Journal of Dairy Science
Volume 89, Issue 10 , Pages 3800-3809, October 2006

Extraction and Partial Characterization of Proteolytic Activities from the Cell Surface of Lactobacillus helveticus Zuc2

Istituto Microbiologia, Università Cattolica del Sacro Cuore (UCSC), via Emilia parmense, 84, 29100 Piacenza, Italy

Received 30 September 2005; accepted 17 April 2006.

Article Outline

Abstract 

Proteolytic activities were extracted from a dairy Lactobacillus helveticus strain and partially characterized. A first cell envelope proteinase (CEP) was extracted using a high ionic strength buffer, both in the presence and in the absence of Ca2+. Moreover, cell treatment by 5 M LiCl allowed for the selective removal of the S-layer protein and CEP, suggesting an enzyme ionic linkage to the cell envelope similar to that observed for the Slayer structure. The enzyme specificity against αs1-CN (f1–23) showed unusual activity on the Lys3-His4 bond compared with other proteinases of the same species. A second proteinase appeared to be linked to the cell membrane because it was extractable only after membrane disgregation by detergents. Its specificity against CN fractions and αs1-CN (f1–23) was different from that of the first CEP; moreover, the measured activity was lower than that of CEP.

Key words: Lactobacillus helveticus, cell envelope, proteinase, specificity

 

Back to Article Outline

Introduction 

Lactobacillus helveticus represents the prevailing species among the thermophilic bacteria of natural whey starter cultures, and it plays an essential role in the ripening of different cheeses, such as Italian hard cheese and Swiss-type cheese. Morishita et al. (1981) demonstrated that L. helveticus strains are auxotrophic for most of the AA. Therefore, because of the low free AA content, its growth in milk depends on an extensive CN breakdown by proteolytic enzymes whose activity is considered the strongest among the lactic acid bacteria (Kunji et al., 1996). For this peculiar characteristic, its use has been suggested as a starter adjunct to Cheddar cheese milk because an attenuated culture enhances the flavor development (Makdor et al., 2000). Casein breakdown results from the action of extracellular proteinase(s), and the hydrolysis products (from 4 to 18 AA) are translocated into the bacterial cell via an oligopetide transport system. Various cytoplasmatic peptidases, with partially overlapping specificities, then degrade the internalized oligopeptides into free AA (Kunji et al., 1996).

Several reports are available on the cell envelope proteinase (CEP) of L. helveticus, which display a strong heterogeneity among such enzymes (Martin-Hernandez et al., 1994; Ono et al., 1997; Hébert et al., 1999).

In some cases, an appreciable release of CEP from L. helveticus strains can be achieved by use of a calcium-free buffer, such as that obtained for lactococcal proteinase (Ono et al., 1997). On the contrary, CEP of the L. helveticus L89 and CRL 1062 strains showed a strong membrane-bound character and was poorly released in a calcium-free extraction buffer (Martin-Hernandez et al., 1994; Hébert et al., 1999). Lactobacillus helveticus CEP show a wide heterogeneity in the relative molecular mass (Mr) of purified enzymes (ranging from 45 to 180 kDa), immunological properties (Yamamoto et al., 1998), and CN breakdown specificity (Ono et al., 1997). Moreover, the proteolytic system of this species is considered to be useful for the production of bioactive peptides; in fact, several bioactive peptides are released from CN either during milk fermentation by strains of this genus (Matar et al., 2001; Scolari et al., 2002; Seppo et al., 2003) or by their isolated proteinases (Maeno et al., 1996).

Pederson et al. (1999) demonstrated the presence of 2 different CEP in a PrtH deletion derivative of the CNRZ 32 strain, obtained by gene replacement. Moreover, a gene (htrA) coding for a stress-inducible Htr-like protein from L. helveticus CNRZ 32 has been cloned, sequenced, and characterized (Smeds et al., 1998). This protein carries the characteristic catalytic domain of trypsin-like proteases, and the authors proposed the location of HtrA of L. helveticus to be on the outer surface of the plasma membrane, on the basis of the AA sequence.

This paper describes the extraction of 2 proteinases from a strongly proteolytic L. helveticus strain and the partial characterization of the most loosely CEP-bound enzyme.

Back to Article Outline

Materials and Methods 

Chemicals and Chromatographic Materials 

Inhibitors, lysozyme, Igepal Ca-360, and alkaline peptidase, as well as the α-CN, β-CN, and κ-CN fractions, were obtained from Sigma (Sigma Chemical Co. St. Louis, MO). The SDS molecular weight standard was from Bio-Rad Laboratories (Hercules, CA). Succi-nyl-Ala-Ala-Phe-7-amido-4-methylcoumarin (AMC), benzoyl-Pro-Phe-Arg-AMC, benzoyl-Phe-Val-Arg-AMC, and benzoyl-Val-Gly-Arg-AMC were obtained from Bachem (Bubendorf, Switzerland).

Strains and Culture Conditions 

The L. helveticus Zuc2 used in this study was previously isolated from natural whey culture for Grana Padano cheese and identified by biochemical tests, randomly amplified polymorphic DNA (RAPD)-PCR analysis, and the electrophoretic mobility of X-prolyl-dipeptidyl aminopeptidase (Scolari and Vescovo, 2004). It was maintained in filter-sterilized cheese whey under liquid nitrogen vapor at the collection of the UCSC Microbiology Institute (Piacenza, Italy) and, when required, was subcultured overnight in the same medium at 44°C.

