“B27 Disease” Is a New Autoimmune

Disease That Affects Millions of People

ALAN EBRINGERa,b AND TAHA RASHIDa
a Schoolof Biomedical and Health Sciences, King’s College London, London,
United Kingdom
b Departmentof Rheumatology, Middlesex Hospital, UCL School of Medicine,
London, United Kingdom

ABSTRACT: “B27 disease” is a new autoimmune disease that afflicts mil-

lions of people throughout the world. “B27 disease” occurs in individuals
who have ankylosing spondylitis (AS) or preankylosing spondylitis and/or
uveitis and are also positive for HLA-B27. Molecular mimicry between
the bowel microbe Klebsiella and the HLA-B27 molecule, as well as the
spinal collagens types I, III, and IV, indicates a pathological mechanism
involving autoimmunity. Antibodies to Klebsiella microbes have been re-
ported in AS patients from 18 different countries. Sera from patients
with AS show complement-dependent cytopathic activity against sheep
red cells coated with HLA-B27 peptide antigens. Diagnosis of B27 disease
can lead to early treatment, involving low-starch diet, sulfasalazine, and
immunosuppressive and biological agents so as to prevent the irreversible
bony changes of established classical AS. The concept of B27 disease pro-
vides a new approach to the study and treatment of these disorders and
needs to be evaluated in prospective studies by the world rheumatological
community.

KEYWORDS: B27 disease; ankylosing spondylitis; Klebsiella; low-starch

diet

INTRODUCTION

“B27 disease” is a term used to describe a new autoimmune disease that

afflicts millions of people throughout the world. The concept of the B27 disease
arose from the discovery in the early 1970s, over 30 years ago, by two groups of
research workers, one from London1,2 and the other one from Los Angeles3 that
the great majority of patients suffering from ankylosing spondylitis (AS) also
possessed or carried the HLA-B27 molecule, which they had inherited from

Address for correspondence: Prof. Alan Ebringer, School of Biomedical and Health Sciences, King’s
College London, 150 Stamford Street, London SE1 9NH, UK. Voice: +020-8997-1883; fax: +020-
7848-4500.
alan.ebrunger@kcl.ac.uk

Ann. N.Y. Acad. Sci. 1110: 112–120 (2007). 

C 2007 New York Academy of Sciences.

doi: 10.1196/annals.1423.013
112
EBRINGER & RASHID 113

their parents. Also, some 30% of these AS patients will undergo an attack of
uveitis during their lifetime.
These two seminal studies showed that HLA-B27 was present in 96% of AS
patients, while the frequency of this HLA antigen in the general population
of the UK or the United States was about 8%. In Northern countries, such as
Sweden or Finland, HLA-B27 is found in 12–14% of the general population.
Thus, it would appear that there are some 5 million individuals in the UK
and approximately 24 million individuals in the USA who possess or carry
HLA-B27.
It had been known for over half a century that AS appeared to run in families.
The discovery that the majority of AS patients carried the HLA-B27 gene
provided a possible explanation for the familial occurrence of this disease.
Furthermore, the observation that a genetic factor was involved suggested
that AS could be a form of “reactive arthritis” occurring after exposure to an
environmental agent, such as that seen in Salmonella, Shigella, or Yersinia
reactive arthritis, since many patients with these conditions also appeared to
possess the HLA-B27 molecule.4
The explanation for the high frequency of HLA-B27 in AS patients is the
central problem in AS research today. Several theories have been proposed
to explain the link: Some suggest that HLA-B27 molecules present an as yet
unidentified environmental antigen; others propose that HLA-B27 is a marker
for a separate AS gene. But the model that has provided the most interesting
results is based on molecular mimicry between some external pathogens and
self-tissues of the patient. The pathological model and precedent that provided
a theoretical framework for studying this condition was rheumatic fever.

