Association between microchimerism and multiple sclerosis in Canadian twins

Journal of Neuroimmunology 179 (2006) 145 – 151 www.elsevier.com/locate/jneuroim

Association between microchimerism and multiple sclerosis in Canadian twins Cristen J. Willer a,1 , Blanca M. Herrera a , Katie M.E. Morrison a , A.D. Sadovnick c , George C. Ebers a,b,⁎ for the Canadian Collaborative Study on Genetic Susceptibility to Multiple Sclerosis a Department of Clinical Neurology, University of Oxford, Oxford, UK Wellcome Trust Center for Human Genetics, University of Oxford, Oxford, UK Department of Medical Genetics and Faculty of Medicine (Neurology), University of British Columbia, Vancouver, Canada b

c

Received 7 August 2005; received in revised form 7 June 2006; accepted 13 June 2006

Abstract Microchimerism, the persistence of foreign cells thought to derive from previous pregnancies, has been associated with autoimmune diseases. A maternal parent-of-origin effect in MS remains unexplained. We tested for microchimerism in monozygotic and dizygotic twin-pairs with MS. Microchimerism was associated with MS in affected females from monozygotic concordant pairs when compared to both affected (p = 0.020) and unaffected (p = 0.025) females in monozygotic discordant pairs. Microchimerism was increased in affected females of dizygotic discordant pairs (p = 0.059). The rate of microchimerism was significantly higher in affected twins than in unaffected co-twins (p = 0.0059). These observations show an association in twins between the presence of microchimerism and having MS. © 2006 Elsevier B.V. All rights reserved. Keywords: Multiple sclerosis; Microchimerism; Twins

1. Introduction Microchimerism, the enduring presence of a small number of non-host cells, is associated with the development of some autoimmune disorders. The most common and plausible source of microchimerism is the presence of fetal stem cells from a previous pregnancy that persist in the maternal bone marrow or lymph nodes and continue to be tolerated by the mother's immune system. Stem cells may also be transferred between dizygotic (DZ) twins during pregnancy but this has not been proven in humans. However, ⁎ Corresponding author. Department of Clinical Neurology, Radcliffe Infirmary, Woodstock Road, Oxford OX2 6HE, UK. Tel.: +44 1865 228579; fax: +44 1865 224757. E-mail address: [email protected] (G.C. Ebers). 1 Present address: Department of Biostatistics, University of Michigan, Ann Arbor, USA. 0165-5728/$ - see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.jneuroim.2006.06.011

transfer between monozygotic (MZ) twins in the twin–twin transfusion syndrome has been confirmed (Bourthoumieu et al., 2005) and transfer from mother to fetus has also been shown (Bianchi et al., 1996; Maloney et al., 1999). Other possible sources of microchimerism include transmission of cells from mother to fetus during pregnancy, from bone marrow or organ transplant, or blood transfusion. Most pregnant women and a large proportion of parous women are positive for microchimeric cells (Table 1) (Bianchi et al., 1996; Thomas et al., 1994). The hypothesis that autoimmune diseases are associated with microchimerism was based in part on studies of the autoimmune disease scleroderma. The clinical features of scleroderma resemble those of graft-versus-host disease (GvHD) (Nelson, 1998), which can occur following transplantation if donor cells attack the cells of the host (Santos and Cole, 1958). This led to the hypothesis that a mechanism involving foreign cells (Nelson, 1996) transferred during a

146

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151

Table 1 Prevalence of microchimerism in peripheral blood of women from published studies Control group

Tissue studied

Normal women, pregnancy history unknown Scleroderma patients and controls Healthy control women who have had sons Japanese women with sons (2–18 years old) Women with sons and nulliparous women Asian women with at least one son

DNA from peripheral blood

Technique

Number positive % Positive Reference for MC/total for MC

PCR for Y-chromosome 1/25 sequences DNA from peripheral blood PCR for HLA alleles 17/31 DNA from peripheral blood PCR for Y-chromosome 12/24 sequences DNA from peripheral blood PCR for Y-chromosome 6/12 sequences DNA from peripheral blood, and PCR for Y-chromosome 2/8 CD34+ enriched fraction sequences DNA from peripheral blood PCR for Y-chromosome 8/41 sequences Women pregnant with female fetuses Plasma and serum from peripheral PCR for Y-chromosome 0/23 blood sequences Normal women pregnant with male DNA from peripheral blood PCR for Y-chromosome 13/19 fetuses at time of study sequences Women pregnant with female fetuses DNA from peripheral blood PCR for Y-chromosome 4/13 at time of study sequences

