Surface Expression of the Conserved C Repeat Region of Streptococcal M6 Protein within the Pip Bacteriophage Receptor of Lactococcus lactis

Abstract

The C repeat region of the M6 protein (M6c) from Streptococcus pyogenes was expressed within the Pip bacteriophage receptor on the surface of Lactococcus lactis. M6c was also detected in the culture medium. The pip-emm6c allele was integrated into the chromosome and stably expressed without antibiotic selection. The level of cell-associated surface expression of PipM6c was 0.015% of total cellular protein. The amount of PipM6c on the cell surface was increased about 17-fold by expressing pip-emm6c from a high-copy-number plasmid. Replacing the native pip promoter with stronger promoters isolated previously from Lactobacillus acidophilus increased surface expression of PipM6c from the high-copy-number plasmid up to 27-fold. Concomitantly, the amount of PipM6c in the medium increased 113-fold. The amount of PipM6c did not vary greatly between exponential- and stationary-phase cultures. Western blots indicated that the full-length PipM6c protein and most of the numerous proteolytic products were found only on the cell surface, whereas only one proteolytic fragment was found in the culture medium.

A recent approach to vaccine development is to stimulate secretion of mucosal immunoglobulins that inhibit colonization of specific bacterial pathogens at the mucosal site of invasion. Because colonization is required to initiate infection in many diseases, blocking of colonization should prevent disease.

Intranasal vaccines against Streptococcus pyogenes, which causes “strep throat,” exemplify this approach. M6 protein is a major surface component and virulence factor of S. pyogenes. Mucosal immunization with the conserved C repeat region of M6 protein (herein referred to as M6c) protects against streptococcal challenge and inhibits pharyngeal colonization by S. pyogenes in mice (3, 4, 11). Moreover, M6-specific secretory immunoglobulin A (IgA) inhibits adherence of S. pyogenes to cultured human pharyngeal cells (12).

Food-grade bacteria are emerging as possible alternatives to attenuated pathogens for the delivery and presentation of heterologous antigens to mucosal immune systems of animals and humans. Lactococcus lactis is a gram-positive bacterium that is used to make fermented dairy foods such as cheese and sour cream, and it has been safely consumed by humans and animals for millennia. L. lactis is designated GRAS (generally recognized as safe) by the Food and Drug Administration and is amenable to molecular biology techniques (14). Many heterologous proteins have been expressed in L. lactis (20, 33, 37, 38, 40, 43, 44), including the full-length M6 protein (32).

L. lactis has been used as a delivery vehicle for mucosal vaccines. It does not colonize, but it survives passage through the gastrointestinal tract (16, 21), which makes it well suited for delivery of antigens to the mucosal immune system (7). Wells et al. have shown that oral or nasal delivery of tetanus toxin fragment C expressed from L. lactis stimulates secretory IgA against tetanus toxin fragment C and protects against a lethal challenge (47). Other antigens expressed from L. lactis and delivered orally or nasally also elicit a mucosal immune response (7, 18, 34).

One of our labs has cloned and sequenced a chromosomal gene (pip) from L. lactis that is required for infection by one species of lactococcal bacteriophage (15). Pip is a membrane protein that serves as a receptor for bacteriophage (27, 42). The physiological function of Pip is unknown, and genetic deletion of Pip causes no phenotypic change in vitro except phage resistance (13, 22).

In this study, M6c was genetically fused to Pip and expressed in L. lactis. Expression was increased by genetic manipulations, and the fused protein was analyzed by Western blotting.

MATERIALS AND METHODS

Bacterial strains, phages, media, growth rates, and plaque assay.

