 Nonnoyl rnitine Dodenoyl rnitine Figure 1. (A) Ethyl nonnote nd ethyl dodenote onentrtions in model wines produed fter the ddition of inresing mounts of nonnoi id or denoi id to the MGJM; (B) Ethyl nonnote nd ethyl dodenote onentrtions in model wines produed fter the ddition of inresing mounts of methyl nonnote or methyl denote to the MGJM; (C) Ethyl nonnote nd ethyl dodenote onentrtions in model wines produed fter the ddition of inresing mounts of nonnoyl rnitine or dodenoyl rnitine to the MGJM. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,,, d or ʹ, ʹ, ʹ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y the Gmes-Howell post ho test. Note tht the y-xis is in log sle.

6 Moleules 215, Methyl Esters of MCFAs s Trnsesterifition Sustrtes The sme onentrtions of the methyl esters of nonnoi nd dodenoi ids s used for the free ftty id experiments were dded to MGJM to test the ility of these ompounds to influene ethyl ester prodution. Methyl nonnote stimulted ethyl nonnote prodution when dded t onentrtions of 1 µm nd greter (Figure 1B). The ddition of 1 mm methyl nonnote to the model must resulted in ethyl nonnote prodution of 234 nm (429 ng L 1 ) nd did not hve the toxi effet oserved when the free ftty id ws dded to the must t this onentrtion. The ddition of methyl dodenote produed similr response to tht of the dodenoi id dditions, with no signifint inrese in ethyl dodenote t ny onentrtion elow 1 mm, ut signifint inrese of ethyl dodenote prodution t tht level of ddition (Figure 1B). These oservtions ould gin e relted to either toxiity response of the yest ove ertin threshold, refletion of the size of the endogenous sustrte pools for these ompounds in reltion to the mount dded to the must, or onsequene of differentil rtes of trnsport into the yest. The metolism of methyl esters of ftty ids y yest is seemingly unknown; however, methyl esters of numer of MCFAs hve een found in grpes [33], nd hve een reported in model wines (e.g., [18]). Yest-medited trnsesterifition, or loholysis, is prole route of formtion of ethyl esters from methyl esters, lthough the onentrtions of these ompounds in the fruit re likely to e too low to gretly influene ethyl ester onentrtions in wine sed on the oservtions in this experiment Ayl-Crnitine Conjugtes, Puttive Trnsport Forms of MCFAs No signifint differene ws oserved in the ethyl nonnote onentrtions of the ontrol wines nd those produed fter the ddition of 1 nd 1 μm nonnoyl rnitine. However, the ddition of 1 μm nonnoyl rnitine to MGJM signifintly inresed the ethyl nonnote onentrtion in the resulting wine (Figure 1C). Inresing the onentrtion of nonnoyl rnitine from 1 μm to 1 mm lso used signifint inreses in ethyl nonnote prodution in the susequent fermenttions (Figure 1C). The ddition of 1 nd 1 μm dodenoyl rnitine to the MGJM used signifint differene in ethyl dodenote ontent of the model wines, nd this ws further enhned when the model must ws supplemented with 1 μm dodenoyl rnitine (Figure 1C). When 1 mm dodenoyl rnitine ws dded to the MGJM the yest ppered to lose viility nd fermenttion of the MGJM did not our, denoted y no loss of weight nd no floulene in the must (dt not shown). The pprent toxiity of the high onentrtion of dodenoyl rnitine ompred to the free ftty id ws unexpeted. Ftty ids pper to e imported into yest ells in proess tht requires tivtion of the moleule with CoA [34]. The rnitine onjugte my e trnsported more redily s it is more solule thn the free ftty id nd euse it is lredy tivted. The inresed trnsporttion of the rnitine onjugte into the ell, followed y liertion of the rnitine, my led to toxi intrellulr levels of the free ftty id tht is not seen when the free ftty id lone is dded to the model must. The mehnism for the trnsport of extrellulr rnitine or onjugtes into the ell in Shromyes erevisie is still unler. Prdoxilly, the nonnoyl rnitine onjugte did not inhiit fermenttion when dded to the must t onentrtion of 1 mm, wheres the free ftty id did. This my reflet the differene in the ftty

