IT July 2000

14/6/00

9:19 am

Page 342

REVIEW I M M U N O L O G Y T O D AY

18 Lombardi, G. et al. Type-1 interferon maintains the survival of anergic CD41 T cells. J. Immunol. (in press) 19 Gombert, W. et al. (1996) Fibroblasts prevent apoptosis of IL-2deprived T cells without inducing proliferation: a selective effect on Bcl-x(L) expression. Immunology 89, 397–404 20 Lenardo, M.J. (1997) The molecular regulation of lymphocyte apoptosis. Semin. Immunol. 9, 1–6 21 Van Parijs, L. and Abbas, A.K. (1998) Homeostasis and self-tolerance in the immune system: turning lymphocytes off. Science 280, 243–249 22 Zhang, X. et al. (1998) Potent and selective stimulation of memoryphenotype CD81 T cells in vivo by IL-15. Immunity 8, 591–599 23 Tough, D.F. et al. (1996) Induction of bystander T cell proliferation by viruses and type 1 interferon in vivo. Science 272, 1947–1950 24 Garcia, S. et al. (1999) Following the development of a CD4T cell response in vivo: from activation to memory formation. Immunity 11, 163–171 25 Emoto, E. et al. (1995) Proteolytic activation of protein kinase C delta by an ICE-like protease in apoptotic cells. EMBO J. 14, 6145–6156 26 Ghayur, T. et al. (1999) Proteolytic activation of protein kinase C delta by an ICE/CED-3-like potease induces characteristics of apoptosis. J. Exp. Med. 184, 2399–2404 27 Hildeman, D.A. et al. (1999) Reactive oxygen species regulate activation-induced T cell apoptosis. Immunity 10, 735–744 28 Siegal, F.P. et al. (1999) The nature of the principal Type-1 interferonproducing cells in human blood. Science 284, 1835–1837 29 Fitzgerald-Bocarsly, P. (1993) Human natural interferon-alpha producing cells. Pharmacol. Ther. 60, 39–62 30 Maini, M. et al. (1999) Virus-induced CD81 T cell clonal expansion is associated with telomerase up-regulation and telomere length

31

32

33

34

35

36

37 38 39

preservation: a mechanism for rescue from replicative senescence. J. Immunol. 162, 4521–4526 Akbar, A.N. et al. (1994) The specific recognition by macrophages of CD81, CD45RO1 T cells undergoing apoptosis: a mechanism for T cell clearance during resolution of viral infections. J. Exp. Med. 180, 1943–1947 Effros, R.B. and Pawelec, G. (1997) Replicative senescence of T cells: does the Hayflick limit lead to immune exhaustion? Immunol. Today 18, 450–455 Salmon, M. et al. (1994) The progressive differentiation of primed T cells is associated with an increasing susceptibility to apoptosis. Eur. J. Immunol. 24, 892–898 Unumatz, D. et al. (2000) Antigen-independent activation of naive and memory resting T cells by a cytokine combination. J. Exp. Med. 180, 1159–1267 Soares, M.V.D. et al. (1998) IL-7-dependent extrathymic expansion of CD45RA1 T cells enables preservation of a naive repertoire. J. Immunol. 161, 5909–5917 Forster, R. et al. (1999) Two subsets of memory T lymphocytes with distinct homing potentials and effector functions. Nature 401, 708–712 McInnes, I.B. et al. (1996) The role of IL-15 in T cell migration and activation in rheumatoid arthritis. Nat. Med. 2, 175–182 Shah, M.H. et al. (1998) A role for IL-15 in rheumatoid arthritis? Nat. Med. 4, 643 Franz, J.L. et al. (1998) Interleukin-16, produced by synovial fibroblasts, mediated chemoattraction for CD41 T lymphocytes in rheumatoid arthritis. Eur. J. Immunol. 28, 2661–2671

Twins: mirrors of the immune system Marco Salvetti, Giovanni Ristori, Roberto Bomprezzi, Paolo Pozzilli and R. David G. Leslie Twin studies are a powerful tool to assess genetic and nongenetic

