Genome-Wide Identification of Virulence Genes in Erysipelothrix rhusiopathiae: Use of a Mutant Deficient in a tagF Homolog as a Safe Oral Vaccine against Swine Erysipelas

Swine erysipelas is caused by the Gram-positive pathogen Erysipelothrix rhusiopathiae. The swine erysipelas live vaccine in Japan, the E. rhusiopathiae Koganei 65-0.15 strain (Koganei), has been reported to cause arthritis and endocarditis. To develop a vaccine with increased safety, we used a virulent Fujisawa strain to construct transposon mutants for a total of 651 genes, which covered 38% of the coding sequence of the genome.

ABSTRACT

Swine erysipelas is caused by the Gram-positive pathogen Erysipelothrix rhusiopathiae. The swine erysipelas live vaccine in Japan, the E. rhusiopathiae Koganei 65-0.15 strain (Koganei), has been reported to cause arthritis and endocarditis. To develop a vaccine with increased safety, we used a virulent Fujisawa strain to construct transposon mutants for a total of 651 genes, which covered 38% of the coding sequence of the genome. We screened the mutants for attenuation by inoculating mice with 10⁸ CFU of each mutant and subsequently assessed protective capability by challenging the surviving mice with 10³ CFU (10² times the 50% lethal dose) of the Fujisawa strain. Of the 23 attenuated mutants obtained, 6 mutants were selected and evaluated for protective capability in pigs by comparison to that of the Koganei strain. A mutant in the ERH_0432 (tagF) gene encoding a putative CDP-glycerol glycerophosphotransferase was found to be highly attenuated and to induce humoral and cell-mediated immune responses in conventional pigs. An in-frame deletion mutant of the gene, the Δ432 mutant, was constructed, and attenuation was further confirmed in germfree piglets; three of four piglets subcutaneously inoculated with 10⁹ CFU of the Δ432 mutant showed no apparent clinical symptoms, whereas all four of the Koganei-inoculated piglets died 3 days after inoculation. It was confirmed that conventional pigs inoculated orally or subcutaneously with the Δ432 strain were almost completely protected against lethal challenge infection. Thus, the tagF homolog mutant of E. rhusiopathiae represents a safe vaccine candidate that can be administered via the oral and subcutaneous routes.

INTRODUCTION

Among current veterinary vaccines, which include live attenuated organisms, inactivated/killed organisms or their components, live vaccines are very effective, with a rapid onset of immunity, and a single dose is usually sufficient to induce long-term, sometimes lifelong, protection (1, 2). Furthermore, live vaccines have advantages over other vaccines in that oral delivery of live vaccines can induce appropriate mucosal immune responses to pathogens at the point of entry (3, 4) and, most importantly, greatly reduce the stress of animals during vaccination if administered via drinking water. Thus, live oral vaccines are suitable for farm animals, for which a low cost is required. In veterinary medicine, many live attenuated bacterial vaccines have been used (1). However, these conventional vaccines have been developed via empirical attenuation, and the nature of the attenuation is based on random mutations; therefore, the mechanisms of attenuation are unknown (1, 5).

Swine erysipelas, caused by infection with the Gram-positive pathogen Erysipelothrix rhusiopathiae, is characterized by acute septicemia or chronic endocarditis and polyarthritis and results in economic losses in the pork industry worldwide (6). Currently, live and inactivated vaccines are available for the control of the disease (6). In Japan, the currently used live vaccine is the acriflavine-resistant E. rhusiopathiae Koganei 65-0.15 strain (Koganei), which was attenuated by 65 passages on agar plates containing 0.15% acriflavine dye and licensed in 1974 for subcutaneous injection. Previously, Imada et al. (7) reported that Koganei-like strains were isolated from diseased pigs that had been given the live vaccine. Recently, we also reported that the vaccine strain causes chronic disease; more than 65% of the clinical isolates from pigs with chronic disease in farms where the Koganei vaccine had been used were determined to be the vaccine strain (8). Importantly, it was also found that acriflavine resistance, which has been regarded as a marker of the strain, has been lost in some of the vaccine strains isolated from diseased pigs. Analysis of the draft genome sequence of the Koganei strain and comparison to the complete genome of the reference strain E. rhusiopathiae Fujisawa revealed that the Koganei strain has 76 strain-specific single nucleotide polymorphisms (SNPs) (9, 10). Thus, the mechanisms of attenuation and acriflavine resistance in this strain have not been clarified. These findings motivated us to develop rationally designed, safe, and effective live vaccines.