7 Moleules 215, id hin length whih my llow more diffusion of the C9 ftty id into the ell ompred to oth nonnoyl rnitine nd free dodenoi id. Certinly the stimultion of ethyl nonnote prodution ws 1-fold lower fter the ddition of 1 or 1 µm nonnoyl rnitine ompred to the equivlent mounts of nonnoi id (Figure 1). Alterntively, the odd numer of rons in the nonnoyl rnitine onjugte my inhiit ny further metolism of this ompound into something tht my e toxi to the yest (for exmple, the free id) ompred to the even-numered dodenoyl rnitine. The onentrtion of yl rnitine onjugtes in grpe erries is unknown. A study tht quntified free rnitine nd rnitine esters in mny different foods found tht fleshy fruits suh s pples, guvs nd ornges ontin onentrtions of totl rnitine esters of etween mg kg 1 [35]. The grpe seed, eing soure of lipids, my ontin more rnitine esters thn the rest of the erry, ut further work is required to determine if these onjugtes from grpes n lter ester omposition in wines Ayl-Amino Aid Conjugtes Potentilly Exploiting Permeses Addition of nonnoyl lnine to the MGJM t onentrtion of 1 μm resulted in signifint inrese in the prodution of ethyl nonnote in the resulting wine (Figure 2A). Similr to when the free ftty id ws dded to the musts, 1 mm nonnoyl lnine ppered to e toxi to the yest s there ws negligile weight loss in these smples; however, ethyl nonnote onentrtions were signifintly higher thn the ontrol. Nonnoyl isoleuine ddition lso used n inrese in ethyl nonnote prodution when dded in onentrtions t 1 μm nd ove (Figure 2B). The toxiity seen when 1 mm nonnoyl lnine ws dded to the MGJM ws not oserved fter the ddition of nonnoyl isoleuine to the MGJM (Figure 2A,B). Ethyl dodenote onentrtions were signifintly higher only in the wine produed fter the ddition of 1 mm dodenoyl lnine to the must, whih led to men wine onentrtion of lmost 1, nmol L 1, or 2.24 mg L 1 (Figure 2A). On the ontrry, no signifint differene ws seen in ethyl dodenote prodution t ny level of dodenoyl isoleuine ddition (Figure 2B). The trnsport mehnism for extrellulr lnine nd isoleuine is vi the generl yest mino id permeses or other permeses with less rod speifiity [36]. However, the onjugted MCFAs re perhps more likely to enter the ell in mnner similr to MCFAs. It would pper tht the lnine onjugtes re trnsported more esily into the ell s the ddition of oth the nonnoyl nd dodenoyl vrints to the musts resulted in wines with greter onentrtions of their respetive ethyl esters thn did the equivlent ddition of the isoleuine onjugtes (Figure 2). Alterntively, the lnine onjugtes ould e more esily esterified or hydrolysed into the free ftty id, whih would inrese the ville sustrte for esterifition. Inresed trnsport or hydrolysis of the onjugte ould lso explin the inresed toxiity of the nonnoyl lnine onjugte t 1 mm ompred to the isoleuine onjugte. The identifition nd quntifition of yl-mino id onjugtes in plnts hs not een reported lthough it is only reently tht suh ompounds hve een desried in mmmls nd insets [24,25]. Future trgeted studies my revel their presene in the plnt kingdom s well Amino Aids Involved in CoA Biosynthesis Four mino ids known to e involved in CoA synthesis (Figure 3) were seleted for studies to determine their effet on ester prodution in MGJM. Vline is ssoited with CoA genertion s it is

8 Moleules 215, onverted to 2-oxoisovlerte, preursor of pntothente [12]. inine is the preursor for the iosyntheti prodution of spermine [13], from whih β-lnine is synthesised y S. erevisie vi polymine degrdtion. β-alnine then ts s the intermedite for pntotheni id iosynthesis [13]. The lter steps of CoA synthesis require the retion of pntothente with ysteine nd ATP (Figure 3). (A) Ayl-lnine onjugtes 1 Control 1μM 1μM 1μM 1mM Ethyl ester (nm) Nonnoyl lnine Dodenoyl lnine (B) Ayl-isoleuine onjugtes 1 Control 1μM 1μM 1μM 1mM Ethyl ester (nm) Nonnoyl isoleuine Dodenoyl isoleuine Figure 2. (A) Ethyl nonnote nd ethyl dodenote onentrtions in model wines produed fter the ddition of inresing mounts of yl-lnine onjugtes to the MGJM. (B) Ethyl nonnote nd ethyl dodenote onentrtions in model wines produed fter the ddition of inresing mounts of yl-isoleuine onjugtes to the MGJM. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,, or ʹ, ʹ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y the Gmes-Howell post ho test. Note tht the y-xis is in log sle. Sine grpe juie ontins vrious hemil ompounds inluding pntotheni id nd other preursors tht ould ffet ester syntheses, it is neessry to eliminte these ftors when studying the role of these four mino ids in ester prodution during fermenttion. Using hemilly defined MGJM ontining simple mixture of sugrs nd nutrients (exluding nitrogen soures) n imitte winemking onditions nd is more suitle when guging the effets of hnging one or more vriles. However, supplementing model musts with ertin mino ids t different levels will lso ontriute to

9 Moleules 215, differenes in totl nitrogen ontents, whih n influene ester synthesis during fermenttion (e.g., [37]). Hene, totl yest ssimille nitrogen (YAN) ws normlised ross the experiments y the ddition of inorgni mmonium in the form of mmonium hloride (NH4Cl). It ws lso neessry to eliminte pntotheni id from the norml MGJM s it is through this pthwy tht we hypothesised tht the mino ids my hve n effet on ester prodution. Figure 3. Simplified pthwy of CoA iosynthesis in the yest S. erevisie to emphsise the potentil role of the four mino ids (oxed) used in this study Stimultion of Ester Prodution y β-alnine. A set of model fermenttions were onduted ontining one CoA-relted mino id t two onentrtions (5 nd 1 mg L 1 β-lnine; 1 nd 3 mg L 1 L-ysteine; 33 nd 65 mg L 1 L-rginine; 3 nd 12 mg L 1 L-vline). The ysteine, rginine nd vline onentrtions fll within, or re slightly higher thn, the typil onentrtions of these mino ids found in grpe juies [38]. The β-lnine onentrtions used were slightly higher thn those previously reported for V. vinifer grpes [39]. Negtive ontrol fermenttions did not ontin mino ids, nd positive ontrol fermenttions inluded ll four mino ids t the lowest level dded in the individul tretments