I

n the late 19th century, Francis In the century following Galton’s publifactors in multifactorial, immuneGalton, Charles Darwin’s cousin, cation, additional methods were used to mediated diseases. Here, Marco first thought of using twins to invesexploit the value of twins, such as comparing tigate the role of heredity and enviidentical twins with and without a disease. Salvetti and colleagues review 1. Since then, twin ronment in human life Immunological investigations in immuneimportant results from such studies studies in biology, medicine and psychology mediated disease are an outstanding examand highlight their potential value. have had a considerable impact on the scienple of the potential of twin studies. Autotific community; this has generally been for immune diseases affect up to 5% of the Future developments that should the better although, in the case of ‘eugenetic’ population and are a major cause of morbidhelp to realize the potential of twin politics, it has also been for the worse2. ity and mortality. The vast majority of twin studies are discussed. Galton realized that, because twin pairs studies in immune-mediated diseases have (whether identical or nonidentical) share the been carried out in autoimmune diseases, same environment, greater concordance for including insulin-dependent diabetes mellia disease in genetically identical twins compared with nonidentical tus (IDDM), multiple sclerosis (MS), rheumatoid arthritis (RA), systwins probably reflects a role for genetic factors leading to the dis- temic lupus erythematosus (SLE), ankylosing spondylitis (AS) and ease. By contrast, similar concordance rates for a disease in twin coeliac disease. These diseases cover a spectrum of organ-specific pairs (whether identical or nonidentical), or differences between and non-organ-specific diseases and will be the focus of this review. identical twins for a given disease, suggest that the disease is probably In such diseases, the impact of genetic factors is so substantial, and due to non-genetically determined factors. the identity of all immune response genes sufficiently uncertain, that 0167-5699/00/$ – see front matter © 2000 Elsevier Science Ltd. All rights reserved.

J U L Y 342

2 0 0 0 Vo l . 2 1

No.7

PII: S0167-5699(00)01658-3

IT July 2000

14/6/00

9:19 am

Page 343

REVIEW I M M U N O L O G Y TO D AY

twin studies are the only way, in an outbred population, to rule out genetic differences contributing to immunological differences. Yet how much of our current knowledge of the pathogenesis of immunemediated diseases comes from twin studies? Disappointingly little, probably due to the difficulty in obtaining sufficient numbers of twins – only about one in every 250 individuals is an identical twin – and the fact that most twin studies have been cross-sectional and few have been prospective (an important point because, on follow-up, twin pairs could show higher concordance rates).

Twin studies and immune-mediated diseases The impact of environmental factors on disease Genetic and environmental factors are both thought to contribute to the pathogenesis of immune-mediated diseases. The most powerful evidence that immune-mediated diseases are due to environmental factors comes from the study of identical twins. The majority of identical twins with an autoimmune disease have an unaffected twin; that is, they are discordant for the disease (Table 1). While the initially unaffected twin may develop the clinical disease some years after the index twin, the majority remain unaffected on prospective study even, as with IDDM, 40 years after the clinical diagnosis of the index twin7. Nevertheless, even identical twins can differ genetically: X-chromosome inactivation in females can lead to different patterns of mosaicism; differential methylation of CpG islands can result in repression of transcription; and novel somatic rearrangements are involved in the development of T-cell receptors (TCRs) and antibodies. Thus, discordance between identical twins may be determined by nongenetic (epigenetic) factors operating on genetic expression. Having established that environmental factors are important, the next step would be to attempt their characterization. Studies in MS have shown that nonidentical twins have a similar concordance rate to their non-twin siblings. Since nonidentical twins would be expected to share a closer environment than non-twin siblings, these observations suggest that clustering of diseases in families (familial aggregation) is genetic and not environmental. Furthermore, this introduces the important distinction, confirmed by studies on adoptees and half-sibs, between environmental factors that determine the familial risk of a disease (most probably with limited influence compared with genetic factors) and environmental factors that act at the population level (and are strong determinants of disease risk)3. While disease-discordant identical twins should be the perfect test-bed to identify such critical environmental factors, no diseasedetermining environmental factor has, as yet, been found using twin studies.

