Production, secretion and purification of a correctly folded staphylococcal antigen in Lactococcus lactis
Lactococcus lactis, a lactic acid bacterium traditionally used to ferment milk and manufacture cheeses, is also, in the biotechnology field, an interesting host to produce proteins of medical interest, as it is “Generally Recognized As Safe”. Furthermore, as L. lactis naturally secretes only one major endogenous protein (Usp45), the secretion of heterologous proteins in this species facilitates their purification from a protein-poor culture medium. Here, we developed and optimized protein production and secretion in L. lactis to obtain proteins of high quality, both correctly folded and pure to a high extent. As proteins to be produced, we chose the two transmembrane members of the HtrA protease family in Staphylococcus aureus, an important extra-cellular pathogen, as these putative surface-exposed antigens could constitute good targets for vaccine development.
A recombinant ORF encoding a C-terminal, soluble, proteolytically inactive and tagged form of each staphylococcal HtrA protein was cloned into a lactococcal expression-secretion vector. After growth and induction of recombinant gene expression, L. lactis was able to produce and secrete each recombinant rHtrA protein as a stable form that accumulated in the culture medium in similar amounts as the naturally secreted endogenous protein, Usp45. L. lactis growth in fermenters, in particular in a rich optimized medium, led to higher yields for each rHtrA protein. Protein purification from the lactococcal culture medium was easily achieved in one step and allowed recovery of highly pure and stable proteins whose identity was confirmed by mass spectrometry. Although rHtrA proteins were monomeric, they displayed the same secondary structure content, thermal stability and chaperone activity as many other HtrA family members, indicating that they were correctly folded. rHtrA protein immunogenicity was established in mice. The raised polyclonal antibodies allowed studying the expression and subcellular localization of wild type proteins in S. aureus: although both proteins were expressed, only HtrA1 was found to be, as predicted, exposed at the staphylococcal cell surface suggesting that it could be a better candidate for vaccine development.
In this study, an efficient process was developed to produce and secrete putative staphylococcal surface antigens in L. lactis and to purify them to homogeneity in one step from the culture supernatant. This allowed recovering fully folded, stable and pure proteins which constitute promising vaccine candidates to be tested for protection against staphylococcal infection. L. lactis thus proved to be an efficient and competitive cell factory to produce proteins of high quality for medical applications.
KeywordsLactococcus lactis Cell factory Secretion Staphylococcus aureus antigen HtrA family Soluble recombinant protein Chaperone
Lactococcus lactis, a Gram-positive lactic acid bacterium and a classical starter for the manufacture of cheeses, can be used as a cell factory to produce proteins of interest [1, 2, 3, 4, 5, 6]. As a long known innocuous, Generally Recognised As Safe (“GRAS”) food-grade species , L. lactis is an interesting host to produce proteins of medical interest [2, 4, 5]. Compared to Escherichia coli, the advantage of L. lactis is that it does not produce endotoxin (lipopolysaccharide) [2, 4, 5] which has to be removed from protein preparations before medical use . In contrast to Bacillus subtilis, L. lactis secretes only one major endogenous protein, Usp45, and no proteases : a strategy combining production and secretion of heterologous proteins in L. lactis is thus interesting as it facilitates protein purification from the culture medium . As secreted heterologous proteins can be degraded by the lactococcal surface protease HtrA, protein yield can be improved by the use of a mutant strain devoid of this surface proteolytic activity .
Several tools have been developed for protein production in L. lactis, in particular expression systems and vectors [11, 12, 13], secretion signals [4, 14, 15] and expression-secretion vectors [4, 13, 16, 17]. Proteins of medical interest have been successfully produced and secreted by L. lactis, in general to be delivered to a host [1, 2], and in a few cases to be purified [16, 18]. In our laboratory, a tightly regulated expression system (ZitR-regulated P zit promoter) [11, 19], an efficient export signal (SPExp4 signal peptide) [4, 20], expression-secretion vectors [4, 21] and mutant host strains devoid of surface proteolytic activity [10, 22, 23] have been developed for L. lactis and used for protein production and secretion [4, 21, 22, 24]. Furthermore, an enzyme of biotechnological interest, the staphylococcal nuclease, which is naturally secreted, could be produced, secreted and purified in L. lactis . Here, we developed, in L. lactis, the secretion of high quality proteins for medical applications. We chose, as model proteins, putative surface-exposed antigens and virulence factors: the two HtrA family members of Staphylococcus aureus, an important extra-cellular pathogen species.
