Metabolism of amino acids, dipeptides and tetrapeptides by Lactobacillus sakei

Food Microbiology 29 (2012) 215e223

Contents lists available at ScienceDirect

Food Microbiology journal homepage: www.elsevier.com/locate/fm

Metabolism of amino acids, dipeptides and tetrapeptides by Lactobacillus sakei Quirin Sinz, Wilfried Schwab* Technische Universität München, Biotechnology of Natural Products, Liesel-Beckmann-Straße 1, D- 85354 Freising, Germany

a r t i c l e i n f o

a b s t r a c t

Article history: Received 2 May 2011 Received in revised form 14 July 2011 Accepted 14 July 2011 Available online 22 July 2011

The microbial degradation of proteins, peptides and amino acids generates volatiles involved in the typical flavor of dry fermented sausage. The ability of three Lactobacillus sakei strains to form aroma compounds was investigated. Whole resting cells were fermented in phosphate buffer with equimolar amounts of substrates consisting of dipeptides, tetrapeptides and free amino acids, respectively. Dipeptides disappeared quickly from the solutions whereas tetrapeptides were only partially degraded. In both approaches the concentration of free amino acids increased in the reaction mixture but did not reach the equimolar amount of the initial substrates. When free amino acids were fed to the bacteria their levels decreased only slightly. Although peptides were more rapidly degraded and/or transported into the cells, free amino acids produced higher amounts of volatiles. It is suggested, that after transport into the cell peptides are only partially hydrolyzed to their amino acids, while the rest is metabolized via alternative metabolic pathways. The three L. sakei strains differed to some extend in their ability to metabolize the substrates to volatile compounds. In a few cases this was due to the position of the amino acids within the peptides. Compared to other starter cultures used for the production of dry fermented sausages, the metabolic impact of the L. sakei strains on the formation of volatiles was very low. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Lactobacillus sakei Resting cells Amino acid metabolism Peptide metabolism Aroma-relevant volatiles Solid phase micro extraction

1. Introduction The flavor of dry fermented sausages derives from the ingredients (meat, spices and smoke) and the chemical changes occurring during the fermentation and drying process. Fermentative flavor formation occurs by bacterial ferments and by meat enzymes which are responsible for the metabolism of fats, carbohydrates and proteins (Dainty and Blom, 1995; Ordóñez et al., 1999). The inside bacterial flora of fermented meat products is dominated by lactic acid bacteria (LAB), mainly Lactobacillus sakei, Lactobacillus curvatus and Lactobacillus plantarum (Hugas et al., 1993; Montel et al., 1998). By lowering the pH, they contribute to the improvement of food safety and stability of fermented foods (Ross et al., 2002). Since the flavor of fermented products could be improved by increasing levels of proteins, peptides and free amino acids, the proteolytic system of several LABs has been investigated (Kunji et al., 1996; Ordóñez et al., 1999; Vermeulen et al., 2005; Savijoki et al., 2006). Peptides are transported into LAB cells by three known systems. An Opp system is capable to transport oligopeptides consisting of

* Corresponding author. Tel.: þ49 8161 712912; fax: þ49 8161 712950. E-mail address: [email protected] (W. Schwab). 0740-0020/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.fm.2011.07.007

up to 35 amino acids (Doeven et al., 2005). A Dpp (previously referred to as DtpP) system delivers di-, tri-, and tetrapeptides with relatively hydrophobic branched-chain amino acids (BCAA) into the cell. This system displays the highest affinity for tripeptides (Foucaud et al., 1995; Sanz et al., 2003). Finally, a DtpT system prefers more hydrophilic and charged di- and tripeptides (Hagting et al., 1994). Free amino acids incorporated into LAB cells by amino acid transporters or released intracellular by peptidase activity can be converted, among other metabolic pathways, to various volatile compounds by transamination and decarboxylation (Christensen et al., 1999; van Kranenburg et al., 2002; Ordóñez et al., 1999; Smit et al., 2009). The transformation of the branched chain amino acids (BCAAs) valine, leucine and isoleucine, the aromatic amino acids tyrosine, tryptophan and phenylalanine and the sulfurcontaining amino acids methionine and cysteine by transamination and decarboxylation leads to volatile compounds which contribute to the sensory perception of fermented food (Ardö, 2006; Smit et al., 2005, 2009; Yvon and Rijnen, 2001). Some have been identified as major flavor contributor in Hungarian salami, such as 3methylbutanal, 3-methylbutanoic acid, 3-(methylthio)-propanal (methional) and phenylacetaldehyde that can result from the degradation of leucine, methionine and phenylalanine, respectively (Söllner and Schieberle, 2009). The metabolites 2-methylpropanal, 2-methylpropanol and 2-methylpropanoic acid formed by valine

216

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

degradation as well as 3-methylbutanol produced by leucine catabolism were also identified in high concentrations in dry fermented sausages (Flores et al., 2004; Marco et al., 2008; Partidário et al., 2006). It has been calculated that 11.8% of the volatile compounds in Milano salami probably originate from amino acid catabolism (Meynier et al., 1999). The lyase-mediated metabolism of cysteine and methionine also results in desired sulfur-containing flavor molecules in cheese (Bruinenberg et al., 1997; Dias and Weimer, 1998; Weimer et al., 1999). However, genes encoding enzymes involved in methionine and cysteine metabolism could not be identified in the genome sequence of L. sakei 23K (Liu et al., 2008). Studies with several Lactococcus lactis strains showed that an aminotransferase initiates the formation of methanethiol from methionine, although specific genes coding lyases (e.g. cystathionine g-lyase) were identified in these bacteria (Gao et al., 1998; Liu et al., 2008). L. sakei, formerly known as Lactobacillus sake, is one of the dominating LAB in dry spontaneously fermented sausages (Hugas et al., 1993; Trüpler and De’Clari, 1997). Therefore, it is commonly used in starter cultures for the production of dry fermented sausages, as it ensures the dominance of the starter during the whole ripening process (Hammes et al., 1990). Proteinase activities of whole cells and cell free extracts from some L. sakei strains have been studied by monitoring the hydrolysis of muscle myofibrillar and pork muscle sarcoplasmic proteins (Toldrá et al., 1992; Sanz et al., 1999; Fadda et al., 1999). Besides, several peptidases have been purified and studied in this species (Montel et al., 1995; Sanz and Toldrá, 1997; Sanz et al., 1998; Sanz and Toldrá, 2001, 2002). The analysis of the genome sequence of L. sakei 23K delivered insights into the strain’s potential concerning transport and proteolysis as well as aroma formation from amino acid substrates (Chaillou et al., 2005; Freiding et al., 2011; Liu et al., 2008, 2010). The studies showed that L. sakei strains possess a putative transport system for oligopeptides (Opp) as well as a di/tripeptides ABC transport system consisting of five subunits (DppA/P, DppB, DppC, DppD and DppF) and a di/tripeptides ion-linked transport system (DtpT). Furthermore, 18 peptidases with different specificities (unique aminopeptidases, endopeptidases, di/tripeptidases and proline peptidases) could be found. Although genes coding typical aminotransferases (araT und bcaT) could not be found in the genome sequence of more than 50 L sakei strains (Freiding et al., 2011; Liu et al., 2008), the formation of volatile amino acid derived metabolites could be demonstrated for L. sakei 23K (Larrouture et al., 2000). Pulsed-field gel electrophoresis analysis and DNAeDNA reassociation analysis showed high genetic heterogeneity (Chaillou et al., 2009; Champomier et al., 1987) whereas biochemical and physiological studies uncovered wide phenotypic heterogeneity within L. sakei strains (ChampomierVergès et al., 2001). This suggests that the metabolic potential of these strains can be as diverse. Metabolic studies to investigate the formation of 3-methylbutanal from leucine also showed a large variation between and within species and that intra-species variation is in many cases larger than that between species (Brandsma et al., 2008; Smit et al., 2004). Contrarily, a recent screening with 51 L sakei strains from different origins showed that all strains were nearly uniform in the genes forming the peptidolytic system (Freiding et al., 2011). Several studies with LAB showed a preferential uptake of peptides and an associated better growth of the microorganisms. (Foucaud et al., 2001; Kunji et al., 1996; Saguir et al., 2008; Smit and Konings, 1990). However, most studies on amino acid metabolism by lactobacilli have employed amino acids rather than peptides as substrates. The possibility to increase amino acid turnover by an optimized supply of peptide substrates has not been analyzed explicitly. Only one study showed that the metabolization of phenylalanine in

