A biologically conjugated polysaccharide vaccine delivered by attenuated Salmonella Typhimurium provides protection against challenge of avian pathogenic Escherichia coli O1 infection

Abstract

Avian pathogenic Escherichia coli (APEC) causes avian airsacculitis and colibacillosis, resulting in significant economic loss to the poultry industry. O1, O2 and O78 are the three predominant serotypes. O-antigen of lipopolysaccharide is serotype determinant and highly immunogenic, and O-antigen polysaccharide-based vaccines have great potential for preventing bacterial infections. In this study, we utilized a novel yeast/bacterial shuttle vector pSS26 to clone the 10.8 kb operon synthesizing APEC O1 O-antigen polysaccharide. The resulting plasmid was introduced into attenuated Salmonella vaccines to deliver this O-antigen polysaccharide. O1 O-antigen was stably synthesized in attenuated Salmonella Typhimurium, demonstrated by slide agglutination, silver staining and western blot. Our results also showed that APEC O1 O-antigen produced in the Salmonella vaccines was attached to bacterial cell surfaces, and the presence of heterologous O-antigen did not alter the resistance to surface-acting agents. Furthermore, birds immunized orally or intramuscularly provided protection against the virulent O1 APEC challenge. Salmonella vaccines carrying APEC O1 O-antigen gene cluster also induced high IgG and IgA immune responses against lipopolysaccharide from the APEC O1 strain. The use of our novel shuttle vector facilitates cloning of large DNA fragments, and this strategy could pave the way for production of Salmonella-vectored vaccines against prevalent APEC serotypes.

Keywords: APEC O1; O-antigen polysaccharide; Salmonella Typhimurium; shuttle vector; yeast recombination.

Figure 1.

Plasmid construction for synthesizing the APEC O1 O-antigen polysaccharide. The amplified linear fragments T1T2-pSC101-asd from pYA3337 and TRP1-ARS4/CEN5-Spec from pYES1L were digested with SbfI and NotI and ligated to generate the shuttle plasmid pSS26. The origin of pSC101 ensured replication in prokaryotic cells, the Asd⁺-balanced lethal host-vector system maintained stability in prokaryotic cells, ARS4/CEN5 ensured replication in MaV203 cells, TRP1 ensured selection in Mav203 cells and Spec was used as a selection marker in prokaryotic cells. T1T2 is the terminator. The O-antigen gene cluster of APEC O1 with an intrinsic promoter was amplified using the primer pairs O1–1F/1R and O1–2F/2R. The linear vector was amplified using the Vector-F/Vector-R primer pair. Each fragment shared an end-terminal homologous sequence with the adjacent fragment. The fragments were assembled into a ring plasmid by recombination in yeast. The recombinant plasmid for synthesizing O1 O-antigen polysaccharide in Salmonella was named pSS27.

Figure 2.

Synthesis of the APEC O1 O-antigen polysaccharide in S. Typhimurium. Bacterial cells of the indicated strains were processed as described in the Materials and Methods. (A) LPS profiles. LPS samples isolated from each strain were subjected to 12% SDS-PAGE followed by silver staining. The Salmonella mutant S740 (Δasd-66 Δcrp-24 Δcya-25 ΔrfbP45) carrying pSS26 produced rough LPS (lane 1). In contrast, the Salmonella mutant SS739 (Δasd-66 Δcrp-24 Δcya-25) carrying either pSS26 (control vector) (lane 3) or pSS27 (APEC O1 gene cluster) (lane 4) and S740 carrying pSS27 (lane 2) produced smooth LPS. LPSs from wild-type Salmonella and E. coli C24–2 exhibited normal ladder-like features. (B) Immunoblot against anti-E. coli O1 serum. The LPSs of Salmonella mutants S740 (Δasd-66 Δcrp-24 Δcya-25 ΔrfbP45) and S739 (Δasd-66 Δcrp-24 Δcya-25) harboring pSS27 and of E. coli C24–2 (lane 6) specifically reacted with anti-E. coli O1 serum (lanes 2 and 4). However, the LPSs from strains with the empty plasmid pSS26 (lanes 1 and 3) and from S. Typhimurium S100 were not reactive to anti-E. coli O1 serum. (C) Immunoblot against anti-Salmonella O4 serum. Only S739 (Δasd-66 Δcrp-24 Δcya-25) (lanes 3 and 4) and Salmonella S100 (lane 5) could produce LPSs that reacted with Salmonella O4 serum.

Figure 3.

Stability of pSS27 plasmid in S. Typhimurium. Strains containing pSS27 were cultured to ∼24 h, 48 h and 72 h to detect plasmid stability in S. Typhimurium. (A) LPS profiles. All LPS samples generated from 24 h (lane 1 and lane 2), 48 h (lane 3 and lane 4) and 72 h (lane 5 and lane 6) cultures produced O-antigen moieties. (B) Western blotting detection against anti-E. coli O1 serum. All strains containing pSS27 reacted with antiserum against E. coli O1 and C24–2 strain. The stability of pSS27 plasmid in S. Typhimurium was not affected by the generations. (C) Plasmid stability of pSS27 carrying APEC O1 O-antigen gene cluster, strains cluster. Strains of S739 and S740 carrying plasmid pSS26 or pSS27 were subcultured every 12 h for 3 days in non-selective medium. Plasmid stability was determined as the percentage of colonies (out of 100 selected) growing on selective media.

Figure 4.

Phenotypic evaluation of Salmonella vaccine expressing O1 O-antigen polysaccharide. (A) Growth curves of the wild-type S100, S. Typhimurium S739 (Δasd-66 Δcrp-24 Δcya-25) and S740 (Δasd-66 Δcrp-24 Δcya-25 ΔrfbP45) with control plasmid pSS26 or pSS27. (B) Swimming motility. (C) Swarming motility. (D) Sensitivity to polymyxin B. (E) Sensitivity to DOC. Heterologous O1 O-antigen polysaccharide expression in S. Typhimurium did not affect its phenotypic characteristics. All assays were repeated three times, and the error bars represent deviations of triplicate tests.

Figure 5.

Serum IgG and mucus IgA responses to LPSs from S. Typhimurium and APEC O1. Samples from five chickens each group were chosen randomly for the use of antibody detection by ELISA. Triplicate wells of every sample were prepared. (A) IgG elicited from serum against S. Typhimurium LPS. (B) IgG elicited from serum against APEC O1 LPS. (C) IgA elicited from intestinal mucus against S. Typhimurium LPS. (D) IgA elicited from intestinal mucus against APEC O1 LPS. The error bars represent variations of triplicate samples. Differences were compared among the vaccine group (pSS27 carrying strain), the empty vector pSS26-treated group and the BSG-treated control group. Antibody type and responses varied with immunized strains and routes. All groups immunized with strains carrying the control vector pSS26 did not show differences with the BSG group.