RHEUMATIC FEVER AS A MODEL

OF AN AUTOIMMUNE DISEASE

Rheumatic fever is an immunologically mediated disease produced by a

high titer of anti-streptococcal antibodies, which bind to cardiac tissue causing
rheumatic heart disease.5 When these antibodies also bind to the basal ganglia
of the brain they give rise to Sydenham’s chorea.6 The antibodies are evoked by
the presence of Streptococcus pyogenes microbes in patients suffering from an
upper respiratory tract infection or tonsillitis. The Streptococcus microbes ex-
hibit molecular mimicry in that they possess protein sequences that cross-react
or resemble cardiac and basal ganglia tissues. The pathological autoimmune
process is driven by the presence of large quantities of anti-streptococcal anti-
bodies which, when binding to cross-reacting self-antigens, activate the com-
plement cascade, causing cell death and tissue damage and thereby leading to
inflammation. Thus rheumatic fever is an autoimmune disease caused by the
patient’s own antibodies acting as autoantibodies.
114 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES

TABLE 1. Country and year of discovery of elevated levels of antibodies to Klebsiella

microbes in AS patients
England 1983
USA 1987
Scotland 1988
Slovakia 1989
Russia 1989
Australia 1990
Finland 1991
Canada 1993
China 1993
Spain 1994
Germany 1994
Turkey 1996
Sweden 1997
Japan 1997
Netherlands 1998
Mexico 1998
Taiwan 1998
India 2002

ANKYLOSING SPONDYLITIS AND HLA-B27

We have adapted the model of rheumatic fever to investigate the link between
HLA-B27 and AS. Rabbits injected with HLA-B27-positive lymphocytes ob-
tained from healthy bone marrow donors showed immunological reactivity
against Salmonella, Shigella, and Klebsiella microbes.7,8
AS patients were tested for the presence of these three microbes in fecal stool
cultures, but only Klebsiella was isolated more frequently when compared to
controls.9 Furthermore, allogeneic HLA-B27 tissue typing sera were found to
bind preferentially to Klebsiella antigens when compared to tissue typing sera
having other HLA specificities.10
Clearly an immunological cross-reactivity suggested a biochemical similar-
ity. A stereobiochemical type of molecular mimicry was described between
HLA-B27 and Klebsiella nitrogenase reductase by Oldstone’s group from La
Jolla,11 involving amino acids DRED and Klebsiella pullulanase D by our
group12 involving amino acids DRDE.
Elevated levels of antibodies to Klebsiella in AS patients have been re-
ported from different countries.13 (TABLE 1) Furthermore, there is also molec-
ular mimicry between collagens types I, III, IV, and Klebsiella pullulanase
A, thereby providing a possible explanation for the spinal locations of the
pathological lesions in AS.12
In two studies, one from Japan14 and the other one from the Netherlands,15
AS patients had antibodies to Klebsiella but not to Proteus, while rheumatoid
arthritis (RA) patients had antibodies against the urinary pathogen Proteus but
not to Klebsiella; thus each disease group was a specificity control for the
EBRINGER & RASHID 115

other condition. In the Dutch study, the AS, RA, and blood donor sera were
coded in Amsterdam, sent to London where antibody assays were carried out,
and results sent back to the Netherlands for decoding. It was suggested that
AS was a type of “Klebsiella-reactive arthritis” while RA could be considered
as a form of “Proteus-reactive arthritis.”
AS sera were found to have complement-dependent cytotoxic activity against
sheep red cells coated with either Klebsiella nitrogenase or Klebsiella pullu-
lanase cross-reacting peptides when compared to RA sera or blood donors.16
Thus anti-Klebsiella antibodies are acting as autoantibodies against HLA-B27
and spinal collagens and therefore AS can be considered as an autoimmune
disease in a similar way to rheumatic fever.
Both diseases are evoked by infectious microorganisms but at the sites of
pathological injury—the heart in rheumatic fever, the brain in Sydenham’s
chorea and the spine and entheses in AS—there are no pathological microbes,
only complement-dependent cytopathic autoantibodies, with activity against
self-antigens.
It would appear that rheumatic fever, Sydenham’s chorea, and AS are autoim-
mune diseases evoked by antibodies to environmental bacteria, the infectious
agents being located in different pathological sites. Such chronic autoimmune
diseases cannot satisfy Koch’s criteria of an infectious disease.

