Med MicrobiolImmunol(1992) 181:165-172
9 Springer-Verlag 1992
Experimental Yersinia enterocolitica infection in rats: analysis of the immune response to plasmid-encoded antigens of arthritis-susceptible Lewis rats and arthritis-resistent Fischer rats Karoline Gaede l, Dietrich Mack 2, and Jiirgen Heesemann 1 l Institut for Hygieneund Mikrobiologie,Universit/itWiirzburg,Josef-Schneider-Strasse2, W-8700 Wtirzburg, Federal Republic of Germany 2 Institut fiir MedizinischeMikrobiologieund Immunologic,W-2000 Hamburg, Federal Republic of Germany Received May 27, 1992
Abstract. Recently, experimentally Yersinia-induced arthritis in Lewis and SHR rats has been described as a potential animal model for analyzing the pathomechanism of reactive arthritis (Toivanen et al. 1986; Hill and Yu 1987). We could confirm that Lewis but not Fischer rats develop aseptic arthritis about 2 weeks after an intravenous inoculation of Yersinia enterocolitica, serotype 0:8 (Hill and Yu 1987). Moreover, we compared the antibody response to virulence-associated Yersinia antigens (Yops) of experimentally infected Lewis rats with that of Fischer rats using an enzyme-linked immunosorbent assay (ELISA) and immunoblotting. The ELISA results revealed a more rapid and vigorous specific IgG, IgA and IgM response for Lewis rats than for Fischer rats in the early state of infection. As demonstrated by immunoblotting, the IgG response was directed predominantly against the plasmid-encoded YopM, YopH, YopD and the V-antigen during the acute phase followed by antibodies against YopE and YadA. Although both rat strains seroconverted against this set of antigens, IgG antibodies to YadA were more prevalent and of higher titer in arthritis-susceptible rats compared to arthritis-resistent rats.
Introduction The enteropathogenic species Yersiniaenterocolitica and Y.pseudotuberculosis cause intestinal infections in humans and rodents. Furthermore, in 10%-30% of the cases extraintestinal manifestations occur, of which reactive arthritis is the most frequent one (Ahvonen et al. 1969; Arvastson et al. 1971). The pathomechanisms of these sequelae are not understood. It is generally accepted that the pathogen cannot be cultivated from affected joints, and thus, sterile reactive arthritis can be clearly distinguished from septic arthritis. However, there is some evidence from immunohistology and -cytology that antigens of the arthritogenic Correspondence to: J. Heesemann
166 p a t h o g e n are p r e s e n t in the i n f l a m e d j o i n t s ( G r a n f o r s et al. 1989; H a m m e r et al. 1990). M o r e o v e r , high specific s e r u m a n t i b o d y titers a n d p r o l o n g e d p e r s i s t a n c e o f specific I g A - s e r u m a n t i b o d i e s a n d o f Yersinia bacilli in the intestinal m u c o s a are further characteristics o f p a t i e n t s suffering f r o m Yersinia-reactive arthritis ( G r a n fors et al. 1980; D e K o n i n g et al. 1989). Recently, an e x p e r i m e n t a l rat m o d e l o f Yersinia-induced arthritis has been p r o p o s e d as a m o d e l o f h u m a n Yersiniatriggered arthritis ( T o i v a n e n et al. 1986; Hill a n d Yu 1987). A f t e r i.v. injection o f the highly virulent Y. enterocolitica s e r o t y p e 0 : 8 ( W A ) - s t r a i n , Lewis a n d S H R rats, b u t not for e x a m p l e F i s c h e r a n d W i s t a r rats, d e v e l o p e d sterile arthritis o f the h i n d p a w s . A striking o b s e r v a t i o n was t h a t the infecting m i c r o o r g a n i s m c o u l d be cultivated f r o m l y m p a t h i c tissue o f arthritis-susceptible rats several m o n t h s after infection, whereas a r t h r i t i s - r e s i s t a n t rats were able to e l i m i n a t e the c h a l l e n g i n g strain within 3weeks. This a n i m a l m o d e l allows e x a m i n a t i o n o f b o t h the i m m u n o l o g i c a l response o f the h o s t following infection a n d the m i c r o b i a l p a t h o g e n i c c o m p o n e n t s involved in triggering the i n f l a m m a t o r y response. As a first a p p r o a c h we a n a l y z e d the specific serum a n t i b o d y response o f Lewis rats (arthritis susceptible) a n d o f F i s c h e r rats (arthritis resistant) after challenge with Y. enterocolitica 0:8. O u r results suggest a link between p e r s i s t a n c e o f the p a t h o g e n , a n t i b o d y h y p e r r e s p o n s i v e n e s s o f the host a n d arthritis susceptibility.
MateriaLs and methods
Animals Specific pathogen-free (SPF) male Fischer and Lewis rats, weighing between 350 and 450g and female Balb/c mice were purchased from Charles River, Sulzfeld, FRG. Two rats were housed per cage in positive-pressure cabinets and maintained on standard diet and water ad libitum.
Bacterial cultivation, enumeration and infection The virulent plasmid-bearing Y. enterocoliticaWA-strain of serotype 0 : 8 was the kind of gift of D. Yu, (Los Angeles, Calif.; Hill and Yu 1987). The strain was passaged in laboratory mice and checked for the presence of the virulence plasmid (Heesemann et al. 1983). Aliquots from 28~ cultures were stored at -80~ in 30% glycerol. Prior to an experiment a sample was recultivated to the late log-phase at room temperature in luria broth. After washing the cultures three times in sterile PBS the bacterial concentration was estimated and adjusted by spectrophotometry to an optical density of 0.3 at 600 nm [about 108 colony-forming units (cfu)/ml] before infection of rats. The actual number of cfu was determined after plating a serial dilution of Mueller-Hinton-agar plates followed by7an incubation at 28 ~C for 24-48 h. Rats were infected by an i.v. injection into the tail vein of 10 cfu in 1.0 ml PBS after application of an inhalant anesthetic. All experiments were repeated at least four times.
