INTRODUCTION

Top Abstract Introduction Materials and Methods Results and Discussion References

Lactic acid bacteria (LAB) play an important role in dairy fermentation processes and have a great influence on the quality and preservation of end products. The primary roles of LAB are to produce lactic acid from lactose, resulting in a pH decrease, and, by proteolysis, to liberate short peptides and free amino acids affecting the flavor and texture of dairy products.

Since the concentration of free amino acids and small peptides is insufficient to support the growth of LAB to high cell densities in milk, these bacteria are dependent on a proteolytic system to liberate free amino acids from milk proteins. The proteolytic system of LAB consists of a cell envelope-associated proteinase, membrane-bound transport systems, and several cytoplasmic peptidase classes. The proteolytic system is particularly important in the development of flavor and texture of cheeses (9). Since Lactococcus strains, along with those of Lactobacillus, are widely used as starters in cheese manufacture, substantial effort has been directed in the last two decades toward elucidating the proteolytic mechanism of Lactococcus lactis. More recently, the proteolytic system of lactobacilli has also been extensively examined.

Over 10 different peptidase types have been identified in various LAB strains, and a large number of peptidase genes have been cloned from different Lactococcus and Lactobacillus species and characterized (reviewed recently by Christensen et al. [4]). For most of the characterized peptidases from Lactobacillus helveticus, an L. lactis counterpart with a similar type of specificity can also be found. However, the overall proteolytic activity of L. helveticus has been found to be higher than that of L. lactis (14, 22). This has led to the use of lysed or heat-shocked L. helveticus cells as flavor adjuncts in cheese processes based on the use of other starters (8). The wide range of peptidases that have been molecularly characterized is now enabling heterologous-expression studies with these peptidases in different lactic acid bacteria. This will be of importance in elucidation of the roles of different peptidases in cheese manufacture, enhancement of the maturation process, and development of new cheeses with improved characteristics.

Based on the favorable proteolytic properties of L. helveticus, our goal has been to study whether the peptidolytic profiles of L. lactis can be changed or improved with peptidases from L. helveticus to provide new strains for testing in cheese processes. In this study, we have constructed new L. lactis strains carrying six previously characterized peptidase genes, i.e., pepN, pepX, pepC, pepI, pepD, and pepR, from industrial L. helveticus strain 53/7 and studied their expression in several peptidase mutants of L. lactis and in different growth media. The results revealed that not all of the promoters from the L. helveticus peptidases are functional in L. lactis. Furthermore, it was shown that expression of the tested L. helveticus peptidases was not affected by the genetic background of L. lactis peptidases. In addition, growth defects of peptidase mutants of L. lactis could be complemented with corresponding L. helveticus genes, and, surprisingly, L. helveticus PepN alone could also complement an L. lactis mutant lacking five peptidases (hereafter termed a fivefold peptidase-negative mutant of L. lactis). Unexpectedly, there were also substantial differences in the responses to the growth media for some of the peptidases.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results and Discussion References

Bacterial strains, plasmids, and growth conditions. The lactococcal strains and plasmids used in this study are listed in Table 1. L. lactis cells were routinely grown at 30°C in M17 medium (Difco) supplemented with 0.5% (wt/vol) glucose or lactose. For determination of enzyme activities, the recombinant L. lactis strains were cultivated to stationary phase in glucose-M17 (GM17) for 12 h and in citrate-buffered milk (Valio Ltd., Helsinki, Finland) for 16 h. Erythromycin and chloramphenicol were used as selection agents at concentrations of 5 and 10 µg/ml, respectively, when needed. Escherichia coli DH5 was grown in Luria broth medium with aeration at 37°C. Erythromycin (300 µg/ml) was added to the growth medium when required. L. helveticus was grown in MRS (Difco) broth at 37°C without aeration.