L. plantarum and L. sanfranciscensis could be enhanced by applying peptides compared to free amino acids (Vermeulen et al., 2006). In this study L. sakei’s potential to form volatile compounds from peptides and amino acids was investigated. Three strains were selected for a more detailed analysis on peptide metabolism by L. sakei (Freiding et al., 2011). Furthermore, two other bacteria commonly employed in starter cultures, L. plantarum and Staphylococcus carnosus (Hammes and Hertel, 1998), were chosen for comparison of the metabolic capacities. Resting cells were incubated in phosphate buffer with equimolar amounts of substrates consisting of dipeptides, tetrapeptides and free amino acids, respectively. Specific peptides were selected to analyze the preference of the cellular transport systems and the specificity of peptidases and enzymes responsible for the metabolism of amino acids. The degradation of peptides and amino acids was monitored by LC-UV/ESI-MSn and GC/MS analysis, the formation of volatile compounds was determined by SPME-GC/MS analysis. The aim of this study was to analyze the metabolic capacity of L. sakei strains to improve their flavor formation potential. 2. Materials and methods 2.1. Chemicals L-Leucine, L-valine, L-phenylalanine, L-methionine, peptone from casein, yeast extract, meat extract, ammonium chloride, cysteineeHCl, TweenÒ 80, D-maltose, D-glucose, K2HPO4  3H2O, KH2PO4, Na2HPO4 and NaH2PO4 were obtained from Carl Roth GmbH & Co. KG (Karlsruhe, Germany). 3-Methylbutanal, 3methylbutanol, 3-methylbutanoic acid, 2-methylpropanal, 2methylpropanoic acid, phenylacetaldehyde, benzaldehyde, methional (3-(methylthio)-propanal), dimethyldisulfide, pyridoxal-5-phosphate, 1,2-dimethoxy- ethane, a-ketoglutaric acid and vitamins and metals for mMRS medium were purchased from Sigma Chemical Co. (St. Louis, Missouri, USA). Peptides were ordered from GenScript USA Inc. (Piscataway, New Jersey, USA).

2.2. Bacterial strains, media and growth The strains used in this study were L. sakei TMW 1.1322, L. sakei TMW 1.1383, L. sakei TMW 1.1393 and L. plantarum TMW 1.708 as well as S. carnosus TMW 2.801 for comparison. They were all obtained from the laboratory collection of Technische Universität München, Lehrstuhl für Technische Mikrobiologie, Freising, Germany. L. sakei TMW 1.1322 and S. carnosus TMW 2.801 are isogenic clones of the genome sequenced strains L. sakei 23K (Chaillou et al., 2005) and S. carnosus TM300, which is phenotypically indistinguishable from the type strain DSM20501 (Rosenstein et al., 2009). In this communication the TMW clone numbers are used for correctness, as strains may change upon prolonged lab propagation. L. sakei TMW 1.1383 and TMW 1.1393 have been isolated from dry fermented sausages with a typical and poor aroma of dry fermented salami, respectively. The sausages have been evaluated by a sensory panel but the strains are probably not to blame for the different aromas (data not shown). The organisms were pre-cultured overnight at 30  C in 2 mL modified De Man Rogosa Sharpe medium (mMRS) (De Man et al., 1960; Stolz et al., 1995). A 20 mL aliquot of mMRS containing 0.3% D-glucose and no D-maltose to mimic conditions prevailing in fermenting sausage were inoculated with the overnight cultures and incubated for 5e7 h at 30  C. Cells were harvested by centrifugation (3000 g for 5 min) when reaching the stationary phase of growth, which was determined by measuring the optical density at 600 nm (O.D.600); 2.3e2.6 for L. sakei TMW 1.1322; 2.7e2.9 for L. sakei TMW 1.1383 and 2.0e2.2 for L. sakei TMW 1.1393.

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

2.3. Resting cells The harvested cells were washed twice with Na2HPO4/NaH2PO4 buffer (0.05 M; pH 6.5) and suspended in the same buffer to an O.D.600 of 1.0 (about 109 cfu/mL). Substrates (Table 1), D-glucose (0.3%), a-ketoglutaric acid (0.01 M) and pyridoxal-5-phosphate (0.002 M) were added (Larrouture et al., 2000). Dipeptides, tetrapeptides or free amino acids were used as substrates in equimolar quantities (Table 1). The chosen peptides contained the amino acids L-leucine, L-valine, L-phenylalanine and L-methionine. For each approach 5 mL aliquots of the resting cells suspensions were filled in three 20 mL headspace vials (VWR International GmbH, Darmstadt, Germany), 100 mL internal standard 1,2dimethoxy- ethane (19.24 mM) was added and incubated for 0, 2 and 4 or 5 days at 30  C for the determination of volatile metabolites by SPME-GC/MS. Controls without cells or without substrates were prepared in the same way. Each experiment and each approach was done in triplicate. To monitor the peptide, amino acid and D-glucose concentrations, 1 mL portions of resting cells were incubated in 1.5 mL reaction vessels for 0, 1, 2 and 5 days at 30  C. The same approaches and the analysis of volatile compounds were also performed in single determination with the strains L. plantarum TMW 1.708 and S. carnosus TMW 2.801. 2.4. Analysis of peptides by LC-UV/ESI-MSn Peptide levels in the reaction mixtures were measured by LC-UV/ ESI-MSn. Cells were removed by centrifugation and 5 mL of the supernatant was injected. The system used for LC-UV/ESI-MSn analysis was a Bruker Esquire 3000 plus mass spectrometer (Bruker, Bremen, Germany), equipped with an Agilent 1100 HPLC system composed of an Agilent 1100 quaternary pump and an Agilent 1100 variable wavelength detector (Agilent, Waldbronn, Germany). The column was a Luna 3u C18(2) 100A, 15 cm  2 mm (Phenomenex, Table 1 Substrate combinations, concentrations and measurements performed with resting cells. Approach

Concentration

Measurement

Dipeptides LEU_MET experiment Resting cells Resting cells þ L-M Resting cells þ M-L Resting cells þ L þ M

10 mM 10 mM Each 10 mM

PEP PEP PEP PEP

AA AA AA AA

Glu Glu Glu Glu

SPME-GC/MS SPME-GC/MS SPME-GC/MS SPME-GC/MS

PHE_VAL experiment Resting cells Resting cells þ F-V Resting cells þ V-F Resting cells þ F þ V

10 mM 10 mM Each 10 mM

PEP PEP PEP PEP

AA AA AA AA

Glu Glu Glu Glu

SPME-GC/MS SPME-GC/MS SPME-GC/MS SPME-GC/MS

LEU_LEU experiment Resting cells Resting cells þ L-L Resting cells þ L þ L

10 mM Each 10 mM

PEP PEP PEP

AA AA AA

Glu Glu Glu

SPME-GC/MS SPME-GC/MS SPME-GC/MS

Tetrapeptides PHE_MET_LEU_VAL experiment Resting cells Resting cells þ F-M-L-V 10 mM Resting cells þ M-F-V-L 10 mM Resting cells þ F þ M þ L þ V Each 10 mM

PEP PEP PEP PEP

AA AA AA AA

Glu Glu Glu Glu

SPME-GC/MS SPME-GC/MS SPME-GC/MS SPME-GC/MS

Resting cells were incubated with the indicated substrates at the indicated concentrations and the medium was analyzed by the indicated methods. PEP analysis of peptide concentrations by LC-UV/ESI-MSn; AA analysis of free amino acid concentrations by GC/MS; Glu determination of D-glucose concentration by enzymatic reaction; SPME-GC/MS analysis of volatile compounds by GC/MS, extraction method: solid phase micro extraction.