Assessment and severity of arthritis Rats were inspected daily for clinical symptoms during the period of the experiment (usually 1 week before the infection up to the end of the experiment). Every 8th or 10th day, respectively, the experimental animals were bled and the incidence of arthritis was recorded. For estimating arthritis development, the circumference of each back paw was measured with a measuring tape at two fixed points. The mean values of four fixed points from two hind paws was taken as an index of the severity of the swelling from arthritis.
167
Serological techniques For serological analysis whole cells, plasmid-encoded proteins [which are released into the culture medium (Yops)] and the purified plasmid-encoded outer membrane protein (YadA) of Y. enterocolitica 0:8 (WA) were used as Yersinia-specificantigens.
Yop ELISA The ELISA was performed as Yop or whole cell ELISA as described (M~iki-Ikola et al. 1991). Briefly, polystyrene plates (96 wells, Greiner, Niirtingen, FRG) were coated overnight at 4~ either with Yop supernatant or with killed whole cells of the plasmid-less Y. enterocolitica 0: 8 (WA)-strain in carbonate-bicarbonate buffer (pH 9.6, washed with 0.05 % (v/v) Tween/PBS, and saturated with 3% (w/v) BSA/PBS for 1 h at room temperature. Diluted rat sera were incubated for 2 h at 37 ~C. Class-specific rat immunoglobulins were detected by incubation with goat antirat IgG, IgM, and IgA (Cooper Biomedical, West Chester, USA; 1: 800), respectively, followed by an incubation with alkaline phosphatase-conjugated anti-goat immunglobulins (IgG/IgM: 1 : 1000, IgA: 1:4000; 1 h 37~ As substrate solution 0.1% (w/v) p-nitro-phenylphosphatase, 9.7% (v/v) diethanolamine, 0.02% (w/v) sodium azide, titrated to pH9.8 with 1M MgCI2" 6H20" HCI was used. The absorbance was determined at 405 nm (ref. 492 nm) in an ELISA photometer.
Isolation and immunoblotting of plasmid-endoced Yersinia antigens Purification of Yops has been described (Heesemann et al. 1986). Partial digestion of YadA by V8 protease was performed as described recently by Mack et al. (1991). YadA fragments are superior for antibody detection than the intact YadA antigen because of reduction of unspecific binding of immunoglobulin. Yops and YadA fragments were separated using SDS-PAGE as described previously (Heesemann et al. 1984). Western blotting and development of immunoblots have been described (Heesemann et al. 1984, 1986). Briefly, IgG-specific antibodies were detected by a subsequent incubation with rat sera (1:200) and alkaline phosphatase-conjugated anti rat IgG (1 : 500).
Results
Class-specific antibody response of Lewis and Fischer rats to Y. enterocolitica released proteins: Yop-ELISA Lewis and Fischer rats were infected with Y. enterocolitica (WA) and b l o o d was taken at week 2, 3, 4, 8 and 12 post infection (p.i.). Figure 1 shows mean values o f the class-specific a n t i b o d y response o f Lewis and Fischer rats ( n = 5 ) to Y. enterocolitica Yops during the course of infection as analyzed by E L I S A . A m o r e rapid and vigorous specific a n t i b o d y response in Lewis rats (arthritis susceptible) than in Fischer rats (arthritis resistant) during the early phase o f infection was evident for all three classes o f immunoglobulins. F u r t h e r m o r e , Lewis rats showed a significantly higher a n t i b o d y titer with respect to I g M and I g G subclasses. D u r i n g arthritis development ( 1 0 - 1 4 d a y s p.i.) a n t i b o d y titers (IgM, IgG, IgA) reached a m a x i m u m . In contrast, the a n t i b o d y response o f Fischer rats is delayed, reaching maximal titers at week 5p.i. (IgM) and week 8p.i. (IgG, IgA), respectively. Twelve weeks p.i. the a n t i b o d y titers of both rat strains declined to c o m p a r a b l e values.
168
weeks after infection
Fig. 1. Class-specificantibody response of Yersiniaenterocolitica-infected Lewis and Fischer rats to Yops (ELISA): sequential sera of five rats have been tested. The dots represent the mean values of one representative experiment. The bars represent the standard errors of the means
Immunoblot analysis of the IgG antibody response to plasmid-encoded Y. enterocolitica released proteins The plasmid-encoded Yops are immune-dominant antigens with known pathogenic functions (particularly YopM, YopH and YopE; Heesemann et al. 1987, 1988). Therefore, we were interested in the detailed analysis of the antibody response against the different Yops. IgG specificity of sequential sera was analyzed by immunoblotting. Figure2 shows the reactivity of sera obtained from Y. enterocolitica-infected Lewis and Fischer rats to Y o p M / H / D / E and V antigen. The diagram in the upper part of Fig. 2 represents the percental reactivity of the tested sera with Y o p M / H / D / E and V antigen during the course of infection. The lower part shows representative immunoblot patterns. The immunoblot results parallel the ELISA data described in the preceding section; the delayed IgG response of Fischer rats to Yop M / H / D / E and V antigen is obvious. However, by immunoblotting we observed that Fischer rats (4th week p.i.: 100%) responded much faster and more frequently against YopM compared to Lewis rats (4th week p.i.: 0%). Moreover, the delayed reactivity of both rat strains to YopE is remarkable.
