New ZealandNorwayPakistanPeruPolandPeurto RicoRomaniaRussiaSingaporeSouth AfricaSudanSwedenThailandTurkeyUnited KingdomUSAUzbekistan

Ankylosing Spondylitis as an Auto-immune Disease Caused by Intestinal Klebsiella Infection: Prospects for a New Therapeutic Approach

A. Ebringer 1'2, T. Rashid1, and C. Wilson1

'Division of Life Sciences, Infection and Immunity Group, Waterloo Campus, King's College London; ? Department of Rheumatology, Middlesex Hospital, UCL School of Medicine, London,

Correspondence to: Professor Alan Ebringer Division of Life Sciences, Infection and Immunity Group, King's College London, Stamford Strecl, London SE1

Tel: 020-78484302 Fax: 020-7848 4500 Email:

Key words: Ankylosing spondylitis, AS and Klebsiella reactive arthritis, Klebsiella, HLA-B27

Ankylosing spondylitis (AS) is the most prominent component of a group of inflammatory arthritides, collectively named as "spondyloarthropathies (SpAs)", It is a chronic disease of insidious onset starting usually in males between ages of 10 and 40 years. It commonly involves the sacroiliac and lower spinal joints and less frequently other large joints such as the hips, knees and ankles as well as the entheses. Non-articular manifestations may also be associated with the disease, such as acute anterior uveitis or iritis, which occurs in 25 percent of patients with AS [1]. The true prevalence of AS appears to be in the region of 0.25 to l% with a peak of 2% in Northern Norway [2],

The role of genetic and environmental factors in AS
The association between AS and HLA-B27 was not the first HLA association to be defined but it remains to this day, the most prominent example linking susceptibility to a rheumatic disease with genes in the major histocompatibility complex. This remarkable association which was reported over 28 years ago [3, 4] has been shown to be universal in world populations. The frequency of HLA-B27 in patients with AS ranges from 81 to 96% with control frequencies ranging between 4 and 12%. Rats transgenic for B27 spontaneously develop a chronic inflammatory disease that resembles, both clinically and histologically, the human SpA, while control rats transgenic for HLA-A2 do not develop such an illness [5]. These observations would signify the importance of this particular HLA gene in the development of AS.
The prevalence of AS correlates with the presence of HLA-B27 in all ethnic groups studied. For example, African Blacks of unmixed ancestry lack B27 genes and AS is very rare among them. On the other hand, certain North American Indian tribes who have a high frequency of B27 also have a high prevalence of AS [6]. The fact that all races show an association of AS with this genetic marker suggests that HLA-B27 is somehow involved in the disease development.

In HLA-B27 negative AS patients, an increased incidence of other B locus antigens, namely HLA-B7, -B22, -BW40, and -B42 which are collectively called the "B7 cross-reacting group" (B7-CREG) has been reported [7] and each of these antigens cross-reacts with HLA-B27 [8]. In general, HLA-B27 negative patients with AS, have a milder disease with an onset occurring in the early thirties as well as an equal sex ratio [9]. It would appear that HLA-B27 negative AS has different features from those found in B27 positive AS patients.
Some twenty subtypes of HLA-B27 (6*2701-2720) have been identified. B*2705 is the ancestral subtype from which the others have evolved, mostly from changes in exons 2 and 3, which encode the alpha 1 and alpha 2 domains of the peptide binding groove respectively. Occurrence of SpA has thus far been documented in subjects possessing any one of the first 10 subtypes, which differ, by one or more amino acids [10].

Studies of the HLA-B27 molecule by two-dimensional gel electrophoresis [11], DNA sequencing [12] and restricted fragment length polymorphism [13] have all shown no molecular differences in this genetic marker between patients with AS and healthy controls. Using HLA-B27 specific monoclonal antibodies, no disease-associated variant of HLA-B27 has been observed in AS [14]. The property mediating susceptibility to disease, therefore, appears to lie in the features of the serologically and biologically determined HLA-B27 epitopes, which are common to all of its subtypes.
Restriction fragment length polymorphism has failed to confirm the presence of any additional genetic material of relevance to disease susceptibility in AS. For example, no difference was detected in the frequencies of tumour necrosis factor between AS patients and healthy controls [15]. Differential linkage disequilibrium analysis with HLA-B27 subtypes suggest that B27 itself remains the primary gene for AS susceptibility and other closely related class I loci, are not involved in the pathogenesis of the disease [16].
In spite of this strong genetic link with AS, other non-genetic, environmental factors, particularly microorganisms, could be involved in triggering the disease;

1) Concordance rate for AS in monozygotic twins is below 50% [17], thus suggesting the intervention of an exogenous factor in conjunction with HLA-B27 necessary to trigger the onset of this disease,
2) Episodic nature of AS, with relapses and remissions weakens the role of a pure and isolated genetic factor acting alone in the development of this condition [18].

3) Transgenic rats possessing the human HLA-B27 gene and raised in germ-free environment do not develop all the features of SpAs. However, when these animals are exposed to Gram negative bacteria, such as Yersinia, then the phenotypic expression of nearly all features of SpAs particularly the arthritic lesions become evident [19]. This observation argues strongly in favour of an environmental bacterial factor acting together with a genetic predisposing factor.
4) Absence of AS in an overwhelming majority of HLA-B27 positive individuals in the general population, and in 80 percent of B27 positive healthy relatives of AS probands [10], suggests that the disease is mainly triggered by the effect of some non-genetic, environmental factor.

Klebsiella pneumoniae as the most prominent microbial factor involved in AS aetiopathogenesis
Infection has long been regarded as a likely cause of AS since evidence linking this disease with chronic prostatic infection emerged in the 1950s [20]. It was nearly twenty years later when infection with Klebsiella microorganisms was first implicated in the possible causation of AS [21]. The evidence linking Klebsiella pathogens to AS have involved a number of experimental studies:

1) Increased antibody titres against Klebsiella but not to other microbes have been detected in the sera of 1250 patients with AS from 13 different countries when compared to 1088 healthy controls (Table 1) and with a total of 446 patients having other inflammatory diseases (Table 2). These observations were demonstrated by independent groups using a number of different immunological assays such as indirect haemagglutination, radioimmunoassay, enzyme-linked immunosorbent assay, indirect immunofluorescent assay and Western immunoblotting.
2) In a study from Finland, the mean values of synovial anti-Klebsiella antibodies, particularly the IgA classes were found to be significantly higher in AS patients when compared to patients with rheumatoid arthritis [22].

