Construction and Immunological Evaluation of Dual Cell Surface Display of HIV-1 Gag and Salmonella enterica Serovar Typhimurium FliC in Lactobacillus acidophilus for Vaccine Delivery

Abstract

Oral vaccines that elicit a mucosal immune response may be effective against human immunodeficiency virus type 1 (HIV-1) because its transmission occurs mainly at the mucosa. The aim of this study was to construct recombinant Lactobacillus for oral delivery of oral vaccines against HIV-1 and to evaluate their immunogenicity. A recombinant Lactobacillus acidophilus strain expressing the HIV-1 Gag on the bacterial cell surface was established by fusion with the signal peptide and anchor motif of a mucus binding protein (Mub) from L. acidophilus with or without coexpression of Salmonella enterica serovar Typhimurium flagellin (FliC) fused to a different Mub signal peptide and anchor. Using HEK293 cells engineered to express Toll-like receptor 5 (TLR5), the biological activity of FliC on the bacterial cell surfaces was determined. The surface-exposed flagellin retained its TLR5-stimulating activity, suggesting that the recombinant strain with Gag and FliC dual display might provide a different immunopotency than the strain expressing only Gag. The immunological properties of the recombinant strains were assessed by coculture with human myeloid dendritic cells (DCs). The heterologous antigens on the cell surface affected maturation and cytokine responses of DCs. Acquired immune responses were also investigated by intragastric immunization of mice. The enzyme-linked immunosorbent spot assay showed induction of gamma interferon-producing cells at local mucosa after immunization of mice with the Gag-producing strain. Meanwhile, the immunization with L. acidophilus displaying both Gag and FliC resulted in an increase of Gag-specific IgA-secreting cells. These results suggested that the Gag-displaying L. acidophilus elicited specific immune responses and the coexistence of FliC conferred an adjuvant effect on local IgA production.

INTRODUCTION

The global epidemic of human immunodeficiency virus (HIV) infection has imposed a heavy burden on our society. In spite of intensive research, no fully effective vaccine has been discovered. The first and only successful phase 3 human trial using a combinational HIV vaccine was reported in 2009; however, the efficacy remains limited (38). Tremendous efforts on the vaccine development have been continuously made by many researchers using various approaches (12). Transmission of HIV usually occurs at mucosal surfaces, especially the rectal and genital mucosae, and mucosa-associated lymphoid tissues are primary sites for virus replication. Mucosal vaccines are potentially capable of stimulating the mucosal immune system and inducing local and systemic immune responses. Hence, mucosal vaccines that elicit HIV-specific immune responses at the local mucosa may be more effective than conventional parenteral vaccines, which usually provide only systemic immunity.

Lactic acid bacteria are potential delivery vehicles for mucosal immunization because they provide immunopotency and safety aspects, which are essential for practical vaccines. Many previous studies have demonstrated that commensal microbes carrying antigens can confer protective immunity (39, 40). For example, recombinant Lactococcus lactis expressing the HIV type 1 (HIV-1) V2-V4 loop of Env induced antigen-specific immune responses that protected mice from challenge with recombinant vaccinia virus carrying the HIV env gene (41). The usefulness of L. lactis as an HIV vaccine, however, remains unclear because cholera toxin, which is a strong but toxic mucosal adjuvant, was required for protection. It is likely that vaccines based on lactic acid bacteria in combination with adjuvant molecules could offer better efficacy (5, 23, 34). In this regard, coexpression of safe adjuvant molecules may be needed for effective immunization when using lactic acid bacteria as delivery vehicles.

Flagellin, the major subcomponent of bacterial flagella, is the ligand of Toll-like receptor 5 (TLR5) and is also recognized by one of the cytosolic nucleotide binding oligomerization domain-like receptors (NLRs), NLRC4/IPAF. The flagellar antigen initiates innate immune responses by activating NF-κB via TLR5 and the caspase-1 inflammasome through NLRC4 (11, 15, 26, 27). Since antigens fused with flagellin have been shown to enhance immune responses, the use of flagellins as adjuvants is under consideration (4, 16, 17, 25). Lactobacillus strains are Gram-positive bacteria, which intrinsically present microbe-associated molecular patterns (MAMPs) such as lipoproteins and lipoteichoic acid (LTA) that are recognized by TLR2 and TLR2/6 (2, 6, 20, 24). Meanwhile, most lactobacilli, including Lactobacillus acidophilus, lack TLR5-stimulating activity due to the absence of flagella. In our previous study, two types of recombinant L. acidophilus displaying Salmonella enterica serovar Typhimurium FliC were engineered and demonstrated to acquire TLR5-stimulating activity (19). Hypothetically, these strains may provide better immunopotency than the wild-type L. acidophilus strain for vaccine delivery.