Immunoblot analysis of the IgG response to Y. enterocolitica YadA (V8) fragments The hydrophobic outer membrane protein YadA (50kDa as monomer) is known to be involved in collagen binding and serum resistance. For specific immunoblotting V8-digested YadA was used as antigen. As shown in Fig. 3 again a delayed IgG antibody response is observed in Fischer rats (2weeks later) compared to Lewis rats. In contrast to the reactivity pattern against Yops no difference could be recognized in the reactivity pattern against YadA fragments in week 4 p.i.
169
Fig. 2. Immunoblot analysis of the IgG antibody response to plasmid-encoded Yops: Lewis (L) and Fischer (F) rats were inoculated intravenously with Y. enterocolitica and observed for 12 weeks. Sequential sera of five rats have been tested. The representative diagram in the upperpart of this figure shows the reactivity of the sera to YopM/H/D/E and V antigen (denoted as 1, 2, 4, 5 and 3, respectively) expressed in percent of tested sera. The lower part shows representative immunoblot patterns
170
Fig. 3. Immunoblot analysis of the IgG response to Y. enterocolitica YadA fragments obtained by partial V8-protease digest: Lewis (L) and Fischer (F) rats were inoculated i.v. with Y. enterocolitica and observed for 12 weeks. Sequential sera of five rats have been tested. The representative diagram in the rightpart of this figure shows the reactivity of the sera to the YadA fragments 08.1-08.5 expressed in percent of tested sera, as well as representative immunoblot patterns. The outer membrane protein pattern with marked YadA in a Comassie blue-stained SDS-PAGE (OMP) and the V8-digested YadA in a silver-stained SDS-PAGE (V8-YadA) are shown on the left side
Y. enterocolitica 0:8 whole cell E L I S A The whole cell E L I S A was used to detect LPS-specific antibodies. The dynamics of the O-antigen-specific I g G response resembled those against Yops. The a n t i b o d y response of Fischer rats was significantly lower than that of Lewis rats (data not shown).
Discussion We could confirm the results o f Hill and Yu (1987) that: (i) Lewis rats but not Fischer rats develop arthritis after i.v. challenge with the highly pathogenic strain of Y. enterocolitica of serotype 0:8, and (ii) susceptibility to arthritis is closely related to susceptibility to infection and persistence o f the p a t h o g e n in lympathic tissues. There is no published information on the h u m o r a l i m m u n e response in the rat model. Therefore, we c o m p a r e d the serum a n t i b o d y response of Lewis and Fischer rats against the plasmid-encoded Yersinia outer proteins (released proteins, Yops) and the outer m e m b r a n e protein Y a d A (Yersinia adhesin). Recently, we have shown that Yops are strong antigens, eliciting a specific IgM-, I g G - and I g A - a n t i b o d y response in orally infected rabbits (Heesemann et al. 1988) and in h u m a n s suffering from Yersinia infection (Heesemann et al. 1987). Here we
171 demonstrate a class-specific antibody response in experimentally infected rats. Surprisingly, the arthritis-susceptible Lewis rats developed a much more vigorous IgM and IgG response against Yops than the arthritis-resistant Fischer rats. A maximal difference in antibody production (IgM and IgG) could be detected during the phase of arthritis, while IgM and IgG titers reach their maxima in Lewis rats. At that time arthritis-resistant Fischer rats have already overcome their maximum on Yersinia bacilli in spleen and liver tissue, this had been reported by Hill and Yu (1987) and could be confirmed in our laboratory (data not shown). The reason for this weaker and delayed antibody production in Fischer rats compared to the susceptible Lewis rats could be due to a more effective capacity to eliminate the injected pathogen. On the other hand, the strong immune response to Yersinia Yops in Lewis rats correlates with the persistence of Yersinia bacilli in lymphatic tissues of infected animals. These findings are in accordance with data obtained from investigations on Yersiniosis patients (de Koning et al. 1989). As mentioned above we could not detect a persistence of IgA antibodies directed against Yops as known from Yersiniosis patients suffering from reactive arthritis. As the rats are challenged parenteraly, involvement of the gut-associated lympathic tissue is less likely. Moreover, the antibody response against surface antigens (particularly lipopolysaccharides) as measured by the whole cell ELISA was comparable to that against Yop antigens. The plasmid-encoded released proteins of Yersinia mediate specific pathogenic functions: the 25-kDa protein YopE is cytotoxic (Rosqvist et al. 1991), the 48 kDa protein YopH mediates protein tyrosine phosphatase activity (Guan et al. 1990), the polymorphic 47- to 57- kDa protein YopM inhibits platelet aggregation (Leung et al. 1990). Since ELISA analysis does not provide information on the spectrum of antibody specificities against the different Yops, we used an immunoblotting technique. The results of the IgG immunoblotting confirm those obtained by ELISA. However, a more detailed analysis reveals that in spite of the enhanced antibody response, Lewis rats do not respond against YopM within the first 4 weeks after challenge, in contrast to Fischer rats. It remains to be clarified whether antibodies against YopM are important for restriction of microbial growth in the host. Moreover, we analyzed the antibody response against the plasmid-encoded outer membrane protein YadA, which mediates cell adherence, serum resistance, collagen binding and resistance to phagocytosis (Pilz 1992; Em6dy et al. 1989). We focussed our interest on this outer membrane protein because of its collagen binding capacity, which could be important for the arthritogenicity of a Y. enterocolitica strain. For immunoblotting we used V8 protease-digested YadA, resulting in five to six polypeptide fragments of different size (Mack et al. 1991). In contrast to the dynamics of the anti-YopM response, Lewis rats responded earlier against YadA fragments than Fischer rats. Four weeks after infection, however, both rat strains exhibited a comparable antibodyreactivity pattern against YadA fragments. These results do not exclude differences of antibody reactivity against native YadA between Lewis and Fischer rats. A striking result was that the antibody titers of rats developing macroscopic arthritis did not differ significantly from those of arthritis-negative Lewis rats. However, these results do not exclude the possibility that the formation of immune complexes, which may be favored by a vigorous antibody response, plays an important role in the pathogenesis of Yersinia-induced arthritis. The role of immune complexes as well as the cellular immune response remains to be clarified and is subject of our current investigations.