Proteinase Extraction and Purification 

Lactobacillus helveticus Zuc2 cells from an overnight culture (250mL) were harvested by centrifugation at 3,000×g, 4°C, for 10min, washed twice in 50mM β-glycerolphosphate buffer, pH 7.0, and 15mM CaCl2 and resuspended to 1:40 of their original volume in each of the extraction buffers listed in Table 1. Bacterial cells were incubated at 37°C for 30min under mild agitation to extract CEP, and were successively removed from suspension by centrifugation at 12,000×g, 4°C, for 10min. Supernatants were assayed for cell lysis by measuring the intracellular lactate dehydrogenase (EC 1.1.1.127) and X-prolyl-dipeptidyl aminopeptidase (EC 3.4.14.5) activities according to Scolari and Vescovo (1996) before storage at −80°C.

Table 1. Buffers used in the cell envelope proteinase extraction trials1
ExtractantpHConcentration (M)Addition
Pi7.00.05, 0.1, 0.25, 0.75, 0.50, 1.00—
Pi7.00.5PMSF (1mM)
Pi7.00.5EDTA (20mM)
MES7.00.5—
MES7.00.5CaCl2 (20mM)
LiCl—5—

1Pi = phosphate buffer; PMSF = phenyl-methyl sulfonyl fluoride; MES = morpholine-ethane sulfonic acid.

The proteinase extracted by 0.5 M phosphate (Pi) buffer, pH 7.0, was loaded (750μL) onto a gel filtration Superdex 200 HR 10/30 column (Pharmacia Biotech, Uppsala, Sweden) previously calibrated by a gel filtration high molecular weight calibration kit (Pharmacia Biotech), and samples were eluted using 50mM Pi buffer, pH 7.0, and 150mM NaCl at a flow rate of 0.4 mL/min. Eluted fractions (5mL) were monitored by a UV detector (280nm) and assayed for proteolytic activity; the active fractions were concentrated sixfold by diafiltration on 10-kDa molecular weight cut-off centricon concentrators (Millipore, Bedford, MA) against the equilibration buffer of the next chromatography. The second purification step was performed on a MonoQ ion-exchange column (Pharmacia Biotech) equilibrated with 20mM Pi buffer, pH 7.5, at a flow rate of 1 mL/min. The enzyme was eluted by a 30-min linear gradient of 0 to 0.5 M NaCl in the equilibration buffer, and the active fractions (containing 3.5μg/mL of the purified enzyme) were stored at −80°C until use (CEP fractions). Moreover, after cell fractionation, the proteolytic activity extracted from the membrane fractions was subjected to the same chromatographic steps as CEP purification, but using a linear gradient of 0 to 3 M NaCl for MonoQ chromatography. All the above steps were performed at 4°C.

Bacterial Cell Fractionation 

Washed cells from an overnight culture were obtained as described in the previous section. To obtain protoplasts and the cell wall fraction, pellets were suspended to 1:25 of their original volume in the digestion buffer and processed as described by Scolari and Vescovo (1996). Protoplasts were separated from the supernatant (cell wall fraction) by centrifugation at 1,300×g, 4°C, for 10min. To remove the high concentration of protoplasts stabilizing sucrose (17%), the cell wall fraction was subjected to dialysis for 24h at 4°C. Pellets containing protoplasts were washed with the digestion buffer and resuspended at 1:25 of the original culture volume in digestion buffer deprived of lysozyme and saccharose. After mechanical disruption of the protoplasts by mild sonication in ice-cold 50mM Pi buffer, pH 7.0, unbroken particles were selectively settled by centrifugation at 1,000×g, 4°C, for 10min; centrifugation of the supernatant at 16,000×g, 4°C, for 10min produced membrane (pellet) and cytoplasmic (supernatant) fractions.

The washed membranes were suspended to the volume of the cytoplasmic fraction using the same buffer with 1% (vol/vol) Igepal Ca-360 added, and were almost completely dissolved by stirring for 2h at 4°C. The detergent was removed from the solubilized membranes by exhaustive diafiltration on 10-kDa molecular weight cut-off centriprep concentrators (Amicon Inc., Beverly MA) using 20mM Pi buffer, pH 7.5. Cell fractions were assayed for proteolytic activity.

Proteinase Characterization 

Cell envelope proteinase fractions were dialyzed against each assay buffer before determination of the enzyme temperature optimum (measured at pH 7.0) and pH optimum (measured at 37°C). Proteolytic activity was measured at 30, 37, 42, and 55°C as indicated in the following sections; measures at the desired pH values (5.0 to 7.0) were performed using the same substrate in 50mM citrate-Pi buffer.

The effect of inhibitors was evaluated by measuring the activity of the same fractions against caseinate at the conditions described above in the presence or absence of each protease inhibitor: 1mM phenyl-methyl sulfonyl fluoride (PMSF), 100μM 3,4-dichloroisocoumarin, 1mM pepstatin, and 10μM E-64 as a thiol-protease inhibitor.

Proteinase Specificity 

Equal volumes of individual CN suspensions (20 mg/mL in 50mM Pi buffer, pH 7.0) and CEP fractions (2mL), previously diafiltered in the same buffer at one-third of their original volume, were used to determine α-CN, β-CN, and κ-CN breakdown by the purified enzyme. Samples for SDS-PAGE analysis were taken after 0, 10, 60, and 120min of mixture incubation at 37°C.

To obtain the αs1-CN (f1–23) peptide, αs1-CN B was purified and the pepsin was hydrolyzed as described by Kaminogawa et al. (1986). The αs1-CN (f1–23) fragment was purified from the hydrolysis mixture by means of semipreparative reversed-phase (RP)-HPLC on a C18 Radial-Pack cartridge (Waters Associates, Milford, MA) in a trifluoroacetic acid–CH3CN solvent system (solvent A: 0.115% trifluoroacetic acid in water; solvent B: 0.1% trifluoroacetic acid–60% CH3CN in water). Samples were eluted using a 60-min linear gradient from 0 to 100% solvent B and recorded at 214nm.

