OUP user menu

Expression of Streptococcus pneumoniae antigens, PsaA (pneumococcal surface antigen A) and PspA (pneumococcal surface protein A) by Lactobacillus casei

Maria Leonor S. Oliveira , Vicente Monedero , Eliane N. Miyaji , Luciana C.C. Leite , Paulo Lee Ho , Gaspar Pérez-Martínez
DOI: http://dx.doi.org/10.1016/S0378-1097(03)00645-1 25-31 First published online: 1 October 2003


A number of recent research works in lactic acid bacteria aim towards the design of new strains that could be used as live vectors for the delivery of antigens for oral vaccination, or other therapeutic molecules. In this work, an inducible expression system based on the Lactobacillus casei lactose operon promoter was used to express three important surface antigens of Streptococcus pneumoniae in this lactic acid bacterium: a virulence-related pneumococcal surface antigen (PsaA) and two variants of the virulence factor PspA (pneumococcal surface protein A). Expression of the three proteins was induced upon growth on lactose and strongly repressed by glucose. These proteins were produced intracellularly. Also, secretion to the growth medium was achieved by means of a fusion to the secreting and processing signals from the L. casei surface proteinase. Interestingly, while secreted PspA proteins were found in the culture supernatants, PsaA remained trapped in the cell wall. Expression of pneumococcal antigens in a food-grade organism opens an alternative for mucosal vaccination against this important pathogen.

  • Lactobacillus casei
  • Pneumococcal antigen
  • Mucosal vaccine

1 Introduction

Streptococcus pneumoniae is a human pathogen that colonizes the upper respiratory tract and is the major cause of important diseases such as pneumonia, otitis media, bacteremia and meningitis [1]. The increasing incidence of antibiotic-resistant S. pneumoniae clinical isolates supports the development of new and more effective streptococcal vaccines. Despite the efficacy of vaccines composed of streptococcal polysaccharides [2], protection is restricted to the serotypes included and their production is expensive, limiting their use in developing countries. For these reasons, a number of laboratories have focused their efforts on the characterization of S. pneumoniae proteins that could induce broader protection against nasopharyngeal carriage and bacteremia [35]. The pneumococcal surface antigen A (PsaA) is a 37-kDa lipoprotein genetically conserved among all S. pneumoniae serotypes [6] and is essential for S. pneumoniae virulence [7]. Immunization with PsaA was shown to protect mice against nasopharyngeal carriage and lung colonization [3,4]. The pneumococcal surface protein A (PspA) is another important virulence factor found attached to the cell wall of all S. pneumoniae strains [8]. PspA is highly immunogenic but it shows some variability among S. pneumoniae strains. According to DNA homology and protein sequences of the N-terminal domain, PspAs can be grouped into three families; 99% of the isolates have PspA belonging to families 1 and 2, in approximately equal proportions, while family 3 PspAs are only present in about 1% of the strains. Recent studies have shown that immunization with different fragments of PspA leads to protection against S. pneumoniae infection in different models [9,10]. Mucosal immunization has proven to be effective against bacterial colonization and further spread to the systemic circulation [11]. Among the advantages of mucosal vaccines, the relative facility and low costs in administration should be stressed. The model lactic acid bacterium Lactococcus lactis has been used as a host for the expression of many bacterial and viral antigens and proved to elicit immune response after inoculation by different mucosal routes [12]. Genetically modified lactobacilli could also be potentially used as live vaccine vectors, because they are also GRAS organisms (‘generally regarded as safe’) used as food starter cultures, but with the additional advantages that they are normally symbiotic organisms associated with human mucosae and some of them have probiotic properties. In addition, lactobacilli have been shown to have intrinsic adjuvant activity, stimulating mucosal and systemic immune responses against associated antigens [13]. Different laboratories have developed genetic tools that allowed the expression of heterologous proteins of clinical interest in lactobacilli [1417]. An interesting example was the expression system based on the lactose-inducible promoter of the lactose operon from Lactobacillus casei, which is controlled by an antitermination mechanism mediated by LacT and lactose availability and regulated by catabolite repression [18,19]. This work describes the molecular cloning, expression and secretion oitaS. pneumoniae PsaA and the N-terminal region of PspAs from families 1 and 2 in L. casei, using an expression-secretion system based on the L. casei lactose operon promoter and the leader sequence from the gene encoding L. casei cell wall proteinase, PrtP [20]. This leader sequence encodes the secretion signal — that is cleaved by the signal peptidase during secretion — and a propeptide that is removed by PrtM (PrtP maturase) to obtain the active form of PrtP [29].