The HtrA family is composed of highly conserved, extra-cytoplasmic serine proteases [25, 26]. In prokaryotes, they are located in the cell envelope: in either the periplasm or the cytoplasmic membrane in Gram-negative bacteria, and in the cytoplasmic membrane in Gram-positive bacteria. Although some HtrA proteases, like E. coli DegS, have a regulatory function, most of them are involved in the protein quality control in the bacterial cell envelope [25, 26, 27]. They are often essential for survival to various stress conditions, notably heat [28, 29] and/or oxidative stress , because they alleviate protein unfolding and misfolding [25, 27]. They can act both as proteases to degrade proteins and as chaperones to assist them in folding, like E. coli DegP/HtrA, the family model [25, 27].
In many pathogens, HtrA proteins are involved in virulence . Several models have been proposed to account for this role. First, HtrA proteases could, under the stress conditions prevailing in the host during infection, degrade unfolded and misfolded proteins and thus indirectly improve cellular fitness and survival . Second, HtrA proteases could play a direct role by processing endogenous, folded, wild type (WT) proteins, as first demonstrated in L. lactis, a food-grade species [10, 32] (and unpublished data), and subsequently confirmed in a pathogenic species, Bacillus anthracis, even though in that case the HtrA target is not a virulence factor . Third, in some Gram-negative pathogens, HtrA proteases could target host proteins like E-cadherin [34, 35, 36] or the interleukin IL8 , even if it remains unclear how intra-cellular HtrA proteins are able to reach their extra-cellular host targets [37, 38]. Finally, HtrA proteins could also contribute to virulence as chaperones, either by enhancing in vivo growth and survival, , or by improving the folding of a virulence factor , or even by contributing to bacteria–host interaction .
In pathogens, HtrA proteins are also often important antigens. In several species, they were found to be immunogenic in vivo, in either infected [42, 43, 44] or convalescent hosts (animals or human patients) [43, 44, 45, 46, 47, 48, 49], even though in Chlamydia trachomatis the meaning of these results with respect to disease remains controversial [50, 51]. Moreover, purified HtrA proteins from some Gram-negative pathogens were shown to be protective against infection [45, 49, 52, 53, 54] even though this was not always the case [55, 56]. Surprisingly, to our knowledge, no such protection studies have been performed using HtrA proteins of Gram-positive extra-cellular pathogens, despite the fact that, as cell surface exposed proteins, they could be recognized by circulating antibodies at an early infection step and might thus constitute good targets for vaccine development.
In S. aureus, an extra-cellular pathogenic species, we previously identified and studied two putative transmembrane members of the HtrA family: HtrA1 and HtrA2 [23, 57]. WT HtrA1 protein from strain RN6390 was found to display extra-cellular proteolytic activity when over-produced in L. lactis . Furthermore, both HtrA proteins were, together, implicated in the virulence and extra-cellular proteome composition of strain RN6390, and each of them was involved in the stress resistance of strain COL . In an independent study, among a peptide library from strain COL, a few HtrA1 peptides were found to be antigenic and to elicit an immunological response in vivo, in infected patients . As predicted surface-exposed proteins, staphylococcal HtrA proteins from strain COL, and in particular the HtrA1 antigen, constitute interesting vaccine candidates , and they were thus chosen to be produced in L. lactis and purified. Recombinant ORFs encoding soluble, proteolytically inactive and tagged HtrA forms were cloned into a lactococcal expression and secretion vector. Recombinant rHtrA protein yield was evaluated after lactococcal growth in rich medium in flasks or in fermenters after medium optimization. After lactococcal growth in the optimized rich medium in fermenters and induction of gene expression, secreted rHtrA proteins were purified in one step from the culture medium. After confirming their identity my mass spectrometry, rHtrA proteins were analysed in vitro for their secondary structures, stability, oligomeric state and chaperone activity, in order to analyze their folding. Finally, their immunogenicity was tested in mice, and the generated antibodies allowed studying the expression and cell surface exposure of WT HtrA proteins in S. aureus.