The impact of genetic factors on disease Twin studies can also be used to estimate the impact of genetic factors on the cause of a disease. A genetic effect is suggested when the concordance rate in identical twins exceeds that in nonidentical twins. This is the case for all the diseases listed in Table 1, where identical twins are more often concordant for the disease than are

Table 1. Concordance rates in identical and nonidentical twin pairs in population-based studies of immune-mediated diseasesa Disease MS RA IDDM SLE

Identical twin pairs (%)

Nonidentical twin pairs (%)

Refs

26.7 12.3 13 33

3.5 3.5 2.5 0

3 4 5 6

Abbreviations: IDDM, insulin-dependent diabetes mellitus; MS, multiple sclerosis; RA, rheumatoid arthritis; SLE, systemic lupus erythematosus. a The concordance rates are consistently less than 100% and are higher in identical than nonidentical pairs.

nonidentical twins, indicating that genetic factors are important. The concordance rates in nonidentical twins are less than 50% of those of identical twins, suggesting a polygenic model for these major immune-mediated diseases (in the case of a single dominant gene, the risk for nonidentical twins would be half that of identical twins; with increasing numbers of genes, the difference between identical and nonidentical twins would be increasingly higher). However, even infectious diseases such as polio or tuberculosis show differences in identical and nonidentical twin pair concordance rates (36% and 6% respectively for polio8, and 51% and 26% respectively for tuberculosis9). Taken together, these studies are consistent with genetic susceptibility influencing the appearance of clinical disease even when environmental factors play a major role in causing the disease. This conclusion is further strengthened by migrant studies10. Children with a Sardinian heritage maintain the same high incidence of IDDM abroad as the indigenous Sardinian population, even if born in a region of low disease incidence.

The impact of HLA genes on autoimmune disease An association between major histocompatibility complex (MHC) HLA type and disease has been noted in many autoimmune conditions. The relative contribution of the HLA locus to disease can be estimated by comparing concordance rates in HLA-identical but dizygotic twins (or siblings) with that in monozygotic twins. Alternatively, the impact of a gene on the development of a disease (penetrance) can be estimated by comparing susceptibility genes in identical twin pairs who are discordant and concordant for the disease. AS is strongly associated with HLA-B27. In one study, 6/8 (75%) HLA-B271 identical twin pairs were concordant for the disease, as compared with only 4/32 (12.5%) nonidentical twins and 4/15 (27%) HLA-B271 nonidentical twin pairs11. No nonidentical twin pairs have been reported to be concordant for AS but discordant for HLA-B27. These observations suggest that HLA-B27 is almost essential to the development of AS but that other genetic, as well as environmental, factors determine which HLA-B271 individuals will manifest the disease.

J U L Y Vo l . 2 1

2 0 0 0 No.7

343

IT July 2000

14/6/00

9:19 am

Page 344

REVIEW I M M U N O L O G Y T O D AY

A similar suggestion can be drawn from twin studies in RA, as demonstrated by the importance of the HLA-DR shared epitope for disease concordance and the gene-dose effect in RA susceptibility12,13. The potential of twin studies to define the role of HLA and non-HLA genes has not been exploited in other immune-mediated diseases, the main problem being the limited numbers of twin pairs available for study. Developments in population-based twin ascertainment allied to the identification of those non-HLA genes that are probably involved in disease susceptibility could lead to the wider application of twin studies in genome research.

Limitations of twin studies in assessing genetic impact on disease There are other problems with assessing the genetic contribution to disease using the twin study method. First, identical twins may have a more similar environment both in utero and in childhood, which would lead to an over-estimate of heritability. Second, in contrast to nonidentical twins, identical twins are always the same sex. Many autoimmune diseases have a clear sex bias; for example, RA, MS, SLE and autoimmune thyroid disease are more common in females. Third, twin studies are hindered by potential biases in ascertainment. In the traditional ‘clinic-based’ approach, identical twin pairs concordant for a disease, or with severe disease, are more likely to be identified. In the ‘population-based’ approach, individuals are identified first as twins and are then assessed for illness. These population-based studies must ascertain large numbers of twin pairs to detect sufficient numbers of affected twins. Finally, most twin studies have been cross-sectional, but, by following twins for a longer period, it might be possible to detect a higher concordance rate3.

Immunological studies Humoral immune responses Identical twins show a similar genetic regulation of the production of antibodies, even when they are reared apart14. A recent study confirmed that such a genetic control remains largely unaltered by chronic autoimmune stimulation14. Moreover, genetic factors can regulate the level and extent of antibody production. Autoantibodies are correlated with certain HLA types (e.g. HLA-DR4 and insulin autoantibodies16 or rheumatoid factor17). A study of IgM and IgG rheumatoid factor isotypes in 70 identical and 84 nonidentical twins discordant for RA concluded that genetic factors were important in determining

