- Email: [email protected]
Susceptibility to demyelinating polyneuropathy in plasma cell dyscrasia may be influenced by amino acid position 9 of the HLA-DRβ chain
Journal of Neuroimmunology, 43 (1993) 139-144 © 1993 Elsevier Science Publishers B.V. All rights reserved 0165-5728/93/$06.00
139
JNI 02319
Susceptibility to demyelinating polyneuropathy in plasma cell dyscrasia may be influenced by amino acid position 9 of the HLA-DR/3 chain Magnus V r e t h e m
a Jan E r n e r u d h a,b, Mabel Cruz c, Olle Olerup a,e, G6ran Solders c, Bo Ekstedt e, Oluf Andersen g and Jan Hillert c'e
Department of Neurology and b Department of Transfusion Medicine and Clinical Immunology, University Hospital, Linkb'ping, Sweden, Departments of c Neurology and d Clinical Immunology, Karolinska Institute at Huddinge Hospital, Huddinge, Sweden, e Center for Bio Technology, Novum, Karolinska Institute, Huddinge, Sweden, f Department of Neurology, Orebro Hospital, Orebro, Sweden and g Department of Neurology, University Hospital, GSteborg, Sweden (Received 9 September 1992) (Accepted 6 October 1992)
Key words: Genetic susceptibility; Monoclonal gammopathy; Polyneuropathy; HLA-D antigen; Restriction fragment length polymorphism
Summary Fifty-five patients with plasma cell dyscrasias were investigated by genomic typing for HLA-DR and -DQ genes by restriction fragment length polymorphism, neurophysiology and for presence of anti-myelin-associated glycoprotein (MAG) antibodies. In 26 patients, a polyneuropathy (PN) of demyelinating type was established. Among these individuals, an association was found with the presence of a tryptophan amino acid residue at position 9 of the DR/3 chain (P < 0.01). This position is part of the first hypervariable region of the DR/3 chain, and may be of importance in determining preferential peptide-binding capacity of the HLA-DR molecule. The presence of anti-MAG antibodies in 15 out of 17 patients with an IgM M-component and demyelinating PN (14 of these 15 individuals carrying a tryptophan at position 9) supports the pathogenic role of an autoimmune response against MAG. The finding of an HLA class II association may indicate a pathogenic role of T cell immunity in this condition.
Introduction Plasma cell dyscrasias are characterized by proliferation of a clone of plasma cells producing a homogeneous paraprotein (M-component) and are statistically associated with demyelinating polyneuropathy (PN) (Latov et al., 1980; McLeod et al., 1984; Kelly, 1985). A pathogenic role of the M-component in the development of some of these cases of PN has been suggested because: the M-component reacts with myelin-associated glycoprotein (MAG) (Braun et al., 1982; Melmed et al., 1983; Nobile-Orazio et al., 1989); direct injection of M-component containing sera causes demyelination in animals (Hays et al., 1987); immunoglobulin M (IgM) deposits have been demonstrated in peripheral nerve
Correspondence to: M. Vrethem, Department of Neurology, University Hospital, S-581 85 Link6ping, Sweden.
(Meier et al., 1983; Stefansson et al., 1983); and beneficial effect of lowered anti-myelin antibody levels by immunosuppressive treatment has been shown (Nobile-Orazio et al., 1988). There is, however, no absolute correlation between either occurrence of antibodies or lowered antibody level and clinical effect (Kelly et al., 1988; Ernerudh et al., 1992). Thus, other mechanisms may be involved and T cell regulation of B cell activity and a direct T cell involvement in the pathogenesis of demyelinating PN has also been suggested (Latov et al., 1985). Both humoral and cell-mediated immune responses may occur, suggesting an autoimmune component in the development of the demyelinating PN. Diseases with autoimmune features are frequently associated with alleles or haplotypes of the HLA class II region. As genomic methods for HLA class II typing have been developed, offering increased accuracy as well as resolution compared to serology, previously known HLA associations have been refined, for exam-
140 TABLE 1 Clinical and laboratory characteristics of patients with plasma cell dyscrasias No.
Age/Sex
Disease
Ig class
Polyneuropathy a
IgM anti-MAG ab. reactivity b
HLA-D haplotype 1
HLA-D haplotype 2
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.
51/M 85/M 66/F 63/F 82/M 50/M 72/F 70/M 75/M 74/F 74/M 77/M 82/M 72/M 66/F 68/M 60/M 71/M 85/M 43/F 63/F 63/M 67/M 62/M 68/M 74/M 76/F 62/M 85/M 61/M 62/M 71/F 65/M 69/F 72/F 71/M 77/M 75/M
MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Waldenstr6m Waldenstr6m Waldenstr6m MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Waldenstr6m Waldenstr/Sm MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Myeloma Myeloma MGUS MGUS Myeloma Myeloma MGUS MGUS MGUS MGUS MGUS MGUS MGUS Myeloma Myeloma MGUS Myeloma
IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM lgM IgM IgM IgM IgM IgG IgG IgG IgG IgG IgG IgG IgA IgA IgM IgM IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgA IgA IgA k-u IgM IgG IgG IgG IgG IgG IgG IgG IgG IgA l-u
Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal
+ + + + + + + + + + + + + + + ND ND -
DR1,DQ5 DR8,DQ4 DR1,DQ5 DR4,DQ8 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR15,DQ6 DR1,DQ5 DR17,DQ2 DRII,DQ7 DR1,DQ5 DR1,DQ5 DR7,DQ2 DR16,DQ5 DR16,DQ5 DR15,DQ6 DR17,DQ2 DR15,DQ6 DR15,DQ6 DR17 DQ2 D R l l DQ7 DRI5 DQ6 DR17 DQ2 DR17 DQ6 DR17 DQ2 DR15 DQ6 DRI DQ5 DR15,DQ6 DR15,DQ6 DRll,DR15,DQ6 DRI7,DQ2 DR17,DQ2 DR1,DQ5 DRI,DQ5 DR10,DQ5 DR4,DQ7 DRI5,DQ6 DR15,DQ6 DR15,DQ6 DR17,DQ2 DR17,DQ2 DR1,DQ5 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR1,DQ5 DR4,DQ8 DR1,DQ5 DR17,DQ2 DRI3,DQ6
DRI5,DQ6 DR7,DQ9 DR4,DQ8 DR7,DQ2 DR7,DQ2 DR7,DQ2 DR7,DQ2 DR4,DQ7 DR15,DQ6 DR15,DQ6 DR17,DQ2 DR4,DQ8 DR4,DQ8 DR13,DQ6 DR17,DQ2 DR15,DQ6 DR9,DQ9 DR13,DQ6 DR4,DQ7 DR4,DQ7 DR17,DQ2 DR4,DQ8 DR4,DQ8 DR7,DQ2 DR13,DQ6 DR11,DQ7 DR7,DQ9 DRI1,DQ7 DR12,DQ7 DR13,DQ6 DR4,DQ8 DR7,DQ9 DR15,DQ6 DR13,DR4,DQ7 DR12,DQ7 DR13,DQ6 DR1,DQ5 DR16,DQ5 DR13,DQ6 DR13,DQ6 DR15,DQ6 DR13,DQ6 DR17,DQ2 DR13,DQ6 DR4,DQ8 DR4,DQ8 DR4,DQ8 DR1,DQ5 DR17,DQ2 DR13,DQ6 DR8,DQ4 DR8,DQ4 DR4,DQ8 DR14,DQ5
77/M 58/M 80/M 76/M 75/M 72/M 60/F 74/M 43/M 61/M 76/M 64/F 57/F 58/M 69/M 51/F 73/F
+ ND ND -
a Clinical polyneuropathy was assumed to exist when both symptoms and signs of bilateral sensory or motor impairment were present in combination with hypo- or areflexia. Demyelinating polyneuropathy was defined as clinical polyneuropathy and decreased motor nerve conduction velocity to less than 60% of normal mean value in at least two motor nerves (Kelly, 1983), or demyelination in nerve biopsies. Axonal polyneuropathy was defined as clinical polyneuropathy but not fulfilling the criteria of neurophysiological demyelinating polyneuropathy, or showing mainly axonal involvement in nerve biopsy. Normal patients, e.g., those without polyneuropathy, did not have clinical polyneuropathy and showed normal neurographic findings. b +, IgM anti-myelin-associated glycoprotein (MAG) antibody reactivity > mean + 3SD of healthy controls in an ELISA assay. ND, not done.