Thus far, we have used attenuated E. rhusiopathiae strains as vectors for delivering foreign antigens (11–13). E. rhusiopathiae is a facultative intracellular pathogen that induces strong cell-mediated immunity in mice (14) and can be orally or intranasally administered to pigs, eliciting cell-mediated immune responses to an expressed foreign antigen (12, 13). Whole-genome sequence analysis revealed that the E. rhusiopathiae genome shows a complete loss of fatty acid biosynthesis pathways and lacks the genes for the biosynthesis of many amino acids, cofactors, and vitamins (15), indicating that this organism has undergone genome reduction and depends on mostly its hosts for nutrients; therefore, this organism cannot propagate if separated from its hosts. Taken together, these characteristics suggest that rationally attenuated E. rhusiopathiae vaccines have potential as safe vectors for the delivery of recombinant antigens from pathogens to mucosal immune systems.

Recently, we successfully established a system for genome-wide analysis of virulence-associated genes of this organism using random transposon mutagenesis (16, 17). In this study, we report the genome-wide identification of virulence genes in E. rhusiopathiae. We propose that a mutant with a deletion of a gene that is homologous to tagF (teichoic acid glycerol F), which is involved in the biosynthesis of wall teichoic acids (WTAs), is a safe and effective vaccine candidate that can be administered orally and subcutaneously (s.c.) to pigs. Our results, however, suggest that E. rhusiopathiae lacks canonical WTAs, and thus the function of the tag homolog remains unknown.

RESULTS

Screening transposon mutants for attenuation and protective capability in mice.

We used the highly virulent Fujisawa strain to construct transposon mutants of a total of 651 genes, which covered 38% of the coding sequence of the genome. We screened all the mutants for attenuation by s.c. inoculation of two mice with 10⁸ CFU (approximately 10⁷ times the 50% lethal dose [LD₅₀] of the parental Fujisawa strain) of each mutant and subsequently assessed their protective capability using the surviving mice. We obtained a total of 23 attenuated mutants; 19 mutants did not cause any clinical symptoms, and 4 mutants caused death in one of the two mice tested per mutant. Among these 23 mutants, 19 mutants induced complete protection against challenge infection with 100 times the LD₅₀ of the parent strain. The balance between safety and immunogenicity is very difficult to achieve, and a high level of attenuation often results in poor protection. In this study, we selected six mutants (Table 1 ) that caused ruffled fur, which is a general clinical sign of the infection, in mice after screening analysis with subcutaneous inoculation with 10⁸ CFU of each mutant and further assessed the virulence and protective capability of the candidates in pigs.

TABLE 1.

E. rhusiopathiae Fujisawa derivatives analyzed in this study

Strain	Locus tag inactivated or deleted	Description	Transposon insertion site or deleted region
Clone 39	ERH_0295	Nitrate/sulfonate/bicarbonate ABC transporter	Transposon insertion in the gene (382_383ins)
Clone 113	ERH_0404	Aspartate aminotransferase	Transposon insertion in the gene (979_980ins)
Clone 117	ERH_0432	CDP-glycerol:poly(glycerophosphate) glycerophosphotransferase	Transposon insertion in the gene (625_626ins)
Clone 129	ERH_0753	Citrate lyase, alpha subunit	Transposon insertion in the gene (126_127ins)
Clone 100	ERH_0862	Bifunctional CTP:phosphocholine cytidylyltransferase/choline kinase	Transposon insertion in the gene (329_330ins)
Clone 146	ERH_1562	Hypothetical protein	Transposon insertion in the gene (180_181ins)
Δ432	ERH_0432	CDP-glycerol:poly(glycerophosphate) glycerophosphotransferase	In-frame deletion of the gene (37_1131del)

Virulence and protective capability of transposon mutants in pigs.

The virulence and protective capability of the six mutants in pigs were assessed by comparison to those of the Koganei strain. Groups of pigs were immunized s.c. with 10⁹ CFU of the strains and observed for 14 days. The intensity of the skin reaction (erythema) to vaccine strains has been reported to correlate well with protective immunity in swine erysipelas (18). After inoculation with the mutant, some of the pigs exhibited erythema at the site of inoculation (Table 2), but none of the vaccinated pigs developed generalized erythema, and all pigs remained healthy during observation. The pigs were subsequently challenged s.c. with 10⁸ CFU of the Fujisawa strain and further observed for 8 days. The pigs inoculated with the Koganei strains and the mutant clones 39, 113, and 117 did not show any clinical signs and were completely protected, whereas in the groups of pigs inoculated with clones 100, 129, and 146 one pig died or was euthanized by day 4 after challenge infection.

TABLE 2.