10 Moleules 215, desried ove (i.e., 5 mg L 1 β-lnine; 1 mg L 1 L-ysteine; 33 mg L 1 L-rginine nd 3 mg L 1 L-vline). The different tretments ffeted the length of the fermenttion lg phse. The model must supplemented with the mixed mino ids hd lg phse of pproximtely two dys, wheres ll the other fermenttions hd lg phse of four dys, despite them hving the sme totl YAN (dt not shown). This ould e due to the differenes in struturl omplexity of the nitrogen soure [4]. It ws hypothesised tht the four mino ids of interest would influene ester prodution through their role in CoA iosynthesis (Figure 3). Therefore, the onentrtions of these mino ids ould theoretilly influene the level of CoA produed y yest during fermenttions, nd CoA ering etyl/yl groups would impt on the synthesis of ethyl/ette esters. The β-lnine-supplemented wines hd signifintly more ethyl hexnote, ethyl otnote nd ethyl denote ( µg L 1 ; µg L 1 ; µg L 1, respetively) thn the ontrol wine (16 µg L 1 ; 39 µg L 1 ; 34 µg L 1 ) or those produed fter supplementtion of the must with the other three mino ids (Figure 4). However, there were no signifint differenes etween the ethyl ester onentrtions in those wines produed from must with either 5 mg L 1 or 1 mg L 1 β-lnine dded. This suggests tht the effet of the ddition of this ompound on ethyl ester prodution my e lredy sturted t 5 mg L 1. The wines tht were produed fter the ddition of ll four mino ids to the must lso ontined signifintly more ethyl esters thn the ontrol, ut in ll three ses the onentrtions in the wines were less thn those oserved in those produed fter the ddition of 5 mg L 1 β-lnine lone (Figure 4). This my e due to the differene in the time it took for the fermenttions to reh dryness, s this hppened pproximtely two dys erlier in the musts with ll four mino ids dded ompred to the other tretments. Voltilistion or metolism of the ethyl esters my hve ourred in these wines while the other tretments were still rehing dryness. Individul dditions of rginine, ysteine nd vline did not hve n impt on ester onentrtions s they were not signifintly different from the negtive ontrol in ll ses. This suggests tht the yest is not limited in its ility to produe ethyl esters during fermenttion y the supply of rginine, ysteine or vline. Signifintly inresed onentrtions of MCFAs (hexnoi id µg L 1 vs. 681 µg L 1 ; otnoi id µg L 1 vs. 184 µg L 1 ; denoi id µg L 1 vs. 174 µg L 1 ) were oserved in fermenttions spiked with 5 nd 1 mg L 1 β-lnine or with ll four mino ids ompred to the ontrol wine (Figure 5). Therefore, the inrese in ethyl ester prodution my e linked to n inrese in the prodution of MCFAs rther thn n inrese in tivtion of existing free ftty ids vi n inrese in CoA onentrtion. The lk of ny signifint differene in hexnoi nd otnoi id ontent etween the wines produed fter the ddition of four mino ids nd those mde fter β-lnine ddition lone my reflet the lower voltility of the MCFAs, ompred to the ethyl esters, whih redues their loss from the wine in the two-dy time differene etween the ompletions of the fermenttions. However, this ws not the se for denoi id, whih ws found in signifintly lower onentrtions when ll four mino ids were dded ompred to those spiked with β-lnine lone (Figure 5C).

11 Moleules 215, (A) Ethyl Hexnote (μm) (B).8.7 Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All Ethyl Otnote (μm) (C).6 Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All.5 Ethyl Denote (μm) Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All Figure 4. The onentrtion of (A) ethyl hexnote, (B) ethyl otnote nd (C) ethyl denote in model wines produed fter the ddition of mino ids s indited on the x-xis. The numer in rkets represents the onentrtion dded in mg L 1. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,, ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y Tukey s post ho test.

12 Moleules 215, (A) Hexnoi Aid (μm) (B) Otnoi Aid (μm) Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All 5 (C) 9 Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All 8 Denoi Aid (μm) Cont β-al β-al (5) (1) (33) (65) (1) Vl Vl (12) All Figure 5. The onentrtion of (A) hexnoi id, (B) otnoi id nd (C) denoi id in model wines produed fter the ddition of mino ids s indited on the x-xis. The numer in rkets represents the onentrtion dded in mg L 1. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,, ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y Tukey s post ho test. The onentrtion of three undnt ette esters ws lso determined in the model wines produed fter mino id dditions to the model must (Figure 6). Both ethyl ette nd phenylethyl ette

13 Moleules 215, prodution ws higher in the wines produed fter β-lnine ddition to the musts, either lone or in omintion with the other three mino ids, ompred to ontrol wine. However, isomyl ette onentrtions were not signifintly higher in the wines produed fter β-lnine ddition ompred to the ontrol. This suggests tht isomyl lohol onentrtions my e limiting prodution of the orresponding ette ester, ut the onentrtion of the lohol moiety of the ette ester is not limiting for ethyl ette nd phenylethyl ette prodution. Presumly, inresed CoA synthesis used y the presene of β-lnine in the must n then inrese prodution of these ette esters during fermenttion. (A) Ethyl Aette (μm) (B).2.18 Cont β-al (5) β-al (1) (33) (65) (1) Vl Vl (12) All Phenylethyl Aette (μm) (C) Isomyl Aette (μm) Cont β-al β-al (5) (1) (33) (65) (1) Vl Vl (12) All.5 Cont β-al β-al (5) (1) (33) (65) (1) Vl Vl (12) All Figure 6. The onentrtion of (A) ethyl ette, (B) isomyl ette nd (C) phenylethyl ette in model wines produed fter the ddition of mino ids s indited on the x-xis. The numer in rkets represents the onentrtion dded in mg L 1. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,, ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y Tukey s post ho test.

