Improved cell surface display of Salmonella enterica serovar Enteritidis antigens in Escherichia coli
© Gustavsson et al.; licensee BioMed Central. 2015
Received: 20 November 2014
Accepted: 12 March 2015
Published: 9 April 2015
Salmonella enterica serovar Enteritidis (SE) is one of the most potent pathogenic Salmonella serotypes causing food-borne diseases in humans. We have previously reported the use of the β-autotransporter AIDA-I to express the Salmonella flagellar protein H:gm and the SE serotype-specific fimbrial protein SefA at the surface of E. coli as live bacterial vaccine vehicles. While SefA was successfully displayed at the cell surface, virtually no full-length H:gm was exposed to the medium due to extensive proteolytic cleavage of the N-terminal region. In the present study, we addressed this issue by expressing a truncated H:gm variant (H:gmd) covering only the serotype-specific central region. This protein was also expressed in fusion to SefA (H:gmdSefA) to understand if the excellent translocation properties of SefA could be used to enhance the secretion and immunogenicity.
H:gmd and H:gmdSefA were both successfully translocated to the E. coli outer membrane as full-length proteins using the AIDA-I system. Whole-cell flow cytometric analysis confirmed that both antigens were displayed and accessible from the extracellular environment. In contrast to H:gm, the H:gmd protein was not only expressed as full-length protein, but it also seemed to promote the display of the protein fusion H:gmdSefA. Moreover, the epitopes appeared to be recognized by HT-29 intestinal cells, as measured by induction of the pro-inflammatory interleukin 8.
We believe this study to be an important step towards a live bacterial vaccine against Salmonella due to the central role of the flagellar antigen H:gm and SefA in Salmonella infections and the corresponding immune responses against Salmonella.
Bacterial surface display of recombinant proteins has been used for various applications in microbiology, molecular biology, biotechnology and vaccinology [1,2]. The field is dominated by display in Escherichia coli, mainly due to the extensive knowledge concerning the genetics, the presence of transformation protocols, the rapid cell growth on simple media, and the multitude of production techniques and protocols for recombinant protein production that are associated with this organism. Further, a range of specific mutant strains with desirable features is available for metabolic engineering.
An area of particular interest is the display of immunogenic peptides at the cell surface for direct use as live vaccine delivery vehicles . This technique offers some interesting advantages over conventional vaccines the first being that a surface display system provides the prerequisites for a more economic process . The reason is that it is designed to circumvent the step of antigen purification and relies on a one-step combination of the production and purification processes. Secondly, the live vaccine systems may have in-built adjuvant effects from endogenous components of the host cell e.g. from the distinctive structure of the E. coli lipopolysaccharides (LPS) but also the lipoteichoic acids, peptidoglycans, lipoproteins etc., which are well recognized by the host’s immune system  and therefore elicit strong immune responses. Finally, these vehicles may have prolonged retention times due to colonization effects and they are probably also safer to use than inactivated or attenuated pathogen-based vaccines, which may revert to their pathogenic form.
Surface expression of H:gm, H:gmd, SefA, and H:gmdSefA in E. coli
Description of the different passenger proteins expressed in this study
As seen in Figure 1, the pAIDA1 vector is constructed to harbor two distinct tags, His6 and c-Myc that are flanking the recombinant passenger protein on the N- and C-terminal sides, respectively. Using fluorescently labeled antibodies specific for the tags allows us to determine the presence and integrity of the different fusion proteins using flow cytometry.
Figure 3 shows also the data for the second detection system based on the His6-tag that is situated outside the antigen facing the medium. All constructs, except H:gm (Additional file 1: Figure S1), stained positive for the presence of this tag at the cell surface. The fluorescence intensity of the H:gm-expressing cells was comparable to the negative control while the new constructs exhibited a significant increase in fluorescence. In contrast to H:gm, the novel construct containing the smaller flagellar antigen H:gmd was expressed very well at the cell surface. In fact, this construct resulted in the highest fluorescence of all constructs tested for both the anti-His6 and the anti-c-Myc antibody (Figure 3B).