3) Immunological and molecular cross-reactivity 'between HLA-B27 Or other self-antigens and Klebsiella microbial antigens have been demonstrated in several independent studies:

A) Sera from rabbits immunised with HLA-B27 positive lymphocytes reacted with the antigenic extracts of Klebsiella, Enterobacter, Salmonella, Shigella and Yersinia microorganisms indicating the existence of shared cross-reactive antigens [21,23].
B) Anti-B27 allogeneic human tissue typing sera were found to bind to Klebsiella antigens more readily than to other tissue typing sera [24],
C) Monoclonal anti-B27 antibodies were found to bind Specifically to Klebsiella, Shigella and Yersinia antigens [25].
D) An anti-HLA-B27 (M2) monoclonal antibody bound specifically to 60 and 80 Kd components of Klebsiella, whereas no such reactivity was demonstrated by five other monoclonal antibodies [26].
E) Although both Klebsiella and Yersinia microbes possess antigens cross-reacting with HLA-B27, AS patients have elevated levels of antibodies only to whole Klebsiella but not to whole Yersinia microorganisms [27].
F) A molecular homology has been demonstrated between a hexameric amino acid sequence, "QTDRED" present in both HLA-B*27Q5 molecules (residues 72-77) and Klebsiella pneumoniae nitrogenase reductase enzymes (residues 188-193). Significantly increased levels of antibodies to a synthetic peptide obtained from Klebsiella nitrogenase and containing this homologous sequence were found in AS patients sera when compared to healthy controls [28].
G) Antibodies against Klebsiella nitrogenase peptides encompassing the "QTDRED" sequences bind more significantly to the synovial tissues taken from AS patients when compared to those with other inflammatory diseases [29].
H) Rabbit anti-Klebsiella sera were able to distinguish HLA-B27 positive lymphocytes from either AS patients or healthy controls when compared to lymphocytes obtained from HLA-B27 negative individuals [30].
1) Molecular similarity has been demonstrated between the "DRDE" amino acid sequence (positions 596-599) present in the PulD secretion proteins of Klebsiella pullulanase enzymes, and the "PulD" sequence (positions 74-77) present in the HLA-B27 molecules [31], Antibody levels against each of these peptides were found to be elevated in AS patients compared to controls.
J) Gly-x-pro repeating sequences were found in the PulA components of Klebsiella pullulanase, which shows molecular similarity with type I, III and IV collagens [31].
K) Japanese AS patients had a significant elevation in the levels of IgA antibodies to the synthetic peptides encompassing "DRDE" or "DRED" amino acid sequences of the Klebsiella PulD proteins or HLA-B27 molecules when compared to patients with rheumatoid arthritis or healthy controls [32].
L) Significant elevation of IgA2 subclass of antibodies against type I, III and IV collagens were observed in Japanese AS patients when compared to healthy individuals [33].
M) AS patients were found to have increased antibody concentrations of IgG and IgA classes to KP2 components of the Klebsiella pneumoniae nitrogenase enzyme when compared to controls [34].

Gastro-intestinal tract as the main source of Klebsiella infection in AS

There is extensive evidence, to suggest an association between gut lesions in AS and presence of Klebsiella as the likely triggering agent across the gut mucosa:

1) HLA-B27 positive individuals suffering from inflammatory bowel diseases (IBD) such as Crohn's disease or ulcerative colitis, have a significantly greater chance of developing AS than HLA-B27 positive individuals not having IBD [35].
2) Patients with AS as well as those with Crohn's disease and to a lesser extent ulcerative colitis but not those with rheumatoid arthritis or healthy controls have significantly higher titres of anti-Klebsiella antibodies, indicating the possibility of a microbial link between AS and these IBDs [36].
3) In a most recent study, patients with AS and Crohn's disease were found to have elevated levels of IgM, IgG and IgA class-specific antibodies to Klebsiella pneumoniae and to collagens type I, III, IV and V. Moreover, a positive correlation was observed between Klebsiella and collagen antibody levels in these patients [37]
4) In reactive arthritis and Reiter's syndrome, an active arthropathy may follow enteropathic infection caused by Yersinia, Shigella, Salmonella and Campylobacter microorganisms [38].
5) Enhanced gut permeability [39, 40], together with macroscopic and microscopic inflammation in the bowel mucosa occurs in AS patients [41].
6) Using a 99m-technetium leucocyte scintigraphy scan 48% of HLA-B27 patients with SpAs showed bowel inflammation; all had active joint disease, whereas no patients with inactive disease had a positive uptake of labelled leucocytes [42].
7) Two longitudinal studies have indicated a temporal relationship between the acquisition of Klebsiella in the faeces and the exacerbation of symptoms in AS patients [43, 44]. Two other groups have also found an increased isolation of Klebsiella in the faecal cultures from active AS patients whether these patients were identified by high CRP levels [45], or defined clinically [46], but others could not find such an association [47]. These results could be due to different methods used for the collection and culture of the faecal specimens.
8) Elevation in the total serum IgA [48-50], total secretory IgA (sIgA) [51] as well as Klebsiella specific sIgA [52] occur in AS patients during active phases of the disease. Moreover, a significant correlation was found between total sIgA and anti-Klebsiella antibody concentrations in patients with AS from Spain [53].
9) Plasma cells in the gut mucosa which are one of the main source for total serum IgA and especially sIgA production were found to be more abundant in patients with AS than in the controls j.54].
10) Total IgA, SIgA and sIgM were all found to be significantly increased in the jejunal lavage fluid of a group of AS patients from Sweden when compared with healthy controls [55].
11) A significant association between increased IgA antibodies to Klebsiella pneumoniae and the gut inflammation was found to occur in Finnish patients with AS [56].
12) AS patients were found to have an increased jejunal production of Klebsiella antibodies when compared to rheumatoid arthritis patients or controls [57],
13) A French group more than three decades ago, has found evidence of inflammatory and sclerotic abnormalities in the pelvic and sacral lymph nodes in AS patients using lymphangiography. The lymph node changes seemed to precede the development of radiological changes in the lumbar spine and sacroiliac joints [58]. The results of these studies support the hypothesis that an important link exists between gut lesions and AS and signs of overt or more commonly occult Klebsiella intestinal infections are frequently encountered in AS.

