- Email: [email protected]
Construction and evaluation of a eukaryotic expression plasmid for stable delivery using attenuated Salmonella
Microbial Pathogenesis 34 (2003) 115–119 www.elsevier.com/locate/micpath
Construction and evaluation of a eukaryotic expression plasmid for stable delivery using attenuated Salmonella Helen S. Garmorya,*, Richard W. Titballa,b, Katherine A. Brownc, Alice M. Bennetta a
Department of Biomedical Sciences, Dstl Chemical and Biological Sciences, Porton Down, Salisbury SP4 0JQ, UK Department of Biological Sciences, Centre for Molecular Microbiology and Infection, Imperial College of Science, Technology and Medicine, London SW7 2AY, UK c Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK b
Received 31 July 2002; accepted 3 October 2002
Abstract An approach to enhancing the stability of eukaryotic expression plasmids for delivery using attenuated Salmonella has been evaluated. The expression apparatus and b-galactosidase gene from the expression plasmid, pCMVb, was cloned into the low copy number plasmid pLG339. The resulting construct, pLGbGAL, was shown to have a lower copy number than pCMVb in Salmonella enterica var Typhimurium aroA strain SL7207. Furthermore, b-galactosidase-specific antibody was induced in mice following intramuscular inoculation with pLGbGAL as naked DNA. Following oral administration of mice with SL7207/pCMVb, recombinants could not be detected in tissues 3 days after inoculation. In comparison, SL7207/pLGbGAL recombinant bacteria could be detected in the Peyer’s patches and spleens indicating that the Salmonella strain was stable. However, both SL7207/pCMVb and SL7207/pLGbGAL failed to induce b-galactosidase-specific IgG in vivo. The mechanism by which attenuated Salmonella are able to release heterologous DNA for antigen processing and presentation is not yet understood. These results suggest that the mechanism needs to be further elucidated in order to rationally improve the system. q 2003 Elsevier Science Ltd. All rights reserved. Keywords: Salmonella; Expression plasmid; Copy number
1. Introduction Naked DNA immunisation has recently provided a promising approach to vaccination [1]. Such DNA usually consists of plasmid vectors encoding antigens under the control of eukaryotic promoters to allow expression in eukaryotic cells. The DNA is commonly dissolved in saline and injected intramuscularly. Alternatively, it may be coated onto gold particles and introduced into the dermis using a gene gun. Such administration of DNA has been shown to stimulate strong immune responses, including both cellular and humoral responses. Naked DNA immunisation offers many advantages which include the relative ease of production, purity and stability of the inoculum, the ability to alter the DNA to manipulate the immune response, * Corresponding author. Tel.: þ 44-1980-614755; fax: þ 44-1980614307. E-mail address: [email protected] (H.S. Garmory).
and the lack of obvious side effects. More recently, reports have described the use of attenuated Salmonella as a carrier system for the in vivo delivery of eukaryotic expression plasmids, including DNA vaccines and gene therapy vectors [2]. Although the use of Salmonella for this purpose was first described in 1997 [3], the technique has apparently not been used widely. Yet this delivery system offers several potential advantages over direct administration of naked DNA. First, Salmonella-based plasmids may be administered via the natural route of the bacterium, i.e. orally [2]. This delivery route may be more acceptable for potentially widespread use. In addition, this delivery system targets the mucosal-associated lymphoid system, improving the possibility of stimulating mucosal immunity which may be important in protection against some infectious agents. Via the natural route of entry, Salmonella infect or are taken up by antigen presenting cells (APC’s) such as macrophages or dendritic cells [4], thereby potentially bringing the expression plasmids into contact
0882-4010/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0882-4010(02)00176-6
116
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
with immune cells for antigen processing and presentation. However, when studied, no recombinant Salmonella were detected in host tissues following oral inoculation with the bacteria [3]. It is known that Salmonella strains may be unstable when harbouring high copy number plasmids [5]. In this study we have set out to evaluate the feasibility of using Salmonella for the delivery of naked DNA plasmids with high or low copy numbers.
