Rheumatoid arthritis (RA) is a common arthritic disorder affecting millions of people throughout the world with women 2-3 times more likely to be affected than men. The prevalence of RA was estimated in the region of 1.3 to 2.1 million [1]. RA could have a considerable impact on the social, psychological [2] and economic [3] conditions of the patients. RA could be associated with extra-articular manifestations and other co-morbidities which might have a negative impact on the patient’s quality of life [4]. A general scientific consensus suggests that RA is an immune-mediated disease that could be triggered by an environmental (mainly microbial) factor in a genetically susceptible individual.

In an important report Peter Stastny from Dallas, showed that some 70% of RA patients belonged to the HLA-DR4 group whilst its frequency in the general population was only about 30% [5]. This result was confirmed by many other centres throughout the world [6]. However, the roles of other, but less predominant genes have not been ruled out [7].

The concordance rate in monozygotic twins was estimated to be 15% among English [8], 12.3% Finnish [9] and 9.1% Danish [10] populations, thereby suggesting the involvement of environmental factors.

The mechanism behind the development of RA could be due to molecular mimicry between an external agent and the HLA-DR4 group, in the same way as there is an association between streptococcal tonsillitis and rheumatic fever-associated conditions such as Sydenham’s chorea, arthritis and carditis [11].

Serum was obtained from a rabbit prior to immunisation to be used as a control sample. The rabbit was then immunised with lymphocytes obtained from a patient with severe RA, whose HLA-DR tissue type was DR4,X. Serum samples were tested against soluble extracts from 18 different bacteria (Table 1).

Table 1: Microorganisms used in the immunodiffusion studies. (QEC = Queen Elizabeth College collection, STHC = St. Stephen’s Hospital collection)

Rabbit antiserum gave 5 precipitin lines against Enterobacter cloacae, Salmonella typhimurium, Alcaligenes faecalis, as well as against Proteus mirabilis and Proteus vulgaris [12]. However, the pre-immunisation serum also produced precipitin lines against the first three microbial sonicates, namely Enterobacter cloacae, Salmonella typhimurium and Alcaligenes faecalis.

The “HLA-DR4 immunised” rabbit serum produced precipitin lines against Proteus bacteria, microorganisms which are the second commonest cause of urinary tract infections after Escherichia coli, especially in the upper urinary tract. It became apparent that a urinary tract infection could readily explain the preponderance of RA in women.

Proteus bacteria are ubiquitous in nature, they are found in soil, on vegetables, in water, sewage, mammalian gut and vagina [13]. The suggestion arises that Proteus bacteria may be involved in RA and if this hypothesis is correct then antibodies to this microbe should be demonstrable in RA patients.

Various independent groups from 15 different locations worldwide have been able to show that antibodies against Proteus mirabilis antigens were higher among RA patients when compared to the corresponding patients with other diseases or healthy controls from the same populations (Table 2).

Table 2: Geographical locations where patients with rheumatoid arthritis (RA) have shown significant elevations in antibodies against Proteus or cross-reactive antigens when compared to the corresponding controls of healthy subjects or non-RA patients *Controls are either healthy individuals or non-RA patients. **P values are referred to statistical results of the levels of antibodies against Proteus or cross-reactive antigens in RA patients compared to controls

England

Serum samples were obtained from 30 RA patients attending the weekly “Gold Clinic” at the Middlesex Hospital in London. Their mean age was 59 years (Range: 40-78 years) and the diagnosis was made according to the American Rheumatism Association criteria [14]. Antibody titres to Proteus mirabilis were measured by a Coombs agglutination assay [12]. The mean Proteus mirabilis agglutination titre in the RA patients was found to be significantly higher than the level found in controls (p<0.001).

A group of 31 English RA patients attending the Lister Hospital in Stevenage were also studied, and antibodies were measured. The RA patients with active disease were also found to have significantly elevated levels of IgG Proteus antibodies when compared to the level in English control subjects (p<0.001), but there was no elevation of antibodies to Escherichia coli [15].

