Production of fermentation aroma compounds by Saccharomyces cerevisiae wine yeasts: e ects of yeast assimilable nitrogen on two model strains

Transcription

1 RESEARCH ARTICLE Prodution of fermenttion rom ompounds y Shromyes erevisie wine yests: e ets of yest ssimille nitrogen on two model strins Frniso M. Crru 1, Krin Medin 1, Lur Frin 1, Edurdo Boido 1, Pul A. Henshke 2 & Edurdo Dellss 3 1 Seion Enologí, Deprtmento de Cieni y Tenologí de Alimentos, Fultd de Quími, Universidd de l Repúli, Montevideo, Uruguy; 2 The Austrlin Wine Reserh Institute, Glen Osmond, Adelide, SA, Austrli; nd 3 Cátedr de Frmognosi y Produtos Nturles, Deprtmento de Quími Orgáni, Fultd de Quími, Universidd de l Repúli, Montevideo, Uruguy Correspondene: Frniso M. Crru, Seion Enologí, Deprtmento de Cieni y Tenologí de Alimentos, Fultd de Quími, Universidd de l Repúli, 118 Montevideo, Uruguy. Tel.: ; fx: ; e-mil: frru@fq.edu.uy Reeived 18 Novemer 27; revised 5 June 28; epted 6 June 28. First pulished online 11 July 28. DOI:1.1111/j x Editor: Ptrizi Romno Keywords Shromyes erevisie ; ssimille nitrogen; fermenttion rom ompounds. Astrt The ontriution of yest fermenttion metolites to the romti profile of wine is well doumented; however, the iotehnologil pplition of this knowledge, prt from strin seletion, is still rther limited nd often ontrditory. Understnding nd modeling the reltionship etween nutrient vilility nd the prodution of desirle rom ompounds y different strins must e one of the min ojetives in the seletion of industril yests for the everge nd food industry. In order to overome the vriility in the omposition of grpe juies, we hve used hemilly defined model medium for studying yest physiologil ehvior nd metolite prodution in response to nitrogen supplementtion so s to identify n pproprite yest ssimille nitrogen level for strin differentition. At low initil nitrogen onentrtions, strin KU1 produed higher quntities of esters nd ftty ids wheres M522 produed higher onentrtions of isoids, g-utyroltone, higher lohols nd 3-methylthio-1-propnol. We propose tht lthough strins KU1 nd M522 hve similr nitrogen onsumption profile, they represent useful models for the hemil hrteriztion of wine strins in reltion to wine qulity. The differentil prodution of rom ompounds y the two strins is disussed in reltion to their pity for nitrogen usge nd their impt on winemking. The results otined here will help to develop trgeted metoli footprinting methods for the disrimintion of industril yests. Introdution The role of yests in the fermenttion of sugrs into lohol nd ron dioxide hs een known for more thn two enturies (Anderson, 1989). However, well over hlf entury elpsed efore the role of yest strin in the prodution of different wines ws estlished y Psteur in 1866 (Psteur, 1866). While the prinipl yest used in tody s food nd loholi everge industries for the prodution of red, eer, spirits, ider nd wine is lssified s Shromyes erevisie, it is well reognized tht not ll S. erevisie strins re suitle for the fermenttion proess nd the ility to produe qulity foods nd everges differs signifintly mong them. Beuse lssil, physiologil nd geneti methods re of limited use in the hrteriztion of wine-mking yest strins, reserhers in wine iotehnology hve een serhing for new methods to revel the geneti dpttions of industril yest strins of the sme speies, i.e. S. erevisie (Pretorius & Hoj, 25). On the other hnd, it is well estlished tht S. erevisie produes different onentrtions of rom ompounds s funtion of fermenttion onditions nd must tretments, for exmple, temperture, grpe vriety, mironutrients, vitmins nd nitrogen omposition of the must. Moreover, mny ommeril yests produe undesirle off-flvors, suh s H 2 S, nd high onentrtion of higher lohols, eti id nd ethyl ette, depending on the onentrtion of ssimille nitrogen present in the grpe must (Bell & Henshke, 25). Additionlly, slow, sluggish nd stuk fermenttions hve often een relted to nitrogen defiieny 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

2 Prodution of fermenttion rom ompounds 1197 (Bisson, 1999). In the lte 197s, these fermenttion prolems were prtilly solved y the ddition of mmonium slts to defiient musts, whih not only inresed fermenttion rte ut lso the sensory desirility of the wines (Bell & Henshke, 25). As onsequene, the routine ddition of mmonil nitrogen to musts to orret nitrogen limittion is widely used y enologists in most wine regions of the world. Although tody some useful methods re essile for mesuring the yest ssimille nitrogen (YAN) ontent of grpe musts in wineries, they re not ommonly used, nd so the use nd effet reltionship tht results in wines with n undesirle flvor is not usully orrelted. Initil studies hve ttempted to relte the yest nitrogen demnd onept with the profile of rom ompounds in wines. Prdoxilly, the vrious studies reported to hrterize the yest rom ompounds of wines mde with vrious wine yests hve not onsidered the importne of nitrogen level of the grpe must or the fermenttion medium utilized (Rnkine, 1967; Rnkine & Pook, 1969; Dudt & Ough, 1973; Di Stefno et l., 1981; Crer et l., 1988; Cvzz et l., 1989; Giudii et l., 199; Herriz et l., 199; Mteo et l., 1991; Delteil & Jrry, 1992; Longo et l., 1992; Lurton et l., 1995; Lem et l., 1996; Zoeklein et l., 1997; Antonelli et l., 1999; Vil et l., 2; Romno et l., 23). Moreover, mny studies were onduted under onditions of exess nitrogen, ove 3 mg N L 1 (Houtmn & Du Plessis, 1986; Crru, 23). This experimentl pproh hs, however, not proved to e suitle euse S. erevisie yest strins tht show