Bacterial strains and growth conditions
Bacterial strains and plasmids used in this study
(A) Bacterial strains
endA1, recA1, gyrA96, thi, hsdR17 (r k − , m k + ), relA1, supE44, Δ(lac-proAB), [F´ traD36, proAB, laqIqZΔM15]
plasmid free derivative of NCDO712
Lowenstein (ATCC 49521)
clinical isolate, capsular polysaccharide CP5-positive
clinical isolate, methicillin resistant
COL htrA 1 :: cat, CmR
COL htrA 2 :: spc, SpcR
T-tailed PCR product Cloning vector, AmpR
pGEMT derivative where recombinant htrA 1r ORF is cloned
htrA 1r encodes HtrA1-ΔTM-Ser255Ala-His6 protein
pGEMT derivative where recombinant htrA 2r ORF is cloned
htrA 2r encodes SPExp4-HtrA2-ΔTM-Ser619Ala-His6 protein
pWV01; CmR; expression-secretion vector where exp4 SP -nuc is cloned under the control of a lactococcal expression system, P zit zitR
exp4 SP -nuc encodes a hybrid protein between a lactococcal signal peptide (SPExp4) and the mature secreted form of the staphylococcal nuclease (NucB)
pLB145 derivative where htrA 1r ORF is cloned in place of nuc ORF to be fused in frame to exp4 SP
exp4 SP -htrA 1r encodes a hybrid precursor leading to secreted rHtrA1 protein
pLB145 derivative where htrA 2r ORF is cloned in place of exp4 SP -nuc ORF
htrA 2r encodes a hybrid precursor leading to secreted rHtrA2 protein
PCR and cloning
High-fidelity PCRs using chromosomal DNA from S. aureus strain COL as a template were performed to obtain recombinant ORFs encoding N-terminally truncated, inactive and tagged forms of staphylococcal HtrA proteins. The recombinant htrA 1r ORF (encoding HtrA1-ΔTM-Ser255Ala-His6) was obtained as follows. (1) The 5′ and the 3′ regions of WT htrA 1 gene were amplified by a high-fidelity Taq polymerase (FINNZYMME) using respectively the following primer couples (see Additional file 1: Table S1 for primer sequences): 1NFΔTM2 (with a NsiI site at its 5′ end) and 1IRS > A (bearing a point mutation for Ser255Ala substitution), or 1IFS > A (bearing a point mutation for Ser255Ala substitution) and 1CRHIS (bearing a sequence encoding His6 with a limited risk of ribosomal slippage, and bearing an EcoRI site at its 3′ end). (2) Using a mix of both the resulting PCR fragments as a template, 1NFΔTM2 and 1CRHIS as primers and a high-fidelity A-tailing Taq polymerase (DNA Expand ROCHE), an overlap PCR fragment, htrA 1r , was obtained.
The recombinant htrA 2r ORF (encoding SPExp4-HtrA2-ΔTM-Ser619Ala-His6) was obtained as follows. (1) The 5′ and the 3′ regions of WT htrA 2 gene were amplified by a high-fidelity Taq polymerase (FINNZYMME) using respectively the following primer couples: either 2NFΔTM2 and 2IRS > A (bearing a point mutation for Ser619Ala substitution), or 2IFS > A (bearing a point mutation for Ser619Ala substitution) and 2CRHIS (bearing the sequence encoding His6 tag and an EcoRI site). (2) Using a mix of both the resulting PCR fragments as a template, and 2NFΔTM2 and 2CRHIS primers, an ORF fragment (containing a NsiI site) was then obtained by high-fidelity overlap PCR. (3) An ES fragment encoding the expression and secretion system (PZn zitR exp4 SP ) of pLB145  was amplified using PznF and SPΔTM2R primers. Using a mix of ES and ORF fragments as a template and PznF and 2CRHIS primers, a high-fidelity overlap PCR was performed. Finally, using the resulting fragment as a template, ZitH2F and 2CRHIS primers and a A-tailing Taq polymerase (DNA Expand ROCHE), htrA 2r PCR fragment was obtained.