14 Moleules 215, β-alnine Conentrtions in Grpes re Suffiient to Stimulte Ester Prodution The previous experiment demonstrted tht dditions of 5 nd 1 mg L 1 β-lnine ould inrese ester onentrtions in model fermenttions, ut ysteine, rginine nd vline did not hve signifint effet. The onentrtions of β-lnine dded to the must were ove the mximum reported in the literture for V. vinifer (33 mg L 1 ; [39]) or Vitis lrusn (13 mg L 1 ; [41]). A olletion of 5 Cernet Suvignon grpe prels otined from eight growing regions ross two vintges ws nlysed for β-lnine ontent to onfirm tht similr onentrtions would e found in Austrlin-grown fruit. The onentrtion of β-lnine in these smples rnged from 14 to 66 mg L 1 with no pprent pttern sed on vintge or growing region. Given tht some of these vlues were elow the lowest onentrtion of β-lnine tested in the experiment ove (5 mg L 1 ), lower onentrtions of β-lnine were ssessed for their ility to stimulte ester prodution. The results indited tht signifint stimultion of ethyl hexnote prodution ws hieved with the ddition of 1 mg L 1 β-lnine to the model must, nd ny further inrese did not signifintly inrese ethyl ester prodution further (Figure 7A). This result ws the sme for ethyl otnote nd ethyl denote prodution (dt not shown). Hexnoi id, otnoi id nd denoi id lso umulted to higher onentrtions in the wines produed fter the ddition of 1 mg L 1 or more β-lnine dt is shown for hexnoi id (Figure 7B). Similrly, ethyl ette ws present in higher onentrtions in wines produed fter ny β-lnine ddition ompred to the ontrol tretment (Figure 7C), lthough the umultion of phenylethyl ette nd isomyl ette ws not signifintly different from the ontrol tretments for some of the low β-lnine onentrtions (dt not shown). This suggests tht 1 mg L 1 or less of β-lnine is suffiient to stimulte the prodution of greter mounts of MCFAs nd their orresponding ethyl esters, ut the ddition of more β-lnine does not signifintly inrese this prodution further. It is possile tht 1 mg L 1 of β-lnine is enough to ount for the requirements of the yest for mximum ftty id nolism nd ny extr is utilised for other forms of nitrogen metolism. The effet of β-lnine dditions on ette ester prodution is proly omplited y the limittion of the lohol moiety of the ester in the se of isomyl ette nd phenylethyl ette; in the se of ethyl ette, where the lohol is not limiting, the umultion is similr to tht seen for the ethyl esters (Figure 7) β-alnine or Pntothente Supplementtion Hve Similr Effet on Ester Prodution To test whether the effet of β-lnine ddition to the model must hd similr effet s dding pntothente, nother set of fermenttions ws prepred with either no dditions, ddition of 1 mg L 1 β-lnine (1.12 mm), ddition of 267 mg L 1 pnthothente (1.12 mm), nd ddition of oth β-lnine nd pntothente (sme onentrtion s ove).

15 Moleules 215, (A).9 Ethyl Hexnote (μm) (B) Hexnoi Aid (μm) Cont β-lnine (mg/l) (C) 2 25 Cont β-lnine (mg/l) Ethyl Aette (μm) Cont β-lnine (mg/l) Figure 7. The onentrtion of (A) ethyl hexnote, (B) hexnoi id nd (C) ethyl ette in model wines produed fter the ddition of inresing mounts of β-lnine s indited on the x-xis. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,, ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y Tukey s post ho test.

16 Moleules 215, (A).9 Ethyl hexnote (μm) (B) Hexnoi id (μm) (C) Cont β-al Pnto β-al/pnto Cont β-al Pnto β-al/pnto 2 Ethyl ette (μm) Cont β-al Pnto β-al/pnto Figure 8. The onentrtion of (A) ethyl hexnote, (B) hexnoi id nd (C) ethyl ette in model wines produed fter the ddition of β-lnine (β-al), pntothente (Pnto) or oth ompounds (β-al/pnto) s indited on the x-xis. Histogrms represent the men vlues (n = 3) nd error rs show stndrd errors. Different letters (e.g.,,, ) denote signifint differenes etween tretments t p <.5 using ANOVA followed y Tukey s post ho test. Ethyl hexnote prodution in the fermenttions ws stimulted to the sme degree y the ddition of β-lnine, pnthothente or oth ompounds to the MGJM (Figure 8A). This ws lso oserved for ethyl denote, lthough ethyl otnote onentrtions were not signifintly different to the ontrol when only pntothente ws dded to the must (dt not shown). Hexnoi id onentrtions were signifintly greter in the wines produed from ll three tretments ompred to the ontrol (Figure 8B)

