Preliminary characterization of wine lactobacilli able to degrade arginine

World Journal of Microbiology & Biotechnology 18: 821 825, 2002. 821 Ó 2002 Kluwer Academic Publishers. Printed in the Netherlands. Preliminary characterization of wine lactobacilli able to degrade arginine G. Spano 1, *, L. Beneduce 1, D. Tarantino 1, G.M. Giammanco 2 and S. Massa 1 1 Institute of Alimentary Productions and Technologies, Agricultural Science, Faculty of Foggia, Foggia University, Via Napoli 25, 71100 Foggia, Italy 2 Department of Hygiene and Microbiology, University of Palermo, Via del Vespro 133, I-90127Palermo, Italy *Author for correspondence: Address: Institute of Alimentary Productions, Agricultural Science, Foggia University, via Napoli 25, 71100 Foggia, Italy, Tel.: +39-881-589303, Fax: +39-881-740211, E-mail: lab.biomol@tiscali.it Received 18 March 2002; accepted 27July 2002 Keywords: Lactococcus lactis ssp. cremoris, NH 3, RAPD-PCR, wine Summary Lactobacillus strains able to degrade arginine were isolated and characterized from a typical red wine. All the strains were gram-positive, catalase-negative and produced both D- and L-lactate from glucose. Strains L2, L3, L4, and L6 were able to produce CO 2 from glucose; however, production of CO 2 from glucose was not observed in strains L1 and L5, suggesting that they belong to the homofermentative wine lactic acid bacteria (LAB) group. All of the lactobacilli were tested for their ability to ferment 49 carbohydrates. The sugar fermentation profile of strain L1 was unique, suggesting that this strain belonged to Lactococcus lactis ssp. cremoris, a non-typical wine LAB. Furthermore, a preliminary typing was performed by using a random amplified polymorphic DNA analysis (RAPD-PCR analysis). Introduction Arginine is one of the major amino acids found in grape juice and wine. The concentration of arginine ranges from a few hundred mg/ml to 2.4 g/l (Spayd & Andersen- Bagge 1996), depending on grape variety, soil, viticultural practices and fermentation conditions. Although arginine in grape juice is mostly metabolized by yeasts during vinification, this amino acid is still present in wine at the end of alcoholic fermentation (Liu & Pilone 1998). Consequently, arginine is generally available for metabolism by wine lactobacilli during the malolactic fermentation (MLF), which is usually performed after alcoholic fermentation by yeasts. The catabolism of arginine via the arginine deiminase (ADI) pathway, is proposed to be an important energy source for bacterial growth and involves three enzymatic reactions: Arginine deiminase Arginine þ H 2 O ƒƒƒƒƒƒƒƒƒƒ! Citrulline þ P i Ornithine transcarbamylase ƒƒƒƒƒƒƒƒƒƒƒƒƒƒ! Carbamoyl P þ ADP þ H 2 O Carbamate kinase ƒƒƒƒƒƒƒƒƒ! Citrulline þ NH 3 Ornithine þ Carbamoyl P ATP þ NH 3 þ CO 2 Wine lactic acid bacteria (LAB) able to derive energy from arginine catabolism may be more competitive in the stressful environment of wine (presence of acid and alcohol) than those stains unable to degrade arginine (Liu & Pilone 1998). Furthermore, arginine can even be converted into ornithine and ammonia by the combined action of arginase and urease enzymes (Kuensch et al. 1974; Sponholz 1992). NH 3 produced within the cell combines with protons to yield NH þ 4. The formation of ammonia from arginine increases wine ph, which may be physiologically important for tolerance and adaptation of wine LAB to the acid environment of wine (ph < 4.0) (Curran et al. 1995). However, wine LAB vary in their ability to degrade arginine. Homofermentative wine lactobacilli are unable to degrade arginine (Edwards et al. 1993; Liu et al. 1995), although arginine is degradable by a number of homofermentative LAB (including Lactobacillus plantarum) from other sources such as fish (Jonsson et al. 1983) and orange juice (Arena et al. 1999). In contrast, all of heterofermentative wine lactobacilli examined hydrolysed arginine, including strains of Lactobacillus buchneri and L. brevis (Weiller & Radler 1976; Edwards et al. 1993; Liu 1993, Liu et al. 1994). The major concern about arginine metabolism by wine LAB is the formation of ethyl carbamate (EC) precursors. EC is a well known animal carcinogen (Mirvish 1968), found in beverages including wine (Canas et al. 1994) and therefore it is important to have the EC level in wine as low as possible. The formation of EC is a spontaneous chemical reaction involving ethanol and EC precursors which include urea produced by