Ireland

Active RA patients attending the Rheumatology Clinic of St. Vincents Hospital in Dublin, Ireland were investigated for the presence of antibodies to Proteus mirabilis and compared to patients with coeliac disease, systemic lupus erythematosus, or sarcoidosis, as well as healthy controls. Antibody levels were measured by ELISA. The levels of antibodies to Proteus in 29 RA patients were significantly elevated (p<0.001) when compared to healthy blood donors as well as when compared to the 3 disease groups: coeliac disease, sarcoidosis, and systemic lupus erythematosus patients [16].

Scotland

Serum samples tested for anti-bacterial antibodies, which have been collected from 100 rheumatoid factor-positive patients with various autoimmune diseases and joint pains, were compared to sera from 100 rheumatoid factor-negative patients attending the Ninewells Hospital in Dundee. The results were shown that RF-positive sera had significantly higher levels of IgM antibodies to P. mirabilis when compared to those of the RF-negative sera (p<0.0001). This response was subsequently found to be associated with sera from patients who clinically had RA. Furthermore, RA patients sera showed to have a significantly higher (p<0.02) IgM antibody levels to P. mirabilis than to other organisms tested, including Escherichia coli, Klebsiella pneumoniae, and Pseudomonas aeruginosa [17]. The IgM response to RA patients was not specific for only one 0 serotype of P. mirabilis but was associated with all eleven 0 serotypes tested. This antibody response was also found to be independent from C-reactive protein and RF.

France

Sera from 50 tissue-typed, active RA patients and 49 healthy controls were obtained from Brest in France. These sera were tested for antibodies against 3 different bacteria, Proteus mirabilis, Escherichia coli and Salmonella typhimurium [18]. The active RA patients from Brest were observed to have elevated levels of IgG Proteus antibodies when compared to the level in control subjects (p<0.001). There was no elevation in IgG antibody titres against Escherichia coli or Salmonella typhimurium. Sera were also collected from 15 patients with active RA in Toulouse. Serum anti-Proteus antibodies in 15 RA patients from Toulouse were measured by indirect immunofluorescence. The mean (± standard error) anti-Proteus antibody titre in the RA patients from Toulouse was 6.30 ± 0.34 and this difference was again statistically significant when compared to the blood donors from Brest (p<0.001) [18].

Netherlands

Antibody studies were carried on Dutch RA and ankylosing spondylitis patients with and without acute anterior uveitis. There were 25 active RA patients whose mean ESR was 56 mm/hour. They showed significant elevations in IgG antibody titre against Proteus when measured by immunofluorescence, compared to either HLA-B27 positive controls (p<0.001) or to HLA-B27 negative controls (p<0.001) [19]. HLA-B27 positive patients with active ankylosing spondylitis showed the highest levels of IgA antibodies to Klebsiella pneumoniae and this was significantly higher than in HLA-B27 positive healthy controls (p<0.001) or HLA-B27 negative healthy controls (p<0.001)or in RA patients (p<0.001). RA patients had antibodies to Proteus, whilst ankylosing spondylitis patients had antibodies to Klebsiella, thereby each disease group was a specificity control for the other condition.

Norway and Spain

In one study, serum samples from RA patients in three different locations, Barcelona in Spain, Oslo in Norway and Epsom in England were examined to determine the levels of anti-Proteus IgG antibodies compared to the corresponding healthy control subjects using indirect immunofluorescence technique: 34 Spanish RA patients recruited from Barcelona, 53 Norwegian RA patients from Oslo, while the third group consisted of 27 English RA patients from Epsom. The results showed that active RA patients had a significantly higher level of anti-Proteus antibodies than their corresponding healthy controls among the Spanish (p<0.001), Norwegian (p<0.001) or English populations (p<0.001) [20].

Finland

There were 129 sera obtained from a Finnish population consisting of 72 patients with “Early RA” and 27 patients with “Advanced RA” attending the Second Department of the Medicine of the University of Helsinki. “Early RA” was defined as a disease that was present for less than 12 months and “Advanced RA” was a disease present in a patient for more than 12 months. Antibodies were measured using indirect immunofluorescence against Proteus mirabilis, Serratia marcescens and Escherichia coli. Significantly elevated levels of Proteus IgG antibodies were found in both the “Early RA” (p<0.001) and “Advanced RA” (p<0.001) compared to healthy Finnish controls [21]. Similar elevations were found in the Proteus IgM antibodies in both the “Early RA” (p<0.02) and “Advanced RA” (p<0.001) compared to the corresponding healthy controls, but no such elevations were found in the IgA class. No significant immunoglobulin elevations were found against Serratia marcescens or Escherichia coli.