17 Moleules 215, nd otnoi nd denoi id responded in the sme mnner to the supplementtion of the must (dt not shown). Of the ette esters, ethyl ette prodution ws signifintly inresed in response to the ddition of pntothente to the model must, ut even greter prodution ws oserved when β-lnine ws dded (Figure 8C). When isomyl ette nd phenylethyl ette were exmined, only those wines produed fter the ddition of β-lnine to the must hd greter mounts of these esters ompred to the ontrol (dt not shown). In generl, oth β-lnine nd pnthothente ould stimulte ester prodution in the model fermenttion system used in these experiments. However, in some ses, β-lnine ws more effiient thn pntothente t inresing ester onentrtions. This oservtion seems ounterintuitive given tht pntothente is produed from pntote nd β-lnine during CoA iosynthesis (Figure 3). However, there my e differenes in the ility of yest to trnsport these ompounds into the ell whih in turn ould influene CoA nd ester prodution. It is known tht pntothente is trnsported y the Fen2p symporter, wheres β-lnine is trnsported minly y the generl mino id permese Gp1p, lthough other trnsporters re involved [42]. The redundny in the β-lnine trnsport system my mke its trnsport more effiient nd hene the onentrtion in the yest ytoplsm my e greter thn tht of pntothente when oth re t the sme onentrtion in the medi. Alterntively, β-lnine my e used in other metoli pthwys tht n lso stimulte ester prodution independent of CoA iosynthesis. A survey of pnthothente onentrtions in grpes grown in the Pifi northwest of the USA found tht they rnged from.18 to 1.26 mg L 1 [43]. This is lower thn the vlues reported for β-lnine (this mnusript, [39]) whih would suggest tht β-lnine my hve more importnt role in CoA prodution during fermenttion of grpes thn pntothente. 3. Experimentl Setion 3.1. Mteril Chemils Commerilly ville ompounds used for fermenttion spiking were purhsed from Sigm-Aldrih (Cstle Hill, NSW, Austrli) nd used without further preprtion. Puttive preursor ompounds were synthesised from ommerilly ville regents nd other hemils outlined elow, suh s mino ids, slts nd sugrs, were nlytil regent grde (Sigm-Aldrih). All solvents used were nlytil regent (Sigm-Alrih) or HPLC grde (Merk, Kilsyth, VIC, Austrli) nd wter ws otined from Milli-Q purifition system (Millipore, North Ryde, NSW, Austrli). Solutions were % v/v with the lne mde up with Milli-Q wter, unless otherwise noted Nuler Mgneti Resonne (NMR) Proton ( 1 H) nd ron ( 13 C) NMR spetr were reorded with Bruker Avne III spetrometer operting t 6 MHz for proton nd 15 MHz for ron nulei. Chemil shifts were reorded s δ vlues in prts per million (ppm). Spetr were quired in methnol-d4 t mient temperture, using the solvent residul pek s internl referene, nd resonnes were ssigned y routine 2D orreltion experiments.

18 Moleules 215, High-Resolution Mss Spetrometry (HR-MS) Spetr were otined on Bruker mirotof-q II with eletrospry ioniztion (ESI) in positive mode. Smples dissolved in 2:1 methnol/wter t onentrtions of pproximtely 1 2 mg L 1 were nlysed y flow injetion HPLC-MS Instrumenttion Anlysis of synthesised ompounds ws undertken using ThermoFinnign Surveyor HPLC onneted to ThermoFinnign LCQ De XP Plus mss spetrometer. Eletrospry ionistion in positive ion mode ws used nd dt quisition nd proessing were performed using Xliur softwre (version 1.3, Thermo Sientifi, Sn Jose, CA, USA) GC-MS Instrumenttion Anlyses of syntheti stndrds nd fermenttion esters were onduted with n Agilent (Plo Alto, CA, USA) 789A gs hromtogrph, equipped with Gerstel (Mülheim n der Ruhr, Germny) MP2 utosmpler fitted with DVB/CAR/PDMS SPME fire (2 m, 5/3 μm, Supelo, Bellefonte, PA, USA), oupled to n Agilent 5975C mss spetrometer s desried previously [19]. Seleted ion monitoring (SIM) ws used for detetion of nlytes for quntittion nd sn mode (m/z 35 35; sn rte, 4.45 sns/s) ws used for synthesised stndrds. Compounds were identified y ompring their mss spetr to spetrl lirries nd from the nlysis of uthenti referene ompounds Melting Points A Buhi melting point B-54 unit ws used, nd melting points were unorreted Synthesis of Compounds d5-ethyl Nonnote To stirred solution of d6-ethnol (5 μl, 8.64 mmol) in nhydrous CH2Cl2 (4 ml) under N2 t room temperture ws dded pyridine (614 μl, 7.59 mmol) followed y 4-dimethylminopyridine (DMAP) (.16 g,.868 mmol). After dissolution, nonnoyl hloride (1.5 ml, 8.32 mmol) ws dded drop-wise to the solution t C, using the formtion of white preipitte. Additionl nhydrous CH2Cl2 (3 ml) ws dded nd the solution ws stirred overnight t room temperture, efore eing quenhed with NHCO3 (1 ml). The phses were seprted, nd the queous phse ws extrted with CH2Cl2 (3 2 ml), nd the omined orgni phses were wshed with rine (2 ml), dried (NSO4), nd filtered efore eing onentrted in vuo to yield yellow/rown oil. Purifition y olumn hromtogrphy on sili (85% hexne/etoa, Rf =.45) gve the title ompound (1.41 g, 7.37 mmol, 89%) s olorless oil (97% pure y GC-MS). 1 H-NMR (6 MHz, CD3OD): δ 2.29 (2H, t, J = 7.6 Hz, H2); 1.62 (2H, pp q, J = 7.4 Hz, H3); (1 H, m, H4-8);.88 (3H, t, J = 7. Hz, H9); 13 C-NMR (15 MHz, MeOH-d4): δ (C1); 59.4 (JC-D = 22 Hz, CD2); (C2); (C7); (C4); (C5); (C6); (C3); (C8); 14.9 (C9); (JC-D = 19.4 Hz, CD3). ESI-HRMS

