APPLIED MICROBIOLOGY, JUIY, 1966 Copyright 1966 Americn Society for Microbiology Vol. 14, No. 4 Printed in U.S.A. Chemicl Chrcteriztion of Wines Fermented with Vrious Mlo-lctic Bcteri' GORDON J. PILONE,2 RALPH E. KUNKEE, AND A. DINSMOOR WEBB Deprtment of Viticulture nd Enology, University of Cliforni, Dvis, Cliforni Received for publiction 7 Mrch 1966 ABSTRACT PILONE, GORDON J. (University of Cliforni, Dvis), RALPH E. KUNKEE, AND A. DINSMOOR WEBB. Chemicl chrcteriztion of wines fermented with vrious mlo-lctic bcteri. Appl. Microbiol. 14:608-615. 1966.-Six mlo-lctic strins of lctic cid bcteri were isolted from Cliforni wines nd identified s Lctobcillus delbrueckii, L. buchneri, L. brevis, Leuconostoc citrovorum, nd two strins of cerevisie. Mlo-lctic fermenttion ws induced in seprte lots of wine by inocultion of ech lot with one of the strins of bcteri. Mlo-lctic fermenttion hd occurred in ech inoculted wine within 2 months. The resultnt wines were subjected to chemicl nlysis, including gs chromtogrphic exmintion of concentrted extrcts of the wines. Only few differences in composition were found when the mlo-lctic wines were compred one with nother. The differences tht were found were in voltile cidity nd in concentrtions of cetoin (plus dicetyl) nd probbly diethyl succinte. As much s one-hlf of the orgnic cid of grpes my be composed of L-mlic cid. The cidity of wine is often reduced by the conversion (mlo-lctic fermenttion) of the mlic cid to lctic cid nd CO2 by certin lctic cid bcteri (mlo-lctic bcteri) some time fter the lcoholic fermenttion (2, 12, 15). Besides the decrese in cidity, other chnges in composition of the wine my occur s the result of mlolctic fermenttion (3, 4, 14, 15). Lctic cid bcteri re vrible in their metbolism from one strin to nother, nd thus the composition of mlo-lctic wines might be expected to be dependent upon the predominnt strin(s) of bcteri present. Six mlo-lctic bcteri were isolted from wines from severl regions of Cliforni. For the most prt, these bcteri were biochemiclly nd txonomiclly diverse. Ech of these bcteri ws used to induce mlolctic fermenttion in seprte wine. The flvor differences in the resultnt wines were found to be smll (10). The wines were then nlyzed for chemicl differences, especilly in the voltile constituents, to determine whether they could be chrcterized s to the bcteri responsible for 'Tken in prt from thesis submitted by the senior uthor s prtil fulfillment for the M.S. degree in Food Science, University of Cliforni. 2 Present ddress: Mont L Slle Vineyrds, Np. Clif. the mlo-lctic fermenttion. The nlyses included gs-liquid chromtogrphy of highly concentrted extrcts of the wines. The mlolctic wines were compred one with nother nd lso with control wines which hd not undergone mlo-lctic fermenttion. MATERIALS AND METHODS Medi. The bsl medium ws modified Rugos medium (5): 2% tryptone (Difco), 0.5% peptone (Difco), 0.5% yest extrct (Difco), 0.1% liver extrct (Wilson & Co., Chicgo, Ill.), nd 0.005% Tween 80 (Nutritionl Biochemicls Corp., Clevelnd, Ohio). For growth of bcteri, the bsl medium ws supplemented with 0.3% glucose nd 0.2% lctose nd mde up in diluted (4.2-fold) cnned tomto juice without preservtive (S & W brnd) which hd been filtered through Hyflo Super Cel filter id. The growth medium ws djusted to ph 4.5 with HCI. For isoltion of bcteri, the growth medium ws solidified with 2% gr (Difco) nd djusted to ph 5.5, rther thn ph 4.5, nd 100 ppm of cycloheximide (Actidione; The Upjohn Co., Klmzoo, Mich.) ws dded to inhibit yest growth. For testing for bility to ferment mlic cid, 0.3% L-mlic cid (Nutritionl Biochemicls Corp.) ws dded to the growth medium, nd it ws djusted to ph 4.5 with KOH. The bsl medium ws used for determintion of fermentble sugr substrtes with the test sugr (sterilized seprtely) present t 2% concentrtion. The bsl 608 medium with 0.5% glucose ws lso used to test for heterolctic fermentbility. Bcteri. Mlo-lctic bcteri were isolted from