DNA methods. Chromosomal DNA from L. helveticus 53/7 and L. lactis was isolated essentially as described by Marmur (17). Plasmid DNA was isolated from L. lactis by alkaline lysis (21) of cells. Purification of plasmid DNA was performed by phenol-chloroform extraction followed by anion-exchange chromatography (DNA Plasmid Midi Kit; Qiagen, Hilden, Germany). Plasmid DNAs were isolated from E. coli by using commercial kits (Pharmacia, Uppsala, Sweden, and Wizard, Madison, Wis.). Oligonucleotides were synthesized with an Applied Biosystems DNA synthesizer (model 392) and purified by gel filtration with NAP-10 columns (Pharmacia). Standard DNA methods were used as described by Sambrook et al. (21). Lactococcal and E. coli strains were electrotransformed by using a Bio-Rad Gene Pulser under conditions of 2.5 kV, 400 , and 25 µF and 2.5 kV, 200 , and 25 µF, respectively.

Plasmid constructions. All of the plasmids constructed in this study are listed in Table 1. A 3.7-kb insert carrying the L. helveticus pepN gene (26) was obtained from pKTH2073 by digestion with XbaI and ligated into pKTH2095 (23). The new plasmid, pKTH2172, was propagated in an E. coli DH5 host. A 2.7-kb insert carrying pepX (31) was obtained from plasmid pKTH2097 by digestion with BamHI and ligated into pKTH2095 to yield a new plasmid, pKTH2171. In L. helveticus, the pepC gene is expressed both as a monocistronic mRNA and as a polycistronic transcript with its adjacent, downstream open reading frame (orf2) (29). For this study, the pepC gene was synthesized without the downstream orf2 by PCR using a pair of specific primers (5'-AAAACTGCAGAGCTTAAGGCAGTTCAATCAGATCAG-3' and 5'-AAAACTGCAGCTAAATTGCTAGCAAATTTTTTGCC-3') containing PstI sites at their 5' ends for cloning. Ligation of the 1.7-kb PstI fragment into the pIL277 vector gave a new recombinant plasmid, pKTH2175. The 1.079-kb L. helveticus pepI (28) coding region was ligated with the L. helveticus pepX promoter by using specific primer pairs (5'-CCGGAATTCGCGTTCAATTTATTATTGCAATTTACG-3'-5'-CA ATAATTTCCATCTTTTTCTCCTTTGTCAGTATTATTACC-3' and 5'-CAAAGGAGAAAAAGATGGAAATTATTGAAGGAAAAATGCC-3'-5'- GGGGAATTCCAGTAACCAACAAACGCTACGTTAAAG-3'), resulting in a transcriptional fusion (PpepX= pepI) (Fig. 1). The new plasmid, based on the vector pKTH2095, was designated pKTH2179. Genes pepD (30) and pepR (27) were digested from pKTH2105 and pKTH2082 with BamHI-SphI and BamHI-SalI, respectively. The new plasmids, pKTH2150 and pKTH2170, which were propagated in E. coli DH5, were constructed by ligating the pepD and pepR inserts into the vector pKTH2095. The recombinant plasmids pKTH2172, pKTH2171, pKTH2175, pKTH2179, pKTH2150, and pKTH2170, harboring pepN, pepX, pepC, PpepX-pepI, pepD, and pepR, respectively, were introduced into the fivefold peptidase mutant L. lactis MG1363 [XTOCN] (Table 1) by electroporation and screened as described by Pedersen et al. (20) or by PCR using specific probes. The pepD and pepR genes were also cloned as translational fusions with the inducible lactococcal nisin promoter (PnisA) in vector pNZ8037 (5). The pepD coding region was synthesized by PCR using a pair of primers (5'-CATGCCATGGCAAAACAAACAGAATGTAC-3' and 5'-CGGGATCCGGAATTGATGTGGTACTTGTTCCAG-3') containing NcoI and BamHI sites at their 5' and 3' ends for cloning. The NcoI cloning site at the translation start codon gave rise to an additional alanine after the initiation methionine on the coding sequence of pepD. The NcoI-BamHI fragment of 1.7 kb, carrying the pepD structural gene, was ligated into the pNZ8037 vector (Fig. 1) and transformed into L. lactis NZ9000 (Table 1). The pepR coding region was synthesized by PCR using a pair of primers (5'-CAATGTCATGAAAACTGGTACTAAAATCATTAC-3' and 5'-CGGGATCCTTGTTATAATTCTAGCATATTAGGGAG-3') containing BspHI and BamHI sites at their 5' and 3' ends for cloning. The BspHI-BamHI fragment of 1.0 kb, carrying the pepR structural gene, was ligated with pNZ8037 (Fig. 1) and transferred into NZ9000. The recombinant plasmids pKTH2182 and pKTH2193, harboring the pepD and pepR inserts, respectively, were screened as described by Pedersen et al. (20).