the levels and frequency of these isotypes; for example, IgM and IgG positivity was higher in healthy identical than healthy nonidentical twins18. In agreement with such a genetic effect, combinations of IDDM-associated autoantibodies were more often found in healthy identical than healthy nonidentical twins of IDDM patients19. Other studies in immune-mediated diseases highlight differences between healthy and diseased co-twins, implying post-zygotic, epigenetic or environmental factors influencing the induction of autoantibodies. In HLA-DR41 identical twin pairs discordant for RA, the twins with RA had higher titres of antibodies towards proteins containing QKRAA, an amino acid sequence in the third hypervariable region common to the HLA-DRb1 alleles associated with the disease20. In IDDM, disease-discordant identical twins are often also discordant for disease-associated autoantibodies. Moreover, autoantibodies are as often detected in healthy identical as nonidentical twins of patients with an autoimmune disease, consistent with a common environmental effect; for example, IDDM-associated autoantibodies were detected in 6/18 identical twins and 6/18 nonidentical twins of IDDM patients19.

Cellular immune responses Cellular immune responses are under strong genetic influence. However, the peripheral TCR repertoire is shaped not only by the availability of germline receptor elements, but also by their rearrangement, thymic selection and clonal stimulation. Studies in identical twins may help define the relative importance of genetic and epigenetic factors on the shaping of the healthy (Table 2) and diseased (Table 3) immune repertoire. Concerning the shaping of the healthy TCR repertoire, shared complementarity-determining region 3 (CDR3) motifs could be identified between healthy identical twins in the response to epitopes of myelin basic protein (MBP) and of Mycobacterium bovis 65 kDa heat shock protein21. Similarly, the Jb repertoires of identical twins were found to be more similar than those of unrelated individuals22. Although shared epigenetic factors cannot be ruled out, these results may be interpreted as suggesting a predominant genetic effect. However, they contrast with a later report, again in healthy identical twins, where a common TCRVa and/or TCRVb gene usage for shared epitope recognition by MBP-specific T-cell lines was infrequent and there was no significant intrapair concordance, stressing the importance of epigenetic events23. In accordance with this view, epitope recognition of MBP-specific T-cell lines in healthy identical

Table 2. Influence of epigenetic factors on the immune repertoire of healthy individuals Condition

Influence on immune repertoire

No influence on immune repertoire

Healthy identical twins

Diverse TCRVa and/or TCRVb gene usage in the response to shared T-cell epitopes23 Diverse patterns of epitope recognition24

Shared CDR3 in the response to T-cell epitopes21 Similar Jb repertoires22 Same Va usage in response to antigens28

Abbreviations: CDR, complementarity-determining region; J, joining; TCR, T-cell receptor; V, variable.

J U L Y 344

2 0 0 0 Vo l . 2 1

No.7

IT July 2000

14/6/00

9:19 am

Page 345

REVIEW I M M U N O L O G Y TO D AY

twins24 was distinct in 6/8 pairs, and there was a complete absence of concordant epitope recognition within two other pairs. In immune-mediated diseases, the resting peripheral T-cell Vb repertoire appears to be virtually identical in identical twins irrespective of discordance for IDDM (Ref. 25) or RA (Ref. 26). Similarly, analysis of six identical twin pairs discordant for RA showed that their Jb gene repertoires were more similar than those of unrelated individuals, and Jb expression was not influenced by RA (Ref. 22). Other studies showed differences in immune repertoire between twin pairs discordant for RA: one suggested that there may be differences in the TCR Vg repertoire26; another showed RA-associated alterations of the TCRBV-BJ combination repertoire in the affected co-twin27. Evidence for disease-associated skewing of the TCR repertoire has also come from studies on the putative antigen in MS (MBP): identical twins can show differences in TCR Va usage after repeated stimulation either with MBP or tetanus toxoid28. Overall, the above data suggest that the gene expression of TCRs by peripheral lymphocytes is largely regulated by genetic factors. However, alterations in the TCR repertoire may occur both in healthy identical twins and in the affected co-twin of pairs discordant for immune-mediated disease. Post-zygotic DNA changes, epigenetic events or environmental factors may account for such discordance in the TCR gene usage. Disappointingly, twin studies have failed to identify diseaserelated epitopes. Characterization of proliferative and cytotoxic responses, as well as the epitope specificity and HLA restriction of a large number (.600) of MBP-specific T-cell lines from six twin pairs, three discordant and three concordant for MS, failed to detect any difference that could be attributed to the disease29. Those peptide epitopes of MBP recognized most frequently by the T-cell lines were already known from previous studies of MS to be immunodominant. Perhaps a similar study, performed serially over a two- or three-year period, will be able to clarify whether responses that remain stable over time differ between affected and non-affected twins. It is possible that the critical difference between identical twins discordant for an autoimmune disease lies not in the T-cell epitope recognition but in the functional response of the cell. The frequency and functional characteristics of CD42CD82Va24JaQ2 T cells [which might be important in regulating the T helper 1 (Th1)-cell-mediated tissue damage in IDDM] were studied in siblings and identical twins with IDDM. Clones from diabetic patients secreted only interferon g (IFN-g) on stimulation30. By contrast, nearly all the clones from