Appearance of skin erythema and agglutinating IgG antibody titers after immunization with E. rhusiopathiae strains (live vaccine and transposon mutants) in conventional pigs^a

Immunization strain (109 CFU)	Pig no.	Erythema at inoculation site	Agglutinating IgG antibody titerb
			Before immunization	Wk 1 after immunization	Wk 2 after immunization	Wk 1 after challenge infection (i.e., wk 3 after immunization)
Koganei	1	No	8	16	16	32
	2	Yes	8	16	32	128
Clone 39	3	No	8	16	16	32
	4	Yes	8	16	64	128
	5	No	8	16	8	8
Clone 100	6	No	8	16	8	64
	7	No	8	16	8	Died
Clone 113	8	No	8	8	8	32
	9	No	16	4	8	256
Clone 117	10	No	8	8	8	16
	11	No	8	4	128	256
	17	No	8	8	8	32
Clone 129	18	Yes	8	4	32	16
	19	No	8	8	8	Died
	20	No	8	8	8	16
Clone 146	21	No	8	4	8	Died
	22	Yes	4	4	8	1,024

^aThe pigs were immunized s.c. with 10⁹ CFU of the mutants and bled on days 0, 7, 14, and 21 after immunization. Serum samples were pretreated with 2-ME at 37°C for 1 h, subjected to twofold serial dilution, and mixed with an equal volume of overnight culture of E. rhusiopathiae Marienfelde diluted 1:100 with culture medium. The agglutination titer was determined after overnight incubation at 37°C.

^bThe results are expressed as the reciprocal of the highest serum dilution showing agglutination.

The humoral immune responses of the pigs were analyzed by determining agglutinating IgG antibodies by a growth agglutination (GA) test. As shown in Table 2, among the pigs that showed low levels of agglutinating IgG antibodies, pigs 7, 19, and 21 died after challenge infection. Clone 117 (carries a transposon insertion in the ERH_0432 gene) induced levels of agglutinating IgG antibodies comparable to those seen with the Koganei strain after challenge infection.

Cell-mediated immune responses were analyzed by a blastogenesis assay (Fig. 1). Compared to the Koganei strain, all the tested mutants induced strong cell-mediated immune responses; it is noteworthy that the cell-mediated immune responses induced by clone 117 were strong on day 14 after immunization and on day 7 after challenge infection. Unexpectedly, the pigs (animals 7, 19, and 21) that showed strong cell-mediated immune responses died after the challenge infection.

FIG 1.

Cell-mediated immune responses on days 7 (A), 14 (B), and 21 (C) after immunization. Lymphoproliferative responses are shown as a stimulation index in response to formaldehyde-inactivated bacterial cells (ER) and concanavalin A (ConA). The numbers shown above the strain names correspond to the pig no. in Table 2. In panel C, the letter D indicates the death of the pigs.

Construction of an in-frame ERH_0432 deletion mutant and its phenotypic characteristics.

Animal experiments using transposon mutants revealed that ERH_0432, which is a homolog of the tag gene that encodes CDP-glycerol:poly(glycerophosphate) glycerophosphotransferase, described as CDP-glycerol glycerophosphotransferase here, and was inactivated in clone 117, may be a promising target for the development of a safe and effective vaccine against swine erysipelas. We constructed an in-frame deletion mutant, designated the Δ432 strain, of this gene in the Fujisawa strain. The Δ432 strain had a growth rate similar to that of the Fujisawa strain (see Fig. S1 in the supplemental material). Draft genome sequence analysis of the mutant confirmed that in-frame deletion of the gene was correct, and there were no unintended insertions/deletions in the genome sequence (data not shown).

The capsule is the most important virulence factor of E. rhusiopathiae (19). It has been shown that the capsular polysaccharide (CPS) of the organism is modified with phosphorylcholine (PCho), and the capsular structure and its modification by PCho are critical for the virulence of this organism (16). Western blot analysis with the monoclonal antibodies (MAbs) ER21 and TEPC-15, which detect the expression of the CPS and PCho (16), respectively, revealed that compared to the parent Fujisawa strain, the Δ432 strain produced greatly reduced levels of the CPS epitope recognized by ER21. The intensities of the reactivity signals for PCho were similar between the strains, but the banding pattern was changed (Fig. 2), suggesting that the molecular integrity of the CPS/PCho was altered in this strain. Electron microscopy analysis also confirmed that compared to the parent strain Fujisawa, the Δ432 strain exhibited reduced levels of the CPS epitope but unaltered levels of PCho (Fig. 3).

FIG 2.

Western blot detection of CPS (A) and PCho (B), as well as a Coomassie blue-stained gel (C). The positions of the protein molecular mass standards (kDa) are shown on the left.

FIG 3.