The strains and plasmids used are listed in Table 1. L. lactis subsp. lactis LM2301 and its isogenic pip-emm6c derivative BG301 were grown at 30°C and maintained on M17 medium (41) supplemented with 0.5% glucose (M17G). pGhost6-based plasmids (Appligene, Pleasanton, Calif.) were maintained in lactococcal strains on M17G with 5 μg of erythromycin per ml. Escherichia coli DH5αmcr (Life Technologies, Rockville, Md.) was grown at 37°C in Luria-Bertani (LB) medium (35) with 20 μg of chloramphenicol per ml or 100 μg of ampicillin per ml for maintenance of pRB04 or pUC19-derived vectors, respectively. pGhost6-derived vectors were maintained in DH5αmcr on brain heart infusion (BHI) (Difco/Becton Dickinson Microbiology Systems, Sparks, Md.) with 250 μg of erythromycin per ml and 100 μg of ampicillin per ml. pTRKH2-derived vectors were maintained in E. coli JM110 and DH5αmcr on BHI with 250 μg of erythromycin per ml. E. coli CC118(pRB04) was grown on LB medium supplemented with chloramphenicol (20 μg/ml) or with kanamycin (30 μg/ml) and chloramphenicol (20 μg/ml) after transposition with TnphoA. Streptococcus gordonii GP1223 was grown at 37°C in BHI with 0.5 mg of streptomycin per ml.

TABLE 1.

Strains and plasmids

Strain or plasmid	Phenotype	Reference or source
lactis
LM2301	Wild type pip, plasmid free	45
BG301	LM2301 with single-copy, chromosomal pip-emm6c	This work
S. gordonii GP1223	Expresses M6c on its surface	8
E. coli
DH5αmcr	Cloning strain	Life Technologies
CC118	phoA strain for screening TnphoA fusions	26
JM110	dam	49
Plasmids
pSMB102	Source of emm6	30
pSMB104	Source of emm6 used for cloning	This work
pSA3	Shuttle vector	9
pRB04	pip cloned in shuttle vector pSA3	15
pUC19	Cloning vector	49
pUCpip	pip cloned in pUC19	This work
pUCpipM6c	pip-emm6c cloned in pUC19	This work
pUCΔpipM6c	Deletion in the pip region of pip-emm6c cloned in pUC19	This work
pTRKH2	High-copy-number shuttle vector	31
pPip	pip with pip promoter in pTRKH2	This work
pTRK568	Promoter P6 in pTRKH2	T. R. Klaenhammer (gift)
pBG568	pTRKH568 with unique XbaI site	This work
pPipM6c	pip-emm6c with pip promoter in pTRKH2	This work
pP6pip	pip with P6 promoter in pBG568	This work
pP6pipM6c	pip-emm6c with P6 promoter in pBG568	This work
pP16pip	pip with P16 promoter in pBG568	This work
pP16pipM6c	pip-emm6c with P16 promoter in pBG568	This work
pΔPipM6c	pPipM6c with large deletion in pip	This work
pΔPipM6c-r	Same as pΔPipM6c except reverse orientation of Δpip-emm6c in pTRKH2	This work
pGhost6	Integration vector used for allelic exchange	25 (Appligene, Pleasanton, Calif.)
pGhPipM6c	pip-emm6c cloned in pGhost6	This work
pPLA16	Source of promoter PLA16	1

Lactococcal bacteriophages were prepared from single plaques and plaque assayed as described previously (41). Phages were stored at −70°C in M17G medium containing 20% glycerol.

Genetic constructions.

All recombinant DNA procedures were done as described by Sambrook et al. (35) or Ausubel et al. (2), except where noted. The C repeat region of emm6 (17) from nucleotide 823 through 1131 (numbering begins at the start of translation) was copied by PCR from pSMB104 using primers 5′-GCTTCCGGAAACAAAGTTTCAGAAGCAA-3′ and 5′-CGATCCGGATAGCTCAGCTTTTTCTTTT-3′, which include BspEI sites (shown in boldface) at their 5′ ends. pSMB104 is a derivative of pSMB102 (30) that encodes M6 amino acids 1 to 7 and 222 to 441. PCR was done with Taq polymerase (Promega, Madison, Wis.) in an Idaho Technology (Idaho Falls, Idaho) 1605 air thermocycler according to the manufacturer's instructions, using the following amplification conditions: 5 min at 94°C; then 30 cycles of annealing at 45°C (0 s), elongation at 72°C (10 s), and denaturation at 94°C (0 s); and finally 2 min at 72°C. The time shown in parentheses is the time that the thermocycler dwelled after reaching the indicated temperature and changing to the next temperature. The temperature transition rate between annealing and elongation temperatures was set to 4 to 5 s/°C.