VOL. 14, 1966 COMPOSITION OF MALO-LACTIC WINES 609 mlo-lctic wines by spreding loop of wine on pltes of the bove isoltion medium. The pltes were incubted t 25 C in desicctors in which the oxygen hd been exhusted by mens of lighted cndle. After 1 to 2 weeks of incubtion, grm-positive-ctlse-negtive (13) colonies were selected nd tested for bility to ferment mlic cid to lctic cid. Bcteri were considered mlo-lctic bcteri if mlic cid hd disppered in the medium by the time of obvious bcteril growth. Mlic nd lctic cids were determined qulittively by pper chromtogrphy (5, 7). Positive mlo-lctic bcteri, which in ll instnces were Lctobcillcee, were clssified by stndrd methods given in Bergey's Mniul of Determintive Bcteriology nd Mnul of Microbiologicl Methods (13). The size nd shpe of the bcteri were determined by exmintion of photomicrogrphs of bcteri imbedded in 1% gr. Heterolctic fermentbility ws determined by gs production in 7 ml of liquid medi (see bove) in tubes (outer dimeter, 13 mm) covered with 2 to 3 ml of sterile molten mixture of equl prts of petroleum jelly nd prffin. Homolctic bcteri were incubted severl weeks fter obvious bcteril growth for confirmtion of lck of gs production. Fermentbility of sugrs ws determined by qulittive mesurement of formtion of lctic cid by pper chromtogrphy (see bove). Optiml tempertures of growth were obtined from Arrhenius plots of growth rtes s mesured by increses in turbidity, with time, of liquid cultures (5 ml) incubted in wter bths t controlled tempertures. The turbidities were red t 600 m, in Spectronic-20 colorimeter. Vinifiction. Thirteen-gllon (49-liter) lots of wine were mde from 5:1 blend of crushed Grenche nd Gmy Beujolis grpes, grown in Dvis, by stndrd methods of this lbortory (5, 10). The lots for mlo-lctic fermenttion were inoculted with mlo-lctic bcteri grown in 500 ml of growth medium, given bove, to which hd been dded 50 ppm of SO2. The bcteri were centrifuged, wshed, nd resuspended in bout 250 ml of sterile distilled wter before use. Lots of crushed grpes with no bcteril ddition served s controls. After completion of mlolctic fermenttion (from 1 to 2 months t 11 to 16 C, depending on the strin), the wines were rcked; 5 gl (18.9 liters) ws tken for preprtion for gsliquid chromtogrphic nlysis, nd the reminder ws stored for sensory nlysis (10). All equipment ws thoroughly clened with hot wter (60 C) to minimize microbiologicl contmintion. A more detiled description of the vinifiction nd induction of mlo-lctic fermenttion hs been given (10). Chemicl nlyses. The wines were nlyzed for ph, totl cidity, voltile cidity, lcohol concentrtion, nd tnnin content by methods given by Amerine (1). Acetoin (cetylmethyl crbinol) plus dicetyl (2, 3-butnedione) concentrtion ws determined colorimetriclly (6, 9). Mlic cid concentrtion ws determined quntittively (8) by enzymtic reduction of nicotinmide denine dinucleotide (Nutritionl Biochemicls Corp.) with mlic dehydrogense (Worthington Biochemicls Corp., Freehold, N.J.). Gs chromtogrphic nlyses. Gs-liquid chromtogrphic nlyses were mde on the voltile constituents of the wines. These frctions were prepred (16) from stem distilltes (t low pressure nd temperture) of 5 gl of wine. The distilltes (bout 10 gl) were extrcted five times with 200-ml volumes of dichloromethne. The extrct ws concentrted by removl of dichloromethne by distilltion. The cids were seprted from the noncidic mterils (the neutrl frction) by extrction with 3% N2CO3 followed by extrction with ether t ph 1. The cids were then methylted with dizomethne. The finl volumes of both the neutrl nd cid frctions were ech bout 2 ml. They were stored t 4 C. The neutrl