The κ-CN (f106–169) glycomacropeptide fragment (GMP) was separated from the cheese whey by semi-preparative RP-HPLC under the same conditions as used for αs1-CN (f1–23) purification and monitored for the presence of sialic acid by the thiobarbituric acid reaction (Sick et al., 1990). Solvent was removed from the GMP-containing fractions by evaporation, and AA analysis was performed according to Scolari et al. (1996) following acid hydrolysis for 24h at 110°C under vacuum in 6 M HCl containing 0.1% (wt/vol) phenol. The peptide was identified by comparing the AA composition with that calculated from the GMP sequence.

Purified αs1-CN (f1–23) and GMP peptides were incubated (final concentration of each: 3μg/mL) with CEP and the membrane proteinase preparations at 37°C, tracking samples after 10 and 30min.

The specificity of CEP against the αs1-CN (f1–23) and GMP fragments was analyzed by RP-HPLC separation of the degradation products according to Exterkate et al. (1991) on an ACCQ·Tag RP-C18 column, 3.9×150mm (Waters Associates). Peaks were collected, dried, and identified by mass spectrometry performed on a laser desorption time-of-flight system (model G2025A; Hewlett-Packard, Palo Alto, CA). Peptides from GMP hydrolysates were identified by N-terminal sequencing and determination of the AA profile.

Proteolytic Activity Measurements and SDS-PAGE Electrophoresis 

Preparations containing proteolytic activity (3.5μg/mL for pure CEP and 0.8μg/mL for protease from membranes) were mixed with the same volume of 1.5% (wt/vol) Na caseinate in 50mM Pi buffer, pH 7.0, and 0.1% (wt/vol) NaN3, dialyzed against the same buffer, and were incubated overnight at 37°C; the activity was measured as the increase in optical density at 340nm (OD340) after o-phthaldialdehyde derivatization according to Scolari et al. (1996).

To detect the presence of endopeptidase activity in the purified fraction of CEP, the dialyzed preparations were separated on native PAGE as described by Scolari and Vescovo (2004); gels were stained after incubation by 0.1mM succinyl-Ala-Ala-Phe-AMC, benzoyl-Pro-Phe-Arg-AMC, benzoyl-Phe-Val-Arg-AMC, and benzoyl-Val-Gly-Arg-AMC in the presence of 0.8 U/mL of alkaline peptidase (EC 3.4.11.10), under the conditions reported by the same authors. Substrates were chosen following the indications of Christensen et al. (2003).

Sodium dodecyl sulfate-PAGE electrophoresis was carried out as described by Scolari et al. (1996) using a Mini-Protean electrophoretic cell (Bio-Rad Laboratories). Hydrolyzed CN and enzymatic activity were separated on 12 and 8% running gels, respectively, and enzyme Mr was determined by a low-range SDS molecular weight standard after Coomassie Brilliant Blue R-250 staining; GMP visualization was possible using 5% (wt/vol) TCA in 10% acetic acid as the destaining solution.

LiCl Treatments 

The LiCl extract was exhaustively diafiltered by 30-kDa molecular weight cut-off centricon against Milli-Q water (Waters Millipore, Milan, Italy) until a white precipitate was formed, which was recovered by centrifugation at 12,000×g for 10min and dissolved to the original extract volume in 50mM Pi, pH 7.0. Both the supernatant and solubilized precipitate were assayed for proteolytic activity, and the specificity of the latter was determined against αs1-CN (f1–23) as described above.

In addition, the proteolytic activity of either intact or LiCl-extracted cells was evaluated. For this purpose, L. helveticus Zuc2 cells, obtained as reported above, were treated twice with 3 M LiCl by mild agitation at 4°C for 20min. Both types of cells were washed and suspended in 50mM β-glycerolphosphate buffer, pH 7.0, added to 7μg/mL of chloramphenicol as a protein synthesis inhibitor and incubated with αs1-CN (f1–23) at 4°C for 30min. The hydrolysis products were separated as described above, and their identification was based on previous results; unknown peaks [from αs1-CN (f1–23)] were identified by AA analysis as reported above.

Back to Article Outline

Results 

CEP Extraction 

Lactobacillus helveticus Zuc2 was characterized by strong proteolytic activity toward CN; however, cell incubation in Ca-free buffers was ineffective in releasing such activity from the cell envelope. The recovery of CEP was enhanced by increasing the ionic strength of the extraction buffer. In particular, when using Pi extraction buffers ranging from 0.05 to 1.00 M (Table 1), a logarithmic relationship was evident between the release of CEP, expressed as OD340, and buffer concentrations up to 0.5 M. Moreover, when using a buffer of still higher ionic strength, less than 40% of the enzymatic activity was retained on the treated cells (Figure 1). These results suggest the use of a 0.5 M Pi buffer for the CEP extraction and purification steps.

  • View full-size image.
  • Figure 1. 

    Percentage of the total proteolytic activity extracted from Lactobacillus helveticus Zuc2 cells by phosphate buffer (Pi) at increasing ionic strength. The activity is expressed as OD340.