2. Results 2.1. Construction of Salmonella containing a low copy number expression plasmid A DNA fragment of pCMVb which includes the human cytomegalovirus major immediate-early (CMV IE) promoter, b-galactosidase gene and poly A sequence, was isolated by digestion with the restriction enzymes EcoRI and SalI and subsequently cloned into EcoRI/SalI digested pLG339, resulting in the low copy number expression plasmid pLGbGAL. Plasmids pCMVb, pLGbGAL and pLG339 were passaged through S. enterica var Typhimurium LB5010 to ensure modification and electroporated into S. enterica var Typhimurium SL7207. The copy numbers of the plasmids in the resulting strains were evaluated by polymerase chain reaction (PCR) amplification of the b-galactosidase gene. PCR of SL7207/pCMVb and SL7207/pLGbGAL colony lysates with primers BGAL/F09 and POLYA/R01 resulted in a PCR product of approximately 800 bp, consistent with the predicted amplification product of 789 bp. The PCR product from SL7207/pLGbGAL appeared to be less intense than that from SL7207/pCMVb, indicating that the copy number of pLGbGAL was less than that of pCMVb (data not shown). In comparison, PCR amplification of aroB from
SL7207/pCMVb and SL7207/pLGbGAL resulted in similar intensity PCR products from both strains, indicating similar copy numbers. 2.2. Inoculation with high and low copy number plasmids expressing b-galactosidase The use of plasmids pCMVb and pLGbGAL to induce an antibody response to b-galactosidase was evaluated. One hundred micrograms of pCMVb or pLGbGAL, prepared from Escherichia coli TOP10 cells, were used to intramuscularly (i.m.) inoculate groups of 6 BALB/c mice on days 0 and 14. Enzyme linked immunosorbant assay (ELISA) analysis of sera removed from the mice on day 56 showed that both plasmids induced b-galactosidase-specific IgG responses in all mice (Fig. 1). 2.3. Evaluation of SL7207 harbouring the expression plasmids The stability of plasmid pLGbGAL in SL7207 was evaluated by studying colonisation of the recombinant strain in murine tissues. Groups of 6 BALB/c mice were intragastrically (i.g.) inoculated with 1 – 5 £ 109 cfu of SL7207, SL7207/pLGbGAL, SL7207/pLG339 or SL7207/pCMVb. Peyer’s patches and spleens were harvested on day 3 and the number of Salmonella present was determined by spreading onto LB agar or LB agar containing antibiotic. Following inoculation with SL7207/pCMVb, no Salmonella could be found in any of the tissues sampled. In comparison, SL7207/pLGbGAL, SL7207/pLG339 and the parent strain SL7207 colonised tissues to a similar level (Fig. 2). There was no significant difference between colonisation of these strains, and enumeration of bacteria in homogenised tissues showed
Fig. 1. b-galactosidase specific IgG responses induced in BALB/c mice (n ¼ 6) i.m. inoculated two times, 2 weeks apart with 100 mg of pCMVb or pLGbGAL. IgG concentrations were determined by ELISA of serum samples removed on day 56. Error bars represent the standard error of the mean.
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
Fig. 2. Colonisation of Peyer’s patches and spleens following i.g. inoculation of BALB/c mice (n ¼ 6) with SL7207 (squares), SL7207/pLG339 (triangles) or SL7207/pLGbGAL (circles). Organs were removed 3 days after inoculation and homogenised in PBS before plating onto LB agar (closed symbols) or LB agar containing kanamycin (open symbols) for enumeration of bacteria. Horizontal lines show the mean value of individual samples.
that pLGbGAL and pLG339 were both stable in SL7207 in vivo (99% and 100%, respectively). The immunogenicity of SL7207/pLGbGAL was also assessed. Groups of six mice were i.g. inoculated four times at 2-week intervals with 1 – 5 £ 109 cfu of SL7207/pCMVb, SL7207/pLGbGAL, SL7207/pLG339, or SL7207. Further groups of mice were again inoculated i.m. with 100 mg of the eukaryotic expression plasmids pCMVb or pLGbGAL. Although b-galactosidase-specific IgG responses were again induced by inoculation with these DNA plasmids, inoculation with SL7207/pCMVb or SL7207/pLGbGAL did not induce detectable b-galactosidase-specific IgG (data not shown). In comparison, SL7207-specific IgG responses were found in all mice inoculated with any of the Salmonella strains used in this study (data not shown).