frctions were chromtogrphed with n Aerogrph model 600 instrument, nd the (methylted) cid frctions with n Aerogrph model 600 B Hi-Fy instrument. Both instruments were equipped with hydrogen flme ioniztion detector. The vrious columns used in the nlyses were prepred with Gs Pck F, the 40-to 60-mesh frctions of Johns- Mnville C-22 Firebrick, or Chromosorb-W. All columns were mde of 0.25-inch (0.6 cm) tubing (outside dimeter). Lengths vried from 6 to 10 ft (182.8 to 304.8 cm). For the neutrl frctions, the following columns nd conditions were used: (i) Gs Pck F crrying 10% (w/w) polyneopentylglycoldipte (NPGA) t 170 C nd 46 ml of N2 per min; (ii) sme s column 1 t 120 C; (iii) Firebrick crrying 10% (w/w) polydiethylene glycolsuccinte (DEGS) t 80 C nd 13 ml of N2 per min; (iv) Firebrick crrying 10% (w/w) diglycerol (DG) t 75 C nd 22 ml of N2 per min; nd (v) Firebrick crrying 10% (w/w) 1,2,3-tris(2- cynoethoxy)propne (TCEP) t 70 C nd 20 ml of N2 per min. For the (methylted) cid frctions, the following columns nd conditions were used: (i) Chromosorb-W crrying 20% (w/w) NPGA t 120 C nd 85 ml of N2 per min; (ii) Sme s 1 t 83 C; nd (iii) Firebrick crrying 10% (w/w) DG t 83 C nd 19 ml of N2 per min. Tenttive identifiction of compounds found in chromtogrphed frctions ws mde by comprison with chromtogrms of known substnces. Quntittive comprison ws mde by multipliction of retention time by pek height nd by ttenution fctor for ech compound. This vlue ws used for clcultion of the percentge of ech pek in terms of ll the peks used in the clcultion. Quntittion of the neutrl compounds ws lso mde by comprison with the mount of ethyl cprylte found. In mny cses, duplicte chromtogrms were mde, nd the results obtined were prcticlly identicl, whether the concentrtion ws bsed on percentge of totl mteril or on reltion to mount of ethyl cprylte. RESULTS Bcteril identifiction. Bcteri selected for this investigtion were: Lctobcillus delbrueckii, L. buchneri, nd cerevisie (strin 1) isolted from Southern Cliforni wines; L. brevis, isolted from Livermore Vlley wine;
610 PILONE, KUNKEE, AND WEBB APPL. MICROBIOL. Lctobcillus delbrueckii L. buchneri L. brevis Orgnism Leuconostoc citrovorum cerevisie (strin no. 1) P. cerevisie (strin no. 2) TABLE 1. Clssifiction of mlo-lctic bcteri Chrcteristics Grm-positive, ctlse-negtive, homofermenttive rods (0.7 by 2 ju), occur singly nd in chins nd occsionlly elongted severl-fold. Lctic cid produced from glucose; lctose nd xylose not fermented. Optiml temperture, 37 C. Grm-positive, ctlse-negtive, heterofermenttive rods (0.9 by 2,). Occur singly nd in chins nd occsionlly elongted severl-fold. Lctic cid produced from glucose, rffinose, nd xylose. Optiml temperture, between 32 nd 35 C. Grm-positive, ctlse-negtive, heterofermenttive rods (0.5 by 1 P), occur singly nd in pirs. Lctic cid produced from glucose nd xylose, rffinose not fermented. Optiml temperture, 32 C. Grm-positive, ctlse-negtive, heterofermenttive cocci (0.5 by 0.7,u), usully occur slightly elongted but often occur in chins of severl sphericl cells. Lctic cid produced from glucose; sucrose nd xylose not fermented. Optiml temperture, between 25 nd 27 C. Grm-positive, ctlse-negtive, homofermenttive cocci (dimeter, 0.9,u), usully occur in pirs nd occsionlly in tetrds; chins never observed. Lctic cid produced from crbohydrtes. Optiml temperture, 25 C. Grm-positive, ctlse-negtive, homofermenttive cocci (dimeter, 1,u), usully occur in pirs nd occsionlly in tetrds; chins never observed. Lctic cid produced from crbohydrtes. Optiml temperture, between 27 nd 30 C. TABLE 2. Chemicl nlyses of the wines Totl cid Voltile cid Achl Acetoin + Tnnin (g Bcteril inoculum ph t(gof tr- Mglic cid (g of cette/ v/v) dicetvl of tnmte/ trtel00g/10 ml) 100 Ml) ~ ~ (PPM3 100 Ml) None: unfermented must (230 Brix)... 