19 Moleules 215, (m/z): Cld for C11H18D5O2 + ([M+H] + ), ; found EI-MS (m/z) (%) 191 (M +, 3), 162 (12), 148 (13), 141 (19), 12 (6), 16 (39), 93 (1), 84 (6), 74 (28), 61 (2), 57 (13), 55 (17) d5-ethyl Dodenote This ompound ws prepred using the proedure desried for d5-ethyl nonnote, using dodenoyl hloride (1.85 ml, 8.29 mmol), d6-ethnol (5 μl, 8.64 mmol), pyridine (614 μl, 7.6 mmol), nd DMAP (15.6 mg, 864 μmol) in CH2Cl2 (13 ml). Purifition y olumn hromtogrphy on sili (85% hexne/etoa, Rf =.59) gve the title ompound (1.85 g, 7.93 mmol, 96%) s olourless oil, whih ws >99% pure (GC-MS). 1 H-NMR (MeOH-d4) δ 2.29 (2H, t, J = 7.6 Hz, H2); 1.62 (2H, pp q, J = 7.2 Hz, H3); (16H, m, H4-8);.88 (3H, t, J = 7. Hz, H12); 13 C-NMR (MeOH-d4) δ (C1); (JC-D = 22.2Hz, CD2); (C13); (C2); (C1); (C4,5); (C6); (C7); (C8); (C9); (C3); (C11); (C12); (C14); 13.2 (JC-D = 19.5 Hz, CD3). ESI-HRMS (m/z): Cld for C14H24D5O2 + ([M+H] + ), ; found, EI-MS (m/z) (%) 233 (M +, 6), 24 (5), 19 (12), 183 (12), 162 (15), 148 (8), 12 (5), 16 (52), 98 (5), 94 (12), 93 (1), 84 (6), 83 (5), 74 (24), 69 (7), 63 (6), 61 (11), 57 (1), 55 (16), 43 (16), 41 (16), 34 (6), 32 (9), 29 (6), 28 (33) (2R/S)-3-Croxy-2-(nonnoyloxy)-N,N,N-trimethylpropn-1-minium (N-Nonnoyl rnitine) Crnitine derivtives were prepred sed on the method of Bøhmer nd Bremer [44]. To stirred solution of trifluoroeti id (TFA) (5 μl) t room temperture ontining (±)-rnitine hloride (1. g, 5 mmol) ws dded dropwise nonnoyl hloride (1.77 g, 1 mmol), t whih time white preipitte formed. The retion ws mintined t 5 C with stirring overnight. After ooling to room temperture, etone (5.5 ml) ws dded, stirring ws mintined for 1 hour nd the preipitte tht formed ws removed vi entrifugtion. Diethyl ether (~3 ml) ws dded dropwise to inipient loudiness, followed y ooling to C. Additionl Et2O (1 ml) ws dded one rystllistion ws well underwy. One rystllistion ws omplete, entrifugtion nd resuspension in Et2O (4 ml) ws used to wsh the rystls (repeted two more times), followed y isoltion nd drying. Purifition y rerystllistion in MeOH/Et2O produed the known [45] title ompound (1.16 g, 3.43 mmol, 69%) s white rystlline solid, whih ws >99% pure (HPLC-MS), mp C. 1 H-NMR (MeOH-d4) δ 5.63 (1H, pp q, J = 6.9 Hz); 4.92 (1H, s); 3.49 (1H, dd, J = 14.4, 8.6 Hz); 3.75 (1H, dd, J = 14.4,.9 Hz); 3.23 (9H, s); (2H, m); 2.39 (2H, m); 1.63 (2H, q, J = 7.35 Hz); (1H, m);.9 (3H, t, J = 7.1 Hz). 13 C-NMR (MeOH-d4) δ , , 69.5, 66.35, 54.68, 37.93, 35.22, 33.11, 3.5, 3.39, 3.31, 25.83, 23.84, (2R/S)-3-Croxy-2-(dodenoyloxy)-N,N,N-trimethylpropn-1-minium (N-Dodenoyl rnitine) Using the sme retion nd isoltion proedure s tht for N-nonnoyl rnitine ove, N-dodenoyl rnitine ws prepred using (±)-rnitine hloride (1. g, 5 mmol) nd dodenoyl hloride (2.19 g, 1 mmol) in (TFA) (5 μl). Purifition y rerystllistion in MeOH/Et2O produed the known [45] title ompound (1.32 g, 3.47 mmol, 69%) s white rystlline solid, whih ws 93% pure (HPLC-MS), mp C. 1 H-NMR (MeOH-d4) δ 5.63 (1H, pp q, J = 6.9 Hz); 3.89 (1H,