In this study, genetically modified L. acidophilus strains displaying Gag from HIV-1 with or without coexpression of FliC were constructed. Subsequently, the immunogenicity of these strains was evaluated using a murine model. The aim of this study was to determine whether or not lactobacilli can serve as a vaccine vector to elicit acquired immune responses against HIV-1.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

Lactobacillus acidophilus NCFM (NCK56) and derivative recombinant strains were grown statically in MRS broth (Difco Laboratories, Detroit, MI) supplemented with or without 5 μg/ml of erythromycin (Em) at 37°C. MRS agar (1.5% agar) plates were incubated anaerobically. Escherichia coli DH5α and other recombinant strains were grown aerobically with shaking in LB medium (Difco) with or without 200 μg/ml of Em at 37°C. All bacterial strains used in this study are shown in Table 1.

Table 1.

Bacterial strains and plasmids

Bacterial strain	Plasmid	Description of bacterial strain and/or plasmid	Reference or source
E. coli DH5α		Cloning host	14
NCK2149	pTRK1029	pTRK882 backbone, unit 3 inserted	19
NCK2151	pTRK1030	pTRK1029 backbone, Ppgm replaced with unit 1	This study
NCK2161	pTRK1035	pTRK1030 backbone, unit 5 inserted	This study
NCK2165	pTRK1037	pTRK1035 backbone, S1392-FliC-A1709 removed	This study
L. acidophilus NCFM		Human intestinal isolate, expression host	3
NCK1895	pTRK882	Erythromycin resistance	10
NCK2152	pTRK1030	Cell surface display of FliC	This study
NCK2162	pTRK1035	Cell surface display of Gag and FliC	This study
NCK2166	pTRK1037	Cell surface display of Gag	This study

Construction of plasmids.

Regular and overlap PCRs were performed using the primers shown in Table 2. The design of an expression cassette is described in Fig. 1. Each subunit was generated by PCR and then connected by overlap PCR to amplify units 1, 3, and 5. First, the unit 3 fragment was cloned into pTRK882 at the EcoRI-HindIII site (pTRK1029). Second, the promoter region of pTRK882 was replaced with unit 1 at the NcoI-EcoRI site (pTRK1030). Third, unit 5 was inserted between units 1 and 3 at the NheI-EagI site (pTRK1035). Finally, pTRK1035 was digested with PstI and self-ligated to remove the open reading frame of fliC (pTRK1037). These plasmids were introduced into L. acidophilus NCFM by electroporation (19).

Table 2.

Primers used in this study

Primer	Sequencea	Restriction site(s)
AK5	ACACAGATCTCCATGGTGCGACAAGTAATAAACTAAAC	BglII, NcoI
AK6	GGATGAAACATTGAATTTAAAAGGAGAAGAATTATGGATAAG
AK11	CGTCCTCTCTTTACTGCGTGGTACCGTGCCAACAGTTACTCCAAC	KpnI
AK12	ACACCTCGAGAAGCTTCTGCAGTTATCTCCGTTTCCTATCAG	PstI, HindIII
AK13	GTTGGAGTAACTGTTGGCACGGTACCACGCAGTAAAGAGAGGACG	KpnI
AK18	CTTATCCATAATTCTTCTCCTTTTAAATTCAATGTTTCATCC
AK21	ACACGAATTCCGGCCGATGGCACAAGTCATTAATAC	EcoRI, EagI
AK22	ACACGAATTCGCTAGCTGGCTCTACACTATCTGCTTG	EcoRI, NheI
AK40	GGTAACGATCCCAGCTCACAAGGATCCACAGTAACAGTAACTTACACACC	BamHI
AK41	GGTGTGTAAGTTACTGTTACTGTGGATCCTTGTGAGCTGGGATCGTTACC	BamHI
AK42	CAAGAAGCGTAGAAAAGAAGACAAATAACTGCAGAAAGTGATTATTAAGACTTAAGGGGC	PstI
AK43	GCCCCTTAAGTCTTAATAATCACTTTCTGCAGTTATTTGTCTTCTTTTCTACGCTTCTTG	PstI
AK44	ATGCCTCGAGCGGCCGGTCACTGTTATCAATATTCTCAGC	XhoI, EagI
AK47	ACACAGATCTGCTAGCATGGACAAATTAAAACACATTGTATGGGC	BglII, NheI