141 TABLE 2 P h e n o t y p i c f r e q u e n c i e s ( % ) of H L A - D R - D Q h a p l o t y p e s a m o n g the total p a t i e n t g r o u p with p l a s m a cell dyscrasia as well as various s u b g r o u p s as c o m p a r e d w i t h controls HLA-DR-DQ haplotypes
All p a t i e n t s (n = 55)
Patients with P N (n = 44)
P a t i e n t s with demyelinating PN (n = 26)
P a t i e n t s of I g M isotype with demyelinating PN (n = 17)
P a t i e n t s with antiM A G reactivity and IgM demyelinating PN (n = 15)
Controls (n = 250)
DR1,DQ5 DR15,DQ6
25 38
23 43
27 46
41 a 47 c
40 b 47 d
18 30
DR16,DQ5 DR17,DQ2 DR4,DQ7 DR4,DQ8 DR7,DQ9 DR7,DQ2 DR8,DQ4 DR9,DQ9 DR10,DQ5 DRll,DQ7 DR12,DQ7 DR13,DQ6 DR14,DQ5
5 27 9 22 5 11 5 2 2 7 4 18 2
5 25 11 16 7 14 2 2 2 9 4 18 .
8 19 12 23 4 23 e 4 4 12 12
12 6 24 6 29 f 6 6 6 6
20 7 27 7 33 g 7 7 -
1 17 12 26 5 7 9 4 2 14 6 29 6
.
.
.
a p < 0.05, R R = 3.2; b p < 0.05, R R = 3.0; c p < 0.02, R R = 3.0; d p < 0.02, R R = 3.5; e p < 0.02, R R = 4.4; f P < 0.02, R R = 5.7; g P < 0.01, R R = 6.9 (for c and d, p values w e r e c a l c u l a t e d for allelic frequences, after s u b t r a c t i o n of the p r i m a r i l y a s s o c i a t e d D R 1 , D Q 5 and D R 7 , D Q 2 haplotypes).
pie in insulin-dependent diabetes mellitus (Todd et al., 1987) and immunoglobulin deficiencies (Olerup et al., 1990). The aim of the present study was to investigate the distribution of HLA-DR and -DQ alleles and DR-DQ haplotypes with a genomic typing procedure in a group of patients with plasma cell dyscrasia and demyelinating PN, also with reference to Ig isotype and presence of antibodies directed against MAG.
Materials and Methods
Patients Fifty-five patients with plasma cell dyscrasia (age 43-85 years, mean age 68 years), including 43 with monoclonal gammopathy of uncertain significance (MGUS), seven with multiple myeloma and five with Waldenstr6m's disease were included in the study. Peripheral nerve function was characterized clinically and neurophysiologically according to Kelly (1983). Eleven patients were found to have no PN, 18 were judged to have mainly axonal PN, and 26 had a mainly demyelinating PN. Of the 26 patients with demyelinating PN, 17 had M-components of IgM, seven of IgG, and two of IgA isotype. Two hundred and fifty randomly collected unrelated healthy individuals were used as controls. Results of genomic HLA class II typing in this group have previously been reported (Olerup and Hillert, 1991).
HLA class H typing Genomic typing for HLA-DR and -DQ was performed by TaqI restriction fragment length polymorphism (RFLP) analysis of the DRB, DQA and DQB genes, as previously described (Bidwell et al., 1987; Carlsson et al., 1987). To avoid local nomenclature, HLA class II haplotypes were designated by their associated serologic specificities (Bidwell et al., 1987; Carlsson et al., 1987). As TaqI RFLP does not distinguish between the haplotypes DR7,DQ9 and DR9,DQ9, this was achieved by polymerase chain reaction by sequence specific primer (PCR-SSP) analysis as described elsewhere (Zetterquist and Olerup, 1992). HLA-DR-DQ haplotypes were deduced from known linkage dysequilibria between DR and DQ alleles.
Anti-MAG antibody analysis Bovine MAG was prepared by a lithium diiodosalicylatephenol method (Quarles et al., 1983) from brain myelin (Norton and Poduslo, 1973). Anti-MAG antibodies were measured as previously described (Cruz et al., 1991), with some modifications. In short, 96 polystyrene microtiter plates (Costar, Cambridge, MA, USA) were coated with 100/zl of MAG (1 /~g/ml). Wells were blocked with 200 p~I/well of phosphate-buffered saline (PBS) containing 2% bovine serum albumin (BSA) (Sigma, St. Louis, MO, USA). 100 /xl of serum samples adjusted to 1 mg/l of IgM were incubated overnight at 4°C. After washing, alkaline phosphatase conjugated rabbit anti-human IgM
142 (my-chain specific) (Dako, Copenhagen, Denmark) was added. After 2 h incubation, plates were washed and incubated with 100 p.1 per well of substrate pnitrophenylphosphate (1 m g / m l ; Sigma) in 10% diaethanolamine and 0.02% MgC1 buffer, p H 9.8. Plates were read after 30 min in an automated E L I S A spectrophotometer at 405 nm. Pooled normal serum (used as intra-assay reference) did not give absorbance values above 0.08. Samples were considered positive when the absorbance values were higher than the mean value + 3 SD of 20 blood donors. The range of absorbance values for these sera was 0-0.11 (mean 0.042) for IgM, yielding a mean value + 3 SD of 0.14 for IgM.
Statistical analysis Comparisons between patients and controls were performed using the chi-square test. All P values are given without correction, but the large number of comparisons made in this study should be borne in mind. However, the group of primary interest was that of patients with demyelinating PN and an M-component of IgM isotype, i.e., with a disease of most probably autoimmune origin.