previous pregnancy may play a role in scleroderma (Famularo and De Simone, 1999). Samples of skin lesions and peripheral blood leukocytes from women with scleroderma were found to contain male cells more often than control women (Artlett et al., 1998; Miyashita et al., 2000; Nelson et al., 1998). Similarly, women with autoimmune Hashimoto's thyroiditis had higher rates of male microchimeric cells than controls with nodular goiter (Klintschar et al., 2001). GvHD is said to be more common when the human leukocyte antigen (HLA) profile of the host matches the donor (Williamson and Warwick, 1995) since similar microchimeric cells presumably reside undetected in the bone marrow or liver (Artlett et al., 1997). Females with scleroderma are more likely to be HLA-compatible with their mothers or children compared to unaffected women (Maloney et al., 1999). We have previously found no evidence for increased HLA-compatibility between multiple sclerosis patients and their mothers (Willer et al., 2005); however, this does not exclude a role for microchimerism in multiple sclerosis, particularly if the primary source of microchimeric cells is from the offspring of affected individuals. HLA-compatibility and microchimerism might be incompletely correlated and therefore difficult to detect in a complex disease. Furthermore, the role of minor histocompatibility antigens was not examined. Here we tested the hypothesis that microchimerism is associated with multiple sclerosis (Willer et al., 2002) by testing for microchimerism in twins diagnosed with MS and their co-twins. These studies were stimulated by the finding of a maternal parent of origin effect in halfsuiblings (Ebers et al., 2004). We compared concordant MZ pairs to discordant MZ pairs to test whether concordance itself is associated with microchimerism. We compared the rates of microchimerism in affected members of discordant pairs, both MZ and DZ to their unaffected

4

Artlett et al. (1998)

55 50

Maloney et al. (1999) Nelson et al. (1998)

50

Murata et al. (1999)

25

Toda et al. (2001)

20

Miyashita et al. (2000)

0

Lo et al. (1998)

68

Bianchi et al. (1996)

31

Bianchi et al. (1996)

co-twins. We observed significantly higher rates microchimerism in women affected with MS than their unaffected co-twins, and significantly higher rates microchimerism in concordant MZ pairs compared discordant MZ pairs.

of in of to

2. Methods 2.1. Ascertainment All twin subjects were identified through the Canadian Collaborative Project on the Genetic Susceptibility to Multiple Sclerosis. Twins with MS, their co-twins, mothers, husbands and twins' husbands were asked to participate in the study. Written consent was obtained from all participants in the study and the research protocol was approved by local Institutional Review Boards at each clinic and research center. A diagnosis of multiple sclerosis was made according to Poser's guidelines (Poser et al., 1983). Probands with a diagnosis of possible MS and their co-twins were excluded from the analysis. If the twin proband was considered to meet the criteria for probable or definite MS and the co-twin had a diagnosis of possible MS then the twin pair was included in the study. 2.2. Study sample There were a total of 190 samples collected under sterile conditions and tested for microchimerism at a minimum of one locus (Figs. 1 and 2). The characteristics of the twin sample from the Canadian Collaborative Study have been previously reported (Willer et al., 2003). Informativity at the beta-globin locus required homozygosity for the genomic DNA, and the Ychromosome test was performed on females only. A detailed history of all previous pregnancies (if female), blood

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151

147

Fig. 1. Rates of Y-chromosome, beta-globin, and total microchimerism in Canadian twins with MS and their mothers.

transfusions and organ transplants was taken from each individual when possible. We tested for the presence of microchimerism in samples from 41 mothers of twin probands and eight unaffected daughters of twins. We collected sterile DNA

samples from ten unaffected husbands of twins, and one unaffected son. We also had available DNA collected under standard (non-sterile) laboratory conditions from 54 husbands of female twins to attempt to determine whether the source of microchimerism was a prior pregnancy.

Fig. 2. Rates of Y-chromosome, beta-globin, and total microchimerism by sex, pregnancy history and affection status.

