Journal of Neuroimmunology, 31 (1991) 91-96

91

© 1991 Elsevier Science Publishers B.V. (Biomedical Division) 0165-5728/91/$03.50 JNI 01036

Antibodies to myelin-oligodendrocyte glycoprotein in cerebrospinal fluid from patients with multiple sclerosisand controls B.-G. Xiao, C. Linington and H. Link Department of Neurology, Karolinska Institutet, Huddinge Unioersity Hospital, Stockholm, Sweden (Received 4 June 1990) (Revised, received 17 August 1990) (Accepted 20 August 1990)

Key words: Multiple sclerosis; Myelin-oligodendrocyte glycoprotein; Antibody; Autoimmunity; Enzyme-linked immunosorbent assay

Summary Myelin-oligodendrocyte glycoprotein (MOG) has been implicated as a target for antibody-mediated immune attack in experimental autoimmune encephalomyelitis (EAE) which has been used extensively as an experimental model of multiple sclerosis (MS). We have screened cerebrospinal fluid (CSF) and plasma from 30 patients with MS, 30 with other neurological diseases (OND) and 30 with tension headache for anti-MOG antibodies of IgG isotype by enzyme-linked immunosorbent assay (ELISA). Such antibodies were detected in CSF from seven of the patients with MS, compared to two with OND and one with tension headache. No anti-MOG IgG antibodies were demonstrable in plasma. Antibody specificity was confirmed by Western blot immunostaining. Antibody levels were higher in MS compared to OND and tension headache. No correlation was observed between anti-MOG IgG antibodies and total IgG levels in CSF. The significance of anti-MOG antibodies demonstrated in MS CSF remains to be defined.

Introduction After more than 150 years of research, the etiology and pathogenesis of multiple sclerosis (MS) is still unknown (Fredrikson and Kam-Hansen, 1989). Based on morpholOgic and clinical similarities with the chronic relapsing variant of experimental allergic encephalomyelitis (EAE),

Address for correspondence: Dr. H. Link, Department of Neurology, Karolinska Institutet, Huddinge University Hospital, S-141 86 Huddinge, Stockholm, Sweden.

which can be induced in juvenile experimental animals by sensitization with spinal cord tissue and Freund's complete adjuvant, autoimmunity has been suggested to be involved in the pathogenesis of MS (Lassmann, 1983; McFarland and Dhib-Jalbut, 1989). EAE can be transferred by myelin basic protein (MBP)-specific T cell lines to naive syngeneic animals (Ben-Nun et al., 1981). Besides MBP, the quantitatively dominating myelin component proteolipid protein (PLP) has also been shown to be encephalitogenic (Waksman et al., 1954; Satoh et al., 1987). However, autoantibody responses against myelin surface epitopes

92 may strongly affect the course of EAE and be responsible for extensive primary demyelination. This has most clearly been documented by systemic injection with a mouse monoclonal antibody 8-18C5 against myelin oligodendrocyte glycoprotein (MOG) (Schluesener et al., 1987; Lassmann et al., 1988; Linington et al., 1988). MOG is a surface marker of oligodendrocyte (Scolding et al., 1989) which was first identified utilizing the monoclonal antibody 8-18C5 (Linington et al., 1984). Oligodendrocytes are responsible for the synthesis and maintenance of myelin in the central nervous system (CNS). The MOG epitope recognized by the monoclonal antibody 8-18C5 is highly conserved and present on rat, human, bovine and guinea pig myelin. Its potential to act as a target for antibody-mediated demyelination has been demonstrated both in vitro (Linington et al., 1988) and in viva (Schluesener et al., 1987; Lassmann et al., 1988; Linington et al., 1988). Whether anti-MOG antibodies occur in MS has not previously been reported. We have adopted a sensitive avidin-biotin enzyme-linked immunosorbent assay (AB-ELISA) to determine anti-MOG IgG antibodies, and present here results from examination of cerebrospinal fluid (CSF) and plasma from patients with MS and controls. Materials and methods

Patients CSF and plasma were obtained from 30 patients (19 females) with clinically definite MS. Their age was 21-69 years (mean 46). None of the patients had been treated with immunomodulatory drugs. Routine CSF studies revealed monanuclear pleocytosis (> 5 x 106/liter) in 14 of the MS patients, slightly elevated CSF/serum plasma albumin ratio (Tibbling et al., 1977) reflecting damage to the blood-brain barrier in five of them, and elevated (> 0.7) IgG index (Link and Tibbling, 1977) in 18. All MS patients had oligoclonal IgG bands in their CSF demonstrable by agarose isoelectric focusing of unconcentrated CSF and diluted plasma, followed by protein transfer to nitrocellulose membrane, immunolabelling, avidin-biotin amplification and peroxidase staining (Olsson et al., 1984).

