Early cytokine profiles in the joint define pathogen clearance and severity of arthritis in Chlamydia-induced arthritis in rats.
код для вставкиСкачатьARTHRITIS & RHEUMATISM Vol. 54, No. 2, February 2006, pp 499–507 DOI 10.1002/art.21643 © 2006, American College of Rheumatology Early Cytokine Profiles in the Joint Define Pathogen Clearance and Severity of Arthritis in Chlamydia-Induced Arthritis in Rats Robert D. Inman and Basil Chiu both synovial tissues and spleen compared with Lewis rats. Local cytokine profiles demonstrated that host resistance was characterized by enhanced synovial expression of tumor necrosis factor ␣, interferon-␥ (IFN␥), and interleukin-4. Conclusion. These studies demonstrated that cytokines thought to be proinflammatory in nature can play an important role in host defense in infectiontriggered arthritis and serve to highlight the dynamic cytokine relationships that constitute effective host– pathogen interactions. Objective. Although Chlamydia trachomatis– induced arthritis is among the most common rheumatic diseases having an identified infectious trigger, the pathogenesis of this arthritis is not well defined. We sought to investigate the host–microbe interactions that contribute to the severity of arthritis initiated by chlamydial infection. Methods. We established an experimental rat model of C trachomatis–induced arthritis that recapitulates many pathologic features of the clinical disease. The severity of the arthritis was defined using an established histopathologic scoring system. Host clearance of the pathogen and local cytokine production were examined by enzyme-linked immunosorbent assays. Results. Lewis rats were susceptible to C trachomatis–induced arthritis, whereas BN rats were relatively resistant to this disease. Significant differences in the histopathologic severity of arthritis were originally observed on day 21, and this prompted an examination of the acute phase of the arthritis. As early as day 5 after the onset of the arthritis, pathologic changes in Lewis rats were more severe than those in BN rats. An evaluation of the role of complement using cobra venom factor treatment excluded complement as being the key to differential sensitivity, because decomplementation did not eliminate the differences in arthritis severity between Lewis and BN rats. Host clearance, in contrast, was significantly different between the rat strains, with BN rats showing more prompt and effective clearance of the pathogen from Infection as a triggering factor in chronic arthritis continues to be an important concept in studies into the pathogenesis of rheumatic diseases, and reactive arthritis (ReA) remains the most well-established example of this interaction (1). In the clinical spectrum of ReA, antecedent infection with Chlamydia trachomatis is the most common association (2). Natural history studies have emphasized the potential for chronicity in ReA, and a previous study documented that chronic symptoms persist 5 years after the onset of ReA in the majority of patients (3). However, the host factors that dictate postinfectious sequelae such as ReA have not been defined and have proved difficult to resolve in the clinical setting. We previously developed an experimental model for C trachomatis–induced arthritis (4). In the chronic phase of this model, the synovial tissues are culturenegative, and the cellular infiltrate in the joint consists of a mononuclear lymphocytic population. From the aspects of histopathology, clinical course, and immune response, this experimental arthritis recapitulates the events seen in clinical Chlamydia-induced arthritis. In the present study, we examined rat strains exhibiting differential sensitivity to C trachomatis–induced arthritis in order to identify the early synovial cytokine profile defining effective host clearance of the pathogen and, in Supported by the Canadian Institutes of Health Research. Robert D. Inman, MD, Basil Chiu, PhD: Toronto Western Research Institute, University of Toronto, Toronto, Ontario, Canada. Address correspondence and reprint requests to Robert D. Inman, MD, Arthritis Center of Excellence, Toronto Western Hospital, 399 Bathurst Street, Toronto, Ontario M5T 2S8, Canada. E-mail: Robert.inman@uhn.on.ca. Submitted for publication July 25, 2005; accepted in revised form November 10, 2005. 499 500 INMAN AND CHIU turn, the chronicity of the arthritis. The choice of the Lewis strain to establish this model initially was based on the fact that it is susceptible to a number of experimental arthritis models (5), including adjuvant-induced arthritis, collagen-induced arthritis (6,7), as well as Yersiniainduced arthritis (8–11). Studies on the rat model of collagen-induced arthritis have mainly focused on the genetic basis for the development of the disease and the antibody response to collagen. This genetics work was pioneered by Griffiths and associates and has been extensively reviewed (12). It was determined that the major histocompatibility complex (MHC) plays an important part in resistance or susceptibility to the disease among rat strains (6,7). In general, high antibody response to type II collagen indicates susceptibility to the development of disease. However, nonarthritic rats have been shown to have high antibody titers (7,13,14). Therefore, the relationship of antibody titer and disease is not absolute. Working with uncommon WA/KIR rats, Rahman and Staines (15) isolated and cultured