Immunogold electron microscopy analysis of the Fujisawa and Δ432 strains. After overnight incubation in BHI-T80 at 37°C, the bacterial cells were loaded onto grids, incubated with the anti-CPS MAb (ER21) or anti-PCho MAb (TEPC-15), and subsequently incubated with a 10-nm colloidal gold particle-conjugated secondary antibody.

Interestingly, cell wall antigen analysis revealed that in contrast to Staphylococcus strains, E. rhusiopathiae strains showed no bands, suggesting that this species lacks canonical WTAs (Fig. 4).

FIG 4.

PAGE analysis of cell wall preparations of E. rhusiopathiae and Staphylococcus aureus strains. The cell wall antigens were extracted with 0.1 N NaOH (A) and 5% TCA (B), and equal amounts of the extracts were subjected to PAGE. Lanes: 1, E. rhusiopathiae Fujisawa; 2, E. rhusiopathiae Δ432; 3, S. aureus Wood 46; 4, S. aureus ATCC 12600.

Virulence of the Δ432 strain in germfree pigs.

The virulence of the Δ432 strain was assessed using germfree pigs and compared to that of the Koganei strain. The piglets (n = 4) s.c. inoculated with 10⁹ CFU of the Δ432 strain survived to the end of the observation period (10 days), whereas all the piglets (n = 4) inoculated with 10⁹ CFU of the Koganei strain died within 3 days after inoculation. In vivo growth of the Δ432 strain in the germfree piglets was examined (Table 3). The Δ432 strain was isolated from the tonsils of all piglets and from all organs examined in piglet 4.

TABLE 3.

In vivo growth of the Δ432 strain in germfree piglets after s.c. inoculation

Organa	No. (CFU) of isolates in:
	Pig 1	Pig 2	Pig 3	Pig 4
Tonsil*	3.8 × 101	1.3 × 101	8.1 × 102	3.1 × 104
Heart*	0	0	0	1.1 × 106
Liver*	0	0	0	3.7 × 102
Spleen*	0	0	0	3.2 × 102
Kidney*	0	0	0	3.7 × 103
Mesenteric lymph nodes*	0	0	0	3.0 × 101
Inguinal lymph nodes*	1.9 × 103	0	5.1 × 103	2.2 × 104
Skin site of inoculation*	0	0	0	7.0 × 103
Lung*	NTb	NT	NT	4.7 × 102
Elbow joint cavity**	0	2 × 101	6.1 × 102	1.6 × 103
Knee joint cavity**	0	0	0	4 × 102
Blood***	1 × 101	1 × 101	1 × 101	3 × 101

^aGerm-free piglets were inoculated s.c. with 10⁹ CFU of the Δ432 strain, and tissue samples were collected 10 days after inoculation. *, CFU/g of sample; **, CFU per swab sample; ***, CFU/ml.

^bNT, not tested.

Protective capability of the Δ432 strain in conventional pigs.

The protective capability of the Δ432 strain via two routes of immunization (subcutaneous and oral) was assessed using conventional pigs and compared to that of the Koganei strain.

After oral immunization, in contrast to the control pigs, which died or were euthanized on day 3 after inoculation, all pigs inoculated with 10¹⁰ CFU of the Δ432 strain or the Koganei strain for 3 consecutive days did not show any signs of disease, i.e., anorexia or depression. All pigs in the two groups challenged with 10⁸ CFU of the Fujisawa strain on day 16 after immunization were completely protected, without any signs of infection (anorexia, depression, lameness, and erythema). Analysis of humoral immune responses revealed that the Δ432 strain could induce levels of agglutinating IgG antibodies comparable to those elicited by the Koganei strain after challenge infection (Table 4).

TABLE 4.

Agglutinating IgG antibody titers after oral inoculation of the Δ432 and Koganei strains in conventional pigs^a

Pig group	Pig no.	Agglutinating IgG antibody titerb
		Before vaccination	8 days after vaccination	14 days after vaccination	18 days after vaccination	7 days after challenge
Nonvaccinated	1	8	<4	<4	4	Died
	2	4	<4	<4	4	Died
Δ432 vaccinated	3	4	4	8	8	16
	4	8	<4	4	4	64
	5	4	4	4	4	32
Koganei vaccinated	6	8	<4	4	8	16
	7	4	4	4	8	16
	8	4	8	32	32	64

^aPigs were inoculated orally with 10¹⁰ CFU of the strains in milk replacer for 3 consecutive days and challenged with 10⁸ CFU of the Fujisawa strain on day 16 after the final vaccination.

^bThe agglutination titer was determined as described in the footnote of Table 2.