Thirty patients (13 females) had other neurological diseases (OND). These comprised cerebravascular diseases in 20 patients, aseptic meningoencephalitis in eight, neuroborreliosis in one and Behcet's disease in one patient. The age of these patients was 28-84 years (mean 55). Thirty subjects (17 females) had tension headache (TH). Their age varied between 19 and 80 years (mean 41). Neurological examination was normal in all these subjects, as were routine CSF findings (see above).

Preparation of antigen MOG was prepared from human brain white matter according to a procedure previously described (Gunn et al., 1989). Purity of MOG was assessed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and ELISA, and specificity determined by Western blot immunostaining with the monoclonal antibody 8-18C5. Purified MOG migrated on SDSPAGE as four bands with approximate molecular weights of 25 kDa, 28 kDa (minor components), 54 kDa and 62 kDa (major components). The 54 kDa and 62 kDa bands represent dimers of the 25 kDa and 28 kDa MOG components, respectively (Xiao et al., manuscript in preparation). Purified MOG did not react with antisera to myelin basic protein (MBP), proteolipid protein (PLP) and myelin-associated glycoprotein (MAG) when examined by ELISA and immunodot blot. Protein concentration of the MOG preparation was determined by a protein-dye binding method (Bradford, 1976). Determination of anti-MOG antibodies by avidinbiotin ELISA Anti-MOG antibodies of the IgG isotype were measured using a modified ELISA with increased sensitivity by including avidin-biotin amplification (AB-ELISA). Briefly, 96-well PVC microtiter plates (Dynatech, Chantilly, VA, U.S.A.) were coated with 100/~1 of a 0.5/xg/ml MOG solution in phosphate-buffered saline, pH 7.4 (PBS) for 2 h at 37°C. Unreactive sites were saturated with 150 t~l of 10% fetal calf serum (Gibco, Paisley, U.K.) in carbonate buffer, pH 9.6, for 2 h at 37°C, and at 4°C overnight. 100 /~1 of CSF and plasma, diluted 1:5 and 1:200, respectively, with 1% bovine serum albumin (BSA) and 0.05% Tween-20

93

in PBS was added to each well and kept at room temperature for 60 rain. After washing, 100 #1 of biotinylated goat anti-human IgG (Vector Lab., Burlingame, CA, U.S.A.), diluted 1/6000 in 3% BSA was added and kept for 2 h at room temperature. BSA was included to decrease non-specific binding. The wells were then incubated with 100 #1 of alkaline phosphatase-conjugated avidin D (Vector), and diluted 1/3000 in 3% BSA for 30 min at 37 ° C. Reaction products were visualized with p-nitrophenyl phosphate (Sigma, St. Louis, MO, U.S.A.) as substrate and read at 405 nm. Results obtained for anti-MOG antibodies were compared to data from wells coated with 10 # g / m l BSA only to control for non-specific binding. Background readings were obtained from wells without CSF present. The limit of detection of IgG in AB-ELISA was approximately 550 pg IgG at an optical density (OD) of 0.1 (Fig. 1). All CSF and plasma samples were divided into three groups by separate randomization schedule, and determinations were made in duplicate and doubleblind. In preliminary tests which were done repeatedly on five individual normal human serum samples, we compared the sensitivity between AB-ELISA and conventional ELISA carried out by alkaline phosphate-conjugated anti-human IgG (Sigma) instead of avidin-biotin amplification. For AB-ELISA, the mean standard deviation of OD for the normal sera was 0.033, and the correlation of variation (CV) was 3.13. Corresponding figures

for conventional ELISA were 0.015 and 4.06, respectively. As is apparent from Fig. 1, sensitivity of AB-ELISA for detection of anti-MOG IgG antibodies was higher. Western blot immunostaining was performed according to Laemmli (1977) and Towbin et al. (1979). MOG was separated by SDS-PAGE on 12.5% gel. After electrophoresis, protein was electrophoretically transferred to nitrocellulose for 4 h at 0.4 mA. The nitrocellulose was then blocked for 2 h at room temperature in 5% BSA-PBS solution. The blot was washed 3 times with PBS, and incubated overnight at room temperature with unconcentrated CSF, 1/100 diluted plasma and 1/50 diluted mouse monoclonal antibody 8-18C5 as positive control, in 1% BSA/PBS solution. The blot was incubated with goat anti-human IgG and sheep anti-mouse IgG Fab fragment conjugated to horseradish peroxidase (Sigma) diluted 1/1000 in PBS. Finally, antibody binding bands were observed by adding 9-ethyl-3-aminocarbazole and H202 as substrate. Owing to the extensive aggregation of MOG, molecular weight increased with time. Western blot bands were mainly located at around 100 kDa and 120 kDa areas after 2-4 months (Xiao et al., manuscript in preparation). Comparisons were made with the MannWilcoxon rank test, and Spearman's rank correlation coefficients were calculated.