normal subjects, and even the nondiabetic identical twins of IDDM patients, secreted both interleukin 4 (IL-4) and IFN-g. Thus, it was the loss of the capacity of these Va24JaQ T cells to secrete IL-4 that was associated with the presence of IDDM. Since identical twins were discordant for this feature, it is unlikely that the change is genetically determined. It remains to be established whether the Th1 bias of these T cells is a consequence of the disease or predisposes to the condition. As with T-cell studies of epitope recognition, the potential of twins in the study of T-cell function is yet to be realized.

Predicting disease Since the healthy identical twin of patients with an autoimmune disease is at high risk of developing the same disease, it has been possible to study twins before the onset of clinical disease. Evidence that IDDM results from an environmental event operating in early childhood derives in part from such twin studies31. Indeed, twin studies first revealed the chronic nature of the autoimmune process in the prediabetic period32. These studies showed immune changes, including disease-associated antibodies in peripheral blood, months and even years before the clinical onset of diabetes. The immune changes were accompanied by a progressive decline in insulin secretory capacity, consistent with a gradual and progressive destruction of the insulin-secreting islet cells. A recent study suggested that the variation in the age at clinical onset of IDDM and, by implication, the disease incubation period following the early induction of autoimmunity, is strongly genetically influenced7. The age of onset in 116 identical twin pairs concordant for IDDM was strikingly correlated (correlation coefficient 0.94) and the correlation for age at diagnosis was higher in identical than nonidentical twins. Extensive twin studies of the co-twins of patients with IDDM have revealed numerous disease-related immune changes with variable predictive power. Changes that increase the risk of developing IDDM in such twins include autoantibodies to islet antigens, activation of T cells (particularly CD81 T cells), increased expression of CD45RA (a marker of naive T cells), increased serum levels of macrophage-derived cytokines, impaired glucose tolerance and decreased insulin response to intravenous glucose challenge33–36. The positive predictive value of combinations of disease-related autoantibodies is as high as 100% (Ref. 18). In broad terms, the predictive power of immune changes in identical twins has been higher than that in siblings, giving us an indicator of the potential of combining genetic

Table 3. Influence of epigenetic factors on the immune repertoire of individuals with immune-mediated diseases Condition IDDM-discordant identical twins RA-discordant identical twins MS-discordant identical twins

Influence on immune repertoire repertoire26

Different T-cell Vg Different TCRBV-BJ repertoire27 Different Va usage in response to antigens28

No influence on immune repertoire Identical T-cell Vb repertoire25 Identical T-cell Vb repertoire26 Jb expression not influenced by RA22

Abbreviations: IDDM, insulin-dependent diabetes mellitus; MS, multiple sclerosis; RA, rheumatoid arthritis; for other abbreviations, see footnote to Table 2.

J U L Y Vo l . 2 1

2 0 0 0 No.7

345

IT July 2000

14/6/00

9:19 am

Page 346

REVIEW I M M U N O L O G Y T O D AY

and immune markers in predicting IDDM in both family and population studies. However, in twins of IDDM patients, not all immune, and even metabolic, changes inevitably result in diabetes. Thus, increased levels of activated T cells had a positive predictive value of only 60%, a decrease in the first-phase insulin response to intravenous glucose gave a positive predictive value of 58% and impaired glucose tolerance only 33% (Refs 33–36). Owing to the small numbers of twins available for study, it has been difficult to extend these observations in twins of IDDM patients to other autoimmune diseases. In MS, cerebrospinal fluid immunoglobulin abnormalities37 (oligoclonal bands), magnetic resonance38,39 and visual evoked response alterations39 (i.e. subclinical defects of the optic pathways that are detectable through electrophysiological techniques) are frequently detected in the healthy cotwin. While the predictive value of such abnormalities is still uncertain, these studies, taken together, suggest that the pathological process associated with autoimmune conditions encompasses a spectrum of immune changes and target organ damage that do not inevitably lead to clinical disease. Moreover, the identification of a long prodrome before the onset of clinical symptoms in IDDM raises the possibility that we can predict, and possibly prevent, many of these diseases.