antibody-producing cells from the spleen, lymph nodes, and bone marrow of rats. They observed no difference in anticollagen antibody production between arthritic and nonarthritic rats. Splenectomy led to more severe disease and earlier onset. They also observed that if experimental animals were given splenocytes obtained from classically immunized rats, the severity of arthritis was greatly reduced. Working mostly with DA rats, Griffiths et al (12,16) identified a series of genetic loci that they collectively named Cia, which modify the severity of disease development. In order to define the pathogenesis of joint inflammation in C trachomatis–induced arthritis, we compared different rat strains for differential susceptibility. We included BN rats as a control, because this rat strain is known to show relative resistance to collageninduced arthritis. Here, we report the contrasting profile of host susceptibility, represented by Lewis rats, and resistance, represented by BN rats, in response to experimental Chlamydia-induced arthritis. MATERIALS AND METHODS Rats. Eight-week-old male rats were purchased from Harlan Sprague Dawley (Indianapolis, IN). The animals were maintained in microisolators in the animal care facility of the Toronto Western Hospital. All animals studied were younger than 12 weeks of age. All studies were approved by the University Health Network Animal Care Committee. Induction of arthritis. Arthritis was induced in the rats by intraarticular injection of synoviocyte-packaged Chlamydia, as previously described (4). Briefly, C trachomatis serotype L2 was inoculated onto monolayers of rat synovial fibroblasts in tissue culture. These stable synovial fibroblast lines were developed as previously described (4). After overnight incubation, cells were harvested and adjusted to 5 ⫻ 105 cells/ml. Rats were anesthetized with isoflurane (Zenoca Pharma, Mississauga, Ontario, Canada), and 0.2 ml of the infected cells containing 2 ⫻ 105 colony-forming units of Chlamydia was injected into one knee joint of the rats of the respective strains. Joint swelling was measured with calipers and recorded in millimeters. Animals were killed on days 2, 7, or 21 postinjection. Mock injections on noninfected synoviocytes demonstrated only a transient inflammation in the joints (4). Histologic analysis. At necropsy, the arthritic knee joints were removed, fixed in formalin, and decalcified. They were then processed and stained with hematoxylin and eosin. The joint sections were evaluated and scored according to a published scoring system (17). Briefly, 5 criteria were used to grade the joints: synovitis of the synovial tissue, inflammatory cell infiltration, joint space exudation, pannus formation, and bone/cartilage damage. Each category was assigned a score of 0–4, where 0 ⫽ normal and 4 ⫽ extreme pathologic change (maximum possible score ⫽ 20). The score for each joint section was then calculated. Chlamydia clearance in vivo. A quantitative assay for Chlamydia in synovial tissues of the injected knee joints was adapted from the IDEIA PCE Chlamydia enzyme-linked immunosorbent assay (ELISA) kit (Dako, Cambridgeshire, UK), which detects the Chlamydia genus only (antilipopolysaccharide specificity). It is a capture ELISA–based assay using a patented polymer chain–conjugated antibody technique and is as sensitive as current polymerase chain reaction (PCR) kits. This kit was originally developed for clinical use but has previously been used successfully in a murine experimental model (18). The joints of the test animals were carefully dissected to harvest the synovial tissues. These tissues were weighed, ground to a pulp, and then resuspended in saline at 40 mg of wet weight per milliliter. The samples were then processed as recommended by the manufacturer. A similar analysis was applied to splenic mononuclear cells that were harvested on day 7 postinjection (see below). Splenocyte clearance of Chlamydia. Lewis and BN rats were killed 7 days postinjection, and the spleens were recovered and homogenized. The amount of Chlamydia was assayed using the Dako IDEIA PCE Chlamydia kit, adapted for use in rats (18). Cytokine analysis of synovial tissues. Synovial tissues were dissected, individually weighed, and dissolved in a Tris– saline–EDTA buffer containing 0.5 mM DTT and 1% Triton X-100, as described by Niki et al (19). The dot enzyme immunoassay procedure is similar in principle to a dot-blot assay, except that the tissue lysates are dotted onto small nitrocellulose discs in the wells of tissue culture plates instead of dotted onto a nitrocellulose membrane in a dot-blot apparatus frame. Five milliliters of the tissue lysate was dotted onto each disc, air-dried in the cold, and stored at ⫺70° until assayed. The assay was performed directly in the wells. Goat polyclonal antibodies to rat interleukin-1 (IL1), IL-4, IL-6, IL-10, interferon-␥ (IFN␥), tumor necrosis factor ␣ (TNF␣), and transforming growth factor  were purchased from Santa Cruz Biotechnology (Santa Cruz, CA) HOST–MICROBE INTERACTIONS IN CHLAMYDIA-INDUCED ARTHRITIS and used at a 1:20 dilution. Antibodies to macrophage inflammatory protein 1␣ and monocyte chemoattractant protein 1 were obtained from Chemicon (Temecula, CA). Discs were incubated for 90 minutes at 37°C. Each well containing a disc was washed and incubated with the secondary peroxidaseconjugated anti-goat IgG antibody (1:8,000 dilution) (Sigma, St. Louis, MO) for 90 minutes at 