AS and Klebsiella reactive arthritis
The world-wide consistency in the finding of increased Klebsiella antibodies in AS patients strongly support the hypothesis that exposure to this microbe occur in these patients. Epitopes expressed on the surface of bacterial molecules are more likely to be involved in the humoral immune responses than epitopes that are internalised and consist of hydrophobic sequences [59]. One of the main targeted antigens to Klebsiella antibodies are to capsular lipopolysaccharides, which constitute the main part of the outer membrane wall of these microbes [60-62]. Thus various antigenic molecules present on the surface of Klebsiella microorganisms including capsular polysaccharides or enzymes produced by these microbes such as nitrogenase reductase and pullulanase, may all be involved in triggering the disease development in AS [63].
The cross-reactive self-antigen sequence (QTDRED) is located on the outer rim of the peptide binding cleft of the HLA-B27 molecules and hence more accessible to the Klebsiella nitrogenase antibodies. The anti-Klebsiella antibodies may bind, although with lower affinity to the cross-reactive epitopes present in the nearby joint tissues such as the sacro-iliac joints, lumbar spine, sites of enthesis and even other remote tissues such as the uveal tracts [64, 65].
The anatomical location of the lesions is determined by the types of auto-antibodies produced following Klebsiella infections. There are two types of cross-reactive auto-antibodies, those cross-reacting with HLA-B27, which may attack the chondrocytes and those cross-reacting with type I, 111 and IV collagens, which are components of spinal tissues. The binding of these Klebsiella cross-reactive antibodies, when present in high titres may trigger the onset of inflammatory cascades such as, the complement system as well as various cytokines resulting in the pathological damages with consequent fibrosis and calcification occurring at these sites (Figure).
The body of cumulative evidence reviewed involving immunogenetic, microbiological and serological studies carried out independently by various groups leads to the suggestion that, AS is a form of "Klebsiella reactive arthritis" (KRA). This entity could be defined by objective tests involving three main lines of experimental findings [66]:

1) Klebsiella microbes can be isolated from the bowel of active AS patients.
2) Klebsiella antibodies can be identified in the sera of active AS patients.
3) Klebsiella cross-reacts with HLA-B27 and collagens I, III and IV self-antigens.

The concept of "KRA" could be placed in the same category as other types of "reactive arthritis" which are caused by gut or urinary bacteria, such as Yersinia, Shigella, Salmonella, Campylobacter, Ureaplasma and Chlamydia. In these conditions the laboratory diagnosis can be confirmed by finding elevated specific anti-bacterial antibodies in the sera with or without isolation of the causative microbes from the bowel or urinary tract in these patients [67], a situation that also occurs in AS patients.

The two main theoretical models to explain the aetiopathogenetic mechanism in AS are the cross reactivity or molecular mimicry theory and the receptor theory. The latter model has been extensively investigated by many research groups, but with inconsistent results, in that no ubiquitous peptide specifically binding to the HLA-B27 groove has been identified [68, 69]. Although it is generally agreed that most class I molecules bind peptides of 8-12 amino acid residues in length, it has also been reported that HLA-B27 molecules could bind peptides of up to 33 amino acids in length [70]. Furthermore, HLA-B27 molecules can form disulphide-bonded heavy chain homodiraers that could result from peptide deficiency or a defective peptide transport mechanisms [71]. The molecular mimicry theory, on the other hand, has been more clearly elucidated based on the results of many experimental studies [72, 73].
Molecular modelling of B27 in the form of disulphide-bonding through Cy67 would result in partial unwinding of the _-l helix in the peptide binding cleft [74], which could make this molecule more accessible to the Klebsiella cross-reactive antibodies.

Importance for an early diagnosis in the management of AS
It requires at least several years for the classical clinical features of AS, especially radiological evidence of sacro-iliitis to be present, before the diagnosis can be established. By the time the disease is diagnosed, the damaging pathological events have occurred and no pharmacological treatment has been demonstrated to induce remission or halt the progression of spinal fusion, and which could reverse the damaging effects at this late stage of the disease.
In Lewis rats with Yersinia-triggered reactive arthritis, the earlier the antibiotic treatment was started the better were the results [75]. This might be comparable to human diseases with a similar pathology and at the same time explain the successful effect of ciprofloxacin on patients with Yersinia reactive arthritis in whom the mean disease duration was relatively short (1.9 years) [76], and the failure of the same treatment in patients with longer disease duration (4.9 years) [77].
AS could develop as the end result of repetitive attacks of KRA. It is proposed that KRA could be defined as a condition occurring in an HLA-B27 positive individual presenting with arthritic symptoms in the spine and/or large joints and having evidence of elevated anti-Klebsiella antibodies. There are two types of KRA that could be recognized under the terms of this definition: "Early KRA" and "Advanced KRA". The taxonomic features of these two conditions are proposed to be as follows:
Early KRA for pre-AS): is said to be present in an individual, who has;

1) Arthritic symptoms in the spine and/or large joints (hips, knees, ankles) for at least three months.
2) "Morning muscle stiffness" relieved by exercise.
3) Presence of HLA-B27 genetic marker.
4) Elevated level of anti-Klebsiella antibodies during episodes of inflammation.
5) No radiological evidence of sacro-iliitis.

Advanced KRA (or classical AS)is said to be present in an individual, who has all the features of "early KRA" (condition 1-4) but also has radiological evidence of sacro-iliitis. Thus "late KRA" would correspond to the classical definition of "AS".

There still remains a small group of "classical AS" patients who are HLA-B27 negative and who fit the criteria defined by the Linssen group [9], and these could be called "HLA-B27 negative AS". Whether Klebsiella is involved in HLA-B27 negative AS is, at the moment, unclear.
It is suggested therefore, that early diagnosis of "AS" or conditions leading to AS such as "KRA", based on the proposed identification criteria, is required for the implementation of a more specific therapeutic management during early stages of the disease.