822 G. Spano et al. yeasts, citrulline and carbamoyl phosphate produced by LAB (Ough 1991). Recently, a clear correlation has been demonstrated between arginine degradation and EC formation during MLF conducted by LAB strains (Liu & Pilone 1998). EC production is strictly dependent from the availability of both arginine and LAB strains able to degrade it. Therefore, wine makers must know the characteristics of indigenous LAB in order to keep as low as possible the level of those strains able to produce arginine. In this paper wine LAB able to degrade arginine were characterized from a typical red wine. Furthermore, a preliminary typing was performed by using a random amplified polymorphic DNA analysis (RAPD-PCR analysis). Materials and methods Isolation of strains and culture coniditions Red wine was used for the isolation of LAB. Samples were taken directly from the vinification tanks at 2 (t 1 ), 19 (t 5 ), 22 (t 6 ), 26 (t 7 ) and 33 (t 9 ) days after the start of alcoholic fermentation. Samples were diluted 1:10 in 0.85% NaCl solution and 0.1 ml were Milan, Italy plated on solidified MRS (De Man et al. 1960) (Oxoid) medium, according to the spread plate Milan, Italy method. Anaerobic incubation (BBL, GasPack-System) was then performed at 30 C for 3 days, and the colonies were counted and isolated. Phenotypic characterization Primary classification was based on results obtained from Gram staining, cell morphology and catalase tests (Harrigan & McCance 1966). The production of gas from glucose was determined with MRS broth (de Man et al. 1960) and the gas that was produced was trapped in Durham tubes. Furthermore, to study arginine degradation by the ADI pathway, cultures of strains on MRS agar were subcultured in MRS broth for 24 h. Ammonia from arginine was evaluated by incubating cultures for 8 days in the medium of Niven (2 g of K 2 HPO 4 per liter, 3 g of L- arginine per liter, 0.1% (v/v) Tween 80, 50 mg of MnSO 4 per liter; ph adjusted to 7) containing 0.3% arginine HCl. The production of NH 3 from arginine was determined by adding Nessler s reagent and monitoring the development of an orange colour. The evaluation of both D()) and L(+) lactate production from glucose was performed using an enzymatic kit (Boheringer Mannheim). The presence of D()) and L(+) lactate isomers was tested by measuring the u.v. absorbance of NADH produced during the enzymatic degradation. Fermentation of 49 carbohydrates was studied by using API 50 CH strips with API 50 CHL medium (biome rieux, Marey-l Etoile, France). Results were recorded after 48 h at 30 C. Two replicates sample were performed for each isolated strain. L. plantarum ATCC 14917T ðnh 3 Þ was used as standard strain. DNA preparation Genomic DNA of LAB was extracted as described by Marmur (1961). PCR amplification of random DNA fragments Amplification was performed in a thermal cycler (Perkin-Elmer 2400, CA, USA) using four different primers of arbitrary nucleotide sequence (Table 1). For PCR amplification 20 ng of genomic DNA was added to a 20 ll PCR mixture contains 1 U/ll of Taq polymerase (Boehringer), 0.25 mm of each dntp (Boehringer Mannehim), 10 buffer (670 mm Tris-HCl ph 8.8, 160 mm (NH 4 ) 2 SO 4, 0.1% (v/v) Tween 20), 3 mm of MgCl 2 and 1.0 lm of a single primer. The reaction mix were overlaid with mineral oil and cycled as shown in Table 1. The PCR reactions were terminated at 75 C for 5 min and thereafter cooled to 4 C. The test reproducibility of RAPD was evaluated by performing two different PCR reactions and running all strains in duplicate on different gels. Gel electrophoresis Gel electrophoresis was carried out by applying 10 ll of sample to 1.5% agarose gels. Gels were run for about 40 min at 80 V in TAE 1 buffer (0.04 M Tris/acetate, EDTA 1 mm). A DNA molecular weight marker VI was used as standard. After electrophoresis the gel was stained in ethidium bromide (1 lg/ml) and thereafter Table 1. Amplification conditions tested in the development of the RAPD protocol for typing LAB NH þ 3 from wine must. Amplification stage Primer P 1 GCGATCCCCA Primer P 2 CCGCAGCCAA Primer P 3 AGCAGCGTGG Primer P 4 GAGGGTGGCGGTTCT Annealing 36 C, 1 min 36 C, 1 min 45 C, 1 min 29 C, 1 min Ramp to 72 C 2 min 2 min 2 min 2 min Extension 72 C, 5 min 72 C, 5 min 72 C, 5 min 72 C, 5 min Cycles 30 30 40 45 Denaturation step was performed at 94 C for 1 min for all of the tested programmes.

