Seric immune complexes in multiple sclerosis do not contain MBP epitopes

Brain Research B~lle~~~,

Vol. 30, pp. 365-368, Printed in the USA. At1rights reserved.

036 I -9230,‘93

1993

WI0

+ .oO

Copyright0 1993PergamonPress Ltd.

Serb Immune Complexes in Multiple Sclerosis Do Not Cousin MBP Epitopes MICHEL GEFFARD*‘,

ANNE BOULLERNE*

AND BRUNO BROCHETt

~Laboratoire d~r~munologie et de Pat~oIogie, Universitt! de Bordeaux II, BP M-33076, Bordeaux C&dex, France jService U~iver~~taire ~eMrologi~ue, capital Peliegrin, Bordea~, France GEFFARD, M.; A. BOULLERNE AND B. BRGCHET. Seric immune compIexes in Multiple Sclerosis do not contain MBP epitopes.BRAIN RES BULL 30 (3/4) 365-368, 1993.-Immune complexes from sera of MS patients, other neurological diseases, and healthy donors were precipitated using polyethyleneglycol and analyzed by sodium dodecylsulfate-polyacrylamide gel electrophoresis. Silver staining evidenced additional protein bands whose molecular weights were 14-16, 38, and 43 kDa. These IC proteins were present in most MS patients studied. To identify their nature, immunoblotting was performed with antihuman immunoglobulins A, M, G antibodies. No immunoreactivity was found below a molecular weight of 66 kDa on a nitrocellulose sheet having the transferred protein pattern of MS IC. Using purified human myelin, MS IC transferred to an immobilon sheet and antihuman myelin basic protein antibodies, an immunoreactivity was seen only on purified human MBP. The small proteins of 14-16 kDa and the others of 38,43 kDa were not immuno~a~ive. Identifi~tion of the nature of these additional proteins in MS IC is in progress. Multiple sclerosis Immunostaining

Immune complex

Myelin basic protein

Protein pattern

THE etiology of Multiple Sclerosis (MS) remains unclear, aithough accumulating evidence suggests that autoimmune processes against central nervous system (CNS) myelin may be triggered by an environmental factor, possibly a virus (3). Studies on autoimmunity have mainly concerned protein and glycolipide components of myelin. Antibodies (Ab) raised against MBP (26), myelin-associated glycoprotein (MAC) (17), myelin lipids (2,8) and immune complexes (IC) (4-7,10,12,20,23-25) have been found in sem and spinal fluid of MS patients. Some components of circulating IC have already been characterized in MS patients (4,5,7,17,24,25). Although autoimmune processes seem to be res~nsible for the genesis of h~o~tholo~c lesions in MS, it is impossible today to incriminate any one antigen and to determine if dysimmunity is the origin or the outcome of other viral or toxic mechanisms. An electrophoretic approach to IC has enabled us to isolate circulating proteins of a possible exogenous or endogenous nature. Our aim was to identify specific proteins in order to characterize myelin epitopes in the demyelinating processes. METHOD

Electrophoresis

either group received corticosteroids before sampling. Nine patients in group 1 and six in group 2 were under long-term immunosuppressive therapy by oral azathioprine. Sera of 10 healthy blood donors (group 3) and 10 patients (group 4) with other neurological diseases (stroke, Guillain Barre; Syndrome, spinal cord injury patients in rehabilitation center for sequellae) were used as controls. Isolation of Circulating IC and Sample Preparation According to a previous report (17) 400 ~1 of 7% PEG (Merck-MW 60~) solution in distilled water were added to 400 pl of serum from each individual. The mixture was incubated overnight at 4°C. Precipited IC were isolated by centrifugation at 5,000 rpm for 1 h (4’C) and resuspended to the starting volume in the following buffer: 25 ml of 0.5 M Tris-HCl qsp 10 ml H20 (pH 6.8). Protein concentration of IC solutions was determined according to the method of Bradford (1). IC were then diluted (v/v) in the sample buffer (2.5 ml of 0.5 M TrisHCI buffer, pH 6.8, 10% (w/v) saccharose, 12% (w/v) SDS qsp 10 ml H20) for polyac~lamide gel electrophoresis.