Results The distribution of H L A - D R - D Q haplotypes among patients and controls is shown in Tables 1 and 2. In the total group of patients with M-components, no significant differences were seen. However, in the group of patients with IgM gammopathy and a demyelinating PN (n = 17), increased frequencies were found for the DR1,DQ5 ( P < 0 . 0 5 , relative risk ( R R ) = 3 . 2 ) and D R 7 , D Q 2 ( P < 0.02, R R = 5.7) haplotypes. Eleven out of 17 (65%) patients in this group carried one or both of these haplotypes, as compared to 24% among controls ( P < 0.01, R R = 5.9). In addition, there was a moderate increase of the DR15,DQ6 haplotype ( P < 0.20, R R = 2.1) (DR15 is a split of DR2). After subtraction of the two primarily associated haplotypes from patients and controls (according to the method described by Payami et al., 1989), the allelic frequency of DR15,DQ6 was found to be significantly increased
( P < 0.02, R R = 3.0). Similar results were obtained in patients with anti-MAG antibodies, since this group was essentially identical to that of IgM M-components with demyelinating PN. Analysis of common features of the three haplotypes DR1,DQ5, DR15,DQ6 and D R 7 , D Q 2 reveals similarities in the first hypervariable region of the DR/3 chain. In particular, these haplotypes differ, together with the DR16,DQ5 and D R 7 , D Q 9 haplotypes, from all other haplotypes in carrying a non-polar tryptophan residue at position 9, where other DR/3 alleles carry a charged glutamic acid residue. Table 3 shows the phenotypic frequency of a DR/3 position 9 tryptophan in the various subgroups. Significant differences compared to the controls were seen for all patients with plasma cell dyscrasia and demyelinating PN as well as for those with IgM isotype. Notably, a similarly increased relative risk, however not statistically significant, was seen among the small group of nine patients with demyelinating PN and M-component of IgG or IgA isotype. When comparing patients with demyelinating PN with individuals either without PN or with an axonal PN (supposedly not of autoimmune origin) there was a statistically significant increase of tryptophan at position 9 ( P < 0.03). We found a strong correlation between demyelinating PN of IgM isotype and anti-MAG antibodies (15 out of 17 patients). For patients with axonal PN or without PN, no difference from the controls was observed in either H L A class II haplotypes or in the presence of anti-MAG antibodies.
Discussion An increasing amount of evidence indicates that demyelinating PN associated with an M-component may be of autoimmune origin. Also, genetic factors have been suspected of playing a role in the susceptibility to this condition and familial cases have been reported (Busis et al., 1985; Jensen et al., 1988). l g G heavy-chain (Gm) allotypes have been reported to differ significantly between patients and healthy controls (Kahn and Pandey, 1987). Furthermore, an increase of
TABLE 3 Frequency (%) of carriers of the amino acid tryptophan at position 9 of the DR~ chain in the different subgroups of patients with plasma cell dyscrasia as compared with controls All patients (n = 55)
Patients with PN (n = 44)
Patientswith demyelinatingPN (n = 26)
71 P < 0.05 RR = 1.9
68 n.s. RR = 1.7
85 P < 0.01 RR = 4.4
Patients of IgA and IgG isotype and demyelinating PN (n = 9) 78 n.s. RR = 3.6
Patients of IgM isotype with demyelinatingPN (n = 17) 88 P < 0.02 RR = 6.0
Patients with antiMAG reactivity and IgM demyelinating PN (n = 15) 93 P < 0.03 RR = 11.0
Controls (n = 250)
56
143
class II antigen expression on Schwann cells has been shown (Mitchell et al., 1991). To our knowledge genomic HLA class II typing has not been performed in these patients, although susceptibility to autoimmune diseases is commonly associated with these genes. In the present study of 26 patients with demyelinating PN, the results indicate a role of a specific amino acid residue of the DR/3 chain. The distribution of HLA-DR-DQ haplotypes in the whole material (n = 55) differed very little in comparison with the controls. Any subdivision of the material increases the number of comparisons and our findings in this relatively rare disorder are not strong enough to etablish this association statistically. On the other hand, associations should not be expected to be found in large heterogeneous groups of patients but in homogeneous, well-defined conditions with clear autoimmune features such as the group of patients with plasma cell dyscrasia of IgM isotype and demyelinating PN (Gosselin et al., 1991). Polymorphism of DRB genes is mostly limited to three hypervariable regions. According to models of the three-dimensional structure of the MHC class II molecules, based on the known structure of HLA class I molecules (Brown et al., 1988), these polymorphic segments make up parts of the walls of the peptidebinding groove. In this way, different alleles differ with regard to their preferential binding of different peptides, although each allelic molecule can bind many different peptides, with varying affinity. Amino acid residue 9 is part of the first hypervariable region and is believed to be located in the /3-pleated sheeth in the bottom of the peptide-binding groove of the class II molecule. Theoretically the substitution of a charged glutamic acid with a highly hydrophobic tryptophan residue at this position should alter the properties of the peptide-binding groove considerably. Previously, most evidence has pointed towards a B-cell immune response in the pathogenesis of PN in IgM plasma cell dyscrasias. Our findings, however, are also compatible with a pathogenic importance of cellmediated immunity. HLA class II associations in general suggest a role of T-cells, and the type of association found here could implicate the binding/presentation of a specific peptide. Thus, the T-cell response could well be more restricted than the B-cell response, which has been shown often to be directed against MAG (Braun et al., 1982; Melmed et al., 1983; Nobile-Orazio et al., 1989) and Po-protein (Bollensen et al., 1988), but also against various gangliosides or chondroitin sulfate not reactive with MAG (Sherman et al., 1983; Miyatani et al., 1987; Ilyas et al., 1988; Yee et al., 1989). The presence of the same HLA class II association in IgA and IgG plasma cell dyscrasias with PN but without anti-MAG specificity of the M-components, could further support this notion, as would the previous demonstration of T-cells in nerve biopsies
from similar patients (Solders et al., unpublished observations). The presence of activated T-cells is also supported by the findings of increased soluble interleukin-2 receptor levels in some patients with M-component-associated PN (Vrethem et al., in press). In conclusion, our results indicate that a genetic predisposition to demyelinating PN in plasma cell dyscrasias associated with the HLA-DR-DQ subregion might be explained by a specific amino acid residue of the DR/~ chain. Since the number of patients in this study was limited, confirmation of this hypothesis has to await the investigation of other similar patients.
Acknowledgements The work was supported by grants from OstergiStlands Liins Landsting, the Swedish Association of Neurologically Disabled, the Swedish Society of Medicine, the Swedish Medical Research Council (Project No. 8709) and the Karolinska Institute.
References Bidwell, J.L., Bidwell, E.A., Laundy, C.J., Klouda, P.T. and Bradley, B.A. (1987) Allogenotypes defined by short DQa and DQ/3 cDNA probeg correlate with, and define splits of HLA-DQ serological specificities. Mol. Immunol. 24, 513-522. Bollensen, E., Steck, A.J. and Schachner, M. (1988) Reactivity with the peripheral myelin glycoprotein Po in serum from patients with monoclonal IgM gammopathy and polyneuropathy. Neurology 38, 1266-1270. Braun, P., Frail, D. and Latov, N. (1982) Myelin-associated glycoprotein is the antigen for monoclonal IgM in polyneuropathy. J. Neurochem. 128, 1261-1265. Brown, J.H., Jardetzky, T., Saper, M.A., Samraoui, B., Bjorkman, P.J. and Wiley, D.C. (1988) A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules. Nature 332, 845-850. Busis, N.A., Halperin, J.J., Stefansson, K., Kwiatkowski, D.J., Sagar, S.M., Schiff, S.R. and Logigian, E.L. (1985) Peripheral neuropathy, high serum IgM, and paraproteinemia in mother and son. Neurology 35, 679-683. Carlsson, B., Wallin, J., B6hme, J. and M611er, E. (1987) HLA-DRDQ haplotypes defined by restriction fragment length analysis: Correlation to serology. Hum. Immunol. 20, 95-113. Cruz, M., Jiang, J.-Y., Ernerudh, J., Solders, G., Olsson, T., Osterman, P.-O. and Link, H. (1991) Antibodies to myelin associated glycoprotein are found in cerebrospinal fluid in peripheral neuropathy associated with monoclonal serum IgM. Arch. Neurol. 48, 66-70. Ernerudh, J., Vrethem, M., Andersen, O., Lindberg, C. and Berlin, G. (1992) Immunochemical and clinical effect of immunosuppressive treatment in monoclonal IgM neuropathy. J. Neurol. Neurosurg. Psychiatry 55, 930-934. Gosselin, S., Kyle, R.A. and Dyck, P.J. (1991) Neuropathy associated with monoclonal gammopathies of undetermined significance. Ann. Neurol. 30, 54-61. Hays, A., Latov, N., Takatsu, M. and Sherman, W. (1987) Experimental demyelination induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37, 242-256.
