INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

In a variety of bacteria, cold shock proteins (CSPs) are the major induced proteins upon exposure to cold shock. Different functions, e.g., as transcriptional activators, RNA chaperones, and anti-freeze proteins, have been attributed to CSPs (for reviews see references 11 and 36). CspA of Escherichia coli and CspB of Bacillus subtilis have been shown to bind single-stranded DNA, and E. coli CspA has been shown to act as a transcriptional activator for the hns and gyrA genes encoding proteins involved in DNA supercoiling (2, 17, 22). E. coli CspA and B. subtilis CspB^B have very similar five-stranded -barrel structures with several outward-facing residues important for single-stranded DNA binding. Furthermore, CSPs contain two highly conserved RNA-binding motifs, named RNP-1 and RNP-2, and indeed, mRNA-binding capacity has been demonstrated for E. coli CspA and B. subtilis CspB (13, 15). It has been proposed that members of the CSP family bind to RNA in a cooperative manner and function as RNA chaperones, thereby facilitating the translation process (13). Disruption of B. subtilis cspB results in a freeze-sensitive phenotype (33) and also affects the level of induction of other cold-induced proteins (CIPs) upon temperature downshock in B. subtilis (12). Deletion of three CSPs in B. subtilis was shown to be lethal (13). Since not all members of the CSP family are cold induced, it has been suggested that CSPs play a role in multiple cellular processes, such as chromosomal condensation and/or cell division (36).

The mesophilic lactic acid bacterium Lactococcus lactis is widely used to start industrial food fermentations. A variety of genes involved in the stress response that probably are important for cell survival under stress conditions have been studied for this organism (7, 26). The L. lactis MG1363 chromosome was found to contain two pairs of tandemly located, cold-inducible csp genes (cspA-cspB and cspC-cspD) and a single, constitutively expressed cspE gene. The CSPs encoded by these genes can be divided in two groups based on isoelectric point (pI) and homology. One group consists of CspA and CspC, which have 80% identical residues, and these CSPs have a pI of 9; the other group includes CspB, CspD, and CspE, which have up to 85% identical residues, and these CSPs have a pI of 5 (35). Upon cold shock of L. lactis IL1403 by transfer from 30 to 15°C, 10-fold induction of cspB-directed -galactosidase activity is observed (4). Similar cold-induced expression has been reported for E. coli cspA, and the data revealed that the transient induction of E. coli CspA occurs at the level of transcription (14) and at the level of mRNA stabilization (1, 8, 9). Furthermore, it has been reported that mRNAs of CIPs are still translated under cold shock conditions because of the presence of a so-called downstream box, which enhances the ability to form the translation initiation complex with nonadapted ribosomes at low temperatures (23). The presence of CSPs in a cell is also determined by the stability of the proteins. The CSPs of B. subtilis undergo very rapid folding and unfolding transitions, and they exhibit low conformational stability in solution. These CSPs are rapidly degraded by proteases in vitro but are protected against proteolysis by binding to RNA (13).

Overproduction of E. coli CspA leads to increases in the levels of three CIPs (16). Moreover, heterologous expression of B. subtilis CspB in E. coli results in a reduction of cellular growth and in production of several proteins that resembles the cold shock response (10). For B. subtilis strains from which csp genes have been deleted compensatory effects of the remaining CSPs have been reported (13), and a similar response might be expected for L. lactis. Moreover, a comparative analysis of the physiological effects of overproduction of different members of a CSP family has never been presented before. For these reasons, we used the nisin-controlled expression system (21) to overproduce the CSPs of L. lactis and, subsequently, to monitor the physiological and regulatory effects of the CSPs. CspB, CspD, and CspE could be overproduced at high levels, whereas for CspA and CspC only low levels of overproduction were detected, probably due to low protein and mRNA stability at 30°C, respectively. Overproduction of specific CSPs resulted in major induction of other CSPs and CIPs, indicating that these proteins have a regulatory function. L. lactis strains overproducing CspB or CspE did not have a shorter lag time upon cold shock but did show enhanced survival after freezing.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains and culture conditions. The L. lactis strains used in this study were cultured at 30°C without aeration in M17 medium containing 0.5% glucose. L. lactis was transformed by electroporation as described by Wells et al. (32). E. coli M1601 was used as a host for cloning experiments and was grown in tryptone-yeast extract medium with aeration at 37°C (3). Chloramphenicol was used as a selection marker at a concentration of 10 µg ml¹. Growth of L. lactis was monitored by measurement of the optical density at 600 nm (OD₆₀₀).