Patients

Polyacrylamide Gel Electrophoresis (SDS-PAGE)

In all patients studied, MS was classified as definite, according to the criteria of Poser et al. (19). Sera from 16 patients with remitting relapsing MS in acute relapse (RRMS-AR) (group 1) and 9 patients with remitting relapsing MS (RRMS) in remission (group 2) were studied. All patients in group 1 had experienced new clinical symptoms in the previous 14 days. No patients in

Electrophoresis was performed according to the anodic discontinuous buffer method (14) using a LKB 200 1 vertical slab unit. Stacking gel (3.5% polyacrylamide, 0.5 M Tris-HCl, pH 6.8, added with 0.1% SDS) and resolving gel (lo%-15% polyacrylamide-gradient, 0.1% SDS, 3.5 M Tris-HCI, pH 8.8) were 0.75 mm thick. Electrophoresis (3.5 M Tris-glycine, pH 8.3,

’ Requests for reprints should be addressed to Michel Geffard, Laboratoire d’Immunologie et Pathologie. Universite de Bordeaux II, BP 66, 146 rue Leo Saignat, 33076 Bordeaux Cedex, France.

365

366

GEFFARD,

0.1% SDS) was carried out for 3-4 h at 22 mA constant current. At the end of the electrophoretic run, proteins were silver-stained according to the method of Merril(16). Before immunoblotting, proteins were immediately transferred to nitr~ellulose paper (pore size: 0.20 pm) or PDVF immobilon using a LKB 2005 Transphor electroblotting unit according to the procedure in the LKB laboratory manual (Tris-glycine buffer. with 20% methanol on anodic side. constant current: 0.8 mA/cmZ).

/mmunctblotting Prowdure (Western blotting) W’ith Rabbit Ah to Human Ig A SDS-PAGE was brooked with IC of 4 patients from group 1 and one from group 3. After 1 hour and a half transfer, the nitrocellulose paper was first washed with Tris-buffered saline (TBS) (0.05 mM Tris, 0.15 M NaCl) pH 7.3, then incubated in TBS plus 0.5% low-fat dried milk (05% LFDM-TBS) for lh at 37°C under agitation in order to block free sites. Nitrocellulose was immersed for 16 h at 4°C under agitation with antihuman Ig G, M, A antibodies in 0.5% LFDM-TBS. After this step, the nitrocellulose was washed twice in TBS-tween and TBS and immersed for 1 h at 37°C under agitation in biotinylated swine Ig to rabbit Ig (Dakopatts) diluted at l/lO~ in 0.5% LFDM-TBS. After 2 washings with TBS-tween and TBS, the nitrocellulose was immersed for one hour and a half at room temperature in peroxidase-conjugated streptavidine (Jackson Immuno Research) diluted at l/10,000 in 0.5% LFDM-TBS. The last 2 washings were performed with TBS alone. Peroxidase was assessed by incubating the nitrocellulose in 150 ml of solution containing TBS and 90 mg 4-chloro- 1-naphtol dissolved in separate 15 ml methanol; 180 11 30% H202 (Merck) were added just prior to use. Immrrnoblotting Procedure (Western Blotting} With Rabbit Ah to Human MBP

BOULLERNE

AND

BROCHEl

different groups of patients and controls (Fig. I). When protein patterns were compared. clear differences were noticed. Three series of additional proteins became evident. Several bands ranging from 14 to 16 kDa were found in 1I sera from group 1, 8 from group 2. none from group 3. and 4 sera from group 4 (all with Guillain Barre Syndrome). A band migrating to about 38 kDa was found in 14 sera from group I, in all sera from group 7. in 1 serum from group 3. and 0 from group 4. A band migrating about 43 kDa was well-stained in 10 sera from group I. in 2 sera from group 2. in 4 sera from group 3.

and in I serum from group 4 (spinal cord injury). To demonst~te whether additional bands found at 14-16, 38 and 43 kDa were not co-precipited with PEG 7% we used the following method. SDS untreated IC solution was filtered through a cutting membrane at 100 kDa. The filtrate analysed by SDS-PAGE did not show additional bands at I4- I6,38, and 43 kDa. Conversely, SDS treated filtrate showed these additional bands. We attempted to identify the nature of these additional protein bands. Four IC proteins and one control IC protein were transferred to a nitroceilulose sheet and antihuman Ig G, M, .A antibodies were applied on the sheet. No immunoreactiv~ty was seen below 66 kDa (Fig. 2A), indicating that these components were not related to human immunoglobulins. Then, antihuman MBP antibodies were applied to an immobilon sheet having the same 6 IC samples (5 MS; I control) and purified human myelin (Fig. 28). No immunoreactivity was found in MS and control IC. Lane 7 (Fig. 2B) shows an immunoreactive band corresponding to human MBP. Moreover, MBP was less silver stained than proteins at 14-l 6 kDa, and MBP was well stained with Coomassie blue. Conversely, all additional proteins were stained to the same degree with both coloration techniques (data not

