INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Erysipelothrix rhusiopathiae (formerly E. insidiosa) is a small gram-positive rod that causes erysipelas mainly in swine (34) and turkeys (2) and less frequently in other animals and humans. E. rhusiopathiae was once thought to be the only member of genus Erysipelothrix and was classified into 23 serotypes and type N based on peptidoglycan antigens of the cell wall (9, 34). The genus now contains two species, E. rhusiopathiae and E. tonsillarum, and other (two) genetically distinct unclassified groups (24, 25). Among 15 serotypes of E. rhusiopathiae (25), serotypes 1 (subdivided into 1a and 1b) and 2 (subdivided into 2a and 2b) are the most important in the pig industry (3, 4, 17, 26, 32, 34). Species other than E. rhusiopathiae are of low virulence in swine (25). Because of the importance of swine erysipelas, killed and attenuated live vaccines having been used extensively. However, despite widely practiced vaccination, the importance of this disease has not decreased (34). In Japan, annually about 2,000 pigs are affected with acute and subacute septicemia and about 2,000 pigs are condemned by meat inspection authorities because of arthritis.

There are many reports on the characterization of protective antigens of E. rhusiopathiae. A mouse protective antigen was identified in culture supernatant (30, 31) and in 10 mM NaOH extracts of bacterial cells (18, 19). Neuraminidase was also considered a protective antigen, because mice were protected by passive immunization with rabbit antiserum against E. rhusiopathiae neuraminidase (16). In 1 mM EDTA extract of T28 (serotype 2), one major polysaccharide capsular antigen of 14.4 to 22 kDa and two main protein antigens of 64 and 48 kDa were revealed by immunoblotting with rabbit antiserum (10). However, the protective activity of these antigens was not examined in the work reported. Groschup et al. (7) showed that protective antisera from pigs recognized prominent bands of 64 to 66 and 39 to 40 kDa in 1 mM EDTA and 10 mM NaOH extracts of T28. Both antigens were trypsin sensitive and contained no detectable polysaccharide. Mice immunized with preparations of the 64- to 66-kDa band were protected against challenge with Frankfurt XI (serotype N). They also described the enhanced production of these protective antigens in serum-free modified Feist broth (6). However, some questions remain as to whether the band of 64 to 66 kDa is the only protective antigen of the bacteria, this band contains only one kind of protective antigen, and this antigen can elicit protection in swine.

Galan and Timoney first identified a mouse protective antigen gene in a 5.4-kb EcoRI fragment of chromosomal DNA of virulent strain E1-6P (serotype 1a) (5). Guinea pig antiserum against the recombinant clone of this gene reacted with E. rhusiopathiae protein antigens of 66, 64, and 43 kDa. These proteins are of the same size as the protective proteins mentioned above. However, the DNA sequences of the gene were not described, and it is also not known whether the clone contained only one gene. Recently, a novel surface protective antigen (SpaA) of E. rhusiopathiae was identified from serotype 2 in a mouse model using a monoclonal antibody recognizing 64-kDa proteins of most serotypes of E. rhusiopathiae (13). Mice immunized with live recombinant Escherichia coli intraperitoneally survived after challenge with the same strain of E. rhusiopathiae. In this study, the presence of 20-amino-acids repeat units at the C terminus was shown to be essential for protection.

In contrast to protein antigens, 14.4- to 22-kDa capsular antigen appears to be not necessary for protection. A live acapsular mutant created by insertion and excision of Tn916, which is avirulent in mice, could confer complete protective immunity to mice (21).

The existence of a common protective antigen among serotypes of E. rhusiopathiae was identified experimentally and practically. Although killed vaccines are prepared from serotype 1a and live vaccines are from serotype 2, both vaccines can cross-protect pigs against challenge with strains of serotypes 1 and 2 (1, 23, 33). In this study, we cloned spaA from a Sau3AI library of virulent strain Fujisawa (serotype 1a), determined that the N-terminal 342 amino acids are necessary to elicit protection in mice, and evaluated the ability of SpaA of strain Fujisawa (SpaA/Fujisawa) to elicit cross-protection in pigs against challenge with serotypes 1 and 2 by the use of fusion of truncated SpaA/Fujisawa with a histidine hexamer (HisSpa1.0).