To assess subcutaneous immunization, the pigs were inoculated s.c. with 10⁹ CFU of each strain and then challenged with 10⁸ CFU of the Fujisawa strain on day 13 after immunization. On day 2 after the challenge infection, the control pigs that received no immunization were euthanized due to severity of symptoms, whereas the pigs immunized with the Δ432 strain or the Koganei strain survived without any signs of infection. The levels of agglutinating IgG antibodies were similar between the two immunized groups on day 13 after immunization and on day 7 after challenge infection (Table 5).

TABLE 5.

Agglutinating IgG antibody titers after s.c. inoculation of the Δ432 and Koganei strains in conventional pigs^a

Pig group	Pig no.	Agglutinating IgG antibody titerb
		Before vaccination	6 days after vaccination	13 days after vaccination	7 days after challenge
Nontreated control	15	4	4	4	Died
	16	4	4	4	Died
Δ432 vaccinated	12	4	4	128	64
	13	4	4	4	64
	14	4	4	4	1,024
Koganei vaccinated	9	4	4	512	512
	10	4	4	4	256
	11	4	4	64	128

^aPigs were inoculated s.c. with 10⁹ CFU of the strain and challenged with 10⁸ CFU of the Fujisawa strain on day 13 after vaccination.

^bThe agglutination titer was determined as described in the footnote of Table 2.

The protective capability of the Δ432 strain via the oral route was further examined with additional conventional pigs. In contrast to the control group, in which two pigs died and one pig was euthanized on day 3 after challenge infection, the single-dose vaccinated group had only one pig die on day 5 after the challenge infection, and the rest of the pigs in the single- and two-dose vaccinated groups were healthy and completely protected (P < 0.001). In a postmortem analysis, E. rhusiopathiae bacteria were recovered from only the tonsils of 15 pigs (5 and 10 from the single- and two-dose groups, respectively) and only the heart of 1 pig in the two-dose group and were not recovered from blood or other organs (lungs, livers, kidneys, spleens, mesenteric lymph nodes, knee joint cavities, and elbow joint cavities) (Table 6). PCR analysis confirmed that all the colonies tested (4 to 32 per organ) were positive for the Δ432 strain and that the challenge strain was not recovered (Table 6).

TABLE 6.

Recovery of E. rhusiopathiae from pig organs after oral vaccination

Group	Pig no.	No. of bacteria, log10 CFU/g (no. of colonies PCR positive for the Δ432 strain/no. of colonies tested), isolated from:
		Tonsil	Heart
Single dose	1	–a	–
	2	–	–
	3	–	–
	4	–	–
	5	4.71 (21/21)	–
	6	3.48 (16/16)	–
	7	3.49 (18/18)	–
	8	4.64 (23/23)	–
	10	5.89 (32/32)	–
Two doses	11	6.45 (28/28)	3.28 (13/13)
	12	5.90 (25/25)	–
	13	5.83 (26/26)	–
	14	6.90 (29/29)	–
	15	5.82 (26/26)	–
	16	6.32 (30/30)	–
	17	5.70 (30/30)	–
	18	6.56 (32/32)	–
	19	4.65 (32/32)	–
	20	2.87 (4/4)	–

^a–, Not recovered. Bacteria were also not recovered from the samples of lungs, livers, kidneys, spleens, mesenteric lymph nodes, knee joint cavities, and elbow joint cavities.

DISCUSSION

Attenuated E. rhusiopathiae vaccines have long been used for oral or nasal immunization of turkeys (20) and pigs (21, 22) and for subcutaneous immunization of pigs in Japan. However, these conventional vaccines can pose a risk of residual virulence and reversion to virulence. In this study, we rationally designed a safe, effective E. rhusiopathiae vaccine that can be delivered via oral or parenteral administration in pigs.

Genome-wide analysis of virulence genes using 651 transposon mutants of E. rhusiopathiae revealed 23 novel virulence genes. Immunogenicity analysis of mutants for these 23 genes in mice revealed that four mutants conferred partial or no protection against challenge infection with the parent strain, suggesting high attenuation of the mutants. Among the rest of the mutants that induced complete protection, six mutants caused ruffled fur, which is a general clinical sign of the infection, in mice after the screening infection with 10⁸ CFU; these mutants were selected and tested for their protective capability in pigs.

The pig experiments revealed that compared to the Koganei vaccine strain, all six mutants tested induced strong cell-mediated immune responses. However, not all the mutants induced a strong humoral immune response, and the pigs (animals 7, 19, and 21) that showed low levels of agglutinating IgG antibodies died after challenge infection. These results are supported by the finding that the protective activity of antiserum against E. rhusiopathiae was found in only the IgG fraction but not in the IgM fraction (23), and IgG antibodies are necessary as opsonins to induce a high level of protective immunity against swine erysipelas (24, 25). Clone 117 (with a transposon insertion in the ERH_0432 gene) induced levels of agglutinating IgG antibodies comparable to those seen with the Koganei strain and induced strong cell-mediated immune responses after the challenge infection, thus indicating that the mutant effectively primed the immune system. It has been suggested that the intensity of skin erythema induced by inoculation of E. rhusiopathiae vaccine strains correlates well with their protective immunity against swine erysipelas (18). However, the pigs immunized with clone 117 did not develop skin erythema. Thus, our results suggest that skin erythema is not always an indicator of strong immune responses for certain attenuated mutants of E. rhusiopathiae.