144 Ilyas, A.A., Li, S.-C., Chou, D.K.H., Li, Y.-T., Jungalwala, F.B., Dalakas, M.C. and Quarles, R.H. (1988) Gangliosides GM2, IV4GalNAcGMlb, and IV4GalNAcGDla as antigens for monoclonal IgM in neuropathy associated with gammopathy. J. Biol. Chem. 263, 4369-4373. Jensen, T.S., Schr6der, H.D., J6nsson, V., Ernerudh, J., Stigsby, B., Kamieniecka, Z., Hippe, E. and Trojaborg, W. (1988) IgM monoclonal gammopathy and neuropathy in two siblings. J. Neurol. Neurosurg. Psychiatry 51, 1308-1315. Kahn, S.N. and Pandey, J.P. (1987) IgG heavy-chain (Gm) allotypes in demyelinating polyneuropathy associated with MAG-binding monoclonal IgM autoantibodies. Ann. Neurol. 21,507-509. Kelly, J. Jr. (1983) The electrodiagnostic findings in peripheral neuropathy associated with monoclonal gammopathy. Muscle Nerve 6, 504-509. Kelly, J. Jr. (1985) Peripheral neuropathies associated with monoclonal proteins: A clinical review. Muscle Nerve 8, 138-150. Kelly, J.J., Adelman, L.S., Berkman, E. and Bhan, I. (1988) Polyneuropathies associated with IgM monoclonal gammopathies. Arch. Neurol. 45, 1355-1359. Latov, N., Sherman, W., Nemni, R., Galassi, G., Shyong, J., Penn, A., Chess, L., Olarte, M., Rowland, L. and Osserman, E. (1980) Plasma cell dyscrasia and peripheral neuropathy with a monoclonal antibody to peripheral nerve myelin. New Engl. J. Med. 303, 618-62l. Latov, N., Godfrey, M., Thomas, Y., Nobile-Orazio, E., Spatz, L., Abraham, J., Perman, G., Freddo, L. and Chess, L. (1985) Neuropathy and anti-myelin associated glycoprotein IgM M-proteins: T cell regulation of M-protein secretion in vitro. Ann. Neurol. 18, 182-188. McLeod, J., Walsh, J. and Pollard, J. (1984) Neuropathies associated with paraproteinemias and dysproteinemias. In: P. Dyck, P. Thomas, E. Lambert and E. Bunger (Eds.), Peripheral Neuropathy, 2nd edn., Vol 2. Saunders, Philadelphia, PA, pp. 1847-1865. Meier, C., Vandevelde, M., Steck, A. and Zubriggen, A. (1983) Demyelinating polyneuropathy associated with monoclonal IgMparaproteinemia. J. Neurol. Sci. 63, 353-367. Melmed, C., Frail, D., Duncan, I., Braun, P., Danoff, D., Finlayson, M. and Stewart, J. (1983) Peripheral neuropathy with IgM kappa monoclonal immunoglobulin directed against myelin-associated glycoprotein. Neurology 33, 1397-1405. Mitchell, G.W., Williams, G.S., Bosh, E.P. and Hart, M.N. (1991) Class II antigen expression in peripheral neuropathies. J. Neurol. Sci. 102, 170-176. Miyatani, N., Baba, H., Sato, S., Nakamura, K., Yuasa, T. and Miyatake, T. (1987) Antibody to sialosyllactosaminylparagloboside in a patient with IgM paraproteinemia and polyradiculoneuropathy. J. Neuroimmunol. 14, 189-196.
Nobile-Orazio, E., Baldini, L., Barbieri, S., Marmiroli, P., Spagnol, G., Francomano, E. and Scarlato, G. (1988) Treatment of patients with neuropathy and anti-MAG lgM M-proteins. Ann. Neurol. 24, 93-97. Nobile-Orazio, E., Francomano, E., Daverio, R., Barbieri, S., Marmiroli, P., Manfredini, E., Carpo, M., Moggio, M., Legmane, G., Baldini, L. and Scarlato, G. (1989) Anti-myelin-associated glycoprotein IgM antibody titers in neuropathy associated with macroglobulinemia. Ann. Neurol. 26, 543 550. Norton, W.T. and Poduslo, S.E. (1973) Myelination in rat nerve: Method of myelin isolation. J. Neurochem. 21. 749-757. Olerup, O. and Hillert, J. (t991) HLA-D region associated genetic susceptibility to multiple sclerosis. Tissue Antigens 38, 1-15. Olerup, O., Smith, C.I.E. and Hammarstr6m, L. (1990) Different amino acid residues at position 57 of the HLA-DQ beta chain are associated with susceptibility and resistance to IgA deficiency. Nature 347, 289-290. Payami, H., Joe, S., Farid, N.R., Stenszky, V., Chart, S.H., Yeo, P.P.B., Cheah, J.S. and Thomson, G. (1989) Relative predispositional effects (RPEs) of marker alleles with disease: HLA-DR alleles and Graves disease. Am. J. Hum. Genet. 45, 541-546. Quarles, R., Barbarash, G., Filewicz, D. and McIntyre, L. (1983) Purification and partial characterization of the myelin associated glycoprotein from adult rat brain. Biochem. Biophys. Acta 757, 140-143. Sherman, W.H., Latov, N., Hays, A.P., Takatsu, M., Nemni, R., Galassi, G. and Osserman, E.F. (1983) Monoclonal IgM-kappa antibody precipitating with chondroitin sulfate C from patients with axonal polyneuropathy and epidermolysis. Neurology 33, 192-201.
Stefansson, K., Marton, L., Antel, J., Wollman, R., Rose, R., Chejfec, G. and Arnason, B. (1983) Neuropathy accompaying IgM-lambda monoclonal gammopathy. Acta Neuropathol. 59, 255-261. Todd, J.A., Bell, J.I. and McDevitt, H.O. (1987) HLA-DQ beta gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature 329, 599-604. Vrethem, M., Ekstedt, B., Skogh, T., Andersen, O., Lycke, J. and Ernerudh, J. (1993) Soluble interleukin-2 receptor levels in serum of patients with demyelinating polyneuropathy associated with monoclonal gammopathy. J. Neurol. Neurosurg. Psychiatry (in press). Yee, W.C., Hahn, A.F., Heran, S.A. and Rupar, A.R. (1989) Neuropathy in IgM paraproteinemia: immunoreactivity to neural proteins and chondroitin sulfate. Acta Neuropathol. 78, 57-64. Zetterquist, H. and Olerup, O. (1992) Identification of the HLADRB1 *04, -DRB1 *07 and -DRBI *09 alleles by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Hum. Immunol. (in press).