Japan

Thirty patients with active RA were recruited from the Otsu region in Japan to be examined for determination of antibodies against different microbes. ELISA studies were carried out against Proteus mirabilis, Escherichia coli and Klebsiella pneumoniae. Patients with active and probably active RA showed elevated levels of IgG antibodies against Proteus mirabilis when compared to controls (p<0.001) [22]. Active RA patients also had elevated levels of IgM antibodies against Proteus mirabilis when compared to controls (p<0.001). There were no IgG, IgA or IgM antibody elevations in active or probably active RA patients when tested against Klebsiella pneumoniae or Escherichia coli.

Russia

In a study from Moscow 27 patients with RA was examined for the levels of antibodies against Proteus mirabilis and compared to 100 healthy control subjects from the same area. Significant elevations of anti-Proteus IgM and IgG antibodies were observed (p<0.0001) in patients with RA compared to healthy subjects [23]. In another study from the same location antibody levels were examined in the sera of 42 Russian RA patients against Proteus mirabilis and two serotypes (03 and 09) of Yersinia enterocolitica. It was found that increased IgG antibody levels were noticed in sera of RA patients when tested against both Yersinia serotypes, especially 03, whilst it was observed that IgM isotypic antibodies against Proteus microbes were more significantly increased among the RA patients group [24]. Increased levels of antibodies against both microbes in patients with RA could be possibly due to the cross-reactive antigens shared between these microbes.

India

Seventy patients with active or inactive RA from Chandigarh in India were studied for determination of the levels of antibodies against Proteus and Salmonella microbial antigens. Other groups in the study that have been recruited from the same area included 18 patients with osteoarthritis and 82 healthy control subjects [25].

Antibody titres against Proteus antigens were found to be significantly higher in active patients with RA when compared to patients with osteoarthritis or to healthy controls (p<0.001). However there was no significant difference in anti-Salmonella antibody levels among various disease and control groups.

Taiwan

Serum samples from 39 patients with RA compared to 52 patients with ankylosing spondylitis attending the China Medical College Hospital in Taichung and 51 corresponding healthy controls were examined for class-specific antibodies against Proteus mirabilis, Escherichia coli, Campylobacter jejuni, Klebsiella pneumoniae, Yersinia enterocolitica, Salmonella typimurium and enteritidis microbes [26].

Although RA patients showed elevations of IgA isotypic antibodies against all bacterial strains, mainly IgG class antibody levels against P. mirabilis (p<0.001) and K. pneumoniae (p<0.01) were found to be significantly elevated when compared to healthy subjects.

Bermuda

Sera were obtained from 34 patients with active RA in Bermuda having erythrocyte sedimentation rates of more than 15 mm/hour. Samples were also obtained from 33 healthy controls. Antibodies were measured by ELISA against Proteus mirabilis and Escherichia coli [15]. The Bermudian RA patients with active disease had an elevated mean level of IgG Proteus antibodies when compared to the level in Bermudian control subjects (p<0.001). However, there was no elevation in antibodies to Escherichia coli.

USA and Canada

A multicenter cohort study was carried out on 246 patients with early inflammatory arthritis (<1 year) attending health centers at Montreal or Winnipeg in Canada and the VA Medical Centre in Philadelphia. Among these patients 113 were identified to have RA, 75 of who were rheumatoid factor-positives. Other groups included in the study were 43 patients with spondylarthropathy and 90 patients with undifferentiated arthritis [27].

The levels of IgM and IgA antibodies against Proteus mirabilis were found to be significantly higher mainly in rheumatoid factor-positive RA patients group when compared to patients with spondylarthropathy (p<0.0005) and undifferentiated arthritis (p<0.005).