217

Torrance, California, USA). The ionization parameters were as follows: the voltage of the capillary was 4000 V and the end plate was set to 500 V. The temperature of the dry gas (N2) was 330  C at a flow of 9 L/min. The full scan mass spectra of the metabolites were measured from m/z 50 to 800 until the ICC target reached 20000 or 200 ms, whichever was reached first. Tandem mass spectrometry was performed using helium as the collision gas, and the collision energy was set at 1 V. The target mass for MS2 spectra was set to m/z 400. Mass spectra were acquired in the negative and positive ionization mode. Auto-tandem mass spectrometry was used to breakdown the most abundant [M þ H]þ or [M  H] ion. The LC solvents were 0.1% formic acid in water (A) and 0.1% formic acid in methanol (B). The gradient went from 0% B and 100% A to 50% B and 50% A in 30 min, then in 5 min to 100% B, remained for 15 min at these conditions, returning to 100% A and 0% B in 5 min at a flow rate of 0.2 mL/min. The detection wavelength was 280 nm. Data analysis was performed using the Data Analysis 3.1 software (Bruker Daltonics, Bremen, Germany). 2.5. Analysis of free amino acids by GC/MS The concentration of free amino acids in the reaction mixtures was measured by GC/MS after precolumn derivatization using the EZ:faastÔ-kit (Phenomenex, Torrance, California, USA). After separation from cells by centrifugation and dilution, 100 mL of the supernatant were used for derivatization as indicated by the manufacturer. Intracellular amino acid levels were determined by analysis of free amino acid concentrations before and after lysis of the cells by sonication. The system used for GC/MS analysis was an Agilent 6890N gas chromatograph equipped with an Agilent 7683B injector, an Agilent 7683 autosampler and an Agilent 5975 mass selective detector (Agilent Technologies Inc., Santa Clara, California, USA). A 2 mL aliquot of the derivatized sample was injected into the injection port of the gas chromatograph at 250  C with the purge valve on (split mode), split ratio 15:1 and split flow 16.5 mL/min. The compounds were separated with a ZB-AAA Zebron Amino Acid column delivered with the EZ:faastÔ-kit (10 m, 0.25 mm i.d.). Helium was used as carrier gas with a constant flow of 1.1 mL/min and an average velocity of 68 cm/s. The initial GC oven temperature was 110  C, ramped to 320  C at 30  C/min and held at 320  C for 5 min. The total run time with a post run time of 5 min at 320  C was 12 min and the GC-mass spectrometer interface was maintained at 310  C. Mass spectra were obtained in the scan mode within a mass range of m/z 45e450 with a threshold of 150 and gain factor of 1. Ionization was performed by electron impact at 70 eV, calibration was performed by autotuning. The amino acids were identified by comparison with the EZ:faastÔ database. Quantification was done by calibration of each amino acid with the delivered standards as indicated by the manufacturer. Data analysis was performed using the MSD ChemStation E.02.00.493 software (Agilent Technologies Inc., Santa Clara, California, USA). 2.6. Analysis of D-glucose by enzymatic reaction Experiments showed that the addition of D-glucose to the resting cells led to an increased formation of volatile compounds from amino acids (data not shown). For this reason 3 g/L D-glucose was added to the reaction mixture. To monitor the concentration of D-glucose during fermentation, the enzymatic reaction kit for Dglucose (R-Biopharm AG, Darmstadt, Germany) was used. After separation from cells by centrifugation and dilution, 100 mL of the supernatant were used for the determination of D-glucose as indicated by the manufacturer. A spectral photometer (Nicolet Evolution 100, Thermo Electron Corporation, Austin, Texas, USA) was used at 340 nm to determine the formation of NADPH/Hþ.

218

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

2.7. Analysis of volatile compounds by SPME-GC/MS GC/MS with solid phase micro extraction (SPME) for sample collection was used to determine the amount of volatile compounds. The reaction mixtures were incubated in the heated agitator of the autosampler at 30  C until sample collection by SPME. The extraction of headspace volatile compounds was done using an SPME device (Supelco, Bellefonte, Pennsylvania, USA), equipped with a 75 mm carboxen/polydimethylsiloxane (CAR/ PDMS) fiber and operated by the autosampler of the GC/MS system. Before analysis the fiber was preconditioned in the injection port of the GC as indicated by the manufacturer. Then, the SPME fiber was exposed to the headspace while maintaining the sample at 30  C for 30 min. The compounds adsorbed by the fiber were desorbed in the injection port of the gas chromatograph (Agilent 7890A) for 10 min at 250  C with the purge valve off (splitless mode). The compounds were separated with a ZB-Wax capillary column (60 m, 0.25 mm i.d., film thickness 0.25 mm; Zebron, Phenomenex, Torrance, California, USA). The GC was equipped with an Agilent 5975C mass selective detector (Agilent Technologies Inc., Santa Clara, California, USA) and a CTC CombiPAL autosampler (CTC Analytics AG, Zwingen, Switzerland) for SPME sample collection. Helium was used as carrier gas with a constant flow of 1.03 mL/min and an average velocity of 25.88 cm/s. The GC oven temperature program started when the fiber was inserted and was held at 30  C for 15 min, ramped to 50  C at 3  C/min, then to 110  C at 4  C/min, to 150  C at 5  C/min, and to 250  C at 10  C/min and finally, held at 250  C for 10 min. The total run time was 64.67 min and the GC/MS interface was maintained at 250  C. Mass spectra were obtained in the scan mode within a mass range of m/z 29e150 with a threshold of 150 and gain factor of 0.15. Ionization was performed by electron impact at 70 eV, calibration was performed by autotuning. The volatile compounds were identified by comparison with mass spectra from a NIST database (NIST MS Search 2.0; FairCom Corporation, Columbia, Missouri, USA) and by comparison of mass spectra and retention times with those of authentic standards. Data analysis was performed using the MSD ChemStation E.02.00.493 software (Agilent Technologies Inc., Santa Clara, California, USA). Quantification of volatiles with the help of relative response factors was not possible when using SPME for sampling. To account for the varying sensitivity of SPME-GC/MS, the internal standard 1,2dimethoxy ethane was added in all the samples and calibration mixtures (19.24 mM). Calibration curves of corresponding relative concentrations using increasing concentrations of volatiles and a defined level of the internal standard 1,2-dimethoxy ethane (19.24 mM) were determined by SPME-GC/MS. The calibration curves could be described by polynomial equations. For each substance, triplicate measurement of the calibration curve was performed and the average was calculated. For the determination of volatile metabolites in the samples, the results from three measuring points of each approach (day 0, 2 and 4 or 5, respectively) were taken to describe a function of amount over time. The maximum amount of phenylacetaldehyde was detected at day 4 and for all other metabolites at day 5. A list of the monitored metabolites is shown in Table 2. 3. Results and discussion 3.1. Fate of peptides and amino acids The ability of three L. sakei strains to convert different substrates into volatile flavor compounds was investigated. Whole resting cells were fermented in phosphate buffer with equimolar amounts of dipeptides, tetrapeptides or free amino acids. LC-UV/ESI-MSn measurement of the dipeptide levels in the reaction mixtures

Table 2 Volatile metabolites monitored by SPME-GC/MS. Amino acid

Volatile compound

L-valine

2-methylpropanal 2-methylpropanoic acid 3-methylbutanal 3-methylbutanol 3-methylbutanoic acid Phenylacetaldehyde Benzaldehyde 3-(methylthio)-propanal (methional) Dimethyldisulfide