139
JNI 02319
Susceptibility to demyelinating polyneuropathy in plasma cell dyscrasia may be influenced by amino acid position 9 of the HLA-DR/3 chain Magnus V r e t h e m
a Jan E r n e r u d h a,b, Mabel Cruz c, Olle Olerup a,e, G6ran Solders c, Bo Ekstedt e, Oluf Andersen g and Jan Hillert c'e
Department of Neurology and b Department of Transfusion Medicine and Clinical Immunology, University Hospital, Linkb'ping, Sweden, Departments of c Neurology and d Clinical Immunology, Karolinska Institute at Huddinge Hospital, Huddinge, Sweden, e Center for Bio Technology, Novum, Karolinska Institute, Huddinge, Sweden, f Department of Neurology, Orebro Hospital, Orebro, Sweden and g Department of Neurology, University Hospital, GSteborg, Sweden (Received 9 September 1992) (Accepted 6 October 1992)
Key words: Genetic susceptibility; Monoclonal gammopathy; Polyneuropathy; HLA-D antigen; Restriction fragment length polymorphism
Summary Fifty-five patients with plasma cell dyscrasias were investigated by genomic typing for HLA-DR and -DQ genes by restriction fragment length polymorphism, neurophysiology and for presence of anti-myelin-associated glycoprotein (MAG) antibodies. In 26 patients, a polyneuropathy (PN) of demyelinating type was established. Among these individuals, an association was found with the presence of a tryptophan amino acid residue at position 9 of the DR/3 chain (P < 0.01). This position is part of the first hypervariable region of the DR/3 chain, and may be of importance in determining preferential peptide-binding capacity of the HLA-DR molecule. The presence of anti-MAG antibodies in 15 out of 17 patients with an IgM M-component and demyelinating PN (14 of these 15 individuals carrying a tryptophan at position 9) supports the pathogenic role of an autoimmune response against MAG. The finding of an HLA class II association may indicate a pathogenic role of T cell immunity in this condition.
Introduction Plasma cell dyscrasias are characterized by proliferation of a clone of plasma cells producing a homogeneous paraprotein (M-component) and are statistically associated with demyelinating polyneuropathy (PN) (Latov et al., 1980; McLeod et al., 1984; Kelly, 1985). A pathogenic role of the M-component in the development of some of these cases of PN has been suggested because: the M-component reacts with myelin-associated glycoprotein (MAG) (Braun et al., 1982; Melmed et al., 1983; Nobile-Orazio et al., 1989); direct injection of M-component containing sera causes demyelination in animals (Hays et al., 1987); immunoglobulin M (IgM) deposits have been demonstrated in peripheral nerve
Correspondence to: M. Vrethem, Department of Neurology, University Hospital, S-581 85 Link6ping, Sweden.
(Meier et al., 1983; Stefansson et al., 1983); and beneficial effect of lowered anti-myelin antibody levels by immunosuppressive treatment has been shown (Nobile-Orazio et al., 1988). There is, however, no absolute correlation between either occurrence of antibodies or lowered antibody level and clinical effect (Kelly et al., 1988; Ernerudh et al., 1992). Thus, other mechanisms may be involved and T cell regulation of B cell activity and a direct T cell involvement in the pathogenesis of demyelinating PN has also been suggested (Latov et al., 1985). Both humoral and cell-mediated immune responses may occur, suggesting an autoimmune component in the development of the demyelinating PN. Diseases with autoimmune features are frequently associated with alleles or haplotypes of the HLA class II region. As genomic methods for HLA class II typing have been developed, offering increased accuracy as well as resolution compared to serology, previously known HLA associations have been refined, for exam-
140 TABLE 1 Clinical and laboratory characteristics of patients with plasma cell dyscrasias No.
Age/Sex
Disease
Ig class
Polyneuropathy a
IgM anti-MAG ab. reactivity b
HLA-D haplotype 1
HLA-D haplotype 2
1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55.
51/M 85/M 66/F 63/F 82/M 50/M 72/F 70/M 75/M 74/F 74/M 77/M 82/M 72/M 66/F 68/M 60/M 71/M 85/M 43/F 63/F 63/M 67/M 62/M 68/M 74/M 76/F 62/M 85/M 61/M 62/M 71/F 65/M 69/F 72/F 71/M 77/M 75/M
MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Waldenstr6m Waldenstr6m Waldenstr6m MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Waldenstr6m Waldenstr/Sm MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS MGUS Myeloma Myeloma MGUS MGUS Myeloma Myeloma MGUS MGUS MGUS MGUS MGUS MGUS MGUS Myeloma Myeloma MGUS Myeloma
IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM IgM lgM IgM IgM IgM IgM IgG IgG IgG IgG IgG IgG IgG IgA IgA IgM IgM IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgG IgA IgA IgA k-u IgM IgG IgG IgG IgG IgG IgG IgG IgG IgA l-u
Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Demyelinating Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Axonal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal Normal
+ + + + + + + + + + + + + + + ND ND -
DR1,DQ5 DR8,DQ4 DR1,DQ5 DR4,DQ8 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR15,DQ6 DR1,DQ5 DR17,DQ2 DRII,DQ7 DR1,DQ5 DR1,DQ5 DR7,DQ2 DR16,DQ5 DR16,DQ5 DR15,DQ6 DR17,DQ2 DR15,DQ6 DR15,DQ6 DR17 DQ2 D R l l DQ7 DRI5 DQ6 DR17 DQ2 DR17 DQ6 DR17 DQ2 DR15 DQ6 DRI DQ5 DR15,DQ6 DR15,DQ6 DRll,DR15,DQ6 DRI7,DQ2 DR17,DQ2 DR1,DQ5 DRI,DQ5 DR10,DQ5 DR4,DQ7 DRI5,DQ6 DR15,DQ6 DR15,DQ6 DR17,DQ2 DR17,DQ2 DR1,DQ5 DR15,DQ6 DR1,DQ5 DR15,DQ6 DR1,DQ5 DR4,DQ8 DR1,DQ5 DR17,DQ2 DRI3,DQ6
DRI5,DQ6 DR7,DQ9 DR4,DQ8 DR7,DQ2 DR7,DQ2 DR7,DQ2 DR7,DQ2 DR4,DQ7 DR15,DQ6 DR15,DQ6 DR17,DQ2 DR4,DQ8 DR4,DQ8 DR13,DQ6 DR17,DQ2 DR15,DQ6 DR9,DQ9 DR13,DQ6 DR4,DQ7 DR4,DQ7 DR17,DQ2 DR4,DQ8 DR4,DQ8 DR7,DQ2 DR13,DQ6 DR11,DQ7 DR7,DQ9 DRI1,DQ7 DR12,DQ7 DR13,DQ6 DR4,DQ8 DR7,DQ9 DR15,DQ6 DR13,DR4,DQ7 DR12,DQ7 DR13,DQ6 DR1,DQ5 DR16,DQ5 DR13,DQ6 DR13,DQ6 DR15,DQ6 DR13,DQ6 DR17,DQ2 DR13,DQ6 DR4,DQ8 DR4,DQ8 DR4,DQ8 DR1,DQ5 DR17,DQ2 DR13,DQ6 DR8,DQ4 DR8,DQ4 DR4,DQ8 DR14,DQ5
77/M 58/M 80/M 76/M 75/M 72/M 60/F 74/M 43/M 61/M 76/M 64/F 57/F 58/M 69/M 51/F 73/F
+ ND ND -
a Clinical polyneuropathy was assumed to exist when both symptoms and signs of bilateral sensory or motor impairment were present in combination with hypo- or areflexia. Demyelinating polyneuropathy was defined as clinical polyneuropathy and decreased motor nerve conduction velocity to less than 60% of normal mean value in at least two motor nerves (Kelly, 1983), or demyelination in nerve biopsies. Axonal polyneuropathy was defined as clinical polyneuropathy but not fulfilling the criteria of neurophysiological demyelinating polyneuropathy, or showing mainly axonal involvement in nerve biopsy. Normal patients, e.g., those without polyneuropathy, did not have clinical polyneuropathy and showed normal neurographic findings. b +, IgM anti-myelin-associated glycoprotein (MAG) antibody reactivity > mean + 3SD of healthy controls in an ELISA assay. ND, not done.