DNA techniques and DNA sequence analysis. PCR amplifications were performed as previously described by Kuipers et al. (20) with 25 cycles (denaturation at 95°C for 30 s, primer annealing at an appropriate temperature for 1 min, and primer extension at 72°C for 2 min). All manipulations with recombinant DNA were carried out by using standard procedures (25). Plasmid DNA and chromosomal DNA of L. lactis were isolated as described previously (31). DNA sequences were determined on both strands with an ALF DNA sequencer (Pharmacia Biotech, Uppsula, Sweden) and were analyzed by using Clone (version 4.0; Clone Manager) and Lasergene (DNASTAR Inc.) software.

Construction of plasmids for overexpression. cspA, cspB, cspC, and cspE were amplified by PCR with the primers listed in Table 1, which contain either an NcoI site (forward primers) or an HindIII site (all reverse primers except OECspERev, which contains an XhoI site because of the presence of an HindIII site 6 bp downstream of the structural cspE gene). The PCR products obtained were digested with NcoI and HindIII (or XhoI) and cloned in pNZ8032 (6) digested with the same restriction enzymes. In this way, a translational fusion of the nisA promoter to the start codon of the respective csp gene that replaced the gusA gene that was originally present in pNZ8032 was obtained. The constructs were made in such a way that each csp gene contained its own putative terminator. Each of the resulting plasmids (Table 1) was transformed into L. lactis NZ3900, which contains the nisR and nisK genes of the two-component nisin transduction pathway integrated on the chromosome (5). Overproduction of CspD was obtained by using L. lactis NZ3900 containing pNZOECspD as described previously (34). To obtain overproduction, the appropriate strain was grown to an OD₆₀₀ of 0.3, and then M17W-nisin, which has a higher induction capacity and a lower growth-inhibitory effect than wild-type nisin (18), was added. For optimal overexpression the strains were incubated at 30°C for 90 min. By inserting of an NcoI site 1 bp was mutated in the second codon of cspA, cspB, and cspC. Consequently, the second amino acid was changed from isoleucine to valine, from threonine to alanine, and from asparagine to aspartic acid in CspA, CspB, and CspC, respectively.

Freeze challenge. To study the effects of CSPs on the survival of L. lactis after freezing, a freeze-thaw challenge experiment with strains overproducing the CSPs at different levels was performed as described previously (34). In short, 1-ml samples of cultures were withdrawn, spun down, and resuspended in fresh medium, the numbers of CFU were determined, and the samples were immediately frozen at 20°C for 24 h. After this freezing period the samples were thawed at 30°C for 4 min, and the numbers of CFU were determined. Next, each sample was frozen again at 20°C, and the cycle was repeated another three times. Freeze-thaw challenges were performed in duplicate, and the results were expressed as the percentage of cells remaining alive relative to the number of cells before the first freeze period (defined as 100%).

Protein extraction and protein analysis using 1D-EF and 2D-EF. Proteins were extracted by homogenization with an MSK cell homogenizer (B. Braun Biotech International, Melsungen, Germany) and zirconium beads (diameter, 0.1 mm; Biospec Products, Bartlesville, Okla.). Protein analysis was performed by using one-dimensional Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis (1D-EF) as described by Schägger and Von Jagow (27) or by using two-dimensional gel electrophoresis (2D-EF) as described by Wouters et al. (34). Samples containing equal amounts of protein were loaded on the 1D-EF gels (20 µg) and 2D-EF gels (40 µg). A low-molecular-weight marker (protein bands at 16.9, 14.4, 10.7, 8.2, and 6.2 kDa) and a high-molecular-weight marker (protein bands at 94, 68, 43, 29, 18.4, and 14.4 kDa) were used as size markers (both were purchased from Pharmacia Biotech, Uppsula, Sweden). 1D-EF gels were stained with Coomassie brilliant blue, and 2D-EF gels were stained with silver stain. The 1D-EF and 2D-EF gels were analyzed by using the Chromoscan program (Joyce Loebl, Gateshead, England) and GEMINI software (Applied Imaging, Sunderland, England), respectively. The intensity of a specific band and the intensity of a specific spot were calculated as a percentage of the total intensity of the bands visualized in a lane and as a percentage of the total intensity of the spots visualized on a gel, respectively. The protein gels were electrophoresed at least in duplicate, and a representative gel is shown below.