Each htrA 1r and htrA 2r PCR fragment was cloned into pGEM-T Easy Vector (Promega), according to the manufacturer’s instructions. After transformation of highly competent JM109 cells, selection on ampicillin, X-Gal and IPTG, and screening for white colonies, recombinant plasmids were extracted, insert size was checked by PCR, and the inserts were sequenced. The resulting plasmids were respectively named pVE8124 and pVE8125 (Table 1B). htrA 1r and htrA 2r fragments were respectively recovered from pVE8124 or pVE8125 plasmids by NsiI + EcoRI or BamHI + EcoRI double digestions. They were subcloned into pLB145  digested by the same enzymes (Additional file 2: Figure S1). After transformation into strain MG1363, the resulting plasmids were checked by PCR and sequenced. The final plasmids were respectively named pVE8126 and pVE8127 (Table 1B).
Recombinant protein production in L. lactis
Recombinant lactococcal cells were grown in 5 mL of M17 medium (buffered by 88 mM β-glycerophosphate) supplemented by 1% glucose, overnight in tubes. For growth in flasks, an overnight culture was diluted 100-fold in 280 mL of the same medium in a flask, and let to grow at 30°C. For growth in fermenters at controlled pH, serial tenfold dilutions of an overnight culture were grown overnight in M17 medium supplemented by 2% glucose), and cultures still in the exponential phase (OD600 between 0.4 and 0.7) were diluted 100-fold in 800 mL of preheated medium, either the same medium or a derivative, more strongly buffered, medium (176 mM β-glycerophosphate) to be grown in a fermenter (Biostat Q, Sartorius), at 30°C and at pH 6.5 (by addition of 5 N NaOH under shaking at 100 rpm).
When cultures reached an OD600 of 0.5 (or in one case, 2), induction was achieved by adding 500 µM EDTA, and cultures were further incubated for 4 h (or for 2 h in the case of the culture induced at OD600 2). After centrifugation at low speed and at 4°C, supernatants were filtered on 0.22 µm, concentrated by about 25-fold by ultrafiltration (under a nitrogen pressure of less than 3 bars and at 4°C) using Millipore filters (Ultrafiltration Membranes NMWL 10,000), and stored at 4°C before purification.
Dry resin (His-Select Nickel Affinity Gel, Sigma; 1 mL) was washed on a Econo-Pac column (Chromatography Columns; BIO-RAD) with 5 volumes of water and 10 volumes of buffer A (50 mM NaH2PO4, 300 mM NaCl, 10 mM Imidazole (Sigma®); pH 8.0), incubated for 15 min at 4°C, equilibrated with 10 volumes of buffer A and incubated for a further 15 min at 4°C. In parallel, each supernatant was diluted twofold in an equal volume of buffer B (100 mM NaH2PO4, 600 mM NaCl, 20 mM imidazole at pH 8.0). Resin (1 mL) and treated supernatant were mixed. After incubation overnight at 4°C, the mix was loaded onto the column and washed twice with buffer A. After addition of 3 mL of buffer C [50 mM NaH2PO4, 300 mM NaCl, 250 mM imidazole (Sigma®); pH 8.0] repeated three times, the three elution fractions were pooled, concentrated nine- to tenfold by ultrafiltration (to reach a final volume of 1 mL), dialysed against PBS buffer overnight at 4°C with shaking (Slide-A-Lyser® Dialysis Cassette 0.5–3 mL/3,500 Da, Pierce®). Finally, glycerol was added to a final concentration of 10% before storage at −20°C. Protein concentration was determined by the Bradford method (Bio-Rad Protein Assay) according to the manufacturer’s instructions, using BSA (Protein Assay Standard II) as a standard. After each concentration or purification step, the protein fractions were analysed for protein purity and stability by Western blotting using antibodies against HtrA1  or His6 (INVITROGEN).
Protein analysis by mass spectrometry
rHtrA proteins were loaded on precasted 4–12% Bis–Tris Mini Gels (Invitrogen, France), SDS-PAGE runs were performed, and gels were stained with Coomassie blue (BioRad, Marnes-la-Coquette, France). Bands of interest were excised and protein identification was performed as previously described  using PAPPSO platform facilities (Jouy-en-Josas, France; http://pappso.inra.fr). Following SDS-PAGE migration and trypsinolysis, protein identification was performed querying MS/MS data against a in-house database containing rHtrA1 and rHtrA2 sequences (Additional file 3: Figure S2) together with the database for L. lactis subsp. cremoris strain MG1363 proteins (Uniprot, 2011/03/04; http://www.uniprot.org/uniprot/?query=organism:mg1363&fil=organism:%22Lactococcus%20lactis%20subsp.%20cremoris%20%28strain%20MG1363%29%20%22&sort=score) and a in-house contaminant protein database. The number of spectras attributed to each rHtrA protein was high, whereas no L. lactis protein could be detected.