In all the extraction experiments, cell lysis was negligible because of the absence of intracellular enzyme activities (EC 1.1.1.127 and EC 1.1.1.127) in culture supernatants. To exclude a Ca2+-triggered autoproteolytic event in the CEP release mechanism, L. helveticus Zuc2 cells were incubated with Pi buffer with 20mM EDTA added, and the same extraction pattern was obtained irrespective of the presence of the additive (Figure 2, lanes 4 and 6). Comparable results were obtained using MES extraction buffer with addition of 20mM CaCl2. Also, an autoproteolytic release of CEP because of structural changes induced by the ionic strength of the buffer was excluded by adding 1mM PMSF (Laan and Konings, 1989) to the 0.5 M Pi extraction buffer. In fact, the same SDS-PAGE profile was obtained in the presence and in the absence of the inhibitor (Figure 2, lanes 4 and 5). Interestingly, this profile showed a band of approximately 45 kDa, with an electrophoretic mobility corresponding to the S-layer monomer, which was also evident in the 5 M LiCl extract (Mozes and Lortal, 1995; Figure 2, lane 3). It is noteworthy that the complete removal of salt from 5 M LiCl extracts by dialysis produced a white precipitate, which retained the original proteolytic activity in its entirety. Sodium dodecyl sulfate-PAGE analysis of the precipitate revealed the presence of 2 main bands corresponding to the S-layer monomer (B) and to the 180-kDa proteinase (A), respectively (Figure 2, lane 2). This could suggest an ability of CEP to associate with the S-layer in L. helveticus, as also hypothesized by Pederson et al. (1999) for the L. helveticus CNRZ 32 strain.

  • View full-size image.
  • Figure 2. 

    Sodium dodecyl sulfate-PAGE separation of various Lactobacillus helveticus Zuc2 cell envelope proteinase (CEP) extracts. Lane 1: purified CEP. Lanes 2 to 6: various extracts: white precipitate from dialysis (2); 5 M LiCl (3); 0.5 M phosphate buffer (Pi) (4); 0.5 M Pi + 1mM PMSF (5); 0.5 M Pi + 20mM EDTA (6). Lane 7: SDS molecular mass marker. A = 180-kDa proteinase; B = S-layer monomer.

Moreover, after complete removal of the S-layer by 3 M LiCl, a small amount (approximately 3 to 5%) of the total extractable proteolytic activity was still removable from the cell surface by thorough washing with 0.5 M Pi. The αs1-CN f(1–23) fragmentation profile of this last extract was the same as that of CEP (data not shown).

Purification and Characterization of CEP 

The purification scheme is shown in Table 2. The high values for specific activity and yield seem to indicate that 0.5 M Pi is a rather selective extractant for CEP, thus allowing for a relatively simple purification process. In fact, the same purification steps permitted only a partial purification of membrane-associated proteinase (MAP) from the membranes, giving a very low yield as well as very low specific activity.

Table 2. Summary of the purification of cell envelope proteinase (CEP) and membrane-associated proteinase (MAP) from Lactobacillus helveticus Zuc2
StepTotal activity,1 ΔOD340Total protein, mgSpecific activity, ΔOD340/mgPurification, foldYield, %
——— (CEP) ———
Crude extract3.091661.86×10−21.00100
Gel filtration chromatography2.472410.29×10−25.5379
Ion-exchange chomatography1.030.1059.81527.4233
——— (MAP) ———
Solubilized membranes3.251,8451.76×10−31.00100
Gel filtration chromatography2.086473.21×10−31.8235
Ion-exchange chomatography0.45371.22×10−26.932

1Activity is expressed as variation in OD340 (optical density at 340nm) after 12h of incubation.

The S-layer associated with CEP, extracted with 0.5 M Pi buffer, eluted from gel filtration chromatography between thyroglobulin (669 kDa) and ferritin (440 kDa), indicating an approximate native Mr of about 560 kDa. Fractions containing proteolytic activity were eluted by subsequent ion-exchange chromatography at an ionic strength of 0.2 M NaCl. On SDS-PAGE they showed a prevailing intense band of 180 kDa Mr (Figure 2, lane 1) that was associated with CEP, thus suggesting a trimeric aggregation of the extracted enzyme, possibly through the W domain of the molecule. The incubation of native gels of purified CEP fractions with the 4 endopeptidase substrates did not reveal the presence of red bands indicating the presence of such an intracellular enzyme class.

The optimal pH of CEP at 37°C is approximately 6, according to values from the literature (Ono et al., 1997). The optimal temperature at pH 7.0 was 44°C, and the enzyme maintained 40% of the activity when incubated for 2h at 60°C. Inhibition of the enzyme was complete and irreversible with either 1mM PSMF or 100μM 3,4-dichloroisocoumarin; 1mM pepstatin gave an inhibition of approximately 10%, but the enzyme activity was almost completely restored after inhibitor removal. No inhibition occurred with the thiol–proteinase inhibitor E-64, thus excluding the presence of a cysteine group in the active site.

CEP Specificity 

Hydrolysis of the CN Fractions 

The specificity of CEP on αs1-CN, β-CN, and κ-CN is reported in Figure 3. The enzyme showed intense activity toward β-CN even from the first 10min of incubation, producing a substantial reduction in the corresponding band. Within this time, a broad band appeared in the hydrolytic profile of κ-CN that had the characteristic shape and electrophoretic mobility of the purified GMP; it was also possible to observe the formation of a fragment with electrophoretic mobility corresponding to the αs1I peptide in the profile of αs1-CN hydrolysates, thus indicating the specificity of the enzyme for the Arg22-Phe23 bond of αs1-CN as well as for the Phe105-Met106 bond of κ-CN. Both fragments were subsequently partially degraded. At the end of incubation, all the fractions were thoroughly hydrolyzed, and the β-CN was completely digested.

  • View full-size image.
  • Figure 3. 

    Digestion of αs1-CN (lanes 1 to 4), β-CN (lanes 5 to 8), and κ-CN (lanes 9 to 12) by Lactobacillus helveticus Zuc2 cell envelope proteinase at different incubation times: 0min, lanes 1, 5, 9; 10min, lanes 2, 6, 10; 60min, lanes 3, 7, 11; 120min, lanes 4, 8, 12. αs1I = αs1-CN (f24–199); GMP = glycomacropeptide.