172 Acknowledgement. We are grateful to U. Vogel for establishing the whole cell ELISA. This work was supported by a grant from the Bundesministerium fur Forschung und Technologie (BMFT 01-8607-44).
References de Koning J, Heesemann J, Hoogkamp-Korstanje 3AA, Festen JJM, Houtman PLM, van Oijen (1989) Yersinia in intestinal biopsy specimens from patients with seronegative spondyloarthropathy: correlation with specific serum IgA antibodies. J Infect Dis 159:109-112 Em6dy L, Heesemann J, Wolf-Watz H, Skurnik M, Kapperud G, O'Toole P, Wadstr6m T (1989) Binding to collagen by Yersiniaenterocolitica and Yersiniapseudotuberculosis: evidence for yopA-mediated and chromosomally encoded mechanisms. J Bacteriol 171:6674-6679 Granfors K, Viljanen M, Tiilikainen A, Toivanen A (1980) Persistence of IgM, IgG and IgA antibodies to Yersinia in Yersinia arthritis. J Infect Dis 141:424-429 Granfors K, Jalkanen S, von Essen R, Lahesmaa-Rantala R, Isom/iki O, Pekkola-Heino K, Merilahti-Palo R, Saario R, Isom/~ki H, Toivanen A (1989) Yersinia antigens in synovial-fluid cells from patients with reactive arthritis. N Engl J Med 320:216-221 Guan KL, Dixon JE (1990) Protein tyrosine phosphatase activity of an essential virulence determinant in Yersinia. Science 249:553-556 Hammer M, Zeidler H, Klimsa S, Heesemann J (1990) Yersinia enterocolitica in the synovial membrane of patients with Yersinia-induced arthritis. Arthritis Rheum 33:1795-1800 Hessemann J, Gaede K (1989) Mechanisms involved in the pathogenesis of Yersinia infections. Rheumatol Int 9 :213-217 Heesemann J, Keller C, Morawa R, Schmidt N, Siemens H J, Laufs R (1983) Plasmids of human strains of Yersiniaenterocolitica: a molecular relatedness and possible importance for pathogenesis. J Infect Dis 147:107-115 Heesemann J, Algermissen B, Laufs R (1984) Genetically manipulated virulence of Yersinia enterocolitica. Infect Immun 46:105-110 Heesemann J, Gross U, Schmidt N, Laufs R (1986) Immunochemical analysis of plasmidencoded proteins released by enteropathogenic Yersinia species grown in calcium deficient media. Infect Immun 54:561-567 Heesemann J, Kalthoff H, Koch F (1986) Monoclonal antibodies directed against plasmidencoded released proteins of entero-pathogenic Yersinia. FEMS Microbiol Lett 36:15-19 Heesemann J, Eggers C, Schr6der J (1987) Serological diagnosis of yersiniosis by immunoblot technique using virulene-associated antigen of enteropathogenic Yersinia. Contrib Microbiol Immunol 9:285-289 Heesemann J, Schr6der J, Ulrich M (1988) Analysis of the class-specific immune response to Yersinia enterocolitica virulence-associated antigens in oro-gastrically infected rabbits. Microb Pathog 5:437-447 Hill JL, Yu DTY (1987) Development of an experimental animal model for reactive arthritis induced by Yersinia enterocolitica infection. Infect Immun 55:721-726 Leung KY, Reisner BS, Strayley SC (1990) YopM inhibits platelet aggregation and is necessary for virulence of Yersiniapestis in mice. Infect Immun 58:3262-3271 Mack D, Pulz M, Heesemann J (1991) Recognition by three different structural groups of outer membrane protein YOP1 of Yersiniaenterocolitica and Yersiniapseudotuberculosis. Med Microbiol Immunol 180:205-211 M/iki-Ikola O, Heesemann J, Lahesmaa-Rantala R, Toivanen A, Granfors K (1991) Combined use of released proteins and lipopolysaccharide in enzyme-linked immunosorbent assay for serological screening of Yersinia infections. J Infect Dis 163:409-412 Pilz D, Vocke T, Heesemann J, Brade V (1992) Mechanism of YadA-mediated serum resistance of Yersinia enterocolitica serotype 0: 3. Infect Immun 60:189-195 Rosqvist R, Forsberg A, Wolf-Watz H (1991) Intracellular targeting of Yersinia YopE cytotoxin in mammalian cells induces actin microfilament disruption. Infect Immun 59:4562-4569 Toivanen A, Merilahti-Palo R, Gripenberg C, Lahesmaa-Rantala R, S6derstr6m KO, Jaakkola U-M (1986) Yersinia-associated arthritis in the rat: experimental model for human reactive arthritis? Acta Pathol Microbiol Immunol Scand [C] 94:261-269
9 Springer-Verlag 1992
Experimental Yersinia enterocolitica infection in rats: analysis of the immune response to plasmid-encoded antigens of arthritis-susceptible Lewis rats and arthritis-resistent Fischer rats Karoline Gaede l, Dietrich Mack 2, and Jiirgen Heesemann 1 l Institut for Hygieneund Mikrobiologie,Universit/itWiirzburg,Josef-Schneider-Strasse2, W-8700 Wtirzburg, Federal Republic of Germany 2 Institut fiir MedizinischeMikrobiologieund Immunologic,W-2000 Hamburg, Federal Republic of Germany Received May 27, 1992