3. Discussion and conclusions The technology of using Salmonella to deliver eukaryotic expression plasmids is relatively new (recently reviewed in [4]). One model eukaryotic expression plasmid used in some studies is pCMVb which expresses b-galactosidase under the control of the CMV IE promoter [3,6 – 9]. In these published studies the aroA attenuated S. enterica var Typhimurium strain SL7207 was used to orally deliver pCMVb to macrophages and dendritic cells in vivo, resulting in the induction of a range of b-galactosidase-specific immune responses, including IgG and CD4þ and CD8þ T cell activity. b-galactosidase activity could be detected 5 weeks after the last boost, although recombinant SL7207 could not be detected in vivo at any time [3]. In our initial studies we were unable to reproduce the results obtained by oral administration of SL7207/pCMVb, using the methodology described by
117
Paglia et al. [6]. No detectable b-galactosidase-specific IgG was induced and, like other researchers [3], we were unable to detect recombinant bacteria in vivo. Thus, in an effort to stabilise the b-galactosidase-expressing plasmid for delivery in Salmonella, we cloned the CMV IE promoter, b-galactosidase gene and polyA sequence from pCMVb into a low copy number plasmid, pLG339. The resulting construct, termed pLGbGAL, was shown to have a copy number less than that of pCMVb in SL7207 and induced a b-galactosidase-specific antibody response when intramuscularly inoculated into mice as naked DNA (Fig. 1). Furthermore, SL7207/pLGbGAL was shown to be stable in vivo and could be recovered from Peyer’s patches and spleens of intragastrically inoculated mice (Fig. 2). However, the improved stability of the b-galactosidaseexpressing eukaryotic expression plasmid in SL7207 in vivo did not improve the induction of b-galactosidasespecific IgG responses in mice. The lack of induction of anti-b-galactosidase IgG following delivery of SL7207/pCMVb in this study contrasts with previously published work [6], indicating that Salmonella delivery of DNA vaccines may not be a robust technology. Furthermore, administration of SL7207/pLGbGAL also failed to induce a detectable antibody response. It is possible that reducing the copy number of the plasmid results in the release of less plasmid from the Salmonella than is required to stimulate immunity. However, the findings of this study raise further questions regarding the methodology by which Salmonella strains are able to deliver eukaryotic expression plasmids. Following oral or nasal administration, recombinant Salmonella invade through the M cells of the Peyer’s patches in the gut or through M-like cells in the nasal passages. Here they encounter and are phagocytosed by resident APC’s, including macrophages and dendritic cells. In general, studies of microbial pathogenesis have increased our understanding of the mechanisms employed by bacteria to access cellular compartments. For example, Shigella flexneri are able to escape from phagolysosomes into the host cell cytosol [10]. In comparison, it is known that Salmonella spp. remain in the phagosomal compartment. Thus, it is somewhat surprising that Salmonella are able to successfully deliver expression plasmids. It has been hypothesised that Salmonella infection may cause leakage of the phagolysosome [2], although the mechanism of DNA transport in vitro appears to operate only in mature cells and not in cell lines [3]. In summary, we were able to produce a low copy number eukaryotic vector expressing b-galactosidase that stimulated a specific immune response in vivo. Furthermore we were able to show that the vector was stable in Salmonella in vivo. Recombinant Salmonella were detected in both Peyer’s patches and spleens of infected mice. Although it is known that APC’s in Peyer’s patches and spleens are involved in the presentation of antigen via both MHC class I and II
118
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
molecules, we were unable to detect any induction of bgalactosidase-specific IgG responses. These results indicate that a further understanding of the mechanism by which Salmonella delivery of DNA occurs is required before a rational approach to vaccine design may be achieved.
4. Materials and methods 4.1. Plasmids, bacterial strains and growth conditions Plasmid pCMVb (copy number , 500 – 700/cell; Clontech Laboratories, Basingstoke, UK) contains the gene encoding b-galactosidase and a gene conferring resistance to ampicillin. Plasmid pLG339 [11] is a low copy number (, 5 copies/cell) plasmid containing a gene conferring resistance to kanamycin. E. coli TOP10 (Stratagene, California, US) was used for cloning and plasmid preparation. S. enterica var Typhimurium LB5010 (r2mþ) was described previously [12]. S. enterica var Typhimurium SL7207 [3] was kindly provided by Professor B.A.D. Stocker (Stanford University School of Medicine, California, US). E. coli and S. enterica var Typhimurium strains were routinely cultured in LB broth or on LB agar, supplemented with antibiotic if required. LB medium was supplemented with 1% (v/v) histidine, 1% (v/v) ‘aromix’ (4 mg/ml tyrosine, 4 mg/ml phenylalanine, 1 mg/ml p-aminobenzoic acid, 1 mg/ml dihydroxybenzoic acid) for growth of SL7207 strains. To produce inocula, SL7207 strains were cultured statically overnight, centrifuged (6000g, 4 8C, 20 min), washed in phosphate buffered saline (PBS) and resuspended in LB broth to a final cell density of approximately 1– 5 £ 1010 cfu/ml. Viable bacteria were enumerated on LB agar plates, supplemented with antibiotic if required. 4.2. Recombinant DNA techniques All enzymes were obtained from Roche Biochemicals (Lewes, UK) and used according to manufacturer’s instructions. DNA restriction fragments were isolated using a QIAquick Gel Extraction kit (QIAGEN Inc., Crawley, UK). Ligation mixtures were electroporated into E. coli TOP10 and recombinants were screened by PCR amplification (95 8C for 2 min; 30 cycles of 95 8C for 20 s, 60 8C for 20 s, 72 8C for 30 s; 72 8C for 10 min) of colony lysates using the primer pair BGAL/F09 (50 -AGGGCCGCAAGAAAACTAT-30 ) and POLYA/R01 (50 -TGTTTATTGCAGCTTATAATGGT-30 ), and Taq DNA polymerase (Roche). Copy numbers of genes were evaluated by PCR amplification (20 cycles, as described above) of colony lysates with BGAL/F09 and POLYA/R01 or with primer pair AROB/F2 (50 -ACATGATTGGCGCGTTTTACC-30 ) and AROB/R2 (5 0 -CTCCGCCGCGCACTTCACTTTTCCCTATG-3 0 ). Recombinant DNA for use in vivo was extracted from TOP10 cells using the Endofree Plasmid Mega Prep kit (QIAGEN Inc., Crawley, UK).