Determination of N termini. Protein (500 µg) was loaded on the 2D-EF gels for determination of the N termini of specific spots by using conditions identical to those used for the analytical gels. The proteins were blotted on an Immobilon-P transfer membrane (Millipore, Bedford, Mass.) by using a Trans Blot unit as recommended by the manufacturer (Bio-Rad, Richmond, Calif.). The proteins were stained with Coomassie brilliant blue, and fragments of the blot were subjected to the Edman procedure and subsequent analysis with the model 476A protein sequencing system (Applied Biosystems, Foster City, Calif.) at the Sequence Center (University of Utrecht, Utrecht, The Netherlands). The N termini derived were screened for sequence similarities by using the BlastP database.

mRNA analysis. RNA was isolated and Northern blot analysis was performed as described previously (19). Equal amounts of RNA were separated on 1% agarose gels and blotted on a GeneScreen Plus membrane (Dupont, NEN Research Products, Boston, Mass.). The blots were hybridized with [-³²P]ATP-labelled probes specific for the individual csp genes (35).

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Overproduction of CSPs in L. lactis. To elucidate the role of lactococcal CSPs in cold adaptation, L. lactis strains overproducing these CSPs were constructed (Table 1). Considerable overproduction was achieved for CspB, CspD, and CspE, which were overproduced so that they accounted for 11, 13, and 19% of the total protein, respectively. Strikingly, for CspA and CspC much lower levels of overproduction were obtained (approximately 0.5 and 0.3% of the total protein, respectively) even when a fourfold-higher concentration of nisin was used (Fig. 1A). Using the nisin-controlled expression system, stepwise overproduction of the CSPs could be achieved by adding increasing concentrations of nisin (see below). Upon addition of increasing concentrations of nisin, the growth rate of L. lactis NZ3900 harboring the overexpression constructs decreased significantly (Fig. 1B [only the data for CspD are shown; identical effects were observed for all CSPs), which was possibly explained by occupation of the transcriptional and translational machinery. The growth of the control strain was also reduced (Fig. 1C), probably for the same reasons; however, the extent of growth reduction was lower than those of the CSP-overproducing strains (Fig. 1B). The growth rate of NZ3900 (plasmid free) was not reduced upon incubation with the same concentrations of nisin (Fig. 1D), eliminating the possibility of an antimicrobial effect of nisin at the concentrations used.

View larger version (20K): [in this window] [in a new window]   FIG. 1.   Protein analysis of cell extracts of L. lactis NZ3900 overproducing CSPs and growth of L. lactis NZ3900 overproducing-sodium CSPs. (A) 1D-EF gels containing cell extracts of L. lactis NZ3900 containing pNZOECspA (lane 2), pNZOECspB (lane 3), pNZOECspC (lane 4), pNZOECspD (lane 5), and pNZOECspE (lane 6) induced with 2.0 ng of nisin/ml for CspA and CspC or with 0.5 ng of nisin/ml for CspB, CspD, and CspE for 90 min at 30°C. Lane 1 contained a low-molecular-weight marker (protein bands at 16.9, 14.4, 10.7, 8.2, and 6.2 kDa). The arrow indicates the position of the CSPs. (B to D) Growth of L. lactis NZ3900 harboring pNZOECspD (B), L. lactis NZ3900 harboring pNZ8020 (C), and L. lactis NZ3900 (D) upon induction with nisin. Growth was measured at 30°C by determining the OD600 without induction () or with induction with 0.1 ng of nisin/ml (), 0.2 ng of nisin/ml (), 0.5 ng of nisin/ml (), 1.0 ng of nisin/ml (×), or 2.0 ng of nisin/ml (). The arrows indicate when nisin was added, and the bulleted arrows indicate when protein was extracted.

