C 2004) Inflammation, Vol. 28, No. 3, June 2004 (


Inflammation Induced by Inoculation of the Joint with Candida albicans Martin Yordanov,1 Svetla Danova,2 and Nina Ivanovska1,3

Abstract—In humans Candida albicans is the most frequently isolated opportunistic fungal pathogen. In immunocompromized host the balance with the commensal fungus easily turns to life-threatening disseminated infection. The asymptomatic Candida persistence in organs and the recurrent infections suggest continuous circulation of yeast cells and their degradation products. Under certain conditions, joints might become one of the infectious sites. More easily a reactivation and destructive process can be provoked in individuals with established arthritis. We have investigated the joint inflammation caused by inoculation of the paw with live C. albicans, in intact mice and mice with collagen-induced arthritis (CIA). The results demonstrate that C. albicans infection when localized into the joints caused rapidly progressing septic arthritis. The effect was associated with a strong swelling, a rapid influx of polymorphonuclear (PMN) cells, and an elevated secretion of TNF-α and IFN-γ by lymph node cells. Joint infection exacerbated the established CIA which correlated with an increased level of anti-collagen antibodies. KEY WORDS: Candida albicans; septic arthritis; collagen-induced arthritis.

INTRODUCTION

of proinflammatory cytokines during severe systemic infections can provoke arthritic process in predisposed individuals (5). Candida albicans is the most frequently isolated opportunistic fungal pathogen. In immunocompromized individuals the balance between the host and the commensal readily turns to severe infection. Studies in mice have shown that the protective response to C. albicans is due to the prevalence of Th1 cytokines and the relative decrease of Th2 cytokines (6, 7). Early in infection the production of TNF-α is essentially high (8) and at the same time TNF-α is one of the major proinflammatory cytokines in arthritis. Experimental arthritis as a consequence of intravenous inoculation with C. albicans has been induced in rats (9). Altogether, these data raised the question whether C. albicans might play a role in the pathogenesis of arthritis. The purpose of the present experiments was to estimate the ability of the yeast to provoke septic form of arthritis and its effect in mice with established collagen-induced arthritis. After inoculation of the paw with C. albicans, the development of swelling, histopathologic changes, and yeast elimination were

The relationships between infections and rheumatic diseases have been an object of recent investigations in the light of Th1/Th2 balance (1, 2). Although reactive arthritis is often associated with bacterial infections of gastrointestinal or urogenital tracts (3) it is not clear whether bacteria causes or exacerbates arthritis in humans. Epidemiologic studies show that despite antibiotic treatment, inflammation and arthritic joint symptoms continue (4). It is supposed that the commensal flora of the gut induces a tolerant state to self-antigens, while the shift to Th1 response and an increased production 1 Department

of Immunology, Institute of Microbiology, Bulgarian Academy of Sciences, 113 Sofia, Bulgaria. 2 Department of Genetics, Institute of Microbiology, Bulgarian Academy of Sciences, 113 Sofia, Bulgaria. 3 To whom correspondence should be addressed at Department of Immunology, Institute of Microbiology, Bulgarian Academy of Sciences, 26 G. Bonchev St., 1113 Sofia, Bulgaria. E-mail: nina@ microbio.bas.bg

127 0360-3997/04/0600-0127/0


C

2004 Springer Science+Business Media, Inc.

128 determined. The secretion of Th1 cytokines by lymph node cells and serum anticollagen antibody level were also examined. MATERIALS AND METHODS Animals Male and female ICR (CD-2) mice were bred in our facility and were kept on standard commercial pellet diet with water ad libitum. The animals use protocols were approved by the Animal Care Commission at the Institute of Microbiology according to the Helsinki declaration. Yeast A virulent clinical isolate 562 of Candida albicans (Institute of Infectious and Parasitic Diseases, Sofia, Bulgaria) was used in the study. Prior to inoculation, the yeast was grown for 48 h on Saubouraud’s broth (Difco, USA) at 37◦ C. After methylene-blue staining no hyphae were observed. The concentration of the yeast was determined by measuring the extinction at OD620 . A standard of absorbancies based on known colony-forming units (CFU) was used to calculate inoculum concentration. C. albicans cells suspended in a volume of 50 µL of phosphate buffer saline (PBS) were injected s.c. into the right hind paw, the left paw was injected with PBS and used as a negative control. All the injections were performed on anesthized with penthobarbital mice. Induction and Evaluation of Collagen-Induced Arthritis (CIA) We have used a previously described model of arthritis induced in newborn mice. Bovine type II collagen (bCII, Sigma Chemicals Co.) was solubilized to 1 mg/mL by stirring overnight in 0.1 M acetic acid at 4◦ C. Animals were immunized subcutaneously at the back with 50 µg bCII 2 days after birth. The disease symptoms started at around 6 weeks of age and were expressed as mild swelling, synovial cell infiltration and hyperplasia, and anti-bCII antibodies (10). Mice with CIA were inoculated into the paw with C. albicans (1 × 104 cells) at 8 weeks of age. Clinical severity of arthritis was qualified according to a grade scale from 0 to 3 for two hind paws according to the severity of erythema and swelling: a score of 0 indicated no visual erythema and swelling, 1 indicated slight swelling, 2 indicated erythema and obvious swelling, and 3 indicated erythema and severe paw swelling.