4.3. In vivo studies Six to eight-week old female BALB/c mice (Charles River Laboratories, Kent, UK) were i.g. inoculated with 100 ml Salmonella using an oral gavage needle. Alternatively, they were i.m. inoculated with DNA plasmids resuspended in 100 ml PBS containing 0.25% of the anaesthetic bupivicaine hydrochloride. Fifty microlitre DNA was injected into each hind leg. To study colonisation of the Salmonella, mice were culled 3 days after inoculation by cervical dislocation and Peyer’s patches (6 per mouse) and spleens were removed. Tissues were homogenised in 1 ml PBS using 50 mm cell strainers (Becton Dickinson Labware, New Jersey, US) and spread onto LB agar or LB agar containing antibiotic for enumeration. To study the immunogenicity of the recombinant Salmonella and expression plasmids, mice were inoculated on days 0, 14, 28 and 42, and serum samples were taken on days 12, 26, 40 for analysis. 4.4. Enzyme-linked immunosorbant assay Serum antibody responses against b-galactosidase were assessed using a modified ELISA. Briefly, 96-well microtitre plates were coated overnight at 4 8C with purified b-galactosidase (5 mg/ml in PBS) and then blocked for 1 h at 37 8C with 1% (w/v) skimmed milk powder in PBS (BLOTTO). The test sera were serially diluted in BLOTTO in duplicate on the plates and incubated for 1 h at 37 8C. Washes between steps were performed three times with PBS containing 0.05% (v/v) Tween-20. Bound antibody was detected following incubation with a HRP-conjugated secondary antibody against mouse IgG, diluted 1:4000 in BLOTTO, for 1 h at 37 8C. 2,20 -azino bis(3-ethylbenzthiazoline-6-sulfonic acid) substrate (ABTS) was added and the plates were incubated at room temperature for 20 min. Serum antibody concentrations were calculated in ng/ml using a standard curve. Concentrations were calculated using Revelatione (Dynex Technologies Ltd, West Sussex, UK).
References [1] Davis HL. Plasmid DNA expression systems for the purpose of immunization. Curr Opin Biotech 1997;8:635–40. [2] Weiss S, Krusch S. Bacteria-mediated transfer of eukaryotic expression plasmids into mammalian host cells. Biol Chem 2001; 382:533–41. [3] Darji A, Guzma´n CA, Gerstel B, et al. Oral somatic transgene vaccination using attenuated S. typhimurium. Cell 1997;91:765– 75. [4] Garmory HS, Brown KA, Titball RW. Salmonella vaccines for use in humans: present and future perspectives. FEMS Microbiol Rev 2002; 339 –353. [5] Sirard J-C, Niedergang F, Kraehenbuhl J-P. Live attenuated Salmonella: a paradigm of mucosal vaccines. Immunol Rev 1999; 171:5–26. [6] Paglia P, Medina E, Arioli I, Guzman CA, Colombo MP. Gene transfer in dendritic cells, induced by oral DNA vaccination with
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119 Salmonella typhimurium, results in protective immunity against a murine fibrosarcoma. Blood 1998;92:3172–6. [7] Paglia P, Terrazzini N, Schulze K, Guzma´n CA, Colombo MP. In vivo correction of genetic defects of monocyte/macrophages using attenuated Salmonella as oral vectors for targeted gene delivery. Gene Ther 2000;7:1725–30. [8] Darji A, zur Lage S, Garbe AI, Chakraborty T, Weiss S. Oral delivery of DNA vaccines using attenuated Salmonella typhimurium as carrier. FEMS Immunol Med Microbiol 2000;27:341–9. [9] Shata MT, Reitz Jr MS, DeVico AL, Lewis GK, Hone DM. Mucosal and systemic HIV-1 Env-specific CD8þ T-cells develop after
119
intragastric vaccination with a Salmonella env DNA vaccine vector. Vaccine 2002;20:623–9. [10] Weiss S, Krusch S. Bacteria-mediated transfer of eukaryotic expression plasmids into mammalian host cells. Biol Chem 2001; 382:533–41. [11] Stoker NG, Fairweather N, Spratt BG. Versatile low-copy-number plasmid vectors for cloning in Escherichia coli. Gene 1982;18: 335–41. [12] Bullas LR, Ryu JI. Salmonella typhimurium LT2 strains which are r2mþ for all three chromosomally located systems of DNA restriction and modification. J Bacteriol 1983;156:471 –4.