The PCR product was gel isolated, cut with BspEI, and ligated to pUCpip that had been cut with the same enzyme and dephosphorylated. The product was named pUCpipM6c. pUCpip was constructed by ligating the 5-kb XbaI fragment of pRB04 (15) to pUC19. The product of the ligation reaction of pUCpipM6c was transformed into E. coli DH5αmcr (Life Technologies). Transformants were selected on LB plates containing ampicillin (50 μg/ml). pUCpipM6c was cut with XbaI, and the pip-emm6c fragment was ligated to pTRKH2 (31) or pGhost6, which had been cut with XbaI. The constructs were named pPipM6c and pGhPipM6c, respectively. The wild-type pip allele, including its apparent promoter, was subcloned into pTRKH2 by ligating the 5-kb XbaI fragment of pRB04 with pTRKH2 that had been cut with XbaI. The resulting plasmid was named pPip. pPip, pPipM6c, and pGhPipM6c were transformed first into DH5αmcr and then into L. lactis 2301.

Δpip-emm6c was constructed by cutting pUCpipM6c with BsmI and isolating the 6.6-kb fragment from an agarose electrophoretic gel. A double-stranded DNA adaptor was prepared using the following synthetic oligonucleotides: 5′-ATCTGG-3′ and 5′-AGATCT-3′. The adaptor was phosphorylated and ligated to the 6.6-kb fragment to restore the correct reading frame to pip-emm6c. The resulting plasmid (pUCΔpipM6c) was transformed into DH5αmcr, and transformants were selected on LBamp. pUCΔpipM6c was cut with XbaI, and the Δpip-emm6c fragment was ligated to pTRKH2 that had been cut with XbaI. Products with either orientation were isolated. The resulting plasmid with Δpip-emm6c in the same orientation as lacZ was named pΔPipM6c. The plasmid with Δpip-emm6c in the reverse orientation as lacZ was named pΔPipM6c-r.

Replacement of the pip promoter with P6 or P16.

The XbaI site between the 3′ end of the P6 promoter and the multiple cloning region was eliminated by restricting with XbaI pTRK568 that had been isolated from E. coli DH5αmcr (dam⁺), treating with Klenow reagent, ligating, and transforming into E. coli JM110 (dam). The resulting construct included a unique XbaI site in the multiple cloning region and was named pBG568. The coding region of pip without its promoter was copied by PCR and cloned into the unique XbaI site of pBG568, creating pP6pip. The fidelity of the PCR-copied pip was tested and verified by restoring phage sensitivity to the phage-resistant strain L. lactis JK101 (13). emm6c was inserted into pP6pip at the unique BspEI site as described above, resulting in pP6pipM6c.

Promoter P6 in pP6pipM6c and pP6pip was replaced by promoter P16 (1). The ≈460-bp SmaI-PstI fragment containing promoter P6 was removed and replaced with the ≈275-bp SmaI-PstI fragment from pPLA16 (1). The resulting plasmids were named pP16pipM6c and pP16pip.

Chromosomal allele replacement.

L. lactis BG301 was constructed by replacing the chromosomal pip in LM2301 with pip-emm6c using pGhPipM6c as described previously (13). Integrants were selected at 37°C. One hundred single colonies were scored for sensitivity to erythromycin and phages c2 and sk1 as described previously (13).

Analytical PCR.

Chromosomal allele replacement of pip with pip-emm6c was verified by analytical PCR as described previously (13) with either of the following primer pairs: 5′-AGTTGTTTGCCCTAGACTGGAAACG-3′ and 5′-ATCCTTTGGTAAGTAAATTCC-3′ or 5′-TGCAAGTAATGGTGAAGATTTAACAAATGAA-3′ and 5′-GAATGCCCGATTTGATTTCAGGCCA-3′. Both PCR products spanned the BspEI site. The PCR products were purified (QIAquick PCR purification kit from Qiagen, Inc.) and sequenced as described below.

Analytical PCR analysis of the ampicillin and erythromycin resistance genes and flanking sequences of pGhost6 was done as described previously (13).

Membrane protein topology mapping.