37°C. Thereafter, color development utilized o-phenylenediamine and H2O2. After the reaction was stopped, the fluid from each well was transferred into wells of a plate for reading in an ELISA reader. Cobra venom factor (CVF) study. In order to evaluate the role of complement in the differential host response to the intraarticular pathogen, we studied the impact of complement depletion in the respective rat strains. Similar approaches have been used in studying arthritis in the rat in different experimental models (20,21). CVF was purchased from Quidel (San Diego, CA). It has a vender-defined stock concentration of 89 units/ml CVF and was administered intraperitoneally at 25 units/kg. The CVF was given at the same time as the intraarticular injection of Chlamydia. On day 7, the rats were killed, and EDTA plasma was assayed for C3a desArg as a marker of C3 activation of C3. The ELISA kit for measuring rat complement C3a desArg was from Cedarlane (Hornby, Ontario, Canada). Statistical analysis. All results are expressed as the mean ⫾ 1SD. Student’s 2-tailed t-test was used for statistical comparisons. 501 observed. Joint spaces were usually clear, with only light to moderate exudate. Pannus was present, but pannus formation in BN rats was never as aggressive as that observed in Lewis rats. Compared with Lewis rats, BN rats infrequently had erosions, and when erosions were observed in BN rats, they were only superficial. Twenty-one days postinjection, synovial tissues from Lewis rats were hypertrophic, with extensive infiltration of leukocytes consisting of both neutrophils and mononuclear cells (Figure 2). Bone and cartilage damage was evident, resulting in a picture of extensively disrupted joint architecture. In contrast, synovial tissues from BN rats showed only mild hypertrophic changes on day 21, and the joint structure approximated that of a normal joint. Quantitative analysis of the histopathologic changes confirmed the significant differences between the 2 strains of rats. Two days after injection, the pathology score for Lewis rats had already reached 10.0 on the Yang-Hamilton scale (17), which has a maximum score of 20. The mean pathology scores rose quickly, to near-maximum scores of 18.6 by day 7 postinjection, and RESULTS Histology. The initial impression of only transient joint swelling in the BN rat prompted an examination of the histopathologic characteristics of the joints in the 2 rat strains. On day 2 postinjection, both strains had moderate infiltration of leukocytes into the synovial tissues, with an impression of neutrophilic predominance in Lewis rats and mononuclear predominance in BN rats. On day 7 postinjection, the contrast in pathologic changes was more dramatic (Figure 1). The injected knee joints of Lewis rats were very densely infiltrated with neutrophils. There were focal points of necrosis in the synovial tissue of some animals. Synovial hyperplasia and hypertrophy were extensive. The joint space contained a significant exudate, which was primarily neutrophilic. Pannus formation was extensive. Significant erosion of bone and articular cartilage was present. In some cases, pannus had advanced into the subchondral bone and was seen proliferating into the bone marrow space. Seven days postinjection, the pathologic changes in BN rats were much milder than those in Lewis rats. The predominant infiltrating cells were mononuclear, and cellular infiltration was much more limited than that seen in Lewis rats. Hypertrophy and hyperplasia of the synovial tissues were minimal, and necrosis was not Figure 1. Joint histopathologic changes on day 7 postinjection in Lewis rats (A) and BN rats (B) with Chlamydia trachomatis–induced arthritis. Hematoxylin and eosin staining shows more aggressive synovitis in Lewis rats than in BN rats. Original magnification ⫻ 10. 502 INMAN AND CHIU Table 1. Effect of decomplementation on differential joint swelling in Lewis rats and BN rats* Treatment C trachomatis only† Rat 1 Rat 2 Rat 3 Rat 4 Rat 5 C trachomatis ⫹ CVF‡ Rat 1 Rat 2 Rat 3 Rat 4 Rat 5 Lewis BN 12.9 12.3 13.1 13.7 13.2 10.8 10.5 10.3 12.5 10.8 13.7 13.7 12.6 12.5 13.5 12.0 12.3 9.9 12.4 10.6 * Values are measurements (in millimeters) of the right knee on day 7. BN rats exhibited significantly less swelling in the Chlamydia trachomatis–injected joint than did Lewis rats, in both the absence (P ⫽ 0.002) and the presence (P ⫽ 0.01) of cobra venom factor (CVF). CVF had no significant impact on swelling in C trachomatis–injected joints in either Lewis rats (P ⫽ 0.66) or BN rats (P ⫽ 0.49). † The mean ⫾ SD values for Lewis rats and BN rats were 13.04 ⫾ 0.50 mm and 10.98 ⫾ 0.87 mm, respectively. ‡ The mean ⫾ SD values for Lewis rats and BN rats were 13.02 ⫾ 0.60 mm and 11.44 ⫾ 1.12 mm, respectively. Figure 2. Joint histopathologic changes on day 21 postinjection in Lewis rats (A) and BN rats (B) with Chlamydia trachomatis–induced arthritis. Hematoxylin and eosin staining shows continued active synovitis in Lewis rats, with BN rats showing no inflammation or joint injury. Original magnification ⫻ 10. remained high (16.2) on day 21. For the BN rats, the day 2 score was 6.0. The highest score reached was only 13.2; this score was reached on day 7 and then declined rapidly back to that of noninflamed joint architecture on day 21. The scores of the BN rats were significantly lower (P ⬍ 0.05) than those of the Lewis rats on days 2, 7, and 21 postinjection (n ⫽ 10 rats at each time point). Influence