Construction and evaluation of a eukaryotic expression plasmid for stable delivery using attenuated Salmonella Helen S. Garmorya,*, Richard W. Titballa,b, Katherine A. Brownc, Alice M. Bennetta a
Department of Biomedical Sciences, Dstl Chemical and Biological Sciences, Porton Down, Salisbury SP4 0JQ, UK Department of Biological Sciences, Centre for Molecular Microbiology and Infection, Imperial College of Science, Technology and Medicine, London SW7 2AY, UK c Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, London WC1E 7HT, UK b
Received 31 July 2002; accepted 3 October 2002
Abstract An approach to enhancing the stability of eukaryotic expression plasmids for delivery using attenuated Salmonella has been evaluated. The expression apparatus and b-galactosidase gene from the expression plasmid, pCMVb, was cloned into the low copy number plasmid pLG339. The resulting construct, pLGbGAL, was shown to have a lower copy number than pCMVb in Salmonella enterica var Typhimurium aroA strain SL7207. Furthermore, b-galactosidase-specific antibody was induced in mice following intramuscular inoculation with pLGbGAL as naked DNA. Following oral administration of mice with SL7207/pCMVb, recombinants could not be detected in tissues 3 days after inoculation. In comparison, SL7207/pLGbGAL recombinant bacteria could be detected in the Peyer’s patches and spleens indicating that the Salmonella strain was stable. However, both SL7207/pCMVb and SL7207/pLGbGAL failed to induce b-galactosidase-specific IgG in vivo. The mechanism by which attenuated Salmonella are able to release heterologous DNA for antigen processing and presentation is not yet understood. These results suggest that the mechanism needs to be further elucidated in order to rationally improve the system. q 2003 Elsevier Science Ltd. All rights reserved. Keywords: Salmonella; Expression plasmid; Copy number
1. Introduction Naked DNA immunisation has recently provided a promising approach to vaccination [1]. Such DNA usually consists of plasmid vectors encoding antigens under the control of eukaryotic promoters to allow expression in eukaryotic cells. The DNA is commonly dissolved in saline and injected intramuscularly. Alternatively, it may be coated onto gold particles and introduced into the dermis using a gene gun. Such administration of DNA has been shown to stimulate strong immune responses, including both cellular and humoral responses. Naked DNA immunisation offers many advantages which include the relative ease of production, purity and stability of the inoculum, the ability to alter the DNA to manipulate the immune response, * Corresponding author. Tel.: þ 44-1980-614755; fax: þ 44-1980614307. E-mail address: [email protected] (H.S. Garmory).
and the lack of obvious side effects. More recently, reports have described the use of attenuated Salmonella as a carrier system for the in vivo delivery of eukaryotic expression plasmids, including DNA vaccines and gene therapy vectors [2]. Although the use of Salmonella for this purpose was first described in 1997 [3], the technique has apparently not been used widely. Yet this delivery system offers several potential advantages over direct administration of naked DNA. First, Salmonella-based plasmids may be administered via the natural route of the bacterium, i.e. orally [2]. This delivery route may be more acceptable for potentially widespread use. In addition, this delivery system targets the mucosal-associated lymphoid system, improving the possibility of stimulating mucosal immunity which may be important in protection against some infectious agents. Via the natural route of entry, Salmonella infect or are taken up by antigen presenting cells (APC’s) such as macrophages or dendritic cells [4], thereby potentially bringing the expression plasmids into contact
0882-4010/03/$ - see front matter q 2003 Elsevier Science Ltd. All rights reserved. doi:10.1016/S0882-4010(02)00176-6
116
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
with immune cells for antigen processing and presentation. However, when studied, no recombinant Salmonella were detected in host tissues following oral inoculation with the bacteria [3]. It is known that Salmonella strains may be unstable when harbouring high copy number plasmids [5]. In this study we have set out to evaluate the feasibility of using Salmonella for the delivery of naked DNA plasmids with high or low copy numbers.
