Journal of Nutritional & Environmental Medicine (December 2004) 14(4),

LABORATORY RESEARCH

Multiple Sclerosis, Sporadic Creutzfeldt–Jakob Disease and

Bovine Spongiform Encephalopathy: Are they Autoimmune

Diseases Evoked by

Acinetobacter

Microbes Showing

Molecular Mimicry to Brain Antigens?

V. CHAMOUN MRCP,2J. R. CROKER FRCP3AND A. BINDER MD4

1

Division of Health and Life Sciences, King’s College, London, UK, 2Institute of Neurology, National Hospital for Neurology and Neurosurgery, Queen Square,

London, UK, 3Department of Geriatric Medicine, UCH School of Medicine, Middlesex Hospital, London, UK, 4Department of Rheumatology, Lister Hospital,

Stevenage, Hertfordshire, UK

Abstract

Purpose and Design:Microbial agents showing molecular mimicry to self-antigens have been reported to initiate an autoimmune pathology in several diseases, including Sydenham’s chorea, rheumatic fever, rheumatoid arthritis (RA) and ankylosing spondylitis (AS). Cows affected by bovine spongiform encephalopathy (BSE) have been shown to have antibodies to Acinetobacter species, microbes that are known to possess bacterial antigens showing molecular mimicry with brain antigens. In this study, the possibility that patients with multiple sclerosis (MS) and sporadic Creutzfeldt–Jakob disease (sCJD) have elevated antibodies to bacterial and brain antigens was investigated.

Materials and Methods: Serum samples from different groups of patients, including 53 with MS, 20 with RA, 20 with AS, two with sCJD, 18 with cerebrovascular accident (CVA) and 20 with viral encephalitis, as well as 29 healthy control subjects were tested using the enzyme-linked immunosorbent assay (ELISA) method for the determination of antibodies against Acinetobacter,Klebsiella,ProteusandEscherichia microbes, as well as myelin basic protein (MBP) brain antigens.

Results:Significantly elevated levels of IgA and IgG antibodies toAcinetobacter and MBP were observed in sera of MS and sCJD patients when compared with patients with encephalitis, CVA, AS, RA as well as 29 healthy blood donors. Furthermore, MS patients had shown a significant correlation between anti-Acinetobacter and anti-MBP antibody levels. Anti-Klebsiella antibodies were found to be elevated only in AS patients and anti-Proteus antibodies elevated only in RA patients.

Conclusions:MS and possibly sCJD could develop as the result of an autoimmune response to Acinetobacterbacteria. The possible role of this microbe in evoking autoimmune responses in MS or sCJD requires further evaluation.

Keywords:Acinetobacter, bovine spongiform encephalopathy, Creutzfeldt–Jakob disease, multiple sclerosis.

This paper is based on a presentation given at the BSAENM summer conference 2004 ‘The Leaky Gut’.

ISSN 1359-0847 print/ISSN 1364-6907 online# _{2004 Taylor & Francis Ltd} DOI: 10.1080/13590840500088131

INTRODUCTION

Some autoimmune diseases appear to be triggered by external environmental agents, especially micro-organisms. Among these diseases, rheumatic fever forms the prototypic example. Antibodies produced against group A haemolyticStreptococcibind to the cross-reactive antigens in the myocardial tissues, synovial joints and the basal ganglia of the brain causing carditis, arthritis and Sydenham’s chorea, respectively [1].

Rheumatoid arthritis (RA), a relatively common and potentially disabling arthritic condition, is another example of an immune-mediated disease. RA appears to be triggered by a urinary pathogen, Proteus mirabilis, which shares cross-reactive epitopes with HLA-DR4 genetic markers as well as with collagen type XI fibres [2], which are present in the small joints of the hands and feet. Patients with RA from 15 different countries worldwide were shown to have elevated levels of antibodies to Proteus[3].