similr profiles of ssimille nitrogen onsumption n nevertheless produe very different profiles of fermenttion rte nd romti ompounds under industril onditions of lower initil nitrogen levels (Jirnek et l., 1991; Crru, 23; Tillndier et l., 27). There re few reports of investigtions into the sensory hrteristis of wines mde with different strins under low nitrogen fermenttion onditions (Crru et l., 1993; Medin et l., 1997; Crru, 23). Only reently, hve the first hemometri studies provided dt on the signifine of yest rom ompounds to pereived flvor in wine (Smyth et l., 25). From the hemil point of view, only few studies hve investigted the yest rom ompounds of wines prepred under defined onditions of YAN ontents similr to winemking onditions (Bosso, 1996; Guitrt et l., 1999; Niolini et l., 2, ; Crru, 23; Bell & Henshke, 25; Beltrn et l., 25; Crru et l., 25; Hernndez- Orte et l., 26; Vilnov et l., 27). Nevertheless, it is diffiult to drw ny firm onlusions onerning the reltionship etween YAN nd the different rom ompounds produed y the yest strins studied due to the different experimentl onditions used. The im of this study ws to hrterize S. erevisie strins for the prodution of key yest rom ompounds (esters, lohols, ids nd ltones) fter fermenttion of hemilly defined medium, whih models the nutrient omposition of grpe must overing wide rnge of initil nitrogen onentrtions. The model yest strins used in this study were seleted from wide group of strins euse oth showed similr nitrogen onsumption when tested under different nitrogen onditions, ut from sensory nd hemil point of view they produed different wines t low nitrogen onentrtion (75 mg N L 1 ). The differentil ehvior of yest strins t different initil nitrogen onentrtions ws exmined for etter understnding of this phenomenon in our model system. Applition of these results for future disrimintion of industril yest strins in reltion to the type of rom ompounds produed will llow the development of dt models for metoli footprinting methods (Kell et l., 25; Crru et l., 28). Mterils nd methods Yest strins Shromyes erevisie strins utilized were: Montrhet UCD 522 (University of Cliforni, Dvis), referred to s M522 in this work, nd KU1 [Uruguyn seleted strin, (Crru et l., 1993; Medin et l., 1997)]. Both strins re used in the ommeril prodution of wine. Inoul were prepred in the sme syntheti medium y inution for 12 h in rotry shker t 15 r.p.m. nd 25 1C. Inoulum size ws ells ml 1 of medium for oth strins. The widely studied wine yest strins, S. erevisie M522 nd KU1, were hrterized s strins hving similr nitrogen demnds (i.e. onsuming similr mount of nitrogen under 5 4 mg N L 1 of YAN). Fermenttion onditions Chemilly defined fermenttion medium (nutrient omponents of grpe juie) ws prepred s desried previously (Henshke & Jirnek, 1993), ut modified s follows: the totl nitrogen ontent ws djusted to si mount of 5 mgn L 1 with eh mino id nd mmonium omponent dded in the sme proportions s indited previously (Henshke & Jirnek, 1993). Medi with YAN onentrtions of 75, 125, 18, 25 nd 4 mg N L 1 were mde y inresing the si onentrtion y supplementtion with dimmonium phosphte (DAP). None of these YAN mounts ws limiting onentrtion for omplete fermenttion of sugrs y the yest strins used. The finl ph of eh medium ws djusted to 3.5 with HCl. Equimolr onentrtions of gluose nd frutose were dded to reh 12 g L 1 nd the mixed vitmins nd slts s desried previously (Henshke & Jirnek, 1993). Tween 8 ws exluded from the medium euse it ws not found to e neessry for omplete fermenttion nd it hd negtive FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218

3 1198 F.M. Crru et l. impt on the sensory hrteristis of the resultnt wines. Ergosterol ws dded s the only supplemented lipid t finl onentrtion of 1 mg L 1. Fermenttions were rried out in 125 ml of medium ontined in 25-mL Erlenmeyer flsks, losed with Muller vlves, filled with pure sulfuri id. YAN ws hosen s the vrile for this investigtion euse it ws found previously tht this ftor signifintly ffeted prodution of fermenttion rom ompounds y these yests under similr experimentl onditions (Crru, 23). Stti th fermenttion onditions were onduted t 2 1C in triplite, simulting winemking onditions. Fermenttion tivity ws mesured s CO 2 weight loss nd expressed in grms per 1 ml, nd totl residul sugrs were nlyzed using the Fehling method (Zoeklein et l., 1995). One dy, smples were tken to mesure ell growth in n improved Neuuer hmer. Smples for sensory nd GC-MS nlysis were tken 2 dys fter the end of fermenttion, filtered through.45-mm pore memrnes nd SO 2 ws dded s 5 mg L 1 of sodium metisulfite. Sensory nlysis Fermenttion produts were sujeted to sensory nlysis through pired preferene test. Fermenttions t 75 mg N L 1 were presented to trined pnel of persons, in order to determine the signifine of the sensory differenes etween the strins. Smples fermented with oth strins with or without ddition of YAN in the hemilly defined medium ontining the si onentrtion of nitrogen (75 mg N L 1 ) were presented to group of five winemkers so s to evlute rom defets nd positive hrteristis. A free desription of desirility nd rom hrteristis ws presented in the tsting sheet. GC nd GC-MS nlysis Higher lohols Higher lohol nd ethyl ette nlysis ws performed y distilltion of 5 ml of smple nd diret injetion of.5 ml of smple of the distillte nd nlyzed using the GC-FID with glss olumn (2 m 2 mm; Cropk C, 6 8 mesh,.2% CW 15, Supelo) in Shimtzu C-17, equipped with EZ-CHROM softwre. The experimentl onditions were s follows: progrm temperture 