The gene deletion mutant Δ432 strain was found to be highly attenuated in germfree piglets, causing no severe symptoms or death. Furthermore, when this strain was tested in conventional pigs via oral and subcutaneous immunization, the pigs were completely protected against challenge infection. A larger-scale pig experiment further confirmed the high efficacy of the Δ432 strain via the oral route, showing that the tag mutant of E. rhusiopathiae represents a safe oral vaccine candidate.

ERH_0432 is a tagF homolog that encodes a CDP-glycerol glycerophosphotransferase (15), which is involved in WTA biosynthesis in Gram-positive bacteria (26, 27). WTAs of Gram-positive bacteria have been shown to play crucial roles in many fundamental aspects of physiology, as well as virulence (26–28). Most of the genes involved in Gram-positive WTA biosynthesis are conditionally essential, and their deletion is lethal (29). However, deletion of the ERH_0432 gene was possible on a wild-type background. Intriguingly, we found that E. rhusiopathiae does not express canonical WTAs. In contrast to other Gram-positive bacteria, including Bacillus subtilis and S. aureus, in which WTA genes are clustered in the genome (30), tag homologs of E. rhusiopathiae are dispersed over the chromosome (15): tagD (ERH_1439), tagF (ERH_0432 and ERH_1440), tagH (ERH_0689), and tagO (ERH_0529), the last of which is fused with mraY (15), a gene essential for peptidoglycan biosynthesis. E. rhusiopathiae lacks clear orthologs of ltaS, which encodes the polyglycerophosphate lipoteichoic acid synthase and is indispensable in S. aureus (31, 32). Given the unique phylogenetic position of E. rhusiopathiae (15), these findings suggest that among Gram-positive bacteria, E. rhusiopathiae possesses a unique cell wall architecture (16, 17). In Western blot and electron microscopy analyses, it was found that the Δ432 strain exhibits greatly reduced levels of a CPS epitope, and the electrophoretic mobility patterns of PCho were different from those of the parent strain Fujisawa. The loss of the CPS epitope and the change in its expression pattern have also been observed in the mutants of the genes ERH_1441, ERH_1442, ERH_1444, ERH_1449, and ERH_1450, all of which were found to be involved in the production of heat-stable peptidoglycan antigens and responsible for determining serovar 1a antigenicity in the Fujisawa strain (17). Thus, the function of the ERH_0432 gene encoding a CDP-glycerol glycerophosphotransferase appears to be very complex, and the role(s) of the gene in the synthesis of cell wall antigens remains unknown. Clarification of the gene function and elucidation of the structure and biosynthesis of the cell wall components of E. rhusiopathiae may provide interesting insights into the roles of WTAs in physiology and virulence in other Gram-positive bacteria.

MATERIALS AND METHODS

Bacterial strains and growth conditions.

The E. rhusiopathiae strain Fujisawa (serotype 1a) was used to generate a transposon mutant library. The Fujisawa strain was originally isolated from a septicemic pig and is highly virulent in mice (the LD₅₀ is approximately 16 CFU for subcutaneous inoculation). The E. rhusiopathiae Fujisawa derivatives used in this study are shown in Table 1. All E. rhusiopathiae strains were grown at 37°C in brain heart infusion broth (BHI; Becton Dickinson, Baltimore, MD) supplemented with 0.1% Tween 80 and 0.3% Tris (pH 8.0; BHI-T80), except in a pig vaccine study where the Δ432 strain was cultured in Trypticase soy broth (Becton Dickinson) supplemented with 0.1% Tween 80 (TSB-T80). The Escherichia coli strains used were JM109 (Toyobo, Tokyo, Japan) and DH5α (Toyobo). Cultivation of the E. coli strains was performed as described previously (25). The Staphylococcus aureus Wood 46 and ATCC 12600 strains, which were used for sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) analysis of WTAs, were cultured in BHI.

DNA methods.

(i) Isolation of DNA. Genomic DNA from the E. rhusiopathiae strains was prepared as described previously (10). Plasmid DNA was isolated from E. coli by using a plasmid miniprep kit (Promega, Madison, WI) according to the manufacturer’s protocol.