A SDS-PAGE was performed with IC of 5 patients from group 1, one from group 3 and purified human myelin (18). Before transfer the immobilon paper was soaked in methanol, then washed in Tris-glycine buffer for 5 mn. After 35 mn transfer the immobilon was dried, then soaked in methanol and washed in water for 5 min. The immobilon was incubated in 2% HSATBS for 30 mn at 37°C under agitation in order to block free sites. It was then immersed for 16 h at 4°C under agitation with antihuman PBM antibodies (9) at l/ 1000 in 2% HSA-TBS. After this step, it was washed again twice in TBS-tween and TBS, then immersed for one hour and a half at 37°C under agitation in biotinylated swine Ig to rabbit Ig (Dakopatts) diluted at l/l000 in 2% HSA-TBS. After 2 washings with TBS-tween and TBS. the immobilon was immersed for 1 hour and a half at room temperature in ~roxidase-conjugated ~reptavidine (Jackson Immuno Research) diluted at l/8,000 in 2% HSATBS. The last 2 washings were performed with TBS alone. Peroxidase was assessed by incubating the immobilon in 200 ml TBS containing 120 mg 4-chloro- I-naphtol dissolved in separate 20 ml methanol ; 240 ~1 30% H202 (Merck) were added just prior to use. Intertest Rel~abi~~t~ ~11 tests were performed at least twice to assess their repro-

ducibility. RESULTS

SDS-PAGE illustrated by silver-staining gels enabled us to visualize protein components involved in the IC found in the

FIG. 1. Silver staining of polyacrylamide get with 7Fg per lane of SDStreated proteins of PEG precipited circulating human IC. Lane I: IC from healthy donor. Lane 2-10: 9 samples of IC from MS patients in acute relapse (a, b, c, d, e, f, g, h, i). A: Pharmacia molecular weight marker (14.4; 20.1; 30; 45; 67; 94 kDa). Symbols location of 14-16 kDa proteins: location of 38 kDa protein: location of 43 kDa protein.

IMMUNE COMPLEXES

IN MULTIPLE

SCLEROSIS

367

to characterize the proteins in these IC. Using isolation of IC by adsorption to Ragi cells, Dasgupta et al. (6) found that IC in 11 sera from 22 MS patients contained MBP and Coyle (5) also found IC containing MBP in MS patients. To confirm these data, we used isolation of IC by PEG precipitation, analysis by SDS-PAGE, and immunoblottings. Seric IC from patients with MS contained proteins migrating to 1416, 38 and 43 kDa. Immunoblotting with antihuman Ig (Ig A, Ig G, Ig M) antibodies showed that these proteins were not Ig or Ig fragments. The molecular weights found of 14- 16 kDa could correspond to a human myelin protein. The molecular weight of the two dominant isoforms of human MBP are 17.2 and 2 1.5 kDa (22). This raises the hypothesis that the small bands we observed between 14 and 16 kDa correspond to MBP degradation products or MBP like material. This hypothesis was tested by immunoblotting; no immunoreactivity for MBP was found in these IC. These results show that 14- 16 kDa proteins do not contain epitopes recognized by a rabbit antihuman MBP serum. Our results suggest that proteins included in seric IC of MS contain neither MBP degradation products, MBP peptides, or exogenous (for

FIG. 2A. Immunoperoxidase staining of a nitrocellulose sheet with SDS treated proteins of PEG precipited circulating human IC (20 &well in

polyacrylamide gel) after incubation with rabbit antihuman Ig G antibodies. Lane 1: IC from healthy donor. Lane 2-5: 4 samples of IC from MS patients in acute relapse (a, d, h, i). A: Biorad biotinylated molecular weight marker (14.4; 2 1.5; 31; 45; 66.2; 97.4 kDa). Symbol location of Ig G revealed by antihuman Ig G antibodies.

shown). This straightforward difference confirmed the absence of related MBP in these proteins. DISCUSSION