141 TABLE 2 P h e n o t y p i c f r e q u e n c i e s ( % ) of H L A - D R - D Q h a p l o t y p e s a m o n g the total p a t i e n t g r o u p with p l a s m a cell dyscrasia as well as various s u b g r o u p s as c o m p a r e d w i t h controls HLA-DR-DQ haplotypes
All p a t i e n t s (n = 55)
Patients with P N (n = 44)
P a t i e n t s with demyelinating PN (n = 26)
P a t i e n t s of I g M isotype with demyelinating PN (n = 17)
P a t i e n t s with antiM A G reactivity and IgM demyelinating PN (n = 15)
Controls (n = 250)
DR1,DQ5 DR15,DQ6
25 38
23 43
27 46
41 a 47 c
40 b 47 d
18 30
DR16,DQ5 DR17,DQ2 DR4,DQ7 DR4,DQ8 DR7,DQ9 DR7,DQ2 DR8,DQ4 DR9,DQ9 DR10,DQ5 DRll,DQ7 DR12,DQ7 DR13,DQ6 DR14,DQ5
5 27 9 22 5 11 5 2 2 7 4 18 2
5 25 11 16 7 14 2 2 2 9 4 18 .
8 19 12 23 4 23 e 4 4 12 12
12 6 24 6 29 f 6 6 6 6
20 7 27 7 33 g 7 7 -
1 17 12 26 5 7 9 4 2 14 6 29 6
.
.
.
a p < 0.05, R R = 3.2; b p < 0.05, R R = 3.0; c p < 0.02, R R = 3.0; d p < 0.02, R R = 3.5; e p < 0.02, R R = 4.4; f P < 0.02, R R = 5.7; g P < 0.01, R R = 6.9 (for c and d, p values w e r e c a l c u l a t e d for allelic frequences, after s u b t r a c t i o n of the p r i m a r i l y a s s o c i a t e d D R 1 , D Q 5 and D R 7 , D Q 2 haplotypes).
pie in insulin-dependent diabetes mellitus (Todd et al., 1987) and immunoglobulin deficiencies (Olerup et al., 1990). The aim of the present study was to investigate the distribution of HLA-DR and -DQ alleles and DR-DQ haplotypes with a genomic typing procedure in a group of patients with plasma cell dyscrasia and demyelinating PN, also with reference to Ig isotype and presence of antibodies directed against MAG.
Materials and Methods
Patients Fifty-five patients with plasma cell dyscrasia (age 43-85 years, mean age 68 years), including 43 with monoclonal gammopathy of uncertain significance (MGUS), seven with multiple myeloma and five with Waldenstr6m's disease were included in the study. Peripheral nerve function was characterized clinically and neurophysiologically according to Kelly (1983). Eleven patients were found to have no PN, 18 were judged to have mainly axonal PN, and 26 had a mainly demyelinating PN. Of the 26 patients with demyelinating PN, 17 had M-components of IgM, seven of IgG, and two of IgA isotype. Two hundred and fifty randomly collected unrelated healthy individuals were used as controls. Results of genomic HLA class II typing in this group have previously been reported (Olerup and Hillert, 1991).
HLA class H typing Genomic typing for HLA-DR and -DQ was performed by TaqI restriction fragment length polymorphism (RFLP) analysis of the DRB, DQA and DQB genes, as previously described (Bidwell et al., 1987; Carlsson et al., 1987). To avoid local nomenclature, HLA class II haplotypes were designated by their associated serologic specificities (Bidwell et al., 1987; Carlsson et al., 1987). As TaqI RFLP does not distinguish between the haplotypes DR7,DQ9 and DR9,DQ9, this was achieved by polymerase chain reaction by sequence specific primer (PCR-SSP) analysis as described elsewhere (Zetterquist and Olerup, 1992). HLA-DR-DQ haplotypes were deduced from known linkage dysequilibria between DR and DQ alleles.
Anti-MAG antibody analysis Bovine MAG was prepared by a lithium diiodosalicylatephenol method (Quarles et al., 1983) from brain myelin (Norton and Poduslo, 1973). Anti-MAG antibodies were measured as previously described (Cruz et al., 1991), with some modifications. In short, 96 polystyrene microtiter plates (Costar, Cambridge, MA, USA) were coated with 100/zl of MAG (1 /~g/ml). Wells were blocked with 200 p~I/well of phosphate-buffered saline (PBS) containing 2% bovine serum albumin (BSA) (Sigma, St. Louis, MO, USA). 100 /xl of serum samples adjusted to 1 mg/l of IgM were incubated overnight at 4°C. After washing, alkaline phosphatase conjugated rabbit anti-human IgM
142 (my-chain specific) (Dako, Copenhagen, Denmark) was added. After 2 h incubation, plates were washed and incubated with 100 p.1 per well of substrate pnitrophenylphosphate (1 m g / m l ; Sigma) in 10% diaethanolamine and 0.02% MgC1 buffer, p H 9.8. Plates were read after 30 min in an automated E L I S A spectrophotometer at 405 nm. Pooled normal serum (used as intra-assay reference) did not give absorbance values above 0.08. Samples were considered positive when the absorbance values were higher than the mean value + 3 SD of 20 blood donors. The range of absorbance values for these sera was 0-0.11 (mean 0.042) for IgM, yielding a mean value + 3 SD of 0.14 for IgM.
Statistical analysis Comparisons between patients and controls were performed using the chi-square test. All P values are given without correction, but the large number of comparisons made in this study should be borne in mind. However, the group of primary interest was that of patients with demyelinating PN and an M-component of IgM isotype, i.e., with a disease of most probably autoimmune origin.