PROBLEMS ASSOCIATED WITH EARLY STAGES OF

ANKYLOSING SPONOYLITIS

Ankylosing spondylitis is a disease that starts between the ages of 15 and

30 years and is characterized by backache, especially in the lumbar area and
generalized muscle stiffness, which is usually relieved by exercise. Relief of
muscle stiffness and lumbar pain by exercise is a distinguishing feature of AS,
which can be used to differentiate it from mechanical backache.
The clinical features of AS are characterized by episodic lumbar backache,
eventually spreading to the thoracic and cervical spine, ribs, and large periph-
eral joints, such as hips, knees, and ankles.
However, in many cases the onset is insidious with a general feeling of
malaise, worse in the mornings and marked by muscle stiffness. The backache
gradually becomes more prominent, but despite many visits to doctors the
diagnosis of AS is obtained usually only after several years or even decades of
suffering.
This “delay to diagnosis” is well recognized by specialists in the field and can
be ascribed to the limitations of classification. To make a diagnosis of AS one
requires evidence of sacroiliitis either on X-ray examination or by magnetic
resonance imaging (MRI). However, the patient will suffer from backache for
several years before pathological changes can be observed by these radiological
techniques.
116 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES

If the patient has clinical evidence of backache but no evidence of sacroiliitis

on X-ray film, the label usually given is idiopathic arthritis, that is, arthritis
of unknown origin or undifferentiated spondylarthropathy, even if there is an
elevation in the erythrocyte sedimentation rate (ESR) or serum C-reactive
protein level.
Clearly the discovery of a link between HLA-B27 and Klebsiella suggests
a new approach to the taxonomy of this disease that might help to identify
patients in the early stages before they had developed the irreversible bony
changes of sacroiliitis with its attendant rigidity of spine and postural changes.
Treatment of advanced cases relies on the use of anti-TNF biologicals, but
the question arises whether such therapies or even some simpler ones, such as
a low-starch diet, could be used in the early cases.
Since it would appear that 20% of HLA-B27-positive individuals have some
symptoms of AS, it seems that there are 1 million individuals in the UK who
have clinical features of AS, while for the United States, with a population
approaching 300 million, there would appear to be 5 million patients with
some symptoms of HLA-B27-associated disease.
It is with these observations in mind that we have proposed new definitions
of preankylosing spondylitis (pre-AS), AS, or uveitis and linked them to the
presence of HLA-B27.17

Preankylosing spondylitis occurs in a patient when he or she has:

(1) Backache present on and off for 3 months.

(2) Patient is HLA-B27 positive.
(3) ESR is elevated > 30 mm/h.
(4) Anti-Klebsiella antibodies are present.
(5) There are no X-ray changes characteristic of spondylitis.

Classical AS occurs in a patient when he or she has:

(1) Backache present on and off for 3 months.

(2) Patient is HLA-B27 positive.
(3) ESR is elevated > 30 mm/h.
(4) Anti-Klebsiella antibodies are present.
(5) Sacroiliitis is present on radiological or MRI examination.

These new definitions exclude an important group of patients who satisfy

the criteria of AS, but do not possess the HLA-B27 gene.
In an important study on a group of HLA-B27-negative patients with AS,
Linssen’s group from the Netherlands18 showed that they differed from HLA-
B27-positive patients in that the disease tended to appear later in their thirties,
the sex incidence was more or less equal, and the severity of the disease was
milder. Whether the condition in HLA-B27-negative patients is a different
disease or associated with different HLA antigens is at the moment unclear.
EBRINGER & RASHID 117

However, the study of HLA-B27 in AS has led to the discovery of a bacterial

microbe, Klebsiella, which appears to be associated with these conditions, and
this suggests that a new classification is required so as to identify susceptible
patients in the early stages of their disease.

DEFINITION OF “B27 DISEASE”

Although the term of “pre-AS” was first proposed by Agarwal from Pitts-
burgh, he did not link this condition to HLA-B27.19 Since AS, pre-AS, and
uveitis are linked to HLA-B27, it is proposed that when patients suffer from
these conditions and they are HLA-B27-positive, they should be assigned to
the same taxonomic group.