65 1C (5 min), C t41c min 1 ; injetor temperture 2 1C; nd detetion temperture, 25 1C. The rrier gs ws nitrogen (2 ml min 1 ). Arom voltile ompounds Extrtion of rom ompounds ws performed using dsorption nd seprte elution from n Isolute (IST Ltd, Mid Glmorgn, UK) ENV1 rtridge pked with 1 g of highly ross-linked styrene-divinyl enzene (SDVB) polymer. Tretment of smples nd GC nlysis were performed s desried previously (Boido et l., 23). Identifition nd quntifition The omponents of wine roms were identified y omprison of their liner retention indies, with pure stndrds or dt reported in the literture. Comprison of mss spetrl frgmenttion ptterns with those stored on dtses ws lso performed. GC-FID nd GC-MS instrumentl proedures using n internl stndrd (1-heptnol) were pplied for quntittive purposes, s desried previously (Boido et l., 23). Sttistil nlysis A stepwise disriminnt nlysis ws rried out with the rom ompounds nlyzed of the 12 wines produed using oth strins in triplite in si fermenttion medium ontining initil YAN of 75 mg N L 1, with nd without n extr YAN ddition of 63 mg N L 1 s DAP. ANOVA for initil YAN onentrtion, yest strin nd rom ompounds were determined with STATISTICA 5.1. Differenes of free voltile ompounds were evluted; the men rting nd lest signifint differenes for initil YAN onentrtions for eh strin were lulted from n ANOVA. Results A preferene sensory test ws performed where signifint differenes were found etween fermenttions of KU1 nd M522 (P o.1) t low nitrogen onentrtion (75 mg N L 1 ), while no signifint differenes were found etween them t higher nitrogen onentrtion (4 mg N L 1 ). At low nitrogen onentrtion, wines were desried s fruity nd plesnt for KU1 nd unplesnt (sopy nd swety) for M522, onfirming our previous work with these two strins (Crru, 23). Tle 1 shows the odor tive vlues (OAV) of the min rom ompounds nlyzed for these two nitrogen onentrtions. On the left side, it n e seen tht from the 7 ompounds tht resulted in higher OAVs for KU1 t low YAN level, four were desried s fruity (esters) nd the others were mediumhin ftty ids, onsidered s preursors of fruity esters. Conversely, seven ompounds tht resulted in lower OAVs for KU1 t this YAN level were the higher lohols nd isoids, orresponding to unplesnt rom desriptions. At the higher nitrogen onentrtion, lthough wines ould not e differentited y the sensory nlysis, M522 produed high onentrtions (OAVs) of prtilly ll the ompounds, inditing tht hemilly, strins hve very 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

4 Prodution of fermenttion rom ompounds 1199 Tle 1. Averge OAV for the studied fermenttions with the model strins t two YAN levels (75 nd 4 mg N L 1 ) OAVs At 75 mg N L 1 At 4 mg N L 1 Referene Threshold Desriptor Mont. 522 KU1 Mont. 522 KU1 Isomyl ette 1 3 Bnn, per Ethyl ette 2 75 Fruity, solvent ethyl hexnote 1 14 Apple, fruit Ethyl otnote 1 5 Pinepple, per, sopy Ethyl denote 1 2 Grpe Methyl-1-propnol 1 4 Wine, solvent Methyl-1-utnol 2 3 Whiskey, mlt, urned Methyl-1-utnol 1 3 Alohol, nil polish Phenylethyl lohol 1 14 Honey, rose, spiy Propnol 3 83 Alohol, pungent.... Isoutyri id 4 5 Rnid, utter, heese Isovleri id 1 33 Swet, id, rnid Butnoi id Rnid,heese, swet Hexnoi id 1 42 Swet, id, rnid Otnoi id 1 5 Swet, heese Denoi id 1 1 Rnid, ftty g-butyroltone 3 1 Crmel, oonut.... Ethyl-4-hydroxyutnote 5 ND Crmel ND ND ND ND -Phenylethyl ette 3 25 Rose, honey, too Methylthio-1-propnol 1 1 Sweet, potto For ref.: 1, the threshold determintion ws performed in 11% wter/ethnol solution ontining 7 g L 1 glyerol, 5 g L 1 trtri id nd ph djusted to 3.4 with 1 M NOH (Ferreir et l., 2); 2, Etievnt (1991); 3, the threshold determintion ws performed in 14% ethnoli solution (Moreno et l., 25); 4, the threshold determintion ws performed in wter (vn Gemert, 23); 5, Crru (23). ND, not determined. different ehvior t this nitrogen onentrtion. On the other hnd, ddition of YAN to the low nitrogen onentrtion (75 mg N L 1 )of63mgnl 1 s DAP (resulting in finl YAN of 138 mg N L 1 ) shows tht the M522 strin requires higher nitrogen onentrtion to produe similr rom profile when ompred with KU1, s shown in Fig. 1. A disriminnt nlysis using 28 ompounds determined using GC-MS shows ler disrimintion in three fermenttion groups. The disriminnt ompounds in this nlysis were: -phenylethyl lohol, -phenylethyl ette, the isoutnoi (isoc4) nd isovleri ids (isoc5), ethyl otnote nd 2-ethyl hexnol. The YAN ddition to M522 resulted in fermenttions with response similr to tht of unsupplemented KU1, s the disriminnt nlysis ould not seprte these two ses. Conditions of growth nd fermenttion tivity ould ffet the profile of voltile yest metolites tht ontriute to wine rom. The growth nd fermenttion kinetis of the two strins were first hrterized over the wide rnge of nitrogen onentrtions utilized in this study (5 4 mg N L 1 ). No signifint differenes in ell growth were found etween strins (Fig. 2), ut signifintly smller totl popultion ws found for oth strins in fermenttions t n initil YAN of 75 mg N L 1. Fermenttion tivity (CO 2 evolution rte) of the two strins ws found to depend on the initil YAN onentrtion up to 125 mg N L 1, (Fig. 2). All fermenttions were ompleted (finl residul sugrs elow 2 g L 1 ) t low nitrogen onentrtions nd fermenttion urves were not signifintly different for strins t higher YAN onentrtions (dt not shown). Prodution of higher lohols nd isoids in response to initil nitrogen onentrtion The sum of the higher lohols, 2-methyl-1-propnol, 2-methyl-1-utnol nd -phenylethyl lohol is depited in Fig. 3. Both strins produed similr profile for higher lohols in reltion to initil nitrogen, exept tht t eh nitrogen onentrtion, strin M522 produed signifintly higher onentrtions of these ompounds. At low initil nitrogen onentrtions (5 75 mg L 1 ), diret reltionship ws oserved wheres t higher initil nitrogen onentrtions n inverse reltionship existed. Aove. 25 mg N L 1, the prodution of higher lohols rehed plteu. 3-Methyl-1-utnol, whih ws produed t onsiderly higher onentrtion thn the remining lohols, showed slightly different prodution profile (Fig. 3) in tht its onentrtion slightly hnged up to 2 mg N L 1 nd ove this onentrtion n inverse reltionship existed FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218