Limited overproduction of CspA and CspC is explained by low protein and mRNA stability. To further investigate the differences in the levels of overproduction obtained for the two groups of lactococcal CSPs, the mRNA levels for lactococcal cspA, cspC, and cspD (positive control) were monitored upon overexpression (Fig. 2A). For cspA and cspD a major increase in the mRNA level was observed upon induction with 0.5 and 2.0 ng of nisin/ml. Thus, despite the similar high mRNA levels for cspA and cspD, a concomitantly high protein level was obtained only for CspD. For cspC a low level of mRNA induction was observed compared to the levels of mRNA induction for cspA and cspD, which provides an explanation for the low amount of CspC obtained. Since the transcription signals for all csp overexpression constructs are identical, the low mRNA level is most likely caused by the low stability of the cspC transcript at 30°C.

View larger version (67K): [in this window] [in a new window]   FIG. 2.   Analysis of L. lactis NZ3900 overproduction of CspA and CspC on the mRNA and protein levels. (A) Northern analysis of cspA (left), cspC (center), and cspD (right) in L. lactis NZ3900 containing pNZOECspA, L. lactis NZ3900 containing pNZOECspC, and L. lactis NZ3900 containing pNZOECspD, respectively, upon induction with 0, 0.5, and 2.0 ng of nisin/ml. Equal amounts of total RNA were loaded on the gel, and the arrow indicates the position of csp transcripts. The film used for detection of the cspC transcript was exposed four times longer to the blot than the films used for detection of cspA and cspD. (B) 1D-EF of cell extracts of L. lactis NZ3900 containing pNZOECspA or pNZOECspA* without induction (lanes 2 and 6) and with induction with 0.2 ng of nisin/ml (lanes 3 and 7), 0.5 ng of nisin/ml (lanes 4 and 8), and 2.0 ng of nisin/ml (lanes 5 and 9). Lane 1 contained a low-molecular-weight marker (16.9, 14.4, 10.7, 8.2, and 6.2 kDa). Equal amounts of protein were loaded on the gel, and the arrow indicates the position of CSPs.

View larger version (83K): [in this window] [in a new window]   FIG. 3.   2D-EF analysis of cell extracts of L. lactis NZ3900 overproducing the lactococcal CSPs. Cell extracts were isolated from L. lactis NZ3900 containing pNZOECspA (A), pNZOECspC (B), pNZ8020 (C), or pNZOECspA* (D) upon induction with 2.0 ng of nisin/ml and were separated at pI 3 to 10. Cell extracts were also isolated from L. lactis NZ3900 containing pNZOECspB (E), pNZOECspD (F), pNZOECspE (G), or pNZ8020 (H) induced with 0.1 ng of nisin/ml and were separated at pI 4 to 7. Equal amounts of protein were loaded on the gel and visualized by silver staining. The positions of molecular weight markers are indicated on the left (high-molecular-weight marker) or on the right (low-molecular-weight marker). Spots corresponding to CSPs of L. lactis are enclosed in squares or rectangles, and the letters indicate the CSP families (F = putative CspF, G = putative CspG); the calculated pIs are 8.5, 5.0, 9.1, 4.5, and 4.6 for CspA, CspB, CspC, CspD, and CspE, respectively. Spots induced by overproduction of the CSPs are enclosed in circles and numbered if they belong to previously identified CIPs (Wouters et al., submitted) or are enclosed in pentagons and designated with letters (X, Y, Z) if they are induced during specific overproduction of CSPs. See text for details.

Regulatory role of CSPs in cold shock adaptation. To study the effect of overproduction of CSPs on protein synthesis patterns in L. lactis, 2D-EF was performed. The overproduced CSPs migrated approximately at their calculated pIs except for CspC, as confirmed by N-terminal sequencing (Fig. 3). Overproduction of CspC (Fig. 3B and Table 2) in L. lactis results in induction of CspB, protein F (putative CspF), and previously unidentified protein G (putative CspG) in the 7-kDa region. Upon induction of control cells (L. lactis harboring pNZ8020) (Fig. 3C and Table 2) with 2.0 ng of nisin/ml, increased levels of production of CspC, CspD, and CspE were observed compared to the levels in noninduced cells (data not shown), which is possibly explained by the reduction in growth (Fig. 1C) and the resulting stress conditions. Overproduction of CspA (Fig. 3A and Table 2) did not result in induction of CSPs, whereas overproduction of CspA* (Fig. 3D and Table 2) resulted in a slight increase in the CspE level compared to that in control cells. Furthermore, overproduction of CspB, CspD, or CspE did not affect the level of any of the other CSPs (Fig. 3E through H and Table 2; data not shown for CspA, which was visualized only on gels at pI 3 to 10). Next, upon overproduction of CspA, CspB, CspC, and CspD a slight decrease in the level of CspE was observed, which might indicate possible downregulation of cspE expression by the respective CSPs. This observation is in agreement with data on the disruption of CspA and CspB, which results in derepression of cspE expression at a low temperature (J. A. Wouters, H. Frenkiel, W. M. De Vos, O. P. Kuipers, and T. Abee, submitted for publication).