Future developments If the potential of twin studies is to be realized, a spirit of close collaboration must exist between clinicians and scientists, including epidemiologists, immunologists and geneticists. To date, twin studies have provided a patchy outline of immune responses associated with autoimmune diseases, usually in limited numbers of twin pairs with little or no prospective analysis. Important issues that twin studies could resolve, such as the genetic impact on disease severity (to explain the variation between patients for arthritis, neurological deficit or vascular damage), have not been addressed. To improve epidemiological and immunological twin research it will be necessary to increase the numbers of pairs under study. This can be attained through the construction of population-based twin registries. Scandinavian countries have a long-standing tradition in this respect and similar registries are now being assembled in other countries40. The creation of centralized facilities for the storage of large amounts of frozen peripheral blood mononuclear cells and sera should be an early goal in establishing these registries, considering also that this would increase the chances of sampling the material from subjects free of any therapy. Recent technological progress highlights the potential of twin studies: a prelude to these developments has been the successful application of techniques such as the differential display of thousands of genes to quantify gene expression; for example, in the study of identical twins discordant for MS (Ref. 41). This analysis has led to the detection of a deficient expression of the inhibitory transcription factor Sp3 in mononuclear blood cells of the affected co-twin; the defect was confirmed by comparing a series of MS patients with control subjects. This study anticipates what might be attained in the near future by employing new and more powerful ‘functional

J U L Y 346

2 0 0 0 Vo l . 2 1

No.7

genomics’ techniques such as microarray analysis. As this technology allows the simultaneous monitoring of hundreds and thousands of genes at their expression level, through the quantitative measurement of the amount of mRNA transcripts, it is well suited for investigating complex diseases. Early applications of the microarray-based approach to the study of human autoimmune disorders, namely RA and inflammatory bowel disease, have been particularly promising. Not only has it allowed identification of novel genes associated with the disease but, more importantly, it has identified distinct patterns of gene expression42. This recognition of distinct gene expression profiles is of particular interest; for instance, using DNA microarrays, Alizadeh and colleagues were able to identify two molecularly distinct forms of diffuse large B-cell lymphoma that had gene expression patterns indicative of different stages of B-cell differentiation43. Cluster analysis of gene expression patterns thus provides a molecular classification of tumours, revealing previously undetected subtypes of cancer that are clinically significant44,45. The recognition of specific features of histological and physiological variation in tumour samples using microarrays suggests that this will be a potent tool to dissect the heterogeneity and complexity of multifactorial disorders. It may be an arduous task to discern differences in mRNA expression that are pathogenetically relevant and distinct from those resulting simply from the many different genetic backgrounds in an outbred population. Performing such studies on identical twins discordant and concordant for a disease could overcome this problem. Since microarray analysis is a relative of differential display in the functional genomics family of techniques, its application in twins should be possible – the more so with the imminent availability of gene sequences. Twin studies might have other potential applications in the assessment of aspects of potential relevance in the pathogenesis of autoimmune diseases. Possible precipitating factors of autoimmune conditions that await definition include epigenetic effects due to DNA or protein modifications (methylation patterns or acetylation of histones), the accumulation of somatic mitochondrial DNA mutations, X chromosome inactivation, and the presence of transcriptionally active endogenous retroviral sequences. Research has come a long way since Francis Galton conceived the potential of twin studies to detect genetic traits. Today, it is possible to define genetic and epigenetic features of a disease using twin studies. New technologies are knocking at the door, and once we have overcome the limited number of twins available for study, the full power of twin studies should be realized.

The authors are supported by the British Diabetic Association, the Diabetic Twin Research Trust, the Istituto Superiore di Sanità and the Fondazione Italiana Sclerosi Multipla. We thank R. Tosi for his critical review of the manuscript.