L-leucine

L-phenylalanine L-methionine

showed that added dipeptides disappeared almost completely from the solutions within one day independent from the strain used. On the contrary, tetrapeptides were partially degraded but still present in large quantities after fermentation. Due to the poor solubility of tetrapeptides in aqueous systems, these substrates were offered as suspensions. Therefore, accurate quantification of tetrapeptide concentrations by LC-UV/ESI-MSn was not feasible. However, a degradation of the tetrapeptides during the 5-day fermentation period could be monitored visually. A decrease in the cloudiness of the suspension could be observed. A possible reason for the lower uptake of tetrapeptides is the lack of transport systems for oligopeptides in L. sakei. Only one putative transport system for oligopeptides could be found in the genome sequence of L. sakei (Freiding et al., 2011; Liu et al., 2010). However, homologous genes coding for a di/tripeptide ABC transport system consisting of five subunits (DppA/P, DppB, DppC, DppD and DppF) and a di/tripeptide ion-linked transport system (DtpT) have been identified. The levels of free amino acids were quantified in the reaction mixtures during the fermentation of the di- and tetrapeptides. Leucine levels of 2.9e6.2 mM could be found in the experiments with the dipeptides (10 mM each) L-M and M-L whereas valine concentrations of 2.2e4.0 mM were determined after fermentation of P-V and V-P (Fig. 1). In the same experiments the methionine levels accounted for 2.4e5.5 mM and the phenylalanine concentrations for 0.6e2.5 mM (data not shown). Therefore, it can be assumed that the peptides were partially hydrolyzed to their amino acid components. Of the strains analyzed L. sakei TMW 1.1383 produced the lowest levels of leucine and methionine in the LEU_MET experiment but released the highest amounts of valine and phenylalanine in the PHE_VAL approach. Higher levels of free amino acids were released from M-L compared to L-M by all strains (Fig. 1). L. sakei strains also differed in their ability to hydrolyze the tetrapeptides. The highest levels of amino acids were released from the tetrapeptide F-M-V-L (10 mM) by the strain L. sakei TMW 1.1393, with a maximum phenylalanine concentration of 1.9 mM. The strains TMW 1.1322 and TMW 1.1383 delivered only 1.0 mM and 1.5 mM of phenylalanine, respectively from this substrate. In the experiments with the isomeric tetrapeptide M-F-V-L, there was a maximum increase in the concentration of phenylalanine of 0.5 mM by L. sakei TMW 1.1383, whereas the other strains released 0.3 mM (TMW 1.1322) and 0.2 mM (TMW 1.1393) of phenylalanine. The other amino acids were released in concentrations of 0.3e1.6 mM from F-M-L-V and 0.1e0.4 mM from M-F-V-L (data not shown). When free amino acids (10 mM) were added to resting cells their levels changed only slightly. During the fermentation period barely 1.8 mM of the acids disappeared from the medium (Fig. 1). Thus, free amino acids represent poor substrates. The controls were always devoid of amino acids (Fig. 1). From these results we concluded that dipeptides are rapidly absorbed unchanged and hydrolyzed extracellularly to some extent by L. sakei strains. As the extracellular proteolytic activity of L. sakei

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

219

medium. It seems that this concentration was sufficient for the energy production of the bacteria because a constant level of Dglucose could be measured after a fermentation period of 48 h (1.5 g/L in the case of dipeptides and 1.0 g/L in the case of free amino acids and the control). Independent of the L. sakei strain used the level of D-glucose in the experiments with peptide addition decreased by about 0.7 g/L to 0.8 g/L, whereas in batches with addition of free amino acids or in controls without substrate it was reduced by 1.3e1.8 g/L (Fig. 2). The lower consumption of D-glucose in the presence of peptides in the fermentation buffer can be explained by a positive energy balance of the peptides in the cell. A lower energy consumption for the uptake of peptides in comparison to free amino acids was also observed in Pediococcus pentosaceus (Aredes Fernández et al., 2003) and differences in the consumption of D-glucose were noticed when dipeptides and free amino acids were added to MRS medium with L. plantarum (Saguir et al., 2008). 3.3. Formation of volatile compounds from amino acid substrates

Fig. 1. Comparative plot showing the net increment of free leucine [mMol/L] in the experiment LEU_MET (top, Table 1) and valine [mMol/L] in the experiment PHE_VAL (bottom, Table 1) by L. sakei TMW 1.1322, TMW 1.1383 and TMW 1.1393. Each bar represents a strain (caption), control I ¼ Na2HPO4/NaH2PO4 buffer inoculated without substrates.

strains appears to be low (Freiding et al., 2011; Sanz et al., 1999), the hydrolysis of the peptides seems to occur by intracellular peptidases that are released from dead and lysed cells. However, amino acids could also be transported out of the cells. Since only an average of 3.3 mM of amino acids was liberated from 10 mM of dipeptides and only 0.9 mM of free amino acids was absorbed by the cells it can be proposed that 5.8 mM of intact dipeptides was taken up. The intracellular level of amino acids was negligible (data not shown). This suggests that a significant proportion of the peptides was transported into LAB cells as intact dipeptides. In the cells they were probably hydrolyzed and converted to metabolites that were not accounted for in this study. Similar results were obtained by Stuart et al. (1999) and Kunji et al. (1993) with L. lactis and Mandelstam (1958) with Escherichia coli. They showed that the ability of the bacteria to survive periods of carbohydrate starvation may depend on their ability to transport and metabolize certain amino acids. As the synthesis of new proteins during starvation was observed, they proposed that amino acids may prolong the survival by providing an energy source, supplying amino acids for cell turnover, and minimizing the breakdown of essential proteins. In L. plantarum higher dipeptide utilization could be observed, suggesting that a peptide hydrolase system with high activity was present in the cells. The accumulation of amino acids from dipeptides internally was higher than from free amino acids confirming the main role of dipeptides as nitrogen sources for microorganism growth (Saguir et al., 2008). In Lactobacillus sanfranciscensis and L. plantarum inefficient transport of single amino acids was demonstrated, which resulted in low intracellular amino acid concentrations and poor amino acid conversion. On the contrary, the peptides used in this study were all quantitatively hydrolyzed by the intracellular peptidases (Vermeulen et al., 2006).

The levels of volatile compounds formed by L. sakei strains TMW 1.1322 (A), TMW 1.1383 (B) and TMW 1.1393 (C) in different fermentation broths were quantified by SPME-GC/MS (Table 3). Levels were compared with those measured in media without substrate (control I) or without a LAB strain (control II). Small amounts of the volatile metabolites could also be found in the control experiments (Table 3). However, the concentration of volatiles in the reaction mixtures with substrates and inoculated with cells was 2e400 times higher, depending on the metabolite. All L. sakei strains were able to metabolize the given substrates to volatile compounds. In the approach LEU_MET (Table 1) the level of 3-methylbutanal produced from the dipeptide M-L was always higher than from L-M for all LAB (Table 3). Similar to these results higher levels of 3(methylthio)-propanal and dimethyldisulfide were produced from the same dipeptide (Table 3). This suggests a hydrolytic preference of the bacterial strains for M-L as they released higher concentrations of free amino acids from M-L compared to L-M (Fig. 1). This observation points to a different metabolism of the structural isomeric peptides. An aminopeptidase has already been purified from L. sakei and characterized (Sanz and Toldrá, 1997). It showed different KM values for Met-AMC (7-amido-4-methylcoumarin) compared to Leu-AMC and therefore different affinities towards Nterminal amino acid residues. The detected amount of dimethyldisulfide varied within the experimental approaches and from strain to strain. Only L. sakei TMW 1.1393 (C) was able to form the metabolite 3-(methylthio)-propanal (methional) from the dipeptide L-M whereas all strains were able to produce this compound from M-L (Table 3). Dimethyldisulfide is assumed to be formed by

3.2. Consumption of D-glucose Preliminary studies have shown that resting cells produced maximum levels of volatiles after addition of 0.3% D-glucose to the

Fig. 2. Comparative plot showing the net decrease of L-glucose [g/L] in the experiment LEU_MET by L. sakei TMW 1.1322, TMW 1.1383 and TMW 1.1393. Each bar represents a strain (caption), control I ¼ Na2HPO4/NaH2PO4 buffer inoculated without substrates.