5 12 F.M. Crru et l. Root Root 1 up to the highest nitrogen onentrtion tested (4 mg L 1 ). The prodution of isoids y the two yest strins in response to initil nitrogen ws lso studied. Figure 3, whih shows the sum of isoutnoi (isoc4) nd isovleri (isoc5) ids, indites ehvior similr to higher lohols for oth strins. Strin M522 produed signifintly higher onentrtions of these ompounds thn KU1. Interestingly, n inverse reltionship etween growth (Fig. 2) nd higher lohols nd isoids profiles (Fig. 3 nd ) ws shown. Prodution of 1-propnol in response to different initil YAN onentrtions M522 KU1+YAN M522+YAN KU1 Fig. 1. In the following experiment, the two seleted strins were fermented with nd without the ddition of 63 mg N L 1 of YAN. The grph shows the disriminnt nlysis, where 28 ompounds determined using GC-MS were used. The ompounds tht were disriminnt in this nlysis were: -phenylethyl lohol, -phenylethyl ette, the isoutnoi nd isovleri ids, ethyl otnote nd 2-ethyl hexnol. A ler exeption mong the higher lohols is 1-propnol s shown in Fig. 4. In this se, positive reltionship etween 1-propnol prodution nd initil YAN ws oserved ross the whole rnge of nitrogen onentrtions studied. In ontrst to the trends shown y the other higher lohols, the ehvior of eh strin is signifintly different t moderte to high YAN onentrtions exeeding. 15 mg N L 1. Strin KU1 produed higher onentrtions of 1-propnol t moderte to higher onentrtions of YAN. Prodution of esters nd ftty ids in response to different initil YAN onentrtions The prodution of esters (sum of isomyl ette, ethyl hexnote, ethyl otnote nd ethyl denote) nd ftty ids (C4, C6, C8 nd C1) y the two strins M522 nd KU1 is shown in Fig. 5. These results indite similr ehvior for the two groups of ompounds; however, the two strins produed different ptterns in reltion to the initil YAN onentrtion. KU1 produed higher onentrtion of these ompounds t lower YAN onentrtions reltive to M522 up to 125 mg N L 1, eyond whih signifintly lower onentrtions of voltiles were produed y KU1. Although ethyl ette showed profile similr to those oserved for the other esters studied, it ws produed t reltively higher onentrtion nd so its response pttern is presented seprtely in Fig. 5. Interestingly, s n e seen, the ehvior of esters nd ftty ids, minly the ethyl ette, is the reverse of tht oserved for 1-propnol for eh strin (see Fig. 4). KU1 showed reverse ehvior ompred with M522 in these ompounds. Other esters nd lohols Phenylethyl ette is presented seprtely euse its formtion pttern, in response to the initil nitrogen onentrtion in the medium, differed from tht for the other esters, showing more similr profile to those of the higher lohols. While 3-methylthio-1-propnol showed response profile similr to the other higher lohols, s shown in Fig. 2, the ehvior of the two strins ws signifintly different for these ompounds s, shown in Fig. 6. KU1 produed lower onentrtions of these ompounds t ll initil YAN onentrtions. Prodution of -utyroltone nd ethyl- 4- hydroxyutnote in response to initil YAN onentrtions Figure 7 shows g-utyroltone prodution, one of the min ltones present in wines, whih showed ehvior similr to the higher lohols. M522 produed higher onentrtions of this ltone t ll YAN onentrtions, with mximum of 125 mg N L 1 ompred with 75 mg NL 1 for the higher lohols. Figure 7 shows the ehvior of ethyl-4-hydroxyutnote, ompound tht ould e metolilly relted to g-utyroltone s disussed elow. The profile of ltone prodution y M522 ws onsistent with this similr metoli reltionship etween the two ompounds. However, this is not the se for KU1, where reverse ehvior is oserved t YAN of 125 mg N L 1. Disussion Reserhes on the hemil identifition of rom ompounds in wine derived from the metoli tivity of yests hve een reported widely in the literture during the lst dedes (Houtmn & Du Plessis, 1986; Rpp & Versini, 1996; Lmrehts & Pretorius, 2; Swiegers et l., 25). From these studies, it n e onluded tht vrious fermenttion produts, inluding ethyl esters, ettes, higher lohols, ftty ids nd thiols, re espeilly importnt to the sensory pereption of different wine types (Ferreir et l., 1996; Guth & Sies, 22; Smyth et l., 25; Cozzolino et l., 26). While this reserh provides importnt informtion on the sensory signifine of yest voltile 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