of complement. We addressed the role of complement in this process by using CVF to decomplement plasma in vivo. It was noted that on day 7 after induction of arthritis, the level of C3a desArg was higher in BN rats (mean ⫾ SD OD units 0.87 ⫾ 0.19) than in Lewis rats (0.44 ⫾ 0.11) (P ⬍ 0.001). After CVF treatment, this difference was not observed (P ⫽ 0.805), with both strains showing comparably increased levels of C3a desArg. Thus, despite eliminating the differential complement levels, there was no impact on the severity of arthritis, and the differential response in swelling between the strains was still observed after decomplementation (Table 1). Chlamydia clearance. The clearance of Chlamydia in synovial tissues on day 7 postinjection was quantitated by ELISA (Figure 3). Lewis rats had a significantly greater Chlamydia load than did BN rats (mean OD values 0.791 and 0.141, respectively; P ⫽ 0.0018). This difference was also reflected in systemic clearance of Chlamydia, using spleen as the target tissue Figure 3. Clearance of Chlamydia in synovial tissues obtained from Lewis (Lew) and BN rats. On day 7, quantitative analysis of Chlamydia showed more effective host clearance of the pathogen from the joints of BN rats (n ⫽ 3) than from the joints of Lewis rats (n ⫽ 3). Values are the mean and SD. OD ⫽ optical density. HOST–MICROBE INTERACTIONS IN CHLAMYDIA-INDUCED ARTHRITIS 503 correlation of synovial tissue weight (reflecting inflammatory infiltration, hypertrophy, and edema) and synovial IFN␥ expression (r ⫽ ⫺0.86). This inverse correlation was also observed for TNF␣ and synovial tissue mass (r ⫽ ⫺0.70). DISCUSSION Figure 4. Effect of cobra venom factor (CVF) on recovery of Chlamydia trachomatis (Ct) from splenocytes. On day 7, clearance of C trachomatis from splenocytes was enhanced in BN rats (n ⫽ 3) compared with Lewis rats (n ⫽ 3), both in the absence (P ⬍ 0.005) and the presence (P ⬍ 0.05) of CVF treatment. Values are the mean and SD. OD ⫽ optical density. to assess dissemination and persistence. Splenocytes from BN rats showed less recovery of the organism than did splenocytes from Lewis rats, and this was true in the presence and in the absence of systemic decomplementation with CVF (Figure 4). Synovial cytokine profiles in Lewis rats and BN rats. An analysis of synovial cytokine expression in Lewis rats and BN rats revealed no differences between the 2 strains for the following cytokines: IL-1, IL-6, IL-10, MIP, and MCP-1 (Figure 5). However, BN rats showed significantly higher synovial expression of IL-4, IFN␥, and TNF␣ (P ⬍ 0.05 for each comparison). When the tissues were analyzed collectively, there was an inverse Figure 5. Cytokine levels in synovial tissue on day 5. Local production of interleukin-4 (IL-4), interferon-␥ (IFN␥), and tumor necrosis factor ␣ (TNF␣) was enhanced in BN rats (n ⫽ 6) compared with Lewis rats (n ⫽ 6) (P ⬍ 0.05 for all 3 cytokines). OD ⫽ optical density; TGF ⫽ transforming growth factor; MIP ⫽ macrophage inflammatory protein; MCP ⫽ monocyte chemoattractant protein. Values are the mean and SD. ReA refers to an aseptic, seronegative arthritic condition in which genetically susceptible individuals experience joint inflammation as a sequela of certain bacterial infections. Patients in whom ReA develops commonly, but not invariably, express HLA–B27 or one of the related cross-reactive group (CREG) antigens of the class I MHC. Arthritis develops following either a gastrointestinal infection with gram-negative organisms such as Salmonella, Yersinia, Shigella, or Campylobacter species or a genitourinary tract infection with C trachomatis. Despite the strong circumstantial clinical evidence linking chlamydial infection to subsequent arthritis, the mechanism underlying this process has not been defined. Recently, there has been mounting evidence that bacteria may indeed be found in the joint in this setting. The demarcation line between septic arthritis and ReA has since been blurred as a result of such evidence. Several groups of investigators, using immunofluorescence techniques, identified Chlamydia-like inclusion bodies in synovial biopsy specimens from patients with ReA (22–24). Schumacher et al (22) also observed characteristic electron-dense bodies of Chlamydia in synovial specimens. Subsequently, it was reported that chromosomal DNA material encoding Chlamydia ribosomal RNA (rRNA) could be demonstrated in synovial samples (25). Besides the gene sequence for 16S rRNA, Bas et al (26) detected a major outer membrane protein (MOMP) gene sequence and 2 chlamydial plasmid sequences. Several investigators then reported obtaining positive PCR results for 16S rRNA (27,28). Gerard et al (29) took synovial specimens that tested positive for 16S rRNA by PCR and then further identified chlamydial gene products (glyQS, r-protein S5 and L5, and Hsp60) within a majority of these biopsy specimens. Nonetheless, as molecular detection technologies began to show the existence and the importance of chlamydial genetic material within the inflamed synovial tissue of ReA joints, several studies raised some question about the specificity of PCR results in synovial tissue. Using PCR, Schumacher et al observed C trachomatis 16S rRNA as well as MOMP gene materials in synovial biopsy specimens from normal volunteers and patients with rheumatoid arthritis (30) and subsequently 504 reported that the chlamydial 16S and MOMP, as well as plasmid gene materials, could be