Construction and evaluation of a O139 Vibrio cholerae vaccine candidate based on a hemA gene mutation

Construction and evaluation of a O139 Vibrio cholerae vaccine candidate based on a hemA gene mutation

Vaccine 24 (2006) 3750–3761 Construction and evaluation of a O139 Vibrio cholerae vaccine candidate based on a hemA gene mutation Manickam Ravichandr...

423KB Sizes 4 Downloads 120 Views

Vaccine 24 (2006) 3750–3761

Construction and evaluation of a O139 Vibrio cholerae vaccine candidate based on a hemA gene mutation Manickam Ravichandran a,∗ , Syed Atif Ali b , Nur Haslindawaty Abdul Rashid a , Sinniah Kurunathan a , Chan Yean Yean a , Lai Chin Ting b , Afifi Sheikh Abu Bakar a , Pattabiraman Lalitha b , Zainul F. Zainuddin b a

Department of Medical Microbiology and Parasitology, School of Medical Sciences, Universiti Sains Malaysia Health Campus, Kubang Kerian 16150, Kelantan, Malaysia b School of Health Sciences, Universiti Sains Malaysia, Kubang Kerian 16150, Kelantan, Malaysia Available online 22 July 2005

Abstract In this paper, we describe the development of VCUSM2, a live metabolic auxotroph of Vibrio cholerae O139. Auxotrophy was achieved by mutating a house keeping gene, hemA, that encodes for glutamyl-tRNA reductase, an important enzyme in the C5 pathway for ␦-aminolevulenic acid (ALA) biosynthesis, which renders this strain dependent on exogenous ALA for survival. Experiments using the infant mouse and adult rabbit models show that VCUSM2 is a good colonizer of the small intestine and elicits greater than a four-fold rise in vibriocidal antibodies in vaccinated rabbits. Rabbits vaccinated with VCUSM2 were fully protected against subsequent challenge with 1 × 1011 CFU of the virulent wild type (WT) strain. Experiments using ligated ileal loops of rabbits show that VCUSM2 is 2.5-fold less toxic at the dose of 1 × 106 CFU compared to the WT strain. Shedding of VCUSM2 in rabbits were found to occur for no longer than 4 days and its maximum survival rate in environmental waters is 8 days compared to the greater than 20 days for the WT strain. VCUSM2 is thus a potential vaccine candidate against infection by V. cholerae O139. © 2005 Elsevier Ltd. All rights reserved. Keywords: O139 Vibrio cholerae; hemA mutation; ALA auxotroph; Vaccine

1. Introduction Cholera is a severe secretory diarrheal disease caused by the Gram-negative bacterium Vibrio cholerae [1]. Previously, only V. cholerae strains of O group 1 (O1 strains) were associated with cholera, but in 1992, a toxigenic non-O1 strain designated as O139 or the Bengal strain was recognized as the etiological agent of epidemic cholera in India and Bangladesh [2]. While vaccines are commercially available against cholera infection caused by V. cholerae O1, studies have shown that these vaccines do not offer protection against infection caused by V. cholerae O139 strains [3]. However in recent years, a number of vaccine candidates for V. cholerae O139 ∗

Corresponding author. Tel.: +60 976 64592; fax: +60 976 48673. E-mail address: [email protected] (M. Ravichandran).

0264-410X/$ – see front matter © 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.vaccine.2005.07.016

have been developed and evaluated but currently no vaccine is commercially available against it [4–6]. An approach shared by all of these vaccine strains is the absence of one or more toxin producing genes. Studies have suggested that a natural infection results in the most efficient immunity; this suggests that the ideal vaccine should mimic the natural infection as closely as possible. Thus, such a vaccine should be live and able to express its entire antigen, but the major limitation in this approach is the associated toxicity. A compromise that combines the advantages of a live vaccine, whilst minimizing or possibly avoiding the problem of toxicity, could be the use of auxotrophic mutants of V. cholerae that make them dependant for growth on substrates unavailable in the human intestine. A number of such auxotrophic vaccine strains that have been developed include CVD102, Peru2glnA (pTIC5), 638T, and V286 [7–10].

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

Two housekeeping genes, thyA and glnA, have been targeted for the construction of auxotrophic mutants of V. cholerae. However, the mutations in both the genes were ‘leaky’ and in the absence of thymidine or glutamine, the auxotrophs were still able to grow in the small intestine of the animal models [7,8,10]. A V. cholerae O1 El Tor (V286) metabolic auxotroph which required ␦-aminolevulinic acid (ALA) for its growth in vitro has been described by Rijpkema et al. [9]. This auxotroph, was not able to grow in the small intestine of rabbits; but elicited good immune response. V286 was created by transposon mutagenesis and penicillin enrichment, hence precise location of the mutation or the gene responsible for the ALA auxotrophy was not known [9]. In E. coli, ALA is synthesized from the five carbon skeleton of glutamate by the C5 pathway. The C5 pathway consists of three steps: glutamate activation to glutamyl-tRNA by glutamyl-tRNA synthetase (GTS), reduction of glutamyltRNA to glutamate semialdehyde (GSA) by glutamyl-tRNA reductase (GTR), and conversion of GSA to ALA by an aminotransferase. ALA is the important intermediate in the synthesis of tetrapyrrole (porphyrin) compounds. Studies have shown that glutamyl-tRNA reductase is the rate limiting enzyme in the C5 pathway of ALA synthesis [11]. The 85 kDa GTR in E. coli, is encoded by a 1.45 kb gene designated as hemA. We proposed that a mutation in the homologous hemA gene of V. cholerae O139 will lead to the creation of an ALA auxotroph that could be used as an attenuated live oral vaccine against this organism. In this paper, we describe the construction, characterization, and evaluation of a V. cholerae O139 vaccine candidate designated as VCUSM2, which was created by mutating the

3751

hemA gene through a series of genetic manipulations. Phenotypic properties of VCUSM2 were characterized and its efficacy as a vaccine was evaluated in animal models.