20 Moleules 215, dd, J = 14.3, 8.6 Hz); 3.74 (1H, d, J = 14.3 Hz); 3.22 (9H, s); (2H, m); 2.39 (2H, pp t, J = 7.5 Hz); 1.63 (2H, q, J = 7.3 Hz); (16H, m);.9 (3H, t, J = 7. Hz). 13 C-NMR (MeOH-d4) δ , , 69.37, 66.21, 37.77, 35.8, 33.7, 3.74, 3.59, 3.48, 3.42, 3.18, 25.71, 23.75, Methyl 2-(Nonnoylmino)propnote (N-Nonnoyl Alnine Methyl Ester) Amide oupling ws performed ording to the method of St [46] nd St et l. [47] using ronyl diimidzole (CDI). To stirred solution of nonnoi id (322.9 μl, 2 mmol) in nhydrous THF (4.39 ml) under N2 ws dded CDI (324.3 mg, 2 mmol), using effervesene from lierted CO2. Stirring ws ontinued t room temperture for 2 h nd methyl L-lninte (279.2 mg, 2 mmol) ws dded, resulting in seprtion of rown oily residue. Stirring t room temperture ontinued overnight, y whih time white preipitte hd formed. The solvent ws removed in vuo, resulting in the formtion of white rystls in n oily residue, whih were then tken up in Et2O (1 ml), EtOA (2 ml), nd H2SO4 (1 M, 15 ml). The phses were seprted nd the orgni phse ws wshed suessively with H2SO4 (1 M, 3 15 ml), wter (2 ml), nd rine (2 ml). The orgni phse ws dried (MgSO4), filtered, nd onentrted in vuo, produing white rystlline solid, (551.3 mg, 2.27 mmol, 113%) whih ws used rude in the susequent retion. A smll portion ws purified for hrteristion y olumn hromtogrphy on sili (8% CH2Cl2/EtOA, Rf =.51) to give the title ompound s ple yellow powder, whih ws 98% pure (HPLC-MS), mp C. [α] 2 D = +4.8 (.62, CHCl3). 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.61 (1H, pp q, J = 7.2 Hz, H11); 2.2 (2H, t, J = 7.6 Hz, H2); 1.63 (2H, pp q, J = 7.4 Hz, H3); 1.4 (3H, d, J = 7.1 Hz, H12); (1H, m, H4-8);.87 (3H, t, J = 7. Hz, H12); 13 C-NMR (MeOH-d4) δ (C1); (C1); (C13); (C11); (C2); (C7); (C4); (C5); 29.9 (C6); (C3); 22.6 (C8); (C12); 14.6 (C9). ESI-HRMS (m/z): Cld for C13H26NO3 + ([M+H] + ), ; found, (Nonnoylmino)propnoi id (N-nonnoyl lnine) To stirred solution of THF (24.4 ml) ontining N-nonnoyl methyl ester (51.3 mg, 1.24 mmol) ws dded queous NOH (1 M, 9.76 ml) nd the solution ws stirred t room temperture for 5 hours. After djustment to ph 3 with HCl (1 M), the solution ws extrted with CH2Cl2 (3 2 ml). The omined orgni frtions were wshed with wter (2 ml) nd rine (2 ml), efore eing dried (MgSO4), filtered nd onentrted in vuo to produe olourless oil. Purifition y rystllistion in MeOH/H2O gve the title ompound (414.8 mg, 1.81 mmol, 9%) s olourless oil, whih ws >98% pure (HPLC-MS). [α] 2 D = 17.2 (.58, CHCl3). 1 H-NMR (MeOH-d4) δ 8.35 (1H, s, CO2H); (1H, m, NH); 4.58 (1H, pp q, J = 7.1 Hz, H11); 2.24 (2H, t, J = 7.7 Hz, H2); 1.63 (2H, pp q, J = 7.2 Hz, H3); 1.46 (3H, d, J = 7.2 Hz, H2); (1H, m, H4-8);.88 (3H, t, J = 7. Hz, H9); 13 C-NMR (MeOH-d4) δ (C1); (C1); (C11); 36.4 (C2); (C7); (C4); (C5); 29.9 (C6); (C3); 22.6 (C8); 18.6 (C12); 14.6 (C9). ESI-HRMS (m/z): Cld for C12H24NO3 + ([M+H] + ), ; found,

21 Moleules 215, Methyl 2-(Dodenoylmino)propnote (N-dodenoyl lnine methyl ester) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine methyl ester ove, N-dodenoyl lnine methyl ester ws prepred using dodenoi id (817.6 μl, 4 mmol), CDI (648.6 mg, 4 mmol) nd methyl L-lninte (558.3 mg, 4 mmol) in THF (8.79 ml), produing white, wxy solid (1.18 g, 3.88 mmol, 97%). Purifition y rerystllistion ws ttempted in CH2Cl2/hexne nd EtOH/hexne, ut the produt formed n orgnogel, so the rude produt underwent the sponifition proedure desried elow. A smll portion ws purified for hrteristion y olumn hromtogrphy on sili (8% CH2Cl2/EtOA, Rf =.47) to give the title ompound s white rystlline solid, whih ws >99% pure (HPLC-MS), mp C. [α] 2 D = +5. (.6, CHCl3). 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.61 (1H, pp q, J = 7.2 Hz, H14); 3.75 (3H, s, H16); 2.2 (2H, t, J = 7.7 Hz, H3); 1.63 (2H, pp q, J = 7.4 Hz, H2); 1.4 (3H, d, J = 7.1 Hz, H15); (16H, m, H4-11);.88 (3H, t, J = 7. Hz, H12); 13 C-NMR (MeOH-d4) δ (C13); (C1); (C16); (C14); (C2); 31.4 (C1); (C4-9); (C3); (C11); 18.7 (C15); 14.6 (C12). Melting point nd spetrl dt were in omplete ord with those previously reported [48] (Dodenoylmino)propnoi id (N-dodenoyl lnine) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine ove, N-dodenoyl lnine ws prepred using N-dodenoyl lnine methyl ester (1.5 g, 3.69 mmol), THF (64.7 ml), nd queous NOH (1 M, 25.9 ml) to produe white powder. Purifition y rerystllistion in MeOH/H2O gve the title ompound (928.8 mg, 3.42 mmol, 86%) s white powder, whih ws >98% pure (HPLC-MS), mp C. 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.57 (1H, pp q, J = 7.1 Hz, H14); 2.24 (2H, t, J = 7.6 Hz, H2); 1.62 (2H, pp q, J = 7.3 Hz, H3); 1.46 (3H, d, J = 7.2 Hz, H15); (16H, m, H4-11);.88 (3H, t, J = 7. Hz, H12); 13 C-NMR (MeOH-d4) δ (C13); (C1); (C14); (C2); (C1); (C4); (C5); (C6); (C7); (C8); (C9); (C3); (C11); 18. (C15); 14.1 (C12). Melting point nd spetrl dt were in omplete ord with those previously reported [48] Methyl 2-(Nonnoylmino)-3-methylpentnote (N-nonnoyl isoleuine methyl ester) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine methyl ester ove, N-nonnoyl isoleuine methyl ester ws prepred using nonnoi id (645.9 μl, 4 mmol), CDI (648.6 mg, 4 mmol) nd methyl L-isoleuinte (726.6 mg, 4 mmol) in THF (8.79 ml), produing ple rown oil (1.22 g, 3.58 mmol, 9%). A smll portion ws purified for hrteristion y olumn hromtogrphy on sili (85% CH2Cl2/EtOA, Rf =.57) produing the title ompound s ple rown oil, whih ws >98% pure (HPLC-MS). 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.62 (1H, dd, J = 8.6, 5. Hz, H11); 2.22 (3H, t, J = 7.6 Hz, H2); (1H, m, H12); 1.64 (2H, pp q, J = 7.3 Hz, H3); 1.43 (1H, dqd, J = 14.8, 7.4, 4.7 Hz, H3); 1.27 (11H, m, H4-8); 1.17 (1H, dqd, J = 14.8, 7.4, 7.3 Hz, H13);.92 (3H, t, J = 7.4 Hz, H14);.9 (3H, d, J = 6.8 Hz, H15);.88 (3H, t, J = 7. Hz, H9). 13 C-NMR (MeOH-d4) δ (C1); (C1); (C11); 52.6 (C16); (C12); 36.7 (C2); (C7); (C4); (C5); (C6); (C3); (C15); (C8); (C13); 14.8 (C9); (C14). ESI-HRMS (m/z): Cld for C16H32NO3 + ([M+H] + ), ; found,