Abstract. Recently, experimentally Yersinia-induced arthritis in Lewis and SHR rats has been described as a potential animal model for analyzing the pathomechanism of reactive arthritis (Toivanen et al. 1986; Hill and Yu 1987). We could confirm that Lewis but not Fischer rats develop aseptic arthritis about 2 weeks after an intravenous inoculation of Yersinia enterocolitica, serotype 0:8 (Hill and Yu 1987). Moreover, we compared the antibody response to virulence-associated Yersinia antigens (Yops) of experimentally infected Lewis rats with that of Fischer rats using an enzyme-linked immunosorbent assay (ELISA) and immunoblotting. The ELISA results revealed a more rapid and vigorous specific IgG, IgA and IgM response for Lewis rats than for Fischer rats in the early state of infection. As demonstrated by immunoblotting, the IgG response was directed predominantly against the plasmid-encoded YopM, YopH, YopD and the V-antigen during the acute phase followed by antibodies against YopE and YadA. Although both rat strains seroconverted against this set of antigens, IgG antibodies to YadA were more prevalent and of higher titer in arthritis-susceptible rats compared to arthritis-resistent rats.
Introduction The enteropathogenic species Yersiniaenterocolitica and Y.pseudotuberculosis cause intestinal infections in humans and rodents. Furthermore, in 10%-30% of the cases extraintestinal manifestations occur, of which reactive arthritis is the most frequent one (Ahvonen et al. 1969; Arvastson et al. 1971). The pathomechanisms of these sequelae are not understood. It is generally accepted that the pathogen cannot be cultivated from affected joints, and thus, sterile reactive arthritis can be clearly distinguished from septic arthritis. However, there is some evidence from immunohistology and -cytology that antigens of the arthritogenic Correspondence to: J. Heesemann
166 p a t h o g e n are p r e s e n t in the i n f l a m e d j o i n t s ( G r a n f o r s et al. 1989; H a m m e r et al. 1990). M o r e o v e r , high specific s e r u m a n t i b o d y titers a n d p r o l o n g e d p e r s i s t a n c e o f specific I g A - s e r u m a n t i b o d i e s a n d o f Yersinia bacilli in the intestinal m u c o s a are further characteristics o f p a t i e n t s suffering f r o m Yersinia-reactive arthritis ( G r a n fors et al. 1980; D e K o n i n g et al. 1989). Recently, an e x p e r i m e n t a l rat m o d e l o f Yersinia-induced arthritis has been p r o p o s e d as a m o d e l o f h u m a n Yersiniatriggered arthritis ( T o i v a n e n et al. 1986; Hill a n d Yu 1987). A f t e r i.v. injection o f the highly virulent Y. enterocolitica s e r o t y p e 0 : 8 ( W A ) - s t r a i n , Lewis a n d S H R rats, b u t not for e x a m p l e F i s c h e r a n d W i s t a r rats, d e v e l o p e d sterile arthritis o f the h i n d p a w s . A striking o b s e r v a t i o n was t h a t the infecting m i c r o o r g a n i s m c o u l d be cultivated f r o m l y m p a t h i c tissue o f arthritis-susceptible rats several m o n t h s after infection, whereas a r t h r i t i s - r e s i s t a n t rats were able to e l i m i n a t e the c h a l l e n g i n g strain within 3weeks. This a n i m a l m o d e l allows e x a m i n a t i o n o f b o t h the i m m u n o l o g i c a l response o f the h o s t following infection a n d the m i c r o b i a l p a t h o g e n i c c o m p o n e n t s involved in triggering the i n f l a m m a t o r y response. As a first a p p r o a c h we a n a l y z e d the specific serum a n t i b o d y response o f Lewis rats (arthritis susceptible) a n d o f F i s c h e r rats (arthritis resistant) after challenge with Y. enterocolitica 0:8. O u r results suggest a link between p e r s i s t a n c e o f the p a t h o g e n , a n t i b o d y h y p e r r e s p o n s i v e n e s s o f the host a n d arthritis susceptibility.
MateriaLs and methods
Animals Specific pathogen-free (SPF) male Fischer and Lewis rats, weighing between 350 and 450g and female Balb/c mice were purchased from Charles River, Sulzfeld, FRG. Two rats were housed per cage in positive-pressure cabinets and maintained on standard diet and water ad libitum.
Bacterial cultivation, enumeration and infection The virulent plasmid-bearing Y. enterocoliticaWA-strain of serotype 0 : 8 was the kind of gift of D. Yu, (Los Angeles, Calif.; Hill and Yu 1987). The strain was passaged in laboratory mice and checked for the presence of the virulence plasmid (Heesemann et al. 1983). Aliquots from 28~ cultures were stored at -80~ in 30% glycerol. Prior to an experiment a sample was recultivated to the late log-phase at room temperature in luria broth. After washing the cultures three times in sterile PBS the bacterial concentration was estimated and adjusted by spectrophotometry to an optical density of 0.3 at 600 nm [about 108 colony-forming units (cfu)/ml] before infection of rats. The actual number of cfu was determined after plating a serial dilution of Mueller-Hinton-agar plates followed by7an incubation at 28 ~C for 24-48 h. Rats were infected by an i.v. injection into the tail vein of 10 cfu in 1.0 ml PBS after application of an inhalant anesthetic. All experiments were repeated at least four times.