Effect of overproduction of CSPs on cold adaptation and survival of L. lactis after freezing. The effect of overproduction of CspB or CspD, the major cold-induced CSPs of L. lactis (34, 35), on adaptation to cold shock was tested. These CSPs were overproduced by using relatively low nisin concentrations (0.05 and 0.1 ng of nisin/ml), yielding levels of expression comparable to those observed under cold shock conditions, after which the cultures were exposed to 10°C (without nisin). For all conditions identical adaptation times of 6 to 7 h were observed (data not shown), indicating no beneficial effect on cold adaptation of elevated CspB or CspD expression prior to cold shock. Also, overproduction of CspA and CspC did not result in differences in cold adaptation compared to that of control cells (NZ3900 harboring pNZ8020).

View larger version (27K): [in this window] [in a new window]   FIG. 4.   Survival of L. lactis NZ3900 after freezing upon overproduction of CspB (A) and CspE (B). Survival is expressed as the percentage of surviving cells compared to the number of cells prior to freezing for L. lactis NZ3900 harboring pNZOECspB or pNZOECspE without induction () or with induction with 0.2 ng of nisin/ml () or 0.5 ng of nisin/ml (). (C) 1D-EF of cell extracts of L. lactis NZ3900 containing pNZOECspB and L. lactis NZ3900 containing pNZOECspE without induction (lanes 1 and 4) and with induction with 0.2 ng of nisin/ml (lanes 2 and 5) and 0.5 ng of nisin/ml (lanes 3 and 6), respectively. Equal amounts of protein were loaded on the gel.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this report, we describe the effects of overproduction of five different CSPs of the lactic acid bacterium L. lactis. The CSPs of L. lactis can be divided into two groups, one consisting of CspA and CspC and the other consisting of CspB, CspD, and CspE, based on amino acid composition and pI (35). Overproduction of CSPs of the latter group resulted in levels of overproduction much higher than those obtained for CspA and CspC (11 to 19% versus 0.3 to 0.5% of the total protein). Most of the CSPs migrated on the 2D-EF gels at their calculated pIs; the only exception was CspC, which had a calculated pI of 8.4 (instead of 9.1) in the overproduction construct because of the Asp-Asn substitution in the second codon. However, the CspC protein was found to migrate at a pI of approximately 6, as confirmed by N-terminal sequencing. No formylation of the N terminus, such as that found for B. subtilis CspB (12), was observed, and it is speculated that the pI shift is caused by posttranslational modifications. Furthermore, two new proteins in the 7-kDa region were identified for L. lactis: one designated putative CspF that was also induced upon cold shock as described previously (34) and the other designated putative CspG that has a pI of 5.5. These putative CSPs might be encoded by genes located on an uncharacterized 3.5-kb HindIII-hybridizing fragment on the L. lactis chromosome (35). The growth rate of L. lactis is reduced upon overproduction of CspB, CspD, and CspE at 30°C. Also, a reduced growth rate was observed for control cells (L. lactis NZ3900 harboring pNZ8020) upon incubation with nisin, but overproduction of CSPs further reduced the growth rate and this might be explained by energy consumption and more intensive occupation of the transcription and translation machinery. Similar growth-inhibitory effects were observed for heterologous expression of B. subtilis CspB in E. coli at 37°C (10).