Results 1.5

1.0

t$ d 0.5

/

0 0.45

t

1.4

4.2

12.6

37

111

333

1000

Ig(; (ng/ml) Fig. 1. Comparison of sensitivity between avidin-biotin ELISA (filled circles) and conventional ELISA (open circles) for detection of IgG in h u m a n normal sera. Arrows indicate mean a m o u n t of IgG detected when the O D at 405 n m is 0.1.

Based on the results obtained from the determination by AB-ELISA of anti-MOG IgG antibodies in CSF and plasma from 30 patients with tension headache, we considered the mean value for OD at 405 nm + 3 SD to represent the upper reference limit. For CSF, this upper reference limit is 0.076, and for plasma 0.243. In MS, seven of the 30 patients had anti-MOG antibodies in CSF. Among the patients with OND, the one with acute encephalitis and one with cerebrovascular disease had anti-MOG antibodies, as had one of the 30 subjects with tension headache (Fig. 2). The patients with MS had higher antiMOG IgG antibody titers in CSF in comparison with the subjects with tension headache ( p < 0.01) or those with OND (p < 0.05), while there was no

94 1.0

TABLE 2 TITERS OF A N T I - M O G ANTIBODIES D E T E R M I N E D BY A V I D I N - B I O T I N ELISA IN THOSE TEN PATIENTS W H O WERE POSITIVE IN CSF F O R THESE ANTIBODIES (OD > 0.076, SEE MATERIALS A N D METHODS), A N D RESULTS FROM WESTERN BLOT 1N I M M U N O S T A I N I N G P E R F O R M E D ON U N C O N C E N T R A T E D CSF

0,9 0.8 0.7 E O.6

P

~. 0.5 e~

6 0.4 0.3

~o 0.2 0.1 0

N MS

ONO

Patient No.

Diagnosis

Antibody titer (OD)

Western blot

1 2 3 4 5 6 7 8 9 10

MS MS MS MS MS MS MS Encephalitis Cerebrovascular disease Tension headache

0.502 0.257 0.237 0.235 0.231 0.172 0.135 0.960 0.566 0.227

+ + + + + + -

TH

Fig. 2. Distribution of anti-MOG IgG antibody titers in CSF determined by avidin-biotin ELISA. The shaded boxes represent ranges of OD at 405 nm up to 0.076 which we consider negative, and the numbers represent the patients being negative among the 30 patients with MS, other neurological diseases (OND) and tension headache (TH), respectively.

difference between O N D and tension headache (Table 1). The specificity of positivity for antiM O G antibodies in several anti-MOG antibodypositive CSF specimens was confirmed by Western blot immunostaining (Table 2), which showed the same staining pattern as the mouse monoclonal antibody 8-18C5 (Fig. 3). When we examined the CSF specimens positive for anti-MOG IgG antibodies by Western blot immunostaining,

utilizing unconcentrated CSF (Table 2), we also observed that the intensity of the antibody response in Western blot was closely related with the antibody titers registered by AB-ELISA. CSF from five patients (three MS, one OND, one TH) negative for anti-MOG antibodies in ELISA were also negative in Western blot. Thus, we consider that the antibodies which we have determined by AB-ELISA are specifically recognizing the M O G molecule. There was no correlation between levels of IgG in CSF and titers of anti-MOG IgG antibodies. All corresponding plasma specimens were negative for anti-MOG IgG antibodies when examined by AB-ELISA. None of 12 randomly sampled

TABLE 1 A N T I - M O G A N T I B O D Y TITERS IN CSF F R O M PATIENTS W I T H MULTIPLE SCLEROSIS (MS), O T H E R N E U R O L O G I CAL DISEASES (OND) A N D T E N S I O N H E A D A C H E (TH) Upper reference value for OD4o5 (SD + 3 SD) is 0.076. Diagnosis

No. of patients

Range

Mean +_ SD (OD)

No. positive (%)

p value

MS OND TH

30 30 30

0.007-0.502 0 -0.960 0 -0.227

0.081 +_ 0.109 0.075 + 0.194 0.025 4-_0.017

7 (23%) 2 (7%) 1 (3%)

< 0.01 a < 0.05 c > 0.05 b

a M S vs. TH. b TH vs. OND. c O N D vs. MS.