This study shows that the electrophoretic profile of proteins contained in circulating IC of patients with MS is different from those of patients with OND and healthy donors. Recent experimental data showing that humoral response plays a major role in chronic experimental demyelination (15) and evidence of a role of the complement in MS (2 1) have revived interest in the study of the humoral mechanism in MS. IC may be a useful approach in assessing the specificity of the human response in MS. Numerous studies have shown that sera of MS patients contain IC (4-7,10,12,20,23,24,25). Most drew attention to the specificity of immunoglobulins (Ig) contained in these IC (4,5,12,24,25). For instance, it has been shown that Ig included in IC from sera or spinal fluid of patients with MS bind with MBP (5), galactocerebroside, ganglioside (12) viral antigens (5), and endothelial cells (24). However, few studies have attempted

FIG. 2B. Immunoperoxidase staining of an immobilon sheet with SDStreated proteins of PEG precipited circulating human IC (20 pg/well in polyacrylamide gel) and SDS treated purified human myelin (10 &well in polyacrylamide gel) after incubation with antihuman MBP antibodies. Lane I: IC from healthy donor. Lane 2-6: 5 samples of IC from MS patients in acute relapse (a, d, e, h, i). Lane 7: purified human myelin. Symbol location of human MBP.

GEFFARD.

368

instance viral) proteins containing MBP-like epitopes. Moreover, unlike MBP proteins our bands were well revealed by silver stain. The 14- I6 kDa proteins were also found in all sera of Guillain Barre Syndrome (GBS) patients studied. IC have been found by several authors in GBS (10,23) but little is known about the characterization of the proteins involved in these IC. Antibodies against several myelin proteins (P2, PO, MAG) have been demonstrated in GBS ( 13). The 14- 16 kDa proteins may correspond to common components of central and peripheral myelin and may be secondary to myelin degradation. The 38 kDa and 43 kDa proteins seem to be more specific to the disease, because no sera of patients with OND contained 38 kDa and there was only one with 43 kDa. In seric IC of 2 out of 3 patients with

BOULLERNE

AND BROC’HE’I

MS, Wiederkehr et al. (27) recently found a similar band between 38 and 40 kDa. Characterization of 38 and 43 kDa proteins using immunoblotting procedure is underway in our laboratory. In the hypothesis of endogenous antigens, these molecular weights could be myelin Wolfgram proteins whose molecular weights are from 45 to 55 kDa (22). These additional proteins could be exogenous antigens possibly leading to retroviral proteins. ACKNOWLEDGEMENTS

We thank Michel Foumier for supplying purified human myelin and antihuman MBP antibodies. This work was supported by grants from “Ligue Francaise contre la S&rose en Plaques”.

REFERENCES 1 Bradford, M. M. A rapid and sensitive method for the quantification of microgram quantities of protein utilizing the principle of protein dve binding. Anal. Biochem. 72:248-254: 1976. 2 Brochet, B: Faiderbe, S.; Audhuy, S.: Gosset, I.; Geffard, M.; Orgogozo, J. M. Antibodies against phosphatidylinositol in multiple sclerosis. In: Wiethiilter, M.; Dichgans, J.; Mertin, J., eds. Proceedings of multiple sclerosis. Amsterdam:Elsevier Science Publishers: 199 I : 97-102. 3 Compston, D. A. S. Genetics and immunology. In: Matthews, W. B.; Compston, A.; Allen, I. V.; Martyn, C. N., eds. McAlpine’s multiule sclerosis. Edinburgh: Churchill Livingstone: I99 I:30I-339. 4. Coyle, P. K. Detection and isolation of immune complexes in multiple sclerosis cerebrospinal fluid. J. Neuroimmunol. 15:97- 107: 1987. 5 Coyle, P. K.; Procyk-Dougherty, 2. Multiple sclerosis immune complexes: An analysis of component antigens and antibodies. Ann. Neurol. 16:660-667; 1984. 6. Dasgputa, M. K.; Catz, I.; Warren, K. G.; McPherson, T. A.; Dossetor. J. B.; Carnegie, P. R. Myelin basic protein: A component of circulating immune complexes in multiple sclerosis. Can. J. Neurol. Sci. 10:239-243; 1983. I. Dasgupta, M. K.; Warren, K. G.; Johny. K. V.; Dossetor, J. B. Circulating immune complexes in multiple sclerosis: Relation with disease activity. Neurology 32: 1000-1004; 1982. 8. Daverat. P.: Geffard. M.: Orgoaozo. J.M. Identification and characterization of anticonjugated azelaic acid antibodies in multiple sclerosis. J. Neuroimmunol. 22: 129- 134; 1989. 9 Garbay, 8.; Foumier, M.; Sallafranque, M. L.; Muller, S.; Boiron, F.; Heappe, A.; Cassagne, C.; Bonnet, J. PO MBP histone and DNA levels in sciatic nerve: Postnatal accumulation studies in normal and Trembler mice. Neurochem. Pathol. 8: 91-107; 1988. 10 Goust, J. M.; Chenais, F.; Cames, J. E.; Hames, C. G.; Fudenberg, H. H.; Hogan, E. L. Abnormal T cell subpopulations and circulating immune complexes in the Guillain-Barre syndrome and multiple sclerosis. Neurology 28:42 l-425; 1978. 11 Guidi, L.; Baroni, R.; Bartoloni, C.; Pellegrino, M.; Tricerri, A.; Marciano, M.; Barone, C.; Gambassi, G. Immune complexes in solid tumours precipitable by 3,5% polyethylene glycol: Analysis of some nonspecific components. Diaanostic and Clinical Immunol. 5:284-288; 1988. _ 12 Kasai, N.; Pachner, A. R.; Yu, R. K. Anti-glycolipid antibodies and their immune complexes in multiple sclerosis. J. Neurol. Sci. 75: 33-42; 1986. 13 Koski, C. L. Characterization of complement-fixing antibodies to peripheral nerve myelin in Guillain Barre Syndrome. Ann. Neural. 27 (suppl): S44-S47: 1990.