^aRestriction sites are underlined.

Fig 1.

Gene map of an expression cassette for HIV Gag and FliC dual-display system. Subunits 1a to 3b are products of the first PCR. Units 1 to 3 represent fragments generated by overlap PCR connecting the subunits. Arrows represent positions of primers. Unique (except for PstI) restriction sites are shown. P_pgm, promoter region of pgm gene from L. acidophilus NCFM; R, region encoding ribosome binding site of Mub from L. acidophilus NCFM (LBA1709 or LBA1392); S₁₇₀₉/S₁₃₉₂, region encoding the signal sequence of Mub (LBA1709 or LBA1392); gag, gag gene from HIV-1; A₁₇₀₉/A₁₃₉₂, region encoding the anchor region of Mub (LBA1709 or LBA1392); fliC, fliC gene from S. enterica serovar Typhimurium.

Western blotting.

Detection of cell wall-associated proteins via Western blotting was described previously (19). In brief, bacterial cells collected from exponential-phase cultures were treated with mutanolysin and homogenized by bead beating. Protease inhibitor cocktail for bacterial cells (Sigma-Aldrich, St. Louis, MO) was supplemented to all buffers used in the preparation. The cell extracts were applied to SDS-polyacrylamide gels using 7.5% TGX precast gel (Bio-Rad, Hercules, CA) and then transferred to a polyvinylidene difluoride membrane by semidry blotting. Precision Plus Protein WesternC protein standard (Bio-Rad) was also loaded as a standard marker. The blot was incubated with a monoclonal anti-p24 IgG antibody (clone 183-H12-5C; from Bruce Chesebro and Kathy Wehrly, NIH AIDS Research and Reference Reagent Program) and horseradish peroxidase (HRP)-conjugated anti-mouse IgG (Cell Signaling Technology, Danvers, MA). Chemiluminescence using Luminata Forte Western HRP substrate (Millipore) was visualized by use of a ChemiDoc XRS+ system (Bio-Rad).

TLR5 stimulation assay.

The reporter assay using TLR5-expressing HEK293 cells was described previously (35). Freshly prepared bacteria were added to the culture of TLR5-expressing cells and incubated for 48 h. TLR5 activation was determined by NF-κB induction using the secreted embryonic alkaline phosphatase (SEAP) pNiFty reporter plasmid (InvivoGen, San Diego, CA). SEAP activity in the coculture supernatants was detected by luminometer after development with the SEAP reporter gene assay per the manufacturer's instructions (Roche Applied Science, Indianapolis, IN).

Isolation of human myeloid DCs and coculture assay.

Untested human blood buffy coats were obtained from the American Red Cross (Durham, NC). Human myeloid dendritic cells (DCs) were isolated in accordance with the method described previously (35). In brief, lymphocytes were separated from whole blood using Ficoll-Plaque Plus (GE Healthcare) and centrifugation. Enrichment of CD1c⁺ cells was performed with a magnetically activated cell sorting CD1c (BDCA-1)⁺ dendritic cell isolation kit (Miltenyi Biotec, Auburn, CA). The isolated DCs suspended in filtered human peripheral blood mononuclear cell (PBMC) medium (10% fetal bovine serum [FBS], 1.1% sodium pyruvate, 1% GlutaMAX, 100 U/ml penicillin, 100 μg/ml streptomycin) were plated in 96-well plates at a concentration of 2 × 10⁵ cells/well. Fresh bacterial cells at stationary phase were washed with phosphate-buffered saline (PBS), suspended in the same PBMC medium, and then inoculated into the DC cultures at a concentration of 2 × 10⁶ CFU/well (DC/bacterial cell ratio, 1:10). After 24 h of incubation, DCs were centrifuged and collected for antibody staining, while supernatants were stored at −80°C for cytokine analysis.