found by PCR in some biopsy specimens from osteoarthritic joints (31). In view of the difficulties defining precise mechanisms of C trachomatis–induced arthritis in the clinical setting, we developed an animal model that recapitulates many of the features seen in clinical ReA (4). What begins as a septic process in the acute phase develops into a chronic arthritis that is resistant to antibiotic treatment and from which viable organisms cannot be recovered. However, the host factors that mediate this process were not defined, and this may shed light on the earliest events in the pathogenesis of C trachomatis– induced arthritis. In the course of our studies, we discovered that BN rats demonstrated a relative resistance to intraarticular challenge with C trachomatis, while Lewis rats demonstrated an aggressive arthritis following this challenge. In BN rats, clinical events following the challenge were much milder than those observed in Lewis rats, with less joint swelling and a transient clinical course. These events were mirrored in significant differences in the histopathologic findings in the affected joints. The BN rats had a mononuclear leukocyte synovial response to the intraarticular injection of Chlamydia, while the acute response in Lewis rats was dominated by neutrophils. In Lewis rats, lymphoid cells were not apparent until the chronic phase of disease. More precise immunostaining was hampered by the decalcification process, which complicates detection of surface markers on infiltrating cells in the joint. Despite the lack of phagocytic cells, BN rats were able to clear bacteria from the injected knees much more effectively than were Lewis rats. In Lewis rats, massive infiltration of phagocytic cells was not effective in this respect and likely contributed to the extensive joint damage seen in these animals. BN rats are very different in this regard and responded in a most effective way to counter the microbial challenge in the joint. The lymphoid infiltration seemed to be part of a self-limiting process that is associated with much less collateral joint damage. The differences in host response to the pathogen occur in the acute phase of arthritis, which suggests that innate immune differences are playing a key role here. In this regard, we evaluated whether differences in host complement might be playing an important part, but decomplementation was not associated with any differential response that would support this notion. Curfs et al (10) found that Lewis rats were susceptible to a Yersinia-induced ReA, while BN and Fischer rats were resistant. Hill and Yu (8) also observed that Fischer, INMAN AND CHIU Buffalo, and DA rats are resistant to Yersinia-induced ReA. It is of interest that the BN rat is recognized for its tendency to develop certain autoimmune diseases. The best-studied model is the asthma model in which BN rats preimmunized with ovalbumin are challenged with an aerosol of the immunogen, resulting in a pulmonary infiltration of eosinophils, neutrophils, and mononuclear cells. Lewis rats are resistant in this model (32). The host response to chlamydial infection is a complex interrelationship of cellular and humoral immunity (2). Increased production and secretion of proinflammatory mediators in the clinical setting of C trachomatis–induced arthritis have been examined by several groups of investigators (33,34). Studies suggest that host cell responses are diverse and dependent on the temporal course. Analyses of clinical samples in the chronic phase have suggested that a mixture of monocytes and T cell cytokines, as well as proinflammatory mediators produced by resident synoviocytes, is present at sites of inflammation. Mononuclear cells in synovial fluid have been shown to express high levels of IL-4 and IL-10. Consistent with culture data, more IL-4– expressing cells than IFN␥-expressing cells were observed in the synovial membrane (35). The time course of most clinical cases of C trachomatis–induced arthritis cannot be defined precisely, but studies to date have focused on chronic changes, in part because it is difficult to examine tissue and serum during the induction of arthritis, because patients are rarely seen at that time. In vitro studies of Chlamydia-infected synoviocytes have demonstrated the production of soluble mediators by resident synoviocytes that could influence local immune and inflammatory responses. Rodel et al (36) showed that chlamydial infection up-regulates indoleamine 2,3-dioxygenase gene expression, and the latter was enhanced by TNF␣. Chlamydial infection induced type I IFN␣/ in these cells, presumably via activation of IFN regulatory factor 1 and IFN-stimulated gene factor 3. The precise cellular difference accounting for the differences between Lewis and BN rats has not been fully defined. We have conducted in vitro studies examining chlamydia invasion into synoviocytes from the 2 rat strains but found no difference in this primary event. The role of cytokines in the pathogenesis of ReA has proven to be a complex issue to resolve definitively. It is known that IFN␥ and TNF␣ are indispensable for an effective defense against bacterial infection. In ReA, failure of effective bacterial elimination at the initiation of the disease may be attributable to a relative lack of Th1 cytokine production. Several studies have demon- HOST–MICROBE INTERACTIONS IN CHLAMYDIA-INDUCED ARTHRITIS strated a relative decrease in Th1 cytokines in peripheral