The method of Manoil (26) was used to map the membrane topology of Pip. E. coli CC118 was transformed with pRB04 (15) and infected with λTnphoA. Plasmid was prepared from pooled, dark blue colonies on indicator plates and used to transform CC118. Plasmids from dark blue transformants on indicator plates were analyzed by restriction using XbaI, AatII, and AflIII. Fusions within pip were sequenced in both directions with phoA (26) and pip (15) sequencing primers.

Immunoblot analysis.

One milliliter of each culture at an optical density at 600 nm (OD₆₀₀) of ca. 0.4 was centrifuged at 12,000 × g for 2 min in a 1.5-ml Microfuge tube. For experiments using stationary-phase cultures, the cultures were harvested after 15 h of incubation, and the OD₆₀₀ was approximately 1.2. The supernatant was saved, and the cell pellet was washed twice with 20 mM Tris-HCl–500 mM NaCl, pH 7.5 (TBS). The washed cells were resuspended in a volume of TBS that adjusted for small differences in the OD at time of harvest and were serially diluted 1:2 or 1:4 in TBS. Likewise, the supernatants were diluted to adjust for small differences at time of harvest. An equivalent volume (0.7 ml) of each dilution and cell-free culture supernatant was vacuum filtered through nitrocellulose membranes (Schleicher and Schuell, Keene, N.H.) using a slot blot filtration manifold (Hoefer Pharmacia, San Francisco, Calif.). Each slot was washed three times with 1 ml of TBS. The membrane was removed from the manifold, blocked for 30 min with TBS containing 3% gelatin, washed for 30 min with TBS containing 0.05% Tween 20 (TTBS), and probed for 15 h with anti-M6c 10F5 monoclonal antibody (19) in TTBS containing 1% gelatin. The membrane was washed three times in TTBS for 5 min, incubated with goat anti-mouse IgG conjugated to alkaline phosphatase (Bio-Rad Laboratories, Hercules, Calif.) for 1 h, and then washed twice in TTBS and once in TBS and developed with alkaline phosphatase substrate (Bio-Rad). Developed slots were quantified by either densitometry or visual comparison with a 1:2 or 1:4 serial dilution of purified M6 protein (generously provided by Kevin Jones, Siga Pharmaceuticals, Corvallis, Oreg.) that was applied to every gel.

Western blotting.

Cells and cell-free culture medium were prepared as described previously (24), except that a mixture of the following protease inhibitors at the indicated final concentrations was added to the washed cells and cell-free medium before addition of trichloroacetic acid: phenylmethylsulfonyl fluoride (35 μg/ml), aprotinin (4 μg/ml), chymostatin (37 μg/ml), leupeptin (700 ng/ml), soybean trypsin inhibitor (1 μg/ml), EDTA (500 ng/ml), benzamidine (35 μg/ml), and acetamide (35 μg/ml). All inhibitors were purchased from Sigma Chemicals (St. Louis, Mo.), U.S. Biochemicals (Cleveland, Ohio), or Boehringer Mannheim (Indianapolis, Ind.). Samples were analyzed on a 10 or 12% discontinuous electrophoretic gel (23), transferred to a nitrocellulose membrane (2), and developed as described above for the immunoblot. Prestained high-molecular-weight standards (Sigma) and horse heart cytochrome c (Sigma) were used as size markers.

DNA sequencing.

DNA was sequenced at Oregon State University Center for Gene Research and Biotechnology using the dideoxy method (36) on an Applied Biosystems 373A DNA sequencer.

RESULTS

Splicing site for M6c within Pip.

An analysis of the deduced amino acid sequence of Pip predicts that the N-terminal two-thirds of the protein would face the external side of the plasma membrane (15). This region might be an ideal site into which M6c could be spliced, so that it would be exposed to the outside of the cell. However, the actual structure of Pip and its topology relative to the plasma membrane have never been tested.

Therefore, the molecular topology of Pip was analyzed by the phoA method of membrane protein mapping. Random genetic fusions between pip and phoA were constructed and expressed in E. coli. Fusion junctions from 10 pip-phoA isolates with high levels of alkaline phosphatase activity were mapped by restriction analysis and sequenced. All but three of the junctions were within the sequence of pip predicted to encode the large, hydrophilic, extracellular region of Pip (Fig. 1). Three junctions (V⁷¹⁴, S⁷¹⁷, and G⁷⁵¹) were found within hydrophobic regions predicted to span the plasma membrane. These data confirm the external topology of the region of Pip encoded by bp 50 to 700 and suggest that this portion of Pip is a suitable site for the introduction of heterologous protein sequences.