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Bacterial strains, vectors, and growth conditions. Bacterial strains and plasmids used in this study are listed in Table 1. E. rhusiopathiae Fujisawa was used for cloning of spaA, preparation of intact SpaA for enzyme-linked immunosorbent assay (ELISA), and challenge of mice and pigs. For challenge of pigs, E. rhusiopathiae 82-875 was also used. Vector plasmid pBluescript II SK+ (Stratagene) was used for cloning of spaA, and expression vector pQE32 (Qiagen) was used to construct HisSpa1.0, in which the histidine hexamer tag was placed at the N terminus of the protein. E. coli XL1-Blue was used as the host strain for these plasmids.

Preparation of antisera. A pig was first immunized with live vaccine and then injected with 10⁷ live cells of Fujisawa intradermally once and with 10⁸ to 10⁹ live cells intravenously three times at weekly intervals. Serum was collected 3 weeks after the final injection. A rabbit was immunized by weekly injection of 1 mg of purified HisSpa1.0 subcutaneously three times with Freund's complete adjuvant and after 2 weeks intravenously twice without adjuvant. Serum was collected 2 weeks after the final injection.

DNA extractions. Chromosomal DNA of E. rhusiopathiae for use in cloning and PCR was extracted and purified as described previously (5, 12). Plasmid DNA was prepared by a modified alkaline lysis-polyethylene glycol precipitation procedure in a dye terminator cycle sequencing protocol (Perkin-Elmer).

Determination of protective region of spaA. Two spaA recombinant plasmids, pA and pB, obtained from a Sau3AI library of Fujisawa were used. Lysates of recombinant E. coli carrying these plasmids could elicit protection in mice. The protective region of the gene was determined by analyzing the Exo-Mung deletion mutants of these plasmids created as instructed by the manufacturer (Stratagene) by immunoblotting and confirmed by a mouse protection test.

Expression and purification of fusion protein. A KpnI fragment of recombinant plasmid pA containing bp 266 to 1,294 of spaA was ligated into the compatible site of expression vector pQE32 to construct pA1.0 as shown in Fig. 1. One KpnI site was located in spaA, and another was located in the multicloning site of pBluescript II. Recombinant fusion protein HisSpa1.0 was expressed in E. coli XL1-Blue(pA1.0) and purified as specified by the manufacturer (Qiagen) under denaturing conditions. The culture was inoculated with a 1:50 dilution of overnight culture of recombinant E. coli, grown at 37°C to mid-exponential phase (A₆₀₀ = 0.8), and then induced with 2 mM isopropyl--D-thiogalactopyranoside for 5 h with vigorous shaking. Cells were harvested and resuspended in 6 M guanidine buffer (pH 8.0) at 0.2 g (wet weight)/ml and stirred for 1 h at room temperature to solubilize the fusion protein. The slurry was centrifuged at 10,000 × g for 15 min; then the fusion protein in the supernatant was filter sterilized and adsorbed on Ni-nitrilotriacetic acid (NTA)-Sepharose column. The Ni-NTA column was washed with guanidine buffer (pH 8.0), 8 M urea buffer (pH 8.0), and 8 M urea buffer (pH 6.3); then the fusion protein was eluted with urea buffer (pH 4.5) and dialyzed against 10 mM phosphate-buffered saline, pH 7.2 (PBS).

View larger version (27K): [in this window] [in a new window]   FIG. 1.   Map of spaA/Fujisawa in recombinant plasmids pA, pB, and pA1.0 and map of HisSpa1.0 in intact SpaA. Protective regions are indicated as dotted boxes.

Mouse immunization and challenge. Four-week-old ddY female mice were immunized subcutaneously with 500 µg of sonicated extract of recombinant E. coli in Freund's incomplete adjuvant or with 50 µg of purified HisSpa1.0 in complete adjuvant twice and challenged with 100 50% lethal doses of E. rhusiopathiae Fujisawa subcutaneously 3 weeks after immunization. The infections were monitored for 12 days, and the cause of death was confirmed by isolation of the organism.