3.5.75.338 None: control I... 3.7.62.332.030 11.4 8.4.11 None: control 2... 3.7.63.330.025 11.4 7.5.10 None: control 3... 3.7.62.320.031 11.4 8.2.10 Lctobcillus delbrueckii.... 3.9.43 Trce.042 11.2 1.10 L. buchneri... 3.9.43 Trce.041 11.3 1.11 L. brevis... 3.9.44 Trce.036 11.5 4.9.11 Leuconostoc citrovorum... 3.9.44 Trce.041 11.4 13.9.11 cerevisie (strin 1)... 3.9.43 Trce.042 11.2 7.3.11 P. cerevisie (strin 2). 3.9.43 Trce.043 11.1 1.11 Leuconostoc citrovorum, isolted from Np Vlley Wine; nd nother P. cerevisie (strin 2) isolted fiom wine mde t Dvis in this deprtment. The chrcteristics leding to their identifiction re summrized in Tble 1. All of these bcteri hd the bility to ferment L-mlic cid to lctic cid. Lctobcillus brevis nd Leuconostoc cirrovorum were isolted by other workers (4) nd hve been designted ML 30 nd ML 34, respectively (18). These ltter two orgnisms were cpble of degrding citric cid dded to the growth medium. Mlo-lctic fermenttion. The bcteri, fter centrifugtion nd resuspension in wter, were dded to the fermenting must t bout 1% inoculum (10). Mlo-lctic fermenttion with Leuconostoc citrovorum ws complete in less thn 1 month, but with the other bcteri, bout 2 months ws required. With Lctobcillus delbrueckii, L. buchneri, nd the two strins, reinocultion ws required bout 6 weeks fter the initil bcteril inocultion. Chemicl nlyses. In Tble 2 re given the results of the chemicl nlyses of the wines used in this study. The bsence of mlic cid (Tble 2) in the wines which hd been inoculted with the bove bcteri showed tht the mlo-lctic fermenttion ws complete. Decidifiction c-
VOL. 14, 1966 COMPOSITION OF MALO-LACTIC WINES TABLE 3. Retention times (minutes) ofpeks for comprisoni of known compounds with unkniown compounds of the wine neutrl frctionis Chromtogrphic conditions Compound NPGA column, NPGA column, DEGS column, DG column, TCEP column, 170 C 120 C 80 C 75 C 70 C Known kun Known Un Known Un Konknown known known Known known Un- Known Unknown Isobutyl lcohol... Norml butyl lcohol... Active myl lcohol... Isomyl lcohol... Hexyl lcohol... Ethyl lctte... Ethyl cprote... Ethyl cprylte... Isobutyl cprote... Isomyl cprote... -y-butyrolctone... Diethyl succinte... Ethyl cprte... Isomyl cprylte... 2-Phenethyl lcohol... 3.4 4.0 5.7 6.0 7.0 8.0 11.1 3.3 4.1 5.4 5.9 6.8 7.9 11.0 2.2 2.0 2.0 5.3 3.8 9.0 12.6 16.0 19.7 25.5 2.2 2.0 2.0 5.4 3.8 9.0 12.5 16.0 19.7 25.2 10.4 13.4 29.0 10.3 13.5 4.1 5.9 7.2 9.4 17.0 16.9 29.0 20.6 See Mterils nd Methods. NPGA = polyneopentylglycoldipte; DEGS = polydiethylene glycolsuccinte; DG = diglycerol; nd TCEP = 1,2,3-tris(2-cynoethoxy)propne. 84.6 43.8 4.1 6.0 7.3 9.4 17.5 17.5 20.9 84.5 43.3 TABLE 4. Percentge mounts of compounds of wine neutrl frctions chromtogrphed onl Gs Pck F-polyneopentylglycoldipte column t 170 C nd 16 ml of N2/min ceejse Pek Compound Lctobcillus L. L brevis LeuconosIoc cerevzste no. delbrueckii buchneri citrovorum 1 2 3 Strin 1 Strin 2 1 Ethyl lctte Ethyl cprote Hexyl lcohol 9.8 8.7 9.6 13.7 12.5 12.0 12.1 11.6 13.1 2 Ethyl cprylte 3.2 2.7 2.6 3.0 3.0 2.8 3.0 2.7 3.0 3 Isomyl cprote 0.7 0.6 0.6 0.8 0.8 0.8 0.8 0.8 0.7 4 Diethyl succinte 0.8 0.6 1.0 0.8 I.8 1.5 0.7 1.2 1.2 5 Ethyl cprte 2.8 2.6 2.2 2.4 3.0 2.2 3.0 2.3 2.5 6 Isomyl cprylte 0.7 0.7 0.6 0.6 0.6 0.5 0.7 0.5 0.7 7 Unknown 1.2 1.1 1.1 1.2 1.3 1.1 1.1 1.3 1.2 8 2-Phenethyl lcohol 80.8 83.0 82.3 77.5 76.9 78.9 78.7 79.0 77.0 Retention time of 9.0 min. 35.2 34.8 compnied the mlo-lctic fermenttion, s seen by the drop in totl cidity nd the rise in ph (Tble 2). The mlo-lctic wines lso showed slight increse in voltile cidity, which often ccompnies mlo-lctic fermenttion (5, 17). The finl lcohol concentrtion nd the tnnin content re prcticlly the sme for ll of the wines. The gretest vrition shown in Tble 2 is in the concentrtion of cetoin (plus