Results The distribution of H L A - D R - D Q haplotypes among patients and controls is shown in Tables 1 and 2. In the total group of patients with M-components, no significant differences were seen. However, in the group of patients with IgM gammopathy and a demyelinating PN (n = 17), increased frequencies were found for the DR1,DQ5 ( P < 0 . 0 5 , relative risk ( R R ) = 3 . 2 ) and D R 7 , D Q 2 ( P < 0.02, R R = 5.7) haplotypes. Eleven out of 17 (65%) patients in this group carried one or both of these haplotypes, as compared to 24% among controls ( P < 0.01, R R = 5.9). In addition, there was a moderate increase of the DR15,DQ6 haplotype ( P < 0.20, R R = 2.1) (DR15 is a split of DR2). After subtraction of the two primarily associated haplotypes from patients and controls (according to the method described by Payami et al., 1989), the allelic frequency of DR15,DQ6 was found to be significantly increased
( P < 0.02, R R = 3.0). Similar results were obtained in patients with anti-MAG antibodies, since this group was essentially identical to that of IgM M-components with demyelinating PN. Analysis of common features of the three haplotypes DR1,DQ5, DR15,DQ6 and D R 7 , D Q 2 reveals similarities in the first hypervariable region of the DR/3 chain. In particular, these haplotypes differ, together with the DR16,DQ5 and D R 7 , D Q 9 haplotypes, from all other haplotypes in carrying a non-polar tryptophan residue at position 9, where other DR/3 alleles carry a charged glutamic acid residue. Table 3 shows the phenotypic frequency of a DR/3 position 9 tryptophan in the various subgroups. Significant differences compared to the controls were seen for all patients with plasma cell dyscrasia and demyelinating PN as well as for those with IgM isotype. Notably, a similarly increased relative risk, however not statistically significant, was seen among the small group of nine patients with demyelinating PN and M-component of IgG or IgA isotype. When comparing patients with demyelinating PN with individuals either without PN or with an axonal PN (supposedly not of autoimmune origin) there was a statistically significant increase of tryptophan at position 9 ( P < 0.03). We found a strong correlation between demyelinating PN of IgM isotype and anti-MAG antibodies (15 out of 17 patients). For patients with axonal PN or without PN, no difference from the controls was observed in either H L A class II haplotypes or in the presence of anti-MAG antibodies.
Discussion An increasing amount of evidence indicates that demyelinating PN associated with an M-component may be of autoimmune origin. Also, genetic factors have been suspected of playing a role in the susceptibility to this condition and familial cases have been reported (Busis et al., 1985; Jensen et al., 1988). l g G heavy-chain (Gm) allotypes have been reported to differ significantly between patients and healthy controls (Kahn and Pandey, 1987). Furthermore, an increase of
TABLE 3 Frequency (%) of carriers of the amino acid tryptophan at position 9 of the DR~ chain in the different subgroups of patients with plasma cell dyscrasia as compared with controls All patients (n = 55)
Patients with PN (n = 44)
Patientswith demyelinatingPN (n = 26)
71 P < 0.05 RR = 1.9
68 n.s. RR = 1.7
85 P < 0.01 RR = 4.4
Patients of IgA and IgG isotype and demyelinating PN (n = 9) 78 n.s. RR = 3.6
Patients of IgM isotype with demyelinatingPN (n = 17) 88 P < 0.02 RR = 6.0
Patients with antiMAG reactivity and IgM demyelinating PN (n = 15) 93 P < 0.03 RR = 11.0
Controls (n = 250)
56
143
class II antigen expression on Schwann cells has been shown (Mitchell et al., 1991). To our knowledge genomic HLA class II typing has not been performed in these patients, although susceptibility to autoimmune diseases is commonly associated with these genes. In the present study of 26 patients with demyelinating PN, the results indicate a role of a specific amino acid residue of the DR/3 chain. The distribution of HLA-DR-DQ haplotypes in the whole material (n = 55) differed very little in comparison with the controls. Any subdivision of the material increases the number of comparisons and our findings in this relatively rare disorder are not strong enough to etablish this association statistically. On the other hand, associations should not be expected to be found in large heterogeneous groups of patients but in homogeneous, well-defined conditions with clear autoimmune features such as the group of patients with plasma cell dyscrasia of IgM isotype and demyelinating PN (Gosselin et al., 1991). Polymorphism of DRB genes is mostly limited to three hypervariable regions. According to models of the three-dimensional structure of the MHC class II molecules, based on the known structure of HLA class I molecules (Brown et al., 1988), these polymorphic segments make up parts of the walls of the peptidebinding groove. In this way, different alleles differ with regard to their preferential binding of different peptides, although each allelic molecule can bind many different peptides, with varying affinity. Amino acid residue 9 is part of the first hypervariable region and is believed to be located in the /3-pleated sheeth in the bottom of the peptide-binding groove of the class II molecule. Theoretically the substitution of a charged glutamic acid with a highly hydrophobic tryptophan residue at this position should alter the properties of the peptide-binding groove considerably. Previously, most evidence has pointed towards a B-cell immune response in the pathogenesis of PN in IgM plasma cell dyscrasias. Our findings, however, are also compatible with a pathogenic importance of cellmediated immunity. HLA class II associations in general suggest a role of T-cells, and the type of association found here could implicate the binding/presentation of a specific peptide. Thus, the T-cell response could well be more restricted than the B-cell response, which has been shown often to be directed against MAG (Braun et al., 1982; Melmed et al., 1983; Nobile-Orazio et al., 1989) and Po-protein (Bollensen et al., 1988), but also against various gangliosides or chondroitin sulfate not reactive with MAG (Sherman et al., 1983; Miyatani et al., 1987; Ilyas et al., 1988; Yee et al., 1989). The presence of the same HLA class II association in IgA and IgG plasma cell dyscrasias with PN but without anti-MAG specificity of the M-components, could further support this notion, as would the previous demonstration of T-cells in nerve biopsies
from similar patients (Solders et al., unpublished observations). The presence of activated T-cells is also supported by the findings of increased soluble interleukin-2 receptor levels in some patients with M-component-associated PN (Vrethem et al., in press). In conclusion, our results indicate that a genetic predisposition to demyelinating PN in plasma cell dyscrasias associated with the HLA-DR-DQ subregion might be explained by a specific amino acid residue of the DR/~ chain. Since the number of patients in this study was limited, confirmation of this hypothesis has to await the investigation of other similar patients.
Acknowledgements The work was supported by grants from OstergiStlands Liins Landsting, the Swedish Association of Neurologically Disabled, the Swedish Society of Medicine, the Swedish Medical Research Council (Project No. 8709) and the Karolinska Institute.
References Bidwell, J.L., Bidwell, E.A., Laundy, C.J., Klouda, P.T. and Bradley, B.A. (1987) Allogenotypes defined by short DQa and DQ/3 cDNA probeg correlate with, and define splits of HLA-DQ serological specificities. Mol. Immunol. 24, 513-522. Bollensen, E., Steck, A.J. and Schachner, M. (1988) Reactivity with the peripheral myelin glycoprotein Po in serum from patients with monoclonal IgM gammopathy and polyneuropathy. Neurology 38, 1266-1270. Braun, P., Frail, D. and Latov, N. (1982) Myelin-associated glycoprotein is the antigen for monoclonal IgM in polyneuropathy. J. Neurochem. 128, 1261-1265. Brown, J.H., Jardetzky, T., Saper, M.A., Samraoui, B., Bjorkman, P.J. and Wiley, D.C. (1988) A hypothetical model of the foreign antigen binding site of class II histocompatibility molecules. Nature 332, 845-850. Busis, N.A., Halperin, J.J., Stefansson, K., Kwiatkowski, D.J., Sagar, S.M., Schiff, S.R. and Logigian, E.L. (1985) Peripheral neuropathy, high serum IgM, and paraproteinemia in mother and son. Neurology 35, 679-683. Carlsson, B., Wallin, J., B6hme, J. and M611er, E. (1987) HLA-DRDQ haplotypes defined by restriction fragment length analysis: Correlation to serology. Hum. Immunol. 20, 95-113. Cruz, M., Jiang, J.-Y., Ernerudh, J., Solders, G., Olsson, T., Osterman, P.-O. and Link, H. (1991) Antibodies to myelin associated glycoprotein are found in cerebrospinal fluid in peripheral neuropathy associated with monoclonal serum IgM. Arch. Neurol. 48, 66-70. Ernerudh, J., Vrethem, M., Andersen, O., Lindberg, C. and Berlin, G. (1992) Immunochemical and clinical effect of immunosuppressive treatment in monoclonal IgM neuropathy. J. Neurol. Neurosurg. Psychiatry 55, 930-934. Gosselin, S., Kyle, R.A. and Dyck, P.J. (1991) Neuropathy associated with monoclonal gammopathies of undetermined significance. Ann. Neurol. 30, 54-61. Hays, A., Latov, N., Takatsu, M. and Sherman, W. (1987) Experimental demyelination induced by serum of patients with neuropathy and an anti-MAG IgM M-protein. Neurology 37, 242-256.