RAPD-PCR 823 washed for 10 min and visualized with a u.v. transilluminator (UV GENä BIORAD). RAPD analysis The similarities among RAPD patterns were calculated by the Dice similarity index (Dice 1945; Brosch et al. 1994) by using the Taxotron software Restrictotyper module (Taxolab software, Institute Pasteur, France). Bands whose molecular weight differed by less than 4% were considered identical. A dendrogram tree was constructed by using the Adanson and the Dendrograf module of the Taxotron software (Taxolab software) and applying the UPGMA algorithm to the distance matrix resulting from the comparison of the RAPD patterns. Results and discussion A total of six LAB able to produce NH 3 from arginine were selected and isolated at 2 (t 1 ), 19 (t 5 ), 22 (t 6 ), 26 (t 7 ) and 33 (t 9 ) days after yielding must (Table 2). Although Nessler s reagent is considered to be less sensitive and then unable to detect all the arg þ strains, its major Table 2. Biochemical characteristics of LAB strains able to degrade arginine. Strain Catalase CO 2 production from glucose NH 3 production from arginine D-lactate L-lactate L1 ) ) + + + L2 ) + + + + L3 ) + + + + L4 ) + + + + L5 ) ) + + + L6 ) + + + + Table 3. Carbohydrate utilization profiles of LAB NH þ 3 Strain as determined with the API 50CHL system. L1 t 1 L2 t 2 L3 t 3 L4 t 4 L5 t 8 L6 t 9 ATCC 14917 T Glycerol ) ) ) ) ) ) ) L-Arabinose ) ++ ++ ++ + ++ ++ Ribose ++ ++ ++ ++ ++ ++ ++ D-Xylose ) ++ ) ) ) ) ) Galctose + ++ ++ ++ ++ ++ ++ D-Glucose ++ ++ ++ ++ ++ ++ ++ D-Fructose ++ ++ ++ ++ ++ ++ ++ D-Mannose ) ++ ++ ++ ++ ++ ++ Rhamnose ) + ++ + + ) ) Mannitol ) ++ ++ ++ ++ ++ ++ Sorbitol ) ++ ++ ++ ++ ++ ++ a-methyl-d-mannoside ) ) ) ++ ) ++ ++ N-Acetylglucosamine ++ ++ ++ ++ ++ ++ ++ Amygdalin ) ++ ++ ++ ++ ++ ++ Arbutin ) ++ ++ ++ ++ ++ ++ Esculin ++ ++ ++ ++ ++ ++ ++ Salicin ) ++ ++ ++ ++ ++ ++ Cellobiose ) ++ ++ ++ ++ ++ ++ Maltose ++ ++ ++ ++ ++ ++ ++ Lactose ++ ++ ++ ++ ++ ++ ++ Melibiose ) ++ ++ ++ ++ ++ ++ Sucrose ++ ++ ++ ++ ++ ++ ++ Trehalose ++ ++ ++ ++ ++ ++ ++ Melezitose ) ++ ++ ++ ++ ++ ++ D-Raffinose ) ++ ++ + + + ++ b-gentiobiose ) ++ ++ ++ ++ ++ ++ D-Turanose ) ++ ++ ++ ++ ++ ++ D-Arabitol ) + + + ) ) + Gluconate ) ++ ++ ++ + + ++ ++, positive reaction; + weak positive reaction; ) negative reaction. None of the strains fermented L-arabitol, erythritol, L-fucose, 2-ketogluconate, 5-keto-gluconate, b-methyl-d-xyloside, xylitol and L-xylose. t=collection time.