Furthermore, patients with ankylosing spondylitis (AS) were shown to have significantly elevated levels of antibodies to Klebsiella microbes, as reported by independent groups from 14 different countries [4].Klebsiellamicrobes possess epitopes, which are found to be homologous and cross-reactive with self-antigens, such as HLA-B27 [5] and collagen types I, III and IV [6].

Bovine spongiform encephalopathy (BSE) is a fatal neurological disease that appeared in the UK in cattle following the introduction of a new form of winter feed. The ban of these feeding materials has successfully resulted in a dramatic drop in the number of infected cattle. Although the main culprit in the causation of BSE was suggested to be an abnormal prion protein [7], another possibility is that the winter feed was contaminated with bacteria containing antigens cross-reacting with brain tissues. Clearly this indicates that BSE, also known as ‘mad cow disease’, would appear to be a gut-mediated disease. The virtual disappearance of BSE in the UK since the removal of ‘meat and bone meal’ from the bovine food chain is consistent with the ‘leaky gut hypothesis’.

Injection of brain tissues into animals produces experimental allergic encephalomyelitis (EAE) as a result of an autoimmune process against the myelin basic proteins (MBP) [8]. Eylar and co-workers [9] have isolated a highly encephalitogenic amino acid peptide, FSWGAEGQK, from bovine myelin. We have used this peptide sequence to search the Genbank and SwissProt databases for similar sequences present in various microbes and found that the enzyme, 4-carboxy-muconolactone decarboxylase of Acinetobacter calcoaceticus, shares a homologous region with bovine myelin protein [10, 11].

Animals affected by BSE were found to have elevated levels of antibodies to Acinetobacter, but not to Escherichia coli, Agrobacterium, Klebsiella, Proteus, Bacillus or Serratia, as well as elevated levels of autoantibodies to MBP and neurofilament (NF) brain antigens when compared with healthy animals [11, 12]. These findings in EAE and BSE have led us to investigate humoral immune responses toAcinetobacterand MBP in patients with multiple sclerosis (MS).

In a previous study, MS patients were found to have elevated levels of antibodies to five strains of Acinetobacter species as well as to MBP and NF autoantigens [13]. In a later study, the same group of MS patients were found to have elevated levels of antibodies to peptides containing homologous sequences present in Acinetobacter, Pseudomonas, MBP and myelin oligodendrocyte glycoprotein (MOG) when compared with patients with cerebrovascular accidents (CVA) or healthy controls. Antisera raised in mice against Acinetobacter peptides were found to be significantly inhibited by homologous peptides from the MBP or Pseudomonas microbe when measured using the enzyme-linked immunosorbent assay (ELISA) method [14].

In another study, Austrian patients with MS and sporadic Creutzfeldt–Jakob disease (sCJD) were found to have significantly elevated levels of antibodies againstAcinetobacter bacteria as well as MBP and NF autoantigens when compared with healthy subjects and patients with other neurological diseases [15].

In an endeavour to evaluate the specificity of this response, we carried out a study in a larger group of MS patients and compared them with patients with other diseases.

MATERIALS AND METHODS

Patients and Controls

Serum samples taken from a total of 123 patients (53 MS, two sCJD, 18 CVA, 20 RA, 20 AS and 10 encephalitis) and 29 healthy controls were investigated. The age and sex distributions of these individuals, as well as their recruitment sources, are shown in Table 1. The selection of each group of patients was defined according to the following diagnostic criteria: New York criteria for AS [17], American College of Rheumatology criteria for RA [18] and the Poser criteria for the diagnosis of MS [19], while the diagnosis of sCJD patients was based on the characteristic pathological appearance involving spongiform changes at post-mortem examination.

