In Vitro and In Vivo Analyses of Constitutive and In Vivo-Induced Promoters in Attenuated Vaccine and Vector Strains of Vibrio cholerae

Abstract

The optimal promoter for in vivo expression of heterologous antigens by live, attenuated vaccine vector strains of Vibrio cholerae is unclear; in vitro analyses of promoter activity may not accurately predict expression of antigens in vivo. We therefore introduced plasmids expressing the B subunit of cholera toxin (CtxB) under the control of a number of promoters into V. cholerae vaccine strain Peru2. We evaluated the tac promoter, which is constitutively expressed in V. cholerae, as well as the in vivo-induced V. cholerae heat shock htpG promoter and the in vivo-induced V. cholerae iron-regulated irgA promoter. The functionality of all promoters was confirmed in vitro. In vitro antigenic expression was highest in vaccine strains expressing CtxB under the control of the tac promoter (2 to 5 μg/ml/unit of optical density at 600 nm [OD₆₀₀]) and, under low-iron conditions, in strains containing the irgA promoter (5 μg/ml/OD₆₀₀). We orally inoculated mice with the various vaccine strains and used anti-CtxB immune responses as a marker for in vivo expression of CtxB. The vaccine strain expressing CtxB under the control of the tac promoter elicited the most prominent specific anti-CtxB responses in vivo (serum immunoglobulin G [IgG], P ≤ 0.05; serum IgA, P ≤ 0.05; stool IgA, P ≤ 0.05; bile IgA, P ≤ 0.05), despite the finding that the tac and irgA promoters expressed equivalent amounts of CtxB in vitro. Vibriocidal antibody titers were equivalent in all groups of animals. Our results indicate that in vitro assessment of antigen expression by vaccine and vector strains of V. cholerae may correlate poorly with immune responses in vivo and that of the promoters examined, the tac promoter may be best suited for expression from plasmids of at least certain heterologous antigens in such strains.

Development of Vibrio cholerae as a vector organism capable of expressing heterologous antigens at mucosal surfaces is attractive. V. cholerae is a noninvasive organism that effectively colonizes the intestinal mucosa of humans, and infection with V. cholerae results in immune responses that are long-lasting (10, 15). Live, attenuated vaccine strains of V. cholerae can be administered orally, and such strains have been well characterized and shown to be both safe and immunogenic in humans (11, 12, 14, 22, 24). V. cholerae vaccines strains can secrete immunoadjuvants in vivo, such as the nontoxic immunoadjuvantive mutant of Escherichia coli heat-labile enterotoxin LT_(R192G) (19), and V. cholerae vaccines strains can express large quantities of heterologous antigens in a balanced lethal plasmid expression system (20). V. cholerae vaccine strains can also efficiently express and secrete both large and small heterologous antigens (2, 3, 17, 18), and a mouse model of V. cholerae infection that permits rapid preliminary evaluation of V. cholerae vaccine and vector strains in vivo has been developed (4, 6). The optimal promoter for in vivo expression of heterologous antigens by vaccine and vector strains of V. cholerae is, however, unclear.

Constitutive promoters can drive high-level expression of certain antigens; however, such expression can be toxic to bacterial cells (5). In contrast, in vivo-induced promoters may have no or low-level activity when evaluated in vitro, but such promoters may be extremely active in vivo (5, 16). Strains expressing heterologous antigens from in vivo-induced promoters may be less compromised than those expressing antigens from constitutive promoters, and in vivo expression of heterologous antigens by in vivo-induced promoters may exceed that of constitutive promoters (5, 16).