Assessment and severity of arthritis Rats were inspected daily for clinical symptoms during the period of the experiment (usually 1 week before the infection up to the end of the experiment). Every 8th or 10th day, respectively, the experimental animals were bled and the incidence of arthritis was recorded. For estimating arthritis development, the circumference of each back paw was measured with a measuring tape at two fixed points. The mean values of four fixed points from two hind paws was taken as an index of the severity of the swelling from arthritis.
167
Serological techniques For serological analysis whole cells, plasmid-encoded proteins [which are released into the culture medium (Yops)] and the purified plasmid-encoded outer membrane protein (YadA) of Y. enterocolitica 0:8 (WA) were used as Yersinia-specificantigens.
Yop ELISA The ELISA was performed as Yop or whole cell ELISA as described (M~iki-Ikola et al. 1991). Briefly, polystyrene plates (96 wells, Greiner, Niirtingen, FRG) were coated overnight at 4~ either with Yop supernatant or with killed whole cells of the plasmid-less Y. enterocolitica 0: 8 (WA)-strain in carbonate-bicarbonate buffer (pH 9.6, washed with 0.05 % (v/v) Tween/PBS, and saturated with 3% (w/v) BSA/PBS for 1 h at room temperature. Diluted rat sera were incubated for 2 h at 37 ~C. Class-specific rat immunoglobulins were detected by incubation with goat antirat IgG, IgM, and IgA (Cooper Biomedical, West Chester, USA; 1: 800), respectively, followed by an incubation with alkaline phosphatase-conjugated anti-goat immunglobulins (IgG/IgM: 1 : 1000, IgA: 1:4000; 1 h 37~ As substrate solution 0.1% (w/v) p-nitro-phenylphosphatase, 9.7% (v/v) diethanolamine, 0.02% (w/v) sodium azide, titrated to pH9.8 with 1M MgCI2" 6H20" HCI was used. The absorbance was determined at 405 nm (ref. 492 nm) in an ELISA photometer.
Isolation and immunoblotting of plasmid-endoced Yersinia antigens Purification of Yops has been described (Heesemann et al. 1986). Partial digestion of YadA by V8 protease was performed as described recently by Mack et al. (1991). YadA fragments are superior for antibody detection than the intact YadA antigen because of reduction of unspecific binding of immunoglobulin. Yops and YadA fragments were separated using SDS-PAGE as described previously (Heesemann et al. 1984). Western blotting and development of immunoblots have been described (Heesemann et al. 1984, 1986). Briefly, IgG-specific antibodies were detected by a subsequent incubation with rat sera (1:200) and alkaline phosphatase-conjugated anti rat IgG (1 : 500).
Results
Class-specific antibody response of Lewis and Fischer rats to Y. enterocolitica released proteins: Yop-ELISA Lewis and Fischer rats were infected with Y. enterocolitica (WA) and b l o o d was taken at week 2, 3, 4, 8 and 12 post infection (p.i.). Figure 1 shows mean values o f the class-specific a n t i b o d y response o f Lewis and Fischer rats ( n = 5 ) to Y. enterocolitica Yops during the course of infection as analyzed by E L I S A . A m o r e rapid and vigorous specific a n t i b o d y response in Lewis rats (arthritis susceptible) than in Fischer rats (arthritis resistant) during the early phase o f infection was evident for all three classes o f immunoglobulins. F u r t h e r m o r e , Lewis rats showed a significantly higher a n t i b o d y titer with respect to I g M and I g G subclasses. D u r i n g arthritis development ( 1 0 - 1 4 d a y s p.i.) a n t i b o d y titers (IgM, IgG, IgA) reached a m a x i m u m . In contrast, the a n t i b o d y response o f Fischer rats is delayed, reaching maximal titers at week 5p.i. (IgM) and week 8p.i. (IgG, IgA), respectively. Twelve weeks p.i. the a n t i b o d y titers of both rat strains declined to c o m p a r a b l e values.
168
weeks after infection
Fig. 1. Class-specificantibody response of Yersiniaenterocolitica-infected Lewis and Fischer rats to Yops (ELISA): sequential sera of five rats have been tested. The dots represent the mean values of one representative experiment. The bars represent the standard errors of the means
Immunoblot analysis of the IgG antibody response to plasmid-encoded Y. enterocolitica released proteins The plasmid-encoded Yops are immune-dominant antigens with known pathogenic functions (particularly YopM, YopH and YopE; Heesemann et al. 1987, 1988). Therefore, we were interested in the detailed analysis of the antibody response against the different Yops. IgG specificity of sequential sera was analyzed by immunoblotting. Figure2 shows the reactivity of sera obtained from Y. enterocolitica-infected Lewis and Fischer rats to Y o p M / H / D / E and V antigen. The diagram in the upper part of Fig. 2 represents the percental reactivity of the tested sera with Y o p M / H / D / E and V antigen during the course of infection. The lower part shows representative immunoblot patterns. The immunoblot results parallel the ELISA data described in the preceding section; the delayed IgG response of Fischer rats to Yop M / H / D / E and V antigen is obvious. However, by immunoblotting we observed that Fischer rats (4th week p.i.: 100%) responded much faster and more frequently against YopM compared to Lewis rats (4th week p.i.: 0%). Moreover, the delayed reactivity of both rat strains to YopE is remarkable.