Marco Salvetti, Giovanni Ristori and Roberto Bomprezzi are at the Dept of Neurosciences of the University of Rome ‘La Sapienza’, 00185 Rome, Italy; Paolo Pozzilli ([email protected]) is at the University Campus Biomedico, 00100 Rome, Italy and at the Dept of Diabetes

IT July 2000

14/6/00

9:19 am

Page 347

REVIEW I M M U N O L O G Y TO D AY

and Metabolism, St Bartholomew’s Hospital and Medical College, London, UK EC1A 7BE; R. David G. Leslie is at the Dept of Diabetes and Metabolism, St Bartholomew’s Hospital and Medical College, London, UK EC1A 7BE.

22

23

References 1 Galton, F. (1875) Frazer’s Magazine 12, 566–576 2 What we learn from twins. The mirror of your soul. (1998) The Economist 1, 74–76 3 Ebers, G.C. and Sadovnick, A.D. (1994) The role of genetic factors in multiple sclerosis susceptibility. J. Neuroimmunol. 54, 1–17 4 Aho, K. et al. (1986) Occurrence of rheumatoid arthritis in a nationwide series of twins. J. Rheumatol. 13, 899–902 5 Kaprio, J. et al. (1992) Concordance for type 1 (insulin-dependent) and type 2 (non-insulin-dependent) diabetes mellitus in a populationbased cohort of twins in Finland. Diabetologia 35, 1060–1067 6 Jarvinen, P. et al. (1992) Systemic lupus erythematosus and related systemic diseases in a nationwide twin cohort: an increased prevalence of disease in MZ twins and concordance of disease features. J. Int. Med. 231, 67–72 7 Fava, D. et al. (1998) Evidence that the age at diagnosis of IDDM is genetically determined. Diabetes Care 21, 925–929 8 Vogel, F. and Motulsky, A.G. (1978) Human Genetics. Problems and Approaches. Springer Verlag 9 Grufferman, S. et al. (1987) Increased sex concordance of sibling pairs with Behçet’s disease, Hodgkin’s disease, multiple sclerosis, and sarcoidosis. Am. J. Epidemiol. 126, 365–369 10 Muntoni, S. et al. (1997) Incidence of insulin-dependent diabetes mellitus among Sardinian-heritage children born in Lazio region, Italy. Lancet 349, 160–162 11 Brown, M.A. et al. (1997) Susceptibility to ankylosing spondylitis in twins: the role of genes, HLA, and the environment. Arthritis Rheum. 40, 1823–1829 12 Jawaheer, D. et al. (1994) ‘Homozygosity’ for the HLA-DR shared epitope contributes the highest risk for rheumatoid arthritis concordance in identical twins. Arthritis Rheum. 37, 681–686 13 De Vries, N. et al. (1997) HLA-DRB1 in eight Finnish monozygotic twin pairs concordant for rheumatoid arthritis. Tissue Antigens 49, 277–279 14 Kohler, P.F. et al. (1985) Genetic regulation of immunoglobulin and specific antibody levels in twins reared apart. J. Clin. Invest. 75, 883–888 15 Kohsaka, H. et al. (1996) The human immunoglobulin V(H) gene repertoire is genetically controlled and unaltered by chronic autoimmune stimulation. J. Clin. Invest. 12, 2794–2800 16 Ziegler, A.G. et al. (1991) HLA-associated insulin autoantibody formation in newly diagnosed type I diabetic patients. Diabetes 40, 1146–1149 17 Silman, A.J. et al. (1991) Rheumatoid factor detection in the unaffected first degree relatives in families with multicase rheumatoid arthritis. J. Rheumatol. 18, 512–515 18 Macgregor, A.J. et al. (1995) Rheumatoid factor isotypes in monozygotic and dizygotic twins discordant for rheumatoid arthritis. J. Rheumatol. 22, 2203–2207 19 Hawa, M. et al. (1997) Value of antibodies to islet protein tyrosine phosphatase-like molecule in predicting type 1 diabetes. Diabetes 46, 1270–1275 20 La Cava, A. et al. (1997) Genetic bias in immune responses to a cassette shared by different microorganisms in patients with rheumatoid arthritis. J. Clin. Invest. 100, 658–663 21 Hawes, G.E. et al. (1995) Limited restriction in the TCR-alpha beta V region usage of antigen-specific clones. Recognition of myelin basic protein (amino acids 84–102) and Mycobacterium bovis 65-kDa heat shock protein (amino acids 3–13) by T cell clones established from