Secondary structure analysis by circular dichroism
Synchrotron radiation circular dichroism experiments, covering the UV spectral range from 190 to 305 nm, were carried out at the DISCO beam line of the Synchrotron SOLEIL (Saint Aubin, France; http://www.synchrotron-soleil.fr). Spectra were acquired at different temperatures, progressively increasing by 5°C in the 25–95°C range. CaF2 (Calcium Fluoride) 50 μm optical pathlength cells were loaded with 2 µL of each protein (rHtrA1 or rHtrA2) in a 50 mM Tris HCl (pH 7.5) and 50 mM NaCl buffer to reach a final protein concentration of 1 mg/mL. For each temperature curve, the mean of three spectra (acquired at 1 nm step per second between 190 and 305 nm) was calculated before subtraction of the baseline (by buffer subtraction) and set to zero between 255 and 260 nm. Mean spectra were calibrated to a standard solution of (+)-camphor-10-sulphonic acid (CSA), normalized and converted to Δε (molar circular dichroism, M−1 cm−1) using the software CDtool . The thermal denaturation curve of each rHtrA protein was calculated at 207 nm, and the melting temperature (Tm) was determined from a sigmoidal fit. Spectra were represented using Origin software (OriginLab, Northampton, MA, USA).
Size exclusion chromatography coupled to multi angle light scattering
Purified rHtrA1 or rHtrA2 proteins (30 μL of samples at 1–4 mg/mL) were loaded on a KW-803 column (Shodex) equilibrated in PBS buffer at a 0.5 mL/min flow rate (Shimadzu HPLC system). Detection was performed using a MiniDAWN TREOS multi angle light scattering detector and an Optilab T-rEX differential refractometer (Wyatt Technology). Molar mass was calculated with the Astra 188.8.131.52 software, using a differential index of refraction (dn/dc) value of 0.183 mL/g.
Protein folding assays
First, the refolding of two proteins, citrate synthase and α-glucosidase, was followed. Denaturation and renaturation reactions were carried out at 20°C. For both proteins, renaturation was initiated by pouring the renaturation solvent onto the unfolded protein under vortex agitation in Eppendorf polyethylene tubes. Citrate synthase was denatured at a concentration of 10 µM in 8 M urea, 50 mM Tris, 2 mM EDTA, 20 mM dithiothreitol pH 8.0 for 30 min. Renaturation was initiated by a 100-fold dilution in 40 mM Hepes, 50 mM KCl, 10 mM (NH4)2SO4, 2 mM potassium acetate, pH 8.0, in the absence of added protein or in the presence of DnaK, rHtrA1 or rHtrA2. The enzymatic activity of citrate synthase was measured as described previously . α-Glucosidase was denatured at a concentration of 2 µM in 8 M urea, 0.1 M potassium phosphate, 1 mM EDTA, 20 mM dithiothreitol, pH 7.0 for 5–10 min. Renaturation was initiated by a 30-fold dilution in 40 mM Hepes–KOH, pH 7.8 at 20°C. The enzymatic activity of α-glucosidase was measured as described previously . Concentrations of substrate proteins and chaperones were similar to those used by us  and others  for investing chaperone activities of GroEL and thioredoxin. Pig heart citrate synthase and Saccharomyces cerevisiae α-glucosidase were from Sigma. DnaK was purified as described in .
Second, the thermal aggregation of citrate synthase was followed. The native enzyme (80 µM) was diluted 100-fold in 40 mM Hepes, 50 mM KCl, 10 mM (NH4)2SO4, 2 mM potassium acetate, pH 8.0, at 44°C, in the absence of added proteins, or in the presence of DnaK, rHtrA1 or rHtrA2. Citrate synthase aggregation was monitored by measuring the absorbance at 650 nm as described in .