220

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

Table 3 SPME-GC/MS analysis of volatiles. Volatile compound Approach L-M A 3-Methylbutanal 3-Methylbutanol 3-Methylbutanoic acid 3-(Methylthio)propanal Dimethyldisulfide

2-Methylpropanal 2-Methylpropanoic acid Phenylacetaldehyde Benzaldehyde

3-Methylbutanal 3-Methylbutanol 3-Methylbutanoic acid 2-Methylpropanal 2-Methylpropanoic acid Phenylacetaldehyde Benzaldehyde 3-(Methylthio)propanal Dimethyldisulfide

C

M-L A

22 (5) 43 (14) 56 (16) 61 (7) 32 (11) 17 (4) 14 (5) 32 (9) 5 (2) 35 (3) 16 (0) 11 (5)

B

C

LþM A

B

86 (9) 19 (1) 27 (2)

94 (4) 10 (2) 23 (1)

104 (8) 91 (15) 32 (6)

102 (12) 163 (9) 64 (2) 71 (18) 46 (18) 36 (1)

C

Control I A B

C

Control II L-M M-L

LþM

1 (2) 1 (1) 1 (2)

2 (3) tra 1 (1) 4 (7) 17 (1) 2 (3)

2 1 nd

4 2 nd

13 (3) 8 (3) 1 (1)

nd

nd

67 (67) 122 (25) 176 (36) 189 (50) 257 (68) 295 (46) 347 (80)

nd

nd

nd

nd

nd

nd

4 (1)

2 (3)

1 (0)

19 (6)

19 (2)

nd

nd

nd

1

tra

2 (1)

C

V-F A

C

Control I A B

C

Control II F-V V-F

FþV

1 (0) nd

nd nd

tra nd

1 nd

tra nd

7 (2) nd

tra 7 (11) nd 38 (1) 34 (12) 7

nd 6

17 (1) 2 (1)

F-V A

B

1 (0) nd

20 (11) 1 (0) 11 (5) nd

7 (6)

18 (4)

13 (1)

B

C

FþV A

8 (5) B

22 (7) 2 (1)

1 (0) nd

28 (19) nd

53 (16) 37 (14) 6 (1) nd

6 (5) 18 (13) 23 (4) 8 (5) 18 (12) 30 (4) 56 (2) 102 (18) 69 (21) nd 143 (51) 247 (94) 214 (40) 200 (73) 272 (85) 424 (23) 329 (101) 601 (88) 630 (102) 22 (2) L-L A

3-Methylbutanal 3-Methylbutanol 3-Methylbutanoic acid

B

B

LþL A

C

147 (23) 138 (2) 153 (42) 31 (18) 20 (4) 7 (4) 40 (10) 46 (10) 22 (4)

B

C

Control I A B

253 (79) 237 (66) 311 (139) 2 (2) 192 (63) 150 (33) 89 (21) tra 61 (23) 53 (10) 43 (12) 15 (0)

F-M-L-V A B

C

14 (5) 6 (3) nd

10 (1) 2 (2) nd

31 (10) 5 (1) 6 (3) 10 (4) nd nd

tra nd

tra nd

1 (0) nd

C

FþMþLþV A B

6 (2) 5 (2) nd

8 (3) 7 (1) nd

tra nd

tra nd

M-F-V-L A B

tra nd

tra nd

C

Control II L-L

LþL

tra tra 9 (0) nd nd 15 (0) 19 (7) 16 (8) nd

23 (2) 16 (1) tra

Control I A B

C

Control II F-M-L-V M-F-V- F þ M þ L þ V L

134 (22) 158 (24) 168 (47) 96 (12) 36 (4) 103 (18) 12 (3) 23 (1) 42 (4)

nd nd nd

nd tra nd

nd 3 (3) nd

nd nd nd

nd nd nd

13 (3) 8 (3) 1 (1)

11 (2) 5 (1)

nd nd

nd nd

nd nd

tra nd

tra nd

7 (2) nd

nd 8 (4) nd

nd 2 (4) nd 24 (11) 23 (21) 34 nd nd nd

nd 22 nd

17 (1) 2 (1) nd

nd

nd

7

2 (1)

4 (2) nd

C

29 (10) 59 (20)

2 (4) nd 45 (14) nd nd 56 (12) 208 (48) 135 (47) 635 (170) 199 (20) 306 (93) 235 (26) 66 (16) 126 (35) 76 (18) 744 (95) 657 (47) 964 (157) nd nd nd nd nd nd 671 (154) 503 (30) 1126 (260) 1 (0) 1 (1) 7 (5) 14 (7) 9 (1) 17 (5) 36 (7) 16 (2) 53 (24)

nd

1

Concentrations in [mg/L] for 2-methylpropanoic acid and 3-methylbutanoic acid; [mg/L] for all other metabolites; standard deviation shown in parenthesis; nd ¼ not detected, tra ¼ in traces (<1 ng/mL and mg/mL) Strains: A ¼ L. sakei TMW 1.1322, B ¼ L. sakei TMW 1.1383, C ¼ L. sakei TMW 1.1393Controls: Control I: without substrates, Control II: without strains.

oxidation of methanethiol which is produced by elimination either from methional or from methionine (Gao et al., 1998). Since dimethyldisulfide could also be found in the experiments with LM-addition, it can be proposed that methional was formed previously. As none of the genes encoding enzymes involved in methionine and cysteine metabolism in LAB could be identified in the genome sequence of L. sakei TMW 1.1322 (Liu et al., 2008), it is unlikely that dimethyldisulfide was produced via an alternative pathway catalyzed by cystathionine g-lyase. However, the highest amount of volatile metabolites was formed when the free amino acids methionine and leucine were used as substrates for the three LAB strains. Although amino acids were poorly absorbed by the bacterial cells they delivered significantly higher levels of aroma compounds. In the experiment PHE_VAL (Table 1), a strain specific preference for the production of the metabolites 2-methylpropanal and 2-methylpropanoic acid was observed. The highest levels of these metabolites were formed by strain L. sakei TMW 1.1383 (B) when FV, V-F or the amino acids F and V were added to the reaction mixture (Table 3), confirming the hydrolytic preference of this bacterial strain (Fig. 1). A similar preference could not be observed for volatile metabolites originating from phenylalanine. Therefore, it is more likely that the strains differ in their ability to metabolize the amino acid valine than in their uptake capacity for the dipeptides or amino acids.