Subcellular localization and size of the fusion proteins
A maybe even more important finding was that H:gmd did not show any visible signs of proteolytic degradation, neither in the FACS fluorescence data (Figure 3) or by SDS-PAGE analysis (Figure 4), in contrast to the previously used H:gm that gave rise to little full-length protein due to extensive proteolysis (Additional file 1: Figure S1) [7,11]. This is a significant improvement as H:gmd contains the serotype-specific specific amino acids, and thus might increase its potential as a live bacterial subunit vaccine against Salmonella.
Functionality of the expressed proteins
The main goal with the present study was to successfully express and display H:gm-derived antigens at the surface of E. coli by using the AIDA-I autotransporter system. Our previous attempts to express and display H:gm were unsuccessful due to extensive proteolytic cleavage in the C-terminal region of that antigen [7,11]. To make H:gm less affected to proteolytic degradation could be achieved through several approaches. One option would be the usage of protease-deficient E. coli strains. However, we already used a mutant strain deficient for the protease OmpT (O17ΔOmpT) and a clear candidate protease could not be identified from our previous experiments. Instead, we chose to engineer the protein to get a H:gm-derivative that could potentially be efficiently displayed at the cell surface and be less sensitive to proteolytic cleavage. For this purpose, we here chose to include only a subpart, called H:gmd, comprising 98 aa out of the total 507 aa in the full-length H:gm (Figure 2). The rationale behind this was threefold. Firstly, the chosen region comprises the serotype-specific part of H:gm needed for development of immunity to SE [21,22]. Secondly, the large size of H:gm can be expected to lead to a longer time required for full translocation over the outer membrane, resulting in an extended exposure to periplasmic proteases. By expressing only a small part of H:gm this exposure should be reduced and consequently the protein should be less susceptible to proteolysis. Thirdly, it has been shown that passenger proteins that form stable folds in the periplasm, for instance due to disulphide formation, are difficult to display using autotransporters due to stalled translocation over the outer membrane [23,24]. As seen in Figure 2, H:gm folds with the N-and C-terminals in close proximity, which could be problematic given the autotransporter translocation mechanism. By expressing only the H:gmd domain this potentially problematic part of the protein is excluded.
This strategy proved to be very successful, as H:gmd generated whole-cell fluorescence intensities much greater than the negative control, both when the presence of the N-terminal His6-tag and the C-terminal c-Myc tag were analyzed (Figure 3B). This is in contrast to the original H:gm construct, which did not generate a signal above the negative control when probed with the His6-reactive antibody, indicating proteolytic degradation (Figure 3A, left panel). Interestingly, not only was the smaller H:gmd protein expressed to full-length at the cell surface, but it also appeared to promote the display and accessibility of the fused protein H:gmdSefA since H:gmd and H:gmdSefA both resulted in higher fluorescence intensities than SefA alone or the full-length H:gm. Although the true reasons behind the observed improvements in expression and cell surface display characteristics of H:gmd over H:gm at present are unclear, they can likely be attributed either to structural reasons as discussed previously, or that the protease sensitive amino acids in H:gm are not present in H:gmd.
With the ultimate goal to generate suitable tools for live vaccine development, we tested the E. coli strains expressing the different Salmonella epitopes for their ability to induce an immune reaction in gut epithelial cells. The response in the in vitro system applied here was overall very low, probably related to the low immunogenicity of the non-pathogenic E. coli K-12 strain O17ΔOmpT, which was unable to stimulate IL-8 production on its own. However, our results suggest that while SefA appeared the most potent immunogen, the truncated H:gmd variant was to some extent even superior to the full-length H:gm construct, in line with the flow cytometric resultss. In contrast, the fusion construct H:gmdSefA did not perform well in this assay. A possible explanation might be that the epitopes of the single proteins are hidden by the structure of the fusion protein.
In this study we engineered a truncated variant of H:gm comprising only the serotype-specific amino acids. H:gmd was resistant to proteolytic cleavage and exhibited highly efficient surface display, in contrast to H:gm. This is a promising result from the perspective of creating a live vaccine against SE, though additional studies are needed to further confirm the immunogenicity of these constructs and to investigate their ability to raise SE-specific antibodies in an animal host.