Pig immunization and challenge. Four-week-old specific-pathogen-free (SPF) pigs obtained from a farm free from swine erysipelas where no pigs were vaccinated against the disease were used. Six pigs were divided into three groups and immunized intramuscularly with 0, 100, and 500 µg of purified HisSpa1.0 in Freund's complete adjuvant twice at 3-week intervals and 2 weeks later challenged intradermally with 4 × 10⁷ Fujisawa (serotype 1a) organisms. Another four pigs were divided into two groups and immunized with 0 and 100 µg of purified HisSpa1.0 twice at 4-week intervals and 2 weeks later challenged with 8 × 10⁷ 82-875 (serotype 2b) bacteria. Dead pigs were autopsied on the day of death, and pigs that survived were euthanized and autopsied 1 week after challenge. Organs (heart, lung, liver, spleen, kidney, lymph nodes, and tonsils) and skin erythema lesions of all pigs were examined by bacterial isolation. Sera were collected from all pigs every week through the experiment, and antibody was titrated by double-antibody sandwich ELISA with intact SpaA, indirect ELISA with HisSpa1.0, and growth agglutination test.

Preparation of intact SpaA in alkaline extract of E. rhusiopathiae. E. rhusiopathiae Fujisawa was cultured in modified Feist broth at 37°C overnight (6). Cells were harvested and washed with distilled water, then resuspended in 10 mM NaOH at 0.1 g (wet weight)/ml, and incubated at 4°C overnight with gentle shaking (7). After neutralization, the suspension was centrifuged at 10,000 rpm for 30 min; then the supernatant was sterilized by filtration and kept at 20°C. The extract was used in double-antibody sandwich ELISA.

SDS-PAGE and immunoblotting. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was performed by the method of Laemmli (11) on 10% gels. After semidry electrophoretic transfer of the antigens to a nitrocellulose membrane, the membrane was blocked with 3% skim milk in PBS supplemented with 0.05% Tween 20 (PBS-Tw) for 30 min, incubated with pig antiserum diluted 1:100 with 1% skim milk in PBS-Tw for 1 h, washed with PBS-Tw three times for 15 min, and then incubated with anti-pig immunoglobulin G (IgG)-peroxidase conjugate (Rockland) diluted 1:1,000 with 1% skim milk in PBS-Tw for 1 h. Membrane was washed as described above and then developed with 0.05% 4-chloro-1-naphthol and 0.01% hydrogen peroxide in PBS-Tw for 15 to 30 min. The reaction was stopped by washing the membrane with PBS-Tw.

ELISA to detect antibody response in pigs against SpaA. In sandwich ELISA, rabbit antiserum against HisSpa1.0 was used to capture intact SpaA in the alkaline extract of E. rhusiopathiae Fujisawa to the ELISA plate well. All steps were done with a 100-µl reaction volume and 1-h incubation at room temperature. Between each step, the plate was washed three times with PBS-Tw.

Growth agglutination test. A modification of the method of Wellmann (29) was used. To prepare the serum dilution and bacterial suspension, tryptic soy broth (pH 7.8) (Difco) supplemented with 0.1% Tween 80, 0.3% Tris, kanamycin (500 µg/ml), and gentamicin (25 µg/ml) (S-TSB) was used. Each 50 µl of pig serum was serially diluted with 50 µl of S-TSB in a sterile round-bottomed microplate and then mixed with an equal volume of overnight culture of E. rhusiopathiae Marienfelde diluted 1:100 with S-TSB. The plate was sealed and incubated at 37°C overnight. The agglutination titer was expressed as the reciprocal of the highest final serum dilution giving agglutination.