The experiment LEU_LEU provided a direct comparison of the transport and metabolism of an amino acid (leucine) and its corresponding dipeptide (Table 1). Although equimolar amounts of substrates were used (10 mM L-L and 20 mM L) the highest levels of volatile metabolites were always detected after the addition of the free amino acid leucine (Table 3). The resulting amounts of metabolites varied between the L. sakei strains. In a similar study the catabolism of leucine by resting cells of L. sakei TMW 1.1322 has yielded mainly a-ketoisocaproic acid and very low amounts of 3methylbutanoic acid (Larrouture et al., 2000). Less than 5% of the initial amount has been degraded. The alternative product ahydroxyisocaproic acid formed from a-ketoisocaproic acid by dehydrogenase has not been observed. The fact, that the L. sakei strains produced the largest amount of volatile metabolites when free amino acids were added was even more pronounced in the experiment PHE_MET_LEU_VAL (Table 1). LAB produced higher levels of 3-methylbutanal from the tetrapeptide F-M-L-V, than from the substrate M-P-V-L (Table 3). This result was in agreement with the previous observation that the dipeptide M-L was more efficiently converted into 3-methylbutanal than L-M. 3-Methylbutanoic acid was only detected in the experiments with the free amino acids. However, the detection limit (mg/ L) for this metabolite is quite high compared to the other volatiles (mg/L). Very small amounts of metabolites originating from valine (2-methylpropanal and 2-methylpropanoic acid) were formed from

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

tetrapeptides by LAB. The strain L. sakei TMW 1.1393 (C) produced larger quantities of both metabolites from the free amino acids than the other two strains used. The same strain showed a preference for the formation of phenylacetaldehyde from tetrapeptides and phenylalanine, consistent with the PHE_VAL experiment (Table 3). The detected amounts of benzaldehyde showed, that the tetrapeptide F-M-L-V was metabolized to a lager extent by LAB than MF-V-L. Similarly, higher levels of amino acids were released from FM-L-V (2.6e6.1 mM free amino acids in total) into the buffer compared to M-F-V-L (0.7e1.5 mM free amino acids in total). Methional (3-(methylthio)-propanal) was not formed from tetrapeptide substrates by any strain used. A preference of the peptide M-F-V-L as a substrate for the formation of dimethyldisulfide could be observed. This is consistent with the results of the LEU_MET experiment in which M-L (methionine at the N-terminus) produced higher levels of the disulfide. Generally, strain L. sakei TMW 1.1393 (C) formed larger amounts of metabolites from the free amino acids leucine, valine, phenylalanine and methionine than the other strains. This strain appears to produce more volatiles from amino acids. In this study free amino acids were identified as the most efficient precursors for the formation of aroma active volatiles by LAB although the amounts were low compared to the amino acids levels used. Similarly, a poor correlation of amino acid catabolism by LAB and the production of corresponding volatile metabolites have been shown (Tavaria et al. (2002)). Most studies, in which the preferential uptake of peptides versus free amino acids has already been demonstrated, were limited to the description of the growth behavior of the microorganisms and the hydrolysis of the peptide substrates (Foucaud et al., 2001; Kunji et al., 1996; Saguir et al., 2008; Smit and Konings, 1990). Only recently, a study was extended to the formation of metabolites (Vermeulen et al., 2006). However, all experiments were performed in chemically defined media, rich media or in food. In this study resting cells were chosen to exclude the effects of media components and to provide only one substrate for the bacteria. Hence, the uptake and metabolism of the substrates by the bacteria in the stationary phase regardless of growth was observed. We were able to demonstrate the preferential uptake of dipeptides in comparison to free amino acids and tetrapeptides. However, this did not lead to an enhanced formation of the monitored aroma-relevant volatile metabolites (Table 2). Therefore it is suggested, that the peptides after transport into the cell are mainly used for other purposes rather than the production of these volatiles. The entry port which mediates the transport (as dipeptide or free amino acid) into the cell appears to govern the fate of the amino acid in the cell.

221

Fig. 3. Comparative plot showing the measured concentrations of 3-methylbutanal [mg/L] from L-leucine in the experiment LEU_MET (top) and 2-methylpropanoic acid [mg/L] from L-valine in the experiment PHE_VAL (bottom) by L. sakei TMW 1.1322, TMW 1.1383, TMW 1.1393, L. plantarum and S. carnosus. Each bar represents an approach (caption), control I ¼ Na2HPO4/NaH2PO4 buffer inoculated without substrates. The control values determined for the substrates (Na2HPO4/NaH2PO4 buffer uninoculated with substrates, control II) were deducted.

methylpropanal and methional and higher levels of 2methylpropanoic acid and dimethyldisulfide than the L. sakei strains (Figs. 3 and 4). Only few studies have been carried out on the metabolism of amino acids by resting cells of different starter

3.4. Comparison of L. sakei with other starter cultures The comparison with other starter cultures used for the production of dry fermented sausage showed significant differences in the extent of the metabolism of the substrates (Fig. 3). In particular S. carnosus TMW 2.801 produced significantly higher amounts of metabolites (e.g. 3-methylbutanal and 2methylpropanoic acid) from the branched-chain amino acids leucine and valine than LAB. This strain also formed higher concentrations of 3-methylbutanal from leucine bound peptides than from the free amino acid. However, less volatile compounds (e.g. phenylacetaldehyde and dimethyldisulfide) were produced by S. carnosus from the aromatic amino acid phenylalanine and the sulfur-containing methionine compared to LAB (Fig. 4). The peptide and amino acid metabolism of L. plantarum TMW 1708 was more related to the L. sakei strains. It produced similar amounts of 3methylbutanol, benzaldehyde and phenylacetaldehyde but formed lower levels of 3-methylbutanal, 3-methylbutanoic acid, 2-

Fig. 4. Comparative plot showing the measured concentrations of dimethyldisulfide [mg/L] from L-methionine in the experiment LEU_MET (top) and phenylacetaldehyde [mg/L] from L-phenylalanine in the experiment PHE_VAL (bottom) by L. sakei TMW 1.1322, TMW 1.1383, TMW 1.1393, L. plantarum and S. carnosus. Each bar represents an approach (caption), control I ¼ Na2HPO4/NaH2PO4 buffer inoculated without substrates. The control values determined for the substrates (Na2HPO4/NaH2PO4 buffer uninoculated with substrates, control II) were deducted.

222

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223

cultures. It has been shown that the catabolism of leucine by LAB was very low (Larrouture et al., 2000). Although leucine was degraded to a-ketoisocaproic acid by L. plantarum the formation of aroma-relevant metabolites were not observed (Larrouture et al., 2000). In accordance with our results, S. carnosus proved to be an efficient producer of the important aroma compounds 3methylbutanal, 3-methylbutanol, and 3-methylbutanoic acid in dry fermented sausages (Masson et al., 1999; Montel et al., 1996; Söllner and Schieberle, 2009; Stahnke, 1999). 4. Conclusion In biochemical and physiological studies a wide phenotypic heterogeneity within L. sakei strains has been reported (Champomier-Vergès et al., 2001). Similarly, genetic heterogeneity within different accessions also suggested that the metabolic potential of these strains should be as diverse (Chaillou et al., 2009; Champomier et al., 1987). By comparing three different strains of L. sakei, some substrate-specific differences in the extent of metabolite formation and substrate preference were observed. Overall, the metabolic potential of L. sakei for the production of aroma-relevant metabolites was low compared to S. carnosus. However, since LAB reach high bacteria counts at the end of the ripening period of dry fermented sausages, their effects on the overall flavor of dry fermented sausages should not be underestimated. We conclude that peptides and amino acids are catabolized differently within L. sakei cells. While dipeptides are more efficiently taken up by LAB than amino acids, higher levels of volatiles were produced from amino acids. Besides, less D-glucose is consumed by the strains in the presence of peptides compared to amino acids. Therefore, it is assumed that the energy-rich amide bonds of the peptides are exploited by the cells. Compared to tetrapeptides, dipeptides were preferred by the cellular transport systems and peptidases. Several studies have already shown the preferential uptake of peptides, an associated better growth of the microorganisms and that peptides are more efficiently converted to metabolites compared to free amino acids. In our experiments L. sakei strains yielded the highest amounts of aroma active metabolites when free amino acid substrates were used. Accordingly, the presence of aminopeptidases and dipeptidases is of vital importance. Proteinase activities of whole cells and cell free extracts from L. sakei strains have been determined and release of free amino acids from muscle myofibrillar and pork muscle sarcoplasmic proteins by LAB have been demonstrated (Toldrá et al., 1992; Sanz et al., 1999; Fadda et al., 1999). Furthermore, structural isomers of peptides were metabolized differently, demonstrating the specificity and complexity of the transport and peptidolytic systems of these microorganisms. In our study L. sakei TMW 1.1322 which is an isogenic clone of the genome sequenced strain 23K did not show striking metabolic differences compared to the other strains used. Genes coding typical aminotransferases or lyases could not be identified in the genetic equipment of L. sakei TMW 1.1322, TMW 1.1383 and TMW 1.1393 (Freiding et al., 2011; Liu et al., 2008). As the formation of volatile metabolites via transamination has been demonstrated for L. sakei TMW 1.1322 (Larrouture et al., 2000), the metabolic pathway of amino acids in these bacteria needs to be elucidated. The poorer formation of volatiles by L. sakei strains compared to S. carnosus could rely on the lack of specific aminotransferases in these strains. Acknowledgments This research project was supported by the German Ministry of Economics and Technology and the FEI (Forschungskreis der