The microbe Proteus mirabilis accounts for only about 10% of all urinary tract infections [28]. However, in some studies the interesting observation was made that no elevation was found in antibodies against the commonest microbe causing about 80% of urinary tract infections namely, Escherichia coli [29]. The elevation of antibodies only against Proteus mirabilis indicated that such immune responses were specific and could not be ascribed to non-specific urinary tract infections occurring in patients whose manual dexterity had been compromised by arthritis or age. Thus, a question arises whether patients with RA have evidence of infections with Proteus mirabilis?

Increased incidence of urinary tract infections in RA have been reported in two independent studies from Edinburgh [30] and Tel Aviv [31]. These findings suggest that there is a possibility for an existing hidden infection expressed in the form of asymptomatic bacteriuria involving Proteus microbes in patients with RA [32].

Therefore, an investigation was carried out to determine whether the urine of patients with active RA contained Proteus bacteria. In a controlled study of 89 RA patients from London, P. mirabilis was isolated more significantly in females (63%) and males (50%) in comparison to age and sex-matched healthy control females (32%) (p<0.001) and males (11%) (p<0.001). However, no significant increase in the isolation rates of Proteus microbes were observed in urine of patients with other rheumatic diseases when compared to healthy control subjects [33].

In another study carried out by a group from Dundee in the U.K., a significantly increased isolation rate of Proteus microbes from the urine of 76 patients (33%) were detected when compared to those of 48 gender-matched healthy individuals (4%) [34].

The association between RA and some sub-types of HLA-DR4 is well established. In Caucasian subjects DR4/Dw4 and DR4/Dw14 subtypes are associated with RA, whereas in Japanese subjects it is DR4/Dw15 that is the susceptibility factor [35]. In Israel where DR4/Dw10 predominates, an association with HLA-DR1 has been reported in patients with RA [36].

Analysis with synthetic oligonucleotides has shown that a particular region of the DRβ1 chain, from positions 70-74 coding for amino acids Gln-Arg-Arg-Ala-Ala (QRRAA), specific for DR1, Dw14 and Dw15, showed a strong association with RA compared to control subjects [37]. The sequence closely resembles that found in DRB1*0401 (DR4/Dw4) individuals, there being only one conservative substitution at position 71, from arginine to lysine (QKRAA). These two amino acids are positively charged and thus the overall shape and charge configuration of these two sequences are similar. This sequence Q(K)RRAA has been described as the “shared epitope” by the Winchester group from New York [38], and this link between RA and “shared epitope” has been confirmed by various other groups over the last two decades [39]. The glutamic acid (E) occupying position 69 is common to all DRβ1 molecules. Furthermore the EQRRAA sequence is also found in DRB1*1402 (DR6/DW16) positive Yakima Indians affected by RA. The RA susceptibility sequence spanning residues 69-74 (EQRRAA) was used to scan published sequences of molecules from Proteus microorganisms. A closely related sequence (ESRRAL) spanning residues 32-37 of the surface membrane haemolysin of Proteus mirabilis (Hpm B polypeptide) was identified which had biochemical and charge similarity to the susceptibility sequence (Figure 1).

Figure 1: Schematic picture depicting molecular similarities between Proteus mirabilis and self-antigens

Cartilage destruction is a major feature of the disease. A search of the protein database was made for any sequence of Proteus mirabilis showing structural similarity to any collagens. One way of distinguishing urine cultures is by determining whether the bacterial colonies are either urease positive or urease negative. Urease is not present in Escherichia coli but is present in Proteus bacteria. Thus the hypothetical question arose if the Proteus urease molecule had any similarity to collagens known to be present in joint tissues.

A computer analysis showed an amino acid homology between Proteus mirabilis urease (IRRET), amino acid residues 337-341 and α2(XI) collagen (LRREI) residues 421-425 (Figure 1) [40,41]. Type XI collagen is a component of hyaline cartilage which is composed of three different polypeptide subunits α1(XI), α2(XI) and α3(XI). This suggests that antibodies to these two Proteus cross-reactive self-components may be involved in the onset of RA.

In an endeavour to assess the cytotoxic activity of antibodies sheep red blood cells were coated with 15-16 mer peptides containing either the (EQRRAA) or (LRREI) sequences, which are homologous to antigens found in Proteus haemolysins or ureases respectively [42].