Proposal for an anti-microbial therapy in patients with AS

Currently the management of patients with AS includes two main approaches. The first one involves the use of one or more of non-steroidal anti-inflammatory drugs (NSAIDs) which are mainly used to alleviate the pain and reduce the associated inflammation. The second approach involves physiotherapy and postural education to prevent the stiffness and deformities which are the main sequelae occurring in this disease. However, the use of NSAIDs are considered as non-specific measures, as they could not reverse or even halt the damaging pathological process of the disease. Furthermore, these drugs have various undesirable side effects.
The use of other therapeutic measures to eliminate or reduce the Klebsiella microbes could, therefore, have beneficial effects on the patients with this disease. These new therapeutic measures involve two main lines, which include antibiotic therapy and dietary measures:

1) Antibiotics
Sulphasalazine (SZN) was and is still considered by many independent groups as one of the most effective and reasonably tolerated drug in the treatment of patients with AS. A 26-week, placebo-controlled trial showed that enteric coated SZN seemed to be effective and well tolerated in AS patients [78]. In a 3-year placebo-controlled trial of SZN a reduced frequency of peripheral arthritis was observed in the treated but not in the control group of AS patients [79]. A 6-month randomised, multi-centre, double-blind, placebo-controlled trial of SZN in spondyloarthritic patients...resistant to treatment with NSAIDs showed that SZN had greater efficiency than placebo [80]. Norwegian AS patients treated with SZN, showed a significant decrease in the total IgA and sIgA in the jejunal perfusion fluid when compared to healthy controls [SS]. A significant decrease in the concentration of K. pneumoniae IgA antibodies during the 26 weeks of SZN treatment has also been reported in Finnish AS patients when compared to controls [81]. Furthermore, many groups have shown that the anti-microbial component, sulphapyridine, is the active moiety of SZN [82-84]. The beneficial effects of SZN could be due to this anti-microbial actions and this may further strengthen the evidence for a role of microbes in pathogenesis of AS. In a most recent study, SZN was also found to be beneficial in preventing recurrences and reducing the severity of anterior uveitis associated with AS [85].

In patients with Yersinia reactive arthritis, a faster remission and relief of pain was found in those receiving ciprofloxacin. Yersinia microorganisms were found to be eliminated from the gut associated lymphoid tissue in six of seven patients receiving ciprofloxacin compared with none of nine patients receiving placebo and patients receiving placebo had more and prolonged circulating IgA antibodies against Yersinia antigens than patients treated with ciprofloxacin [76].

Klebsiella infections in more than 75% of cases respond to one of these antibiotics: Cephalothin, nalidixic acid, nitrofurantoin, sulfonamides, tetracycline and trimethoprim-sulfamethoxasole [86]. A search for finding other antibiotics, which might have specific and effective actions on Klebsiella microbes is clearly necessary.

2) Low starch diet:
It has been observed that the majority of AS patients, who were on a diet, low in starch and high in proteins, have claimed a drop in the severity of their symptoms as well as a reduction in their requirement for NSAID intake. This clinical improvement was found to be correlated with a decrease in the total serum IgA levels and erythrocyte sedimentation rates [87].
Starch is an essential nutrient material necessary for the growth of enterobacterial species, and Klebsiella is considered as one of the main members in this group. Finegold and co-workers have observed that subjects on low carbohydrate diets have decreased levels of Klebsiella microorganisms in their stools [88]. The substrates available in the gut for fermentation are one of the essential factors in determining the microbial population in the various regions of the gastro-intestinal tract. These substrates include simple sugars and polysaccharidss derived from starch. Klebsiella microbes do not seem to grow on plant cellulose [89]. Mean number of Klebsiella microbes for three different sugars (glucose, sucrose, and lactose) per gram of substrate was found to be significantly higher when compared to the value obtained following incubation with eleven different amino acids [90].

Therefore, the use of a diet, low in starch and high in proteins, could help in preventing the growth of these bacteria in the bowel and thus in reducing the incidence of intestinal Klebsiella infection. The sheets describing the "London AS Diet" which we have given, for the last 18 years, to our patients attending the "AS Research Clinic" at the Middlesex Hospital, are given at the end of this article (Appendix).

It is suggested that a combination of Klebsiella specific antibiotics given in short courses, especially during attacks of Klebsiella reactive arthritis or as a low-dose prophylactic measure particularly in genetically susceptible individuals with history of recurrent attacks of KRA together with a low starch diet may have a therapeutic effect on AS patients especially if started during early stages of the disease. Prospective controlled studies are required to determine the relevance of these measures in the treatment of patients with AS.

1) Klebsiella microorganisms have a direct role in the pathogenesis of AS based on the results of microbiological, serological and molecular studies carried out by different independent research groups throughout the world.
2) AS would appear to be the end result of recurrent episodes of "Klebsiella reactive arthritis".
3) The disease process could possibly be halted by identifications of early cases and treatment with anti -Klebsiella therapy, involving low starch diet, SZN and antibiotics.
4) Prospective controlled studies are required to evaluate anti-Klebsiella therapy in the management of AS.


We thank the Trustees of the Middlesex Hospital, the Arthritis Research Campaign (Grant EO514) and "American Friends of King's College London" for their support.


1. Rosenbaum JT. Acute anterior uveitis and spondyloarthropathies. Rheum Dis Clin NAml992;18:143-51.

2. Calin A. Ankylosing spondylitis. In: Maddison PJ, Isenberg DA, Woo P, and Glass
DN, eds. Oxford Textbook of Rheumatology. Oxford: Oxford University Press,,

3. Brewerton DA, Cafferty MFP, Hart FD, James DCO, Nicholls A, Sturrock RD. Ankylosing spondylitis and HLA-B27. Lancet 1973;i:904-7.

4. Schlosstein L, Terasaki PI, Bluestone E, Pearson CM. High association of HLA antigen W27 with ankylosing spondylitis. N Engl J Med 1973;288:704-6.

5. Hammer KE, Maika SD, Richardson JA, Tang JP, Taurog JD. Spontaneous inflammatory disease in transgenic rats expressing HLA-B27 and human 62m: An animal model of HLA-B27-associated human disorders. Cell 1990;63:1099-112.