14. Laemmli, V. K. Cleavage of structural proteins during the assembly of head bacteriophage T4. Nature 227:680-685; 1970. 15. Linington, C.; Bradl, M.; Lassman, H.; Brtmner, C.; Vass, K. Augmentation of demyelinating in rat acute allergic encephalomyelitis by circulating mouse monoclonal antibodies directed against a myelin/oligodendrocyte glycoprotein. Am. J. Pathol. 130:443-454: 1988. 16. Merril, C. R.; Goldman, D.; Van Keuren, M. L.; Gel protein stain: silver stain. Electrophoresis 3: 17; 1982. 17. Nobile-Orazio, E.; McIntosh C.; Latov N. Anti-MAG antibody and antibody complexes: detection by radioimmunoassay. Neurology 35:988-992; 1985. 18. Norton, W. T. Isolation of myelin from nerve tissue. In: Sleischer, S.: Parker, I., eds. Method in enzymology. New York: Academic Press; 1974, XxX1:435-444. 19. Poser, C. M.; Patty, D. W.; Scheinberg, L.; MacDonald, 1.; Davis, F.; Ebergs, G.; Johnson, K.; Sibley, W. A.; Silbergerg, D. H.; Tourtelotte, W. New diagnostic criteria for multiple slcerosis. Ann. Neurol. 13~227-231; 1983. 20. Rudick, R. A.; Bidlack, J. M.; Knutson, D. W. Multiple sclerosis: Cerebrospinal fluid immune complexes that bind Clq. Arch. Neurol. 42:856-858; 1985. 21. Scolding, N. J.; Morgan, B. P.; Masten, W. A.; Linington, C.; Campbell, A. M.; Compston, D. A. S. Vesicular removal by oligodendrocytes and membrane attack complexes formed by activated complement. Nature 339:620-622; 1989. 22. Stoffel, W. The myelin membrane of the central nervous system: Essential macromolecular structure and function. Angew. Chem. Int. Ed. Engl. 29:958-976; 1990. 23. Tachovsky, T. G.; Lisak, R. P.; Koprowski, H.; Theophilopoulos. A. N.: Dixon, F. J. Circulating immune complexes in multiple sclerosis and other neurological diseases. Lancet 6:997-999: 1976. 24. Tanaka, Y.; Tsukada, N.; Koh, Ch.-S.; Yanagisawa, N. Antiendothelial cell antibodies and circulating immune complexes in sera of patients with multiple sclerosis. J. Neuroimmunol. 17:49-59: 1987. 25. Wajgt, A.; Gorny, M. K.; Jenek, R. The influence of high-dose prednisone medication on autoantibody specific activity and on circulating immune complex level in cerebrospinal fluid of multiple sclerosispatients. Acta Neurol. Stand. 68:378-385; 1983. 26. Warren. K. G.; Catz, 1. Cerebrospinal fluid autoantibodies to myelin basic protein in multiple sclerosis patients. J. Neural. Sci. 91: i43151; 1989. 27. Wiederkehr, F.; Bueler, M. R.; Vonderschmitt, D. J. Analysis of circulating immune complexes isolated from plasma, cerebrospinal fluid and urine. Electrophoresis 12: 478-486; 1991.