2. Results 2.1. Construction of Salmonella containing a low copy number expression plasmid A DNA fragment of pCMVb which includes the human cytomegalovirus major immediate-early (CMV IE) promoter, b-galactosidase gene and poly A sequence, was isolated by digestion with the restriction enzymes EcoRI and SalI and subsequently cloned into EcoRI/SalI digested pLG339, resulting in the low copy number expression plasmid pLGbGAL. Plasmids pCMVb, pLGbGAL and pLG339 were passaged through S. enterica var Typhimurium LB5010 to ensure modification and electroporated into S. enterica var Typhimurium SL7207. The copy numbers of the plasmids in the resulting strains were evaluated by polymerase chain reaction (PCR) amplification of the b-galactosidase gene. PCR of SL7207/pCMVb and SL7207/pLGbGAL colony lysates with primers BGAL/F09 and POLYA/R01 resulted in a PCR product of approximately 800 bp, consistent with the predicted amplification product of 789 bp. The PCR product from SL7207/pLGbGAL appeared to be less intense than that from SL7207/pCMVb, indicating that the copy number of pLGbGAL was less than that of pCMVb (data not shown). In comparison, PCR amplification of aroB from
SL7207/pCMVb and SL7207/pLGbGAL resulted in similar intensity PCR products from both strains, indicating similar copy numbers. 2.2. Inoculation with high and low copy number plasmids expressing b-galactosidase The use of plasmids pCMVb and pLGbGAL to induce an antibody response to b-galactosidase was evaluated. One hundred micrograms of pCMVb or pLGbGAL, prepared from Escherichia coli TOP10 cells, were used to intramuscularly (i.m.) inoculate groups of 6 BALB/c mice on days 0 and 14. Enzyme linked immunosorbant assay (ELISA) analysis of sera removed from the mice on day 56 showed that both plasmids induced b-galactosidase-specific IgG responses in all mice (Fig. 1). 2.3. Evaluation of SL7207 harbouring the expression plasmids The stability of plasmid pLGbGAL in SL7207 was evaluated by studying colonisation of the recombinant strain in murine tissues. Groups of 6 BALB/c mice were intragastrically (i.g.) inoculated with 1 – 5 £ 109 cfu of SL7207, SL7207/pLGbGAL, SL7207/pLG339 or SL7207/pCMVb. Peyer’s patches and spleens were harvested on day 3 and the number of Salmonella present was determined by spreading onto LB agar or LB agar containing antibiotic. Following inoculation with SL7207/pCMVb, no Salmonella could be found in any of the tissues sampled. In comparison, SL7207/pLGbGAL, SL7207/pLG339 and the parent strain SL7207 colonised tissues to a similar level (Fig. 2). There was no significant difference between colonisation of these strains, and enumeration of bacteria in homogenised tissues showed
Fig. 1. b-galactosidase specific IgG responses induced in BALB/c mice (n ¼ 6) i.m. inoculated two times, 2 weeks apart with 100 mg of pCMVb or pLGbGAL. IgG concentrations were determined by ELISA of serum samples removed on day 56. Error bars represent the standard error of the mean.
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
Fig. 2. Colonisation of Peyer’s patches and spleens following i.g. inoculation of BALB/c mice (n ¼ 6) with SL7207 (squares), SL7207/pLG339 (triangles) or SL7207/pLGbGAL (circles). Organs were removed 3 days after inoculation and homogenised in PBS before plating onto LB agar (closed symbols) or LB agar containing kanamycin (open symbols) for enumeration of bacteria. Horizontal lines show the mean value of individual samples.
that pLGbGAL and pLG339 were both stable in SL7207 in vivo (99% and 100%, respectively). The immunogenicity of SL7207/pLGbGAL was also assessed. Groups of six mice were i.g. inoculated four times at 2-week intervals with 1 – 5 £ 109 cfu of SL7207/pCMVb, SL7207/pLGbGAL, SL7207/pLG339, or SL7207. Further groups of mice were again inoculated i.m. with 100 mg of the eukaryotic expression plasmids pCMVb or pLGbGAL. Although b-galactosidase-specific IgG responses were again induced by inoculation with these DNA plasmids, inoculation with SL7207/pCMVb or SL7207/pLGbGAL did not induce detectable b-galactosidase-specific IgG (data not shown). In comparison, SL7207-specific IgG responses were found in all mice inoculated with any of the Salmonella strains used in this study (data not shown).