Specificity Against αs1-CN (f1-23) and GMP 

A deeper knowledge of the specificity of CEP was obtained by analyzing the hydrolysis products of both αs1-CN (f1–23) and GMP. Reversed-phase HPLC digestion profiles of αs1-CN (f1–23) after 10 and 30min of incubation, respectively (Figures 4A and 4B), showed a complex mixture of hydrolysis products as well as almost complete substrate degradation within the first 10min of incubation, confirming the strong activity demonstrated by the enzyme toward the CN fractions. This digestion profile was also obtained by using the white precipitate from the LiCl extraction, but in a shorter incubation time (data not shown).

  • View full-size image.
  • Figure 4. 

    Reversed-phase HPLC patterns of αs1-CN (f1–23) hydrolysis products after 10min (A) and 30min (B) of incubation at 37°C with purified Lactobacillus helveticus Zuc2 cell envelope proteinase.

The accumulation of αs1-CN (f1–9), (f10–16), and (f17–23) fragments in the first 10min of incubation could suggest Gln9-Gly10 and Leu16-Asn17 bonds as being the primary cleavage sites. Accordingly, the peptide αs1-CN (f1–16) released from splitting of the latter bond could be readily converted to the αs1-CN (f1–9) and (f10–16) fragments, because the amount of αs1-CN (f1–16) in the RP-HPLC profile was consistently low at both incubation times. Similarly, the fragment αs1-CN (f10–23), containing the sensitive Leu16-Asn17 bond, could be degraded to the peptides αs1-CN (f10–16) and (f17–23), because it was not detectable after 10min of incubation.

Moreover, it is possible to speculate that within the first 10min, a small amount of the substrate was hydrolyzed at the Asn17-Glu18 bond, releasing αs1-CN (f18–23) and (f10–17), whereas the appearance of the αs1-CN (f1–6) and (f7–13) peptides could suggest a similar sensitivity of the Ile6-Lys7 and Gln13-Glu14 bonds to CEP activity. One of the possible degradation products of these peptide bond cleavages, the tetra peptide Glu14-Val15-Leu16-Asn17, was not detectable, probably because it did not bind to the column (Oberg et al., 2002).

Among the probable secondary hydrolytic reactions (after 30min), the conversion of αs1-CN (f1–9) at the sites Lys3-His4, Ile6-Lys7, and His8-Gln9 could explain the accumulation of the αs1-CN (f 4–9), (f1–8), and (f1–6) fragments, whereas αs1-CN (f10–17) and (f10–16) degradation (at the Gln13-Glu14 bond) produced mainly αs1-CN (f10–13).

The accumulation of αs1-CN (f17–21) might originate from cleavage of the Leu21-Arg22 bond of αs1-CN (f17–23). The charged αs1-CN (f22–23) and (f1–3) fragments probably eluted early in the chromatogram and were thus not detectable.

Amino acid analysis of the purified GMP showed more than 89% identity with the calculated composition, except for Thr, Leu, and Gly (70, 62, and 79%, respectively). Moreover, the contents of Tyr, Phe, His, and Arg were negligible, allowing GMP to be considered pure.

The incubation of GMP with CEP up to 30min did not produce a spectrum of hydrolyzed peptides as wide as that obtained from αs1-CN (f1––23) hydrolysis, because approximately 70% of the substrate (GMP and glycosylated GMP fractions) remained undigested (Figure 5). The main peptide fragments detected in the reaction mixture were κ-CN (f106–111), (f106–113), (f112–116), and (f117–146) as well as the carboxy-terminus (f112–149) and (f147–169) (Figure 5), thus suggesting that Lys111-Lys112, Lys116-Thr117, and Leu146-Glu147 were the sole sensitive bonds. A similar degradation profile was obtained by Sick et al. (1990) during GMP digestion by tripsin.

  • View full-size image.
  • Figure 5. 

    Reversed-phase HPLC patterns of GMP hydrolysis products after 30min of incubation at 37°C with purified Lactobacillus helveticus Zuc2 cell envelope proteinase. GMP = undigested glycomacropeptide (not glycosylated fraction); glycosylated GMP = undigested glycomacropeptide (glycosylated fractions).

Cell Fractionation and Partial Characterization of Membrane-Fraction Proteinase 

Cell lysis during the protoplastization step, expressed as the percentage of intracellular lactate dehydrogenase released, was less than 5% of total cells. Cell fractionation demonstrated that proteolytic activity was partitioned between the membrane (46%) and the cytoplasmatic (17%) fractions, respectively, whereas the remaining 37% was associated with the cell wall fraction. This last percentage appeared to be in contrast to that obtained for CEP proteinase activity extracted by 0.5 M Pi. A possible explanation could be the environmental differences in residual activity measured on bacterial cells after Pi buffer treatment and those detected on free enzymes in the membrane fractions. Moreover, enzyme autodigestion is highly probable because of the long dialysis process of the cell wall fraction.

More than 80% of the activity in the membrane fraction was not extractable by 0.5 M Pi but only after membrane solubilization by the treatment with Igepal CA-360, thus suggesting an enzyme anchorage to the membrane. The enzyme contained in the solubilized and dialyzed membrane fraction was eluted from the anion-exchange MonoQ column at 3.0 M NaCl of the gradient; this could suggest that the MAP is distinct from the CEP observed in the 0.5 M Pi extraction buffer. Approximate Mr of MAP was about 215 kDa, as estimated by gel filtration chromatography. The 2 chromatographic steps mentioned permitted only partial purification of MAP from the membrane fraction, so in this preliminary study, detection of the enzyme Mr on SDS-PAGE was not possible.