Ernährungsindustrie e.V., Bonn) in project AiF 15458 N. We thank Prof. Dr. Vogel, Lehrstuhl für Technische Mikrobiologie, Technische Universität München and his staff for providing the bacterial strains and their help. Appendix. Abbreviations

BCAAs LAB mMRS SPME TMW

branched-chain amino acids lactic acid bacteria modified MRS solid phase micro extraction Technische Mikrobiologie Weihenstephan

References Ardö, Y., 2006. Flavour formation by amino acid catabolism. Biotechnology Advances 24, 238e242. Brandsma, J.B., Floris, E., Dijkstra, A.R.D., Rijnen, L., Wouters, J.A., Meijer, W.C., 2008. Natural diversity of aminotransferases and dehydrogenase activity in a large collection of Lactococcus lactis strains. International Dairy Journal 18, 1103e1108. Bruinenberg, P.G., de Roo, G., Limsowtin, G.K.Y., 1997. Purification and characterization of cystathionine g-lyase from Lactococcus lactis subsp. cremoris SK11: possible role in flavor compound formation during cheese maturation. Applied and Environmental Microbiology 63, 561e566. Chaillou, S., Champomier-Vergès, M.-C., Cornet, M., Crutz-Le Cop, A.-M., Dudez, A.-M., Martin, V., Beaufils, S., Darbon-Rongère, E., Bossy, R., Loux, V., Zagorec, M., 2005. The complete genome sequence of the meat-borne lactic acid bacterium Lactobacillus sakei 23K. Nature Biotechnology 23, 1527e1533. Chaillou, S., Daty, M., Baraige, F., Dudez, A.-M., Anglade, P., Jones, R., Alpert, C.-A., Champomier-Vergès, M.-C., Zagorec, M., 2009. Intraspecies genomic diversity and natural population structure of the meat-borne lactic acid bacterium Lactobacillus sakei. Applied and Environmental Microbiology 75, 970e980. Champomier, M.-C., Montel, M.-C., Grimont, F., Grimont, P.A.D., 1987. Genomic identification of meat lactobacilli as Lactobacillus sake. Annales de l’Institut Pasteur/Microbiologie 138, 751e758. Champomier-Vergès, M.-C., Chailou, S., Cornet, M., Zagorec, M., 2001. Lactobacillus sakei: recent developments and future prospects. Research in Microbiology 152, 839e848. Christensen, J.E., Dudley, E.G., Pederson, J.A., Steele, J.L., 1999. Peptidases and amino acid catabolism in lactic acid bacteria. Antonie van Leeuwenhoek 76, 217e246. Dainty, R., Blom, H., 1995. Flavour chemistry of fermented sausages. In: CampbellPlatt, G., Cook, P.E. (Eds.), Fermented Meats. Blackie Academic and Professional, London, pp. 176e193. Dias, B., Weimer, B., 1998. Conversion of methionine to thiols by Lactococci, Lactobacilli, and Brevibacteria. Applied and Environmental Microbiology 64, 3320e3326. Doeven, M.K., Kok, J., Poolman, B., 2005. Specificity and selectivity determinants of peptide transport in Lactococcus lactis and other microorganisms. Molecular Microbiology 57, 640e649. Fadda, F., Sanz, Y., Vignolo, G., Aristoy, M.-C., Oliver, G., Toldrá, F., 1999. Hydrolysis of pork muscle sarcoplasmic proteins by Lactobacillus curvatus and Lactobacillus sake. Applied and Environmental Microbiology 65, 578e584. Aredes Fernández, P.A., Saguir, F.M., Mance de Nadra, M.C., 2003. Effect of amino acids and peptides on growth of Pediococcus pentosaceus from wine. Latin American Applied Research 33, 225e229. Flores, M., Durá, M.-A., Marco, A., Toldrá, F., 2004. Effect of Debaryomyces spp. on aroma formation and sensory quality of dry-fermented sausages. Meat Science 68, 439e446. Foucaud, C., Kunji, E.R.S., Hagting, A., Richard, J., Konings, W.N., Desmazeaud, M., Poolman, B., 1995. Specificity of peptide transport systems in Lactococcus lactis: evidence for a third system which transports hydrophobic di- and tripeptides. Journal of Bacteriology 177, 4652e4657. Foucaud, C., Hemme, D., Desmazeaud, M., 2001. Peptide utilization by Lactococcus lactis and Leuconostoc mesenteroides. Letters in Applied Microbiology 32, 20e25. Freiding, S., Gutsche, K.A., Ehrmann, M.A., Vogel, R.F., 2011. Genetic screening of Lactobacillus sakei and Lactobacillus curvatus strains for their peptidolytic system and amino acid metabolism, and comparison of their volatilomes in a model system. Systematic and Applied Microbiology 34, 311e320. Gao, S., Mooberry, E.S., Steele, J.L., 1998. Use of 13C nuclear magnetic resonance and gas chromatography to examine methionine catabolism by Lactococci. Applied and Environmental Microbiology 64, 4670e4675. Hagting, A., Kunji, E.R.S., Leenhouts, K.J., Poolman, B., Konings, W.N., 1994. The diand tripeptide transport protein of Lactococcus lactis: a new type of bacterial peptide transporter. The Journal of Biological Chemistry 269, 11391e11399. Hammes, W.P., Hertel, C., 1998. New developments in meat starter cultures. Meat Science 49, 125e138. Hammes, W.P., Bantleon, A., Min, S., 1990. Lactic acid bacteria in meat fermentation. FEMS Microbiology Reviews 87, 165e174.