B27 disease occurs in an HLA-B27-positive patient who has any of the

following conditions:
(1) Classical AS
(2) Pre-AS
(3) Uveitis
The value of the concept of B27 disease is that it suggests a novel approach
to therapy. Since the link between AS and HLA-B27 has led to the discovery of
the bowel microbe Klebsiella to be somehow involved, one can ask the question
whether possible therapeutic interventions can reduce the quantity of bowel
bacteria.

BOWEL FLORA AND LOW-STARCH DIET

It has been demonstrated that bowel flora bacteria grow on dietary car-
bohydrates entering through the ileocecal junction.20 The majority of these
carbohydrate molecules are derived from dietary starch21 and a low-starch diet
has been found to be effective in reducing the symptoms in AS patients.22 It
would appear that if an HLA-B27-positive individual has a high consumption
of starch, such as that found in bread, cakes, pasta, rice, and potatoes, this will
lead to proliferation of bowel bacteria.
One of the microbes in the bowel flora is Klebsiella, which will then evoke
a rise in anti-Klebsiella antibodies. Binding of such antibodies to the lumbar
spine collagens and chondrocytes carrying HLA-B27 will produce inflamma-
tion, with backache and B27 disease, especially in individuals having a high
starch consumption (FIG. 1).
An excellent book written by a patient with AS describes in detail some
of the recipes patients can use in the low-starch diet.23 Furthermore, another
sufferer with B27 disease has set up a website called “Kick-AS,” in which
118 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES

HLA-B27
INDIVIDUALS

HIGH STARCH
CONSUMPTION

INCREASED ENTEROBACTERIAL LOAD

(ESPECIALLY KLEBSIELLA MICROBES)

B27 DISEASE

FIGURE 1. The role of high starch consumption in the development of B27 disease.

the low-starch diet has been popularized especially among AS patients in the
USA.
A low-starch diet has been used at the Ankylosing Spondylitis Research
Clinic of the Middlesex Hospital since 1982. Some 500 AS patients have been
treated with this diet and the majority have claimed some benefits.24

CONCLUSIONS

A new autoimmune disease has been identified, occurring in individuals

who also possess the HLA-B27 antigen. There would appear to be over 20
million people with B27 disease throughout the world. Early identification of
patients with B27 disease is required so that therapy can be started using low-
starch diet, exercises, sulfasalazine, immunosuppressive drugs, and anti-TNF
biologicals in an endeavor to avoid the irreversible sequelae of established AS
and in generally helping the sufferers to reduce the severity of their symptoms.

ACKNOWLEDGMENTS

This work was supported by the American Friends of King’s College.