Yordanov, Danova, and Ivanovska For histopathologic examination mice were sacrificed, and their limbs were fixed in 10% buffered formalin and decalcified in HNO3 for 12 days. The tissue was then processed and embedded in paraffin. Five-micrometer sections were stained with hematoxylin and eosin (H&E) using standard methodology. Candida Clearance and Cell Populations in the Infected Paw In some experiments subgroups of five mice were killed at different intervals after paw inoculation with 1 × 104 yeast cells. To assess the outgrowth of the fungus, the skin of the paws was removed and the paws were homogenized in 2 mL of phosphate-buffered saline (PBS). After centrifugation at 1200 rpm the number of viable Candida cells in the supernatants was determined by plating serial dilutions on Saubouraud’s agar, and the CFU were counted after 48 h incubation at 37◦ C. Joints exudates from injected and uninjected paws were collected by washing the joint space with 25 µL of PBS. After routine Giemsa-Romanovsky staining, the percentage of lymphocytes, polymorphonuclear cells (PMN) and basophiles plus mast cells was calculated per 200 cells. PCR Analysis of Joint Exudates The PCR reactions were performed in a Progene cycler (Techne, UK) with 0.5 pmol of each primer and 10 µL of samples (joint exudates) in a final volume of 25 µL, using “Ready To Go PCR Beads” (Amersham Pharmacia Biotech, UK). Forward primer (5 -GCA TCG ATG AAG AAC GCA GC-3 ), corresponding to 5.8S rRNA gene and reverse primer (5 -TCC TCC GCT TAT TGA TAT GC-3 ), corresponding to 28S rRNA gene of fungi were purchased from MWG Biotech (Germany). Two reactions for each time point were performed using the protocol described by Wahyuningsih et al. (11). For PCR positive and negative controls 10 µL containing 2 ng of purified Candida DNA or 10 µL distilled sterile water were used, respectively. Generated single PCR product of 340 bp is characteristic only for C. albicans strains. Cytokine Assays Cell suspensions were prepared from freshly removed mesenterial and popliteal lymph nodes. After two washings in RPMI-1640 media (Gibco BRL) cells (5 × 106 cells/ml) were incubated in 24-well Falcon plates (Becton Dickinson, USA) in RPMI-1640 media

Inflammation Induced by Inoculation of the Joint with Candida albicans

129

containing 5% fetal bovine serum (Sigma), 100 U/mL penicillin, 100 µg/ml streptomycin (Sigma), 2 mM Lglutamine (BDH Chemicals), 25 mM 2-mercaptoethanol, and 10 mM HEPES (Sigma). Cells were cultured in the presence or absence of 50 µg/mL of heat-inactivated (1 h, 70◦ C) C. albicans. After 4 days supernatants were collected and frozen at −70◦ C. Cytokine concentration of TNF-α and IFN-γ was determined by ELISA kits (Euroclone, Paington-Devon, UK). Minimum detectable concentration was less than 25 pg/mL for TNF-α and 15 pg/ml for IFN-γ . ELISA Assay for Detection of Serum Anticollagen IgG Antibodies Mice were bled at day 60 after injection of C. albicans and individual sera were analyzed by standard ELISA. Briefly, 96-well flat-bottom ELISA plates (Falcon, USA) were coated with bCII at 100 µg/mL in PBS by incubation overnight at 4◦ C. After washing with PBS/0.05% Tween 20, serum samples (1:100 dilution) were added for 2 h at 37◦ C. The plates were then washed and incubated for 1 h at 37◦ C with peroxidase-labeled goat antimouse IgG (Cappel) diluted 1:10,000. The reaction was proceeded by o-phenylenediamine for 45 min and stopped by adding of 4 N H2 SO4 . The absorbance was measured at 492 nm in ELISA-reader. Statistics Student’s t test was used to determine significance of values among experimental groups.