Bacterial Preparations

Acinetobacter calcoaceticus microbes were obtained from the National Collections of Industrial and Murine Bacteria (Aberdeen, UK), while Klebsiella pneumonia, Proteus mirabilisandEscherichia colimicrobes were provided by the Department of Microbiology, King’s College London. All these bacteria were cultured using the same method. Briefly, cultures were grown in 2 l flasks on an orbital shaker for 16 hours at 37‡C in 200 ml of nutrient broth (Oxoid; 25 g l21) and at 30‡C for 2 days for Acinetobacter. The flasks were inoculated with 10 ml of the corresponding starter culture and were left shaking at 37‡C for 6 hours. The bacterial cultures were then harvested by centrifugation at 6000 rpm for 20 min at 4‡C (Beckman JA-20 rotor, six 250 ml cuvettes). The pellets of cells were then washed three times with 0.15 M phosphate-buffered saline (PBS, pH 7.4) before being finally resuspended in 20 ml of PBS.

ELISA

An optimal concentration of each bacterial sample was prepared by dilution in 0.05 M carbonate buffer (pH 9.6) to give an optical density reading of 0.25 (106bacterial cells ml21)

TABLE 1. Sources of the serum samples and demographic characteristics of the patient and control groups included in the study

Number

Age (years):

mean (range) Male/female Source of sera

Multiple sclerosis 53 46 (29–68) 26/27 National Hospital for Neurology and Neurosurgery, London

Encephalitis 10 38 (3–66) 8/2 National Hospital for Neurology and Neurosurgery, London

Creutzfeldt–Jakob disease 2 66 (58–74) 1/1 National Hospital for Neurology and Neurosurgery, London

Cerebrovascular accident 18 82 (58–94) 7/11 Department of Geriatric Medicine, UCL, London

Ankylosing spondylitis 20 49 (32–71) 16/4 AS Research Clinic, Middlesex Hospital, London

Rheumatoid arthritis 20 68 (47–88) 4/16 Department of Rheumatology, Lister Hospital, Stevenage Healthy controls 29 43 (19–66) 15/14 Blood Donor Bank, London

on the spectrophotometer (Corning, model 258). The ELISA method was carried out as previously described [20]. Briefly, 96-well polystyrene microELISA plates (Dynatech) were coated with bacterial suspensions (200ml well21) overnight at 4‡C and the non-specific sites were blocked with PBS containing 0.1%Tween 20 (Sigma) and 0.2%bovine serum albumin (Sigma). The plates were then washed three times and 200ml (1/200 dilution) of each test or control serum sample was added. After incubation for 2 hours at 37‡C, the plates were washed again and 200ml of rabbit anti-human alpha-chain or gamma-chain-specific horseradish peroxidase conjugate (Dako) in 1/500 dilution was added to each well and incubated for 1 hour at 37‡C. The washing procedure was repeated, and 200ml of substrate solution 2,2’-azinobis-(3-ethylbenzthiazoline-6-sulfonic acid) (0.5 mg ml21; Sigma) in citrate phosphate buffer, pH 4.1, containing 0.98 mM H2O2, was added to each well. The plates were developed in the dark at room temperature for 25 min. After stopping the reaction by adding 100ml of sodium fluoride (2 mg ml21; Sigma) solution to each well, absorbances were measured on a microtitre plate reader (Dynatech MR606) at 630 nm. Each serum sample was tested in duplicate and all studies were carried out under code, so that the tester was blind to whether the sample was a test or a control.

The same method was also used to determine the level of autoantibodies against bovine MBP (Sigma) with an optimal antigen concentration of 25mg ml21.

Statistical Analysis

The mean optical density units of patients with MS or other disease controls were compared with the mean values of healthy controls using Student’s t-test, and 95%

confidence limits of the control group were calculated. Pearson’s correlation coefficient r was calculated using the statistical package Prism 3.0 (GraphPad software).