FIG. 1.

Diagram of membrane protein topology analysis. The membrane topology of Pip was analyzed by the alkaline phosphatase fusion method described in Materials and Methods. Pip is represented as a line. Fusion junctions from colonies that expressed high levels of alkaline phosphatase were mapped to the amino acid residues indicated by arrowheads. The residue code and number are shown below the corresponding arrowhead. The shaded boxes above the line indicate regions of hydrophobicity that are potential membrane-spanning segments. The diagram is drawn to scale.

Construction of L. lactis with pip-emm6c.

The coding region of M6c (emm6c) was copied by PCR and inserted in frame after bp 321 in pip (Fig. 2A and B). The genetic fusion (pip-emm6c) included the putative native promoter and transcriptional terminator of pip.

FIG. 2.

Diagram of fusion allele constructs. (A) Unmodified pip. (B) pip-emm6c. Three hundred twelve base pairs of the C repeat region of emm6 (shaded box) was inserted into the unique BspEI site of pip at nucleotide 324. (C) The Δpip-emm6c allele. The coding region of pip between the two BsmI sites was deleted (dashed lines). An adaptor (stippled box) was inserted between the BsmI sites to maintain the correct reading frame. The promoter (promo) is shown as an open arrow.

L. lactis LM2301 was transformed with pip-emm6c, and its single-copy, wild-type pip allele was replaced with the recombinant allele. The allele-exchanged strain (BG301) was isolated by selection on growth plates that contained phage c2 (data not shown). Analytical PCR using primers that flank the inserted emm6c amplified one band of ≈650 bp from chromosomal DNA of BG301 and one band of ≈330 bp from chromosomal DNA of LM2301 (data not shown). The predicted sizes are 653 and 331 bp, respectively. DNA sequencing of the PCR products confirmed the exchange of alleles (data not shown). Neither the ermAM nor the bla marker of the recombinant plasmid was detected by analytical PCR using DNA from BG301, although control reactions with the same primers and recombinant plasmid yielded the predicted products (data not shown).

M6c expression in L. lactis BG301.

M6c was expressed in exponential-phase cultures of L. lactis BG301. Equivalent amounts of washed cells and cell-free growth medium from the same cultures were analyzed by immunoblotting using a monoclonal antibody against M6c. The amount of M6c exposed to the surface of BG301 was 0.014 μg/ml/OD₆₀₀ unit, which is equivalent to 0.007 μg/10⁸ cells or 0.015% of total cellular protein (Table 2). M6c was below the limit of detection in the culture medium. There was no detectable expression of M6c from control cultures of LM2301.

TABLE 2.

Expression of M6c

Strain	Mean M6c level (μg/OD unit/ml) ± SD (n)
	Cell surface	Medium
LM2301	0a (5)	0a (5)
BG301	0.014 ± 0.007 (3)	0a (3)
LM2301(pPipM6c)	0.24 ± 0.1 (5)	1.0 ± 0.14 (2)
LM2301(pPipM6c)b	0.23 ± 0.05 (3)	1.4 ± 2.3 (3)
LM2301(pΔPipM6c)	0.21 ± 0.06 (2)	0.26 ± 0.02 (2)
LM2301(pΔPipM6c-r)	0.047 ± 0.05 (2)	0.031 ± 0.011 (2)
LM2301(pP6pipM6c)	3.3 ± 0.2 (2)	56.7 ± 2.6 (2)
LM2301(pP16pipM6c)	6.6 ± 0.4 (2)	113 ± 5 (2)
GP1223	12 ± 2.6 (6)	0.28 ± 0.10 (4)

^aBelow the limit of detection, which was about 0.002 μg/OD unit/ml. 

^bStationary phase. 

M6c expression from a multicopy plasmid.

The pip-emm6c allele was subcloned to a high-copy-number vector (pPipM6c) and used to transform L. lactis LM2301. Surface expression of PipM6c from LM2301(pPipM6c) was 17-fold higher than that from BG301 (Table 2). The amount of PipM6c in the growth medium of LM2301(pPipM6c) was 1.0 μg/ml/OD unit, which is fourfold higher than the amount on the cell surface. No M6c was detected from a control that expressed the wild-type pip allele from the same plasmid.