95

(kD)

....

94 76

:1=

43

i

:~1=

30 20

1

2

3

4

Fig. 3. Specificstaining of anti-MOG IgG antibodies by Western blot. (1) Molecular weight standard (Pharmacia, Sweden); (2) mouse monoclonalantibody 8-18C5 as positivecontrol; (3) CSF which is positivefor anti-MOG IgG antibodies in avidinbiotin ELISA; (4) CSF negative for anti-MOG antibodies in avidin=biotin ELISA. Arrowsindicate immunostainingbands.

plasma specimens (four MS, four O N D and four TH) showed positive reactivity to MOG when examined by Western blot.

Discussion

Using a sensitive ELISA, we have for the first time demonstrated that low titers of anti-MOG IgG antibodies are present in the CSF of patients with clinically definite untreated MS. Antibody specificity was confirmed by Western blot immunostaining. Presence of anti-MOG antibodies is not limited to patients with MS, but the antibody titers were higher in MS compared to O N D or tension headache. Anti-MOG IgG antibodies were detected in CSF only, indicating that they were probably synthesized intrathecally. A major but unproven hypothesis regarding the pathogenesis of MS is that autoaggressive T cells with capacity to react with white matter antigens migrate into the CNS and initiate the inflammatory reaction that results in demyelination. In

EAE, MBP and PLP represent autoantigens for T cell attack, but other CNS myelin proteins might be targets as well. EAE is T cell mediated, and passive transfer with MBP-reactive T cell lines has been well documented (Ben-Nun et al., 1981). However, an antibody response to a myelin component could contribute to the development of the disease and lead to extensive primary demyelination (Linington et al., 1988). Indirect evidence suggests that the immune response against CNS antigens others than MBP and PLP may play an additional role in the pathogenesis of EAE, especially by augmenting demyelination (Linington et al., 1988). Based on its exposed localization (Brunner et al., 1989) and strong immunogenic properties compared to MBP (Lebar et al., 1986), M O G fulfills two crucial requirements for an antigen that may be target for antibody-dependent autoimmune demyelination. In chronic relapsing EAE, there is also good correlation between antiM O G antibody titers and in vivo demyelinating activity (Linington and Lassmann, 1987) and, furthermore, an anti-MOG monoclonal antibody mediates demyelination in vivo (Schluesener et al., 1987; Lassmann et al., 1988; Linington et al., 1988). Our present results provide direct evidence that an anti-MOG IgG antibody response may also occur in MS, although the role of this response in the pathogenesis of demyelination in this disease remains unsettled. We found anti-MOG antibodies in CSF from one patient with encephalitis and another with cerebrovascular disease. It is known that both encephalitis and acute ischemia can cause axonal stasis, swelling, attenuation and secondary demyelination (Nukada and Dyck, 1987), which may lead to a secondary immune response to myelin antigens. It is not excluded that the anti-MOG antibodies which we detected in MS patients also reflect an immune response secondary to demyelination. Utilizing an immunospot assay which makes possible the detection of antibody secretion at the level of individual cells, we have recently presented evidence that most patients with MS have a B cell response to MBP and M A G (Link et al., 1989; Olsson et al., 1990). This autoantibody response is strongly compartmentalized to the CSF. The present data indicate that even M O G belongs

96

to the spectrum of autoantigens which induce an autoimmune response in MS. The anti-MOG antibodies which we detected in CSF from about a fourth of our MS patients could represent antibodies which are not taken up locally by myelin, indicating that they do not recognize a major pathogenic epitope. In conclusion, we have shown presence of antiMOG IgG antibodies in CSF of MS patients, less frequently and at lower levels in patients with O N D or tension headache. Anti-MOG antibodies could not be detected in the patients' plasma. MOG may be a target antigen for immune attack in MS, and anti-MOG antibodies may be important in the pathogenesis a n d / o r development of demyelination.