Bacterial strains and plasmids
Escherichia coli K-12 strain O17ΔOmpT was used for surface expression using the AIDA-I autotransporter, as previously reported . The plasmid pAIDA1-SefA  was used to produce a recombinant fusion of SE fimbriae protein SefA and AIDA-I on the surface of the host cell. This plasmid encodes the C-terminal AIDA β-barrel (AIDAC) followed by a linker consisting of the first 54 amino acids (aa) of the native AIDA passenger, the recombinant SefA passenger and finally the native AIDA signal peptide. Additionally, the plasmid contains two epitope tags (c-Myc and His6) located on the C- and N-terminal sides of SefA, respectively (Figure 1). pAIDA1-H:gm [6-8] encodes the SE flagellar antigen H:gm instead of SefA. For construction of pAIDA1-H:gmd and pAIDA1-H:gmdSefA, standard recombinant DNA techniques were used . Restriction endonucleases, T4 DNA ligase, plasmid purification kits (miniprep), and DreamTaq DNA polymerase were purchased from Fermentas. Primers were supplied by Eurofins MWG Operon while GE Healthcare supplied the PCR DNA and Gel band Purification Kit. H:gmd_forw (5′-acaggtaccactaaatctactgctggaaccgctgaag-3′) and H:gmd_rev (5′-actgagctcctggatccacgaccctccgcctcctgaacccccgcctccgtttttggttttatcatcaaaag-3′) were used to PCR amplify a H:gmd-encoding DNA fragment from pAIDA1-H:gm. The resulting amplicon was then purified, digested with KpnI and SacI, gel purified and finally ligated into KpnI- and SacI-digested pAIDA1 to generate pAIDA1-H:gmd. To create pAIDA1-H:gmdSefA, the purified H:gmd-amplicon was instead cleaved with KpnI and BamHI, gel purified and then ligated into pAIDA1-SefA previously digested with KpnI and BclI, thus yielding pAIDA1-H:gmdSefA. These novel constructs were all verified by DNA sequencing. Table 1 shows a summary of the properties of the different antigens.
All cultivations were performed in minimal salts medium, consisting of (per liter): 7.0 g (NH4)2SO4, 1.6 g KH2PO4, 6.6 g Na2HPO4 .2H2O, and 0.5 g (NH4)2-H-Citrate. In addition, all media were supplemented with 1 ml l−1 of both 1 M MgSO4 and trace element solution that were sterile filtered into the different reactors. The trace element stock solution consisted of (per liter): 0.5 g CaCl2 · 2H2O, 16.7 g FeCl3 · 6H2O, 0.18 g ZnSO4 · 7H2O, 0.16 g CuSO4 · 5H2O, 0.15 g MnSO4 · 4H2O, 0.18 g CoCl2 · 6H2O, 20.1 g Na-EDTA.
All cultivations were performed in a batch format initialized by shake flask growth in 1 L bottles containing 100 ml of minimal salts medium supplemented with 10 g l−1 glucose. These were inoculated from a frozen glycerol stock (−80°C) and placed inside a shaking incubator (37°C, 180 rpm) over night. In the following morning, six samples were withdrawn and used to inoculate a six-parallel stirred-tank bioreactor unit (Greta, Belach Bioteknik AB, Sweden) where each reactor contained 800 ml of minimal medium to which 10 g l−1 of separately sterilized glucose had been added. After inoculation, the cultures were grown for two generations to an optical density of 0.4, before induction by addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to a concentration of 200 μM. Zero-samples for surface expression were withdrawn immediately before induction. The cultures were then grown for approximately four generations, after which final samples were withdrawn for subsequent analysis and comparison.
The dissolved oxygen concentration (DOT) in the reactors was controlled at 40% by automatic adjustment of the stirrer speed. The airflow was initially set to 0.1 VVM, and then incrementally increased to 1.0 VVM. The temperature was automatically controlled at 37°C and pH was kept at 7.0 by titration with NH4OH (12.5% w/v). Foaming was kept at a minimum by manual addition of antifoaming agent when necessary.
Cell mass accumulation was followed by sampling at regular intervals. The optical density at 600 nm (OD600) was measured using a spectrophotometer (Genesys 20, Thermo Scientific). All samples were diluted in saline solution (0.9% NaCl) to an approximate OD600 of 0.1 prior to measurement to compensate for the non-linearity of OD600 measurements, and actual OD600 values were derived by multiplication with the dilution factor.