M6c expression from S. gordonii.

M6c was measured on the surface and in the cell-free culture medium of S. gordonii GP1223, which is genetically engineered to express the C repeat region of M6 on its surface (8). Cells were harvested from cultures at exponential phase. Compared to LM2301(pPipM6c), S. gordonii GP1223 expressed 50 times more antigen on its cell surface (Table 2). However, M6c in the culture medium of GP1223 was only 28% of that in the medium of LM2301(pPipM6c) (Table 2).

Effect of growth phase on expression of M6c.

LM2301(pPipM6c) was grown to stationary phase, and the expression of M6c was measured. The amounts of M6c on the cells and in the culture medium were nearly the same as those in the equivalent fractions from exponential-phase cultures when normalized for differences in cell densities (Table 2).

M6c expression from deletion allele of pip.

Detection of M6c within the context of the chimeric protein may be limited by inaccessibility of the monoclonal antibody to its epitope. If the amount of Pip in the chimera was significantly reduced, the protein conformation might change, which might increase accessibility of the antibody to the M6c epitope. With this in mind, 451 aminoacyl residues of Pip adjacent to the M6c epitope were genetically deleted from PipM6c (Fig. 2C).

The deletion allele (Δpip-emm6c) was expressed in L. lactis from the same multicopy plasmid used to construct pPipM6c. Immunoblot analysis of the resulting strain, LM2301(pΔPipM6c), showed that the amount of M6c on the cell surface was similar to that on LM2301(pPipM6c) (Table 2). The cell-free culture medium of LM2301(pΔPipM6c) contained 26% of that in the medium of LM2301(pPipM6c).

The Δpip-emm6c allele was subcloned in the opposite orientation on the same plasmid and expressed in L. lactis. Cultures of the resulting strain, LM2301(pΔPipM6c-r), expressed about 22 and 12% of the amounts of PipM6c on the cell surface and in the cell-free medium, respectively, expressed by LM2301(pΔPipM6c) (Table 2).

Replacement of pip promoter.

The level of expression of PipM6c was increased by replacing the pip promoter with promoters P6 and P16, which were previously isolated from Lactobacillus (1, 10). In L. lactis, the expression of the cat-86 gene was about 1.5-fold higher from P16 than from P6 (1, 10). Immunoblot analysis showed that in L. lactis(pP6pipM6c), P6 increased surface expression of PipM6c about 14-fold compared to that from the native pip promoter (Table 2). P16 increased cell surface expression of PipM6c another twofold above that from P6. The level of surface expression of PipM6c from promoter P16 in L. lactis was approximately half of that from S. gordonii GP1223.

The greatest increase in expression of M6c occurred in the culture medium. When expressed from promoters P6 and P16, the amounts of M6c in the culture medium increased 57- and 113-fold, respectively (Table 2). This was 201- and 403-fold higher, respectively, than the levels expressed in the culture media from S. gordonii GP1223.

M6c was not detected on the surface or in the medium of cultures of LM2301(pP6pip) or LM2301(pP16pip).

Molecular size of M6c on cell surface and in culture medium.

Washed cells and cell-free culture media were analyzed by Western blotting with anti-M6c monoclonal antibody. The results (Fig. 3) showed seven intensely stained bands associated with the cell, which had mean molecular masses (± standard deviations) of 111 (±2), 106 (±4), 100 (±8), 52 (±1), 49 (±1), 32 (±0), and 29 (±1) kDa. This agrees closely with the hypothetical size of the chimeric protein (106 kDa), considering that coiled-coil proteins like M6 and Pip (B. Geller, unpublished analysis) are notorious for running at aberrant sizes on denaturing gels (17). The relative intensities of the three largest bands varied significantly among blots, and often the top band was most intense. There were numerous bands of lesser intensity that varied in number depending on the amount of time that the blot was left in the chromagenic reagent, suggesting that PipM6c may have been proteolytically degraded during sample preparation, despite the inclusion of eight protease inhibitors. It appears unlikely that the fragments were produced in vivo, because many of the fragments are too small to include both M6c and the cell membrane anchoring region (15) at the carboxy terminus.