6. Khan MA. Spondyloarthropathies. Editorial overview. Curr Opin Rheum 1996;8:267-6S.

7. Khan MA, Kushner I, Braun WE. Genetic heterogeneity in primary ankylosing spondylitis. J Rheumatol 1980;7:383-86.

8. Joysey VC, and Wolf E. HLA-A, -B and -C antigens: Their serology and cross-reaction. Dr Med Bull 1978;34:217-22.

9. Linssen A. B27-positive versus B27-negative disease. Scand J Rheumatol 1990;87(Suppl.);lll-19.

10. Khan MA. HLA-B27 polymorphism and association with disease. J Rheumatol 2000;27:1110-14.

11. Karr RW, Hahn Y, Schwartz BD. Structural identity of human histocompatibility leukocyte antigen-B27 molecules from patients with ankylosing spondylitis and normal individuals. J Clin Invest 1982;69:443-50.

12. Coppin HL, and McDevitt HO. Absence of polymorphism between HLA-B27 genomic exon sequences isolated from normal donors and ankylosing spondylitis patients. J Immunol 1986;137:2168-72.

13. Higgins CM, Lund T, Shipley ME, Ebringer A et al. Ankylosing spondylitis and HLA-B27: restriction fragment length polymorphism and sequencing of an HLA-B27 allele from a patient with ankylosing spondylitis. Ann Rheum Dis 1992;51:855-62.

14. Grumet FC, Fendley BM, Engleman EG. Monoclonal anti-HLA B27 antibody (B27M): Production and lack of detectable typing difference between patients with ankylosing spondylitis, Reiter's syndrome and normal controls. Lancet 1981;ii:174-76.

15. Verjans GM, van der Linden SM, van Eys GJM, de Waal LP, Kijlstra A. Restriction fragment length polymorphism of the tumour necrosis factor region in patients with ankylosing spondylitis. Arthritis Rheum 1991;34:486-89,

16. Martinez-Borra J, Gonzalez S, Lopez-Vazquez A et al. HLA-B27 alone rather than B27-relatcd class I haplotypes contributes to ankylosing spondylitis susceptibility. Hum Immunol 2000;61:131-39.

17. Eastmond CJ and Woodrow JC. Discordance for ankylosing spondylitis in monozygotic twins, Ann Rheum Dis J977;36:360-4.

18. Ebringer A, Ghuloom M, Ptaszynska T, Wilson C. Role of micro-organisms in the aetiopathogcnesis of HLA-B27 diseases. In: Veys E and Mielants H, eds. Spondyloarthropathies-involvement of the gut. Amsterdam: Experts Medica, Elsevier, 1987;235-47.

19. Taurog JD, Richardson JA, Crofl JT et al. The germ-free state prevents development of gut and joint inflammatory disease in HLA-B27 transgenic rats. J ExpMed 1994; 180:2359-64.

20. Mason RM, Murray RS, Oates JK, Young AC. Prostatitis and ankylosing spondylitis. Br Med J 1958; 1:748-51.

21. Ebringer A, Cowling P, Ngwa-Suh N, James DCO, Ebringer RW. Cross-reactivity between Klebsiella aerogenes species and B27 lymphocyte antigens as an actiological factor in ankylosing spondylitis. In: Dausset J, Svegaard A, eds. HLA and Disease. Paris: INSERM 1976;58:27.

22. Maki-Ikola O, Penttinen M, von Essen R, Gripenberg-Lerche C, Isomaki H, Granfors K. IgM, IgG and IgA class enterobacterial antibodies in serum and synovia! fluid in patients with ankylosing spondylitis and rheumatoid arthritis. Br J Rheumatol 1997;36;l051-53.

23. Welsh J, Avakian II, Cowling P, Ebringer A, Wooley P, Panayi G, Ebringer R. Ankylosing spondylitis, HLA-B27 and Klebsiella. I: Cross-reactivity studies with rabbit antisera. Br J Exp Pathol 1980;61:85-91.

24. Avakian H, Welsh J, Ebringer A, Entwistle CC. Ankylosing spondylitis, HLA-B27 and Klebsiella, II. Cross-reactivity studies with human tissue typing sera. Br J Exp Pathol 1980;61:92-96.
25. Van Bohemen CG, Grumet FC, Zanen HC. Identification of HLA-B27 Ml and M2 cross-reactive antigens in Klebsiella, Shigella and Yersinia. Immunology 1984;52:607-9.

26. Ogasawara M, Kono DH, Yu DTY. Mimicry of human histocompatibility HLA-B27 antigens by Klebsiella pneumonias. Infect Immun 1986;51:901-8.

27. Toivanen A, Stahlberg TH, Granfors K, Ebringer A. Absence of antibodies to Yersinia enterocolitica in patients with ankylosing spondylitis in London, Lancet 1986:2:1400-01.

28. Schwimmbeck PL, Yu DTY, Oldstone MBA. Autoantibodies to HLA-B27 in the sera of HLA-B27 patients with ankylosing spondylitis and Reiter's syndrome: molecular mimicry with Klebslella pneumoniae as potential mechanism of autoimmune disease. J Exp Med 1987; 166:173-81,

29. Husby G, Tsuchiya N, Schwimmbeck PL et al, Cross-reactive epitope with Klebsiella pneumoniae nitrogenase in articular tissue of HLA-B27 positive patients with active ankylosing spondylitis. Arthritis Rheum 1989;32:437-45.

30. Baines M, Ebringer A, Avakian H, Samuel D. The use of enzyme immunoassay (EIA) and radiobinding assay to investigate the cross-reactivity of Klebsiella antigens and HLA-B27 in ankylosing spondylitis patients and healthy controls. Scand J Rheumatol 1990;19:341-49.

31. Fielder M, Pirt SJ, Tarpey I, Wilson C, Cunningham P, Ettelaie C, Binder A, Bansal S, Ebringer A. Molecular mimicry and ankylosing spondylitis: possible role of a novel sequence in pullulanase of Klebsiella pneumoniae FEES 1995; 369:243-48.

32. Tani Y, Tiwana H, Hukuda S, Nishioka 3, Fielder M, Wilson C, Bansal S, Ebringer A. Antibodies to Klebsiella, Proteus and HLA-B27 peptides in Japanese patients with ankylosing spondylitis and rheumatoid arthritis. J Rheumatol 1997;24:109-14.