dicetyl). Here the control wines were nerly like, but the concentrtion found in the mlo-lctic wines vried from bout one-hlf of tht found in the control to more thn one nd one-hlf times s much. Gs chromtogrphic nlyses. The wines were concentrted bout 10,000-fold (see Mterils nd Methods) to determine, by gs-liquid chromtogrphic techniques, whether differences in mounts of voltile mterils could be detected in the wines fermented with different bcteri. To prevent esterifiction of the cids nd lcohols of the concentrted smples, the cids were seprted from the noncidic mteril, nd the
612 PILONE, KUNKEE, AND WEBB APPL. MICROBIOL. TABLE 5. Percentge mounts of compounds of wine neutrl frctions chromtogrphed on Gs Pck F-polyneopentylglycoldipte column t 120 C nd 46 ml of N2/min Pek Compound Lctobcillus L. L brevis Leuconostoc cerevisie no. _ delbrueckii buchneri cisrovorum 1 2 3 Strin 1 Strin 2 1 Ethyl cprylte 2.5 2.2 2.2 2.3 2.4 2.2 2.6 2.3 2.4 2 Unknown TR TR TR TR TR TR 3 Diethyl succinte 0.5 0.4 0.6 0.7 1.3 1.1 0.4 0.7 0.7 4 Ethyl cprte 2.0 1.7 2.3 1.8 2.1 1.8 5 Isomyl cprylte 0.9 1.0 1.0 1.1 1.1 1.0 1.1 1.2 0.9 6 2-Phenethyl lcohol 94.3 94.3 94.5 94.1 92.1 93.9 93.9 92.9 94.1 Retention time of 7.1 min (not isomyl cprylte):-= none, TR = trce. TABLE 6. Percentge mounts of compounds of wine neutrl frctions chromtogrphed on Firebrickpolydiethyleneglycosuccinte column t 80 C nd 13 ml of N2/min Pek Compound Lctobcillus Lu L. buchneri revis Leuconostoc cerevisie no. delbrueckis citrovorum 1 2 3 Strin 1 Strin 2 1 Hexyl lcohol 67.6 67.0 65.9 58.7 56.2 57.9 54.5 54.6 54.5 2 Ethyl lctte 4.8 7.0 9.5 22.3 23.0 22.6 22.9 24.6 24.5 3 Ethyl cprylte 27.6 26.0 24.4 19.0 20.8 19.5 22.6 20.8 21.0 TABLE 7. Percentge mounts of compounds of the wine neutrl frctions chromtogrphed on Firebrickdiglycerol column t 75 C nd 22 ml of N2/min Pek Lctobcillus L. L. Leuconostoc cerevisie no. Compound delbrueckii buchneri brevis citrovorum 1 2 3 Strin 1 Strin 2 1 Isobutyl lcohol 8.7 7.3 8.2 7.9 7.4 8.0 7.1 7.9 7.7 2 Norml butyl lcohol 3.0 3.1 2.8 3.0 2.3 2.2 2.4 1.8 2.6 3 Active myl lcohol 58.1 58.4 57.4 56.2 56.8 56.0 53.6 53.2 54.0 4 Hexyl lcohol Ethyl lctte 13.9 15.1 16.2 20.5 21.4 21.1 20.7 21.0 21.7 5 Ethyl cprylte 3.6 3.4 3.7 3.2 3.3 3.5 3.8 3.8 3.4 6 Unknown Diethyl succinte 9.0 9.0 7.7 5.8 4.2 5.5 8.1 7.7 6.6 7 Unknowns 0.7 0.7 1.2 0.8 1.8 1.5 0.8 1.3 1.3 8 'y-butyrolctone 2.9 3.1 2.7 2.7 2.8 2.1 3.6 3.2 2.7 Retention time of 43.3 min. b Retention time of 54.6 min. two frctions from ech wine were exmined seprtely. Chromtogrphy of neutrl frctions. Eighteen mjor components were found on chromtogrmi of the concentrted neutrl frctions of the extrcted wines. Tenttive identifiction of 15 of these compounds is shown in Tble 3 by comprison of retention times of known compounds on severl columns. The percentge mounts of the vrious compounds found with four different chromtogrphic conditions re shown in Tbles 4 through 7. In generl, only smll differences were found mong the wines, either when the mlo-lctic wines were compred with one nother or when mlo-lctic wines were compred with control wines. Two differences, however, re noteworthy. (i) An obvious difference mong the wines ws in the concentrtion of ethyl lctte. Ethyl lctte ws seprted from other components of the neutrl frction on the DEGS column (Tble 6). Mlo-lctic wines, which of course hve reltively higher concentrtions of