2. Materials and methods 2.1. Strains and media Bacterial strains and plasmids used in this study are described in Table 1. All bacterial strains were grown in Luria-Bertani (LB) medium. WT V. cholerae O139 Bengal was maintained on LB agar containing polymyxin B (0.75 ␮g/ml or 450 unit/ml). E. coli strains were maintained on LB agar and plasmid containing transformants were selected on LB agar containing ampicillin (100 ␮g/ml) or kanamycin (50 ␮g/ml). VCUSM1 and VCUSM2 were maintained on LB agar supplemented with ALA (40 ␮g/ml or 240 ␮M). 2.2. Construction of hemA mutants The strategy for constructing hemA mutants, VCUSM1 and VCUSM2, is shown in Fig. 1. 2.3. Construction of VCUSM1 We had previously isolated and characterized a region of the V. cholerae genome which contained the hemA and the neighboring hemM genes (unpublished; GenBank accession number: AF227752) whose sequences were used to design the primers for this study. This region of 2.18 kb in size was

Table 1 Bacterial strains and plasmids used in this study Strains or plasmids E. coli Top10 E. coli DH5␣ ␭-pir E. coli BW 20767 ␭-pir WT VCUSM1 VCUSM2 Plasmids pCR2.1 TOPO pTOPO-hemA/M pARO 180 pARO-hemA/M pARO-hemA/M-aphA pARO-hemA/M-gfp pWM91 pWM91-hemA/M-aphA pWM91-hemA/M-gfp pWM91-hemA/M a HUSM:

Description

Source or reference

F- merA ∆(mrv hsdRMS mcrBC) φ80lacZ∆M15 ∆lac X74 deOR recA1 araD139 ∆(ara-leu)7697 galU galK rpsL(strR ) endA1 nupG F-φ80d 1acZ∆ m15∆(lacZYA-argF) U169 deOR recA1 endA1 hsdR17 (rk − , mk + ) phoA suppE44 thi-1 gyrA96 relA1 RP4 2tet: mu-1kan::Tn7integrant leu 63::IS10 rec A1 cre(510 hsdR17 end A1 Zbf-5 uid) (∆MluI): pir thi Wild type, V. cholerae O139 Bengal isolated from a patient in HUSMa V. cholerae O139 ALA auxotroph (hemA:: aphA) V. cholerae O139 ALA auxotroph (hemA)

Invitrogen

LacZα m13R T7 promoter m13F-20 M13F-40 MCS f1 origin KanR , AmpR ColE1 origin pTOPO with hemA/M gene from V. cholerae O139 Bengal, AmpR LacZ oriT/bom pMBI AmpR pARO180 with hemA/M at EcoRI site pARO-hemA/M with aphA at polished BstXI site of hemA gene, AmpR, KanR pARO-hemA/M with gfp at polished BstXI site of hemA gene, AmpR f1(+) ori lacZ␣ of pBluescript II (SK+ ) mobRP4, oriR6K,SacB and AmpR pWM91 with hemA::aphA/hemM fragment cloned at SmaI site, AmpR , KanR pWM91 with hemA::gfp/hemM fragment cloned at SmaI site, AmpR pWM91 with frame-shift mutated hemA

Hospital University Sains Malaysia.

Gift from Dr. William Metcalf, University of Illinois [33]

This study This study This study Invitrogen This study [34] This study This study This study Gift from Dr. William Metcalf, University of Illinois [33] This study This study This study

3752

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

Fig. 1. Schematic diagram of the construction of VCUSM1 and VCUSM2.

PCR amplified using primers VHF (5 -GAC CTG TGA TGT AAA GGA AC-3 ) and VHR (5 -CTT CAT AGC GCT CAA CAA GG-3 ). The hemA/M gene fragment was then cloned into the pCR2.1-TOPO cloning vector. The resultant recombinant plasmid, pTOPO-hemA/M, was digested with EcoRI and the hemA/M gene was subcloned into the EcoRI site of the pARO180 vector to create the recombinant plasmid pAROhemA/M. In order to knockout hemA gene, a 1.1 kb aphA gene cassette that encodes for kanamycin resistance was PCR amplified from pCR2.1-TOPO (Invitrogen) using primers, KS1 (5 -TCG AGC TCT AGA AGC TTC AGG GCG CAA GGG CTG CT-3 ) and KR1 (5 -TCG AGC TCT

AGA AGC TTC AGA AGA ACT CGT CAA GAA G-3 ) which were based on the published sequence (http://www. invitrogen.com/content/sfs/vectors/pcr2 1topo seq.txt). The aphA PCR product was blunt-ended using T4 DNA polymerase and blunt-end ligated at BstXI site (located at position 1419 with reference to the GenBank sequence) of the hemA gene in pARO-hemA/M to create the recombinant plasmid pARO-hemA/M-aphA. PCR was then performed on pARO-hemA/M-aphA using the internal primers 18R3 (5 -CTG TTG GTC GGG GCT GGT GAA-3 ) and VHA-AS4 (5 -AGC GAC GGG CCT TTA GTG C-3 ) which would amplify a region of approximately 256 bp in size inclusive of the BstXI site in the WT

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

hemA gene. The presence of a PCR product of approximately 1.4 kb (results not shown) confirmed the insertion of the aphA gene within the BstXI site of the hemA gene of this recombinant plasmid. The 3.36 kb aphA-mutated hemA/M fragment was excised from pARO-hemA/M-aphA using EcoRI and the excised fragment was blunt-ended using T4 DNA polymerase. The fragment was subcloned into the plasmid pWM91 at the SmaI site by blunt-end ligation to create the recombinant plasmid pWM91-hemA/M-aphA. Transformants were selected by growth on LB-kanamycin and the presence of the aphAmutated hemA in pWM91-hemA/M-aphA was confirmed by colony PCR using 18R3 and VHA-AS4 primers. pWM91-hemA/M-aphA was conjugatively transferred to WT V. cholerae O139 by filter mating. Briefly, donor E. coli cells (BW20767-␭ pir) containing pWM91-hemA/M-aphA and the recipient WT O139 V. cholerae were grown on LB broth with appropriate antibiotics until the OD600 reached 0.4. One milliliter of each donor and recipient cultures were added to a sterile micro-centrifuge tube and centrifuged at 1500 × g for 10 min at room temperature. The pellet was resuspended in 75 ␮l of prewarmed LB broth and spread on a membrane filter (pore size, 0.2 ␮m; 25 mm in diameter, Millipore) placed on the surface of LB agar plate. After incubation at 37 ◦ C for 60 min, the cells were washed from the filter and the merodiploids obtained as a result of the first crossover event were selected on LB agar containing ampicillin, kanamycin and polymyxin B. Finally, merodiploids were grown in modified LB medium containing no NaCl but with 15% sucrose to drive the second recombination event that will result in the replacement of the WT hemA with that of the mutated hemA. Colonies that grew were screened for sensitivity to ampicillin, resistant to kanamycin and polymyxin B and requirement for ALA for growth. One such clone was selected and designated as VCUSM1. To confirm the presence of the aphA-mutated hemA in