It was observed that the percentage lysis of sera from RA patients against cells coated with (EQRRAA) peptides was significantly higher when compared to sera from patients with ankylosing spondylitis (p<0.0001) or healthy controls (p<0.0001) (Figure 2). Similar results were also found when RA patient’s sera have been tested against cells coated with (LRREI) peptides.

Figure 2: Percentage lysis of sheep red blood cells coated with HLA-DRB1*0404 (EQRRAA), DRB1*0402 (EDERAA – an HLA sequence not linked with RA) and type XI collagen (LRREI) peptides via complement mediated cytotoxicity by control sera, and sera from ankylosing spondylitis (AS) and RA patients. (With kind permission from Wilson et al. 2003)

Antibodies to the urinary microbe Proteus mirabilis have been demonstrated to be present in RA patients from several countries. Molecular mimicry between Proteus haemolysin and HLA-DR4 and between Proteus urease and hyaline cartilage, provide a possible explanation for the pathogenesis of the disease. Moreover, anti-Proteus antibodies have cytopathic properties against joint components. Since several studies have shown that antibiotics, such as minocycline [43], sulfasalazine, clarithromycin and levofloxacin [44-47] are effective in the treatment of RA, the question arises whether multi-centre studies should be carried out using Proteus-sensitive antibiotics together with a high fluid intake as well as the currently used medical treatments including biologicals, to assess their therapeutic effectiveness in this disease.

This study was supported by the Trustees of the Middlesex Hospital and the “American Friends of King’s College London”.