6 Prodution of fermenttion rom ompounds Growth response to YAN Totl Cells x KU1 M M522 7 KU1 6 6 g CO 2 1 ml mg L 1 M mg L 1 M mg L 1 M Dys g CO 2 1 ml Dys 5 mg L 1 KU1 75 mg L 1 KU1 125 mg L 1 KU Fig. 2. () Growth response to different nitrogen onentrtions for M522 nd KU1 strins. Results re the verge of duplite fermenttions. () Fermenttion rte urves, determined y ulture weight loss due to CO 2 evolution for M522 nd KU1 t low YAN onentrtions (5 125 mg N L 1 ). Results re the verge of duplite fermenttions nd SEs of the men o 5%. ompounds, more foused reserh is required on the physiologil regultion of these ompounds for the enefit of the loholi everge industries, suh s wine nd eer prodution. A lk of reognition of the importne of defining the nitrogen ontent of medi in reltion to rom ompounds hs produed onsiderle disrepnies nd misunderstndings in the literture. The results presented in this work will filitte etter understnding of the importne of the pproprite YAN onentrtion for yest hrteriztion in reltion to rom ompounds. Tle 1 shows the OAV of the ompounds tht re shown in the different figures. Although most of the ompounds studied showed signifint hemil nd OAV differenes etween strins, some of them my not ontriute to the sensory hrteristis of wine t the studied onentrtions. This is the se for ethyl denote, 1-propnol nd g- utyroltone. However, from the hemil point of view some interesting hnges in the ehvior of oth strins for ll the ompounds studied my ontriute to hemil yest disrimintion if n pproprite nitrogen onentrtion is used in hemilly defined medium. In the se of higher lohols s prodution, exept for 1-propnol, the response to the initil nitrogen onentrtion vries etween reports, ut, in generl, is desried y n inverse reltionship. However, nitrogen-dependent iphsi response pttern of higher lohol prodution involving diret positive reltionship t low initil nitrogen onentrtions nd n FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218

7 122 F.M. Crru et l. () 5 4 Higher lohols Propnol 1 1 mg L () mg L () 12 mg L 1 Isoids C4 nd C d d 1 d d d inverse reltionship t higher initil onentrtions ws desried for eer yests mny yers go y Ayrp (1967, 1968, 1971). Interestingly, little ttention hs een direted towrd this phenomenon sine these studies were pulished. In ft, most of the reent studies were performed using higher onentrtions of nitrogen (exeeding 1 mg N L 1 ), with whih only n inverse response is usully oserved Methyl 1-utnol d Fig. 3. Reltionship etween the umultion of higher lohols nd isoids in wines mde with two Shromyes erevisie yest strins in response to the initil nitrogen onentrtion. () Sum of 2-methyl-1- propnol, 2-methyl-1-utnol nd -phenylethyl lohol. () Sum of the isoutnoi nd isovleri ids nd () the mjor lohol 3-methyl-1- utnol. These ompounds were produed y yest strins M522 ( ) nd KU1 (B) in n rtifiil grpe must medium nd mesured using GC nd GC-MS s desried in Mteril nd methods. Letters t eh point indite the signifint differenes (P o.5) ording to n LSD test of ANOVA lulted for eh strin. Error rs indite SD. 5 mg L Moreover, we elieve tht our report is the first to desrie omprehensively the ehvior of isoid prodution in response to different initil nitrogen onentrtions. The nitrogen-dependent ommon trend for prodution of isoids nd higher lohols suggests tht their metolism nd prodution my e oordinted s reported previously (Nordstrom, 1964; Rezin et l., 197). On the other hnd, we suggest tht the signifintly higher prodution of higher lohols nd isoids y M522 ould reflet less effiient usge of nitrogen, resulting in n inrese of ron flux relted to rnhed-hin mino id metolism y this strin. These model strins would e useful for further studies for etter understnding of the different geneti (BAT nd ARO genes) nd/or physiologil [NAD(P)H/NAD(P) lnes] dpttions tht ould explin the differenes found here. Our results suggest tht KU1, whih produes less higher lohols nd isoids t ll the nitrogen onentrtions tested, when ompred with M522, might regulte more effetively the ron flux t ny given nitrogen vilility resulting in the exretion of less quntities of ron metoli wstes (Riereu-Gyon et l., 2) from the ell t ll initil nitrogen onentrtions. In ddition, the diret positive reltionship of these ompounds t low initil onentrtion ould e explined y the very low yest fermenttion tivity in this sitution, s shown in Fig. 2, whih is indued y the low iosyntheti pity of the ell. These results re in greement with the first gene expression nlysis onduted in n industril strin (Bkhus et l., 21), where, t low nitrogen level (53 mg N L 1 ) ompred with high nitrogen level (4 mg N L 1 ), ultures disply greter expression of genes involved in trnsltion nd in oxidtive ron metolism, Fig. 4. Reltionship etween 1-propnol prodution nd initil YAN onentrtion for M522 ( ) nd KU1 (B) yest strins. Letters t eh dt point indite the level of signifint differene (P o.5) ording to n LSD test of ANOVA lulted for eh strin. Error rs indite SD of the men vlue. 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

8 Prodution of fermenttion rom ompounds 123 suggesting tht respirtion is more nitrogen onserving thn fermenttion (Bkhus et l., 21). In ontrst to other higher lohols, 1-propnol is known to e formed y the ondenstion of pyruvi id nd etyl CoA (Nyknen, 1986). Our results re in greement with previous studies (Ayrp, 1968; Mrgheri et l., 1984); however, the formtion of 1-propnol vried in reltion to nitrogen level for the two strins (Fig. 4). Strin M522 produed reltively higher onentrtions of 1-propnol t lower nitrogen onentrtions ut reltively less t higher nitrogen levels. This reversl in prodution in reltion to nitrogen vilility does not pper to hve een reported efore. Moreover, the ft tht the reltive onentrtion of 1-propnol produed in response to nitrogen is generlly reversed for ethyl ette with respet to eh of the strins would e n interesting topi for further reserh. In ddition, our results indited n indiret reltionship etween higher lohol nd isoid prodution nd growth (see Figs 2 nd 3). The effet of some