33. Tani Y, Sato H, Tanaka N, Mori K, Doida Y, Hukuda S. Serum IgAl and IgA2 antibodies against collagcns in patients with ankylosing spondylitis. Scand J Rheumatol l997;26:380-82,

34. Ahmadi K, Wilson C, Tiwana H, Ebringer A. Antibodies to Klebsiella pneumoniae mtrogenase rcductase in patients with ankylosing spondylitis. Ann Rheum Dis 1998; 57:441.

35. Wright V. Seronegative polyarthritis: a unified concept. Arthritis Rheum 1978; 21:619-33.

36. Tiwana H, Wilson C, Walmsley RS, Wakefield AJ, Smith MS, Cox NL, Hudson MJ, Ebringer A. Antibody responses to gut bacteria in ankylosing spondylitis, rheumatoid arthritis, Crohn's disease and ulcerative colitis. Rheumatol Int 1997;17:11-16.

37. Tiwana H, Natt RS, Benitez-Brito R, Shah S, Wilson C, Bridger S, Harbord M, Samer M, Ebringer A. Correlation between the immune responses to collagens type I, III, IV and V and Klebsiella pneumonias in patients with Crohn's disease and ankylosing spondylitis. Rheumatol 2001 ;40:15-23.

38. Amor Bf and Toubert AA. Reactive arthropathy, Reiter's syndrome, and enteric arthropathy in adults. In: Maddison PJ, Isenberg DA, Woo P and Glass DN, eds. Oxford Textbook of Rheumatology. Oxford: Oxford University Press, 1998; 1084-97.

39. Smith DM, Gibson RA, Drookes PM. Abnormal bowel permeability in ankylosing spondylitis and rheumatoid arthritis. J Rheumatol 1985;12:299-30S.

40. Martinez-Gonzalez O, Cantero-Hinojosa J, Paule-Sastre P et al. Intestinal permeability in patients with AS and their healthy relatives. Br J Rheumatol 1994;33:644-47.

41. Mielants H, Veys EM, Cuvelier C, De Vos M. Ileocolonoscopic findings in seronegative spondyloarthropathies. Br J Rheumatol 1988;27 (Suppl. 2): 195-205.

42. Lionetti P, Pupi A, Veltroni M et al. Evidence of sub-clinical inflammation by 99m technetium leukocyte scintigraphy in patients with HLA-B27 positive juvenile onset active spondyloarthropathy. J Rheumatol 2000;27:1538-41.

43. Ebringer R, Cook D, Cawdell DR, Cowling P, Ebringer A. Ankylosing spondylitis: Klebsiella and HLA-B27. Rheumatol Rehabil 1977;16:190-6.

44. Eastmond CJ, Wilshaw HE, Burgess SE et al. The frequency of faecal Klebsiella aerogenes in patients with ankylosing spondylitis and controls with respect to individual features of the disease. Ann Rheum Dis 1980;39:118-23,

45. Hunter T, Harding GK, Kaprove RE, Schroeder ML. Fecal carriage of various Klebsiella and Enterobacter species in patients with active ankylosing spondylitis. Arthritis Rheum 1981;24:106-8.

46. Kuberski TT, Morse HG, Rate RG, Bonnell MD. Increased recovery of Klebsiella from the gastrointestinal tract of Reiler's syndrome and ankylosing spondylitis patients. Br J Rheumatol 19S3;22 (Suppl. 2):85-90.

47. Van Kregten E, Huber-Bruning O, Vandenbroucke JP, Willers JMN. No conclusive evidence of an epidemiologies! relation between Klebsiella and ankylosing spondylitis. J Rheumatol 1991;!8:384-8.

48. Cowling P, Ebringer R, Ebringer A. Association of raised serum IgA in ankylosing spondylitis. Ann Rheum Dis 1980;39:545-49.

49. Collado A, Sanmarti R, Brancos MA, Kanterewicz E, Qullart T, Rotes-QueroI J, Cobos A. Immunoglobulin A and C-reactive protein levels in ankylosing spondylitis. Ann Rheum Dis 1987;46:719-20.

50. Hocini H, Isoaki S, Benlahrache C et al. Increased levels of serum IgA as IgAl monomers in ankylosing spondylitis. Ann Rheum Dis 1992;51:790-2.

51. Collado A, Sanmarti R, Serra C, Gallart T, Canete JD, Gratacos J, Vives 3, Munoz-Gomez J. Serum levels of secretory IgA in ankylosing spondylitis. Scand J Rheumatol !991;20:153-58.

52. Maki-Ikola O, Nissila M, Lehtinen K, Leirisalo-Repo M, Granfors K, IgAl and IgA2 subclass antibodies against Klebsiella pneumoniae in the sera of patients with peripheral and axial types of ankylosing spondylitis. Ann Rheum Dis 1995;54: 631-35.

53. Collado A, Gratacos J, Ebringer A, Rashid T, Marti A, Sanmarti R, Munoz-Gomez J, Serum IgA &n\\-Klebsiella antibodies in ankylosing spondylitis patients from Catalonia. Scand J Rheumatol 1994;23:119-23.

54.. Stodell MA, Butler RC, Zemelman VA, Henry K, Brewerton DA. Increased number of IgG-containing cells in rectal lamina propria of patients with ankylosing spondylitis. Ann Rheum Dis 1984;43:172-76.

55. Feltelius N, Hvatum M, Brandtzaeg P, Knutson L, Hallgren R. Increased jejunal secretory IgA and IgM in ankylosing spondylitis: Normalization after treatment with sulfasalazine. J Rheumatol 1994; 21: 2076-81.

56. Maki-Ikola O, Leirisalo-Repo M, Turanen U, Granfors K. Association of gut inflammation with increased serum IgA class Klebsiella antibody concentration in patients with axial ankylosing spondylitis (AS); Implication for different aetiopathogenetic mechanisms for axial and peripheral AS. Ann Rheum Dis 1997;56:180-S3.

57. Maki-Ikola O, Hallgren R, Kanerud L, Feltelius N, Knutsson L, Granfors K. Enhanced jejunal production of antibodies to Klebsiella and other Enterobactcria in patients with ankylosing spondylitis and rheumatoid arthritis. Ann Rheum Dis 1997;56:421-25.