3753

VCUSM1, PCR was performed on the genomic DNA using 18R3 and VHA-AS4 primers. The 1.4 kb PCR product was purified and sequenced and the expected result was obtained (Table 2). 2.4. Construction of VCUSM2 It would be ethically important for future human studies to remove the kanamycin resistance marker from VCUSM1. Towards this end, a 992 bp fragment of Green fluorescent protein gene gfp was PCR amplified using primers GFP-F (5 AGTCATTTAAATGATTCATTAATGCAGCTGGC-3 ) and GFP-R (5 -AGTCATTTAAATTTATTTGTAGAGCTCATCCA-3 ) bearing SmiI restriction sites (boldfaced) on their 5 -ends and using pGFPCR plasmid (Gift from Somssich, I.E, Max-Planck-Institut, GenBank accession no. AF007834) as template. The amplified gfp gene was cloned onto BstXI site of pARO-hemA/M and subsequently hemA::gfp-hemM construct was cloned onto pWM91 using similar protocol as described under construction of VCUSM1 (Fig. 1). This recombinant plasmid was then digested with SmiI which resulted in the excision of the gfp gene. The larger fragment was gel-purified and re-ligated to recreate the plasmid but without the gfp gene. This caused a AGT CAT TTA AAT GAC TTG AA sequence insertion at 1419 bp position which in turn resulted in the formation of a stop codon at position 1435 and hence a truncation of the hemA gene. The loss of the gfp gene was confirmed by PCR using 18R3 and VHA-AS4 primers, restriction digest analysis using SmiI and finally by sequencing as shown in Table 2. These genetic manipulations had in effect created a markerless frame-shift mutation in the hemA gene and the recombinant plasmid was designated as pWM91-hemA/M. pWM91-hemA/M was then conjugatively transferred to VCUSM1 using filter mating. The merodiploids obtained as

Table 2 Location of hemA gene mutation in VCUSM2

The nucleotide positions are based on the GenBank sequence, AF227752. The restriction enzyme sites are italized, the inserted sequences that caused +1 frame-shift mutation is underlined and the stop codon formed is in bold. The primer locations are marked as arrows. 18R3 is the forward primer and VHA AS4 is the reverse primer.

3754

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

a result of first crossover event were selected on LB agar plates containing ampicillin, kanamycin, polymyxin B, and ALA. Finally, the merodiploids were grown in modified LB medium containing no NaCl but with 15% sucrose to drive the replacement of the hemA::aphA with the frame-shift mutated hemA. The potential mutants were screened for sensitivity to ampicillin and kanamycin, resistant to polymyxin B and required ALA for growth. One such clone was selected and named as VCUSM2. The presence of the frame-shift hemA mutant gene was confirmed by sequencing as previously described for VCUSM1.

The zone of clearance formed around the colonies were measured. 2.7. Intestinal colonization assay Intestinal colonization assay was performed as described previously [14]. Briefly, WT or VCUSM2 was suspended in LB broth to a final concentration of 1 × 106 cells/50 ␮l and inoculated intragastrically to 3–5-days-old BALB/c mice. The mice were euthanized after 16–18 h and viable vibrios were counted by plating dilutions of homogenized whole gut.

2.5. Phenotypic assays 2.8. Oral immunization 2.5.1. Growth characteristics of VCUSM2 2.5.1.1. Determination of optimal coasting time. Freshly grown VCUSM2 cells in the presence of ALA would contain an endogenous supply of ALA. To determine the time it would take to deplete this supply of ALA (coasting), an overnight culture of VCUSM2 was inoculated in minimal broth (with out ALA) to 1 × 106 CFU and incubated at 37 ◦ C with shaking at 200 rpm. Growth was monitored every 1 h by measuring the turbidity at OD600 . 2.5.1.2. Determination of optimal concentration of ALA. VCUSM2 undepleted or depleted of cellular ALA (by coasted for 6 h) were inoculated into minimal broth to 1 × 106 CFU in the absence and presence of various concentrations of ALA and incubated at 37 ◦ C for 12 h while shaking. Growth was compared with the WT grown in minimal broth in the absence of ALA. Growth was monitored every 1 h by measuring the turbidity at OD600 . 2.5.2. Determination of cholera toxin (CT) production CT production was determined using a GM1 ganglioside dependent enzyme linked immunosorbent assay (ELISA) as described previously [12]. Monoclonal antibodies against CTB subunit were used as primary antibody and anti-rabbit IgG-HRP was used as secondary antibody. The plates were developed with ABTS (22 -azino-di[3-ethyl benzthiazoline sulfate]) solution (1 mg/ml) and absorbance was recorded at 405 nm with reference to 495 nm using Multiskan EX microtiter plate reader. 2.6. Assays for hemagglutinin/hemolysin activity and protease activity

Locally supplied New Zealand white adult rabbits were used in this study. A week before planned immunization, the rabbits were orally administered 125 mg/kg of Metronidazole and 464 mg/kg of Sulfaquinoxaline on days 1 and 3. Three groups of rabbits, comprising of four animals (1–1.5 kg) per group were oro-gastrically inoculated with 1 × 1010 CFU of either VCUSM2 or WT on days 0 and 14. Rabbits were immunized using the protocol described previously [15] except that in order to retard peristalsis, morphine (10 mg/kg) was injected intraperitoneally instead of tincture of opium. Fecal material was also collected every 24 h in alkaline peptone water (APW) and cultured on thiosulphate citrate bile salt (TCBS) agar supplemented with/without ALA to detect the presence of viable VCUSM2/WT cells. 2.9. Serological studies From the immunized rabbits, blood samples were collected for up to 5 weeks at 1 week intervals. Serum was separated and stored at −20 ◦ C in small aliquots without addition of any preservatives. 2.10. Determination of anti-CT IgG and IgA Anti-cholera toxin (CT) antibody levels were measured by ELISA as described previously [16]. Pre and post immune sera (1:10–1:1280 dilution) from rabbits were used as primary antibody and anti-rabbit IgG-HRP or IgA-HRP were used as secondary antibody. Titer was calculated as the highest dilution of test serum producing an OD greater than that of the preimmune sera. 2.11. Determination of anti-LPS IgG and IgA

Hemagglutinin and hemolysin assays were performed as described previously [13]. The titers were defined as the reciprocal of the highest bacterial dilution that caused the hemagglutination and hemolysis after 1 and 12 h, respectively, at 25 ◦ C in the presence of 1% chicken RBCs. To determine the protease activity, freshly subcultured WT or VCUSM2 were streaked on 2.5% skimmed milk agar and the plates were incubated at 30 ◦ C for 24–48 h.