22 Moleules 215, (Nonnoylmino)-3-methylpentnoi id (N-nonnoyl isoleuine) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine ove, N-nonnoyl isoleuine ws prepred using N-nonnoyl isoleuine methyl ester (.972 g, 3.41 mmol), THF (4.25 ml), nd queous NOH (1 M, 3.4 ml) to produe white solid. Rerystllistion in 1:1 EtOA/hexne produed the title ompound (59 mg, 2.17 mmol, 64%) s fine white rystls tht were 98% pure (HPLC-MS). [α] 2 D = ( 1.135, CHCl3). 1 H-NMR (MeOH-d4) δ (1H, s, CO2H); (1H, m, NH); 4.63 (1H, dd, J = 8.5, 4.7 Hz, H11); (2H, m, H2); (1H, m, H12); 1.63 (2H, pp q, J = 7.4 Hz, H3); 1.5 (1H, dqd, J = 13.4, 7.4, 4.5 Hz, H3); (1H, m, H4-8); (1H, m, H13);.95 (3H, d, J = 6.8 Hz, H15);.94 (3H, t, J = 7.4 Hz, H14);.88 (3H, t, J = 7. Hz, H9). 13 C-NMR (MeOH-d4) δ (C1); (C1); (C11); (C12); (C2); (C7); 29.23, 29.17, (C4,5,6); (C3); 25.3 (C15); (C8); (C13); 14.7 (C9); (C14). ESI-HRMS (m/z): Cld for C15H3NO3 + ([M+H] + ), ; found, Methyl 2-(dodenoylmino)-3-methylpentnote (N-dodenoyl isoleuine methyl ester) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine methyl ester ove, N-dodenoyl isoleuine methyl ester ws prepred using dodenoi id (81.6 mg, 4 mmol), CDI (648.6 mg, 4 mmol), methyl L-isoleuinte (726.6 mg, 4 mmol) in THF (8.79 ml), produing white solid (1.269 g, 3.85 mmol, 96%). A smll portion ws purified for hrteristion y olumn hromtogrphy on sili (85% CH2Cl2/EtOA, Rf =.59) produing the title ompound s white powder, whih ws 98% pure (HPLC-MS). [α] 2 D = (.76, CHCl3). 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.62 (1H, dd, J = 8.6, 5. Hz, H14); 2.22 (2H, pp t, J = 7.6 Hz, H2); (1H, m, H15); 1.63 (2H, pp q, J = 7.3 Hz, H3); 1.44 (1H, dqd, J = 14.8, 7.4, 4.7 Hz, H16); (16H, m, H4-11); 1.16 (1H, dqd, J = 14.8, 7.4, 7.3 Hz, H13);.92 (3H, t, J = 7.4 Hz, H17);.91 (3H, d, J = 6.9 Hz, H18);.88 (3H, t, J = 7. Hz, H9). 13 C-NMR (MeOH-d4) δ (C1); (C3); (C14); 52.7 (C19); (C15); (C2); (C1), 3.92, 29.58, 29.46, 29.3, (C4,5,6,7,8,9); (C3); (C18); (C11); (C16); 14.1 (C12); (C17). ESI-HRMS (m/z): Cld for C19H38NO3 + ([M+H] + ), ; found, (Dodenoylmino)-3-methylpentnoi id (N-dodenoyl isoleuine) Using the sme retion nd isoltion proedure s tht for N-nonnoyl lnine ove, N-dodenoyl isoleuine ws prepred using N-dodenoyl isoleuine methyl ester (1.96 g, 5.98 mmol), THF (4.25 ml), nd queous NOH (1 M, 6.1 ml) to produe white solid. Purifition y rerystllistion in 9:1 MeOH/H2O overnight t 4 C produed n orgnogel, whih dissolved t room temperture, nd vuum filtrtion t C fforded the title ompound (737 mg, 2.35 mmol, 61%) s very fine white rystls tht were >98% pure (HPLC-MS). [α] 2 D = ( 1.2, CHCl3). 1 H-NMR (MeOH-d4) δ (1H, m, NH); 4.63 (1H, dd, J = 8.3, 4.7 Hz, H14); (2H, m, H2); (1H, m, H15); 1.64 (2H, pp q, J = 7.1 Hz, H3); (1H, m, H16); (16H, m, H4-11), (1H, m, H16);.96 (3H, d, J = 6.8 Hz, H18);.94 (3H, t, J = 7.4 Hz, H17);.88 (3H, t, J = 7. Hz, H12). 13 C-NMR (MeOH-d4) δ (C13); (C13); 56.4 (C14); (C15); (C2); (C1),