(ii) PCR and sequencing. PCR was performed with KOD FX DNA polymerase (Toyobo) and a Bio-Rad T-100 thermal cycler (Bio-Rad, Richmond, CA). The sequence of amplified DNA fragments was confirmed by direct sequencing with an ABI Prism 3130xl genetic analyzer (Applied Biosystems, Foster City, CA) using a BigDye Terminator v3.1 cycle sequencing kit (Applied Biosystems) according to the manufacturer’s instructions. All oligonucleotides were purchased from Hokkaido System Science (Sapporo, Japan).

(iii) Construction of transposon mutants. A transposon mutant library of the Fujisawa strain was constructed using a mariner-based transposition system with pMC plasmids, as previously described (16, 17, 33). Briefly, the thermosensitive pMC plasmids were introduced into the Fujisawa strain by electroporation, as described previously (11), and transformants were selected at 30°C on BHI-T80 plates supplemented with erythromycin (1 μg/ml). Individual colonies were grown in BHI-T80 with erythromycin at 30°C. The cultures were diluted and plated on BHI-T80 containing erythromycin at 30°C and then shifted to 40°C to eliminate the pMC plasmids. Erythromycin-resistant clones resulting from integration of the transposon into the chromosome were picked and passaged on BHI-T80 plates supplemented with erythromycin. The transposon insertion site was determined as previously described (17) by arbitrary PCR and subsequent sequencing of the transposon-flanking DNA regions of transformants that contained a single transposon insertion.

(iv) Construction of the Δ432 strain. Construction of a gene fragment with deletion of the ERH_0432 gene was performed using the PCR-based overlap extension method (34) using genomic DNA from the Fujisawa strain. Briefly, the first PCR was performed to amplify the 5′ and 3′ ends of a chromosomal region encompassing the ERH_0432 gene and its flanking regions in parallel, using two primer sets, ERH_0429BglII_F (5′-GAAGATCTATGCTTAAAGAAACAGTTTT-3′)/delERH_0432R (5′-CAAATACAACATTGTGCTTTAATATATAAT-3′) and delERH_0432F (5′-AAAGCACAATGTTGTATTTGGATAAATTAA-3′)/ERH_0435SalI_R (5′-ACGCGTCGACATGAAAACAACAATTTATCT-3′); the two inner primers were designed to overlap each other at the underlined sequences and eliminate 365 amino acids (i.e., amino acids 13 to 377) of the ERH_0432 gene product. The second PCR was performed with the two outermost primers, ERH_0429BglII_F and ERH_0435SalI_R, using a mixture of gel-purified PCR products from the first step as the template. PCR amplification was performed with the following conditions: an initial denaturation step at 94°C for 2 min and three amplification steps (35 cycles) consisting of 98°C for 10 s, 50°C for 30 s, and 68°C for 2 min (first PCR) or 4 min (second PCR).

Inactivation of the ERH_0432 gene in the E. rhusiopathiae Fujisawa strain was performed using the shuttle vector plasmid pMAD, as previously described (11, 17, 35). To confirm the in-frame gene deletion in the Fujisawa strain, a draft genome sequence was generated using the Illumina HiSeq platform. The draft genome sequence of the resulting Δ432 strain was analyzed as described elsewhere (10) by comparison to the whole-genome sequence of the Fujisawa strain (accession no. AP012027).

Phenotypic characterization of the Δ432 strain.

(i) Expression of CPS and PCho. The expression of the CPS antigens and CPS modification with PCho were examined as previously described (16) using the MAbs ER21 and TEPC-15 (Sigma-Aldrich, St. Louis, MO), which detect the CPS and PCho, respectively. Briefly, E. rhusiopathiae strains were cultured in 10 ml of BHI-T80 at 37°C for 20 h. Bacterial cells harvested by centrifugation were washed with 20 mM Tris-HCl (pH 7.6), suspended in 0.5 ml of 20 mM Tris (pH 7.6) containing 0.5% Triton X-100 and incubated at 37°C for 1 h with rotation. After incubation, the bacterial cells were removed by centrifugation, and the supernatants containing crude capsular antigens were used for immunoblot analysis. Immunogold electron microscopy was performed as previously described (16).

(ii) SDS-PAGE analysis of cell wall antigen preparations of E. rhusiopathiae and S. aureus strains. Cell wall antigen preparations intended for WTA extraction were performed using NaOH or trichloroacetic acid (TCA), mostly following the protocol of Pollack and Neuhaus (36), except for the following modifications: purified bacterial cell pellets were treated with 0.1 N NaOH for 16 h at room temperature or 5% TCA at 4°C for 40 h, and both of the treatments were performed with constant inversion.

Animal experiments.