RESULTS

IgA Antibacterial Antibodies

IgA antibodies to Acinetobacter calcoaceticus(Fig. 1) were significantly elevated in serum samples from patients with MS when compared with healthy controls (t~4.12,pv0.001);

patients with AS (t~3.46, pv0.001); RA (t~3.12, pv0.005); CVA (t~3.84, pv0.005);

and those with encephalitis (t~2.56, pv0.02). IgA Acinetobacter antibodies were also

elevated significantly in two sCJD patients when compared with healthy controls (t~6.25, pv0.0001), patients with AS (t~10.24, pv0.0001), RA (t~6.97, pv0.0001), CVA

(t~7.99,pv0.0001) and encephalitis (t~8.39,pv0.0001), but not when compared with the

MS patients. There was no significant difference in the IgA anti-Acinetobacterlevel between controls and the other disease groups. IgA antibodies toKlebsiella(data not shown) were significantly elevated in AS patients when compared with healthy controls (t~4.22, pv0.001); patients with RA (t~3.49, pv0.005); CVA (t~3.44, pv0.005); encephalitis

(t~2.54,pv0.02); and MS (t~5.89, pv0.001). There was no significant difference in the

level of IgA anti-Klebsiellaantibodies between other disease groups. Furthermore, there was no difference in the level of IgA anti-Escherichia coli antibodies between all disease groups and healthy controls.

IgG Antibacterial Antibodies

IgG antibodies to Acinetobacter calcoaceticus (data not shown) were also significantly elevated in MS patients when compared with healthy controls (t~2.65,pv0.01); patients

with AS (t~2.57, pv0.02); RA (t~2.86, pv0.01); CVA (t~3.08, pv0.005); and

encephalitis (t~2.96, pv0.005). IgG antibodies were significantly elevated in sCJD

(t~3.62, pv0.01). IgG antibodies to Proteus mirabilis were significantly elevated in RA

patients when compared with healthy controls (t~3.93, pv0.001); patients with AS

(t~3.31, pv0.005); CVA (t~3.15, pv0.005); encephalitis (t~2.30, pv0.05); and MS

(t~5.17, pv0.001). There was no significant difference in the level of IgG anti-Proteus

antibodies between the other disease groups. Furthermore, there was no significant

FIG. 1. Anti-AcinetobacterIgA levels (bars~means) in sera of 29 healthy controls and 123 patients with different diseases: 20 with ankylosing spondylitis (AS), 20 with rheumatoid arthritis (RA), 18 with cerebrovascular accident (CVA), 10 with encephalitis (ENC), 53 with multiple sclerosis (MS) and two with Creutzfeldt–Jakob disease (CJD). The dashed line represents 95%confidence limits for the mean of healthy controls (two-tailed test).

difference in the level of IgG anti-Escherichia coliantibodies between all disease groups and healthy controls.

IgA Anti-myelin Antibodies

IgA antibodies to MBP were significantly elevated in MS patients (Fig. 2) when compared with healthy controls (t~4.88, pv0.001); patients with AS (t~3.75, pv0.001); RA

(t~3.10, pv0.005); and CVA (t~3.04, pv0.005), but not when compared with patients

with encephalitis. Anti-MBP IgA antibodies were also significantly elevated in sCJD patients when compared with healthy controls (t~6.82, pv0.0001); patients with AS

(t~6.78, pv0.0001); RA (t~6.67, pv0.0001); CVA (t~4.37, pv0.001); encephalitis

(t~4.93,pv0.001); and MS (t~2.71,pv0.01). Longitudinal Assessment of IgA Antibacterial Antibodies

A longitudinal study of one 38-year-old female patient with MS showed three peaks of IgA anti-Acinetobacterantibody levels during weekly measurements, over a period of 10 weeks (Fig. 3), but no such elevations were observed with IgA anti-Klebsiellaor anti-Escherichia coliantibodies. The peaks of anti-Acinetobacteractivity appeared to be preceded either by a urinary tract or upper respiratory tract infection.

Correlation Analyses

The correlation coefficientr was calculated between IgA antibody levels to Acinetobacter calcoaceticus and MBP in 53 MS patients and 87 non-MS individuals (20 AS, 20 RA, 18 CVA and 29 healthy controls). There was a significant positive correlation among these groups between antibodies to Acinetobacter and autoantibodies to MBP (r~z0.20, pv0.01). No significant correlation was seen when each of the groups of patients, MS and

non-MS, were studied separately.