Flow cytometric analysis.

For analysis of cell surface antigens on lactobacilli, an overnight culture of the bacteria at stationary phase was collected and washed with PBS. The cells were incubated with RY544 antibody (19) and human anti-HIV-1 polyclonal Ig (catalog number 3957, NIH AIDS Research and Reference Reagent Program, NABI, and NHLBI), followed by treatment with Alexa 488-conjugated anti-human IgG (Invitrogen, CA) and phycoerythrin (PE)-conjugated anti-rabbit IgG (Jackson ImmunoResearch Laboratories, PA). In order to stain human DCs, fluorescein isothiocyanate-conjugated anti-human TLR5, PE-Cy7-conjugated anti-human CD80, PE-conjugated anti-human CD86, peridinin chlorophyll protein (PerCP)-conjugated anti-human CD83, and PerCP-conjugated anti-human CD40 (BioLegend, CA) suspended in PBS were used. Flow cytometry was performed using an LSRII apparatus and analyzed with FACS Diva software. Detailed procedures have been described previously (19).

Multiple cytokine quantification.

Multiple cytokines in culture supernatants of DCs were detected with a customized Milliplex Map kit (Millipore, Billerica, MA), in accordance with the manufacturer's instructions. In brief, antibody-immobilized beads for detection of interleukin-1β (IL-1β), IL-6, IL-8, IL-10, IL-12 (p70), tumor necrosis factor alpha (TNF-α), and alpha 2 interferon (IFN-α2) were incubated with 5-times-diluted culture supernatants of DCs. The beads were subsequently treated with detection antibodies as well as PE-conjugated streptavidin and analyzed by use of a Bio-Plex system (Bio-Rad).

Immunization of mice.

Six 9-week-old female BALB/c mice (NCI, Raleigh, NC) per group were immunized intragastrically with recombinant lactobacilli. Bacterial cells were collected from an overnight culture and suspended in immunization buffer (0.1 M NaHCO₃, 20 mg/ml soybean trypsin inhibitor, saline) in order to protect cell surface antigens (19). Three daily doses of the bacterial suspensions (2 × 10⁹ CFU each) were administered to mice on weeks 0, 2, and 4. On week 6, mice were euthanized. Spleen, Peyer's patches, colon, and female reproductive tract were collected and lymphocytes were prepared. Single-cell suspensions of spleen and Peyer's patch cells were prepared by mashing the tissues with the plunger end of syringes and filtering through cell strainers. Lymphocytes from the colon and female reproductive tract were isolated in accordance with the procedure described previously (35). The care and use of experimental animals complied with the guidelines of North Carolina State University.

ELISpot assay.