romti lohols, suh s tryptophol, isomyl lohol nd -phenylethyl lohol, ting s uto-signling moleules ple of stimulting morphogenesis in S. erevisie (Hzelwood et l., 28), my explin this ehvior. Furthermore, quorumsignling pthwy linking environmentl sensing nd entry into the sttionry phse in S. erevisie hs een desried reently (Chen & Fink, 26). As n e seen in our dt, these results might help to explore the ourrene of other puttive quorum-sensing moleules in S. erevisie. Prodution of esters nd ftty ids Yests synthesize ftty ids y the hydrolysis of the yl- CoA derivtives nd esters y esterifition of tivted ftty ids nd lohols. The results otined here showed ler metoli orreltion etween these ompounds (Fig. 5). Aette esters re desried s hving fruity roms nd re onsidered s plesnt flvors from sensory point of view (Smyth et l., 25). The results otined in the present work my provide n explntion for the more plesnt sensory hrter of the wines otined with KU1 t low nitrogen onentrtion ompred with M522, whih re lso in greement with the OAVs shown in Tle 1. The ehvior of strin KU1, in whih higher onentrtions of these ompounds re produed when nitrogen vilility is low, ontrdits the onept rised in mny reports stting tht the inrese in ester prodution is diretly relted to the inrese of nitrogen in the must. On the other hnd, this ontrditory ehvior of KU1 ould lso explin why severl studies did not oserve onsistent orreltion etween YAN grpe musts nd esters nd ftty id onentrtion (Rpp & Versini, 1996; Lmrehts & Pretorius, 2). Esters nd ftty ids hve lso een onsidered metoli wstes with potentilly toxi effet on the ell () μg L () 3 μg L 1 Esters (Peddie, 199). A limited prodution of these ompounds y KU1 t high nitrogen onentrtion ompred with M522 n e understood s n effiient ehvior of this strin in reltion to nitrogen usge. More interestingly, the profiles of ftty ids nd esters otined with KU1 re quite similr to the higher lohol nd isoid profiles shown in Fig. 3,. Ftty Aids () 3 μg L Ethyl ette Fig. 5. Reltionship etween the prodution of esters nd ftty ids nd initil YAN onentrtion for M522 ( ) nd KU1 (B) yest strins. () Sum of esters (isomyl ette, ethyl hexnote, ethyl otnote nd ethyl denote); () sum of ftty ids (C4, C6, C8 nd C1); nd () ethyl ette. Letters t eh dt point indite the level of signifint differene (P o.5) ording to n LSD test of ANOVA lulted for eh strin. Error rs indite SD of the men vlue. FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218

9 124 F.M. Crru et l. () 1 β-phenyl ethyl ette () 3 γ-utyroltone µg L () 1 5 µg L Methylthio-1- propnol Prodution of -utyroltone, 3-methylthio-1- propnol nd ethyl 4-hydroxyutnote Limited informtion out the prodution of these ompounds y yest n e found in the literture. Although g-utyroltone ws hrterized s eing negtively relted to YAN level in grpe must (Bosso, 1996), our results, whih were otined over very wide rnge of nitrogen onentrtions, showed diret reltionship with low nitrogen levels. Suh low nitrogen levels hve not een onsidered in other reports. The profile of g-utyroltone prodution lso resemles tht of higher lohols nd isoids. The reltion of g-utyroltone with ethyl-4-hydroxyutnote prodution ws proposed previously (Muller et l., 1973); however, n importnt strin differene ws oserved for KU1 in ethyl-4-hydroxyutnote prodution. Reently, 3-methylthio-1-propnol prodution ws desried to e negtively relted to the nitrogen onentrtion of the medium (Moreir et l., 22). However, our results showed different ehvior for M522 t lower nitrogen levels, inditing positive reltion with this thiolohol (Fig. 6), Fig. 6. Prodution of other voltile ompounds tht show vriility in response to the initil YAN onentrtion y M522 ( ) nd KU1 (B) yest strins. () -phenylethyl ester nd () the 3-methylthio-1-propnol. Letters t eh dt point indite the level of signifint differene (P o.5) ording to n LSD test of ANOVA lulted for eh strin. Error rs indite SD of the men vlue. µg L 1 () µg L ethyl 4-hydroxyutnote whih is more onsistent with the generl ehvior of higher lohols shown in Fig. 3. In summry, depending on the yest strin utilized, the formtion of rom ompounds presented different responses when nitrogen ws dded to the medium. Nitrogen ddition ould lso derese ester nd ftty id onentrtion during fermenttion of strin, suh s KU1, phenomenon not desried previously to the est of our knowledge. From hemil point of view, in our rtifiil medium, it ws possile to otin ler disrimintion etween oth model strins evluted through the nlysis of key ompounds produed fter fermenttion like higher lohols, ftty ids, ethyl esters nd ltones. Mjor differenes etween strins in reltion to rom ompound onentrtions were otined t higher nitrogen level (4 mg N L 1 ). Bsed on these two strins nd the pproprite onditions determined in this work, our results will llow the development of dt models for metoli footprinting methods for industril yest strin disrimintion. From iotehnologil point of view, the results otined hd shown etter dpttion of strins suh s KU Fig. 7. Prodution of g-utyroltone () nd ethyl 4-hydroxyutnote () in response to the initil YAN onentrtion y M522 ( ) nd KU1 (B) yest strins. Letters t eh dt point indite the level of signifint differene (P o.5) ording to n LSD test of ANOVA lulted for eh strin. Error rs indite SD of the men vlue. 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