144 Ilyas, A.A., Li, S.-C., Chou, D.K.H., Li, Y.-T., Jungalwala, F.B., Dalakas, M.C. and Quarles, R.H. (1988) Gangliosides GM2, IV4GalNAcGMlb, and IV4GalNAcGDla as antigens for monoclonal IgM in neuropathy associated with gammopathy. J. Biol. Chem. 263, 4369-4373. Jensen, T.S., Schr6der, H.D., J6nsson, V., Ernerudh, J., Stigsby, B., Kamieniecka, Z., Hippe, E. and Trojaborg, W. (1988) IgM monoclonal gammopathy and neuropathy in two siblings. J. Neurol. Neurosurg. Psychiatry 51, 1308-1315. Kahn, S.N. and Pandey, J.P. (1987) IgG heavy-chain (Gm) allotypes in demyelinating polyneuropathy associated with MAG-binding monoclonal IgM autoantibodies. Ann. Neurol. 21,507-509. Kelly, J. Jr. (1983) The electrodiagnostic findings in peripheral neuropathy associated with monoclonal gammopathy. Muscle Nerve 6, 504-509. Kelly, J. Jr. (1985) Peripheral neuropathies associated with monoclonal proteins: A clinical review. Muscle Nerve 8, 138-150. Kelly, J.J., Adelman, L.S., Berkman, E. and Bhan, I. (1988) Polyneuropathies associated with IgM monoclonal gammopathies. Arch. Neurol. 45, 1355-1359. Latov, N., Sherman, W., Nemni, R., Galassi, G., Shyong, J., Penn, A., Chess, L., Olarte, M., Rowland, L. and Osserman, E. (1980) Plasma cell dyscrasia and peripheral neuropathy with a monoclonal antibody to peripheral nerve myelin. New Engl. J. Med. 303, 618-62l. Latov, N., Godfrey, M., Thomas, Y., Nobile-Orazio, E., Spatz, L., Abraham, J., Perman, G., Freddo, L. and Chess, L. (1985) Neuropathy and anti-myelin associated glycoprotein IgM M-proteins: T cell regulation of M-protein secretion in vitro. Ann. Neurol. 18, 182-188. McLeod, J., Walsh, J. and Pollard, J. (1984) Neuropathies associated with paraproteinemias and dysproteinemias. In: P. Dyck, P. Thomas, E. Lambert and E. Bunger (Eds.), Peripheral Neuropathy, 2nd edn., Vol 2. Saunders, Philadelphia, PA, pp. 1847-1865. Meier, C., Vandevelde, M., Steck, A. and Zubriggen, A. (1983) Demyelinating polyneuropathy associated with monoclonal IgMparaproteinemia. J. Neurol. Sci. 63, 353-367. Melmed, C., Frail, D., Duncan, I., Braun, P., Danoff, D., Finlayson, M. and Stewart, J. (1983) Peripheral neuropathy with IgM kappa monoclonal immunoglobulin directed against myelin-associated glycoprotein. Neurology 33, 1397-1405. Mitchell, G.W., Williams, G.S., Bosh, E.P. and Hart, M.N. (1991) Class II antigen expression in peripheral neuropathies. J. Neurol. Sci. 102, 170-176. Miyatani, N., Baba, H., Sato, S., Nakamura, K., Yuasa, T. and Miyatake, T. (1987) Antibody to sialosyllactosaminylparagloboside in a patient with IgM paraproteinemia and polyradiculoneuropathy. J. Neuroimmunol. 14, 189-196.
Nobile-Orazio, E., Baldini, L., Barbieri, S., Marmiroli, P., Spagnol, G., Francomano, E. and Scarlato, G. (1988) Treatment of patients with neuropathy and anti-MAG lgM M-proteins. Ann. Neurol. 24, 93-97. Nobile-Orazio, E., Francomano, E., Daverio, R., Barbieri, S., Marmiroli, P., Manfredini, E., Carpo, M., Moggio, M., Legmane, G., Baldini, L. and Scarlato, G. (1989) Anti-myelin-associated glycoprotein IgM antibody titers in neuropathy associated with macroglobulinemia. Ann. Neurol. 26, 543 550. Norton, W.T. and Poduslo, S.E. (1973) Myelination in rat nerve: Method of myelin isolation. J. Neurochem. 21. 749-757. Olerup, O. and Hillert, J. (t991) HLA-D region associated genetic susceptibility to multiple sclerosis. Tissue Antigens 38, 1-15. Olerup, O., Smith, C.I.E. and Hammarstr6m, L. (1990) Different amino acid residues at position 57 of the HLA-DQ beta chain are associated with susceptibility and resistance to IgA deficiency. Nature 347, 289-290. Payami, H., Joe, S., Farid, N.R., Stenszky, V., Chart, S.H., Yeo, P.P.B., Cheah, J.S. and Thomson, G. (1989) Relative predispositional effects (RPEs) of marker alleles with disease: HLA-DR alleles and Graves disease. Am. J. Hum. Genet. 45, 541-546. Quarles, R., Barbarash, G., Filewicz, D. and McIntyre, L. (1983) Purification and partial characterization of the myelin associated glycoprotein from adult rat brain. Biochem. Biophys. Acta 757, 140-143. Sherman, W.H., Latov, N., Hays, A.P., Takatsu, M., Nemni, R., Galassi, G. and Osserman, E.F. (1983) Monoclonal IgM-kappa antibody precipitating with chondroitin sulfate C from patients with axonal polyneuropathy and epidermolysis. Neurology 33, 192-201.
Stefansson, K., Marton, L., Antel, J., Wollman, R., Rose, R., Chejfec, G. and Arnason, B. (1983) Neuropathy accompaying IgM-lambda monoclonal gammopathy. Acta Neuropathol. 59, 255-261. Todd, J.A., Bell, J.I. and McDevitt, H.O. (1987) HLA-DQ beta gene contributes to susceptibility and resistance to insulin-dependent diabetes mellitus. Nature 329, 599-604. Vrethem, M., Ekstedt, B., Skogh, T., Andersen, O., Lycke, J. and Ernerudh, J. (1993) Soluble interleukin-2 receptor levels in serum of patients with demyelinating polyneuropathy associated with monoclonal gammopathy. J. Neurol. Neurosurg. Psychiatry (in press). Yee, W.C., Hahn, A.F., Heran, S.A. and Rupar, A.R. (1989) Neuropathy in IgM paraproteinemia: immunoreactivity to neural proteins and chondroitin sulfate. Acta Neuropathol. 78, 57-64. Zetterquist, H. and Olerup, O. (1992) Identification of the HLADRB1 *04, -DRB1 *07 and -DRBI *09 alleles by PCR amplification with sequence-specific primers (PCR-SSP) in 2 hours. Hum. Immunol. (in press).