To examine optimal promoter activity in V. cholerae vaccine and vector strains, we compared in vitro and in vivo activities of a number of promoters. We used derivatives of V. cholerae vaccine strain Peru2 (V. cholerae O1 El Tor C6709 ΔattRS1) expressing a nontoxic B subunit (CtxB) of cholera toxin from a plasmid; strains used in this study were identical except for the identity of the promoter driving CtxB expression. We evaluated the tac promoter, which is constitutively active in V. cholerae since V. cholerae lacks the lacI^q repressor (1). We also examined two in vivo-induced V. cholerae promoters: the heat shock htpG promoter (induced under conditions of environmental stress) (13) and the V. cholerae iron-regulated irgA promoter (induced under low-iron conditions) (7, 9). We confirmed appropriate in vitro regulation of these promoters, and we analyzed systemic and mucosal immune responses to CtxB in mice inoculated with the various vaccine strains of V. cholerae.

MATERIALS AND METHODS

Bacterial strains and media.

The bacterial strains and plasmids used in this study are described in Table 1. All strains were maintained at −70°C in Luria-Bertani (LB) broth (21) containing 15% glycerol. All cultures contained either streptomycin (100 μg/ml) or ampicillin (100 μg/ml). Cultures were grown at indicated temperatures with aeration.

TABLE 1.

Bacterial strains and plasmids used in this study

Strain or plasmid	Relevant genotype or phenotypea	Reference or source
V. cholerae
C6709	O1, El Tor, Inaba, wild type; Smr	23
Peru2	C6709 ΔattRS1; Smr	23
O395	O1, classical strain; Smr	Lab collection
E. coli JM105	thi rpsL endA sbcB15 hsdR4 supE Δ(lac-proAB) F′ [traD36 proAB+ lacIqlacZ ΔM15]; Smr	Pharmacia P-L Biochemicals Inc., Milwaukee, Wis.
Plasmids
pVC100	pUC19-based plasmid, containing the toxR-htpG intergenic region of V. cholerae El Tor E7946 as an EcoRI-DraI fragment cloned into HincII; Ampr	13
pMBG126	pUC18-based plasmid containing a 510-bp HindIII-SmaI DNA fragment that includes the iron-regulated irgBA genes from V. cholerae O395, blunt ended into HindIII and SphI sites of the polylinker; Ampr	2
pKK223-3	pBR322-based plasmid containing a multiple cloning site between the tac promoter and the rrnB transcriptional terminator; Ampr	Pharmacia P-L Biochemicals
pETR1	pKK223-3-based plasmid, with a 393-bp EcoRI-PstI insert containing a modified ctxB derived from V. cholerae C6709 by recombinant mutagenic PCR. ctxB contains a unique NheI site 6 bp 3′ to the sequence of the carboxy terminus of the CtxB signal peptide. The tac promoter drives ctxB expression; Ampr	This study
pMCSETR1B	pETR1 derived, with replacement of a BamHI-EcoRI fragment containing the tac promoter with a polylinker (BamHI-NsiI-XbaI-EcoRV-BglII-SpeI-EcoRI-PstI). The 3′ end of the polylinker contains stop codons in all three reading frames; Ampr	18
pETR13	pMCSETR1B derived, with a 600-bp BamHI-XbaI fragment containing the heat shock htpG promoter of V. cholerae C6709 controlling expression of ctxB; Ampr	This study
pETR12	pMCSETR1B derived, with a 255-bp HindIII-NcoI blunt-ended insert containing the iron-regulated irgA promoter of V. cholerae O395 cloned into the EcoRV site of pMCSETR1B, such that the irgA promoter controls expression of ctxB; Ampr	This study

^aSm^r, streptomycin resistant; Amp^r, ampicillin resistant. 

Recombinant DNA methods.

Isolation of plasmid DNA, restriction digestions, and agarose gel electrophoresis were performed using standard procedures (21). DNA sequencing was performed at the DNA Sequencing Core Facility, Department of Molecular Biology, Massachusetts General Hospital, using ABI Prism DiTerminator cycle sequencing with AmpliTaq DNA polymerase FS with an ABI 377 DNA sequencer (Perkin-Elmer Applied Biosystems Division, Foster City, Calif.) (17).

Plasmids were transformed into E. coli JM105 by using standard techniques or were electroporated into V. cholerae with a Gene Pulser (Bio-Rad Laboratories, Richmond, Calif.) as instructed by the manufacturer and modified for electroporation into V. cholerae as previously described (8). Electroporation conditions were 2,500 V at 25-mF capacitance, producing time constants of 4.8 to 4.9 ms.