Ninety-six-well MultiScreen-IP filter plates (Millipore) were pretreated with 70% ethanol and washed three times with 200 μl sterile distilled water, before being coated with 10 μg/ml anti-mouse IgA (Mabtech, Mariemont, OH). The plates were incubated overnight at 4°C. Before addition of cells, plates were washed five times in sterile PBS and blocked with cytotoxic T lymphocyte (CTL) medium (35) at 37°C for 2 h. Cells were added in duplicate to the wells at a concentration of 2.5 × 10⁵ per well, and then plates were incubated for 16 h for total IgA and 20 h for Gag (p24)-specific IgA at 37°C. To detect spots, plates were initially washed in PBS containing 0.05% Tween 20 (PBST) to remove cells. For the total IgA enzyme-linked immunosorbent spot (ELISpot) assay, 1 μg/ml biotinylated polyclonal goat anti-mouse IgA (Mabtech) was added in blocking buffer (PBS containing 1% FBS). For the p24-specific IgA, 0.5 μg/ml biotinylated p24 (Prospec, East Brunswick, NJ) diluted in blocking buffer was added. Biotinylated antibodies were incubated in the plates for 2 h at room temperature. Plates were washed three times in PBST and three times in PBS before addition of 100 μl of filtered 3,3,5,5-tetramethylbenzidine (Mabtech). Plates were incubated for 10 min at room temperature and then washed 10 times in water and air dried. Spots were counted in an ImmunoSpot analyzer (Cellular Technology Ltd., Shaker Heights, OH). For detection of IFN-γ-producing cells, isolated lymphocytes were stimulated with HIV-1 lysate at 10 ng/ml. The virus lysate was prepared from aldritiol-2 (AT-2)-inactivated HIV-1 IIIB, including isotypic Gag, by sonication for 1 min. The protein concentration of the virus lysate was measured by enzyme-linked immunosorbent assay (ELISA). Plates were blocked with 200 μl CTL medium and then coated with 50 μl of IFN-γ capture antibody (Mabtech) at a concentration of 10 μg/ml overnight at 4°C. Control wells consisted of unstimulated cells or cells stimulated with phorbol myristate acetate (50 ng/ml) and ionomycin (300 ng/ml). Cells were added at a concentration of 2.5 × 10⁵ per well and incubated for 48 h at 37°C in 5% CO₂. Plates were then washed six times with PBST and incubated with biotinylated IFN-γ detection antibody (Mabtech) at 1 μg/ml for 3 h at room temperature. Plates were washed 6 times, and 100 μl/well streptavidin-HRP (1 μg/ml; Pierce Biochemical, Rockford, IL) was added for 1 h. Plates were washed with PBST and PBS and then developed with 3-amino-9-ethylcarbazole substrate (Sigma, St. Louis, MO). The reaction was stopped after 8 min by rinsing the plates with deionized water. The assay was run in duplicate, and spot-forming cells were counted and analyzed on the ImmunoSpot analyzer. To determine the number of spot-forming cells, the counts for duplicate wells were averaged, background counts from unstimulated wells were subtracted, and the total was normalized to 1 × 10⁶ cells.

Statistical analysis.

The Wilcoxon signed-rank test was applied to evaluate significant differences in assays using DCs. For multiple comparisons, the level of significance was adjusted by use of the Bonferroni correction. Tukey's multiple-comparison test was used for in vivo experiments.

RESULTS

Expression and cell surface display of Gag and FliC.

Three recombinant L. acidophilus strains, NCK2152, NCK2162, and NCK2166, were constructed in this study. NCK2152 and NCK2166 harboring pTRK1030 and pTRK1037, respectively, were expected to express either FliC or Gag on the cell surface, respectively. NCK2162 carrying pTRK1035 was engineered to display both FliC and Gag. Production of the Gag protein by these strains was confirmed by Western blotting. As shown in Fig. 2a, Gag-specific bands were detected in NCK2162 and NCK2166. The estimated molecular mass of the recombinant protein on the cell surface is approximately 71 kDa where the major bands appeared. The TLR5-stimulating activity of the FliC-producing strains was determined by a reporter gene assay. As expected, NCK2152 and NCK2162 activated NF-κB through TLR5 (Fig. 2b). There was no significant difference in the activity between the two strains. Cell surface display of the antigens was determined by flow cytometry. The three recombinant strains and a reference strain, NCK1895, were treated with antigen-specific antibodies and fluorescence-conjugated secondary antibodies. As shown in Fig. 2c, either FliC-specific or Gag-specific fluorescence was detected from NCK2152 and NCK2166, while both signals were observed in NCK2162. These data indicate that the three L. acidophilus strains express and display FliC, Gag, or both.

Fig 2.

Recombinant L. acidophilus strains producing/displaying Gag, FliC, or both. Production and cell surface display of Gag, FliC, or both were confirmed by combination of different techniques. (a) Gag-specific bands were detected by Western blotting. Cell extracts of recombinant L. acidophilus strains and purified Gag protein were analyzed. Molecular masses were estimated using protein standards and are shown on the left margin. (b) TLR5-stimulating activity was determined by a reporter gene assay. Mean values of the fold increase in NF-κB activation with standard errors are calculated from triplicate assays. (c) Cell surface display of the proteins was detected by flow cytometry. Bacterial cells stained with antigen-specific antibodies followed by fluorochrome-conjugated secondary antibodies were analyzed in two-dimensional dot plots.