Immunization of mice and Western blotting
OF1 mice were immunized with 10 μg of rHtrA1 or rHtrA2 co-injected with SP02 adjuvant (proprietary adjuvant of Sanofi Pasteur) by the subcutaneous route (0.2 mL) in the scapular girdle region at day 0, day 21 and day 36. Blood samples were collected under anaesthesia (see above) at day 0 and at day 61 at the retro-orbital sinus.
Staphylococcal culture pellets were lysed according to optical density. An equivalent of OD680nm = 20 was lysed, for each pellet, in 250 µL of 2X lysis buffer: Tris–HCl 20 mM, Triton 1.2%, PMSF 1 mM, Halt Protease Inhibitor Single-Use Cocktail (Thermo Scientific) 1X, benzonase (Sigma) 5 U/µL and lysostaphin (Sigma) 100 µg/µL and incubated over night at 37°C. This suspension was heat inactivated 15 min at 95°C.
Samples were resolved by SDS 4–12%-PAGE (NuPAGE, Invitrogen) and transferred to nitrocellulose membranes (Transblot Transfert Medium, BioRad). Membranes were saturated with 5% skim milk in PBS buffer (2.7 mM KCl, 137 mM NaCl) overnight at 4°C. Filters were then incubated with 1:200 dilution of the primary antibody in 1% skim milk in PBS buffer for 1 h at room temperature, washed three times for 5 min with PBS with 0.05% Tween 20 (PBST). Filters were then incubated with 1:3,000 dilution of a Peroxidase AffiniPure F(ab’)2 fragment goat anti-mouse IgG (H + L) (Jackson ImmunoResearch Laboratories, USA) with 1% skim milk in PBS buffer for 1 h at room temperature and washed three times for 5 min as described above. Finally, the filters were washed 10 min in deionized water. Immunopositive bands were visualized using the Amplified Opti-4CN Kit (Bio-Rad Laboratories, Inc., USA) and were quantified using a densitometer (Genetools, Syngene, UK).
Flow cytometry assay
The ability of polyclonal antisera elicited by the recombinant proteins to bind to the surface of live S. aureus strains was determined using flow cytometric detection of indirect fluorescence. Strain Lowenstein was grown at 37°C with shaking. Frozen bacteria were inoculated into 50 mL of appropriate medium and grown till exponential, late exponential or late stationary phases. A culture sample was centrifuged and washed once with PBS (Eurobio, Courtaboeuf, France). The final pellet was resuspended in PBS with 1% bovine serum albumin (BSA, Eurobio, Courtaboeuf, France) at a density of 108 CFU/mL. 20 µL of dilutions of pooled serum were added to 20 µL of bacteria in 96 deep-well plates (Ritter, Schwabmunchen, Germany). For each serum, three dilutions were tested: 1/200, 1/2,000 and 1/20,000. The plate was incubated for 1 h at 37°C with shaking. The bacteria were centrifuged, washed once with PBS 1% BSA and resuspended with 100 µL of goat anti-mouse IgG F(ab’)2 conjugated to PE (Southern Biotech, Birmingham, USA) diluted 100-fold. The plate was incubated for 1 h at 37°C with shaking in the dark. The bacteria were washed twice with PBS 1% BSA. The fluorescent staining of bacteria was analyzed on a Cytomics FC500 flow cytometer (Beckman Coulter, Fullerton, USA). The fluorescent signal obtained for bacteria incubated with the specific polyclonal antisera was compared to the signal obtained for bacteria incubated with the corresponding negative control serum (buffer + SP02 adjuvant alone).
Results and discussion
Design of recombinant staphylococcal HtrA proteins
For each HtrA protein of the staphylococcal strain COL, a soluble, proteolytically inactive and tagged form was produced (see Additional file 3: Figure S2). (1) The soluble, C-terminal region (devoid of the transmembrane domain, Figure 1) was fused in frame to a lactococcal signal peptide (see below, ) to be produced as a secreted protein. (2) The conserved catalytic Ser residue (Ser255 or Ser619 in the case of HtrA1 or HtrA2 respectively) was replaced by an Ala residue to abolish proteolytic activity, and thus to avoid the self-degradation previously observed for WT proteins when overproduced in a lactococcal htrA mutant strain . (3) A His6 tag was fused at the C-terminus of each recombinant protein to facilitate its purification.