The protocol was essentially the same as described above, except that the plates were coated with 50 ng of purified O139 LPS (kind gift from Fournier MJ, Pasteur Institute) in carbonate buffer (pH 9.6). Pre and post immune sera (1:10–1:1280 dilution) from rabbits were used as primary antibody and anti-rabbit IgG-HRP or IgA-HRP were used as secondary antibody. Titer was calculated as the highest dilution of test

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

serum producing an OD greater than that of the preimmune sera.

3755

nated rabbits were used. The loops were then inoculated with 102 –108 CFU of VCUSM2 or the WT. Further manipulations were identical to that described above.

2.12. Determination of vibriocidal antibodies 2.15. Environmental survival assay Vibriocidal antibodies were detected as previously described [17]. Briefly, WT (indicator strain) were grown in oxoid nutrient broth (ONB) for 4–6 h and diluted with sterile PBS containing 20% pooled normal rabbit serum as complement source to a final concentration of 4 × 104 cells/ml. Serum from immunized or unimmunized rabbits was heat inactivated at 56 ◦ C for 30 min and then diluted in sterile PBS to a final serum dilution of 1:10–1:1280. In a round bottom 96-well microtiter plate, 50 ␮l of indicator cells were mixed with 50 ␮l of diluted immunized or unimmunized rabbits serum and incubated at 37 ◦ C for 1 h. After incubation, 50 ␮l of cell suspension was plated on LB agar plates and the number of viable cells was determined. Titer was calculated as the highest dilution of test serum capable of killing 50% of cells as compared to the preimmune sera. 2.13. Protective efficacy in RITARD and ligated ileal loop models To evaluate the protective efficacy of VCUSM2, orally vaccinated rabbits (on days 0 and 14 with 1010 CFU of VCUSM2 or WT strain) were challenged with 1 log higher dose (1011 CFU) of virulent WT strain using the removable intestinal tie-adult rabbit diarrhoea (RITARD) model as described previously [18] 2 weeks after the second vaccination dose. A control group consisting of unvaccinated animals was similarly challenged but with 109 CFU of the WT strain because our earlier studies showed that a higher dose was too lethal and meaningful data could not be collected. The animals were monitored for diarrhoea or death every 6 h for up to 5 days. The protective efficacy of VCUSM2 was also tested using the ligated ileal loop model. Two weeks after the second vaccination dose, the rabbits were prepared as previously reported [19]. Briefly, several 5 cm loops separated by 1 cm intervals were prepared and the loops were inoculated separately with 102 –108 CFU of the virulent WT strain. The loops were returned to the peritoneum and the abdominal cavity was closed. Rabbits were given a limited amount of water ad libitum. A group of unvaccinated rabbits was also treated in an identical manner. After 18 h, the rabbits were euthanized by intravenous injection of sodium pentobarbitone and the ligated loops were recovered. The volume of fluid accumulated in each loop was recorded as fluid (ml) accumulated per loop by dividing the length of the loop (5 cm). 2.14. Reactogenicity studies Reactogenicity studies were performed using the ligated ileal loop assay as described above except that unvacci-

Environmental survival assays were performed as described previously [10]. Water samples (sea, river, tap and sewage) were collected from different places in Kota Bharu Kelantan, Malaysia. WT and VCUSM2 strains were grown in LB broth (with appropriate supplements) and diluted to 1 × 106 CFU in autoclaved sea, river, tap and sewage water. Inoculated water samples were kept at room temperature and aliquots were plated on appropriate media up to 20 days postinoculation.

3. Results 3.1. Construction of VCUSM1 and VCUSM2 The initial strain for developing the recombinant vaccine strains VCUSM1 and VCUSM2 was a clinical isolate belonging to the O139 Bengal serogroup recovered from a patient with cholera admitted to Hospital Universiti Sains Malaysia (HUSM). This strain was screened and was positive for the presence of ctxAB, ace, zot, tcp and hap genes as revealed by PCR analysis using specific primers. VCUSM1 is mutated at the hemA gene by inserting kanamycin resistance conferring gene, aphA at BstXI site. VCUSM2 is a derivative of VCUSM1 in which hemA::aphA was replaced with frameshift mutated hemA (Fig. 1). Both VCUSM1 and VCUSM2 were unable to grow in the absence of ALA in minimal or LB media. PCR screening of both strains for the presence of various toxin and toxin-associated genes confirmed that these strains were essentially identical to the WT strain apart from the hemA gene. We also observed no differences between VCUSM2 and WT strain in terms of production of CT, hemagglutination, hemolysis and proteolytic activity (data not shown). 3.2. Phenotypic characterization of VCUSM2 VCUSM2 and WT were streaked on LB agar plates with or without 40 ␮g/ml ALA and incubated at 37 ◦ C for 12 h. VCUSM2 failed to grow on LB agar plates in the absence of ALA, whereas LB agar plates supplemented with ALA supported the growth of VCUSM2 comparable to the growth of WT strain. The ability of VCUSM2 to grow in vitro in the absence of exogenous ALA by utilizing its endogenous supply of ALA (coasting) was tested. VCUSM2 continues to multiply for ∼7 generations (5–6 h) in the absence of ALA until its cellular pool of ALA becomes too low for further growth. But the growth of VCUSM2 significantly declined when coasted for 6 and 12 h (Fig. 2).

3756

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761 Table 3 Intestinal colonization capability of VCUSM2 before and after coasting when compared with WT in infant mouse model Strain

Coasting time (h)

Average CFU recovered after inoculating 1 × 106 CFU

Wild type



1.5 × 107

VCUSM2

0 3 6 12

2.45 × 106 1.3 × 106 3.05 × 105 1.5 × 105

Minimum of four suckling mice were included for each strain and for different coasting time. The mice were orally inoculated with 50 ␮l (1 × 106 CFU) of VCUSM2 or WT.