Immunoblot analysis of the IgG response to Y. enterocolitica YadA (V8) fragments The hydrophobic outer membrane protein YadA (50kDa as monomer) is known to be involved in collagen binding and serum resistance. For specific immunoblotting V8-digested YadA was used as antigen. As shown in Fig. 3 again a delayed IgG antibody response is observed in Fischer rats (2weeks later) compared to Lewis rats. In contrast to the reactivity pattern against Yops no difference could be recognized in the reactivity pattern against YadA fragments in week 4 p.i.
169
Fig. 2. Immunoblot analysis of the IgG antibody response to plasmid-encoded Yops: Lewis (L) and Fischer (F) rats were inoculated intravenously with Y. enterocolitica and observed for 12 weeks. Sequential sera of five rats have been tested. The representative diagram in the upperpart of this figure shows the reactivity of the sera to YopM/H/D/E and V antigen (denoted as 1, 2, 4, 5 and 3, respectively) expressed in percent of tested sera. The lower part shows representative immunoblot patterns
170
Fig. 3. Immunoblot analysis of the IgG response to Y. enterocolitica YadA fragments obtained by partial V8-protease digest: Lewis (L) and Fischer (F) rats were inoculated i.v. with Y. enterocolitica and observed for 12 weeks. Sequential sera of five rats have been tested. The representative diagram in the rightpart of this figure shows the reactivity of the sera to the YadA fragments 08.1-08.5 expressed in percent of tested sera, as well as representative immunoblot patterns. The outer membrane protein pattern with marked YadA in a Comassie blue-stained SDS-PAGE (OMP) and the V8-digested YadA in a silver-stained SDS-PAGE (V8-YadA) are shown on the left side
Y. enterocolitica 0:8 whole cell E L I S A The whole cell E L I S A was used to detect LPS-specific antibodies. The dynamics of the O-antigen-specific I g G response resembled those against Yops. The a n t i b o d y response of Fischer rats was significantly lower than that of Lewis rats (data not shown).
Discussion We could confirm the results o f Hill and Yu (1987) that: (i) Lewis rats but not Fischer rats develop arthritis after i.v. challenge with the highly pathogenic strain of Y. enterocolitica of serotype 0:8, and (ii) susceptibility to arthritis is closely related to susceptibility to infection and persistence o f the p a t h o g e n in lympathic tissues. There is no published information on the h u m o r a l i m m u n e response in the rat model. Therefore, we c o m p a r e d the serum a n t i b o d y response of Lewis and Fischer rats against the plasmid-encoded Yersinia outer proteins (released proteins, Yops) and the outer m e m b r a n e protein Y a d A (Yersinia adhesin). Recently, we have shown that Yops are strong antigens, eliciting a specific IgM-, I g G - and I g A - a n t i b o d y response in orally infected rabbits (Heesemann et al. 1988) and in h u m a n s suffering from Yersinia infection (Heesemann et al. 1987). Here we
171 demonstrate a class-specific antibody response in experimentally infected rats. Surprisingly, the arthritis-susceptible Lewis rats developed a much more vigorous IgM and IgG response against Yops than the arthritis-resistant Fischer rats. A maximal difference in antibody production (IgM and IgG) could be detected during the phase of arthritis, while IgM and IgG titers reach their maxima in Lewis rats. At that time arthritis-resistant Fischer rats have already overcome their maximum on Yersinia bacilli in spleen and liver tissue, this had been reported by Hill and Yu (1987) and could be confirmed in our laboratory (data not shown). The reason for this weaker and delayed antibody production in Fischer rats compared to the susceptible Lewis rats could be due to a more effective capacity to eliminate the injected pathogen. On the other hand, the strong immune response to Yersinia Yops in Lewis rats correlates with the persistence of Yersinia bacilli in lymphatic tissues of infected animals. These findings are in accordance with data obtained from investigations on Yersiniosis patients (de Koning et al. 1989). As mentioned above we could not detect a persistence of IgA antibodies directed against Yops as known from Yersiniosis patients suffering from reactive arthritis. As the rats are challenged parenteraly, involvement of the gut-associated lympathic tissue is less likely. Moreover, the antibody response against surface antigens (particularly lipopolysaccharides) as measured by the whole cell ELISA was comparable to that against Yop antigens. The plasmid-encoded released proteins of Yersinia mediate specific pathogenic functions: the 25-kDa protein YopE is cytotoxic (Rosqvist et al. 1991), the 48 kDa protein YopH mediates protein tyrosine phosphatase activity (Guan et al. 1990), the polymorphic 47- to 57- kDa protein YopM inhibits platelet aggregation (Leung et al. 1990). Since ELISA analysis does not provide information on the spectrum of antibody specificities against the different Yops, we used an immunoblotting technique. The results of the IgG immunoblotting confirm those obtained by ELISA. However, a more detailed analysis reveals that in spite of the enhanced antibody response, Lewis rats do not respond against YopM within the first 4 weeks after challenge, in contrast to Fischer rats. It remains to be clarified whether antibodies against YopM are important for restriction of microbial growth in the host. Moreover, we analyzed the antibody response against the plasmid-encoded outer membrane protein YadA, which mediates cell adherence, serum resistance, collagen binding and resistance to phagocytosis (Pilz 1992; Em6dy et al. 1989). We focussed our interest on this outer membrane protein because of its collagen binding capacity, which could be important for the arthritogenicity of a Y. enterocolitica strain. For immunoblotting we used V8 protease-digested YadA, resulting in five to six polypeptide fragments of different size (Mack et al. 1991). In contrast to the dynamics of the anti-YopM response, Lewis rats responded earlier against YadA fragments than Fischer rats. Four weeks after infection, however, both rat strains exhibited a comparable antibodyreactivity pattern against YadA fragments. These results do not exclude differences of antibody reactivity against native YadA between Lewis and Fischer rats. A striking result was that the antibody titers of rats developing macroscopic arthritis did not differ significantly from those of arthritis-negative Lewis rats. However, these results do not exclude the possibility that the formation of immune complexes, which may be favored by a vigorous antibody response, plays an important role in the pathogenesis of Yersinia-induced arthritis. The role of immune complexes as well as the cellular immune response remains to be clarified and is subject of our current investigations.