2 Materials and methods

2.1 Bacterial strains and growth conditions

L. casei CECT 5275 (formerly ATCC 393 [pLZ15]) was routinely grown in MRS medium (Oxoid), at 37°C, without shaking. For the analysis of heterologous protein expression, L. casei was grown in basal MRS medium (10 g peptone, 8 g beef extract, 4 g yeast extract, 1 ml Tween 80, 2 g potassium phosphate, 5 g sodium acetate, 2 g di-ammonium citrate, 0.2 g magnesium sulfate and 0.05 g manganese sulfate, per liter) buffered with 0.2 M potassium phosphate pH 6.7 and supplemented with 0.5% glucose or lactose, as required. Escherichia coli DH5α was grown in Luria–Bertani medium (LB), at 37°C, with vigorous shaking. Plating of bacteria was performed on their respective media with 1.8% agar. Antibiotic concentrations used for the selection of transformants were 300 µg ml−1 of erythromycin for the selection of E. coli transformants and 5 µg ml−1 for L. casei.

2.2 Plasmids and recombinant DNA procedures

Expression plasmids were derived from the pIAβ5 vector in which the lac promoter region was introduced [19,21]. The gene encoding PsaA from S. pneumoniae serotype 6B (strain St 472/96 from the Instituto Adolpho Lutz, São Paulo, SP, Brazil) was isolated from plasmid pCI-psaA [22] as a 867-bp fragment by restriction with BamHI and EcoRV. This fragment was then cloned into pAIlacGFP [23], previously digested with BamHI and EcoRI (made blunt with the Klenow fraction of E. coli DNA polymerase I) enzymes, yielding pIA-psaA (intracellular expression). The same psaA fragment was cloned into pIA1133PVA [24], pre-digested with BamHI and KpnI (made blunt with the Klenow enzyme), obtaining pIAS-psaA. This created a translational fusion of PsaA to the first 201 amino acids of PrtP from L. casei [20] including the signal sequence and the maturase processing sequence (pro-sequence).

The N-terminal regions (α-helix region plus the proline-rich domain) of PspA1 and PspA3 (PspA′1 and PspA′3, respectively) were amplified by polymerase chain reaction (PCR) from the plasmids pTG-pspA′1 and pTG-pspA′3 ([10]; pspA′1 accession number AYO82387; pspA′3 accession number AY082389) which carried genes isolated from S. pneumoniae St 435/96 and St 259/98 strains, respectively (Instituto Adolpho Lutz, São Paulo, SP, Brazil). PCRs were performed using the Expand High Fidelity PCR system (Roche Molecular Biochemicals), in a 50-µl volume containing 200 µM of each deoxynucleoside triphosphate and 20 pmol of each primer (pspA1forw: 5′-GGGAGATCTACCATGATCTTAGGG-3′; pspA1rev: 5′-GGGCCATGGTTATCTAGATGGTTG-3′; pspA3forw: 5′-GGGAGATCTAGAACCATGATCTTAGGG-3′; pspA3rev: 5′-GGACCATGGTTATTTTGGTGCAGGAGC-3′). PCR amplification conditions were as follows: 94°C, 4 min; 30 cycles of 94°C, 30 s; 55°C, 30 s; 72°C, 1.5 min; 72°C, 7 min for final extension. Both PCR products, of pspA′1 (1092 bp) and pspA′3 (1200 bp), were cleaved at the 5′ terminus with BglII restriction endonuclease and then ligated to BamHI- and EcoRI- (made blunt with the Klenow enzyme) digested pIAlacGFP and BamHI/KpnI- (made blunt with the Klenow enzyme) digested pIA1133PVA, yielding pIA-pspA′1, pIA-pspA′3, pIAS-pspA′1 and pIAS-pspA′3 (Fig. 1). Restriction endonucleases and Klenow enzyme were purchased from Gibco BRL (Invitrogen Corp., Carlsbad, CA, USA) and used as recommended by the manufacturer. Nucleic acid manipulation and general cloning procedures were performed according to laboratory manuals [25].

Figure 1

Schematic representation of the different constructions used in this work. plac represents the lactose promoter from L. casei; SD, Shine–Dalgarno sequence; LacT′, six first codons of lacT; prtPSP, fragment encoding the signal peptide from L. casei PrtP; prtPPro, fragment encoding the pro-sequence from PrtP. The black triangle represents the cleavage site by PrtM in the PrtP sequence.