To delete the N-terminal transmembrane domain of each HtrA protein without affecting the overall protein folding, we exploited the strategy previously applied to produce a recombinant, soluble form of the E. coli trans-membrane protein DegS (sharing 31% identity with each of the staphylococcal HtrA proteins) whose structure could be solved . The precise boundary of the deletion in the staphylococcal HtrA proteins was chosen to preserve, upstream of the region homologous to recombinant DegS (Figure 1), an N-terminal negatively charged residue to contribute to the negative charge at the N-terminus of the mature secreted rHtrA protein (see below).
Lactococcal system for protein production and secretion
For protein production and secretion in L. lactis, we chose to clone the recombinant ORFs in pLB145, a lactococcal expression and secretion vector . In the resulting plasmids (Additional file 2: Figure S1), recombinant ORFs are under the control of the lactococcal ZitR-regulated P zit promoter [11, 19] so that their expression can be induced by the addition of EDTA [4, 11]. They are fused in frame to the coding sequence of the lactococcal signal peptide SPExp4 (bearing two positively charged residues at its N-terminus) plus the two negatively charged residues present at its C-terminus in the endogenous mature Exp4 protein [4, 20]. This fusion strategy preserves both the natural charges around SPExp4 to ensure its correct insertion into the membrane (according to the positive-inside rule [65, 66]), and its natural cleavage site for the lactococcal signal peptide peptidase. Finally, this fusion strategy should ensure efficient secretion both at the translocation and release steps.
Production and secretion of recombinant proteins in L. lactis, and purification
rHtrA protein identity
rHtrA proteins are correctly folded
rHtrA proteins are active chaperones
We then investigated the function of rHtrA1 and rHtrA2 under heat shock conditions. As reported previously [63, 72], citrate synthase loses its native conformation and undergoes aggregation during incubation at 44°C. As shown in Figure 7b, 5 µM rHtrA1 or 5 µM rHtrA2 reduced citrate synthase aggregation by 30%, whereas 5 µM DnaK reduced citrate synthase aggregation by 49%. On the contrary, ovalbumin and lysozyme, as previously reported , were inefficient in protecting citrate synthase from thermal denaturation (not shown). These results suggest that rHtrA1 and rHtrA2 can interact with partially unfolded proteins and protect them against thermal denaturation.
Together, these results demonstrate for the first time that the soluble domains of staphylococcal HtrA proteins possess chaperone activity in vitro. This activity is in the case of HtrA2 protein independent of the long, N-terminal domain of unknown function , at least in vitro, and further studies will be needed to determine if this domain could contribute to envelope protein folding in S. aureus and/or to virulence in vivo. Finally, our results indicate that rHtrA proteins are correctly folded, and thus strongly suggest that they should expose conformational epitopes relevant for vaccine applications.
Purified rHtrA proteins are immunogenic in mice and WT HtrA1 protein is exposed at the staphylococcal cell surface
These antibodies were then used to study the expression and cell surface localization of HtrA proteins in S. aureus after growth under different conditions. In a preliminary experiment, expression was studied in two strains (strain COL and strain Lowenstein which is responsible for systemic infections) grown in two media: a defined medium (containing a high salt concentration), and a complex standard medium supplemented or not with 2.2′ Dipyridyl (a chelator leading to iron depletion) as high salt concentration and iron depletion are close to the conditions faced by S. aureus in the host. Western blotting allowed detecting full-length HtrA1 and HtrA2 proteins in the cells in similar amounts under all conditions (Figure 8), indicating a constitutive expression under the tested conditions and the absence of extensive degradation, as confirmed by the analysis of culture supernatants (data not shown).
In this study, the C-terminal region of staphylococcal HtrA transmembrane proteins could efficiently be produced and secreted in L. lactis as correctly folded and stable forms that were easily purified from the culture medium in one step. L. lactis was demonstrated to be an efficient cell factory with respect to protein quality, in terms of both purity and folding, in particular in the case of a surface-exposed antigen, like staphylococal HtrA1 protein. Our results indicate that to produce proteins of high quality and purity for medical applications, L. lactis is an efficient and competitive alternative to E. coli and B. subtilis hosts.