VOL. 14, 1966 COMPOSITION OF MALO-LACTIC WINES 613 TABLE 8. Retentioni times (minutes) of peks for comprisoll of known compounds with unknown compounds of the methylted wine cids frctions Chromtogrphic conditions Compound NPGA column, 120 C NPGA column, 83 C DG column, 83 C Known Unknown Known Unknown Known Unknown Methyl cprote 4.0 4.0 Ethyl lctte........ 5.8 5.8 12.0 12.0 Methyl cprylte..... 4.0 3.9 15.6 15.3 Methyl cprte... 1 11.8 58.0 58.4 See Mterils nd Methods. NPGA = polyneopentylglycodipte nd DG = diglycerol. TABLE 9. Percentge mounts of compounids of methylted wine cids frctions chromtogrphed oln Chromosorb-W-polyneopentylglycoldipte column t 120 C nd 85 ml of N2/min Pek Compound Lclobcillus L b L is Leuconostoc cerevisie no. - - ~~~~~~~delbruieckii ce.rvs citrovorum no.1 2 3 Strin 1 Strin 2 1 Methyl cprote 22.5 24.0 22.2 26.1 27.3 26.4 24.1 23.8 23.7 2 Ethyl lctte - - TR TR TR TR TR TR 3 Methyl cprylte 59.5 61.0 61.4 60.5 59.1 60.6 64.4 60.5 59.4 4 Methyl cprte 18.0 15.0 16.4 13.4 13.6 13.0 11.5 15.7 16.9 Not confirmed under these conditions s ethyl lctte; - = none, TR = trce. TABLE 10. Percenitge mounts of compounids of methylted winie cids frctions chromtogrphed on Chromosorb-W-polyneopentylglycoldipte columnt t 83 C nd 85 ml of N2/min Pek Lctobcills L. buchneri L. brevis Leuconostoc cereviste no. Compound delbrueckii-bcnr.bei citrovorum 1 2 3 Strin 1 Strin 2 1 Methyl cprote 27.8 28.1 28.1 51.7 33.3 27.1 28.2 26.3 25.5 2 Ethyl lctte - - 6.2 2.3 2.4 2.7 2.2 2.1 3 Methyl cprylte 56.3 56.1 56.4 38.5 48.0 53.3 52.7 55.4 56.3 4 Methyl cprte 15.8 15.8 15.5 3.5 16.4 16.9 16.4 16.2 16.2 None. lctic cid, were seen to hve percentges of ethyl lctte severl times tht found in the controls. The higher concentrtion ws lso found in Tbles 4 nd 7, but here the degree of difference ws obscured by the presence of other compounds with the sme retention times. (ii) A vrition in concentrtion of presumed diethyl succinte ws found from wine to wine in the mlo-lctic wines. The percentge of diethyl succinte found in ny of the wines ws smll, but it vried from bout equl to tht of the control in the wine fermented with Leuconostoc citrovorum to nerly double tht in the wine fermented with Lctobcillus buchneri. These differences re seen both in Tbles 4 nd 5 (NPGA columns t two different tempertures). (The sme vrition ws seen if the diethyl succinte dt were normlized with respect to ethyl cprylte.) Diethyl succinte should lso be detected on the DG column (Tble 7), but here rther lrge unknown compound with the sme retention time s diethyl succinte obscured the ltter. Another unknown compound (pek no. 7 in Tble 7) ws suspected of being diethyl succinte becuse the percentge mount of it vried in the sme wy, but its retention time ws too gret to be diethyl succinte. Chromtogrphy of cid frctions. Four compounds were obtined nd identified (Tble 8) by gs chromtogrphy of cid frctions of the extrcted wines. One of the compounds found ws ethyl lctte (Tble 8), which indicted
614 PILONE, KUNKEE, AND WEBB APPL. MICROBIOL. incomplete seprtion of the cids nd the neutrl frctions. Agin, the concentrsion of ethyl lctte ws higher in the mlo-lctic wines nd ws ctully too smll to be detected in the control wines (Tbles 9 nd 10). Generlly, only smll differences were found mong the wines for the methylted cids (Tbles 9 nd 10). [Unusul concentrtions of methylted cids were found in the wine fermented with Lctobcillus delbrueckii (Tble 10), but this ws pprently erroneous since the berrtion did not pper when the frction ws chromtogrphed on the sme support column t different temperture (Tble 9)]. DIscussioN The purpose of the reserch ws to discover whether mlo-lctic wines could be chrcterized, by chemicl nlysis, by the strin of bcteri which ws responsible for the mlo-lctic fermenttion. A comprison of the composition of mlo-lctic wines reveled only smll differences. The most noticeble vrition ws in the cetoin (plus dicetyl) concentrtion. Two of the mlolctic wines hd concentrtions of cetoin (plus dicetyl) the sme or lower thn the controls, but the others were higher. The mlo-lctic wine with the lowest cetoin (plus dicetyl) concentrtion (tht fermented with Lctobcillus brevis) lso hd reltively lower voltile cidity, thn