In vitro phagocytosis assay. Peripheral blood was collected from an SPF pig not immunized against swine erysipelas. Mononuclear cells and neutrophils were isolated by standard density gradient centrifugation on Ficoll-Paque (Pharmacia). The mononuclear cell fraction was washed and resuspended at 5 × 10⁶/ml (monocytes = 10⁶/ml) in 5 ml of RPMI 1640 medium (Sigma) supplemented with 10% fetal calf serum and 25 mM HEPES (pH 7.2) (RPMI-FCS). Monocytes were regarded as 20% of the mononuclear cells and were not separated from lymphocytes. Neutrophils were purified by ammonium chloride lysis of erythrocytes and after washing resuspended in 5 ml of RPMI-FCS at 10⁶/ml. An overnight culture of Fujisawa was centrifuged and resuspended at 10⁸/ml in 0.5 ml of a 1:1 mixture of inactivated pig serum, collected from an immunized pig or nonimmunized control pig at the time of challenge, and fresh germfree pig serum. The suspension was incubated at 37°C for 1 h with gentle agitation to opsonize the bacteria, then added to the each phagocyte suspension at a final concentration of 10⁷ CFU/ml, and incubated at 37°C with gentle shaking. After 30 min, the phagocytes were washed with medium twice, resuspended in 10 ml of RPMI-FCS containing penicillin (10 U/ml) to inhibit the growth of E. rhusiopathiae in the medium, and reincubated at 37°C. Each 2 ml of the suspension was collected after 0, 2, 4, and 18 h; 1 ml was used for Giemsa staining, and another 1 ml was used for counting the bacteria surviving in phagocytes.

PCR cloning of protective region of spaA. Each DNA fragment flanking the 1,026-bp protective region of spaA was amplified by PCR from chromosomal DNA of Koganei (serotype 1a, Japanese official live vaccine strain), SE-9 (serotype 2a, U.S. official vaccine strain), and ATCC 19414^T (serotype 2b, type strain of E. rhusiopathiae), and cloned into pBluescript II. DNA sequences of these fragments were determined by cycle sequencing on a model 373S automated DNA sequencer (Applied Biosystems).

PCR-RFLP of spaAs of diverse serotypes. spaAs of almost full size (bp 16 to 1871) were amplified by PCR with a primer set designed from sequences of spaA/Fujisawa from chromosomal DNA of Fujisawa (serotype 1a), Koganei (serotype 1a), SE-9 (serotype 2a), ATCC 19414^T (serotype 2b) and E. rhusiopathiae reference strains of all 12 serotypes harboring this gene. Each 1 µg of the purified PCR products was digested with restriction enzymes EcoRI, HpaII, KpnI, PstI, and SacI, respectively, and restriction fragment length polymorphism (RFLP) was analyzed by 1% agarose gel electrophoresis.

Nucleotide sequence accession number. The nucleotide sequence of spaA/Fujisawa will appear in the DDBJ/EMBL/GenBank nucleotide sequence databases with accession no. AB019124. Nucleotide sequences of protective regions of spaA/Koganei, spaA/ATCC 19414^T, and spaA/SE-9 will appear in the DDBJ/EMBL/GenBank nucleotide sequence databases with accession no. AB024082, AB024083, and AB024084, respectively.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Determination of protective region of spaA. The spaA/Fujisawa recombinant plasmids pA and pB had bp 1 to 1,294 of spaA in a 1.7-kb insert and bp 1 to 1,881 of full-length spaA in a 3.8-kb insert, as shown in Fig. 1. Analysis of Exo-Mung deletion mutants of them showed that an approximately 1.0 kb C-terminal region of the insert of pA was necessary to elicit protection in mice. This was confirmed by immunizing mice with 50 µg of purified HisSpa1.0 in complete adjuvant twice and successive challenge. After challenge, four of five mice survived.

Purification of truncated SpaA by affinity chromatography. The elution profile of HisSpaA1.0 at each stage of the purification procedure analyzed by SDS-PAGE and immunoblotting (Fig. 2). Purified HisSpa1.0 showed the predicted molecular size of 45.5 kDa and was reactive with pig antiserum immunized with E. rhusiopathiae. Both mouse and pig antisera against HisSpa1.0 reacted well with the 69.0-kDa intact SpaA in the alkaline extract and with a 43-kDa SpaA fragment in the culture supernatant of Fujisawa (data not shown).

View larger version (46K): [in this window] [in a new window]   FIG. 2.   SDS-PAGE and immunoblot analysis of affinity purification of HisSpa1.0 on an Ni-NTA-agarose column. Lanes: 1, crude extract; 2, flowthrough fraction; 3, purified protein eluted at pH 4.5. (A) Coomassie blue staining of polyacrylamide gel; (B) immunoblot detection with pig serum immunized with live E. rhusiopathiae.