Maturation and cytokine responses of human myeloid DCs induced by recombinant L. acidophilus.

The immunological impact of stimulation with the L. acidophilus strains on human DCs was assessed in vitro. Freshly isolated CD1a⁺ myeloid DCs from human blood were incubated with the recombinant lactobacilli. Before incubation, the DCs expressed low levels of CD40, CD83, CD80, and CD86 (Fig. 3a). These maturation markers were upregulated by 24 h, regardless of the bacterial strain used. For CD40 and CD83, however, stimulation with L. acidophilus caused additional upregulation. In order to compare the expression levels of CD40 and CD83 induced by different L. acidophilus strains, the median fluorescence intensity of CD40 and the percentage of CD83⁺ cells were analyzed. As shown in Fig. 3b, the levels of those molecules were slightly different depending on the strain. The DCs stimulated with NCK2162 expressed higher levels of CD40 than those stimulated with NCK2152 in all donors. NCK2166 induced CD83⁺ cells more efficiently than NCK2162 or NCK1895. For cytokine responses, NCK2162-stimulated DCs secreted larger amounts of IL-1β and IL-6 than NCK1895-stimulated cells (Fig. 4). IL-12 production by NCK2162 stimulation was higher than that by NCK2152 stimulation. TNF-α was induced more efficiently by NCK2166 than by NCK1895. Extremely high levels of IL-8 production and low levels of IFN-α2 secretion were also observed. Overall, the immune responses of the DCs stimulated with the recombinant lactobacilli varied widely, depending on the donors. This may be due to dissimilar genetic and immunological backgrounds. Since each datum obtained using DCs from the same donor was correlated, a statistical analysis for dependent samples was applied (Wilcoxon signed-rank test).

Fig 3.

Upregulation of maturation markers in human myeloid DCs after stimulation with the L. acidophilus strains. Freshly prepared human myeloid DCs (from six different donors) were exposed with or without the recombinant lactobacilli and incubated for 24 h. (a) Four maturation markers expressed on the stimulated or nonstimulated DCs were compared in overlaid histogram plots. The experiment was repeated 6 times using DCs from different donors. The bar in the histogram plot of CD83 represents the region of CD83⁺ cells. Filled gray, DCs before incubation (0 h); dotted line, DCs after incubation (24 h) without stimulation; solid line, DCs after incubation with NCK1895. (b) The expression levels of CD40 and CD83 among DCs stimulated with (NCK1895, NCK2152, NCK2162, and NCK2166) or without (medium only) recombinant L. acidophilus strains were compared. The median fluorescence intensity (MFI) of CD40⁺ cells and counts of CD83⁺ cells were plotted. The statistical significance of differences between each of the L. acidophilus-stimulated DCs was analyzed by the Wilcoxon signed-rank test with the Bonferroni correction. A P value of <0.01 is considered a significant difference in this multiple comparison.

Fig 4.

Cytokine profiling in cultures of human myeloid DCs stimulated with or without the recombinant lactobacilli. Seven cytokines in culture supernatants of stimulated DCs (24 h) were detected by use of a Milliplex suspension bead assay system. The concentrations of each cytokine are plotted individually. The median fluorescence intensity values of IL-8 and IFN-α2 are shown only for reference because the values were out of the certified detection range. The statistical significance of differences between each group was analyzed by the Wilcoxon signed-rank test with the Bonferroni correction. A P value of <0.01 is considered a significant difference in this multiple comparison.

Induction of antigen-specific IgA and IFN-γ-producing cells by oral immunization of mice.