Fig. 1. Mean arthritic score after injection of 1 × 104 or 1 × 102 yeast cells into the paw of na¨ıve mice or mice with CIA (CIA + 1 × 104 ). Data represent the mean ± SD (n = 5 − 9).

At different time points exudates from the joints of inoculated paws (1 × 104 live yeast) and untreated paws were taken and after routine staining, different cell types were counted. While the ratio between the lymphocytes and PMN showed no significant variation in the untreated joint, a great number of PMN was observed in the exudate from the inoculated joint at the first day (Fig. 2). One week later the lymphocyte populations were predominant, simultaneously with a slight elevation of the number of mast cells and basophiles. At the end of observation (day 60) the percentage of PMN was increased in the injected paws in comparison to the uninjected one. One week after the injection of 1 × 104 Candida cells 100% of mice

RESULTS To examine the ability of C. albicans to induce arthritis live yeast cells were injected directly into the right hind paw of na¨ıve and mice with CIA, using left hind paw, injected with saline as a control. A high dose of 1 × 104 cells caused strong paw swelling and erythema with a peak at day 7 and a slow decrease until day 35 (Fig. 1). Less exerted swelling with erythema only at first 48 h was measured when 1 × 102 live cells were injected into the paw. After 21 days the swelling was fully resorbed. Mice with CIA developed rapid inflammatory reaction, remaining nearly unchanged for the whole period of examination. More expressed signs of arthritis were observed in the joint of PBS-injected paw of mice with CIA + 104 Candida as compared to CIA group (data not shown).

Fig. 2. Different cell populations in the joint exudates from healthy mice (untreated) and from mice injected with 1 × 104 live C. albicans into the paw. The percentage of lymphocytes (Ly), polymorphonuclears (PMN), and basophiles (Ba) plus mast cells (Ma) were calculated per 200 cells (7 mice per group).

130

Fig. 3. Ethidium bromide stained agarose gel (2%) of PCR product (B), obtained from joint exudates of mice at day 4 after i.a. injection of 1 × 102 live C. albicans cells. Lanes: 1, negative control; 2, molecular weight marker (100 bp); 3, sample; 4, C. albicans DNA (positive control).

showed paw contamination (1 × 103 CFU in injected paw). Three weeks later 20% of animals expressed paw contamination (1.8 × 102 CFU) and after 5 weeks no live yeast was isolated from the injected paw. We also inoculated mice with 1 × 102 live cells into the paw and it was established that after the second day live Candida cells were not isolated from the joint. Parallelly, PCR analysis of the joint exudates was performed. It was shown that yeast DNA persisted in the joint for 4 days (Fig. 3), while it was not detected in the joint of PBS-injected paw. When mice with CIA were injected with a higher dose of 1 × 104 Candida cells we established the presence of yeast DNA in both joints until day 4 after no live Candida was isolated. To examine the effect of C. albicans inoculation on the secretion of proinflammatory cytokines, TNF-α and IFN-γ were measured. As shown in Fig. 4, lymph node cells obtained 1 week after yeast injection secreted high levels of both cytokines after in vitro stimulation with inactivated C. albicans. Histopathologic examination showed a lot of infiltrating cells in the synovium and deep in the cartilage at day 3 after the injection of yeast at a dose of 1 × 104 /paw (Fig. 5b). Sixty days postinjection of live Candida bone tissue was severely damaged (Fig. 5c). The development of arthritis is attended with accumulation of anticollagen antibodies. We assessed the presence of anti-CII antibodies in mice with CIA, injected with C. albicans and mice with CIA, injected with Candida (Fig. 6). As previously was established animals with CIA expressed low anti-CII antibody synthesis (10). Inter-

Yordanov, Danova, and Ivanovska

Fig. 4. Secretion of TNF-α and IFN-γ . Mice (n = 5) were injected with PBS (control) or with 1 × 104 C. albicans cells into the paw. After 1 week mesenterial and popliteal lymph node cells were cultured in vitro in the presence of heat inactivated Candida cells (50 µg/mL) for 4 days. The level of both cytokines in the culture supernatants in the absence of stimuli was under the detectable limits of the kits. Data represent mean ± SD from three determinations. ∗ p < 0.001 by Student’s t test.

estingly, in approximately 50% of mice injected with the yeast, anticollagen antibodies were presented . Moreover, mice with CIA which were injected with live Candida showed elevated anti-CII IgG titer as compared to the group with CIA.