DNA restriction endonucleases, T4 DNA ligase, calf intestinal alkaline phosphatase, and Klenow fragment of DNA polymerase I were used as specified by manufacturers. Pfu DNA polymerase (Stratagene, La Jolla, Calif.) was used for thermocyclic DNA amplification, using reaction mixtures and protocols as previously described (18). Restriction enzyme-digested plasmid DNA fragments were fractionated on 1% agarose gels; required DNA fragments were removed under UV illumination and recovered using GenElute agarose spin columns (Supelco Inc., Bellefonte, Pa.).

Plasmid constructions.

Plasmid pETR1 was constructed by recombinant mutagenic PCR using two oligonucleotide primer pairs (primers 1 plus 3 and primers 2 plus 4) to amplify ctxB from the genome of V. cholerae C6709 and to introduce a unique NheI site 6 bp 3′ to the coding sequence of the CtxB leader peptide. Primer 1 (5′-ACTGACAGCTGGGAATTCTAAGGATGAATTATGATTAAATTAAAA-3′) and primer 3 (5′-AGTAATATTTTGGCTAGCAGGTGTTCCATGTGCATATGC-3′), including EcoRI and NheI restriction enzyme sites, respectively (underlined), were initially used to amplify a 110-bp PCR product extending from the Shine-Dalgarno sequence to that of the leader peptide of CtxB, inclusively. Primer 2 (5′-TCAGACTTCAGACTGCAGGGGCAAAACGGTTGCTTCTCATCATCG-3′) and primer 4 (5′-CATGGAACACCTGCTAGCCAAAATAATACTGATTTGTGT-3′), including PstI and NheI restriction enzyme sites, respectively (underlined), were used to amplify a 406-bp PCR product that extended from the leader sequence to the stop codon of ctxB, inclusively. Overlap extension PCR, using primers 1 and 2, was performed with equimolar concentrations of gel-purified 110- and 406-bp PCR products for 10 cycles (94°C, 1 min; 50°C, 1 min; 72°C, 30 s). The final 393-bp fragment was cloned into the EcoRI and PstI sites of pKK223-3 immediately downstream of the tac promoter.

To construct pMCSETR1B, a BamHI-EcoRI DNA fragment containing the tac promoter of plasmid pETR1 was replaced with a polylinker that included BamHI-NsiI-XbaI-EcoRV-BglII-SpeI-EcoRI-Pst-I restriction enzyme sites (18). To prohibit translational read-through into ctxB, the 3′ end of the multiple cloning site included stop codons in all three reading frames. Plasmid pMCSETR1B, therefore, contained a multiple cloning site immediately upstream of a 393-bp promoterless ctxB gene from V. cholerae C6709.

Plasmid pVC100 contains the toxR-htpG intergenic region of V. cholerae El Tor E7946 as an EcoRI-DraI DNA fragment inserted into the HincII restriction site of the polylinker of pUC19 (13). Plasmid pETR13 was constructed by cloning an approximately 600-bp BamHI-XbaI DNA fragment containing the toxR-htpG intergenic region from pVC100 into pMCSETR1B, such that the modified ctxB gene was placed under the transcriptional control of the htpG promoter.

Plasmid pMBG126 contains a 510-bp HindIII-SmaI DNA fragment that includes the iron-regulated irgBA genes from V. cholerae O395; this fragment had previously been blunt-end ligated into the HindIII-SphI sites of the polylinker of pUC18 (2). To construct plasmid pETR12, a 438-bp HindIII-EcoRV DNA fragment from pMBG126 containing the irgA promoter was blunt-end ligated into the EcoRV site of pMCSETR1B. The 438-bp insert was then truncated with NcoI and EcoRV, treated with the Klenow fragment of DNA polymerase I, and religated; this truncation removed a 183-bp nonpromoter fragment of DNA and placed ctxB under the transcriptional control of a 255-bp insert containing the irgA promoter.