Flow cytometric analyses of surface expression
Cell samples for analysis of surface expression were aseptically withdrawn and stored at −80°C, as previously described . At the day of analysis, 50 μl of each sample was labeled using fluorescent antibodies against the two detection tags (His6 and Myc) flanking the passenger protein, as described previously . Surface expression levels were evaluated based on measurement of fluorescence from the two antibodies using a flow cytometer (Gallios, Beckman Coulter). 10,000 events per sample were recorded. The excitation wavelength was 488 nm for the His6-tag (FITC) analysis and emission was detected at 525/40 nm, while the c-Myc-analysis (SureLight® Allophycocyanin) used 638 nm and 660/20 nm for the excitation and emission wavelengths, respectively.
Analysis of OM proteins
E. coli proteins were separated into soluble, inner membrane and OM fractions as previously reported . The resulting outer membrane protein samples were separated using SDS-PAGE on 10 % Bis-Tris gels (NuPage, Invitrogen). The gels were then stained using PageBlue protein staining solution (Fermentas) according to the manufacturer’s protocol.
Cell stimulation assay
Colonrectal adenocarcinoma cell line HT-29 (ATCC, HTB-38) were grown in McCoy’s 5A medium supplemented with 10% fetal bovine serum at 37°C in a humidified incubator with 5% CO2. For experiments, cells were seeded on 24-well plates and used when confluent. Bacteria were collected at the time point of optimal protein expression by centrifugation, washed once in phosphate-buffered saline (PBS) and adjusted to an OD600 of 0.125 in PBS. Bacteria were diluted 1:100 in complete cell culture medium to reach a final concentration of approximately 106 colony forming units (CFU)/ml. Medium was removed from the cells and 1 ml of fresh medium with bacteria was added. The plate was centrifuged at 600 g for 5 min to accelerate bacterial contact to the cells and correct for potential differences in motility of the strains. The infected cells were incubated at 37°C in a humidified incubator with 5% CO2 for 4 hours. Then, the supernatants were collected, floating cells were removed by centrifugation at 300 g for 5 min and the cleared supernatants were stored at −80°C until analysis. Secreted interleukin (IL) 8 was quantified by enzyme-linked immunosorbent assay (ELISA) according to the manufacturer’s protocol (R&D Systems).
The Swedish International Development Agency (SIDA) is acknowledged for their financial support.
- Benhar I. Biotechnological applications of phage and cell display. Biotechnol Adv. 2001;19:1–33.View ArticleGoogle Scholar
- Georgiou G, Stathopoulos C, Daugherty PS, Nayak AR, Iverson BL, Curtiss R. Display of heterologous proteins on the surface of microorganisms: from the screening of combinatorial libraries to live recombinant vaccines. Nat Biotechnol. 1997;15:29–34.View ArticleGoogle Scholar
- Dertzbaugh MT. Genetically engineered vaccines: an overview. Plasmid. 1998;39:100–13.View ArticleGoogle Scholar
- Weidemann B, Schletter J, Dziarski R, Kusumoto S, Stelter F, Rietschel ET, et al. Specific binding of soluble peptidoglycan and muramyldipeptide to CD14 on human monocytes. Infect Immun. 1997;65:858–64.Google Scholar
- Benz I, Schmidt M. Adhesin (AIDA-I) Involved in Diffuse Adherence of Enteropathogenic Escherichia coli. Infect Immun. 1989;57:1506–11.Google Scholar
- Gustavsson M, Bäcklund E, Larsson G. Optimisation of surface expression using the AIDA autotransporter. Microb Cell Fact. 2011;10:72.View ArticleGoogle Scholar
- Jarmander J, Gustavsson M, Do T-H, Samuelson P, Larsson G. A dual tag system for facilitated detection of surface expressed proteins in Escherichia coli. Microb Cell Fact. 2012;11:118.View ArticleGoogle Scholar
- Gustavsson M, Muraleedharan MN, Larsson G. Surface expression of ω-transaminase in Escherichia coli. Appl Environ Microb. 2014;80:2293–8.View ArticleGoogle Scholar