DISCUSSION

Significantly elevated levels of IgA and IgG antibodies to Acinetobacter and IgA autoantibodies to MBP were found in the majority of MS patients and the two sCJD patients when compared with healthy controls as well as with patients with AS, RA, CVA and those with encephalitis. Furthermore, one MS patient had shown bouts of antibody elevations against Acinetobacter but not Klebsiella or Escherichia microbes following attacks of upper respiratory tract infections. Moreover, a significant positive correlation was observed between anti-Acinetobacter and anti-MBP antibodies.

Similar results have also been shown previously, where patients with MS had significantly elevated levels of autoantibodies to MBP and NF brain antigens as well as antibodies to five different strains ofAcinetobacterandPseudomonas aeruginosa, but not to Escherichia colimicrobes when compared with healthy controls or patients with CVA [14]. In a later study, antibodies to mimicry peptides from Acinetobacter, Pseudomonas aeruginosa, MBP and MOG were significantly elevated in MS patients compared with healthy controls or patients with CVA, and these antibacterial and anti-brain antibodies were found to be cross-reactive to each other [15].

MS is the most common immune-mediated, demyelinating disorder of the central nervous system. Like other presumed autoimmune diseases, MS is more common in females and clinical symptoms often appear during young adulthood [21]. Patients with MS show relatively high morbidity and mortality rates, especially in chronic progressive forms

of the disease and particularly among those living in high latitudes [22, 23], both in the northern and southern hemispheres.

Epidemiological studies on variant CJD show a higher incidence of this disease in Scotland compared with southern England [24]. A similar increased incidence of MS occurs in Scotland compared with other parts of the UK [25].

A hypothesis has been proposed, that individuals having genetic susceptibility factor(s)

FIG. 2. IgA anti-myelin basic protein antibodies in the same groups of patients and healthy controls shown in Fig. 1. The value parameters used are the same as in Fig. 1.

will develop the disease through exposure to some environmental factors. Various viruses have been proposed as possible causative factors in MS [26], but no such agents have been directly linked to the onset of sCJD.

Another alternative possibility, however, is that the development of either MS or even CJD could occur in susceptible individuals who develop autoantibodies to neuronal antigens, such as myelin, NF and possibly prion proteins, as the result of overt or subclinical upper respiratory tract infections with Acinetobacter and/or Pseudomonas bacteria [14,15], microbes that carry antigens cross-reacting with brain tissues. Recent studies have shown that active complement components such as C1q, C3b and membrane attack complexes, which denote signs of inflammation, were found in the pathological

FIG. 3. Serial IgA antibody levels to Acinetobacter, Klebsiella and Escherichia coli in one patient with multiple sclerosis followed at weekly intervals for over 2 months. URTI, upper respiratory tract infection; UTI, urinary tract infection.

lesions of the diseased brains of both MS [27] and CJD [28] patients. More recently, a comprehensive literature evaluation carried out on MS and sCJD showed considerable similarities in various clinical, pathological and immunological features of these diseases [29], which raises the question of whether sCJD is a severe form of MS.

Sinusitis is present in around 70% of MS patients and the rate of MS exacerbations appears to be doubled during the sinusitis attacks [30]. Furthermore, using magnetic resonance imaging scans of the nasal sinuses, 53%of MS patients had evidence of sinusitis [31]. In another study, antral tap and endoscopically directed tissue culture (ETC) were carried out on acute sinusitis patients and the commonest microbial agents were found to beAcinetobacterspecies, involving 37 and 33%of the isolates from the antral tap and the ETC, respectively [32].