Plasmids pMCSETR1B, pETR1, pETR13, and pETR12 were confirmed by restriction enzyme digestion and sequence analysis; plasmids were electroporated into V. cholerae vaccine strain Peru2.

In vitro analysis of various promoters controlling CtxB expression.

In vitro expression of CtxB by the various V. cholerae vaccine strains was analyzed by measuring CtxB concentrations in culture supernatants as previously described (6, 18, 19). Overnight cultures of V. cholerae Peru2(pMCSETR1B), Peru2(pETR1), and Peru2(pETR13) were grown at 25, 30, 37, and 42°C. Overnight cultures of Peru2(pMCSETR1B) and Peru2(pETR12) were grown at 37°C in LB containing ampicillin (normal iron conditions) and in LB containing ampicillin and 0.150 mM 2,2′-dipyridyl, an iron chelator producing low-iron conditions (7). Cell-free supernatants of overnight cultures were serially diluted in phosphate-buffered saline (PBS)–0.05% Tween 20 (Sigma Chemical Co., St. Louis, Mo.) (PBS-T; undiluted to 1:128), applied to 96-well microtiter plates previously coated with 100 ng of type III ganglioside (Sigma) per well in 50 mM carbonate buffer (pH 9.6), and subsequently blocked with PBS–1% bovine serum albumin (Sigma). Plates were incubated at 37°C for 1 h and washed with PBS-T, and a 1:2,000 dilution of goat anti-CtxB (List Biological Laboratories, Inc., Campbell, Calif.) in PBS-T was applied to each well. Following incubation at 37°C for 1 h, plates were washed with PBS-T, and a 1:2,000 dilution of anti-goat immunoglobulin G (IgG)-horseradish peroxidase conjugate (Southern Biotechnology Associates Inc., Birmingham, Ala.) was added to each well. Plates were then incubated at 37°C for 1 h, washed with PBS-T, and developed with a 1-mg/ml solution of 2,2′-azinobis(ethylbenzthiazolinesulfonic acid) (ABTS; Sigma) with 0.1% H₂O₂ (Sigma). The optical density at 405 nm (OD₄₀₅) was read in a Vmax microplate reader (Molecular Devices Corp., Sunnyvale, Calif.) and compared with that of a standard curve generated using dilutions of purified CtxB (List) in PBS-T.

Inoculation of germfree mice.

Immediately upon removal from their shipping container, four groups of 5 to 19 germfree female Swiss mice, 3 to 4 weeks old (Taconic Farms, Inc., Germantown, N.Y.), were orally inoculated by gastric intubation with 250 μl of inoculum containing approximately 10⁹ CFU of Peru2(pMCSETR1B), Peru2(pETR1), Peru2(PETR13), or Peru2(pETR12) resuspended in 0.5 M NaHCO₃ (pH 8.0) (6). Mice were subsequently housed in non-germfree conditions. All groups of mice received a primary oral vaccination series administered on days 0, 2, 4, and 6; oral booster inoculations of 10⁹ CFU in 125 μl were administered on days 28, 42, 56, and 70.

Immunological sampling.

Mice were sacrificed on day 84, at which time blood, bile, and stool samples were collected, processed, aliquoted, and stored as previously described (19).

Detection of immune responses.

Serum vibriocidal antibody titers were measured in a microassay as previously described (18). Anti-CtxB antibody responses were detected using microtiter plates previously coated with ganglioside and CtxB (18, 19). To detect anti-CtxB IgG and IgA antibodies in sera, duplicate samples of 1:200 dilutions of sera in PBS-T were added to wells previously coated with ganglioside-CtxB. Following overnight incubation, goat anti-mouse IgG antibody conjugated to biotin or goat anti-mouse IgA antibody conjugated to biotin (Kirkegaard & Perry Laboratories, Gaithersburg, Md.) was applied to each well. After the addition of a 1:4,000 dilution of streptavidin-horseradish peroxidase conjugate (Zymed Laboratories, Inc., South San Francisco, Calif.) in PBS-T, plates were developed for peroxidase activity as described above. The OD₄₀₅ was determined kinetically with the Vmax microplate reader. Plates were read for 5 min at 19-s intervals, and the maximum slope for an OD change of 0.2 U was reported as milli-OD units per minute (6, 17).