Casein degradation by the partially purified MAP was much lower than that observed for the purified CEP, showing no evident specificity against the 3 CN fractions. Also, unlike CEP, the degradation of αs1-CN and κ-CN by MAP did not produce the respective αs1-I and GMP fragments (data not shown).

LiCl-Treated Cells 

The RP-HPLC profile of αs1-CN (f1-23) after 30min of incubation with LiCl-treated whole cells is presented in Figure 6. The hydrolytic pattern is completely different from that produced by untreated cells, which substantially reproduces the specificity profile of CEP (data not shown). Only the His8-Gln9 and Gln9-Gly10 bonds were cleaved at the assay conditions, resulting in a lower proteolytic activity of treated cells compared with untreated ones. The fragment αs1-CN (f1–8) was further hydrolyzed at the Ile6-Lys7 bond.

Back to Article Outline

Discussion 

Lactobacillus helveticus Zuc2 seems to carry 2 different proteinases, showing 2 distinct cell envelope anchors. The presence of 2 cell surface proteinases in L. helveticus strains was first suggested by Pederson et al. (1999).

CEP 

The inability of Ca-free buffers to remove CEP was also reported by Martin-Hernandez et al. (1994) for the L. helveticus L89 strain. A substantial release of CEP from L. helveticus Zuc2 was achieved only by extracting buffer at an ionic strength higher than 250mM; such a value is reported as the limit beyond which electrostatic interactions are maximally weakened (Dill, 1990). The use of a specific inhibitor during the release process allowed us to exclude an autoproteolytic event promoted by conformational changes, possibly consequent to increasing the ionic strength of the extractant. Moreover, the stabilizing effect of the Ca2+ ion reported for the lactococcal CEP anchor (Laan and Konings, 1989) was excluded for the L. helveticus Zuc2 CEP, which is extractable in the presence of high CaCl2 concentrations. The same results were also observed for CEP release from L. acidophilus BGRA 43 and L. delbrueckii ssp. bulgaricus BGPF1 (Fira et al., 2001). These considerations and the low recovery of the proteinase in the cell wall fraction suggest the absence of the LPXTG sorting motif, which leads to the covalent linkage of lactococcal cell-wall proteinase (PrtP) to the peptidoglycan (Navarre and Schneewind, 1999), whose lack was observed by Pederson et al. (1999) and Germond et al. (2003) in proteinase PrtH from L. helveticus CNRZ 32 and PrtB proteinase from L. delbrueckii ssp. bulgaricus NCDO 1489, respectively, causing the authors to hypothesize a cell wall attachment of the CEP by electrostatic forces.

The reassociation of proteinase with S-layer monomers during the exhaustive dialysis of the 5 M LiCl extract supports the hypothesis of an S-layer involvement in the anchorage of the enzyme to the cell surface. S-Layer proteins could be the adhesion site for the CEP, as previously reported for a high Mr amylase in Bacillus stearothermophilus DSM 2358 (Egelseer et al., 1995), subtilase from Bacillus and Clostridium spp. (Schmidt et al., 1995), and pullulanase in Thermoanaerobacterium thermosulfurigenes EM1 (Matuschek et al., 1994). However, this is in contrast with the observation that a small amount (5%) of CEP is released by thoroughly washing L. helveticus Zuc2 cells in 0.5 M Pi buffer also after the complete removal of the S-layer by 3 M LiCl. Alternatively, both CEP and S-layer proteins could be linked to the underlying peptidoglycan or to the secondary cell-wall polymers by electrostatic interactions (Sara and Sleytr, 2000) through homologous sequences. In fact, Pederson et al. (1999) demonstrated that the W domain of proteinase PrtH from L. helveticus CNRZ 32 is homologous (up to 33%) to the C-terminal domain of the S-layer protein family of Lactobacillus, thus suggesting its possible role in cell-wall binding by a mechanism similar to S-layer protein anchorage (Boot et al., 1995).

The pH and the temperature optima were in accordance with those reported for L. helveticus CEP (Martin-Hernandez et al., 1994; Ono et al., 1997). The retained activity at 60°C assumes a practical usefulness, because the L. helveticus Zuc2 strain belongs to the starter microbiota of Grana cheese, whose technology requires a cooking temperature of approximately 55°C.

The complete inhibition obtained by PMSF includes CEP in the serine proteinase class. Moreover, reversible inhibition of pepstatin was less than 10%, suggesting that the enzyme does not belong to the aspartyl protease family, although some enzymes of this type are not influenced by this inhibitor—this in spite of its ability to produce GMP and αs1-I peptides, which are more typically produced by the action of aspartyl proteases on Phe105-Met106 and Arg22-Phe23 bonds of κ- and αs1-CN, respectively. However, a high susceptibility of the former bond to the lactococcal serin-type proteinase was also reported by Kunji et al. (1996). The relative accumulation of such a peptide in the first step of κ-CN hydrolysis (Figure 3) appears to be consistent with the resistance of the isolated GMP to the CEP activity (Figure 5).

Cell envelope proteinase specificity resembles that of proteinase PrtP I, primarily degrading β-CN, and to a lesser extent κ-CN and αs1-CN, whereas the MAP specificity cannot be assigned to either the P I or the P III type (Kunji et al., 1996). The behavior of both the proteinases against αs1-CN was different from that of L. helveticus CRL 1062, which preferentially digests this protein (Hébert et al., 1999).

Cell envelope proteinase cleaves αs1-CN (f1–23) differently from other L. helveticus proteinases (Kunji et al., 1996; Ono et al., 1997; Pederson et al., 1999; Oberg et al., 2002). In fact, L. helveticus Zuc2 CEP hydrolyzes all the cleavage sites of all the proteinases of the same species, and it is also active on the Lys3-His4 bond in the peptide amino-terminus domain. Such a bond was not even split by the CEP of the strain L. helveticus L 89, isolated from Grana cheese whey (Martin-Hernandez et al., 1994), as was the strain used in this study.