3. Discussion and conclusions The technology of using Salmonella to deliver eukaryotic expression plasmids is relatively new (recently reviewed in [4]). One model eukaryotic expression plasmid used in some studies is pCMVb which expresses b-galactosidase under the control of the CMV IE promoter [3,6 – 9]. In these published studies the aroA attenuated S. enterica var Typhimurium strain SL7207 was used to orally deliver pCMVb to macrophages and dendritic cells in vivo, resulting in the induction of a range of b-galactosidase-specific immune responses, including IgG and CD4þ and CD8þ T cell activity. b-galactosidase activity could be detected 5 weeks after the last boost, although recombinant SL7207 could not be detected in vivo at any time [3]. In our initial studies we were unable to reproduce the results obtained by oral administration of SL7207/pCMVb, using the methodology described by
117
Paglia et al. [6]. No detectable b-galactosidase-specific IgG was induced and, like other researchers [3], we were unable to detect recombinant bacteria in vivo. Thus, in an effort to stabilise the b-galactosidase-expressing plasmid for delivery in Salmonella, we cloned the CMV IE promoter, b-galactosidase gene and polyA sequence from pCMVb into a low copy number plasmid, pLG339. The resulting construct, termed pLGbGAL, was shown to have a copy number less than that of pCMVb in SL7207 and induced a b-galactosidase-specific antibody response when intramuscularly inoculated into mice as naked DNA (Fig. 1). Furthermore, SL7207/pLGbGAL was shown to be stable in vivo and could be recovered from Peyer’s patches and spleens of intragastrically inoculated mice (Fig. 2). However, the improved stability of the b-galactosidaseexpressing eukaryotic expression plasmid in SL7207 in vivo did not improve the induction of b-galactosidasespecific IgG responses in mice. The lack of induction of anti-b-galactosidase IgG following delivery of SL7207/pCMVb in this study contrasts with previously published work [6], indicating that Salmonella delivery of DNA vaccines may not be a robust technology. Furthermore, administration of SL7207/pLGbGAL also failed to induce a detectable antibody response. It is possible that reducing the copy number of the plasmid results in the release of less plasmid from the Salmonella than is required to stimulate immunity. However, the findings of this study raise further questions regarding the methodology by which Salmonella strains are able to deliver eukaryotic expression plasmids. Following oral or nasal administration, recombinant Salmonella invade through the M cells of the Peyer’s patches in the gut or through M-like cells in the nasal passages. Here they encounter and are phagocytosed by resident APC’s, including macrophages and dendritic cells. In general, studies of microbial pathogenesis have increased our understanding of the mechanisms employed by bacteria to access cellular compartments. For example, Shigella flexneri are able to escape from phagolysosomes into the host cell cytosol [10]. In comparison, it is known that Salmonella spp. remain in the phagosomal compartment. Thus, it is somewhat surprising that Salmonella are able to successfully deliver expression plasmids. It has been hypothesised that Salmonella infection may cause leakage of the phagolysosome [2], although the mechanism of DNA transport in vitro appears to operate only in mature cells and not in cell lines [3]. In summary, we were able to produce a low copy number eukaryotic vector expressing b-galactosidase that stimulated a specific immune response in vivo. Furthermore we were able to show that the vector was stable in Salmonella in vivo. Recombinant Salmonella were detected in both Peyer’s patches and spleens of infected mice. Although it is known that APC’s in Peyer’s patches and spleens are involved in the presentation of antigen via both MHC class I and II
118
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119
molecules, we were unable to detect any induction of bgalactosidase-specific IgG responses. These results indicate that a further understanding of the mechanism by which Salmonella delivery of DNA occurs is required before a rational approach to vaccine design may be achieved.
4. Materials and methods 4.1. Plasmids, bacterial strains and growth conditions Plasmid pCMVb (copy number , 500 – 700/cell; Clontech Laboratories, Basingstoke, UK) contains the gene encoding b-galactosidase and a gene conferring resistance to ampicillin. Plasmid pLG339 [11] is a low copy number (, 5 copies/cell) plasmid containing a gene conferring resistance to kanamycin. E. coli TOP10 (Stratagene, California, US) was used for cloning and plasmid preparation. S. enterica var Typhimurium LB5010 (r2mþ) was described previously [12]. S. enterica var Typhimurium SL7207 [3] was kindly provided by Professor B.A.D. Stocker (Stanford University School of Medicine, California, US). E. coli and S. enterica var Typhimurium strains were routinely cultured in LB broth or on LB agar, supplemented with antibiotic if required. LB medium was supplemented with 1% (v/v) histidine, 1% (v/v) ‘aromix’ (4 mg/ml tyrosine, 4 mg/ml phenylalanine, 1 mg/ml p-aminobenzoic acid, 1 mg/ml dihydroxybenzoic acid) for growth of SL7207 strains. To produce inocula, SL7207 strains were cultured statically overnight, centrifuged (6000g, 4 8C, 20 min), washed in phosphate buffered saline (PBS) and resuspended in LB broth to a final cell density of approximately 1– 5 £ 1010 cfu/ml. Viable bacteria were enumerated on LB agar plates, supplemented with antibiotic if required. 4.2. Recombinant DNA techniques All enzymes were obtained from Roche Biochemicals (Lewes, UK) and used according to manufacturer’s instructions. DNA restriction fragments were isolated using a QIAquick Gel Extraction kit (QIAGEN Inc., Crawley, UK). Ligation mixtures were electroporated into E. coli TOP10 and recombinants were screened by PCR amplification (95 8C for 2 min; 30 cycles of 95 8C for 20 s, 60 8C for 20 s, 72 8C for 30 s; 72 8C for 10 min) of colony lysates using the primer pair BGAL/F09 (50 -AGGGCCGCAAGAAAACTAT-30 ) and POLYA/R01 (50 -TGTTTATTGCAGCTTATAATGGT-30 ), and Taq DNA polymerase (Roche). Copy numbers of genes were evaluated by PCR amplification (20 cycles, as described above) of colony lysates with BGAL/F09 and POLYA/R01 or with primer pair AROB/F2 (50 -ACATGATTGGCGCGTTTTACC-30 ) and AROB/R2 (5 0 -CTCCGCCGCGCACTTCACTTTTCCCTATG-3 0 ). Recombinant DNA for use in vivo was extracted from TOP10 cells using the Endofree Plasmid Mega Prep kit (QIAGEN Inc., Crawley, UK).