Electroporation of L. casei was carried out as previously described [26]. E. coli competent cells were prepared by the RbCl procedure and transformed as described in laboratory manuals [25]. Antibiotic-resistant E. coli and L. casei clones were screened for the presence of the insert of interest by PCR, using specific primers as described above, or by digestion of plasmids with restriction endonucleases. Positive clones were frozen in MRS (L. casei) or LB (E. coli) containing 15% glycerol, at −80°C.

2.3 Protein expression and Western blot analysis

Isolated clones were grown overnight in basal MRS medium supplemented with either 0.5% glucose or 0.5% lactose and 0.2 M potassium phosphate buffer pH 6.7. Cultures (10 ml) were collected by centrifugation at 4000×g for 10 min and the bacterial pellet was suspended in 1 ml of 100 mM Tris–HCl, pH 8.0. Cell suspensions were transferred to 2-ml tubes containing the same volume of glass beads and lysates were prepared by vigorous shaking in a Bead-Beater (Biospec, Bartlesville, OK, USA) (four cycles of 30 s at maximal speed). Lysates were centrifuged at 15 000×g for 5 min and supernatants were maintained at −20°C for further analysis. For secretion analysis, overnight cultures were centrifuged at 4000×g for 10 min and supernatants were concentrated 10-fold under vacuum or 100-fold by centrifugation in a Millipore Ultrafree-15 filter device (10 kDa molecular mass cut-off), according to the instructions of the manufacturer.

To assay the localization of the secreted PsaA fusion, lactose-induced cells from a 10-ml culture were treated with 5 mg ml−1 lysozyme in 1 ml of Tris–HCl 100 mM pH 7.5 for 1 h at room temperature, cells were centrifuged and supernatants were kept for analysis. In a parallel experiment, cells were treated with 1 mg ml−1 of pronase (Sigma-Aldrich) for 1 h at room temperature, washed three times to remove the protease and treated with lysozyme as above. Protein extracts or supernatants were analyzed by sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS–PAGE) and electrotransferred to nitrocellulose membranes using a Mini Protean II equipment (Bio-Rad, Life Science Research Products, Hercules, CA, USA). Mouse polyclonal anti-PsaA antiserum was developed against recombinant S. pneumoniae PsaA (from strain St 472/96) expressed in E. coli. Mouse polyclonal anti-PspA1 antiserum was developed against purified PspA from S. pneumoniae strain St 435/96 and anti-PspA3 antiserum against the purified protein from St 259/98, according to the choline chloride wash method described by Briles et al. [27]. Horseradish peroxidase-conjugated goat anti-mouse IgG or alkaline-phosphatase-conjugated goat anti-mouse IgG (Sigma, St. Louis, MO, USA) were used according to instructions of the manufacturer. Detection was performed using the chemiluminescent ECL kit (Amersham-Pharmacia Biotech, Little Chalfont, UK) or nitroblue tetrazolium chloride/5-bromo-4-chloro-3-indolyl phosphate toluidine salt reagents from Roche.

3 Results and discussion

3.1 Construction of expression and secretion vectors

Expression vectors used in this work were based on the regulatory properties of the L. casei lactose operon [19]. Previous research on immunization using recombinant lactobacilli has revealed differences in immune response after mucosal administration depending on the different cellular location of the expressed antigen [28]. For that reason, both intracellular and extracellular expression vectors were constructed to produce streptococcal antigens in L. casei. For intracellular expression, DNA fragments encoding mature PsaA (amino acids 21–309), as well as PspA′1 (family 1) and PspA′3 (family 2) [10], were cloned downstream of the lac promoter fused to the first six codons of lacT, so that translation was driven by the lacT ribosomal binding site (Fig. 1). The N-terminal regions of PspA′1 and PspA′3 were chosen to avoid the possible toxic effect of the downstream regions (choline binding region) and because it was shown to be sufficient to induce protective antibodies [10].

In addition, a set of secretion vectors (Fig. 1) was constructed that contained the secretion signal (sp, 34 codons) and pro-sequence (pro, 167 codons) of the L. casei prtP gene [20], downstream of the lac promoter. This should mimic secretion and processing of native PrtP, an efficiently secreted protein in L. casei. The secretion signal would promote translocation of the fusion proteins through the cell membrane, where it would be cleaved by the leader peptidase. Then, at the cell wall, fusion proteins would be processed by the L. casei PrtM, which would cleave PrtP's pro-sequence [29]. Expression of the desired protein can be induced by growing bacteria in lactose and repressed by using glucose as carbon source, thus this system allows the expression control of cloned genes preserving the stability of constructs when working with potentially toxic proteins.

