| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Editorial |
Mount Sinai School of Medicine New York, New York 10029
Address all correspondence and requests for reprints to: Dr. T. F. Davies, Mount Sinai School of Medicine, Box 1055, 1 Gustave L. Levy Place, New York, New York 10029. E-mail: terry.davies{at}mssm.edu.
Men are simple creatures. They have one X chromosome and one Y chromosome with often distinct functions and controls. Women are more complex in that they have two similar X chromosomes, resulting in a double dose of often similar genes among the 1098 genes coded for on the X chromosome. These chromosomes appear to battle it out within the mammalian cell to see which will predominate and actually function. The loser is inactivated by the winner, and this phenomenon of dosage compensation is called X inactivation (1). Most of the battle for X chromosome superiority appears to take place in early embryonic life and results in different cells expressing the paternal or the maternal X chromosome genes. The resulting stable state is inherited through subsequent mitotic cell divisions, resulting in females becoming internal mosaics in addition to the external mosaicism resulting from fetal cell transfer later in life (2).
A multifunctional domain on the X chromosome to be inactivated, called the X inactivation center (XIC), is present in the Xq13 region (3). Using transgenic and knockout mice, this domain has been further refined to reveal the XIST gene, which encodes an untranslated RNA that coats the X chromosome and, together with associated factors, silences it (4). The mechanism of choice—deciding which chromosome becomes inactive—is complex and appears to be mostly random. However, in mice, another element within XIC has been identified that may skew inactivation toward the maternal or paternal chromosome (the X controlling element, Xce). Importantly, different alleles of Xce have been identified that display different strengths of skewing (5).
Because X chromosome choice is assumed to be mostly random, the result is generally 50% of cells expressing the paternal genes and 50% expressing the maternal genes. Under such circumstances, both paternal and maternal antigens will be recognized by the immune system within the thymus, and any T cells that have high affinity for such antigens will be deleted by apoptosis (6). It is immediately obvious that, if such deletion of self-reactive T cells was not successful for one set of X antigens, or even just a few antigens, their expression in the periphery may be seen as foreign and the stage would be set for an autoimmune response as suggested for other antigens (7). This has become an attractive explanation for the autoimmune diseases, including autoimmune thyroid disease (AITD).
The autoimmune diseases are associated with genes that, by definition, are of course Mendelian but result in non-Mendelian inheritance. This unpredictable form of inheritance has been called complex and distinguishes them from those diseases inherited by a single gene. In fact, almost all the single genes responsible for a single disease have now been identified, but the complex diseases have remained difficult to analyze. Although progress has been made in some, for example Crohn’s disease, with the identification of NOD2 (8), the majority have resulted in the identification of many weakly influential genes that, presumably, when acting together result in recognizable disease. We cannot even be sure that entirely different complexes of weakly contributing genes may result in similar clinical disease because pathology may have limited readout ability. The AITDs are some of the most common examples of complex genetic diseases that certainly appear at this time to be in a quagmire of information. Unlike insulin-dependent diabetes, the influence of the HLA gene region in AITD is limited (9, 10) as are the two other genes common to autoimmune diseases in general, CTLA-4 and LYP (9, 11). Furthermore, the results of whole-genome screening have been surprisingly disappointing with only a few sites being confirmed in multiple studies, and even larger numbers of families are needed to clarify the data (12). In addition to some as-yet-uncharacterized sites, such screening has identified the thyroglobulin gene as a disease-specific influence, but the degree of influence in all but a small percentage of patients is also limited (13, 14).
Hence, the search is on for a better understanding of the genetics of AITD. The phenomenon of X chromosome inactivation is attractive because of the unexplained female predominance of autoimmune disease in females and, furthermore, the relatively low concordance between identical (monozygotic) twins—because the twinning event may take place in some twins before X chromosome inactivation occurs in embryogenesis (15). As discussed earlier, the skewing of chromosome selection has been linked to a region on the X chromosome, suggesting X-linked inheritance of a potential bias in the normally random selection process (1).
Up to 50% of Turner’s syndrome patients show stigmata of AITD, and studies on the correlation between karyotype and phenotype have suggested that a gene on chromosome Xq—the same region as reported for the XIC discussed earlier—may play a role in AITD (16). Although other reports have linked AITD to X chromosome loci we were not able to confirm this in our studies to date (9), but association studies on the XIC region alleles have not yet been performed in humans.
Reports from studies of patients with systemic lupus erythematosus, insulin-dependent diabetes, and rheumatoid arthritis have not shown any skewing of X chromosome function away from the theoretical 50:50, although it has been reported in scleroderma (17). It is, therefore, an important observation when such skewing is observed in autoimmune thyroid disease as reported by Brix et al. (18) in this issue. This group has been studying AITD in twins for a number of years (19). They were able to study their twin cohort, concordant for AITD, for peripheral blood lymphocyte X chromosome inactivation and found that 34% of the twin pairs concordant for AITD showed more than an 80% skew compared with 11% of twin controls. Similar data were derived from examining twins discordant for AITD. Even more impressive was the derivation of an odds ratio for the increased risk of developing AITD of 9.0, one of the highest such ratios ever reported in AITD.
Confirmation of these types of data must make sure that this phenomenon is not just reality but also not just restricted to twins. Furthermore, the final say is most likely to be found at the tissue level. Demonstrating skewing of X chromosome expression within the thyroid gland of patients with AITD is experimentally straight forward. In addition, observations that thyroid antigen (thyroglobulin) haplotypes are associated with AITD are consistent with the lack of tolerance for a maternal or paternal form of thyroid antigen (14). So X inactivation may be another plum for the pie of thyroid knowledge, but the pie itself is not yet there to eat.
Footnotes
Abbreviations: AITD, Autoimmune thyroid disease; Xce, X controlling element; XIC, X inactivation center.
Received September 19, 2005.
Accepted September 22, 2005.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