View larger version (29K): [in this window] [in a new window]   FIG. 1.   Promoter fusion regions of pepI, pepD, and pepR.

RNA isolation and Northern hybridization. Total RNA was isolated from L. lactis and L. helveticus cells, grown in GM17 and MRS, respectively, by using an RNeasy Midi Kit (Qiagen). RNA gel electrophoresis and Northern blots were performed as described previously (11). A 0.24- to 9.5-kb RNA ladder (Gibco BRL Life Technologies, Rockville, MD) was used as a standard. For use as a hybridization probe, an 864-bp HpaI-BamHI fragment of L. helveticus pepN, a 994-bp HaeIII fragment of L. helveticus pepX, a 1.0-kb PCR fragment of pepC (primers 5'-AGGTCCCGGGTAAAGGAGGATTTTTAATGG-3' and 5'-CACGGCGGTAAAGATTGG-3'), and a 1.079-kb PCR fragment of the pepI coding region (primers 5'-CAAAGGAGAAAAAGATGGAAATTATTGAAGGAAAAATGCC-3' and 5'-GGCGAATTCCAGTAACCAACAAACGCTACGTTAAAG-3') were labeled with digoxigenin-dUTP (DIG; Boehringer Mannheim, Mannheim, Germany). A DIG luminescence detection kit (Boehringer) was used for hybrid detection.

Enzyme assays. L. lactis cells were disrupted with an Ultrasonic 2000 sonicator (B. Braun, Melsungen, Germany), cell debris was removed, and the PepN, PepX, PepC, and PepI activities of the cell extracts were determined by the method of El Soda and Desmazeaud (7). The substrates used were 16.4 mM L-lysine p-nitroanilide (Sigma) for PepN and PepC, 16.4 mM L-proline p-nitroanilide for PepI, and 16.4 mM L-glysine-proline p-nitroanilide for PepX. The buffers and temperatures used were 50 mM Tris-HCl (pH 7.5) and 45°C for PepN, 50 mM 2-(N-morpholino) ethanesulfonic acid (MES; pH 6.5) and 45°C for PepX, 50 mM Tris-HCl (pH 7.0) and 40°C for PepC, and 50 mM Tris-HCl (pH 7.5) and 37°C for PepI, respectively. To determine PepC activity, PepN activity was inhibited with 5 mM EDTA (26, 29). PepD activity was determined from cell extracts by the Cd-ninhydrin method (6) with 2 mM Leu-Leu in 50 mM MES (pH 6.0) at 55°C. PepR activity was determined as described by Baankreis and Exterkate (1) at 37°C with 2 mM Pro-Leu as a substrate. In determining PepD and PepR activity, 1 U is defined as the amount of enzyme activity producing a variation of 0.01 in absorbance at 505 or 480 nm, respectively, per minute. The protein concentrations were determined with the Bio-Rad protein assay reagent based on the Bradford dye-binding procedure (2). Bovine serum albumin (Sigma) was used as a protein standard. All enzyme activities presented are averages of two to four parallel measurements.

SDS-PAGE. The production of PepD and PepR under nisin-induced conditions was monitored by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (11% [wt/vol] acrylamide gels) using the procedure of Laemmli (15). The gel was stained with Coomassie brilliant blue R250. LMW-SDS proteins (Pharmacia) were used as molecular weight markers.

	RESULTS AND DISCUSSION

Top Abstract Introduction Materials and Methods Results and Discussion References

Construction of strains for expression of L. helveticus genes in L. lactis. The pepN, pepX, pepD, and pepR genes of L. helveticus were subcloned from the plasmids described in Materials and Methods into E. coli DH5 by using the vector pKTH2095. The new plasmids, designated pKTH2172, pKTH2171, pKTH2150, and pKTH2170, respectively, were transferred into a fivefold peptidase-negative mutant strain, [XTOCN], of L. lactis MG1363 (Table 1). The pepC and pepI genes reside in operon structures in L. helveticus (7, 28). For this work, the pepC gene and pepI under the control of the pepX promoter (PpepXpepI) were isolated by PCR, transferred into pKTH2095, and cloned as plasmid constructs pKTH2179 and pKTH2175, respectively, directly into L. lactis. Due to stability problems encountered with pepC in pKTH2095, its cloning vector was changed to pIL277. The pepN, pepX, pepC, and pepI genes (pKTH2172, pKTH2171, pKTH2175, and pKTH2179) were also transferred into the wild-type strain L. lactis MG1363 with and without the lactose protease plasmid pLP712. Furthermore, pepN, pepX, and pepC were transferred into the MG1363[XTOCN] mutant carrying pLP712. The pepN, pepX, and pepC genes were also introduced into pLP712-carrying L. lactis mutants MG1363[N], MG1363[X], and MG1363[NC], respectively (Table 1).

