By David N. Leff

There's a "perfect-murder" whodunnit waiting to be written, based on the one and only way to foil forensic DNA detection. Expert witnesses, prosecutors and defense lawyers point to the likelihood of matching the DNA in a blood or sperm stain with another person as one in a million or a billion. The one exception to this scenario is that there's one perpetrator whose DNA is identical to that of his or her identical twin.

By the opposite - life-saving - token, there's only a single organ donor with a perfect guarantee against immune rejection of a transplanted graft - again, an identical twin.

So you might expect that if one of a pair of monozygotic twins - by definition cloned from the union of a single egg and sperm - contracts Type I diabetes, so will his or her carbon-copy sibling. It's not necessarily so. Early-onset or insulin-dependent diabetes mellitus (IDDM) - Type I (once called "juvenile") - is an autoimmune disorder in which the body's immune T lymphocytes destroy the pancreatic beta cells that secrete insulin.

Even that exception has an exception:

When the immune system is involved, not all identical twins are truly indistinguishable - concordant. "Monozygotic twins," observed research immunologist S. Brian Wilson, "were thought to be about 30 percent to 40 percent concordant. More recent studies are suggesting that it's closer to 50 percent to 60 percent concordance in terms of developing diabetes." That is, a little more than half of them will share the Type I disease.

Given their embryonic origin, why not 100 percent?

"That's the 64 million dollar question," said Wilson. "In fact, identical twins are identical from the immunologic perspective with two important differences: T lymphocyte rearrangements and immunoglobulin [Ig] rearrangements are random. So they do not share the same T-cell repertoire or Ig repertoire.

"Nor do identical twins absolutely share whatever the nefarious environmental trigger is that helps initiate IDMM. You can imagine two twins," Wilson explained. "One went to camp, one didn't, and the unfortunate one got some sort of enteric virus, which triggered Type I diabetes."

Fingering Antigen In MHC Lineup

"Autoimmune insulin dependence," he went on, "is a polygenic disorder, dominated by the MHC Class II locus, and that significant environmental component." The chromosomal region of MHC - the major histocompatibility complex - contains genes encoding cell-surface glycoproteins that regulate interactions among immune-system cells. Thus, its Class II molecules deal with the growth and differentiation of T lymphocytes and B cells. It also manages antigenic presentation, in which an attacked target cell expresses antigen on its surface for the attention of antibody-generating B cells. And certain cell-killing T cells recognize antigens brought forward by Class I MHC molecules.

Tracking down these infinitely devious autoimmune cabals, and bringing them to clinical justice, is a major challenge to the immunologic research community in which Wilson labors. He holds joint appointments at Massachusetts General Hospital and the Dana Farber Cancer Institute, both in Boston. Today's issue of the Proceedings of the National Academy of Sciences (PNAS), dated June 20, 2000, carries an article, of which Wilson is lead author, titled "Multiple differences in gene expression in regulatory . . . cells from identical twins discordant for type 1 diabetes."

"The overall finding of this work," Wilson told BioWorld Today, "is an analysis of a set of very interesting T-cell clones, in that they are thought to control the flavor of immunologic responses to various tumors, pathogens and autoimmune disorders. We had previously published the alterations in their function in paired T-cell clones in autoimmune diabetes discordant twins."

When a killer T cell goes into action against its target cell, it turns loose a phalanx of inflammatory cytokines, led by interferon (INF) and interleukin (IL).

"These cells make predominantly INF-gamma and IL-4," Wilson pointed out, "and promote cell-mediated immune responses - in this case, inappropriately directed at the pancreatic beta cell. If we modulate the T-helper (T_H) cell subset in animal models, we can markedly ameliorate the autoimmune disorder. So one of the going hypotheses is that there is a dysfunction in the regulation of T-cell development, which does not stem the tide of the T_H 1-like phenotype.

"So what we did," he recounted, "was first look at a mouse cell as a candidate for this type of function. So we studied this family of T cells in an unusual set of four human monozygotic twins, plus one set of triplets, that were discordant for autoimmune diabetes. We quantitated them, cloned them, and looked at their functional phenotype in terms of their cytokine secretion. And we discovered," Wilson went on, "that all the clones derived from the non-diabetic individuals made INF-gamma and IL-4 equally, whereas all the clones derived from the diabetic monozygotic twin made INF-gamma exclusively, and no IL-4."

From DNA Chips, T Cells By The Million

"The ultimate role of these T cells, we determined, was to regulate the T_H phenotype in part by secreting IL-4. You can see how the ones that came from the diabetics could not. So in this study we decided to investigate what was the molecular basis for that discordant phenotype. We noticed right away, from inhibitor and calcium flux studies, that it was not a simple problem, so we approached it by doing gene chip analysis. What we did was expand individual T-cell clones, such that we were able to obtain synchronous cultures of 10 million to 20 million T cells per data point.

"And what we discovered was that there were essentially six patterns of genes - those induced in the non-diabetic, and unchanged in the diabetic.

"What we've done," he summed up, "is identify a reproducibly accessible family of T cells that have an important regulatory function, which is aberrant in Type I diabetics. So what we're going to do now is look at the expression profile that we think explained some of the discordance, and predict a function for these T cells that wasn't previously known, and build a nice, representative database.

"Interestingly," Wilson concluded, "the family of T cells that we studied by gene chips are present in reduced numbers - and their functional status has not been determined, for patients with other autoimmune diseases, namely, rheumatoid arthritis multiple sclerosis, probably inflammatory bowel disease. The data is not yet available for lupus patients."