blood and synovium of patients with ReA (35,37). In a study of 11 patients with ReA, it was shown that stimulation of synovial fluid mononuclear cells resulted in low levels of IFN␥ and TNF␣ but high amounts of IL-10 (35). In T cell clones derived from the synovial fluid of patients with ReA, disease-related bacterial antigens predominantly induced Th1 cytokine secretion (38–40). It is interesting to note that in Chlamydia-induced ReA, synovial fluid levels of IFN␥ were lower in HLA– B27–positive patients than in HLA–B27–negative patients (41). This bears an interesting relationship to our findings in the present study, in which as-yetuncharacterized genetic susceptibility conferring sensitivity to experimental C trachomatis–induced arthritis was also associated with a relative lack of IFN␥ production locally. A comparable analogy can be seen in the significant associations observed between low TNF␣ secretion and a more chronic course of ReA in HLA– B27–positive patients, but that is not always the case (47,42). This also could represent the clinical parallel to that seen in the current study, in which relatively impaired host TNF␣ production was associated with the chronicity and severity of the experimental C trachomatis–induced arthritis. It is interesting to note that the TNF alleles that have been associated with ankylosing spondylitis are those that confer lower, not higher, TNF production (43). This paradoxical finding has been difficult to explain. Our findings in experimental arthritis suggest that an impairment during the acute phase of arthritis might set the stage for a more chronic course. In comparison with synovial fluid levels of TNF␣ in rheumatoid arthritis, synovial fluid levels of TNF␣ in ReA are lower, despite comparable levels of IL-2 receptor. A lower expression of TNF␣ has been observed in peripheral blood lymphocytes from HLA–B27–positive individuals; therefore, impairment of such microbicidal cytokines may set the stage for an altered host response to infection. This possibility is supported by our experimental studies in which deficiency of TNF␣ was associated with more severe infection and more severe arthritis (44). It is difficult in the clinical setting to resolve whether the relative shift to Th2-type cytokine production in ReA (in contrast to rheumatoid arthritis) reflects a genetic predilection for a blunted host response to such infections, or whether this is a profile that reflects the host attempt at gradual, but eventual, clearance of the pathogen. The role of innate immunity in the host response to Chlamydia has just recently come under scrutiny (45). Netea et al reported that Chlamydia 505 pneumoniae stimulates IFN␥ synthesis through MyD88dependent, Toll-like receptor 2 (TLR-2)– and TLR-4– independent induction of IL-18 release (46). Polymorphisms in CD14 may confer differential host responsiveness to chlamydia (47). Eng et al recently reported a CD14 promoter polymorphism associated with CD14 expression and Chlamydia-stimulated TNF␣ production (48). Recent studies have demonstrated that neutrophil Rac and TLR-4 may play important roles in this context (49). The role of cytokines in the joint has become an important issue, not only for understanding the pathogenesis of arthritis but also because of the advent of anticytokine therapies. Such treatment in patients with spondylarthropathies may have a tendency to normalize impaired Th1 cytokine production, but this issue has not been finally resolved (50–52). Our findings highlight the fact that cytokine expression in the joint is dynamic, with beneficial and deleterious effects varying according to the time of observation. Furthermore, the complex microenvironment of the joint reflects a finely balanced system, and interrupting this cytokine circuitry may have unexpected effects. This is not yet resolved in clinical ReA. REFERENCES 1. Inman RD, Perl A, Phillips PE. Infectious agents in chronic rheumatic diseases. In: Koopman WJ, Moreland LW, editors. Arthritis and allied conditions: a textbook of rheumatology. 15th ed. Philadelphia: Lippincott Williams & Wilkins; 2005. p. 647–8. 2. Inman RD, Whittum-Hudson JA, Schumacher HR, Hudson AP. Chlamydia and associated arthritis. Curr Opin Rheumatol 2000; 12:254–60. 3. Thomson GT, DeRubeis DA, Hodge MA, Rajanayagam C, Inman RD. Post-Salmonella reactive arthritis: late clinical sequelae in a point source cohort. Am J Med 1995;98:13–21. 4. Inman RD, Chiu B. Synoviocyte-packaged Chlamydia trachomatis induces a chronic aseptic arthritis. J Clin Invest 1998;102:1776–82. 5. Joe B, Griffiths M, Remmers EF, Wilder RL. Animal models of rheumatoid arthritis and related inflammation. Curr Rheumatol Rep 1999;1:139–48. 6. Griffiths MM, Eichwald EJ, Martin JH, Smith CB, DeWitt CW. Immunogenetic control of experimental type II collagen–induced arthritis. I. Susceptibility and resistance among inbred strains of rats. Arthritis Rheum 1981;24:781—94. 7. Griffiths MM, DeWitt CW. Genetic control of collagen-induced arthritis in rats: the immune response to type II collagen among susceptible and resistant strains and evidence of multiple gene control. J Immunol 1984;132:2830–6. 8. Hill JL, Yu DT. Development of an experimental animal model for reactive arthritis induced by Yersinia enterocolitica infection. Infect Immun 1987;55:721–6. 9. Toivanen A, Merillahti-Palo R, Gripenberg C, Lahesman-Rantala R, Soderstrom KO, Jaakkola UM. Yersinia-associated arthritis in the rat: experimental model for human reactive arthritis. Acta Path Microb Immun Scand 1986;94:261–9. 