1. Gibofsky A (2012) Overview of epidemiology, pathophysiology, and diagnosis of rheumatoid arthritis. Am J Manag Care 18: S295-302.
2. Benka J, Nagyova I, Rosenberger J, Calfova A, Macejova Z, et al. (2014) Social support as a moderator of functional disability's effect on depressive feelings in early rheumatoid arthritis: a four-year prospective study. RehabilPsychol 59: 19-26.
3. Huscher D, Mittendorf T, von Hinüber U, Kötter I, Hoese G, et al. (2014) Evolution of cost structures in rheumatoid arthritis over the past decade. Ann Rheum Dis .
4. Cutolo M, Kitas GD, van Riel PL (2014) Burden of disease in treated rheumatoid arthritis patients: going beyond the joint. Semin Arthritis Rheum 43: 479-488.
5. Stastny P (1978) Association of the B-cell alloantigen DRw4 with rheumatoid arthritis. N Engl J Med 298: 869-871.
6. Julià A1, Marsal S (2013) The genetic architecture of rheumatoid arthritis: from susceptibility to clinical subphenotype associations. Curr Top Med Chem 13: 720-731.
7. Korczowska I (2014) Rheumatoid arthritis susceptibility genes: An overview. World J Orthop 5: 544-549.
8. Silman AJ, MacGregor AJ, Thomson W, Holligan S, Carthy D, et al. (1993) Twin concordance rates for rheumatoid arthritis: results from a nationwide study. Br J Rheumatol 32: 903-907.
9. Aho K, Koskenvuo M, Tuominen J, Kaprio J (1986) Occurrence of rheumatoid arthritis in a nationwide series of twins. J Rheumatol 13: 899-902.
10. Svendsen AJ, Kyvik KO, Houen G, Junker P, Christensen K, et al. (2013) On the origin of rheumatoid arthritis: the impact of environment and genes--a population based twin study. PLoS One 8: e57304.
11. Burke RJ, Chang C (2014) Diagnostic criteria of acute rheumatic fever. Autoimmun Rev 13: 503-507.
12. Ebringer A, Ptaszynska T, Corbett M, Wilson C, Macafee Y, et al. (1985) Antibodies to proteus in rheumatoid arthritis. Lancet 2: 305-307.
13. Armbruster CE, Mobley HL (2012) Merging mythology and morphology: the multifaceted lifestyle of Proteus mirabilis. Nat Rev Microbiol 10: 743-754.
14. Ropes Mw, Bennett Ga, Cobb S, Jacox R, Jessar Ra (1958) 1958 Revision of diagnostic criteria for rheumatoid arthritis. Bull Rheum Dis 9: 175-176.
15. Subair H, Tiwana H, Fielder M, Binder A, Cunningham K, et al. (1995) Elevation in anti-Proteus antibodies in patients with rheumatoid arthritis from Bermuda and England. J Rheumatol 22: 1825-1828.
16. Rogers P, Hassan J, Bresnihan B, Feighery C, Whelan A (1988) Antibodies to Proteus in rheumatoid arthritis. Br J Rheumatol 27 Suppl 2: 90-94.
17. Senior BW, McBride PD, Morley KD, Kerr MA (1995) The detection of raised levels of IgM to Proteus mirabilis in sera from patients with rheumatoid arthritis. J Med Microbiol 43: 176-184.
18. Fielder M, Tiwana H, Youinou P, Le Goff P, Deonarain R, et al. (1995) The specificity of the anti-Proteus antibody response in tissue-typed rheumatoid arthritis (RA) patients from Brest. RheumatolInt 15: 79-82.
19. Blankenberg-Sprenkels SH, Fielder M, Feltkamp TE, Tiwana H, Wilson C, et al. (1998) Antibodies to Klebsiellapneumoniae in Dutch patients with ankylosing spondylitis and acute anterior uveitis and to Proteus mirabilis in rheumatoid arthritis. J Rheumatol 25: 743-747.
20. Rashid T, Darlington G, Kjeldsen-Kragh J, Forre O, Collado A, et al. (1999) Proteus IgG antibodies and C-reactive protein in English, Norwegian and Spanish patients with rheumatoid arthritis. ClinRheumatol 18: 190-195.
21. Rashid T Leirisalo-Repo M, Tani Y, Hukuda S, Kobayashi S, et al. (2004) Antibacterial and antipeptide antibodies in Japanese and Finnish patients with rheumatoid arthritis. ClinRheumatol 23: 134-141.
22. Tani Y, Tiwana H, Hukuda S, Nishioka J, Fielder M, et al. (1997) Antibodies to Klebsiella, Proteus, and HLA-B27 peptides in Japanese patients with ankylosing spondylitis and rheumatoid arthritis. J Rheumatol 24: 109-114.
23. Ushakova MA, Muravijuv UV, Vaneeva NP, Vaneeva NP, Jastrebova NE, et al. (2000) The Proteus mirabilis antibody levels in patients with rheumatoid arthritis. Rheumatologia 14: 78.
24. Maslova II, Khorobrykh NE, Ushakova MA, Ovechko NN, Murav'evIuV (2004) [Antibodies to Yersinia enterocolitica and Proteus mirabilis in blood sera of rheumatoid arthritis patients]. ZhMikrobiolEpidemiolImmunobiol : 71-72.