Pig protection assay. All pigs immunized with purified HisSpa1.0 in Freund's complete adjuvant were protected completely against intradermal challenge with virulent strain Fujisawa (serotype 1a) and strain 82-875 (serotype 2b) (Table 2). They showed no clinical symptoms and no urticarial lesions at the injection site. Bacterial isolation 1 week after challenge was also negative by both direct culture and enrichment culture.

View larger version (35K): [in this window] [in a new window]   FIG. 3.   Antibody response of pigs immunized with two injections of various doses of HisSpa1.0 at weeks 0 and 3 of the experiment. Sera were collected at 0, 1, 2, 3, and 5 weeks, and pigs were challenged at 5 weeks of the experiment. (A) Sandwich ELISA with intact SpaA; (B) indirect ELISA with HisSpa1.0; (C) growth agglutination test. Pigs 1 and 2, 0 µg; pigs 3 and 4, 100 µg; pigs 5 and 6, 500 µg.

Effect of pig serum immunized with HisSpa1.0 on in vitro phagocytosis. In vitro phagocytosis activity of swine neutrophils observed by Giemsa staining was significantly enhanced by opsonizing Fujisawa cells with pig antiserum immunized with HisSpa1.0. Nearly 50% of neutrophils showed phagocytosis when the cells were opsonized, but only 20% showed phagocytosis when the cells were not opsonized. On the other hand, phagocytosis of monocytes was not affected by opsonization of the bacteria; 18 and 29% of monocytes showed phagocytosis.

View larger version (54K): [in this window] [in a new window]   FIG. 4.   Effect of pig serum immunized with HisSpa1.0 on in vitro phagocytosis of monocytes (columns a and b) and neutrophils (columns c and d) of a nonimmunized SPF pig. (A) Percentage of cells showing the indicated phagocytosis score; (B) number of E. rhusiopathiae bacteria surviving in phagocytes. Bacteria were treated with control serum (columns a and c) or immunized serum (columns b and d) collected at the time of challenge.

DNA sequence analysis of diverse spaAs. The DNA sequences of the protective regions of spaAs of four strains of serotypes 1 and 2, Fujisawa, Koganei, SE-9, ATCC 19414^T, and Tama, were almost identical, as shown by alignment of amino acid sequences (Fig. 5).

View larger version (55K): [in this window] [in a new window]   FIG. 5.   Alignment of the 342-amino-acid sequences of protective regions (amino acids 76 to 438) of SpaA proteins of E. rhusiopathiae Fujisawa (serotype 1a), Koganei (1a), SE-9 (2a), ATCC 19414T (2a), and Tama-96 (2). Identical and different amino acids are marked with asterisks and dots, respectively.

View larger version (34K): [in this window] [in a new window]   FIG. 6.   RFLP analysis of PCR products of spaA (bp 16 to 1,871) of Fujisawa (serotype 1a), Koganei (1a), SE-9 (2a), ATCC 19414T (2b), and reference strains of E. rhusiopathiae, representing all 12 serotypes harboring the gene. Products of all strains but two showed the same restriction fragment size as Fujisawa with all enzymes used; products of R32E11 (2a) and Kaparek (9) gave about 300-bp longer and about 100-bp shorter C-terminal fragments, respectively, with all enzymes used. Restriction fragments showing size variation are indicated as checked boxes.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

In this study, we showed that purified fusions of truncated SpaA/Fujisawa, constructed with the N-terminal 342 amino acids (90 to 431) and histidine hexamer, could elicit complete protection in swine against challenge with virulent strains of E. rhusiopathiae of serotypes 1 and 2, and the C-terminal amino acid repeats of SpaA were not necessary for protection. In contrast, Makino et al. (13) emphasized the importance of the C-terminal amino acid repeats for protection, because in their study only recombinant E. coli expressing complete SpaA could elicit protection in mice. Although they created many Exo-Mung deletion derivatives, including two clones harboring complete spaA, they could not show protection with any of them (13). The contradiction in results can be attributed to the experimental method used by Makino et al. They immunized mice by intraperitoneal injection of a large dose of viable recombinant E. coli, and most mice died from endotoxin shock. In the paper they mentioned that their protection assay was very difficult to perform. It appears difficult to obtain reproducible results by their method.