Because IgG antibodies can cross the blood–brain barrier, inflammation and associated immune reactions would then follow as the result of antibody-mediated cytotoxic damage to the neuronal tissues, similar to the mechanism proposed for Sydenham’s chorea in rheumatic fever [33]. Furthermore, Proteus and Klebsiella antibody-specific complement-mediated cellular damages were observed more significantly in RA and AS patient groups, respectively, when compared with control subjects [34]. A similar process could occur in MS, with pathological changes causing episodic attacks of inflammation and demyelination with progressive neuronal loss, giving rise to relapsing remitting or progressive forms of MS. These pathological lesions might remain quiescent but progressively lead to neuronal loss and/or vacuolation with spongiform changes, giving rise to the histopathological characteristics of these diseases.

The number of CJD patients in our study is too small to draw any conclusions, but the presence of elevated levels of antibodies toAcinetobacter/Pseudomonasand autoantibodies to brain antigens in MS patients suggests that these could be triggering or perpetuating factors in the development of this disease. Further studies, however, are required to assess the role of these bacteria in both diseases.

ACKNOWLEDGEMENTS

This work was supported by the Department of the Environment, Food and Rural Affairs (DEFRA)(CSA 4302), Trustees of the Middlesex Hospital and American Friends of King’s College London.

REFERENCES

[1] Gibofsky A, Zabriskie JB. Rheumatic fever. In: Isenberg D, Maddison PJ, Woo P, Glass D, Breedveld FC (eds). Oxford Textbook of Rheumatology. Oxford: Oxford University Press, 2004, 669277.

[2] Wilson C, Ebringer A, Ahmadi K et al. Shared amino acid sequences between major histocompatibility complex class II glycoproteins, type XI collagen and Proteus mirabilis in rheumatoid arthritis. Ann Rheum Dis 1995; 54: 216–20.

[3] Ebringer A, Rashid T, Wilson C. Rheumatoid arthritis: proposal for the use of anti-microbial therapy in early cases. Scand J Rheumatol 2003; 32: 2–11.

[4] Ebringer A, Rahsid T, Wilson C, Ptaszynska T, Fielder M. Ankylosing spondylitis, HLA-B27 and

Klebsiella—an overview: proposal for early diagnosis and treatment. Curr Rheumatol Rev (in press).

[5] 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 Klebsiella pneumoniaeas potential mechanism of autoimmune disease. J Exp Med 1987; 166: 173–81. [6] Fielder M, Pirt SJ, Tarpey Iet al.Molecular mimicry and ankylosing spondylitis: possible role of a

novel sequence in pullulanase ofKlebsiella pneumoniae. FEBS 1995; 369: 243–8.

[7] Prusiner SB. Biology and genetics of prion diseases. Ann Rev Microbiol 1994; 48: 655–86. [8] Patterson PY. Experimental allergic encephalomyelitis and autoimmune disease. Adv Immunol

[9] Eylar EH, Caccam J, Jackson JJet al.Experimental allergic encephalomyelitis: synthesis of disease-inducing site of the basic protein. Science 1970; 168: 1220–3.

[10] Ebringer A, Pirt SJ, Wilson C et al. Bovine spongiform encephalopathy: is it an autoimmune disease due to bacteria showing molecular mimicry with brain antigens? Environ Health Perspect 1997; 105: 1172–4.

[11] Ebringer A, Hughes L, Rashid T, Wilson C.Acinetobacterimmune responses in multiple sclerosis: etiopathogenetic role and its possible use as a diagnostic marker. Arch Neurol 2005; 62: 33–6. [12] Tiwana H, Wilson C, Pirt J et al. Autoantibodies to brain components and antibodies to

Acinetobacter calcoaceticusare present in bovine spongiform encephalopathy. Infect Immun 1999; 67: 6591–5.

[13] Wilson C, Hughes LE, Rashid T et al. Antibodies to Acinetobacter bacteria and bovine brain peptides, measured in bovine spongiform encephalopathy (BSE) in an attempt to develop an ante-mortem test. J Clin Lab Immunol 2003; 52: 23–40.

[14] Hughes LE, Bonell S, Natt RSet al.Antibody responses toAcinetobacterspp. andPseudomonas aeruginosa in multiple sclerosis: prospects for diagnosis using the myelin-Acinetobacter -neurofilament antibody index. Clin Diagn Lab Immunol 2001; 8: 1181–8.