25. Wanchu A, Deodhar SD, Sharma M, Gupta V, Bambery P, et al. (1997) Elevated levels of anti-proteus antibodies in patients with active rheumatoid arthritis. Indian J Med Res 105: 39-42.
26. Chou CT, Uksila J, Toivanen P (1998) Enterobacterial antibodies in Chinese patients with rheumatoid arthritis and ankylosing spondylitis. ClinExpRheumatol 16: 161-164.
27. Newkirk MM, Goldbach-Mansky R, Senior BW, Klippel J, Schumacher HR Jr, et al. (2005) Elevated levels of IgM and IgA antibodies to Proteus mirabilis and IgM antibodies to Escherichia coli are associated with early rheumatoid factor (RF)-positive rheumatoid arthritis. Rheumatology (Oxford) 44: 1433-1441.
28. Fairley KF, Carson NE, Gutch RC, Leighton P, Grounds AD, et al. (1971) Site of infection in acute urinary-tract infection in general practice. Lancet 2: 615-618.
29. Rashid T, Ebringer A (2007) Rheumatoid arthritis is linked to Proteus--the evidence. ClinRheumatol 26: 1036-1043.
30. Lawson AA, Maclean N (1966) Renal disease and drug therapy in rheumatoid arthritis. Ann Rheum Dis 25: 441-449.
31. Tishler M, Caspi D, Almog Y, Segal R, Yaron M (1992) Increased incidence of urinary tract infection in patients with rheumatoid arthritis and secondary Sjögren's syndrome. Ann Rheum Dis 51: 604-606.
32. Rashid T, Ebringer A (2011) Rheumatoid arthritis is caused by asymptomatic Proteus urinary tract infections. Clinical Management of Complicated Urinary Tract Infections 171-180.
33. Ebringer A, Wilson C, Ahmadi K, Corbett M, Rashid T, et al. (1993) Rheumatoid arthritis as a reactive arthritis to Proteus infection: Prospects for therapy. In: Machtey I (editor): Sixth International Seminar on the Treatment of Rheumatic Diseases”. ProgRheumatol 5: 77-83.
34. Senior BW, Anderson GA, Morley KD, Kerr MA (1999) Evidence that patients with rheumatoid arthritis have asymptomatic ‘non-significant’ Proteus mirabilis bacteriuria more frequently than healthy controls. J Infect 38: 99-106.
35. Maeda H, Juji T, Mitsui H, Sonozaki H, Okitsu K (1981) HLA DR4 and rheumatoid arthritis in Japanese people. Ann Rheum Dis 40: 299-302.
36. Schiff B, Mizrachi Y, Orgad S, Yaron M, Gazit E (1982) Association of HLA-Aw31 and HLA-DR1 with adult rheumatoid arthritis. Ann Rheum Dis 41: 403-404.
37. Watanabe Y, Tokunaga K, Matsuki K, Takeuchi F, Matsuta K, et al. (1989) Putative amino acid sequence of HLA-DRB chain contributing to rheumatoid arthritis susceptibility. J Exp Med 169: 2263-2268.
38. Gregersen PK, Silver J, Winchester RJ (1987) The shared epitope hypothesis. An approach to understanding the molecular genetics of susceptibility to rheumatoid arthritis. Arthritis Rheum 30: 1205-1213.
39. Holoshitz J (2010) The rheumatoid arthritis HLA-DRB1 shared epitope. CurrOpinRheumatol 22: 293-298.
40. Kimura T, Cheah KS, Chan SD, Lui VC, Mattei MG, et al. (1989) The human alpha 2(XI) collagen (COL11A2) chain. Molecular cloning of cDNA and genomic DNA reveals characteristics of a fibrillar collagen with differences in genomic organization. J BiolChem 264: 13910-13916.
41. Wilson C, Ebringer A, Ahmadi K, Wrigglesworth J, Tiwana H, et al. (1995) Shared amino acid sequences between major histocompatibility complex class II glycoproteins, type XI collagen and Proteus mirabilis in rheumatoid arthritis. Ann Rheum Dis 54: 216-220.
42. Wilson C, Rashid T, Tiwana H, Beyan H, Hughes L, et al. (2003) Cytotoxicity responses to peptide antigens in rheumatoid arthritis and ankylosing spondylitis. J Rheumatol 30: 972-978.
43. O'Dell JR, Paulsen G, Haire CE, Blakely K, Palmer W, et al. (1999) Treatment of early seropositive rheumatoid arthritis with minocycline: four-year followup of a double-blind, placebo-controlled trial. Arthritis Rheum 42: 1691-1695.
44. Ogrendik M (2013) Antibiotics for the treatment of rheumatoid arthritis. Int J Gen Med 7: 43-47.
45. Tiwana H, Wilson C, Walmsley RS, Wakefield AJ, Smith MS, et al. (1997) Antibody responses to gut bacteria in ankylosing spondylitis, rheumatoid arthritis, Crohn's disease and ulcerative colitis. RheumatolInt 17: 11-16.
46. Deighton CM, Gray JW, Bint AJ, Walker DJ (1992) Anti-Proteus antibodies in rheumatoid arthritis same-sexed sibships. Br J Rheumatol 31: 241-245.
47. Tiwana H, Wilson C, Cunningham P, Binder A, Ebringer A (1996) Antibodies to four gram-negative bacteria in rheumatoid arthritis which share sequences with the rheumatoid arthritis susceptibility motif. Br J Rheumatol 35: 592-594.