Q. Sinz, W. Schwab / Food Microbiology 29 (2012) 215e223 Hugas, M., Garriga, M., Aymerich, T., Monfort, J.M., 1993. Biochemical characterization of lactobacilli from dry fermented sausages. International Journal of Food Microbiology 18, 107e113. Kunji, E.R.S., Ubbink, T., Martin, A., Poolman, B., Konings, W.N., 1993. Physiological responses of Lactococcus lactis ML3 to alternating conditions of growth and starvation. Archives of Microbiology 159, 372e379. Kunji, E.R.S., Mierau, I., Hagting, A., Poolman, B., Konings, W.N., 1996. The proteolytic systems of lactic acid bacteria. Antonie van Leeuwenhoek 70, 187e221. Larrouture, C., Ardaillon, V., Pépin, M., Montel, M.C., 2000. Ability of meat starter cultures to catabolize leucine and evaluation of the degradation products by using an HPLC method. Food Microbiology 17, 563e570. Liu, M., Nauta, A., Francke, C., Siezen, R.J., 2008. Comparative genomics of enzymes in flavor-forming pathways from amino acids in lactic acid bacteria. Applied and Environmental Microbiology 74, 4590e4600. Liu, M., Bayjanov, J.R., Renckens, B., Nauta, A., Siezen, R.J., 2010. The proteolytic system of lactic acid bacteria revisited: a genomic comparison. BMC Genomics 11, 1e15. De Man, J.C., Rogosa, M., Sharpe, M.E., 1960. A medium for the cultivation of lactobacilli. The Journal of Applied Bacteriology 23, 130e135. Mandelstam, J., 1958. Turnover of protein in growing and non-growing populations of Escherichia coli. The Biochemical Journal 69, 110e119. Marco, A., Navarro, J.L., Flores, M., 2008. The sensory quality of dry fermented sausages as affected by fermentation stage and curing agents. European Food Research and Technology 226, 449e458. Masson, F., Hinrichsen, L., Talon, R., Montel, M.-C., 1999. Factors influencing leucine catabolism by a strain of Staphylococcus carnosus. International Journal of Food Microbiology 49, 173e178. Meynier, A., Novelli, E., Chizzolini, R., Zanardi, E., Gandemer, G., 1999. Volatile compounds of commercial Milano salami. Meat Science 51, 175e183. Montel, M.-C., Seronie, M.-P., Talon, R., Hébraud, M., 1995. Purification and characterization of a dipeptidase from Lactobacillus sake. Applied and Environmental Microbiology 61, 837e839. Montel, M.-C., Reitz, J., Talon, R., Berdagué, J.-L., Rousset-Akrim, S., 1996. Biochemical activities of Micrococcaceae and their effects on the aromatic profiles and odours of a dry sausage model. Food Microbiology 13, 489e499. Montel, M.-C., Masson, F., Talon, R., 1998. Bacterial role in flavour development. Meat Science 49, 111e123. Ordóñez, J.A., Hierro, E.M., Bruna, J.M., de la Hoz, L., 1999. Changes in the components of dry-fermented sausages during ripening. Critical Reviews in Food Science and Nutrition 39, 329e367. Partidário, A.M., Padilha, M., Roseiro, C., Silva, L., Santos, C., 2006. Volatile compounds produced during ripening of Paínho de Portalegre dry fermented sausage. Revista Portuguesa de Ciencias Veterinárias 101, 115e120. Rosenstein, R., Nerz, C., Biswas, L., Resch, A., Faddatz, G., Schuster, S.C., Götz, F., 2009. Genome analysis of the meat starter culture bacterium Staphylococcus carnosus TM300. Applied and Environmental Microbiology 75, 811e822. Ross, R.P., Morgan, S., Hill, C., 2002. Preservation and fermentation: past, present and future. International Journal of Food Microbiology 79, 3e16. Söllner, K., Schieberle, P., 2009. Decoding the key aroma compounds of a Hungarian-type salami by molecular sensory science approaches. Journal of Agricultural and Food Chemistry 57, 4319e4327. Saguir, F.M., Loto Compos, I.E., Manca de Nadra, M.C., 2008. Utilization of amino acids and dipeptides by Lactobacillus plantarum from orange in nutritionally stressed conditions. Journal of Applied Microbiology 104, 1597e1604. Sanz, Y., Toldrá, F., 1997. Purification and characterization of an aminopeptidase from Lactobacillus sake. Journal of Agricultural and Food Chemistry 45, 1552e1558. Sanz, Y., Toldrá, F., 2001. Purification and characterization of an X-prolyl-dipeptidyl peptidase from Lactobacillus sakei. Applied and Environmental Microbiology 67, 1815e1820.

223

Sanz, Y., Toldrá, F., 2002. Purification and characterization of an arginine aminopeptidase from Lactobacillus sakei. Applied and Environmental Microbiology 68, 1980e1987. Sanz, Y., Mulholland, F., Toldrá, F., 1998. Purification and characterization of a tripeptidase from Lactobacillus sake. Journal of Agricultural and Food Chemistry 46, 349e353. Sanz, Y., Fadda, S., Vignolo, G., Aristoy, M.-C., Oliver, G., Toldrá, F., 1999. Hydrolysis of muscle myofibrillar proteins by Lactobacillus curvatus and Lactobacillus sake. International Journal of Food Microbiology 53, 115e125. Sanz, Y., Toldrá, F., Renault, P., Poolman, B., 2003. Specificity of the second binding protein of the peptide ABC-transporter (Dpp) of Lactococcus lactis IL1403. FEMS Microbiology Letters 227, 33e38. Savijoki, K., Ingmer, H., Varmanen, P., 2006. Proteolytic systems of lactic acid bacteria. Applied Microbiology and Biotechnology 71, 394e406. Smit, E.J., Konings, W.N., 1990. Relationship between utilization of proline and proline-containing peptides and growth of Lactococcus lactis. Journal of Bacteriology 172, 5286e5292. Smit, B.A., Engels, W.J.M., Wouters, J.T.M., Smit, G., 2004. Diversity of L-leucine catabolism in various microorganisms involved in dairy fermentations, and identification of the rate-controlling step in the formation of the potent flavour component 3-methylbutanal. Applied Microbiology and Biotechnology 64, 396e402. Smit, G., Smit, B.A., Engels, W.J.M., 2005. Flavour formation by lactic acid bacteria and biochemical flavour profiling of cheese products. FEMS Microbiology Reviews 29, 591e610. Smit, B.A., Engels, W.J.M., Smit, G., 2009. Branched chain aldehydes: production and breakdown pathways and relevance for flavour in foods. Applied Microbiology and Biotechnology 81, 987e999. Stahnke, L.H., 1999. Volatiles produced by Staphylococcus xylosus and Staphylococcus carnosus during growth in sausage minces part II. The influence of growth parameters. Lebensmittel-Wissenschaft und -Technologie 32, 365e371. Stolz, P., Böcker, G., Hammes, W.P., Vogel, R.F., 1995. Utilization of electron acceptors by lactobacilli isolated from sourdough: I. Lactobacillus sanfrancisco. Zeitschrift für Lebensmittel-Untersuchung und -Forschung 201, 91e96. Stuart, M.R., Chou, L.S., Weimer, B.C., 1999. Influence of carbohydrate starvation and arginine on culturability and amino acid utilization of Lactococcus lactis subsp. lactis. Applied and Environmental Microbiology 65, 665e673. Tavaria, F.K., Dahl, S., Carballo, F.J., Malcata, F.X., 2002. Amino acid catabolism and generation of volatiles by lactic acid bacteria. Journal of Dairy Science 85, 2462e2470. Toldrá, F., Rico, E., Flores, J., 1992. Activities of pork muscle proteases in model cured meat systems. Biochimie 74, 291e296. Trüpler, H.G., De’ Clari, L., 1997. Taxonomic note: necessary correction of specific epithets formed as substantives (nouns) “in apposition”. International Journal of Systematic Bacteriology 47, 908e909. van Kranenburg, R., Kleerebezem, M., van Hylckama Vlieg, J., Ursing, B.M., Boekhorst, J., Smit, B.A., Ayad, E.H.E., Smit, G., Siezen, R.J., 2002. Flavour formation from amino acids by lactic acid bacteria: predictions from genome sequence analysis. International Dairy Journal 12, 111e121. Vermeulen, N., Pavlovic, M., Ehrmann, M.A., Gänzle, M.G., Vogel, R.F., 2005. Functional characterization of the proteolytic system of Lactobacillus sanfranciscensis DSM 20451T during growth in sourdough. Applied and Environmental Microbiology 71, 6260e6266. Vermeulen, N., Gänzle, M., Vogel, R.F., 2006. Influence of peptide supply and cosubstrates on phenylalanine metabolism of Lactobacillus sanfranciscensis DSM20451T and Lactobacillus plantarum TMW1.468. Journal of Agricultural and Food Chemistry 54, 3832e3839. Weimer, B., Seefeldt, K., Dias, B., 1999. Sulfur metablolism in bacteria associated with cheese. Antonie van Leeuwenhoek 76, 247e261. Yvon, M., Rijnen, L., 2001. Cheese flavour formation by amino acid catabolism. International Dairy Journal 11, 185e201.