All animal experiments performed in this study were approved by the Animal Ethics Committee of the National Institute of Animal Health (NIAH), Tsukuba, Ibaraki, Japan.

The mice used in this study were 6- to 8-week-old female ddY mice purchased from Japan SLC, Inc. (Hamamatsu, Japan). Conventional pigs were bred in the NIAH and used at the age of 3 to 4 weeks. Germ-free pigs were produced by hysterectomy in the NIAH and used at the age of 2 weeks.

(i) Screening transposon mutants for attenuation and protective capability in mice. Transposon mutants were screened for attenuation and protective capability. The mice were s.c. inoculated with 10⁸ CFU of each of the mutants (two mice per mutant) and observed for 14 days. The surviving mice were subsequently challenged s.c. with 10³ CFU of the parent Fujisawa strain, and the protective capability was assayed.

(ii) Virulence and protective capability of the transposon mutants in pigs. The virulence and protective capability of six transposon mutants, which caused ruffled fur in mice after screening inoculation with 10⁸ CFU, were tested in pigs and compared to those of the Koganei strain. Groups of pigs were inoculated s.c. with 10⁹ CFU of each transposon mutant and observed for clinical symptoms for 14 days. The pigs were subsequently challenged s.c. with 10⁸ CFU of the Fujisawa strain and further observed for 8 days. During the observation period, humoral and cell-mediated immune responses were monitored by a GA test (24) and a blastogenesis assay (12, 13), respectively. In the GA test, serum was pretreated with the same amount of 0.2 M 2-mercaptoethanol (2-ME), which breaks disulfide bonds and depolymerizes IgM, for the determination of agglutinating IgG antibody titers (37).

(iii) Virulence of the Δ432 strain in germfree pigs. The virulence of the Δ432 strain was tested using germfree pigs, which are highly susceptible to E. rhusiopathiae infection, and compared to that of the Koganei strain. Germ-free piglets (n = 4 for each strain) were inoculated s.c. with 10⁹ CFU of the Δ432 strain or the Koganei strain and observed for 10 days. At necropsy, tissue samples were collected, and the in vivo growth of the Δ432 strain was examined.

(iv) Protective capability of the Δ432 strain in conventional pigs. The protective capability of the Δ432 strain was compared to that of the Koganei strain using conventional pigs via two routes of immunization. To assess oral immunization, the pigs were inoculated with 10¹⁰ CFU of each strain in milk replacer (SPF-LAC; Weyerhaeuer, Eaton, OH) for 3 consecutive days. On day 16 after the final day of administration, the pigs were challenged s.c. with 10⁸ CFU of the Fujisawa strain. To assess s.c. immunization, the pigs were inoculated with 10⁹ CFU of each strain. On day 13 after immunization, the pigs were challenged s.c. with 10⁸ CFU of the Fujisawa strain. In these experiments, the challenged pigs were observed for 10 days. During the observation period, serum agglutinating IgG antibody titers were determined by the GA test, as described above.

The vaccine efficacy of the Δ432 strain via the oral route was further examined with additional conventional pigs. To meet the requirements for biological products, the Δ432 strain was grown in Trypticase soy broth that contained minimal animal-derived products. Three-week-old piglets divided into three groups of 10, 10, and 3 pigs per group were housed separately in different isolation rooms. The pigs were allowed to freely consume milk replacer mixed with the bacterial suspension in plastic containers; 10 pigs in group 1 were fed once, and 10 pigs in group 2 were fed twice on two consecutive days. The theoretical dose of vaccination was 1.0 × 10¹⁰ CFU of the strain per pig per day. As controls, 3 pigs were fed milk replacer containing no bacteria. The pigs were challenged intradermally with 1.0 × 10⁸ CFU of the Fujisawa strain 23 days after the final day of vaccination of group 2 and monitored for clinical signs and death for 14 days. In a postmortem analysis, blood, tonsil, heart, lung, liver, kidney, spleen, mesenteric lymph node, knee joint cavity, and elbow joint cavity samples were examined for the presence of E. rhusiopathiae bacteria by plating on BHI-T80 agar supplemented with 0.001% crystal violet, 0.03% sodium azide, 400 μg/ml kanamycin, and 25 μg/ml gentamicin. Colonies that appeared on the selective plates were checked (4 to 32 colonies per organ) by a PCR assay with the primers ERH_0431F (5′-CGTGATCGTAACGATAAGCC-3′) and ERH_0433R (5′-GATGCAGTAAAGCCTGGGTC-3′) for differentiation of the Δ432 strain and the challenge Fujisawa strain.

Statistical analysis between the mortality of the groups was performed by Kaplan-Meier analysis, followed by the log-rank test, using Prism 5.0 software (GraphPad, La Jolla, CA).