Growth patterns of lactococcal strains. The growth curves of the lactococcal strains and strain-plasmid combinations tested in this study revealed no significant differences between the strains when grown in GM17. In milk, the growth of the fivefold peptidase-negative mutant MG1363[XTOCN] and the pepN and pepN pepC deletion mutants MG1363[N] and MG1363[NC] was impaired as described earlier by Mierau et al. (19). L. helveticus pepN (pKTH2172) in L. lactis restored the growth of MG1363[N] and MG1363[NC] to the wild-type level. In contrast to overexpressed L. lactis pepN in pNZ1120, which can restore the pepN deficiency but, cannot complement the growth defects of other pep mutations in MG1363[XTOCN] in milk (19), L. helveticus pepN (pKTH2173) was able to restore the growth of this fivefold mutant almost to the wild-type L. lactis level (Fig. 2). This may suggest broader substrate specificity for L. helveticus pepN than for L. lactis. This is also in agreement with our earlier observation that L. helveticus PepN can to some extent hydrolyze, for example, proline-containing peptides (P. Varmanen and A. Palva, unpublished data). As expected, L. helveticus pepC and pepX alone could not compensate for the growth defects of MG1363[XTOCN] in milk.

View larger version (25K): [in this window] [in a new window]   FIG. 2.   Growth curves of L. lactis MG1363 () and its derivatives MG1363[N] (), MG1363[XTOCN] (), MG1363 + pepN (), MG1363[N] + pepN (), and MG1363[XTOCN] + pepN () in milk.

Expression of L. helveticus peptidase genes in L. lactis. Peptidase activities were studied at different time points of growth. Enzyme activities were determined from cell lysates of milk- or GM17-grown L. lactis MG1363 and MG1363[XTOCN] hosts in the presence or absence of the pLP712 plasmid and from their transformants harboring L. helveticus pepN, pepX, pepC, pepI, pepD, or pepR. In addition, PepN, PepX, and PepC activities were also determined in L. lactis MG1363[N], MG1363[X], and MG1363[NC] peptidase mutants, respectively. Recombinant strains harboring pepN, pepX, pepC, or pepI gave rise to detectable enzyme activities of the L. helveticus-derived peptidases both in M17- and in milk-grown cells. In contrast, no L. helveticus-derived PepD or PepR activity exceeding the background level of L. lactis was found. The genetic background of any of the L. lactis derivatives tested did not affect the expression level of any of the L. helveticus peptidases studied. Since the amounts of each L. helveticus peptidase activity were equal in all lactococcal hosts, only the enzyme activities of MG1363 and its different L. helveticus peptidase transformants are shown.

View larger version (25K): [in this window] [in a new window]   FIG. 3.   Expression of L. helveticus pepN in L. lactis. (A) Peptidase activity of L. lactis MG1363 and its pepN+ transformant strain in GM17- and milk-grown cells. GM17-grown cells: MG1363 (), MG1363 + pepN (). Milk-grown cells: MG1363 (), MG1363 + pepN (). (B) Northern blot hybridization. Lanes a, samples (5 µg of total RNA) taken at an OD600 of 0.5; lanes b, samples taken 3 h thereafter. Lanes 1, MG1363[XTOCN]; lanes 2, pepN+ transformant of MG1363[XTOCN]; lanes 3, L. helveticus 53/7; lanes 4, MG1363; lanes 5, pepN+ transformant of MG1363. The size of pepN mRNA is indicated with an arrow.