148

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151

2.3. DNA extraction

3. Results

Blood samples were obtained from all twins by venipuncture in sterile, sealed EDTA-containing vacutainers. DNA was extracted with a salting-out method (Qiagen) using sterile conditions in a class II cabinet. Water blanks were handled in an identical manner to the samples for each extraction. Two different areas in separate tissue-culture rooms each containing separate class II cabinets were used for: (i) DNA extraction, (ii) DNA dilution and PCR preparation. A third separate room was used for post-PCR gel electrophoresis.

To assess whether an association between multiple sclerosis and microchimeric cells was detectable in our sample of twins, we tested 190 individuals for the presence or absence of microchimeric cells (Figs. 1 and 2). At the beta-globin locus, which can detect microchimerism in both males and females, we tested 80 women and seven were positive. None of the 31 samples from males were positive (Fig. 2). The difference in rate of microchimerism in females and males was of marginal significance (p = 0.09). Having shown the absence of microchimerism in males, the remaining analyses focus on females only. Microchimerism for Y-chromosome DNA was tested in nearly all samples from females. Of 154 females tested, 12 were positive for Y-chromosome microchimerism (Fig. 2). All of the individuals found to be positive for Y-chromosome microchimerism had reported having had at least one biological son. The rate of Y-chromosome microchimerism in women that reported having a son was 11.9% (12/101). We observed no individuals that were positive for microchimerism at both loci tested. However, of 12 individuals that tested positive for Y-chromosome DNA, only 6 were homozygous and were therefore tested for beta-globin microchimerism. Of the seven women that tested positive for non-genomic beta-globin DNA, four had previously been pregnant, one had no offspring, and the pregnancy history of two women was unknown (Fig. 2). We examined whether the rate of microchimerism was different in female twins with a diagnosis of MS compared to those considered to be unaffected at the time of the analysis (Fig. 1). Among all twins, we observed that 15.9% (11/69) of affected females were microchimeric, which was significantly higher than the finding of no microchimeric individuals among unaffected female co-twins (0/38, p = 0.0059). An additional individual that tested positive for microchimerism was considered to have ‘possible MS’ and was not included in comparison between affected and unaffected individuals. We next examined the influence of twin zygosity and concordance status on the rate of microchimerism. We observed that 33.3% of females from concordant MZ pairs tested positive for microchimerism (6/18), which was significantly increased over the rate observed in discordant female MZ twins (0/27, p = 0.0023, Fig. 1). In females from discordant unlike-sex DZ pairs, we found 28.6% to be microchimeric (4/14) which although showing a higher rate than the zero positive cases among four unaffected females from unlike-sex pairs was not significantly different. We tested for microchimerism in mothers of twin pairs where at least one twin had definite or probable MS (Fig. 1). We observed a rate of microchimerism of 14.6% (6/41) in mothers of twins. In only one family did we observe microchimerism in both the mother of twins and one of her affected MZ concordant twin daughters. Both the mother and twin daughter tested positive for Y-chromosome DNA.

2.4. Microchimerism detection Amplification refractory mutation detection system polymerase chain reaction (ARMS-PCR) was performed for two loci (Lo, 1994). For the bi-allelic beta-globin locus, individuals that were homozygous for either allele were tested for the opposite allele. Genotype determination at this locus was performed using primers 5′CCTTTTTGTTTCAGCTTTTCACTGTG and 5′GGAAAGAATAGCATCTACTCTTGTTCAGGA with sizes of 639 for the ‘R’ allele and 538 for the ‘T’ allele. For the detection of the ‘R’ allele in ‘T/T’ individuals, the first round of PCR was performed with primers 5′ CCTGTATAATCCAATTCCCAAG and 5′TCAAGTTGATCTGCCTGCCTC and the second round primers were 5′TCTGATTTCTAACCTCTGACCT and 5′ATCCAGCCAGGATAAGGTCTT. For the detection of the ‘T’ allele in individuals with the ‘R/R’ genotype, the first round primers were 5′CTGTATAATCCCAATTCCCAAC and 5′TCAAGTGATCTGCCTGCCTT and the second round primers were 5′ GTCATTAGATAGCTACTTGCCC and 5′ATCCAGCCAGGATAAGGTCTA. We also tested for the presence of Ychromosome locus DYS14 in females using first round primers 5′CTAGACCGCAGAGGCGCCAT and 5′TAGTACCCACGCCTGCTCCGG and the second round primers 5′CATCCAGAGCGTCCCTGGCTT and 5′CTTTCCACAGCCACATTTGTC. All DNA was diluted to 100 ng/μL and 1 μg of DNA per individual was tested in each first-round PCR reaction which had a total volume of 100 μL. A total of 2 μL from round 1 was transferred to the second-round PCR reaction. Female laboratory technicians performing extractions, genotyping and microchimerism assays were blinded to affection status, sex, pregnancy history and all other clinical information for all samples. Twins were tested for the beta-globin locus in the same batch if their genotypes were identical. Mothers were always tested separately from their offspring to prevent cross-contamination between mother and twin samples. To confirm a subset of our results, any samples that tested positive for microchimerism at the beta-globin locus were repeated (N = 7), and in no cases were the two tests discordant. Comparisons of rates of microchimerism were performed using Fisher's exact test because cell counts were typically small.