FS designed and performed all experiments for recombinant rHtrA protein production in L. lactis under the supervision of IP. FT and DS, under the supervision of BR, carried out Western blotting and flow cytometry experiments. DC performed MS/MS experiments. CV and PF were in charge of SRCD and SEC-MALS analysis. VG performed chaperone activity tests under the supervision of GR. BR, PF, DC and GR wrote the relevant sections of the manuscript they were involved in. IP conceived and supervised the whole project and wrote the manuscript. All authors read and approved the final manuscript.
This work benefited from the facilities and expertise of the Imagif Structural and Proteomic Biology Pole of the Centre de Recherche de Gif (https://www.imagif.cnrs.fr) biophysics platform. We thank Sophie Ruiz (Sanofi Pasteur) for her participation in this study, Vincent Juillard (Micalis, Jouy-en-Josas) and Mireille Yvon (Micalis, Jouy-en-Josas) for their help in ultracentrifugation and fermenter experiments respectively. Thanks to Maarten van de Guchte (Micalis, Jouy-en-Josas) for critical reading of the manuscript.
INRA is a government research organism that has two patents on protein production by L. lactis: i) Zinc-regulated prokaryotic expression cassettes: US 8,354,272 B2, CA 2496350, EP 1537215 B1 and WO 2004/020640; and ii) Gram-positive bacteria deprived of HtrA protease activity and their uses: US 6,994,997 B1, EP 1141337 B1 and WO 2000/039309.
Bachra Rockbi, Delphine Seguin and Fabienne Telles are employees of Sanofi Pasteur an international pharmaceutical company, notably involved in vaccine development. Sanofi Pasteur has a patent: Method for the Production of Overproducing Staphylococcus aureus Strains, US20100880566 20100913.
- 20.Poquet I, Ehrlich SD, Gruss A (1998) An export-specific reporter designed for Gram-positive bacteria: application to Lactococcus lactis. J Bacteriol 180:1904–1912Google Scholar
- 23.Rigoulay C, Poquet I, Madsen SM, Gruss A (2004) Expression of the Staphylococcus aureus surface proteins HtrA1 and HtrA2 in Lactococcus lactis. FEMS Microbiol Lett 237:279–288Google Scholar
- 28.Lipinska B, Fayet O, Baird L, Georgopoulos C (1989) Identification, characterization, and mapping of the Escherichia coli htrA gene, whose product is essential for bacterial growth only at elevated temperatures. J Bacteriol 171:1574–1584Google Scholar
- 29.Strauch KL, Johnson K, Beckwith J (1989) Characterization of degP, a gene required for proteolysis in the cell envelope and essential for growth of Escherichia coli at high temperature. J Bacteriol 171:2689–2696Google Scholar
- 43.Roop RM 2nd, Fletcher TW, Sriranganathan NM, Boyle SM, Schurig GG (1994) Identification of an immunoreactive Brucella abortus HtrA stress response protein homolog. Infect Immun 62:1000–1007Google Scholar
- 53.Loosmore SM, Yang YP, Oomen R, Shortreed JM, Coleman DC, Klein MH (1998) The Haemophilus influenzae HtrA protein is a protective antigen. Infect Immun 66:899–906Google Scholar
- 56.Bae JE, Schurig GG, Toth TE (2002) Mice immune responses to Brucella abortus heat shock proteins. Use of baculovirus recombinant-expressing whole insect cells, purified Brucella abortus recombinant proteins, and a vaccinia virus recombinant as immunogens. Vet Microbiol 88:189–202CrossRefGoogle Scholar
- 59.Rokbi B, Lafont C (2010) Method for the production of overproducing Staphylococcus aureus strains. US 20100880566 20100913 C12N1/20Google Scholar
- 63.Jakob U, Gaestel M, Engel K, Buchner J (1993) Small heat shock proteins are molecular chaperones. J Biol Chem 268:1517–1520Google Scholar
- 66.von Heijne G (1986) The distribution of positively charged residues in bacterial inner membrane proteins correlates with the trans-membrane topology. EMBO J 5:3021–3027Google Scholar
- 68.Cates GA, Yang YP, Klyushnichenko V, Oomen R, Loosmore SM (2000) Properties of recombinant HtrA: an otitis media vaccine candidate antigen from non-typeable Haemophilus influenzae. Dev Biol (Basel) 103:201–204Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.