View larger version (29K): [in this window] [in a new window]   FIG. 4.   Expression of L. helveticus pepX in L. lactis. (A) Peptidase activity of L. lactis MG1363 and its pepX+ transformant strain in GM17- and milk-grown cells. GM17-grown cells: MG1363 (), MG1363 + pepX (). Milk-grown cells: MG1363 (), MG1363 + pepX (). (B) Northern blot hybridization. Lanes a, samples (20 µg of total RNA) taken at an OD600 of 0.5; lanes b, samples taken 3 h thereafter. Lanes 1, MG1363[XTOCN]; lanes 2, pepX+ transformant of MG1363[XTOCN]; lanes 3, L. helveticus 53/7; lanes 4, MG1363; lanes 5, pepX+ transformant of MG1363. The size of pepX mRNA is indicated with an arrow.

View larger version (33K): [in this window] [in a new window]   FIG. 5.   Expression of L. helveticus pepC in L. lactis. (A) Peptidase activity of L. lactis MG1363 and its pepC+ transformant strain in GM17- and milk-grown cells. (A) GM17-grown cells: MG1363 (), MG1363 + pepC (). Milk-grown cells: MG1363 (), MG1363 + pepC (). (B) Northern blot hybridization. Lanes a, samples (20 µg of total RNA) taken at an OD600 of 0.5; lanes b, samples taken 3 h thereafter. Lanes 1, MG1363[XTOCN]; lanes 2, pepC+ transformant of MG1363[XTOCN]; lanes 3, L. helveticus 53/7; lanes 4, MG1363; lanes 5, pepC+ transformant of MG1363. The sizes of mono- and polycistronic pepC mRNAs are indicated with arrows.

View larger version (20K): [in this window] [in a new window]   FIG. 6.   Expression of L. helveticus pepI in L. lactis. (A) Peptidase activity of L. lactis MG1363 and its pepI+ transformant strain in GM17- and milk-grown cells. GM17- grown cells: MG1363 (), MG1363 + pepI (). Milk-grown cells: MG1363 (), MG1363 + pepI (). (B) Northern blot hybridization. Lanes a, samples (40 µg of total RNA) taken at an OD600 of 0.5; lanes b, samples taken 3 h thereafter. Lanes 1, MG1363[XTOCN]; lanes 2, pepI+ transformant of MG1363[XTOCN]; lanes 3, L. helveticus 53/7. The size of pepI mRNA is indicated with an arrow.

Expression of pepD and pepR under the control of PnisA. The pepD and pepR structural genes were isolated by PCR with primers which were designed according to previously characterized L. helveticus genes and contained at their initiation codons restriction sites suitable for cloning. After ligation of these genes into pNZ8037, the resulting plasmids, pKTH2182 and pKTH2193 (Table 1), were transferred into L. lactis NZ9000. Expression of pepD and pepR in these recombinant strains was induced with nisin at different concentrations. Recombinant peptidase activity assays and SDS-PAGE analyses of nisin-induced samples withdrawn as a function of growth were carried out. The highest recombinant peptidase activities were obtained with a nisin concentration of 5 ng/ml, whereas the higher nisin concentrations resulted in retarded growth and a decline of peptidase activities. SDS-PAGE of the soluble fractions of lysates showed increasing amounts of PepD and PepR with nisin concentrations up to 20 and 5 ng/ml, respectively (Fig. 7B and C). SDS-PAGE carried out from insoluble fractions of lysates showed that recombinant PepD and PepR began to accumulate in the cell as insoluble aggregates at nisin concentrations of 0.5 and 5 ng/ml, respectively.

View larger version (29K): [in this window] [in a new window]   FIG. 7.   Expression of L. helveticus pepD and pepR in L. lactis NZ9000. (A) Relative recombinant PepD and PepR activities produced under the control of PnisA in GM17-grown cells after induction with nisin. Shown are PepD activity in NZ9000 + pepD at 1 h () and 3 h () after induction with nisin and PepR activity in NZ9000 + pepR at 1 h () and 3 h () after induction with nisin. (B) SDS-PAGE of the recombinant PepD-producing L. lactis transformant 3 h after induction with nisin. (C) SDS-PAGE of the recombinant PepR-producing L. lactis transformant 3 h after induction with nisin. Lanes 1, uninduced cells; lanes 2 to 6, induction with 0.05, 0.5, 5, 20, and 50 ng of nisin, respectively.