58. Fournier AM, Denizet D, Delagrange A. La lymphographie dans la spondylarthrite ankylosante. Journal de Radiologie et d Electrologie 1969;50:773-84.

59. Goodman JW. Antigen Presentation and the Major Histocompatibility Complex. In: Stites DP, ed. Basic and Clinical Immunology. USA: Lange Med. Books, 1994; 58-67.

60. Sahly H, Podshun R, Sass R, Broker B, Kekow J, Gross WL, Ullmann U. Serum antibodies to Klebsiella capsular polysaccharides in ankylosing spondylitis. Arthritis Rheum 1994;37:7S4-59.

61. Maki-Ikola O, Nissila M, Lehtinen K, Granfors K. IgA class serum antibodies against three different Klebsiella serotypes in ankylosing spondylitis. Br J Rheumatol 1998;37:1299-302.

62. Ahmadi K, Wilson C, Tiwana H, Binder A, Ebringer A. Antibodies to Klebsiella pneumoniae lipopolysaccharide in patients with ankylosing spondylitis. Br J Rheumatol 199S;37:1330-33.

63. Ebringer A and Wilson C. HLA molecules, bacteria and autoimmunity. J Med Microbiol 2000; 49: 305-11.

64. Maki-Ikola O, Lehtinen K, Toivanen P, Granfors K. Antibodies to Klebsiella pneumoniae, Escherichia coli and Proteus mirabilis in the sera of ankylosing spondylitis with/without iritis and enthesitis. Br J Rheumatol 1995;34:418-20.

65. Sahly H, Rodsehun R, Rekow J, Nolle B, Gross WL, Ullmann U. Humoral immune response to Klebsiella capsular polysaccharides in HLA-B27 positive patients with acute anterior uveitis and ankylosing spondylitis. Autoimmunity 1998;28:209-15.

66. Ebringer A. Ankylosing spondylitis is caused by Klebsiella: Evidence from immunologic, microbiologic, and serologic studies. Rheum Dis Clin North Am 1992; 18: 105-21.

67. Wollenhaupt J, Schnarr A, Kuipers JG. Bacterial antigens in reactive arthritis and spondarthritis. Rational use of laboratory testing in diagnosis and follow-up. In: Zeidler H and Schumacher HR, eds. Spondyloarthropathies. Baillieres Clin Rheumatol 1998; 12:627-47.

68. Khare SD, Bull MJ, Hanson J, Luthra HS, David CS. Spontaneous inflammatory disease in HLA-B27 transgenic mice is independent of MHC class II molecules: A direct role for B-27 heavy chain and not B27-derived peptides. J Imrnunol 1998:160:101-6.

69. Lopez De Castro JA. The pathogenetic role of HLA-B27 in chronic arthritis. Curr Opinlmmun 1998;10: 59-66.

70. Urban RG Chicz RM, Lane WS, Strorainger JL et al. A subset of HLA-B27 molecules contains peptides much longer than nonamers. Proc Nat Acad Sci U.S.A. 1994;91:1534-38.

71. Alien RL, O'Callaghan CA, McMichael AJ, Bowness P. Cutting edge: HLA-B27 can form a novel beta 2-microglobuIin free heavy chain homodimer structure. J Immunol 1999; 162:5045-48.

72. Ebringer A, Baines M, Childerstone M, Ghuloom M, Ptaszynska T. Etiopathogenesis of ankylosing spondylitis and the cross-tolerance hypothesis. In: Ziff M and Cohen SB, eds. Advances in inflammation research. The spondyloarthropathies. Vol 9. New York: Raven Press, 1985;101-2g.

73. Raybourne RB, Bunning VK, Williams KM. Reaction of anti-HLA-B monoclonal antibodies with envelope proteins of Shigella species. Evidence for molecular mimicry in the spondyloarthropathies. J Immunol 1988;140:3489-95.

74. Edwards JCW, Bowness P, Archer JR. Jekyll and Hyde: the transformation of HLA-B27. Immunol Today 2000;21:256-60.

75. Zhang Y, Toivanen A, Toivanen P. Experimental Yersinia-lriggered reactive arthritis; effect of a 3-week course of ciprofloxacin. Br J Rheumatol 1997;36:541-46.

76. Hoogkamp-Korstanje JA, Moesker H, Bruyn QW. Ciprofloxacin v placebo for treatment of Yersinia enterocolitica triggered reactive arthritis. Arm Rheum Dis 2000;59:914-17.

77. Toivanen A, Yli-Kerttula T, Luukainen R et al. Effect of antimicrobial treatment on chronic reactive arthritis. Clin Exp Rheumatol 1993;! 1:301-7.

78. Nissila M, Lehtinen K, Leirisalo-Repo M, Luukainen R, Mutru O, Yli-Kerttula U. Sulfasalazine in the treatment of ankylosing spondylitis: A twenty-six-week, placebo-controlled clinical trial. Arthritis Rheum 1988;31:1111-15.

79. Kirwan J, Edwards A, Huitfeldt B, Thompson P, Currey H. The course of established ankylosing spondylitis and the effects of sulphasalazine over 3 years. Dr J Rheumatol 1993;32:729-33.

80. Dougados M, van der Linden S, Leirisalo-Repo M et al. Sulfasalazine in the treatment of spondyloarthropathy: A randomized, multicentre, double-blind, placebo-controlled study. Arthritis Rheum 1995;38:618-27.

81. Nissila M, Lahesmaa R, Leirisalo-Repo M, Lehtinen K, Toivanen P, Granfors K. Antibodies to Klebsiella pneumoniae, Escherichia coli and Proteus mirabilis in ankylosing spondylitis: Effect of sulfasalazine treatment. J Rheumatol 1994;21:2082-87.

82. Pullar T, Hunter JA, Cappell HA. Which component of sulphasalazine is active in rheumatoid arthritis? Br Med J 1985;290:1535-38.

83. Neumann VC, Taggart AJ and leGallez P, Asbury C, Hill J, Bird HA. A study to determine the active moiety of sulphasalazine in rheumatoid arthritis. J Rheumatol 1986;13:285-87.