The amino-terminus region of αs1-CN [αs1-CN (f1–9)] is hydrolyzed only by intracellular postproline PepO2 and PepO3 endopeptidases from L. helveticus CNRZ 32, at the Pro5-Ile6 bond, under cheese conditions (Sridhar et al., 2005); moreover, those authors demonstrated that the Lys3-His4 and Lys7-His8 bonds of the same peptide were sensitive to the action of endopeptidase PepE. This underlines the unique specificity of L. helveticus Zuc2 CEP.

The low affinity of CEP toward αs1-CN seems to be in contrast with the high activity observed on the αs1-CN (f1––23) substrate, probably resulting from a different accessibility between the free Arg1-Phe23 sequence and the whole protein. In fact, an altered cleavage in the Arg1-Phe23 sequence of ααs1-CN compared with the free αs1-CN (f1-23 peptide was observed previously (Martin-Hernandez et al., 1994).

MAP 

The resuspension of the cell membrane fraction in Ca2+ free buffer did not produce an appreciable release of MAP, thus demonstrating its resistance to autoproteolysis in situ. The use of a membrane destructuring detergent to release MAP suggests the strong association of the enzyme with the membrane structure, as previously hypothesized for L. helveticus L89 (Martin-Hernandez et al., 1994) and L. helveticus CRL 1062 (Hébert et al., 1999).

The hydrolyzing activity toward whole CN fractions as well the lack of sensitive bonds in the amino-terminus of the αs1-CN (f1–23) peptide clearly distinguished MAP from the endopeptidases previously identified in L. helveticus (Fenster et al., 1997; Sridhar et al., 2005).

The stronger activity of CEP compared with that of MAP observed in intact cells could be the result of either the observed lower enzyme activity or a more peripheral location of the former, but also a difference in the specific activity.

Work is in progress to confirm the presence of the second proteinase in the L. helveticus Zuc2 strain and to better characterize it in order to establish the role of the 2 different proteinases during the growth of such species in cheese whey. Also, the simple extraction method used in this research will be applied to selectively recover the CEP from different L. helveticus strains, with the aim of checking the correlation between the ability of proteinases to produce angiotensin-converting enzyme–inhibiting peptides (or antihypertensive peptides, or both) from CN and determine their substrate specificity.

Back to Article Outline

Conclusions 

The results of this study suggest the presence of 2 different proteinases on the cell surface of the L. helveticus Zuc2 strain, although further evidence should be acquired to better characterize the MAP. The different mechanisms of enzyme linkage to the cell surface we observed allow for selective extraction of the 2 proteinases.

In particular, because of the ability, demonstrated elsewhere, for L. helveticus whole cells to produce bioactive (antihypertensive) peptides after CN digestion, the possibility of using an easy, discriminative extraction method for the cell envelope–anchored enzyme (CEP) constitutes an important prerequisite for obtaining a deeper knowledge of the role of the enzyme involved in production of the bioactive (antihypertensive) peptides.

Back to Article Outline

Acknowledgment 

Zucchelli cheese factory, which belongs to the Grana Padano consortium, is acknowledged for kindly providing natural whey cultures used in this study.