FIG. 3.

Western blotting. Exponential-phase cultures were separated into cells (lanes C) and cell-free culture medium (lanes M). The cells and medium were prepared as described in Materials and Methods. The amount of medium applied was equivalent to one-fourth of the amount of the cells. The blot was probed with monoclonal antibody against M6c. Lanes 1 and 2, LM2301(pPip); lanes 4 and 5, LM2301(pPipM6c); lanes 3 and 6, size standards as indicated to the right in kilodaltons.

Only one band was detected in the culture medium, and its apparent size was 79 (±2) kDa. The patterns from LM2301(pPipM6c) and LM2301(pP16pipM6c) were qualitatively the same (data not shown).

DISCUSSION

We have shown that Pip can be used to direct surface expression of a heterologous protein in L. lactis. The results of the pip-phoA fusions suggested that the region within Pip between residues 64 and 684 is external to the plasma membrane. This guided our selection of the site at residue 108 for fusing Pip to M6c. Indeed, PipM6c was detected on the surface of intact cells, which shows that the region of Pip at residue 108 is outside the plasma membrane. Further analysis, including lacZ fusions, will be necessary to map the topology of Pip in more detail.

Insertion of M6c disrupted the phage receptor activity of Pip. This phenotypic change was exploited in this study to select strains that had undergone pip-emm6c allelic exchange. A rapid, direct selection for exchange of pip alleles has been previously demonstrated using a pUC-based integration vector that lacks a gram-positive origin of replication but encodes a selectable marker for L. lactis (22). The pip locus may be an ideal location for inserting recombinant alleles into the chromosome of L. lactis, because pip is not required for viability or rapid growth in vitro (13), and allelic replacement can be selected directly on plates that contain phage.

Expression of cell surface-associated M6c was increased 17-fold by moving pip-emm6c from the chromosome to a multicopy plasmid. Replacing the native promoter increased total (cell-associated plus cell-free medium) expression of PipM6c up to about 100-fold, of which cell-associated PipM6c was about 7% of total cellular protein. Most of the increase was found in the cell-free culture medium. The reason for the appearance of much of the additional PipM6c in the cell-free culture medium is unknown. One possibility is that overexpression of PipM6c exceeded the capacity of the cell to sort or attach PipM6c to the surface of the cell, resulting in release of the excess to the medium. Another possibility is that proteolytic enzymes or the acidic conditions of the lactococcal culture released the protein from the cell surface. In any case, finding a 79-kDa fragment of PipM6c in the culture medium shows that a large fragment of PipM6c is not anchored to the cell.

The amount of antigen delivered by L. lactis to the mucosal immune system may be limited to that expressed in vitro before vaccination (46). This is because L. lactis does not colonize the gastrointestinal tract (16, 21), and its metabolic activity in the gastrointestinal tract is not required for effectiveness as a mucosal vaccine (29, 34). The relatively high level (7% of total cellular protein) of surface-expressed antigen (M6c) that we report here is similar to levels of expression reported for other antigens in effective lactococcal vaccines (28, 29, 34, 46, 47) and may be sufficient for eliciting a mucosal immune response.

Surface expression of immunogens may be the most effective way of presenting lactococcus-based vaccines to the immune system (28, 46). Particulate-associated proteins evoke a stronger response and are less likely to produce tolerance in the mucosal immune system than soluble proteins (5, 6, 48). One report showed that subcutaneous injection of tetanus toxin fragment C elicited a stronger systemic immune response when attached to the surface of L. lactis than an equivalent dose of the intracellular antigen (28).

Another factor that may be important for lactococcus-based vaccines is access of surface-attached antigens to the immune system. One report has shown that a heterologous antigen can be attached to the surface of L. lactis and is accessible on intact cells (39). Another report showed that although surface-attached tetanus toxin fragment C was immunogenic, it was not accessible on intact lactococcal cells by using immunolabeling techniques (28). Clearly, we have shown in this report that M6c is accessible on the surface of intact L. lactis. It remains to be learned if this accessibility translates to higher immunogenicity or if this could create a liability for exposure to proteases or stomach acid.