3.2 Intracellular expression of S. pneumoniae antigens in L. casei

L. casei cells were transformed with the pIA-psaA, pIA-pspA′1 or pIA-pspA′3 plasmids. Several clones isolated from each L. casei transformation were grown in basal MRS medium supplemented with either glucose or lactose and a representative of each construct was chosen for further experiments. Overnight cultures were collected and cell lysates were analyzed by SDS–PAGE and Western blot, using specific antisera. Coomassie blue gel staining showed a high expression level of a 37-kDa protein in lysates from L. casei carrying the pIA-psaA construct induced by lactose, but not when the same clone was grown in glucose (Fig. 2A). This protein reached 14.9% of the total protein amount and was specifically recognized by the anti-PsaA antisera in extracts of lactose-grown clones (Fig. 2B, lane 2), but not when the same clones were grown in glucose (Fig. 2B, lane 1) or when extracts from lactose-grown wild-type cells were used (Fig. 2B, lane 3).

Figure 2

Intracellular expression of S. pneumoniae PsaA, PspA′1 and PspA′3 by L. casei. Total cell lysates corresponding to the same number of cells were analyzed by SDS–PAGE and Western blot with specific antibodies. A: Coomassie blue gel staining shows the expression of a 37-kDa protein in lysates from an L. casei [pIA-psaA] clone grown in lactose (lane 2), but not in glucose (lane 1); MW, molecular mass marker. Western blot analysis shows the induction of the three recombinant proteins upon L. casei growth in lactose (lanes 2) but not in glucose (lanes 1). B: pIA-psaA. C: pIA-pspA′1. D: pIA-pspA′3. Wild-type L. casei grown in lactose was used as negative control (B–D, lanes 3). Arrows indicate the S. pneumoniae proteins.

Similarly, expression of the 50-kDa protein reactive with antibodies against PspA′1 (Fig. 2C) and PspA′3 (Fig. 2D) was induced upon growth of the clones in lactose (lanes 2) and its expression was repressed when bacteria were grown in the presence of glucose (lanes 1). The negative control, wild-type L. casei, did not display the corresponding 50-kDa band (lanes 3). A smaller band could be observed in all lanes (Fig. 2C) due to the unspecific reaction of anti-PspA′1 serum against an undetermined L. casei protein. These results show that the lactose promoter from L. casei can be used to efficiently induce the production of streptococcal heterologous proteins. However, expression of PsaA was more efficient than PspA, possibly due to a more compatible codon usage or a better structural stability of this protein in L. casei.

3.3 Secretion of S. pneumoniae proteins by L. casei

Clones isolated from transformations with the secretion constructs (pIAS-psaA, pIAS-pspA′1 and pIAS-pspA′3) were grown in basal MRS medium supplemented with either glucose or lactose. Overnight cultures were centrifuged and the supernatants were concentrated for analysis by Western blot.

PsaA could not be detected in the supernatants of any of the clones analyzed, even after 100-fold concentration in a Millipore Ultrafree-15 column (data not shown). However, a 55-kDa protein corresponding to the unprocessed form (pro::PsaA) as well as the mature PsaA (37 kDa) were detected by Western blot in cell lysates from lactose-grown cells (Fig. 3A, lane 2), but not when cells were grown in glucose (lane 1). These results indicated that the pro::PsaA fusion protein was expressed and might be targeted to the cell wall, where it could be partially processed by the L. casei PrtM, resulting in the 37-kDa PsaA band. However, PsaA was not released to the culture medium, suggesting that it remained attached to the cell surface, as it is in S. pneumoniae. In order to test this hypothesis, cells from an induced culture were incubated with lysozyme, which would release cell wall-anchored proteins, and the supernatants were analyzed by Western blot. As can be observed in Fig. 3B, lysozyme treatment released both 55- and 37-kDa bands to the supernatant (lane 2), which could not be detected in supernatants of cells only treated with protease (lane 4) or when protease was added after lysozyme treatment (lane 5). Treatment with protease, previous to lysozyme, gave a single band smaller than 37 kDa (lane 3), suggesting that a fragment of PsaA may be exposed to the outer side of the cell wall, possibly the N-terminal part of it. None of the bands was observed in supernatants from untreated cells (lane 6). PsaA appeared to be embedded in the cell wall and only partly accessible to protease digestion. PsaA is believed to function as a component of a Mn2+ and Zn2+ transporter in S. pneumoniae. It is presumably a lipoprotein anchored to the pneumococcal surface via an N-terminal cysteine covalently attached to the lipid moiety of the membrane [7]. As this N-terminal anchoring sequence is not present in the cloned PsaA, retention of PsaA to the cell surface of L. casei is most probably due to other interactions with the cell wall.