the other mlo-lctic wines. One might hve expected lrger differences in composition of the wines. The bcteri which induced the mlolctic fermenttion were different in their specificities for nd metbolisms of sugrs, in their optiml growth tempertures, nd in the velocities of mlo-lctic fermenttion (10). It hs been shown tht mlo-lctic bcteri re vrible in their formtion of succinic cid nd cetoin (19). One ssumes tht the smll differences in chemicl composition tht were found mong the mlo-lctic wines resulted from the bcteri responsible for the mlo-lctic fermenttion. However, differences might lso result from vinifiction procedures, such s incomplete mixing of crushed grpes prior to division into individul lots. Bigger differences in composition of the wines were found in the comprison of the wines without mlo-lctic fermenttion with those which hd undergone mlo-lctic fermenttion. Mlolctic fermenttion resulted, of course, in loss of mlic cid nd increse in ph, nd lso n increse of both lctic cid nd ethyl lctte. In ddition, the cetoin (plus dicetyl) concentrtions nd voltile cidities were generlly higher in the mlo-lctic wines; this hs usully been found in the comprison of wines with nd without mlo-lctic fermenttion (5, 6, 11, 17). But side from the expected differences, the composition of the control wines ws nerly the sme s tht of the mlo-lctic wines. It would seem tht the metbolic end products of the bcteri were generlly in such smll mounts tht ny differences in the wines s the result of the bcteri were obscured by the presence of end products of yest metbolism or mterils from the grpes. The low concentrtion of metbolic end products from the bcteri might occur becuse the mlo-lctic fermenttion followed the yest fermenttion nd only low concentrtions of certin substrtes, residul sugrs etc., were vilble for metbolism by the bcteri. Greter differences in composition of mlo-lctic wines fermented by vrious bcteri might be obtined with use of other vrieties of grpes which hd lrger mounts of certin substrtes. The wines hve lso been subjected to sensory evlution by n experienced tsting pnel (10). Less thn hlf of the judges could mke consistent differentition of the wines. The wines were tsted nd chemiclly nlyzed bout 4 months fter vinifiction. The smll orgnoleptic nd chemicl differences found my become more noticeble fter the wines hve been ged. Collectively, 4 of the 10 members of the tsting pnel who were ble to differentite the wines rnked highest the wine fermented with Lctobcillus brevis. This wine lso hd the lowest voltile cidity nd lowest cetoin (plus dicetyl) concentrtion. Since some of the tste pnel members were ble to differentite the wines, greter differences thn were observed in chemicl composition of the wines might hve been expected. Contrrily, Webb et l. (16) reported tht two sherries, fermented in two different wys, hd lrge differences in rom nd tste but qulittively lmost identicl compositions of voltile mteril. The percentges of voltile mterils found by chromtogrphy represent the reltive mounts of chemicls present in the concentrted extrct, not necessrily the percentge mounts in the originl wine. Becuse of the mny mnipultions required to obtin the extrct, perhps it is surprising tht the reltive compositions vry from one extrct to nother s little s they do. But it is this consistency in the reltive mounts of most of the compounds found tht gives importnce to the differences found for some of the compounds. The identifiction of most of the voltile compounds is only tenttive. This is especilly true of diethyl succinte, for which comprison of retention times with known di-