10 Prodution of fermenttion rom ompounds 125 for the fermenttion of grpe vrieties with onsiderly vrile hrter suh s most of the red Vitis vinifer nd some white grpes suh s Suvignon Bln, Chrdonny, Riesling nd Musts. On the ontrry, strins like M522 would e more suitle for fermenttion of neutrl vrieties, in whih fermenttion roms suh s esters nd ftty ids would ontriute towrd improving the fruity intensity of the wine when the nitrogen level of these grpe musts is inresed. Aknowledgements We would like to thnk Prof. Ptrik Moyn nd Dr Crin Gggero. for the ritil omments mde for this mnusript We would lso thnk PDT, INIA nd CSIC of UdelR for finnilly supporting this projet. Referenes Anderson RG (1989) Yest nd the Vitorin rewers: inidents nd personlities in the serh for the true ferment. J. Inst Brew 95: Antonelli A, Cstellri l, Zmonelli C & Rossi C (1999) Yest influene on voltile omposition of wines. J Agri Food Chem 47: Ayrp T (1967) Formtion of higher lohols from 14 C-lelled vline nd leuine. J Inst Brew 73: Ayrp T (1968) Formtion of higher lohols y vrious yests. J Inst Brew 74: Ayrp T (1971) Biosyntheti formtion of higher lohols y yest. Dependene on the nitrogenous nutrient level of the medium. J Inst Brew 77: Bkhus LE, DeRisi J, Brown P & Bisson LF (21) Funtionl genomi nlysis of ommeril wine strin of Shromyes erevisie under differing nitrogen onditions. FEMS Yest Res 1: Bell SJ & Henshke PA (25) Implitions of nitrogen nutrition for grpes, fermenttion nd wine. Aust J Grpe Wine Res 11: Beltrn G, Esteve-Zrzoso B, Rozes N, Ms A & Guillmon JM (25) Influene of the timing of nitrogen dditions during syntheti grpe must fermenttions on fermenttion kinetis nd nitrogen onsumption. J Agri Food Chem 53: Bisson LF (1999) Stuk nd sluggish fermenttions. Am J Enol Viti 5: Boido E, Lloret A, Medin K, Friñ L, Crru F, Versini G & Dellss E (23) Arom omposition of Vitis vinifer v. Tnnt: the typil red wine from Uruguy. J Agri Food Chem 51: Bosso A (1996) Influenz dell ggiunt di dosi resenti di zoto mmonile i mosti sull omposizione in sostnze voltili di origine fermenttiv e sulle prinipli rtteristihe olfttive di luni vini inhi. Riv Viti Enol 3: Crer MJ, Morero J, Orteg JM & Medin M (1988) Formtion of ethnol, higher lohols, esters nd terpenes y five strins in musts from Pedro Ximenez grpes in vrious degrees of ripeness. Am J Enol Viti 39: Crru FM (23) Chrteriztion of yest in reltion to the ility to utilize nitrogen Studies of rom ompounds. Ph.D. Thesis, pp 312. Universidd de l Repuli, Uruguy. Crru FM, Neirotti E & Gioi O (1993) Stuk wine fermenttions: effet of killer/sensitive yest intertions. J Ferment Bioengin 76: Crru FM, Medin K, Boido E, Frin L, Gggero C, Dellss E, Versini G & Henshke PA (25) De novo synthesis of monoterpenes y Shromyes erevisie wine yests. FEMS Miroiol Lett 243: Crru FM, Boido E & Dellss E (28) Terpenoids in grpes nd wines: origin nd mirometolism during the vinifition proess. Nt Produt Commun 3: Cvzz A, Versini G, Dll Serr A & Romno F (1989) Chrteriztion of six S. erevisie strins on the sis of their voltile ompound prodution, s found in wines of different rom profiles. Proeedings of the 7th Interntionl Symposium on Yests, pp John Wiley & Sons, Chihester, UK. Chen H & Fink GR (26) Feedk ontrol of morphogenesis in fungi y romti lohols. Genes Dev 2: Cozzolino D, Smyth HE, Lttey KA, Cynkr W, Jnik L, Dmergs RG, Frnis IL & Gishen M (26) Comining mss spetrometry sed eletroni nose, visile-ner infrred spetrosopy nd hemometris to ssess the sensory properties of Austrlin Riesling wines. Anl Chim At 563: Dudt CE & Ough CS (1973) A method for quntittive mesurement of voltile ette esters from wine. Am J Enol Viti 24: Delteil D & Jrry JM (1992) Chrteristi effets of two ommeril yest strins on Chrdonny wine voltiles nd polyshride omposition. Wine Ind J 2: Di Stefno R, Ciolfi G & Delfini C (1981) Composti voltili prodotti dei lieviti. Riv Viti Enol 34: Etievnt PX (1991) Wine. Voltile Compounds in Foods nd Beverges, 2nd edn (Mrse H, ed), pp Mrel Dekker, New York. Ferreir V, Fernndez P & Cho JF (1996) A study of ftors ffeting wine voltile omposition nd its pplition in disriminnt nlysis. Food Si Tehnol 29: Ferreir V, Lopez R & Cho JF (2) Quntittive determintion of odornts of young red wines from different grpe vrieties. J Si Food Agri 8: Giudii P, Romno P & Zmonelli C (199) A iometri study of higher lohol prodution in Shromyes erevsie. Cn J Miroiol 36: Guitrt A, Hernndez Orte P, Ferreir V, Pen C & Cho J (1999) Some oservtions out the orreltion etween the mino id ontent of musts nd wines of the Chrdonny vriety nd their fermenttion roms. Am J Enol Viti 5: Guth H & Sies A (22) Flvour of wines: towrds n understnding y reonstitution experiments nd n nlysis of ethnol s effet on odour tivity of key ompounds. FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218

11 126 F.M. Crru et l. Proeedings of the 11th Austrlin Wine Industry Tehnil Conferene (Blir RJ, Willims PJ & Høj PB, eds), pp Adelide, South Austrli. Hzelwood LA, Drn JM, vn Mris AJA, Pronk JT & Dikinson JR (28) The Ehrlih pthwy for fusel lohol prodution: entury of reserh on Shromyes erevisie metolism. Appl Environ Miroiol 74: Henshke PA & Jirnek V (1993) Yest: metolism of nitrogen ompounds. Wine Miroiology nd Biotehnology (Fleet GH, ed), pp Hrwood Ademi Pulishers, Chur, Switzerlnd. Hernndez-Orte P, Bely M, Cho J & Ferreir V (26) Impt of mmonium dditions on voltile idity, ethnol, nd romti ompound prodution y different Shromyes erevisie strins during fermenttion in ontrolled syntheti medi. Aust J Grpe Wine Res 12: Herriz T, Reglero G, Herriz M, Mrtin-Alvrez PJ & Cezudo MD (199) The influene of the yest nd type of ulture on the voltile omposition of wines fermented without sulfur dioxide. Am J Enol Viti 41: Houtmn