824 G. Spano et al. advantage is the capacity to detect strains that strongly degrade arginine (Pilone et al. 1991). Morphologically, cells of L1 types were spherical and occurred singly, while cells of strains L2, L3, L4, L5 and L6 were rod-shaped, in pairs, and sometimes in short chains. All the strains were gram-positive, catalasenegative and produced both D- and L-lactate from glucose (Table 2). Strains L2, L3, L4, and L6 were able to produce CO 2 from glucose; however, production of CO 2 from glucose was not observed in strains L1 and L5, suggesting that they belong to homofermentative wine LAB. All of the lactobacilli were tested for their ability to ferment 49 carbohydrates (Table 3). We found that the sugar profile of strain L1 was unique, suggesting that this strain belong to Lactococcus lactis ssp. cremoris, a non-typical wine LAB, according to API 50CH manufacturer. The homofermentative pedicococci and lactobacilli isolated from the wine environment do not seem to degrade arginine; however, strains of Lactococcus lactis ssp. cremoris isolated from other sources (such as cheese) are able to degrade arginine (Crow & Thomas 1982). Furthermore, some arg ) strains of L. lactis ssp. cremoris may became arg + (Coventry et al. 1984; Davey & Heap 1993) and this feature is particularly important when arginine hydrolysis is used as biochemical tool in bacterial classification. To date, this is the first evidence in which homofermentative lactococci isolated from wine are able to produce NH 3. The sugar profile of strain L5, which was unable to produce gas, was completely different from that of L1. Therefore, it was concluded that this strain does not belong to Lactococcus lactis ssp. cremoris species. RAPD-PCR patterns of each strain were obtained with primers P 1,P 2,P 3 and P 4 in separate reactions. RAPD analysis revealed clear differences between isolated LAB strains which were almost indistinguishable by previously phenotypic tests (see strains L3 and L5 for example, Table 3). The primer P 1 (GCGATCCCCA) gave the largest number of bands, the other primers resulted in fewer bands (data not shown). The band pattern obtained by using P 1 primer is shown in Figure 1. Bands were either common to all the strains tested (monomorphic), common to two or more strains (shared), or Figure 1. Band patterns obtained from six lactobacilli strains using primer P 1. Lanes 1! 6, strains L1, L2, L3, L4, L5 and L6, respectively, which were able to degrade arginine. M: molecular weight marker IV. unique to only one of the strains tested (strain specific). The RAPD profiles was different for strains L1, L2, L4 and L5, while it was identical for strains L3, and L6 suggesting that they were genotypically related. Indeed, in the dendrogram produced by the UPGMA algorithm three isolates (L1, L3 and L6) were clustered together. Two of them, L3 and L6, having identical RAPD patterns, seem to share the same clonal origin, while the third, L1, differing by only one band, can be considered as closely related to them. In contrast, the other three RAPD patterns (L2, L4, and L5) were clearly unrelated since they showed >50% difference according to the Dice similarity index (Figure 2). Strain L1 was previously identify as Lactococcus lactis ssp. cremoris and primer P 1 (which had greater resolution then other primers) may be useful to type this species in wine sample. Although it has been suggested that the ability to produce ammonia from arginine can only be used as supportive evidence for a strain being heterofermentative (Liu & Pilone 1998), we were able to select some homofermentative NH þ 3 strain from red wine (including unusual LAB wine such as Lactococcus lactis ssp. Figure 2. UPGMA dendrogram based on RAPD analysis of LAB strains able to degrade arginine.