References Ben-Nun, A., Otmy, H. and Cohen, I.R. (1981) Genetic control of autoimmune encephalomyelitis and recognition of the critical nonapeptide moiety of myefin basic protein in guinea pigs are exerted through interaction of lymphocyte and macrophage. Eur. J. Immunol. 11, 311-316. Bradford, M. (1976) A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-254. Brunner, C., Lassmann, H., Waehneldt, T.V., Matthieu, J.M. and Linington, C. (1989) Differential ultrastructural localization of myelin basic protein, myelin oligodendroglial glycoprotein, and 2',3'-cyclic nucleotide 3'-phosphodiesterase in the CNS of adult rats. J. Neurochem. 52, 296-304. Fredrikson, S. and Kam-Hansen, S. (1989) The 150-year anniversary of multiple sclerosis: does its early history give an etiological clue? Perspect. Biol. Med. 32, 237-243. Gunn, C., Suckling, A. and Linington, C. (1989) Identification of a common idiotype on myelin oligodendroeyte glycoprotein-specific autoantibodies in chronic relapsing experimental allergic encephalomyelitis. J. Neuroimmunol. 23, 101108. Lassmann, H. (1983) A Comparative Neuropathology of Chronic Relapsing Experimental Allergic Encephalomyelitis and Multiple Sclerosis. Neurology Series, Springer Verlag, Berlin. Lassmann, H., Brunner, C., Bradl, M. and Linington, C. (1988) Experimental allergic encephalomyelitis: the balance between encephalitogenic T lymphocytes and demyelinating antibodies determines size and structure of demyelinated lesions. Acta Neuropathol. 75, 566-576.

Lebar, R., Lubetzki, C., Vincent, C., Lombrail, P. and Boutry, J.M. (1986) The M2 autoantigen of central nervous system myelin, a glycoprotein present in oligodendrocyte membrane. Clin. Exp. Immunol. 66, 423-443. Linington, C. and Lassmann, H. (1987) Antibody responses in chronic relapsing experimental allergic encephalomyelitis: correlation of serum demyelinating activity with antibody titre to MOG. J. Neuroimmunol. 17, 61-69. Linington, C., Webb, M. and Woodhams, P.L. (1984) A novel myelin associated glycoprotein defined by a mouse monoclonal antibody. J. Neuroimmunol. 6, 387-396. Linington, C., Bradl, M., Lassrnann, H., Brunner, C. and Vass, K. (1988) Augmentation of demyelination in rat acute allergic encephalomyelitis by circulating mouse monoclonal antibody directed against a myelin oligodendrocyte glycoprotein. Am. J. Pathol. 130, 443-454. Link, H. and Tibbling, G. (1977) Principles of albumin and IgM analyses in neurological disorders. III. Evaluation of lgG synthesis within the central nervous system in multiple sclerosis. Scand. J. Clin. Lab. Invest. 37, 397-401. Link, H., Baig, S., Jiang, Y.-P., Olsson, O.~ HtSjeberg, B., Kostulas, V. and Olsson, T. (1989) B cells and antibody in MS. Res. Immunol. 140, 219-226. McFarland, H.F. and Dhib-Jalbut, S. (1989) Multiple sclerosis: possible immunological mechanism. Clin. Immunol. Immunopathol. 50, 96-105. Nukada, H. and Dyck, P.J. (1987) Acute ischemia causes axonal stasis, swelling, attenuation, and secondary demyelination. Ann. Neurol. 22, 311-318. Olsson, T., Kostulas, V. and Link, H. (1984) Improved detection of oligoclonal IgG in cerebrospinal fluid by agarose isoelectric focusing, double-antibody peroxidase and avidin-biotin amplification. Clin. Chem. 30, 1246-1249. Olsson, T., Baig, S~, Ht~jeberg, B. and Link, H. (1990) Antimyelin basic protein and anti-myelin antibody producing cells in multiple sclerosis. Ann. Neurol. 37, 132-136. Satoh, J., Sakai, K., Endoh, M., Koike, F., Kunishita, T., Namikawa, T., Yamamura, T. and Tabira, T. (1987) Experimental allergic encephalomyelitis mediated by murine encephalitogenic T cell lines specific for myelin proteolipid apoprotein. J. Immunol. 138, 179-184. Schluesener, H., Sobel, R., Linington, C. and Weiner, H. (1987) A monoclonal antibody against a myelin oligodendrocyte glycoprotein induces relapses and demyelination in CNS autoimmune disease. J. Immunol. 139, 40164021. Tibbling, G., Link, H. and (~hman, S. (1977) Principles of albumin and IgG analyses in neurological disorders. I. Establishment of references values. Scand. J. Clin. Lab. Invest. 37, 385-390. Waksman, B.H., Porter, H., Lees, M.D., Adams, R.D. and Folch, J. (1954) A study of the chemical nature of components of bovine white matter effective in producing allergic encephalomyelitis in the rabbit. J. Exp. Med. 100, 451-471.