10. Curfs JH, Meis JF, van der Lee HA, Mulder JA, Kraak WA, 506 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. INMAN AND CHIU Hoogkamp-Korstanje JA. Persistent Yersinia enterocolitica infection in three rat strains. Microb Pathog 1995;19:57–63. Odinot RT, Curfs JH, Meis JF, Melchers WJ, HoogkampKorstanje JA. Local expression of cytokine mRNA in spleen and Peyer’s patches of rats is involved in resistance against infection with Yersinia enterocolitica. Cytokine 1998;10:206–12. Joe B, Griffiths MM, Remmers EF, Wilder RL. Animal models of rheumatoid arthritis and related inflammation. Curr Rheumatol Rep 1999;1:139–48. Clague RB, Morgan K, Shaw MJ, Holt PJ. Native type II collagen induced arthritis in rat. J Rheumatol 1980;7:775–82. Larsson P, Kleinan S, Holmdahl R, Klareskog L. Homologous type II collagen–induced arthritis in rats: characterization of the disease and demonstration of clinically distinct forms of arthritis in two strains of rats after immunization with the same collagen preparation. Arthritis Rheum 1990;33:693–701. Rahman J, Staines NA. Contribution of the spleen, lymph nodes and bone marrow to the antibody response in collagen induced arthritis in the rat. Clin Exp Immunol 1991;85:48–54. Griffiths MM, Wang J, Joe B, Dracheva S, Kawahito Y, Cannon GW, et al. Identification of four new quantitative trait loci regulating arthritis severity and one new quantitative trait locus regulating autoantibody productions in rats with collagen-induced arthritis. Arthritis Rheum 2000;43:1278–89. Yang YH, Hamilton JA. Dependence of interleukin-1–induced arthritis on granulocyte–macrophage colony-stimulating factor. Arthritis Rheum 2001;44:111–9. Haranaga S, Yamaguchi H, Friedman H, Izumi SI, Yamamoto Y. Chlamydia pneumoniae infects and multiplies in lymphocytes in vitro. Infect Immun 2001;69:7753–9. Niki Y, Yamada H, Seki S, Kikuchi T, Takaishi H, Toyama Y, et al. Macrophage- and neutrophil-dominant arthritis in human IL-1 transgenic mice. J Clin Invest 2001;107:1127–113. Gaede K, Baumeister E, Heeseman J. Decomplementation by cobra venom factor suppresses Yersinia-induced arthritis in rats. Infect Immun 1995;63:3697–701. Goodfellow RM, Williams AS, Levin JL, Williams BD, Morgan BP. Soluble complement receptor one (sCR1) inhibits the development and progression of rat collagen induced arthritis. Clin Exp Immunol 2000;19:210–6. Schumacher HR Jr, Magge S, Cherian PV, Sleckman J, Rothfuss S, Clayburne G, et al. Light and electron microscopic studies on the synovial membrane in Reiter’s syndrome: immunocytochemical identification of chlamydial antigen in patients with early disease. Arthritis Rheum 1988;31:937–46. Taylor-Robinson D, Thomas BJ, Dixey J, Osborn MF, Furr PM, Keat AC. Evidence that Chlamydia trachomatis causes seronegative arthritis in women. Ann Rheum Dis 1988;47:295–9. Inman RD, Chiu B, Katz A. Immunological features of “reactive arthritis” due to Chlamydia trachomatis. Arthritis Rheum 1989;32 Suppl 1:S113. Rahman MU, Cheema MA, Schumacher HR, Hudson AP. Molecular evidence for the presence of chlamydia in the synovium of patients with Reiter’s syndrome. Arthritis Rheum 1992;35:521–9. Bas S, Griffais R, Kvien T, Glennas A, Melby K, Vischer TL. Amplification of plasmid and chromosome Chlamydia DNA in synovial fluid of patients with reactive arthritis and undifferentiated seronegative oligoarthropathies. Arthritis Rheum 1995;38: 1005–13. Nanagara R, Li F, Beutler A, Hudson A, Schumacher HR Jr. Alteration of Chlamydia trachomatis biologic behavior in synovial membranes: suppression of surface antigen production in reactive arthritis and Reiter’s syndrome. Arthritis Rheum 1995;38:1410–7. Branigan PJ, Gerard HC, Hudson AP, Schumacher HR Jr, Pando J. Comparison of synovial tissue and synovial fluid as the source of nucleic acids for detection of Chlamydia trachomatis by polymerase chain reaction. Arthritis Rheum 1996;39:1740–6. 29. Gerard HC, Branigan PJ, Schumacher HR Jr, Hudson AP. Synovial Chlamydia trachomatis in patients with reactive arthritis/ Reiter’s syndrome are viable but show aberrant gene expression. J Rheumatol 1998;25:734–42. 30. Schumacher HR, Arayssi T, Crane M, Lee J, Gerard H, Hudson AP, et al. Chlamydia trachomatis nucleic acid can be found in the synovium of some asymptomatic subjects. Arthritis Rheum 1999; 42:1281–4. 31. Olmez N, Wang GF, Li Y, Zhang H, Schumacher HR. Chlamydial nucleic acid in synovium in osteoarthritis: what are the implications. J Rheumatol 2001;28:1874–80. 32. Hylkema MN, Hoekstra MO, Luinge M, Timens W. The strength of the OVA-induced airway inflammation in rats is strain dependent. Clin Exp Immunol 2002;29:390–6. 33. Kotake S, Schumacher HR, Arayssi TK, Gerard HC, Branigan PJ, Hudson AP, et al. Gamma interferon and interleukin-10 gene expression in synovial tissues from patients with early stages of Chlamydia-associated arthritis and undifferentiated oligoarthritis and from healthy volunteers. Infect Immun 1999;67:2682–6. 34. Muller B, Gimsa U, Mitchison NA, Radbruch A, Sieper J, Yin Z. Modulating the Th1/Th2 balance in inflammatory arthritis. Springer Semin Immunopathol 1998;20:181–96. 35. Yin Z, Braun J, Neure L, Wu P, Liu L, Eggens U, et al. Crucial role of interleukin-10/interleukin-12 balance in the regulation of the type 2 T helper cytokine response in reactive arthritis. Arthritis Rheum 1997;40:1788–97. 