DISCUSSION C. albicans is considered to be a rare case of septic arthritis (12–14). The fungus could be introduced into the joint space as a result of surgical procedures or during disseminated infection in immunocompromized patients (15). When the yeast settles directly into the joint it activates inflammatory host defense mechanisms aimed to destroy the pathogen, even at the expense of joint destruction. Data show that not only live yeast but also inactivated C. albicans triggers T-cell proliferation, activates NK cells and induces the production of different cytokines (16, 17). Our experiments showed that after paw inoculation with live C. albicans, the fungus resided into the joint for a long period of 5 weeks, maintaining strong inflammation. One reason for the plosive edema and erythema was the rapid influx of PMN cells. The increase of PMN was characteristic not only for the early phase of inflammation but it was also observed 60 days postinoculation when live pathogen was not isolated. A very low dose of 1 × 102 live cells proved to be sufficient for the initiation of continuos local inflammatory reaction, even in the absence of live pathogen after the second day. PCR analysis showed

Inflammation Induced by Inoculation of the Joint with Candida albicans

131

Fig. 5. Hematoxylin–eosin staining of ankle sections from (a) healthy mice; (b) mice injected with 1 × 104 live Candida cells into the paw with severe synovitis at day 3; (c) mice injected with 1 × 104 live Candida cells into the paw with synovitis and bone destruction at day 60.

that the yeast DNA was detectable in the joint exudates of injected paws after the elimination of the live pathogen. In mice with CIA both joints were susceptible to infection and retaining of yeast DNA in the case of higher inoculation dose. Although under physiologic conditions the microorganisms may be rapidly eliminated, and may not trigger specific immune responses, they may have a proinflammatory role. Through their components such as cell wall components and DNA (18) pathogens possibly can affect the innate immune system. It is well known that arthritic disorders are mediated by the elevation of Th1 cytokines, such as TNF-α and IFN-γ (19–21). TNF-α is able to change the nature of synovial fibroblasts to uncontrolled proliferation and evasion into cartilage and bone

Fig. 6. Anti-CII IgG antibodies in the sera of mice with CIA, of mice injected with 1 × 104 Candida cells into the paw and of mice with CIA and injected with 1 × 104 Candida Sera were collected at day 60 after the yeast injection and tested at 1:100 dilution (n = 8–9). ∗p < 0.05 estimated by unpaired t test vs the group with CIA.

tissue (22). IFN-γ secretion begins early in the development of arthritis and is elevated simultaneously with the disease progression (23). Our data showed that the paw infection with the yeast increased TNF-α and IFN-γ secretion by lymph node cells. The results point that TNF-α and IFN-γ might be functional in Candida infection except as a protective factors, also as an arthritogenic stimuli. Infection into the joint of mice with established arthritis exacerbated the disease. The clinical symptoms were associated with enhanced production of anti-CII antibodies and they correlated with the histopathologic data which showed severe bone destruction established in CIA mice injected with the yeast. It is not clear whether the elevated anti-CII antibody titers are the result of cross-reactivity of yeast antigens or the joint inflammation triggered by the fungus leads to an increased liberation of self-antigens. The main problem in septic and reactive arthritis is that the predisposed individuals could not be determined before the disease development. The use of animal models gives the opportunity to investigate the crucial factors for initiating the disease and also to look for new therapeutic approaches. The findings presented provide an insight into what happens during the course of septic arthritis and indicate that chronic Candida infection may underlie the reactive arthritis in predisposed individuals. The etiology of arthritis cannot be summarized to a unique mechanism and most probably, a variety of pathogen’s factors provoke inflammation resulting in similar joint pathology.

Acknowledgment—This work was supported by a Grant L-1208 by the National Fund for Scientific Research (Bulgaria).