The C-terminal 20-amino-acid repeat region of E. rhusiopathiae SpaA was shown by Makino et al. (13) to be necessary for SpaA to bind tightly to the bacterial surface like other gram-positive bacteria. This region has high sequence homology with the C-terminal amino acid repeats of pneumococcal surface protein A (PspA) (44.9% over 225 amino acids) and Streptococcus pneumoniae secretory IgA binding protein (SpsA) (40.1% over 227 amino acids) (8, 35). On the other hand, the protective region of SpaA is located at the N-terminal region, like that of PspA. In PspA, epitopes eliciting protection in mice were present in the 43-kDa -helical N-terminal half of the native 84-kDa molecule (27) and in amino acids 192 to 260 and 192 to 588 (14, 28). Despite the high diversity of the -helical protective region of PspA (15, 22), a recombinant PspA of one strain can elicit cross-protection against pneumococci of different capsular types and PspA serological types (15, 28). In contrast to PspA, the DNA sequences of protective region of SpaA of E. rhusiopathiae of five strains of serotypes 1 and 2, the most important serotypes in pigs, were almost identical and highly conserved. Also, nucleic acid sequences of spaA of all serotypes of E. rhusiopathiae harboring this gene seemed to be well conserved when examined by PCR-RFLP. Although Makino et al. (13) showed strain differences of EcoRI fragment size in the spaA gene, such as 0.7 or 2 kb, by Southern hybridization, in this study all of these stains gave the same 0.7-kb fragment upon EcoRI digestion by PCR-RFLP. These results indicate that spaA genes are highly conserved among different serotypes of E. rhusiopathiae and this characteristic provides a major mechanism of the cross-protection activity of SpaA against challenge with different serotypes. Although the number of C-terminal amino acid repeats appears to vary among strains, this repeat has no role in protection.

Pigs immunized with recombinant truncated SpaA were completely protected against challenge with virulent strains of E. rhusiopathiae. By in vitro phagocytosis assay, we determined that the major protection mechanism in these pigs seemed to be the enhancement of the activity of neutrophils to phagocytose and kill the bacteria by effective opsonization with antibody produced against SpaA. On the other hand, the bacteria phagocytosed by monocytes of nonimmunized pigs tended to resist killing. This result suggested that activation of monocytes and macrophages may be also necessary for ready clearance of the bacteria. In immunized pigs, macrophages seemed to be also activated readily after challenge.

The antibody response of all pigs immunized with HisSpa1.0 was sensitively detected by ELISA with intact SpaA/Fujisawa and HisSpa1.0 from 2 weeks after the first immunization. In contrast, the conventional growth agglutination test widely used in Japan did not detected the antibody response to HisSpa1.0, although the test is considered useful to assay the protective antibody in pigs (20). These results indicates that the growth agglutination test cannot directly detect the antibody response against SpaA, the protective antigen of E. rhusiopathiae, and can detect antibody responses against other antigens.

By immunoblotting analysis with pig antiserum against HisSpa1.0, it was shown that SpaA, like the 64- to 66-kDa protective protein described by Groschup and Timoney (6), was produced in larger amounts in modified Feist broth than in brain heart infusion, and SpaA was produced consistently in larger amounts by SE-9 than by Fujisawa.

In this study, we found that purified recombinant SpaA/Fujisawa can elicit protective immunity in pigs, the N-terminal 342 amino acids are necessary for protection in pigs, the nucleic acid sequences of this region are highly conserved among strains of serotypes 1 and 2, which are the most important serotypes in pigs, truncated SpaA of serotype 1 can elicit cross-protective immunity against challenge with serotype 2, and the sequences of full-size spaA also seemed to be highly conserved among all serotypes of E. rhusiopathiae harboring this gene. From these results, we conclude that this truncated SpaA may be useful for development of new types of vaccines such as component, vector, and DNA vaccines and of new diagnostic techniques such as ELISA to assay protective antibody of vaccinated pigs and maternal protective antibody of piglets.