RAPD-PCR 825 cremoris) and a preliminary molecular characterization was performed. Work is in progress in order to type all the LAB strains which were able to degrade arginine. Furthermore, it is worth identifying as many LAB NH þ 3 as possible, for example by collecting sample from several wine makers, in order to establish their real contribute to EC production. References Arena, M.E., Saguir, F.M. & Manca de Nadra, M.C. 1999 Arginine dihydrolase pathway in Lactobacillus plantarum from orange. International Journal of Food Microbiology 47, 203 209. Brosch, R., Chen, J. and Luchansky, J.B. 1994 Pulsed-fleld fingerprinting of Listeriae: identification of genomic divisions for Listeria monocytogenes and their correlation with serovar. Applied and Environmental Microbiology 60, 2584 2592. Canas, B.J., Joe, F.L. Jr., Diachenko, G.W. & Burns, G. 1994. Determination of ethyl carbamate in alcoholic beverages and soy sauce by gas chromatography with mass selective detection: collaborative study. Journal of AOAC International 77, 1530 1536. Coventry, M.J., Hillier, A.J. & Jago, G.R. 1984 Changes in the metabolism of factory-derived bacteriophage resistant derivatives of Streptococcus cremoris. Australian Journal of Dairy Technology 39, 154 159. Crow, V.L. & Thomas, T.D. 1982 Arginine metabolism in lactic streptococci. Journal of Bacteriology 150, 1024 1032. Curran, T.M., Lieou, J. & Marquis, R.E. 1995 Arginine deiminase system and acid adaptation of oral streptococci. Applied and Environmental Microbiology 61, 4494 4496. Davey, G.P. & Heap, H.A. 1993 Appearance of the arginine phenotype in Lactococcus lactis subsp. Cremoris 2204 following phage transduction. Canadian Journal of Microbiology 39, 754 758. De Man, J.C., Rogosa, M. & Sharpe, M.E. 1960 A medium for the cultivation of Lactobacilli. Journal of Applied Bacteriology 23, 130 135. Dice, L.R. 1945 Measures of the amount of ecological association between species. Ecology 26, 297 302. Edwards, C.G., Powers, J.R., Jensen, K.A., Weller, K.M. & Peterson, J.C. 1993 Lactobacillus spp. from Washington State wines: isolation and characterization. Journal of Food Science 58, 453 458. Harrigan, W.F. and McCance, M.E. 1966 Laboratory Methods in Microbiology, 3rd ed., London: Academic Press. Kuensch, U., Temperli, A. & Mayer, K. 1974 Conversion of arginine to ornithine during malolactic fermentation in a red Swiss wine. American Journal of Enology and Viticulture 25, 191 193. Jonsson, S., Clausen, E. & Raa, J. 1983 Amino acid degradation by a Lactobacillus plantarum strain from fish. Systematic and Applied Microbiology 4, 148 154. Liu, S.-Q. 1993 Arginine metabolism in Malolactic Wine Lactic Acid Bacteria and Its Oenological Implications. PhD thesis, Massey University, New Zealand. Liu, S.-Q. & Pilone, G.J. 1998 A review: arginine metabolism in wine lactic acid bacteria and its practical significance. Journal of Applied Microbiology 84, 315 327. Liu, S.Q., Pritchard, G.G., Hardman, M.J. & Pilone, G.J. 1994 Citrulline production and ethylcarbamate (urethane) precursor formation from arginine degradation by wine lactic acid bacteria, Leuconostoe oenos and Lactobacillus buchneri American Journal and Viticulture. Liu, S.-Q., Pritchard, G.G., Hardman, M.J. & Pilone G.J. 1995 Occurrence of arginine deiminase pathway enzymes in arginine catabolism in wine lactic acid bacteria. Applied and Environmental Microbiology 61, 310 316. Marmur, J. 1961 A procedure for the isolation of DNA from microorganisms. Journal of Molecular Biology 3, 208 218. Mirvish, S.S. 1968 The carcinogenic action and metabolism of urethan and N-hydroxyurethan. Advances in Cancer Research 11, 1 42. Ough, C.S. 1991 Influence of nitrogen compounds in grapes on ethyl carbamate formation in wines. In Proceedings of the International Symposium on Nitrogen in Grapes and Wine 18 19 June, 1991, Seattle, Washington, ed. Rantz, J.M. pp. 165 171. California: American Society for Enology and Viticulture. ISBN 0-96307110-6. Pilone, G.J., Clayton, M.G. & van Duivenboden, R.J. 1991 Characterization of wine lactic acid bacteria: single broth culture for tests of heterofermentation, mannitol from fructose and ammonia from arginine. American Journal of Enology and Viticulture 42, 153 157. Spayd, S.E. & Andersen-Bagge, J. 1996 Free amino acid composition of grape juice from 12 Vitis vinifera cultivars in Washington. American Journal of Enology and Viticulture 47, 389 402. Sponholtz, W.R., 1992 The breakdown of arginine by lactic acid bacteria and its relation to ethylcarbamate production. Biol. Oggi 6, 15 24. Weiller, H.G. & Radler, R. 1976 On the metabolism of amino acids by lactic acid bacteria isolated from wine. Zeitschrift fu r Lebensmittel- Untersuchung und Forschung 161, 259 266.