24

25

26

27

28 29

30 31 32 33

34 35

36 37 38 39 40 41

42

43 44 45

peripheral blood mononuclear cells of monozygotic twins and HLA-identical individuals. J. Immunol. 154, 555–566 Nanki, T. et al. (1996) Genetic control of T cell receptor BJ gene expression in peripheral lymphocytes of normal and rheumatoid arthritis monozygotic twins. J. Clin. Invest. 98, 1595–1601 Shanmugam, A. et al. (1996) TCR alpha beta gene usage for myelin basic protein recognition in healthy monozygous twins. J. Immunol. 156, 3747–3754 Shanmugam, A. et al. (1994) Healthy monozygous twins do not recognize identical T cell epitopes on the myelin basic protein autoantigen. Eur. J. Immunol. 24, 2299–2303 Malhotra, U. et al. (1992) Variability in T cell receptor V beta gene usage in human peripheral blood lymphocytes. Studies of identical twins, siblings, and insulin-dependent diabetes mellitus patients. J. Immunol. 149, 1802–1808 Kohsaka, H. et al. (1993) The expressed T cell receptor V gene repertoire of rheumatoid arthritis monozygotic twins: rapid analysis by anchored polymerase chain reaction and enzyme-linked immunosorbent assay. Eur. J. Immunol. 23, 1895–1901 Mizushima, N. et al. (1997) HLA-dependent peripheral T cell receptor (TCR) repertoire formation and its modification by rheumatoid arthritis. Clin. Exp. Immunol. 110, 428–433 Utz, U. et al. (1993) Skewed T-cell receptor repertoire in genetically identical twins correlates with multiple sclerosis. Nature 364, 243–247 Martin, R. et al. (1993) Myelin basic protein-specific T-cell responses in identical twins discordant or concordant for multiple sclerosis. Ann. Neurol. 34, 524–535 Wilson, S.B. et al. (1998) Extreme Th1 bias of invariant Valpha24JalphaQ T cells in type 1 diabetes. Nature 391, 177–181 Leslie, R.D.G. and Elliott, R.B. (1994) Early environmental events as a cause of IDDM. Evidence and implications. Diabetes 43, 843–850 Srikanta, S. et al. (1983) Type I diabetes mellitus in monozygotic twins: chronic progressive beta cell dysfunction. Ann. Intern. Med. 99, 320–326 Tun, R.Y.M. et al. (1994) Importance of persistent cellular and humoral immune changes before diabetes develops: prospective study of identical twins. Br. Med. J. 308, 1063–1068 Peakman, M. et al. (1996) Persistent activation of CD81 T-cells characterizes prediabetic twins. Diabetes Care 19, 1177–1184 Beer, S.F. et al. (1990) Impaired glucose tolerance precedes but does not predict insulin-dependent diabetes mellitus: a study of identical twins. Diabetologia 33, 497–502 Lo, S.S. et al. (1992) Altered islet beta-cell function before the onset of type 1 (insulin-dependent) diabetes mellitus. Diabetologia 35, 277–282 Williams, A. et al. (1980) Multiple sclerosis in twins. Neurology 30, 1139–1147 Kinnunen, E. et al. (1988) Genetic susceptibility to multiple sclerosis. A co-twin study of a nationwide series. Arch. Neurol. 45, 1108–1111 Thorpe, J.W. et al. (1994) British Isles survey of multiple sclerosis in twins: MRI. J. Neurol. Neurosurg. Psychiatry 57, 491–496 Salvetti, M. et al. (1997) Italian population yields world’s largest twin registry. Nat. Med. 3, 1176 Grekova, M.C. et al. (1996) Deficient expression in multiple sclerosis of the inhibitory transcription factor Sp3 in mononuclear blood cells. Ann. Neurol. 40, 108–112 Heller, R.A. et al. (1997) Discovery and analysis of inflammatory disease-related genes using cDNA microarrays. Proc. Natl. Acad. Sci. U. S. A. 94, 2150–2155 Alizadeh, A.A. et al. (2000) Distinct types of diffuse large B-cell lymphoma identified by gene profiling. Nature 403, 503–511 Golub, T.R. et al. (1999) Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531–537 Perou, C.M. et al. (1999) Distinctive gene expression patterns in human mammary epithelial cells and breast cancers. Proc. Natl. Acad. Sci. U. S. A. 96, 9212–9217

J U L Y Vo l . 2 1

2 0 0 0 No.7

347