84. Taggart A, Gardiner P, McEvoy F, Hopkins R, Bird H. Which is the active moiety of sulfasalazine in ankylosing spondylitis? Arthritis Rheum 1996;39:400-5.

85. Benitez-Del-Castillo JM, Garcia-Sanchez J, Iradier T, Banares A, Sulfasalazine in the prevention of anterior uveitis associated with ankylosing spondylitis. Eye 2000; 14:340-43.

86. Kapusnik-Uner, Abdominal infections. In: Hernfindal ET, Gourley DR and Hart LL, eds. Clinical Pharmacy and Therapeutics. USA: Williams and Wilkins, 1992; 1125-43.

87. Ebringer A and Wilson C. The use of low starch diet in the treatment of patients suffering from ankylosing spondylitis. Clin Rheumatol 1996;15 (Suppl. l):61-65.

88. Finegold SM, Sutler VL, Sugihara PT, Elder HA, Lehmann SM, Philips RL. Fecal microbial flora in Seventh Day Adventist populations and control subjects. Amer J Clin Nut 1977:30:1781-92,

89. Deschamps AM, Richard C, Lebeault JM. Bacteriology and nutrition of environmental strains in Klebsiella pneumonias involved in wood and bark decay. Ann Microbiol (mst Past) 1983; 134A: 189-96.

90. Ebringer A, Baines M, Ptaszynska T. Spondylarthritis, uveitis, HLA-B27 and Klebsiella. Immunol Rev 1985; 86: 101-16.

91. Trull A, Ebringer R, Panayi OS, Colthorpe J, James DCO, Ebringer A. IgA antibodies in ankylosing spondylitis. Scand J Rheumatol 1983;I2:249-53.

92. Trull A, Ebringer A, Panayi GS, Ebringer R, James DCO. HLA-B27 and the immune response to enterobacterial antigens in ankylosing spondylitis. Clin Exp Immunol 1984;55:74-80.
93. Shodjai-Moradi F, Ebringer A, Abuljadayel I (1992) IgA antibody response to Klebsiella in ankylosing spondylitis measured by immunoblotting. Ann Rheum Dis 51:233-37.

94. Tiwana H, Walmsley RS, Wilson C, Yiannakou JY, Ciclitira PJ, Wakefield AJ, Ebringer A. Characterization of the humoral immune responses to Klebsiella species in inflammatory bowel disease and ankylosing spondylitis. Br J Rheumatol 1998;3 7:525-31.

95. Cooper R, Fraser SM, Sturrock RD, Gemmel CO. Raised titres of anti-KIebsiflla IgA in ankylosing spondylitis, rheumatoid arthritis, and inflammatory bowel disease. Br Mcd J 1988;296; 1432-34.

96. O'Mahony S, Anderson N, Nuki G, Ferguson A. Systemic and mucosal antibodies to Klebsiella in patients with ankylosing spondylitis and Crohn's disease Ann Rheum Dis 1992;51:1296-300.

97. Mateicfca F, Kozakova D, Cebecauer M, Ondrasik M, Bosmansky K. Protilatky proti Klebsiella pneumoniae pri ankylozujucej spondylitide. Cas. Lek. Ces 1989;12S:132-34.

98. Thomson GTD, Alfa M, Orr K, Inman RD, Olson N, Thomson BRJ. Humoral immune response to bacterial lipopolysaccharide in adult idiopathic spondyloarthropathies. Amer College of Rheumatol, 57th Annual Meeting. San Antonio. 1993; Abstract Vol. S224, C218.

99. Yuan GH, Shi GY, Ding Y/i. Antibodies to Klebsiella pneumoniae in ankylosing spondylitis. Chung-Hua Nei Ko Tsa Chih Chinese J Int Med 1993;32:467-69 (Medline).

100. Sahly H, Kekow J, Podschun R, Schaff M, Gross W, Ullmann. Comparison of the antibody responses to the Klebsiella capsular types in ankylosing spondylitis and various rheumatic diseases. Infect Immun 1994;62:4838-43.

101. Ardicoglu O, Atay MB, Ataoglu H, Etiz N, Ozeci H. IgA antibodies to Klebsiella in ankylosing spondylitis. Clin Rheumatol 1996; 15:573-76.

102. Canoino-Diaz ME, Perez-Salazar JE, Dominguez-Lopez L et al. Antibody response to Klebsiella pneumoniae 60 kDa protein in familial and sporadic ankylosing spondylitis: role of HLA-B27 and characterization as a GroEL-like protein. J Rheumatol 1998;25:1756-64.

103. Blankcnberg-Sprenkels SHD, Fielder M, Feldkamp TEW, Tiwana H, Wilson C, Ebringer A. Antibodies to Klebsiella pneumoniae in Dutch patients with ankylosing spondylitis and acute anterior uveitis and to Proteus mirabilis in rheumatoid arthritis. J Rheumatol 199S;25:743-47.

104. Chou CT, Uksila J, Toivanen P. Enterobacterial antibodies in Chinese patients with rheumatoid arthritis and ankylosing spondylitis. Clm Exp Rheumatol 1998;16:161-64.

Table 1: Geographical distribution and analytical review of anti-Klebsiella antibodies in AS patients compared to healthy controls (HC)



Number of AS patients

Number of HCs

P value





















London &. Stevenage London


25 50

O.005 O.001















Chapel Hill




































Nuremberg & Kiel Ankara



<0,01 <0.001









Helsinki & Turku Helsinki


100 100

0.001 <0.001



Mexico DF

















Total number of subjects
1250 1088    


Table 2; Geographical distribution and analytical review anti-Klebsiella antibodies in patients with AS when compared to other rheumatic diseases (ORD)



Number of Number of AS ORD

P value





21 PsA






31 RA




London & Stevenage


25 RA






50 RA












50 RA












28 RA




Lubeck & Kiel


24 RA, 20 ReA,
24 SLE, 24 PsA






25 RA



Total number of subjects
446 643    

*RA; Rheumatoid arthritis, PsA: Psoriatic arthritis, NIA: Non-inflammatoty arthropathy, ReA: Reactive arthritis, SLE: Systemic lupus erythematosus, OA: Osteoarthritis

ED NOTE: The London AS Diet can be found here; originally included with this document.

Proposed sequential pathogentic events in AS development

Valid HTML 4.0 Transitional