172 Acknowledgement. We are grateful to U. Vogel for establishing the whole cell ELISA. This work was supported by a grant from the Bundesministerium fur Forschung und Technologie (BMFT 01-8607-44).
References de Koning J, Heesemann J, Hoogkamp-Korstanje 3AA, Festen JJM, Houtman PLM, van Oijen (1989) Yersinia in intestinal biopsy specimens from patients with seronegative spondyloarthropathy: correlation with specific serum IgA antibodies. J Infect Dis 159:109-112 Em6dy L, Heesemann J, Wolf-Watz H, Skurnik M, Kapperud G, O'Toole P, Wadstr6m T (1989) Binding to collagen by Yersiniaenterocolitica and Yersiniapseudotuberculosis: evidence for yopA-mediated and chromosomally encoded mechanisms. J Bacteriol 171:6674-6679 Granfors K, Viljanen M, Tiilikainen A, Toivanen A (1980) Persistence of IgM, IgG and IgA antibodies to Yersinia in Yersinia arthritis. J Infect Dis 141:424-429 Granfors K, Jalkanen S, von Essen R, Lahesmaa-Rantala R, Isom/iki O, Pekkola-Heino K, Merilahti-Palo R, Saario R, Isom/~ki H, Toivanen A (1989) Yersinia antigens in synovial-fluid cells from patients with reactive arthritis. N Engl J Med 320:216-221 Guan KL, Dixon JE (1990) Protein tyrosine phosphatase activity of an essential virulence determinant in Yersinia. Science 249:553-556 Hammer M, Zeidler H, Klimsa S, Heesemann J (1990) Yersinia enterocolitica in the synovial membrane of patients with Yersinia-induced arthritis. Arthritis Rheum 33:1795-1800 Hessemann J, Gaede K (1989) Mechanisms involved in the pathogenesis of Yersinia infections. Rheumatol Int 9 :213-217 Heesemann J, Keller C, Morawa R, Schmidt N, Siemens H J, Laufs R (1983) Plasmids of human strains of Yersiniaenterocolitica: a molecular relatedness and possible importance for pathogenesis. J Infect Dis 147:107-115 Heesemann J, Algermissen B, Laufs R (1984) Genetically manipulated virulence of Yersinia enterocolitica. Infect Immun 46:105-110 Heesemann J, Gross U, Schmidt N, Laufs R (1986) Immunochemical analysis of plasmidencoded proteins released by enteropathogenic Yersinia species grown in calcium deficient media. Infect Immun 54:561-567 Heesemann J, Kalthoff H, Koch F (1986) Monoclonal antibodies directed against plasmidencoded released proteins of entero-pathogenic Yersinia. FEMS Microbiol Lett 36:15-19 Heesemann J, Eggers C, Schr6der J (1987) Serological diagnosis of yersiniosis by immunoblot technique using virulene-associated antigen of enteropathogenic Yersinia. Contrib Microbiol Immunol 9:285-289 Heesemann J, Schr6der J, Ulrich M (1988) Analysis of the class-specific immune response to Yersinia enterocolitica virulence-associated antigens in oro-gastrically infected rabbits. Microb Pathog 5:437-447 Hill JL, Yu DTY (1987) Development of an experimental animal model for reactive arthritis induced by Yersinia enterocolitica infection. Infect Immun 55:721-726 Leung KY, Reisner BS, Strayley SC (1990) YopM inhibits platelet aggregation and is necessary for virulence of Yersiniapestis in mice. Infect Immun 58:3262-3271 Mack D, Pulz M, Heesemann J (1991) Recognition by three different structural groups of outer membrane protein YOP1 of Yersiniaenterocolitica and Yersiniapseudotuberculosis. Med Microbiol Immunol 180:205-211 M/iki-Ikola O, Heesemann J, Lahesmaa-Rantala R, Toivanen A, Granfors K (1991) Combined use of released proteins and lipopolysaccharide in enzyme-linked immunosorbent assay for serological screening of Yersinia infections. J Infect Dis 163:409-412 Pilz D, Vocke T, Heesemann J, Brade V (1992) Mechanism of YadA-mediated serum resistance of Yersinia enterocolitica serotype 0: 3. Infect Immun 60:189-195 Rosqvist R, Forsberg A, Wolf-Watz H (1991) Intracellular targeting of Yersinia YopE cytotoxin in mammalian cells induces actin microfilament disruption. Infect Immun 59:4562-4569 Toivanen A, Merilahti-Palo R, Gripenberg C, Lahesmaa-Rantala R, S6derstr6m KO, Jaakkola U-M (1986) Yersinia-associated arthritis in the rat: experimental model for human reactive arthritis? Acta Pathol Microbiol Immunol Scand [C] 94:261-269