VOL. 14, 1966 COMPOSITION OF MALO-LACTIC WINES 615 ethyl succinte ws mde on only one kind of gs chromtogrphic column. All of the voltile compounds found re those tht hve generlly been found in fermented mteril (16). The percentges of two compounds found by gs chromtogrphy re not given. Isomyl lcohol (3-methyl 1-butnol) ws found s seprte pek on the DG column, but it ppered in such high concentrtion tht it could not be mesured ccurtely by this method, nd thus ws not included in Tble 7. Isobutyl (2-methyl 1-propnyl) cprote ws identified on the NPGA column, but it ppered only s shoulder of nother pek nd is not given in Tble 5. The unknown pek in Tble 5 (pek no. 2-not isobutyl cprote) is seen only in mlo-lctic wines, but it ws of some interest becuse it ws found only in trce mounts. In some of the tbles of the gs chromtogrphic results, the compounds which were present in lrge mounts often ppered in ll the mlo-lctic wines t somewht lower percentges thn in the control wines; for exmple, see 2-phenethyl lcohol (pek no. 8) in Tble 4. The results were given s percentges of the totl mteril found on the chromtogrms nd, becuse of the lrger mount of ethyl lctte in the mlo-lctic wines, the other compounds in the mlo-lctic wines ppered t correspondingly lower percentge of the totl. When the dt were normlized with respect, e.g., to ethyl cprylte, these differences between the control nd mlo-lctic wines disppered. LITERATURE CITED 1. AMERINE, M. A. 1965. Lbortory procedures for enologists. Univ. Cliforni, Dvis. 2. AMERINE, M. A., AND W. V. CRUESS. 1960. The technology of wine mking, p. 523-525. Avi Publishing Co., Inc., Westport, Conn. 3. INGRAHAM, J. L., AND G. M. COOKE. 1960. A survey of the incidence of mlo-lctic fermenttion in Cliforni tble wines. Am. J. Enol. Viticult. 11:160-163. 4. INGRAHAM, J. L., R. H. VAUGHN, AND G. M. COOKE. 1960. Studies of the mlo-lctic orgnisms from Cliforni wines. Am. J. Enol. Viticult. 11:1-4. 5. KUNKEE, R. E., C. S. OUGH, AND M. A. AMERINE. 1964. Induction of mlo-lctic fermenttion by inocultion of must nd wine with bcteri. Am. J. Enol. Viticult. 15:178-183. 6. KUNKEE, R. E., G. J. PILONE, AND R. E. COMBS. 1965. The occurrence of mlo-lctic fermenttion in Southern Cliforni wines. Am. J. Enol. Viticult. 16:219-223. 7. LUGG, J. W. H., AND B. T. OVERELL. 1948. "One-" nd "two-dimensionl" prtition chromtogrphic seprtions of orgnic cids on n inert sheet support. Austrlin J. Sci. Res. Sec. A 1:98-111. 8. MAYER, H., AND I. BUSCH. 1963. Uber eine enzymtische Apfelsurebestimmung in Wein und Trubensft. Mitt. Gebiete Lebensm. Hyg. 54:60-65. 9. NEISH, A. C. 1952. Anlyticl methods for bcteril fermenttions. 2nd rev. Nt. Res. Council Cnd Rept. No. 46-8-3. 10. PILONE, G. J., AND R. E. KUNKEE. 1965. Sensory chrcteriztion of wines fermented with severl mlo-lctic strins of bcteri. Am. J. Enol. Viticult. 16:224-230. 11. RADLER, F. 1962. Die Bildung von Acetoin und Dicetyl durch die Bkterien des biologischen Surebbus. Vitis 3:136-143. 12. RIBIREAu-GAYON, J., AND E. PEYNAUD. 1961. Trite d'oenologie, vol. 2, Libririe Polytechnique Ch. Bernger, Pris. 13. SOCIETY OF AMERICAN BACTRIOLOGISTS. 1957. Mnul of microbiologicl methods. McGrw- Hill Book Co., Inc., New York. 14. SUVERKROP, B., AND A. TCHELISTCHEFF. 1949. Mlo-lctic fermenttion in Cliforni wines. Wines nd Vines 30(7) :19-22. 15. VAUGHN, R. H., AND A. TCHELISTCHEFF. 1957. Studies on the mlic cid fermenttion in Cliforni tble wines. I. An introduction to the problem. Am. J. Enol. 8:74-79. 16. WEBB, A. D., R. E. KEPNER, AND W. G. GALETTO. 1964. Comprison of the roms of flor sherry, bked sherry, nd submerged-culture sherry. Am. J. Enol. Viticult. 15:1-10. 17. WEBB, R. B. 1962. Lbortory studies of the mlo-lctic fermenttion. Am. J. Enol. Viticult. 13:189-195. 18. WEBB, R. B., AND J. L. INGRAHAM. 1960. Induced mlo-lctic fermenttions. Am. J. Enol. Viticult. 11:59-63. 19. WHITING, G. C., AND R. A. COGGINS. 1963. Metbolism of mlte nd citrte by lctic cid bcteri. Long Ashton Res. St. Ann. Rept., p. 157-167.