[15] Hughes L, Smith PA, Bonell S et al. Cross-reactivity between related sequences found in

Acinetobacter sp., Pseudomonas aeruginosa, myelin basic protein and myelin oligodendrocyte glycoprotein in multiple sclerosis. J Neuroimmunol 2003; 144: 105–15.

[16] Allerberger F, Berger T, Reindl M et al. Anti-Acinetobacter and anti-neuronal antibodies in Austrian patients with multiple sclerosis and other neurological diseases. Ann Neurol (submitted). [17] Van der Linden S, Valkenburg HA, Cats A. Evaluation of diagnostic criteria for ankylosing

spondylitis. Arthritis Rheum 1984; 27: 361–7.

[18] Arnett FC, Edworthy SM, Bloch DAet al.The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 315–24. [19] Poser CM, Paty DW, Scheinberg Let al.New diagnostic criteria for multiple sclerosis: guidelines

for research protocols. Ann Neurol 1983; 13: 227–31.

[20] Khalafpour S, Ebringer A, Abuljadayel I et al. Antibodies to Klebsiella and Proteus micro-organisms in ankylosing spondylitis and rheumatoid arthritis patients measured by ELISA. Br J Rheumatol 1988; 27(Suppl.II): 86–9.

[21] Hafler DA. Multiple sclerosis. J Clin Invest 2004; 113: 788–94.

[22] Landtblom AM, Riise T, Boiko Aet al.Distribution of multiple sclerosis in Sweden based on mortality and disability compensation statistics. Neuroepidemiology 2002; 21: 167–79.

[23] Warren S, Warren KG, Svenson LWet al.Geographic and temporal distribution of mortality rates for multiple sclerosis in Canada, 1965–1994. Neuroepidemiology 2003; 22: 75–81.

[24] Cousens SN, Linsell L, Smith PGet al. Geographical distribution of variant CJD in the UK (excluding Northern Ireland). Lancet 1999; 353: 18–21.

[25] Rothwell PM, Charlton D. High incidence and prevalence of multiple sclerosis in south east Scotland: evidence of a genetic predisposition. J Neurol Neurosurg Psychiatr 1998; 64: 730–5. [26] Gilden DH. Viruses and multiple sclerosis. JAMA 2001; 286: 3127–9.

[27] Kulkarni AP, Kellaway LA, Lahiri DK et al. Neuroprotection from complement-mediated inflammatory damage. Ann NY Acad Sci 2004; 1035: 147–64.

[28] Kovacs GG, Gasque P, Strobel Tet al.Complement activation in human prion disease. Neurobiol Dis 2004; 15: 21–8.

[29] Ebringer A, Rashid T, Wilson C, Boden R, Thompson E. A possible link between multiple sclerosis and Creutzfeldt–Jakob disease based on clinical, genetic, pathological and immunological evidence involvingAcinetobacterbacteria. Med Hypoth 2005; 64: 487–94.

[30] Gay D, Dick G, Upton G. Multiple sclerosis associated with sinusitis: case-controlled study in general practice. Lancet 1986; 1: 815–19.

[31] Jones RL, Crowe P, Chavda SVet al.The incidence of sinusitis in patients with multiple sclerosis. Rhinology 1997; 35: 118–9.

[32] Casiano RR, Cohn S, Villasuso Eet al.Comparison of antral tap with endoscopically directed nasal culture. Laryngoscope 2001; 111: 1333–7.

[33] Kirvan CA, Swedo SE, Heuser JS et al. Mimicry and autoantibody-mediated neuronal cell signalling in Sydenham chorea. Nat Med 2003; 9: 914–20.

[34] Wilson C, Rashid T, Tiwana Het al.Cytotoxicity responses to peptide antigens in rheumatoid arthritis and ankylosing spondylitis. J Rheumatol 2003; 30: 972–8.