Figure 3

A: Analysis of PsaA expression from the secretion vector pIAS-psaA. Recombinant L. casei clones were grown in lactose or glucose and analyzed by Western blot. Expression of mature (37 kDa) and non-processed PsaA (55 kDa) is observed when lactose is used in the medium (lane 2). Lysate from glucose-grown bacteria is shown in lane 1. Wild-type L. casei was used as negative control (lane 3). B: Analysis of PsaA localization by Western blot. Lane 1, total lysate of lactose-grown cells expressing PsaA fusion; lane 2, supernatant of the same cells treated with lysozyme; lane 3, supernatant of cells treated with protease followed by lysozyme treatment; lane 4, supernatant of cells treated with lysozyme before the protease treatment (control); lane 5, supernatant of cells treated with protease (control); lane 6, supernatant of untreated cells (control).

On the other hand, when 10-fold concentrated supernatants of L. casei [pIAS-pspA1] were analyzed by Western blot, secretion of PspA1 could be observed under lactose inducing conditions (Fig. 4A, lane 2). However, analysis of total cell extracts of the same clones showed that a detectable amount of unprocessed PspA′1 was accumulated intracellularly in L. casei (Fig. 4B, lane 4). It could be observed that both intracellular and secreted forms migrated to a position slightly higher than would correspond to their theoretical size (sp::pro::PspA′1, 71.4 kDa and pro::PspA′1, 68.0 kDa). Since the PspA′1 mature form of 50 kDa could not be found in the supernatants, it could be deduced that the fusion protein pro::PsPA′1 was not processed by L. casei PrtM. Similarly, only the unprocessed form of PspA′3 was secreted by lactose-grown cells (Fig. 5A, lane 2). When total cell extracts of the same cells were analyzed (Fig. 5B), again, only the unprocessed protein was observed (Fig. 5B, lane 4). As observed before, an unspecific signal could be observed upon overexposure of PspA′3 Western blots. Also, both fused forms of PspA′3 displayed an anomalous migration, as seen in PspA′1.

Figure 4

Secretion of PspA′1 by L. casei [pIAS-pspA′1]. A: Western blot analysis of 10-fold concentrated supernatant from cells grown in glucose (lane 1) and in lactose (lane 2). Supernatants from wild-type L. casei grown in lactose were used as negative control (lane 3). B: Analysis of cell lysates grown in lactose (lane 4) and glucose (lane 3). An L. casei clone expressing the intracellular PspA′1 was used as a control (lane 1, glucose; lane 2, lactose). Proteins reacting with anti-PspA′1 serum are indicated (arrows).

Figure 5

Secretion of PspA′3 by L. casei [pIAS-pspA′3]. A: Western blot analysis of 10-fold concentrated supernatant from cells grown in glucose (lane 1) and in lactose (lane 2). A fusion protein reacting with anti-PspA′3 serum is indicated (arrow). Supernatants from wild-type L. casei grown in lactose were used as negative control (lane 3). B: Analysis of cell lysates grown in lactose (lane 4) and glucose (lane 3). Proteins reacting with anti-PspA′3 serum are indicated (arrows). An L. casei clone expressing the intracellular PspA′3 was used as a control (lane 1, glucose; lane 2, lactose).

To date, a number of lactobacilli expressing different pathogen antigens have been constructed, such us Bacillus anthracis protective antigen, tetanus toxin fragment C (TTFC) and cholera toxin B subunit [1517]. However, satisfactory immune responses and protection have only been achieved in the case of TTFC [30]. The results reported here show that L. casei provides a good system for the expression of antigens from the important pathogen S. pneumoniae, which could be used in the design of oral vaccines. However, its secretion efficiency could yet be improved, perhaps through the screening and use of more efficient secretion signals, or secretion-proficient strains. Theoretically, a vaccination mixture composed of PspAs from families 1 and 2 could be sufficient to cover most S. pneumoniae serotypes. Moreover, a mixture of PsaA and PspA has been shown to induce a better protection against colonization, bacteremia and pulmonary infection [3,4]. Research is currently under way to determine if chimeric proteins accumulated, secreted or attached to the L. casei cell wall could be used to stimulate the production of protective anti-PsaA and anti-PspA antibodies after mucosal vaccination in an animal model.


We wish to thank FAPESP for the grant supporting M.L.S.d.O. This work was supported by FAPESP and by the Commission of the European Communities, specific RTD programme ‘Quality of Life and Management of Living Resources’, QLK1-2000-0146 ‘DEPROHEALTH’. It does not necessarily reflect the Commission's views and in no way anticipates the Commission's future policy in this area.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
View Abstract