The concentration of ALA required for the optimal growth of VCUSM2 was also studied and is depicted in Fig. 3. The concentration of ALA required for optimal growth, i.e. growth equal to the WT, was found to be 40 ␮g/ml. At this concentration, VCUSM2 and the WT reached the stationary phase after 8 h of growth. This optimal value has also been previously reported for ALA auxotrophs of V. cholerae and E. coli [9,20]. This concentration of ALA was kept constant for all subsequent experiments with VCUSM2.

VCUSM2 was able to colonize the infant mouse intestine, although at reduced levels as compared to the WT (Table 3). On average 2.45 × 106 CFU of VCUSM2 was recovered as compared to 1.5 × 107 CFU of WT and therefore showed a ∼10-fold decrease in the colonization capacity. Similar results have been reported in a thyA mutant of O139 V. cholerae [4]. The effect of VCUSM2 coasting on in vivo colonization was also studied in infant mouse model. VCUSM2 coasted for 0, 3, 6 and 12 h were intragastrically inoculated into infant mice. The results show that the colonization efficiency was reduced further by 10-fold in the 6 and 12 h coasted VCUSM2 strain. This also suggests that infant mouse intestines do not contain free ALA or heme that could support the growth of the VCUSM2.

3.3. Intestinal colonization and effect of coasting time

3.4. Immunological analysis

Intestinal colonization of infant mice has been shown to correlate with the colonization and immunogenicity of live cholera vaccines in humans [21]. Since colonization is critical for elicitation of immune response, VCUSM2 was examined for its colonization ability in the infant mouse model.

The immune response of the VCUSM2 and WT immunized rabbits were evaluated by measuring anti-CT IgG/IgA, anti-LPS IgG/IgA and vibriocidal antibodies. The antibody analysis results are shown in Fig. 4A–E. After immunization with VCUSM2, specific anti-CT IgG/IgA and anti-LPS IgG/IgA titers were appreciably higher by the third week and peak titers were obtained by the fourth week (Fig. 4A–D). A similar antibody profile was seen in animals immunized with the WT strain. These results are very interesting in the sense that despite VCUSM2 having a short life span compared to WT in vivo due to the ALA auxotrophy, it elicited more or less similar immune responses as that of WT. These results also indicate that VCUSM2 is able to colonize and induce reasonably good immune responses against two major antigens of V. cholerae, namely CT and LPS. Vibriocidal antibody is used as an indicator of protection against cholera [21]. In our study, the vibriocidal antibodies were enumerated before and after immunization with VCUSM2 and WT strain. The results are shown in Fig. 4E. The preimmune sera did not show any detectable level of vibriocidal antibody titer confirming that the rabbits were not exposed to V. cholerae or related organisms. A rise in vibriocidal antibody titer was recorded by second week postvaccination (GMT = 225 and 275 for WT and VCUSM2,

Fig. 2. Effect of various coasting times on the growth of VCUSM2 in minimal medium with out ALA. Each value represents the average of triplicate samples. The standard deviations are represented by error bars.

Fig. 3. Effect of different concentration of ALA (10–80 ␮g/ml) on the growth of VCUSM2 were studied. Each OD value represents the average of duplicate samples.

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

3757

Fig. 4. Geometric mean titer (GMT) of anti-CT IgG (A), anti-CT IgA (B), anti-LPS IgG (C), anti-LPS IgA (D) and vibriocidal antibody (E) elicited in rabbits (n = 4) that were orally vaccinated with wild type ( ) or VCUSM2 () (1 × 1010 CFU on days 0 and 14). The error bars represent standard deviations.

respectively) and rose to the peak titer on third week (GMT = 275 for WT and VCUSM2) of post immunization with VCUSM2 and WT strain. However, the titer for the both groups of vaccinated animals declined on the fourth week (GMT = 200 and 175 for WT and VCUSM2, respectively) post immunization. The peak vibriocidal antibody titer on

the third week was the same for the VCUSM2 and WT, indicating that VCUSM2 elicits a protective immune response which was supported by the RITARD results (Table 4). Vibriocidal antibody titer profiles obtained with VCUSM2 were comparable to results obtained with other vaccine strains [4,22].

Table 4 Evaluation of protective efficacy of VCUSM2 in RITARD model Challenge outcome

Mild diarrhea Moderate diarrhea Severe diarrhea Animals death Percent mortality Average fluid present in small intestine (ml) Culture on TCBS (+/−ve) N/A: not applicable, since no fluid was collected.

Challenge dose = 1 × 1011 V. cholerae O139 (WT) Rabbit vaccinated with VCUSM2 n = 4

Rabbit vaccinated with wild type n = 4

Challenge dose = 1 × 109 V. cholerae O139 (WT) Unvaccinated rabbits (control group) n = 4

0 of 4 0 of 4 0 of 4 0 of 4 0% Not present N/A

0 of 4 0 of 4 0 of 4 0 of 4 0% Not present N/A

0 of 4 0 of 4 0 of 4 4 of 4 100% 25–30 ml +ve

3758

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

Table 5 Evaluation of protective efficacy and reactogenicity of VCUSM2 and WT in ligated ileal loops of rabbits Strains

VCUSM2 Vaccinated Unvaccinated (reactogenicity) Wild type Vaccinated Unvaccinated (reactogenicity)

Challenge dose of WT strain and fluid accumulation (ml/cm) 102

103

104

105

106

107

108

<0.1 <0.1

<0.1 <0.1

<0.1 <0.1

<0.1 0.5

<0.1 0.8

<0.1 1.6

<0.1 1.9

<0.1 0.7

<0.1 1.2

<0.1 1.8

<0.1 1.5

<0.1 1.9

<0.1 2.1

<0.1 2.2

The rabbits were vaccinated with 1 × 1010 CFU of VCUSM2 or WT on days 0 and 14.