AC & Du Plessis CS (1986) The effet of grpe ultivr nd yest strin on fermenttion rte nd onentrtion of voltile omponents in wine. S Afr J Enol Viti 6: Jirnek V, Lngridge P & Henshke PA (1991) Yest nitrogen demnd: seletion riterion for wine yests for fermentting low nitrogen musts. Proeedings of the Interntionl Symposium on Nitrogen in Grpes nd Wine (Rntz J, ed), pp Amerin Soiety for Enology nd Vitiulture, Dvis, CA. Kell DB, Brown M, Dvey HM, Dunn WB, Spsi I & Oliver SG (25) Metoli footprinting nd systems iology: the medium is the messge. Nt Rev Miroiol 3: Lmrehts MG & Pretorius IS (2) Yest nd its importne to wine rom review. S Afr J Enol Viti 21: Lem C, Gri-Jres C, Orriols I & Angulo L (1996) Contriution of Shromyes nd non-shromyes popultions to the prodution of some omponents of Alrino wine rom. Am J Enol Viti 47: Longo E, Velzquez JB, Sieiro C, Cnsdo J, Clo P & Vill TG (1992) Arom ompounds of Shromyes erevisie wine strins isolted from the Slnes region (Glii, Spin). World J Miro Biot 8: Lurton I, Snkkers G, Roullnd C, Gly B & Versvud A (1995) Influene of the fermenttion yest strin on the omposition of wine spirits. J Si Food Agri 67: Mrgheri G, Versini G, Ginotti L, Pellegrini R & Tnon D (1984) Fttori di qulit dei vini inhi giovnni: influenz delĺzot ssimilile dei mosti e dei omponenti romtii dei vini. L Revist So Itl Alimentzione 13: Mteo J, Jimenez M, Huert T & Pstor A (1991) Contriution of different yests isolted from musts of Monstrell grpes to the rom of wine. Int J Food Miroiol 14: Medin K, Crru FM, Gioi O & Brseso N (1997) Nitrogen vilility of grpe juie limits killer yest growth nd fermenttion tivity during mixed-ulture fermenttion with sensitive ommeril yest strins. Appl Environ Miroiol 63: Moreir N, Mendes F, Pereir O, Guedes de Pinho P, Hogg T & Vsonelos I (22) Voltile sulphur ompounds in wines relted to yest metolism nd nitrogen omposition of grpe musts. Anl Chim At 458: Moreno JA, Ze L, Moyno L & Medin M (25) Arom ompounds s mrkers of the hnges in sherry wines sujeted to iologil geing. Food Control 16: Muller CJ, Kepner RE & We AD (1973) Ltones in wines. Am J Enol Viti 24: 5 9. Niolini G, Mohiutti R, Lrher R & Moser S (2) Lieviti ed romi dei vini: omprzione tr eppi ommerili di lrg diffusione. L Enotenio 36: Niolini G, Volonterio G, Lrher R, Moser S & Dll Serr A (2) Prestzioni fermmenttive ed romtihe di lieviti sudfrini di reente immissioe in Itli. L Enotenio 36: Nordstrom K (1964) Studies on the formtion of voltile esters in fermenttion with rewer s yest. Svensk Kemisk Tidskrift 76: Nyknen L (1986) Formtion nd ourrene of flvour ompounds in wine nd distilled loholi everges. Am J Enol Viti 37: Psteur L (1866) Etudes sur le vin. Imprimeurs Imperils, Pris. Peddie HAB (199) Ester formtion in rewery fermenttions. J Inst Brew 96: Pretorius IS & Hoj PB (25) Grpe nd wine iotehnology; hllenges, opportunities nd potentil enefits. Aust J Grpe Wine Res 11: Rnkine BC (1967) Influene of yest strin nd ph on pyruvi id ontent of wines. J Si Food Agri 18: Rnkine BC & Pook KF (1969) Influene of yest strin on inding of sulphur dioxide in wines, nd on its formtion during fermenttion. J Si Food Agri 2: Rpp A & Versini G (1996) Influene of nitrogen ompounds in grpes on rom ompounds of wines. Viti Enol Si 51: Rezin G, Sles H & Andresen A (197) Mehnism of mjor ongener formtion in loholi grin fermenttions. J Agri Food Chem 18: Riereu-Gyon P, Duourdieu D, Donehe B & Lonvud A (2) Biohemistry of loholi fermenttion nd metoli pthwys of wine yests. Hndook of Enology, The Miroiology of Wine nd Vinifitions, Vol. 1, pp John Wiley & Sons, New York. Romno P, Cruso M, Cpee A, Lipni G, Prggio M & Fiore C (23) Metoli diversity of Shromyes erevisie strins from spontneously fermented grpe musts. World J Miro Biot 19: Smyth HE, Cozzolino D, Herderih MJ, Sefton MA & Frnis IL (25) Relting voltile omposition to wine rom: identifition of key rom ompounds in Austrlin white wines. Proeedings of the 12th Austrlin Wine Industry Tehnil Conferene (Blir RJ, Willims PJ & Pretorius IS, 28 Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved FEMS Yest Res 8 (28) Downloded from y guest on 13 July 218

12 Prodution of fermenttion rom ompounds 127 eds), pp Austrlin Wine Industry Tehnil Conferene In., Melourne, VIC, Adelide. Swiegers JH, Brtowsky PA, Henshke PA & Pretorius IS (25) Yest nd teril modultion of wine rom nd flvour. Aust J Grpe Wine Res 11: Tillndier P, Portugl FR, Fuster A & Strehino P (27) Effet of mmonium onentrtion on loholi fermenttion kinetis y wine yests for high sugr ontent. Food Miroiol 24: Vn Gemert LJ (23) Compiltions of Odour Threshold Vlues in Air, Wter nd Other Medi. Oliemns Punter & Prtners BV, The Netherlnds. Vil I, Slyrolles JM, Gerlnd C, Bumes R, Byonove C & Brre P (2) Comprison of romti nd neutrl yest strins: influene of vinifition onditions. Wein Wiss 55: Vilnov M, Uglino M, Vrel C, Sieert T, Pretorius IS & Henshke PA (27) Assimille nitrogen utilistion nd prodution of voltile nd non-voltile ompounds in hemilly defined medium y Shromyes erevisie wine yests. Appl Miroiol Biotehnol 77: Zoeklein BW, Fugelsng K, Gump BH & Nury FS (1995) Wine Anlysis nd Prodution. Chpmn & Hll, New York. Zoeklein BW, Mry JE, Willims JM & Jsinski Y (1997) Effet of ntive yests nd seleted strins of Shromyes erevisie on glyosyl gluose, potentil voltile terpenes, nd seleted glyones of white Riesling (Vitis vinifer L.) wines. J Food Comp Anl 1: FEMS Yest Res 8 (28) Federtion of Europen Miroiologil Soieties Pulished y Blkwell Pulishing Ltd. All rights reserved Downloded from y guest on 13 July 218