4.3. In vivo studies Six to eight-week old female BALB/c mice (Charles River Laboratories, Kent, UK) were i.g. inoculated with 100 ml Salmonella using an oral gavage needle. Alternatively, they were i.m. inoculated with DNA plasmids resuspended in 100 ml PBS containing 0.25% of the anaesthetic bupivicaine hydrochloride. Fifty microlitre DNA was injected into each hind leg. To study colonisation of the Salmonella, mice were culled 3 days after inoculation by cervical dislocation and Peyer’s patches (6 per mouse) and spleens were removed. Tissues were homogenised in 1 ml PBS using 50 mm cell strainers (Becton Dickinson Labware, New Jersey, US) and spread onto LB agar or LB agar containing antibiotic for enumeration. To study the immunogenicity of the recombinant Salmonella and expression plasmids, mice were inoculated on days 0, 14, 28 and 42, and serum samples were taken on days 12, 26, 40 for analysis. 4.4. Enzyme-linked immunosorbant assay Serum antibody responses against b-galactosidase were assessed using a modified ELISA. Briefly, 96-well microtitre plates were coated overnight at 4 8C with purified b-galactosidase (5 mg/ml in PBS) and then blocked for 1 h at 37 8C with 1% (w/v) skimmed milk powder in PBS (BLOTTO). The test sera were serially diluted in BLOTTO in duplicate on the plates and incubated for 1 h at 37 8C. Washes between steps were performed three times with PBS containing 0.05% (v/v) Tween-20. Bound antibody was detected following incubation with a HRP-conjugated secondary antibody against mouse IgG, diluted 1:4000 in BLOTTO, for 1 h at 37 8C. 2,20 -azino bis(3-ethylbenzthiazoline-6-sulfonic acid) substrate (ABTS) was added and the plates were incubated at room temperature for 20 min. Serum antibody concentrations were calculated in ng/ml using a standard curve. Concentrations were calculated using Revelatione (Dynex Technologies Ltd, West Sussex, UK).
References [1] Davis HL. Plasmid DNA expression systems for the purpose of immunization. Curr Opin Biotech 1997;8:635–40. [2] Weiss S, Krusch S. Bacteria-mediated transfer of eukaryotic expression plasmids into mammalian host cells. Biol Chem 2001; 382:533–41. [3] Darji A, Guzma´n CA, Gerstel B, et al. Oral somatic transgene vaccination using attenuated S. typhimurium. Cell 1997;91:765– 75. [4] Garmory HS, Brown KA, Titball RW. Salmonella vaccines for use in humans: present and future perspectives. FEMS Microbiol Rev 2002; 339 –353. [5] Sirard J-C, Niedergang F, Kraehenbuhl J-P. Live attenuated Salmonella: a paradigm of mucosal vaccines. Immunol Rev 1999; 171:5–26. [6] Paglia P, Medina E, Arioli I, Guzman CA, Colombo MP. Gene transfer in dendritic cells, induced by oral DNA vaccination with
H.S. Garmory et al. / Microbial Pathogenesis 34 (2003) 115–119 Salmonella typhimurium, results in protective immunity against a murine fibrosarcoma. Blood 1998;92:3172–6. [7] Paglia P, Terrazzini N, Schulze K, Guzma´n CA, Colombo MP. In vivo correction of genetic defects of monocyte/macrophages using attenuated Salmonella as oral vectors for targeted gene delivery. Gene Ther 2000;7:1725–30. [8] Darji A, zur Lage S, Garbe AI, Chakraborty T, Weiss S. Oral delivery of DNA vaccines using attenuated Salmonella typhimurium as carrier. FEMS Immunol Med Microbiol 2000;27:341–9. [9] Shata MT, Reitz Jr MS, DeVico AL, Lewis GK, Hone DM. Mucosal and systemic HIV-1 Env-specific CD8þ T-cells develop after
119
intragastric vaccination with a Salmonella env DNA vaccine vector. Vaccine 2002;20:623–9. [10] Weiss S, Krusch S. Bacteria-mediated transfer of eukaryotic expression plasmids into mammalian host cells. Biol Chem 2001; 382:533–41. [11] Stoker NG, Fairweather N, Spratt BG. Versatile low-copy-number plasmid vectors for cloning in Escherichia coli. Gene 1982;18: 335–41. [12] Bullas LR, Ryu JI. Salmonella typhimurium LT2 strains which are r2mþ for all three chromosomally located systems of DNA restriction and modification. J Bacteriol 1983;156:471 –4.