3.5. Protective efficacy in RITARD and ligated ileal loop models The findings of the challenge study using the RITARD model is summarized in Table 4. No mortality was seen in the rabbits vaccinated with VCUSM2 and WT strain even though these rabbits were challenged with higher dose of the WT strain (1011 ). On the other hand, all rabbits in the unvaccinated group died following the challenge with a dose of 109 WT strain. Necroscopic examination of the dead animals revealed the presence of 25–30 ml of rice water fluid in the small intestine. When cultured on TCBS plates, the fluid contained large number of WT strain (V. cholerae O139). The results clearly indicate that VCUSM2 is as good as WT strain in eliciting natural immune response and has an excellent potential to protect the animals from experimental cholera. The protective efficacy of VCUSM2 and WT strain was also assessed in ligated ileal loops of unvaccinated and vaccinated adult rabbits. The results obtained are summarized in Table 5. Animals vaccinated with VCUSM2 and WT yielded no significant amount of fluid in their intestinal loops at any challenge dose. On the other hand, 0.7–2 ml/cm of fluid was accumulated in the loops of the unvaccinated rabbits. The contents of the loop were culture positive for the presence of WT vibrios. In the case of vaccinated rabbits, we observed the presence of live vibrios in the small amount of fluid that was found in the loops. This showed that the vibrios were not killed, rather they were not able to colonize the small intestine or only at a very low level. Multiplication and secretion of cholera toxin is tightly linked with the establishment of the infection foci. Even if there were some vibrios able to colonize the small intestine, the presence of anti-CT antibodies had neutralized the CT and thus prevented fluid accumulation in the intestinal loops [21]. 3.6. Assessment of the bacterial shedding following the oral vaccination The best marker of intestinal colonization is the excretion of vibrios in rabbit fecal material. A prolonged shedding of

Fig. 5. Bacterial shedding in the fecal pellets of rabbits vaccinated with wild type (WT) and VCUSM2 on day 0 ( ) and 14day () of immunization. The shedding days are obtained from the average of four rabbits in each group. The standard deviations are represented by error bars.

vibrios in coproculture is indicative of successful colonization and subsequent multiplication of the vibrios in the small intestine (the same is true for humans as well). The shedding of the WT and VCUSM2 was assessed by collecting the fecal pellets every 24 h in APW (alkaline peptone water). As can be seen in Fig. 5, the coprocultures of the rabbits orally vaccinated with WT (primary inoculation on day 0) were positive for 9–10 days. However, bacterial shedding lasted for 3 days following the second inoculation on day 14. By contrast, the coprocultures of rabbits vaccinated with VCUSM2 were positive for bacterial shedding for 3–4 days after the initial dose. The bacteria cleared even more rapidly following the second oral inoculation on day 14. These results are in agreement with earlier findings [23,24] that the vibrios are cleared more rapidly following the second inoculation due to the elicitation of the immune response. It has been reported that rabbits vaccinated with 1010 CFU of V. cholerae are resistant to recolonization with the homologous (serovar) strain for up to 8 weeks [23]. There was a marked difference in bacterial shedding between the rabbits orally inoculated with the WT and VCUSM2 on day 0 (10 days versus 4 days). These results confirmed our earlier findings in infant mice and ligated illeal loops that VCUSM2 survive for only a limited period of time due to the unavailability of ALA. 3.7. Reactogenicity of VCUSM2 Reactogenicity studies were performed in ligated ileal loops of unvaccinated rabbits. Table 5 shows the results for reactogenicity studies. The volume of fluid that was accumulated in the loops inoculated with 107 –108 CFU of either WT or VCUSM2 was in the range of 1.6–2.2 ml/cm. However, at the inoculum of 106 CFU, VCUSM2 produced 50% less fluid as compared to the WT (0.8 ml/cm versus 1.9 ml/cm). The difference was more pronounced at lower doses and fluid accumulation in loops injected with 105

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

3759

4. Discussion

Fig. 6. Survival of VCUSM2 () and WT V. cholerae () in river water (A) and sea water (B). Each colony forming units (CFU) value represents the average of triplicate samples. The standard deviations are represented by error bars.

VCUSM2 dropped to 0.5 ml as compared to 1.5 ml/cm in loops injected with the WT strain. No fluid accumulation was found in the loops inoculated with 104 CFU or less of VCUSM2 as compared to the loops inoculated with the WT which still contained 0.7–1.8 ml/cm of fluid. Another important observation was the presence of the bloody mucus in the loops inoculated with the WT strain suggesting damage of epithelial lining of the ileum. This phenomenon was absent in the loops inoculated with VCUSM2 even at doses as high as 108 CFU. Although the reason for this is not clear, these results support the observation that VCUSM2 has greatly reduced reactogenicity compared to the WT strain. 3.8. Environmental safety features The evaluation of the survival capacity of VCUSM2 in environmental waters show that they did not survive for more than 2–3 days in tap or sewage water (data not shown). Slightly better growth was observed in river (Fig. 6A) and sea water (Fig. 6B). Nevertheless, VCUSM2 was severely limited in growth and it could not be detected beyond 8 days in these relatively nutrient rich water samples. On the other hand, WT was detectable beyond 10 days and even after the 20th day post inoculation in river and sea water. These results suggest that biological containment of VCUSM2 is better as compared to the thymidine auxotroph of V. cholerae El Tor which was able to survive for 11 and 18 days in sewage and sea water, respectively [10].

The use of metabolic auxotrophs has been proposed by some researchers as an alternative approach to reduce toxicity in the development of vaccines. However, only a few examples of metabolic auxotrophs exist and they largely failed to gain interest of researchers due to undesired under or over attenuation. One metabolic auxotroph that appeared to be a potential cholera vaccine candidate was V286 originally described by Rijpkema et al. [9]. V286 was a derivative of CVD101 that was isolated after transposon mutagenesis and required exogenous ALA for its survival [9]. The authors noted that V286 was able to colonize and elicit an immune response in adult rabbits, but the genetic basis of ALA auxotrophy in V286 was not known. Much work has been done on ALA auxotrophy in other bacterial species such as E. coli and Salmonella typhimurium [11,20,25,26]. The available data suggested that mutation in the hemA gene might be responsible for the ALA auxotrophy. For the first time in the literature, we have isolated and cloned the hemA and hemM genes of V. cholerae O139 Bengal (Accession number AF227752). In this study, we have successfully developed an ALA auxotrophic vaccine strain, VCUSM2, a mutant of O139 V. cholerae where the wild type hemA gene is replaced with a markerless, frame-shifted version of hemA. VCUSM2 elicited high titer of anti-LPS IgG/IgA antibodies on the fourth week post-vaccination, similarly an increase in titer of anti-CT IgG/IgA was also observed (Fig. 4A–D). These antibody titers were comparable with the antibody titer elicited by the WT strain. The mechanism by which naturally immunized (after an earlier episode of cholera) and vaccinated humans become resistant to a subsequent cholera infection is believed to involve immune responses against anti-LPS and other surface antigens. Anti-CT immunity also does play a role in the prevention of the onset of cholera by neutralizing the CT secreted by the vibrios [21]. High levels of vibriocidal antibody titer have been used as a marker for protection against cholera [21,27]. These antibodies are shown to be primarily directed towards ‘O’ or LPS antigens of V. cholerae. The importance of ‘O’ antigen is obvious by the fact that immunity against the O1 organism does not cross protect against O139 cholera and vice versa [3]. Moreover, studies by Mukhopadhyay et al. have shown that anti-LPS antibodies and their IgG fractions can mediate protection through inhibition of intestinal adherence and colonization activities of vibrios [28]. Our results indicate that VCUSM2 was able to elicit an early onset of high levels of vibriocidal antibodies. The data show that vibriocidal antibody titer began to rise by the second week of immunization and remains high up to the third week. The vibriocidal antibody level elicited by VCUSM2 was similar to that elicited by the WT strain on the third week. Our results are also comparable with two other O139 vaccine candidate, L911T and L912T, which attained peak vibriocidal titers of 1:320 and 1:80, respectively [4]. These data strongly suggest