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151

The mothers of twins who were positive for microchimerism did not have a history of having more pregnancies compared to the women who tested negative for microchimerism (average of 4.8 versus 4.4, p = 0.75) or having pregnancies more recently; the average time from most recent pregnancy was 43.4 years for five mothers of twins with microchimerism (range 36–52) who provided pregnancy history information and 38 years for 28 mothers negative for microchimerism (range 27–66, p = 0.63). After testing samples obtained from eight daughters of twins, one was microchimeric (12.5%). The microchimeric daughter had an unknown pregnancy history and was in her fourth decade at the time of testing. We evaluated the potential sources of microchimerism for all positive cases (Table 2). In most cases, genotypes were available from the mothers and husbands of female twins. We chose to examine husbands of female twins (the fathers of their children) to eliminate the need to collect DNA from potentially young children, and also to be able to determine the potential genotypes of any unavailable offspring. In nearly all cases, the husband carried the allele that was present in the microchimeric individual or a woman with Ychromosome microchimerism had previously had a son, which suggests that the offspring were the most likely source of microchimerism. All seven mothers of twins with microchimerism were positive for the Y-chromosome test, and all of these women reported having sons. Among affected members of monozygotic twin pairs, six were microchimeric, and there were two pairs in which both members showed microchimerism. In one pair with two microchimeric twins, the source may have been the mother of the twins, or offspring in both cases. In another microchimeric pair, the mother's genotype suggested she was not the source of microchimeric cells, and the husband of one twin carried the allele detected. The other twin, positive for the non-genomic allele at the beta-globin locus after duplicate tests, reported no pregnancies. Of the 5 microchimeric DZ twins, one tested positive for Ychromosome microchimerism and was a member of a female–female DZ pair, suggesting the likely source was her offspring. Of the four microchimeric women who have male DZ co-twins, two were positive with the Ychromosome test and may represent twin–twin transfer, the other two were positive with the beta-globin assay. One may represent twin–twin transfer since the male twin

149

carried the allele found in the female twin, however, since we did not find Y-chromosome microchimerism, this suggests a female may be the source of the microchimeric cells. 4. Discussion A maternal effect has been shown in MS, which is of uncertain origin (Ebers et al., 2004). We sought to determine whether there was an association between the presence of microchimerism and the diagnosis of multiple sclerosis. We ascertained twins since unaffected co-twins are matched for maternal and childhood environments, and matched for genes in the case of MZ twins, yet remain discordant for MS status. We also ascertained concordant twins because we hypothesized that MZ twins concordant for MS may have been exposed to maternal factors that increase the risk for MS in both twins. Discordant MZ twins served as a uniquely matched control group. We ascertained discordant DZ twins to test for transfer of cells from the unaffected co-twin to the affected twin. An increased rate of microchimerism was found in affected women from concordant MZ and discordant DZ pairs (Fig. 2). There were no positive cases of microchimerism in affected women from MZ discordant pairs. Microchimerism in concordant MZ pairs may be reflective of a shared environment and/or genetic factors that put them at risk for both microchimerism and multiple sclerosis. The lack of microchimerism in MZ discordant twins suggests that concordance for MS may increase the likelihood of twins being microchimeric but we cannot exclude the possibility that microchimerism is unrelated to concordance. It is also possible that the biological changes related to multiple sclerosis, such as changes in the blood–brain barrier and/or non-specific immune activation may allow or cause microchimerism to persist. Discordant affected female DZ twins from unlike-sex pairs had higher rates of microchimerism than those from like-sexed pairs. This may be reflective of the usage of the Ychromosome to assess microchimerism. Cells that were transferred between DZ twins would be more easily detected with our method if they came from a male co-twin than from a female. We observed no cases of microchimerism in males, so if transfer between DZ twins were common, it is possible that foreign cells are better tolerated in females compared to