The observation that CspB, CspD, and CspE of L. lactis can be overproduced to obtain large quantities at 30°C is remarkable. Artificial overexpression of the E. coli cspA gene was very low at 37°C due to the low mRNA stability (9). The low level of overexpression obtained for the lactococcal cspC gene is most likely explained by the low stability of the transcript at 30°C, as shown by Northern blotting. Since high mRNA induction for cspA is observed, the low level of CspA overproduction should be explained by other factors. CspC of B. subtilis contains an Ala residue at position 58 whereas CspB and CspD of B. subtilis contain a Pro residue at this position, and indeed, B. subtilis CspB and CspD were far more stable than B. subtilis CspC (28). At 30°C CspA* can be overproduced to obtain much larger quantities than the quantities of CspA, which is most likely explained by the high stability of the protein as a consequence of the reduced entropy. In addition to the specific Arg-Pro mutation, the decrease in pI may also contribute to CspA* stability. Recently, it was shown that the stabilities of the CSPs of B. subtilis increased significantly in the presence of a nucleic acid ligand. It has been suggested that the stabilities of these CSPs in vivo are mediated by binding to mRNA (28), which might be largely dependent on the overall charge of the protein. Schröder et al. (30) suggested that CSPs may act as RNA chaperones since they possess a positively charged RNA-binding epitope that is backed by a negatively charged surface that would prevent approach of RNA by charge repulsion. Mutation of the surface-exposed Phe residues at positions 15, 17, and 27, which are important for nucleic acid binding, has been found to result in decreased stability of the protein, probably due to a decrease in the nucleic acid-binding capacity that makes the protein more susceptible to protease action (28, 29). This would provide an alternative explanation for the increased stability of CspA* and could also explain why the CSPs of L. lactis that have a low pI can be overproduced to obtain large quantities. Overproduction was also examined at 10°C, and still no overproduction of CspA and CspC was observed, indicating that mRNA and/or proteins were not sufficiently stabilized to yield detectable protein levels (data not shown). For L. lactis NZ3900 harboring pNZOECspD moderate CspD overproduction was observed at 10°C compared to that of control cells (data not shown), which shows that the nisin-controlled expression system is functional at a low temperature but with reduced efficiency.

Upon overproduction of CspC the levels of CspB, putative CspF, and putative CspG increased, whereas overproduction of CspA, CspB, CspD, or CspE did not affect the expression of other CSPs. CspC might directly stimulate the expression of the other CSPs by transcriptional activation, as has been reported for genes regulated by E. coli CspA for which Y-box motifs (ATTGG or CCAAT) have been shown to be important (2, 17, 22). Several of these elements are observed in the upstream regions of the lactococcal csp genes (34). The observation that all lactococcal CSPs induce expression of certain proteins, including several CIPs, indicates a regulatory function for this group of proteins. The production of CIP2, CIP4, CIP5, and CIP9 seems to be regulated by several CSPs, indicating that there is overlap in regulatory pathways. The N terminus of CIP2 was identified and shows homology to 50S ribosomal protein L9 of B. subtilis. Cold-induced ribosomal proteins have been reported for both E. coli (S1, S6, L7/L12) (16) and B. subtilis (S6, L7/L12) (12), and it has been suggested that they are essential for correct assembly of rRNA at low temperatures. For overproduction of CspB, CspD, and CspE induction is also observed for non-cold-induced proteins X, Y (CelA), and Z. The observed induction of protein Y, homologous to a cellobiose-specific enzyme II subunit of the phosphotransferase system of B. subtilis (24), might be an indication that CSPs play a role in processes other than cold adaptation (e.g., sugar metabolism).

Overproduction of CSPs did not stimulate adaptation of L. lactis to cold shock conditions. However, similar to overproduction of CspD (34), overproduction of CspB and CspE resulted in 10- and 5-fold-greater survival after freezing compared to that of control cells, respectively. For B. subtilis disruption of B. subtilis cspB resulted in decreased survival after freezing (a 14-fold decrease compared to survival of wild-type cells), indicating an essential role for this gene in protection against freezing (33). CspB, CspD, and CspE may enhance the survival of L. lactis after freezing either directly by protecting RNA or DNA by the moderate, nonspecific binding activities mentioned for CSPs (e.g., RNA stabilization) or indirectly by inducing other factors that provide cryoprotection. The proteins that are induced upon overproduction of the lactococcal CSPs might be involved in cryoprotection of L. lactis.

In this work overproduction of specific CSPs of L. lactis and several physiological effects were studied. Overproduction of different quantities could be obtained for all CSPs depending on mRNA and protein stability. 2D-EF revealed that overproduction of CSPs at 30°C resulted in induction of a specific group of proteins, including CIPs. It is concluded that CSPs of L. lactis play a regulatory role in the cold shock response and that they control the production of both CSPs and CIPs.