3760

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761

that VCUSM2 vaccine candidate induces good antibacterial immunity. Previous studies by two independent groups of researchers have also shown that a combination of V. cholerae LPS and CT significantly raises protection against cholera and that the antibacterial and antitoxic immunities acted synergistically by interfering with two separate events in cholera pathogenesis [29,30]. The immunological studies of VCUSM2 show that is able to elicit both the antibacterial and antitoxic affect which is an important prerequisite for a potential vaccine candidate. The immunological studies of VCUSM2 were further supported by animal studies. RITARD challenge assays show that VCUSM2-vaccinated animals were 100% protected (Table 4) when challenged with the high dose (1011 CFU) of WT strain. The results of the ligated ileal loop assay were also consistent with the results obtained by the RITARD assay. The presence of live WT strains in the loops of vaccinated rabbits, although no significant fluid accumulation was observed, is suggestive of anti-LPS IgG antibodies in the lumen which inhibits intestinal adherence and colonization of vibrios [28]. It has been shown that natural infection derived immunity offers the best protection against cholera [31]. VCUSM2 has all virulence genes intact, allowing the full repertoire of antigens to be presented, which mimics natural infection, and was thus able to elicit a good immune response. Despite the functionality of these genes, the reactogenicity of this strain is greatly reduced due to the limited multiplication in vivo because of ALA auxotrophy. In addition, VCUSM2’s limited life span in environmental waters is a useful containment property. The ALA requirement for VCUSM2 (hemA mutant of V. cholerae) was similar to that of hemA mutants of E. coli and S. typhimurium indicating that these organisms have similar heme metabolic processes [20,26]. V. cholerae are reported to have at least three heme receptors, i.e. HutA, HutR and HasR. The presence of three functional heme receptors in this organism suggests that heme is an important nutrient for V. cholerae [32]. Survival of hemA mutants (such as VCUSM2) in the presence of potential sources of heme in the small intestine is definitely not a desired feature. It is well documented that vibrios secrete a number of proteases including hemolysins [21]. There is a theoretical possibility that once hemA mutants colonize the small intestine, they will secrete these proteases and cause damage to the epithelial cells. As a result, heme will be released from the damaged cells that will help the hemA mutants to survive. However, experiments carried out with VCUSM2 in the infant mice and ligated ileal loops of the rabbits suggest that this did not happen. The results obtained in this study clearly indicated a compromised growth of VCUSM2 in the small intestines of infant mice and adult rabbits. Although colonization efficiency was reduced, VCUSM2 was still able to induce antibody responses comparable to that of the WT strain. It is hoped that VCUSM2 vaccine candidate will exhibit similar characteristics in the human intestine. If compromised

growth of VCUSM2 is not seen in human trials, mutating the hutA, hutR and hasR genes encoding for HutA, HutR and HasR heme receptors would be considered. In conclusion, we believe that VCUSM2 is a promising, orally administrable, highly immunogenic, minimally toxic and environmentally safe vaccine candidate against V. cholerae O139. Work is ongoing in our laboratory to further reduce the reactogenicity of VCUSM2 and to construct similar mutants for the El Tor biotype of V. cholerae.

Acknowledgements This study was supported by a grant from the Ministry of Science, Technology and Innovation, Malaysia. We thank F. Mooie for discussions with Zainul F. Zainuddin which resulted in this project.

References [1] Faruque SM, Albert MJ, Mekalanos JJ. Epidemiology, genetics, and ecology of toxigenic Vibrio cholerae. Microbiol Mol Biol Rev 1998;62(4):1301–14. [2] Albert MJ, Siddique AK, Islam MS, Faruque AS, Ansaruzzaman M, Faruque SM, et al. Large outbreak of clinical cholera due to Vibrio cholerae non-O1 in Bangladesh. Lancet 1993;341(8846):704. [3] Albert MJ, Alam K, Ansaruzzaman M, Qadri F, Sack RB. Lack of cross-protection against diarrhea due to Vibrio cholerae O139 (Bengal strain) after oral immunization of rabbits with V. cholerae O1 vaccine strain CVD103-HgR. J Infect Dis 1994;169(1): 230–1. [4] Ledon T, Valle E, Valmaseda T, Cedre B, Campos J, Rodriguez BL, et al. Construction and characterisation of O139 cholera vaccine candidates. Vaccine 2003;21(11–12):1282–91. [5] Tacket CO, Losonsky G, Nataro JP, Comstock L, Michalski J, Edelman R, et al. Initial clinical studies of CVD 112 Vibrio cholerae O139 live oral vaccine: safety and efficacy against experimental challenge. J Infect Dis 1995;172(3):883–6. [6] Coster TS, Killeen KP, Waldor MK, Beattie DT, Spriggs DR, Kenner JR, et al. Safety, immunogenicity, and efficacy of live attenuated Vibrio cholerae O139 vaccine prototype. Lancet 1995;345(8955):949–52. [7] Attridge SR. Thymine auxotrophy as an attenuating marker in Vibrio cholerae. Microb Pathog 1995;19(1):11–8. [8] Ryan ET, Crean TI, Kochi SK, John M, Luciano AA, Killeen KP, et al. Development of a Delta glnA balanced lethal plasmid system for expression of heterologous antigens by attenuated vaccine vector strains of Vibrio cholerae. Infect Immun 2000;68(1):221–6. [9] Rijpkema SG, Bik EM, Jansen WH, Gielen H, Versluis LF, Stouthamer AH, et al. Construction and analysis of a Vibrio cholerae delta-aminolevulinic acid auxotroph which confers protective immunity in a rabbit model. Infect Immun 1992;60(6):2188–93. [10] Valle E, Ledon T, Cedre B, Campos J, Valmaseda T, Rodriguez B, et al. Construction and characterization of a nonproliferative El Tor cholera vaccine candidate derived from strain 638. Infect Immun 2000;68(11):6411–8. [11] Verderber E, Lucast LJ, Van Dehy JA, Cozart P, Etter JB, Best EA. Role of the hemA gene product and delta-aminolevulinic acid in regulation of Escherichia coli heme synthesis. J Bacteriol 1997;179(14):4583–90. [12] Gardel CL, Mekalanos JJ. Regulation of cholera toxin by temperature, pH, and osmolarity. Methods Enzymol 1994;235:517–26.