36. Rodel J, Groh A, Hartmann M, Schmidt KH, Lehmann M, Lungerhausen W, et al. Expression of interferon regulatory factors and indoleamine 2,3-dioxygenase in Chlamydia trachomatis-infected synovial fibroblasts. Med Microbiol Immunol (Berl) 1999; 187:205–12. 37. Braun J, Yin Z, Spiller I, Siegert S, Rudwaleit M, Liu L, et al. Low secretion of tumor necrosis factor ␣, but no other Th1 or Th2 cytokines, by peripheral blood mononuclear cells correlates with chronicity in reactive arthritis. Arthritis Rheum 1999;42:2039–44. 38. Lahesmaa R, Yssel H, Batsford S, Luukkainen R, Mottonen T, Steinman L, et al. Yersinia enterocolitica activates a T helper type 1-like T cell subset in reactive arthritis. J Immunol 1992;148: 3079–85. 39. Schlaak J, Hermann E, Ringhoffer M, Probst P, Gallati H, Meyer zum Buschenfelde KH, et al. Predominance of Th1-type T cells in synovial fluid of patients with Yersinia-induced reactive arthritis. Eur J Immunol 1992;22:2771–6. 40. Simon AK, Seipelt E, Wu P, Wenzel B, Braun J, Sieper J. Analysis of cytokine profiles in synovial T cell clones from chlamydial reactive arthritis patients: predominance of the Th1 subset. Clin Exp Immunol 1993;94:122–6. 41. Bas S, Kvien TK, Buchs N, Fulpius T, Gabay C. Lower level of synovial fluid interferon-␥ in HLA-B27-positive than in HLA-B27negative patients with Chlamydia trachomatis reactive arthritis. J Rheumatol 2003;42:461–7. 42. Butrimiene I, Jarmalaite S, Ranceva J, Venalis A, Jasiuleviciute L, Zvirbliene S. Different cytokine profiles in patients with chronic and acute reactive arthritis. Rheumatology (Oxford) 2004;43:1300–4. 43. Rudwaleit M, Siegert S, Yin Z, Eick J, Thiel A, Radbruch A, et al. Low T cell production of TNF␣ and IFN␥ in ankylosing spondylitis: its relation to HLA-B27 and influence of the TNF-308 gene polymorphism. Ann Rheum Dis 2001;60:36–42. 44. Zhao YX, Zhang H, Chiu B, Payne U, Inman RD. Tumor necrosis factor receptor p55 controls the severity of arthritis in experimental Yersinia enterocolitica infection. Arthritis Rheum 1999;42: 1662–72. 45. Prebeck S, Brade H, Kirschning CJ, da Costa CP, Durr S, Wagner H, et al. The Gram-negative bacterium Chlamydia trachomatis L2 stimulates tumor necrosis factor secretion by innate immune cells independently of its endotoxin. Microbes Infect 2003;5:463–70. 46. Netea MG, Kullberg BJ, Jacobs LE, Verver-Jansen TJ, van der HOST–MICROBE INTERACTIONS IN CHLAMYDIA-INDUCED ARTHRITIS Ven-Jongekrijg J, Galama JM, et al. Chlamydia pneumoniae stimulates IFN␥ synthesis through MyD88-dependent, TLR2- and TLR4-independent induction of IL-18 release. J Immunol 2004; 173:1477–82. 47. Rupp J, Goepel W, Kramme E, Jahn J, Solbach W, Maass M. CD14 promoter polymorphism –159C⬎T is associated with susceptibility to chronic Chlamydia pneumoniae infection in peripheral blood monocytes. Genes Immun 2004;5:435–8. 48. Eng HL, Wang CH, Chen CH, Chou MH, Cheng CT, Lin TM. A CD14 promoter polymorphism is associated with CD14 expression and Chlamydia-stimulated TNF␣ production. Genes Immun 2004; 5:426–30. 49. Zhang X, Glogauer M, Zhu F, Kim TH, Chiu B, Inman RD. Innate immunity and arthritis: neutrophil Rac and Toll-like recep- 507 tor 4 expression define outcomes in infection-triggered arthritis. Arthritis Rheum 2005;52:1297–304. 50. Baeten D, Van Damme N, Van den Bosch F, Kruithof E, De Vos M, Mielants H, et al. Impaired Th1 cytokine production in spondyloarthropathy is restored by anti-TNF␣. Ann Rheum Dis 2001;60:750–5. 51. Zou J, Rudwaleit M, Brandt J, Thiel A, Braun J, Sieper J. Down-regulation of the nonspecific and antigen-specific T cell cytokine response in ankylosing spondylitis during treatment with infliximab. Arthritis Rheum 2003;48:780–90. 52. Baeten D, Vandooren B, De Rycke L, Veys EM, De Keyser F. Effect of infliximab treatment on T cells cytokine responses in spondylarthropathy: comment on the article by Zou et al [letter]. Arthritis Rheum 2004;50:1015–6. DOI 10.1002/art.21515 Clinical Images: Dysphagia is sometimes easier to swallow than spinal surgery The patient, a 71-year-old man, was referred with a long history of dysphagia. Lateral radiographs of the cervical spine had unexpectedly revealed a pair of huge, beak-shaped osteophytes arising from the anterior vertebral bodies of C2 and C3 (left). The patient described a 10-year-history of intermittent dysphagia to solids, exacerbated in certain neck positions. Videofluoroscopic examination (right) revealed that the epiglottis (E) did not move past the osteophytes during swallowing. This interrupted the flow of the bolus of barium (B) into the esophagus (O) and resulted in silent aspiration (A) of barium into the trachea. The patient declined neurosurgical referral, explaining that he was usually able to swallow normally. While anterior osteophytes cause dysphagia due to compression of the esophagus in rare instances (Benhabyles M, Brattstrom H, Sunden G. Dysphagia and dyspnea as complications in spondylarthritis anklyopoetica with cervical osteophytes. Acta Orthop Scand 1970;41:396–401), this symptom is extremely common after surgery to the anterior cervical spine (Smith-Hammond CA, New KC, Pietrobon R, Curtis DJ, Scharver CH, Turner DA. Prospective analysis of incidence and risk factors of dysphagia in spine surgery patients: comparison of anterior cervical, posterior cervical, and lumbar procedures. Spine 2004;29:1441–6). R. Rajendram, MRCP, MBBS, BSc, AKC J. Ehtisham, MRCP, MBBChir, PhD R. W. Smith, FRCP Milton Keynes General Hospital Milton Keynes, UK
1/--страниц