- Benz I, Schmidt MA. Structures and functions of autotransporter proteins in microbial pathogens. Int J Med Microbiol. 2011;301:461–8.View ArticleGoogle Scholar
- Leyton DL, Rossiter AE, Henderson IR. From self sufficiency to dependence: mechanisms and factors important for autotransporter biogenesis. Nat Rev Microbiol. 2012;10:213–25.View ArticleGoogle Scholar
- Nhan NT, Gonzalez de Valdivia E, Gustavsson M, Hai T, Larsson G. Surface display of salmonella epitopes in Escherichia coli and Staphylococcus carnosus. Microb Cell Fact. 2011;10:22.View ArticleGoogle Scholar
- Jarmander J, Janoschek L, Lundh S, Larsson G, Gustavsson M. Process optimization for increased yield of surface-expressed protein in Escherichia coli. Bioproc Biosyst Eng. 2014;37:1685–93.View ArticleGoogle Scholar
- Strindelius L, Filler M, Sjöholm I. Mucosal immunization with purified flagellin from Salmonella induces systemic and mucosal immune responses in C3H/HeJ mice. Vaccine. 2004;22:3797–808.View ArticleGoogle Scholar
- Robertson JMC, McKenzie NH, Duncan M, Allen-Vercoe E, Woodward MJ, Flint HJ, et al. Lack of flagella disadvantages Salmonella enterica serovar Enteritidis during the early stages of infection in the rat. J Med Microbiol. 2003;52:91–9.View ArticleGoogle Scholar
- Edwards RA, Schifferli DM, Maloy SR. A role for Salmonella fimbriae in intraperitoneal infections. Proc Natl Acad Sci U S A. 2000;97:1258–62.View ArticleGoogle Scholar
- Clouthier SC, Müller KH, Doran JL, Collinson SK, Kay WW. Characterization of three fimbrial genes, sefABC, of Salmonella enteritidis. J Bacteriol. 1993;175:2523–33.Google Scholar
- Peralta RC, Yokoyama H, Ikemori Y, Kuroki M, Kodama Y. Passive immunisation against experimental salmonellosis in mice by orally administered hen egg-yolk antibodies specific for 14-kDa fimbriae of Salmonella enteritidis. J Med Microbiol. 1994;41:29–35.View ArticleGoogle Scholar
- de Vries N, Zwaagstra KA, Veld JH H i’t, van Knapen F, van Zijderveld FG, Kusters JG. Production of monoclonal antibodies specific for the i and 1,2 flagellar antigens of Salmonella typhimurium and characterization of their respective epitopes. Appl Environ Microb. 1998;64:5033–8.Google Scholar
- Bäcklund E, Reeks D, Markland K, Weir N, Bowering L, Larsson G. Fedbatch design for periplasmic product retention in Escherichia coli. J Biotechnol. 2008;135:358–65.View ArticleGoogle Scholar
- Rochon M, Römling U. Flagellin in combination with curli fimbriae elicits an immune response in the gastrointestinal epithelial cell line HT-29. Microbes Infect. 2006;8:2027–33.View ArticleGoogle Scholar
- Mizumoto N, Toyota-Hanatani Y, Sasai K, Tani H, Ekawa T, Ohta H, et al. Detection of specific antibodies against deflagellated Salmonella Enteritidis and S. Enteritidis FliC-specific 9 kDa polypeptide. Vet Microbiol. 2004;99:113–20.View ArticleGoogle Scholar
- Li J, Nelson K, McWhorter AC, Whittam TS, Selander RK. Recombinational basis of serovar diversity in Salmonella enterica. Proc Natl Acad Sci U S A. 1994;91:2552–6.View ArticleGoogle Scholar
- Jong WSP, ten CM H-J, Blaauwen den T, Slotboom DJ, Tame JRH, Wickström D, et al. Limited tolerance towards folded elements during secretion of the autotransporter Hbp. Mol Microbiol. 2007;63:1524–36.View ArticleGoogle Scholar
- Leyton DL, Sevastsyanovich YR, Browning DF, Rossiter AE, Wells TJ, Fitzpatrick RE, et al. Size and conformation limits to secretion of disulfide-bonded loops in autotransporter proteins. J Biol Chem. 2011;286:42283–91.View ArticleGoogle Scholar
- Sambrook J, Fritsch EF, Maniatis T. Molecular cloning. 1989.Google Scholar
- Kelley LA, Sternberg MJE. Protein structure prediction on the Web: a case study using the Phyre server. Nat Protoc. 2009;4:363–71.View ArticleGoogle Scholar
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.