132 REFERENCES 1. Thompson, S. J. and C. J. Elson. 1993. Susceptibility to pristaneinduced arthritis is altered with changes in bowel flora. Imunol. Lett. 36:227–231. 2. Kato, I., K. Endo-Tanaka, and T. Yokokura. 1998. Suppressive effects of the oral administration of Lactobacillus casei on type II collagen-induced arthritis in DBA/1 mice. Life Sci. 63:635–644. 3. Sieper, J., J. Braun, and G. H. Kingsley. 2000. Report on the Fourth International Workshop on Reactive Arthritis. Arthritis Rheum. 43:720–734. 4. Turgut, B., O. Vural, M. Demir, and M. Kaldir. 2002. Candida arthritis in a patient with chronic myelogenous leukemia (CML) in blastic transformation, unresponsive to fluconazole, but treated effectively with liposomal amphotericin B. Ann. Hematol. 81:529– 531. 5. Costalonga, M., J. S. Hodges, and M. C. Herzberg. 2002. Streptococcus sanguis modulates type-II collagen-induced arthritis in DBA/1J mice. J. Immunol. 169:2189–2195. 6. Montagnoli, C., S. Bozza, A. Bacci, R. Gaziano, P. Mosci, J. Morschhauser, L. Pitzurra, M. Kopf, J. Cutler, and L. Romani. 2003. A role for antibodies in the generation of memory antifungal immunity. Eur. J. Immunol. 33:1193–1204. 7. Romani, L. 1999. Immunity to Candida albicans: Th1, Th2 cells and beyond. Curr. Opin. Microbiol. 2:363–367. 8. Feldmann, M., F. Brennan, and R. Maini. 1996. Role of cytokines in rheumatoid arthritis. Annu. Rev. Immunol. 14:397–440. 9. Nakamura, Y., S. Noda, T. Masuhara, S. Ito-Kuwa, K. Nakamura, and S. Aoki. 1993. Radiographic features of experimental Candida arthritis in rats. Mycopathologia 121:65–75. 10. Ivanovska, N., M. Yordanov, and V. Raykovska. 2003. Single immunization of newborn mice with heterologous type-II collagen induces arthritic disease. Autoimmunity 36:205–210. 11. Wahyuningsih, R., H. J. Freisleben, H. G. Sonntag, and P. Schnitzler. 2000. Simple and rapid detection of Candida albicans DNA in serum by PCR for diagnosis of invasive candidiasis. J. Clin. Microbiol. 38:3016–3021. 12. Cuellar, M. L., L. H. Silveira, and L. Espinoza. 1992. Fungal arthritis. Ann. Rheum. Dis. 51:690–697.

Yordanov, Danova, and Ivanovska 13. Del Val Del Amo, N., E. C. Arruabarrena, R. Blanco, A. V. Pinillos, T. V. Martinez, and V. V. Rodriguez. 1997. Study of 112 patients with septic arthritis caused by pyogenic organisms and fungi: Changes in the clinical spectrum during the last 2 decades. Rev. Clin. Esp. 197:540–544. 14. Harris, M., G. Pereira, M. Myers, A. Cardin, B. Razdan, J. Pleasure, and L. Bell. 2000. Candidal arthritis in infants previously treated for systemic candidiasis during the newborn period: Report of three cases. Pediatr. Emerg. Care 16:249–251. 15. Merrer, J., B. Dupont, A. Nieszkowska, B. De Jonghe, and H. Outin. 2001. Candida albicans prosthetic arthritis treated with fluconazole alone. J. Infect. 42:208–209. 16. Arancia, G., A. Stringaro, P. Crateri, A. Torosantucci, C. Ramoni, F. Urbani, C. Ausiello, and A. Cassone. 1998. Interaction between human interleukin-2-activated natural killer cells and heat-killed germ tube forms of Candida albicans. Cell. Immunol. 186:28–38. 17. Cassone, A., M. Boccanera, D. Adriani, G. Santoni, and F. De Bernardis. 1995. Rats clearing a vaginal infection by Candida albicans acquire specific, antibody-mediated resistance to vaginal reinfection.Infect. Immun. 63:2619–2624. 18. Deng, G., I. M. Nilsson, M. Verderngh, L. Collins, and A. Tarkowski. 1999. Intra-articularly localized bacterial DNA containing CpG motifs induces arthritis. Nat. Med. 5:702–705. 19. Nicholson, L. B. and V. K. Kuchroo. 1996. Manupulation of the Th1/Th2 balance in autoimmune disease. Curr. Opin. Immunol. 8:837–842. 20. Mauri, C., R. O. Williams, M. Walmsley, and M. Feldmann. 1996. Relationship between Th1/Th2 cytokine patterns and the arthritogenic response in collagen-induced artritis. Eur. J. Immunol. 26:1511–1518. 21. Park, S. H., D. J. Min, M. L. Cho, W. U. Kim, J. Youn, W. Park, C. S. Cho, and H. Y. Kim. 2001. Shift toward T helper 1 cytokines by type II collagen-reactive T cells in patients with rheumatoid arthritis. Arth. Rheum. 44:561–569. 22. Firestein, G. 1999. Starving the synovium: Angiogenesis and inflammation in rheumatoid arthritis. J. Clin. Invest. 103:3–4. 23. Manoury-Schwartz, B., G. Chiocchia, N. Bessis, O. Abehsira-Amar, F. Batteux, S. Muller, S. Huang, M. C. Boissier, and C. Fournier. 1997. High susceptibility to collagen-induced arthritis in mice lacking IFN-gamma receptors. J. Immunol. 158:5501–5506.