M. Ravichandran et al. / Vaccine 24 (2006) 3750–3761 [13] Nandy RK, Sengupta TK, Mukhopadhyay S, Ghose AC. A comparative study of the properties of Vibrio cholerae O139, O1 and other non-O1 strains. J Med Microbiol 1995;42(4):251–7. [14] Angelichio MJ, Spector J, Waldor MK, Camilli A. Vibrio cholerae intestinal population dynamics in the suckling mouse model of infection. Infect Immun 1999;67(8):3733–9. [15] Butterton JR, Beattie DT, Gardel CL, Carroll PA, Hyman T, Killeen KP, et al. Heterologous antigen expression in Vibrio cholerae vector strains. Infect Immun 1995;63(7):2689–96. [16] Cooper GN, McNab CE, Jackson GD, Walker P. Enzyme-linked immunoassays for antibodies against Vibrio cholerae. Aust J Exp Biol Med Sci 1983;61(Pt 6):713–25. [17] Attridge SR, Qadri F, Albert MJ, Manning PA. Susceptibility of Vibrio cholerae O139 to antibody-dependent, complement-mediated bacteriolysis. Clin Diagn Lab Immunol 2000;7(3):444–50. [18] Spira WM, Sack RB, Froehlich JL. Simple adult rabbit model for Vibrio cholerae and enterotoxigenic Escherichia coli diarrhea. Infect Immun 1981;32(2):739–47. [19] Liang W, Wang S, Yu F, Zhang L, Qi G, Liu Y, et al. Construction and evaluation of a safe, live, oral Vibrio cholerae vaccine candidate, IEM108. Infect Immun 2003;71(10):5498–504. [20] Chen W, Russell CS, Murooka Y, Cosloy SD. 5-Aminolevulinic acid synthesis in Escherichia coli requires expression of hemA. J Bacteriol 1994;176(9):2743–6. [21] Kaper JB, Morris Jr JG, Levine MM. Cholera. Clin Microbiol Rev 1995;8(1):48–86. [22] Eko FO, Schukovskaya T, Lotzmanova EY, Firstova VV, Emalyanova NV, Klueva SN, et al. Evaluation of the protective efficacy of Vibrio cholerae ghost (VCG) candidate vaccines in rabbits. Vaccine 2003;21(25–26):3663–74. [23] Cray Jr WC, Tokunaga E, Pierce NF. Successful colonization and immunization of adult rabbits by oral inoculation with Vibrio cholerae O1. Infect Immun 1983;41(2):735–41. [24] Pierce NF, Kaper JB, Mekalanos JJ, Cray Jr WC, Richardson K. Determinants of the immunogenicity of live virulent and mutant Vibrio cholerae O1 in rabbit intestine. Infect Immun 1987;55(2):477–81.

3761

[25] Choi P, Wang L, Archer CD, Elliott T. Transcription of the glutamyl-tRNA reductase (hemA) gene in Salmonella typhimurium and Escherichia coli: role of the hemA P1 promoter and the arcA gene product. J Bacteriol 1996;178(3):638–46. [26] Elgrably-Weiss M, Park S, Schlosser-Silverman E, Rosenshine I, Imlay J, Altuvia S. A Salmonella enterica serovar typhimurium hemA mutant is highly susceptible to oxidative DNA damage. J Bacteriol 2002;184(14):3774–84. [27] Glass RI, Svennerholm AM, Khan MR, Huda S, Huq MI, Holmgren J. Seroepidemiological studies of El Tor cholera in Bangladesh: association of serum antibody levels with protection. J Infect Dis 1985;151(2):236–42. [28] Mukhopadhyay S, Nandi B, Ghose AC. Antibodies (IgG) to lipopolysaccharide of Vibrio cholerae O1 mediate protection through inhibition of intestinal adherence and colonisation in a mouse model. FEMS Microbiol Lett 2000;185(1):29–35. [29] Bondre VP, Sinha VB, Srivastava BS. Evaluation of different subcellular fractions of Vibrio cholerae O139 in protection to challenge in experimental cholera. FEMS Immunol Med Microbiol 1998;19(4):323–9. [30] Svennerholm AM, Holmgren J. Synergistic protective effect in rabbits of immunization with Vibrio cholerae lipopolysaccharide and toxin/toxoid. Infect Immun 1976;13(3):735–40. [31] Levine MM, Nalin DR, Craig JP, Hoover D, Bergquist EJ, Waterman D, et al. Immunity of cholera in man: relative role of antibacterial versus antitoxic immunity. Trans R Soc Trop Med Hyg 1979;73(1):3–9. [32] Mey AR, Payne SM. Haem utilization in Vibrio cholerae involves multiple TonB-dependent haem receptors. Mol Microbiol 2001;42(3):835–49. [33] Metcalf WW, Jiang W, Daniels LL, Kim SK, Haldimann A, Wanner BL. Conditionally replicative and conjugative plasmids carrying lacZ alpha for cloning, mutagenesis, and allele replacement in bacteria. Plasmid 1996;35(1):1–13. [34] Parke D. Construction of mobilizable vectors derived from plasmids RP4, pUC18 and pUC19. Gene 1990;93(1):135–7.