To detect specific IgA antibody responses in stool and bile, concentrations of total IgA in stool and bile were first measured. Duplicate samples of serial twofold dilutions of processed stool (1:100 to 1:800) or bile (1:800 to 1:6,400) in PBS-T were added to wells previously coated with 100 ng of rat monoclonal anti-mouse IgA antibody R5-140 (Pharmingen, San Diego, Calif.) (19). After the addition of a 1:2,000 dilution of goat anti-mouse IgA antibody conjugated to biotin and streptavidin-horseradish peroxidase conjugate, plates were developed for peroxidase activity as described above. Comparison was made to a mouse IgA standard (Kappa TEPC 15; Sigma). To detect specific anti-CtxB antibodies in stool and bile, 200 ng of total biliary IgA (single) or 200 ng of total stool IgA (duplicate) in PBS-T was added to wells previously coated with ganglioside-CtxB (19, 20). Plates were incubated overnight at room temperature and washed with PBS-T, following which goat anti-mouse IgA antibody conjugated to biotin and streptavidin-horseradish peroxidase conjugate were added. Plates were developed for peroxidase activity, and the OD₄₀₅ was determined kinetically.

Statistics and graphs.

Statistical analysis for the comparison of geometric means was performed with the Mann-Whitney U test for nonparametric data using SPSS for Windows 8.0. Data were plotted using GraphPad PRISM, version 3 (GraphPad Software Inc., San Diego, Calif.).

RESULTS AND DISCUSSION

In vitro analysis of CtxB expression.

Cultures of V. cholerae Peru2 containing plasmid pETR1, pETR13, pETR12, or control plasmid pMCSETR1B were grown under a number of in vitro conditions, and supernatants were analyzed for CtxB (Table 2). Peru2(pETR1), with the tac promoter controlling CtxB expression, secreted CtxB to the supernatant at a range of temperatures (25, 30, and 37°C). Peru2(pETR1) expressed approximately 2 to 5 μg/ml/OD₆₀₀ at 30 and 37°C. Lower expression (0.5 to 1 μg/ml/OD₆₀₀) occurred at 25°C. CtxB was not detected in culture supernatants of Peru2(pETR1) grown at 42°C. Strain Peru2(pETR13), with the V. cholerae heat shock htpG promoter controlling CtxB expression, did not secrete detectable amounts of CtxB when grown at 25 or 30°C but did express CtxB (approximately 60 ng/ml/OD₆₀₀) when grown at 37°C. Peru2(pETR12), with the irgA promoter controlling CtxB expression, secreted approximately 40- to 50-fold more CtxB into supernatants when grown under low-iron conditions (approximately 4 to 5 μg/ml/OD₆₀₀) compared to expression during normal iron conditions (approximately 100 ng/ml/OD₆₀₀). Peru2(pMCSETR1B), containing a promoterless ctxB gene, produced no CtxB under any in vitro condition.

TABLE 2.

In vitro CtxB expression under the control of various promoters^a

Promoter/plasmid	Mean CtxB expression (μg/ml/OD600) ± SE
	Temp (°C)				With 0.150 mM 2-2′-dipyridyl
	25	30	37	42
tacp/pETR1	0.67 ± 0.67	4.64 ± 2.82	1.93 ± 0.8	0	—
htpGp/pETR13	0	0	0.6 ± 0.04	0	—
irgAp/pETR12	—	—	0.120 ± 0.007	—	4.50 ± 1.10
Control (pMCSETR1Bb)	0	0	0	0	0

^aAll experiments were performed in V. cholerae Peru2. Peru2(pETR1) and Peru2(pETR13) were grown at indicated temperatures. Peru2(pETR12) was grown at 37°C without or with 0.150 mM 2,2′-dipyridyl (an iron chelator). —, not determined. 