To investigate whether Gag-displaying recombinant L. acidophilus would elicit specific immune responses, mice were immunized through a mucosal route by oral gavage. After repeated doses, mice were sacrificed to collect lymphoid tissues/organs. For detection of IgA-producing cells at local mucosa, total lymphocytes from Peyer's patches, large intestine, and female reproductive tract were prepared and incubated to perform ELISpot assay. Total IgA-producing cells were present at 1.0 × 10⁴ to 1.7 × 10⁴ cells/million cells in Peyer's patches, 1.0 × 10³ to 6.5 × 10³ cells/million cells in female reproductive tract, and 3.0 × 10⁴ to 6.3 × 10⁴ cells/million cells in large intestine. No significant difference in total IgA-producing cells was observed except for cells from the female reproductive tract from the NCK1895-immunized group (Fig. 5a). The isolated lymphocytes secreting Gag-specific IgA were also counted. As shown in Fig. 5b, a significant increase in Gag-specific IgA-producing cells was present in the female reproductive tract and large intestine from the NCK2162-immunized group. Meanwhile, Gag-specific antibodies were not titrated by ELISA due to their weak or nonspecific signals (data not shown). IFN-γ-producing cells were counted after incubation with HIV-1 lysate. Relatively large numbers of IFN-γ-producing cells were detected in mucosal lymphoid tissues, while few such cells were found in spleen (Fig. 6). In Peyer's patches and female reproductive tract, significantly higher frequencies of IFN-γ-producing cells were observed in the NCK2166-immunized group. Especially in Peyer's patches, a remarkable difference between the NCK2166-stimulated group and the other groups was shown. A relatively high background of IFN-γ-producing cells was observed in the female reproductive tract.

Fig 5.

Counts of lymphocytes secreting IgA (total or Gag specific). Lymphocytes were isolated from Peyer's patches (PP), female reproductive tract (FRT), and large intestine (LI) and analyzed by ELISpot assay. (a) Total IgA-producing cells were counted after 48 h incubation without stimulation. (b) Lymphocytes secreting Gag-specific IgA were detected after stimulation with HIV-1 lysate. Each column in the charts represent lymphocytes prepared from nonimmunized (gray open bars), NCK1895-immunized (gray filled bars), NCK2162-immunized (black open bars), and NCK2166-immunized (black filled bars) mice. Tukey's multiple-comparison test was applied for statistical analysis. A P value of <0.05 is considered a significant difference. SFU, spot-forming units.

Fig 6.

Detection of IFN-γ-producing lymphocytes in the immunized mice. Lymphocytes were isolated from spleen (SP), Peyer's patches (PP), female reproductive tract (FRT), and large intestine (LI) and applied for ELISpot assay. IFN-γ-producing cells were counted after 48 h stimulation with HIV-1 lysate. Each column in the chart represents lymphocytes prepared from nonimmunized (gray open bars), NCK1895-immunized (gray filled bars), NCK2162-immunized (black open bars), and NCK2166-immunized (black filled bars) mice. Tukey's multiple-comparison test was applied for statistical analysis. A P value of <0.05 is considered a significant difference.

DISCUSSION

In the current study, recombinant L. acidophilus strains displaying HIV-1 Gag with or without coexpression of FliC as an adjuvant on the cell surface were constructed and were tested for in vitro and in vivo immunogenicity. To date, many Lactococcus lactis and Lactobacillus strains have been engineered to be potential mucosal vaccine vectors. Most of these studies investigated the immunological properties of recombinant lactic acid bacteria producing antigen alone. A few have demonstrated the effectiveness of providing additional adjuvanticity via genetic modification of the bacteria. Adjuvant has more often been provided in trans and rarely in cis. For example, single-chain mouse IL-12 was produced by L. lactis and administered to mice along with another recombinant L. lactis strain expressing E7 antigen of human papillomavirus type 16 (5). Similarly, L. casei engineered to secrete mouse IL-1β was mixed with heat-killed Salmonella Enteritidis and dosed orally (18). An alternative approach employed by Mohamadzadeh et al. was to construct lactobacilli expressing antigen fused with a dendritic cell-targeting peptide to enhance uptake of antigen by DCs. This approach successfully protected mice against Bacillus anthracis using L. acidophilus as a delivery vehicle (29). Another study tested recombinant L. lactis secreting either murine IL-2 or murine IL-6 with coexpression of tetanus toxin fragment C (34). All of these approaches improved the efficacy and expanded the potential of lactic acid bacteria as delivery agents.