EBRINGER & RASHID 119

REFERENCES

1. CAFFREY M.F.P. & D.C.O. JAMES. 1973. Human lymphocyte antigen association in
ankylosing spondylitis. Nature 242: 121.
2. BREWERTON D.A., F.D. HART, E. BLUESTONE, et al. 1973. Ankylosing spondylitis
and HL-A 27. Lancet i: 904–907.
3. SCHLOSSTEIN L., P.I. TERASAKI, E. BLUESTONE, et al. 1973. High association of an
HLA antigen, W27, with ankylosing spondylitis. N. Engl. J. Med. 288: 704–
706.
4. LEIRISALO-REPO M. 2005. Reactive arthritis. Scand. J. Rheumatol. 34: 251–259.
5. GUILHERME L., J. KALIL & M. CUNNINGHAM. 2006. Molecular mimicry in the
autoimmune pathogenesis of rheumatic heart disease. Autoimmunity 39: 31–39.
6. KIRVAN C.A., S.E. SWEDO, D. KURAHARA, et al. 2006. Streptococcal mimicry
and antibody-mediated cell signaling in the pathogenesis of Sydenham’s chorea.
Autoimmunity 39: 21–29.
7. EBRINGER A., P. COWLING, N. NGWA-SUH, et al. 1976. Crossreactivity between
Klebsiella aerogenes species and B27 lymphocyte antigens as an aetiological
factor in ankylosing spondylitis. In HLA and Disease. J. Dausset & A. Svegaard,
Eds.: 27. INSERM. Paris.
8. WELSH J., H. AVAKIAN, A. EBRINGER, et al. 1980. Ankylosing spondylitis, HLA-
B27 and Klebsiella. I. Crossreactivity studies with rabbit antisera. Brit. J. Exp.
Pathol. 61: 85–91.
9. EBRINGER R., D. COOKE, D.R. CAWDELL, et al. 1977. Ankylosing spondylitis, Kleb-
siella and HLA-B27. Rheumatol. Rehabil. 16: 190–196.
10. AVAKIAN H., J. WELSH, A. EBRINGER, et al. 1980. Ankylosing spondylitis, HLA-
B27 and Klebsiella. II. Crossreactivity studies with human tissue typing sera.
Brit. J. Exp. Pathol. 61: 92–96.
11. SCHWIMMBECK P.L., D.T.Y. YU & M.B.A. OLDSTONE. 1987. Autoantibodies to
HLA-B27 in the sera of HLA-B27 patients with ankylosing spondylitis and
Reiter’s syndrome: molecular mimicry with Klebsiella pneumoniae as potential
mechanism of autoimmune disease. J. Exp. Med. 166: 173–181.
12. FIELDER M, S.J. PIRT, I. TARPEY, et al. 1995. Molecular mimicry and ankylos-
ing spondylitis: possible role of a novel sequence in pullulanase of Klebsiella
pneumoniae. FEBS. Lett. 269: 243–248.
13. EBRINGER A., T. RASHID, C. WILSON, et al. 2006. Ankylosing spondylitis, HLA-
B27 and Klebsiella—an overview: proposal for early diagnosis and treatment.
Curr. Rheumatol. Rev. 2: 55–68.
14. TANI Y., H. TIWANA, S. HUKUDA, 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.
15. BLANKENBERG-SPRENKELS S.H.D., M. FIELDER, T.E.W. FELTKAMP, et al. 1998. An-
tibodies to Klebsiella pneumoniae in Dutch patients with ankylosing spondylitis
and acute anterior uveitis and to Proteus mirabilis in rheumatoid arthritis. J.
Rheumatol. 25: 743–747.
16. WILSON C., T. RASHID, H. TIWANA, et al. 2003. Cytotoxicity responses to peptide
antigens in rheumatoid arthritis and ankylosing spondylitis. J. Rheumatol. 30:
972–978.
17. EBRINGER A., T. RASHID, C. WILSON, et al. 2005. Pre-ankylosing spondylitis is a
clinical disease entity: diagnostic criteria and therapeutic implications. Rheuma-
tologia 19: 91–96.
120 ANNALS OF THE NEW YORK ACADEMY OF SCIENCES

18. LINSSEN A. 1990. B27 positive versus B27 negative disease. Scand. J. Rheumatol.
87(Suppl.): 111–119.
19. AGARWAL A. 1980. Pre-ankylosing spondylitis. In Ankylosing Spondylitis. J.M.H.
Moll, Ed.: 69–77. Churchill Livingstone. London.
20. EBRINGER A., M. BAINES, M. CHILDERSTONE, et al. 1985. Etiopathogenesis of
ankylosing spondylitis and the crosstolerance hypothesis. In Advances in In-
flammation Research: The Spondyloarthropathies. M. Ziff & S.B. Cohen, Eds.:
101–128. Raven Press. New York.
21. ANDERSON I.H., A.S. LEVINE & M.D. LEVITT. 1981. Incomplete absorption of the
carbohydrate in all-purpose wheat flour. N. Engl. J. Med. 304: 891–892.
22. EBRINGER A. & C. WILSON. 1996. The use of a low starch diet in the treatment of
patients suffering from ankylosing spondylitis. Clin. Rheumatol. 15(Suppl. 1):
62–66.
23. SINCLAIR C. 2004. IBS and the Low Starch Diet. Ninox Press. Wellington.
24. EBRINGER A. & C. WILSON. 2002. Ankylosing spondylitis and diet. In Food Al-
lergy and Intolerance. J. Brostoff & S.J. Challacombe, Eds.: 761–768. Saunders.
London.