Back to Article Outline

References 

  1. Boot HJ, Kolen CPHM, Pouwels PH. Identification, cloning and nucleotide sequence of a silent S-layer protein gene of Lactobacillus acidophilus ATCC 4356 which has extensive similarity with the S-layer gene of this species. J. Bacteriol. 1995;177:7222–7230
  2. Christensen JE, Broadbent JR, Steele JL. Hydrolysis of casein-derived peptides αs1-casein (f1–9) and β-casein (f193–209) by Lactobacillus helveticus peptidase deletion mutants indicates the presence of a previous undetected endopeptidase. Appl. Environ. Microbiol. 2003;69:1283–1286
  3. Dill KA. Dominant forces in protein folding. Biochemistry. 1990;29:7133–7155
  4. Egelseer E, Schocher I, Sara M, Sleytr UB. The Slayer from Bacillus stearothermophilus DSM 2358 funtions as an adhesion site for high molecular-weight amylase. J. Bacteriol. 1995;177:1444–1451
  5. Exterkate FA, Alting AC, Slangen CJ. Specificity of two genetically related cell-envelope proteinases of Lactococcus lactis subsp. cremoris towards αs1-casein (f1–23)-fragment. Biochem. J. 1991;273:135–139
  6. Fenster KM, Parkin KL, Steele JL. Characterization of thiol-dependent endopeptidase from Lactobacillus helveticus CNRZ 32. J. Bacteriol. 1997;179:2529–2533
  7. Fira D, Kojic M, Banina A, Spasojevic I, Strahinic I, Topisirovic L. Characterization of cell envelope-associated proteinases of thermophilic lactobacilli. J. Appl. Microbiol. 2001;90:123–130
  8. Germond JE, Delley M, Gilbert C, Atlan D. Determination of the domain of the Lactobacillus delbrueckii subsp. bulgaricus cell surface proteinase PrtB involved in attachment to the cell wall after heterologus expression of the prtB gene in Lactococcus lactis. Appl. Environ. Microbiol. 2003;69:3377–3384
  9. Hébert EM, Raya RR, de Giori GS. Characterization of cell-envelope proteinase from Lactobacillus helveticus. Biotechnol. Lett. 1999;21:831–834
  10. Kaminogawa S, Yan TR, Azuma N, Yamauchi K. Identification of low molecular weight peptides in Gouda-type cheese and evidence for the formation of these peptides from 23 N-terminal residues of αs1-casein by proteinases of Streptococcus cremoris H61. J. Food Sci. 1986;51:1253–1256
  11. Kunji ERS, Mierau I, Hagting A, Poolman B, Konings WN. The proteolytic system of lactic acid bacteria. Antoine van Leeuwenhoek. 1996;70:187–221
  12. Laan H, Konings WN. Mechanism of proteinase release from Lactococcus lactis subsp. cremoris Wg2. Appl. Environ. Microbiol. 1989;55:3101–3106
  13. Maeno M, Yamamoto N, Takano T. Identification of an antihypertensive peptide from casein hydrolisate produced by a proteinase from Lactobacillus helveticus CP 790. J. Dairy Sci. 1996;79:1316–1321
  14. Makdor SA, Tong PS, El Soda M. Ripening of Cheddar cheese with added attenuated adjunct cultures of lactobacilli. J. Dairy Sci. 2000;83:1684–1691
  15. Martin-Hernandez C, Alting A, Extertake F. Purification and characterization of the mature, membrane associated cell-envelope proteinase of Lactobacillus helveticus L89. Appl. Microbiol. Biotechnol. 1994;40:828–834
  16. Matar C, Valdez JC, Medina M, Rachid M, Perdigon G. Immunomodulating effects of milks fermented by Lactobacillus helveticus and its non-proteolytic variant. J. Dairy Res. 2001;68:601–609
  17. Matuschek M, Burchhardt G, Sahm K, Bahl H. Pullulanase of Thermoanaerobacterium thermohydrosulfurigenes EM1 (Clostridium thermosulfurogenes): Molecular analysis of the gene, composite structure of the enzyme, and a common model for its attachment to the cell surface. J. Bacteriol. 1994;176:3295–3302
  18. Morishita T, Deguchi Y, Yajima M, Sakurai T, Yura T. Multiple nutritional requirements of lactobacilli: Genetic lesions affecting amino acid biosynthetic pathways. J. Bacteriol. 1981;148:67–71
  19. Mozes N, Lortal S. X-ray photoelectron spectroscopy and biochemical analysis of the surface of Lactobacillus helveticus ATCC 12046. Microbiology. 1995;141:11–19
  20. Navarre WW, Schneewind O. Surface proteins of gram-positive bacteria and mechanisms of their targeting to the cell wall envelope. Microbiol. Mol. Biol. Rev. 1999;63:174–229
  21. Oberg CJ, Broadbent JR, Strickland M, Mc Mahon DJ. Diversity in specificity of the extracellular proteinases in Lactobacillus helveticus and Lactobacillus delbrueckii subsp. bulgaricus. Lett. Appl. Microbiol. 2002;34:455–460
  22. Ono H, Yamamoto N, Maeno M, Takano T, Momose H. Purification and characterization of a cell-wall associated proteinase of Lactobacillus helveticus CP53. Milchwissenschaft. 1997;52:373–377
  23. Pederson JA, Mileski GJ, Weimer BC, Steele JL. Genetic characterization of a cell envelope-associated proteinase from Lactobacillus helveticus CNRZ32. J. Bacteriol. 1999;181:4592–4597
  24. Sara M, Sleytr UB. Minireview—S-layer proteins. J. Bacteriol. 2000;182:859–868
  25. Scolari G, Bottazzi V, Battistotti B. Antihypertensive effect of peptides produced by Lactobacillus helveticus in fermented milk. Proc. Cheese: Technology, Ripening and Characterization. October 29, 2002, Parma, Italy. Reggio Emilia, Italy: Associazione Italiana Tecnici del Latte; 2002;Pages 47–53
  26. Scolari G, Vescovo M. Peptidase localization in Lactobacillus casei 5Mn373 by stable protoplast formation. Biotechnol. Techn. 1996;10:687–692
  27. Scolari G, Vescovo M. Use of X-prolyl-dipeptidyl aminopeptidase electrophoretic mobility as a tool to differentiate Lactobacillus species of dairy relevance. Ann. Microbiol. 2004;54:529–542
  28. Scolari G, Vescovo M, Rossi G. Substrate models for L. casei peptidases profiling. Int. Dairy J. 1996;6:921–925
  29. Seppo L, Jauhiainen T, Poussa T, Korpela R. A fermented milk high in bioactive peptides has a blood pressure-lowering effect in hypertensive subjects. Am. J. Clin. Nutr. 2003;77:326–330
  30. Schmidt BF, Woodhouse L, Adams RM, Ward T, Mainzer SE, Lad PJ. Alkalophilic Bacillus sp. strain LG12 has a series of serine proteinase genes. Appl. Environ. Microbiol. 1995;61:4490–4493
  31. Sick H, Kaiser K, Henne A. The action of glycomacropeptide (GMP) on trypsin activity in vitro. Milchwissenschaft. 1990;45:76–78
  32. Smeds A, Varmanen P, Palva A. Molecular characterization of a stress inducible gene from Lactobacillus helveticus. J. Bacteriol. 1998;180:6148–6153
  33. Sridhar VR, Hughes JE, Welker DL, Broadbent JR, Steele JL. Identification of endopeptidase genes from the genomic sequence of Lactobacillus helveticus CNRZ32 and the role of these genes in hydrolysis of model bitter peptides. Appl. Environ. Microbiol. 2005;71:3025–3032
  34. Yamamoto N, Ono H, Maeno M, Ueda Y, Takano T, Momose H. Classification of Lactobacillus helveticus strains by immunological differences in extracellular proteinases. Biosci. Biotechnol. Biochem. 1998;62:1228–1230

PII: S0022-0302(06)72421-3

doi:10.3168/jds.S0022-0302(06)72421-3

Journal of Dairy Science
Volume 89, Issue 10 , Pages 3800-3809, October 2006

Article Tools

* Image Usage & Resolution