The strategy employed here was based on two lines of study. The first was the clear demonstration that stimulation of multiple TLRs can greatly enhance immune responses (21, 32). The second was the growing evidence that flagellin can be a potent adjuvant (28). In both cases, it is also clear that antigen delivery and TLR activation must be targeted to the same antigen-presenting cell as a fusion product or coformulation (1, 16, 25, 31). We employed the unique approach of displaying both antigen and adjuvant simultaneously on the bacterial cell surface using FliC as the adjuvant. The recombinant surface FliC induced NF-κB via TLR5 in a reporter assay, thereby proving that it was biologically active. Lactic acid bacteria produce lipoteichoic acid, diacylated peptidoglycans, and muramyl dipeptides that lead to activation of TLR2, TLR2/6, and NOD2, respectively (22). Expression of FliC resulted in an expansion of the TLR activation profile, as L. acidophilus and the majority of other Lactobacillus strains are intrinsically devoid of flagella.

DCs are professional antigen-presenting cells and play a crucial role in initiating acquired immune responses. Myeloid DCs are especially important in relation to HIV-1 infection and host defense (8). In the current study, immune responses elicited in human myeloid DCs stimulated with the recombinant lactobacilli were analyzed. The bacterial stimulation caused maturation and cytokine production of the DCs, regardless of the recombinant cell surface antigens. Previous studies showed that maturation and the cytokine responses of mouse or human DCs were induced by lactobacilli in a species/strain-dependent manner (7, 9, 30, 42). This is apparently the result of variation of the MAMP composition and structure (13, 24, 33). In this study, additional effects on DC maturation and cytokine production resulted from expression of recombinant antigens/adjuvants on the cell surface. Although there was no drastic alteration in DC maturation or cytokine production, several responses were enhanced by the Gag- and FliC-expressing recombinant L. acidophilus. The strain displaying Gag alone induced greater CD83 upregulation and TNF-α secretion in the DCs than the reference strains. Relatively large amounts of IL-1β and IL-6 were produced in the DCs stimulated with the strain displaying both Gag and FliC compared to the amounts produced in the DCs stimulated by the control strain. It is critical to keep in mind that the DC is not the only cell that will ultimately be influenced by an orally delivered recombinant lactobacillus vaccine vector and that other immune cells and, importantly, the mucosal epithelium may play a prominent role in the immune response.

To this end, in vivo evaluation of immunogenicity is essential, and therefore, recombinant L. acidophilus strains were assessed by intragastric immunization in a murine model. Gag-specific IgA-producing cells were found in the large intestines and female reproductive tract of mice immunized with the recombinant strain displaying Gag and FliC but not those immunized with the recombinant strain displaying Gag alone. This result implies that the coexpression of FliC on the bacterial cell surface promotes induction of Gag-specific IgA-producing cells at the local mucosa. The fact that Gag-specific IgA cells were found in the lamina propria and not in the Peyer's patches is consistent with a TLR5-mediated mechanism of induction. As shown by Uematsu and colleagues, CD11c^hi CD11b^hi TLR5-positive DCs reside in the lamina propria and induce T cell-independent antigen-specific IgA class switching of resident naive B cells, and this requires flagellin signaling through TLR5 (36, 37). In contrast to the specific IgA responses, the L. acidophilus strain expressing Gag alone induced specific IFN-γ-producing cells most efficiently at the local mucosa. This result suggests that this strain promotes cellular immunity rather than humoral immunity. Thus, we show here that two different Gag-displaying L. acidophilus strains have dissimilar immunogenicity depending on the coexpression of FliC. Additional studies will be required to fully elucidate the induction pathways of the cellular versus humoral responses and to determine whether both responses can be induced either with a mixed inoculum containing both recombinant strains or through temporal separation of immunization with each individual recombinant strain.

The major routes of HIV-1 transmission are mucosal surfaces of the lower intestine and genital tracts. Thus, vaccines which can confer appropriate HIV-1-specific immunity at these sites may provide robust protection. This study provides proof of concept that recombinant L. acidophilus without external mucosal adjuvant could be applied for development of oral vaccines against HIV-1 as a delivery agent. The next step is to determine whether immunogenic HIV-1 envelope antigens can be expressed by recombinant lactobacilli and whether a combination of local mucosal IgA- and cell-mediated responses can provide protection against challenge.