Table 2 Probable sources of microchimerism Microchimeric individual

Total number positive

Offspring

Offspring or mother

Offspring or DZ twin

Offspring or mother or DZ twin

Offspring or non-familial

No clear source

MZ affected twin DZ affected twin Mother of twins Possibly affected twin Unaffected daughter

6 5 7 1 1

2 2 6 1 1

2 0 0 0 0

0 2 0 0 0

0 1 0 0 0

1 0 1 0 0

1 0 0 0 0

Non-familial sources include blood transfusion, bone marrow transplant or organ transplant.

150

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151

males perhaps related to the state of relative immunosuppression against foreign antigens which characterizes pregnancy. We assessed the likely source of microchimerism whenever possible (Table 2). We found that the most likely source was usually offspring but mothers were also a potential source in some cases. In some DZ pairs, the co-twin was also a potential source. In only one nulliparous woman was there no clear source of Y-chromosome microchimerism. The absence of microchimerism in the male individuals was reassuring because it shows that laboratory contamination, a major hazard in such studies, could be discounted as a cause of the positive results we observed. We found very limited evidence of microchimerism in women who reported having no pregnancies. The few positive cases could be due to microchimerism derived from their mother, or the selfreported pregnancy status may have been inaccurate or an early miscarriage might have gone unnoticed or unrecognized. However, the finding of overall lower rates in selfreported nulliparous women compared to multiparous women was also reassuring with respect to the validity of the methodology. Overall, we observed lower rates of microchimerism in unaffected women than has been observed in controls in other studies (Table 1 and Fig. 2). Microchimerism detection is highly dependent on the technique used, and we chose to use a hemi-nested PCR procedure that would minimize the risks of contamination (Lo, 1994). It is also possible that unaffected women that have some genetic risk of MS, such as the unaffected twins of affected individuals, have a lower rate of microchimerism than the general population. Another plausible explanation of lower rates of microchimerism in our study compared to other studies is that our sample was primarily of post-menopausal women and many of the previously reported rates of microchimerism were in women pregnant or recently pregnant at the time of study. A higher rate of microchimerism in female twins with MS compared to unaffected co-twins, which was statistically significant, was unexpectedly found. Similarly, an apparent association with concordance in MZ twins is intriguing. Although the results come from a relatively small sample, it is not likely that this twin sample size will be easily matched. Nevertheless, replication of these results in other non-twin cohorts could help place these findings in proper context. They support the hypotheses that microchimerism is related both to having MS and to concordance in MZ twin pairs but more studies will be needed to determine the biology underlying this phenomenon and if it may be involved in susceptibility to MS in the population. Acknowledgements The authors would like to thank the participants who kindly participated in this research. Special thanks to Allison Robson, Holly Armstrong, Beverly Scott, Sandra Noble-