^bContains a promoterless ctxB gene. 

These results demonstrate that the tac, heat shock htpG, and iron-regulated irgA promoters were all active in vitro and that they responded to environmental conditions as expected: the tac promoter drove constitutive expression of CtxB under a number of in vitro conditions, the irgA promoter was most active in low-iron conditions, and the htpG promoter was most active at higher temperatures. Under the in vitro conditions tested, the constitutively active tac promoter and the V. cholerae iron-regulated irgA promoter (in low-iron conditions) both drove high-level and equivalent expression of CtxB.

Measurement of vibriocidal antibody responses.

Vibriocidal antibodies were measured in serum samples collected on day 84 (Fig. 1). Vibriocidal antibodies are directed against V. cholerae organisms themselves and reflect the ability of V. cholerae strains to both survive in vivo and colonize the intestinal surface (20). Vibriocidal antibody levels were equivalent in all groups of animals, suggesting that in vivo survival patterns of the various V. cholerae vaccine strains, including Peru2(pETR1), in which the tac promoter drives high-level constitutive expression of CtxB, were equivalent, as well.

FIG. 1.

Geometric mean titers (GMT) of vibriocidal antibody responses on day 84 samples. Groups of mice were inoculated with Peru2(pMCSETR1B) (CONTROL), Peru2(pETR1) (tacp → ctxB), Peru2(pETR13) (htpGp → ctxB), or Peru2(pETR12) (irgAp → ctxB). See text for details. Error bars represent standard errors of the mean for each group. Vibriocidal responses were comparable in all groups of mice.

Measurement of systemic and mucosal anti-CtxB antibodies.

Systemic (serum) and mucosal (stool and bile) anti-CtxB antibodies were measured in samples collected on day 84 (Fig. 2). Compared to responses in mice that received control strain Peru2(pMCSETR1B), mice that received vaccine strains expressing CtxB under the control of the tac or irgA promoter had significant serum anti-CtxB IgG responses (P ≤ 0.05), confirming previous results (3). Compared to responses in mice that received Peru2(pETR13) expressing CtxB under the control of the heat shock htpG promoter, and compared to responses in mice that received Peru2(pETR13) expressing CtxB under the control of the iron-regulated irgA promoter, anti-CtxB responses were highest in mice that received Peru2(pETR1), the strain expressing CtxB under the control of the tac promoter. Mice that received V. cholerae Peru2(pETR1) had the highest level of anti-CtxB responses in all samples: serum IgG (P ≤ 0.05), serum IgA (P ≤ 0.05), stool IgA (P ≤ 0.05), and bile IgA (P ≤ 0.05).

FIG. 2.

Anti-CtxB enzyme-linked immunosorbent assay results on day 84 samples. Mice were inoculated orally with Peru2(pMCSETR1B) (CONTROL), Peru2(pETR1) (tacp → ctxB), Peru2(pETR13) (htpGp → ctxB), or Peru2(pETR12) (irgAp → ctxB). The geometric mean and standard error of the mean are reported for each group. ∗, P ≤ 0.05 compared to mice that received Peru2(pETR13) and compared to mice that received Peru2(pETR12).

In summary, we used anti-CtxB antibody responses in mice as a surrogate marker for in vivo activity of various promoters in attenuated vaccine and vector strains of V. cholerae. Despite the fact that optimal activities of the irgA and tac promoters were equivalent in vitro, we found that strains expressing CtxB from the constitutively active tac promoter resulted in the most prominent immune responses in animals. Changing the identity of the expressed antigen may alter the relative efficiencies of the promoters that we examined, as may changing from a plasmid to a chromosomally based antigen expression system; however, for plasmid-based systems in vaccine and vector strains of V. cholerae, the tac promoter appears to be well suited for high-level in vivo expression of at least certain heterologous antigens. Since the level of expression of an antigen correlates with the prominence of induced immune responses, our data should facilitate development of effective vaccine and vector strains of V. cholerae.