Topham for sample preparation and general laboratory advice. We would like to acknowledge Dave Dyment and the late Steve Poland for helpful discussion. These studies were made possible through the continued collaboration of the Canadian Collaborative Study on Genetic Susceptibility to Multiple Sclerosis and its participants. References Artlett, C.M., Welsh, K.I., Black, C.M., Jimenez, S.A., 1997. Fetal–maternal HLA compatibility confers susceptibility to systemic sclerosis. Immunogenetics 47, 17–22. Artlett, C.M., Smith, J.B., Jimenez, S.A., 1998. Identification of fetal DNA and cells in skin lesions from women with systemic sclerosis. N. Engl. J. Med. 338, 1186–1191. Bianchi, D.W., Zickwolf, G.K., Weil, G.J., Sylvester, S., DeMaria, M.A., 1996. Male fetal progenitor cells persist in maternal blood for as long as 27 years postpartum. Proc. Natl. Acad. Sci. U. S. A. 93, 705–708. Bourthoumieu, S., Yardin, C., Terro, F., Gilbert, B., Laroche, C., Saura, R., Vincent, M.C., Esclaire, F., 2005. Monozygotic twins concordant for blood karyotype, but phenotypically discordant: a case of “mosaic chimerism”. Am. J. Med. Genet., A 135, 190–194. Ebers, G.C., Sadovnick, A.D., Dyment, D.A., Yee, I.M., Willer, C.J., Risch, N., 2004. Parent-of-origin effect in multiple sclerosis: observations in half-siblings. Lancet 363, 1773–1774. Famularo, G., De Simone, C., 1999. Systemic sclerosis from autoimmunity to alloimmunity. South. Med. J. 92, 472–476. Klintschar, M., Schwaiger, P., Mannweiler, S., Regauer, S., Kleiber, M., 2001. Evidence of fetal microchimerism in Hashimoto's thyroiditis. J. Clin. Endocrinol. Metab. 86, 2494–2498. Lo, Y.M., 1994. Detection of minority nucleic acid populations by PCR. J. Pathol. 174, 1–6. Lo, Y.M., Tein, M.S., Lau, T.K., Haines, C.J., Leung, T.N., Poon, P.M., Wainscoat, J.S., Johnson, P.J., Chang, A.M., Hjelm, N.M., 1998. Quantitative analysis of fetal DNA in maternal plasma and serum: implications for noninvasive prenatal diagnosis. Am. J. Hum. Genet. 62, 768–775. Maloney, S., Smith, A., Furst, D.E., Myerson, D., Rupert, K., Evans, P.C., Nelson, J.L., 1999. Microchimerism of maternal origin persists into adult life. J. Clin. Invest. 104, 41–47. Miyashita, Y., Ono, M., Ueki, H., Kurasawa, K., 2000. Y-chromosome microchimerism in rheumatic autoimmune disease. Ann. Rheum. Dis. 59, 655–656. Murata, H., Nakauchi, H., Sumida, T., 1999. Microchimerism in Japanese women patients with systemic sclerosis. Lancet 354, 220. Nelson, J.L., 1996. Maternal–fetal immunology and autoimmune disease: is some autoimmune disease auto-alloimmune or allo-autoimmune? Arthritis Rheum. 39, 191–194. Nelson, J.L., 1998. Microchimerism and the pathogenesis of systemic sclerosis. Curr. Opin. Rheumatol. 10, 564–571. Nelson, J.L., Furst, D.E., Maloney, S., Gooley, T., Evans, P.C., Smith, A., Bean, M.A., Ober, C., Bianchi, D.W., 1998. Microchimerism and HLAcompatible relationships of pregnancy in scleroderma. Lancet 351, 559–562. Poser, C.M., Paty, D.W., Scheinberg, L., McDonald, W.I., Davis, F.A., Ebers, G.C., Johnson, K.P., Sibley, W.A., Silberberg, D.H., Tourtellotte, W.W., 1983. New diagnostic criteria for multiple sclerosis: guidelines for research protocols. Ann. Neurol. 13, 227–231. Santos, G.W., Cole, J.L., 1958. Effects of donor and host lymphoid and myeloid tissue injections in lethally X-irradiated mice treated with rat bone marrow. J. Natl. Cancer Inst. 21, 279. Thomas, M.R., Williamson, R., Craft, I., Rodeck, C.H., 1994. The time of appearance, and quantitation, of fetal DNA in the maternal circulation. Ann. N.Y. Acad. Sci. 731, 217–225.

C.J. Willer et al. / Journal of Neuroimmunology 179 (2006) 145–151 Toda, I., Kuwana, M., Tsubota, K., Kawakami, Y., 2001. Lack of evidence for an increased microchimerism in the circulation of patients with Sjogren's syndrome. Ann. Rheum. Dis. 60, 248–253. Willer, C.J., Sadovnick, A.D., Ebers, G.C., 2002. Microchimerism in autoimmunity and transplantation: potential relevance to multiple sclerosis. J. Neuroimmunol. 126, 126–133. Willer, C.J., Dyment, D.A., Risch, N.J., Sadovnick, A.D., Ebers, G.C., 2003. Twin concordance and sibling recurrence rates in multiple sclerosis. Proc. Natl. Acad. Sci. U. S. A. 100, 12877–12882.

151

Willer, C.J., Dyment, D.A., Sadovnick, A.D., Ebers, G.C., 2005. Maternal– offspring HLA-DRB1 compatibility in multiple sclerosis. Tissue Antigens 66, 44–47. Williamson, L.M., Warwick, R.M., 1995. Transfusion-associated graftversus-host disease and its prevention. Blood Rev. 9, 251–261.