Wednesday, January 8, 2014

The "New World Syndrome" genetic basis: has it been found?

In several papers in the early 1980s, I was working with colleagues Ranajit Chakraborty and Bob Ferrell and others, on disease problems that we had noticed in Mexican Americans.  Among these were a particular distribution of body fat (a characteristic obesity pattern), gallstone disease and type 2 (at the time, adult onset) diabetes.

In that work we also noticed a correlation between risk of these problems in Mexican Americans and a pandemic of the same in Amerindians in North America (data from Central and South America were less and less clearly interpretable).  We developed a hypothesis that these various conditions, and perhaps other consequences of them, were not only due to vulnerable genotypes but that the genotypes may have arisen in the ancestors of Amerindians.  Because various traits were variously associated in the available data, our idea was that this was a syndrome, or collection of physiologically associated traits, that had a New World origin.  Each person might have a different set of the associated conditions.

Because the problem seemed continental in scope, we thought this may be due not just to social conditions or poverty, but to specific susceptible genotypes.  In a 1984 paper, we called this a 'New World syndrome' (NWS) of conditions.

There was little in the way of actual genes that could be identified at the time of this work, and we used mainly epidemiological data on prevalence and age of onset, and on family correlations.  But a few genetic markers (variable genes identified by blood-typing and protein analysis) were available and we could estimate the European and Amerindian fractions of ancestry in the Texas Mexican-Americans that were the main object of our particular studies.

Even in the absence of actual genotypes in detail, and certainly not any specifically related to the biology of these traits, we could demonstrate associations between Native American admixture and  prevalence in a way that was consistent with a single-gene effect.  Even if traits like diabetes are usually genomically complex, the severity, early onset, high prevalence and syndromic nature of the NWS, and continental dispersion of a trait in such a recently settled part of the world, made it seem plausible that the inherited effect was in some key enzyme or pathway--that is, might be genetically simple. 

The restriction of this particular set of traits to Native Americans and those admixed with them, and their apparent absence (that is, as a syndrome or at the same level of risk), and the admixture association, suggested to us that the genetic variant(s) responsible had come over into the New World through the initial Bering land bridge route, at substantially high frequency.  That is why the NWS is specific to the New World, in our hypothesis.

Bering Land Bridge, 21,000 years BP to present; Wikimedia Commons

The historic evidence from 19th century and earlier anthropological and traveler reports showed that the pandemic of diabetes and associated obesity was not the aboriginal state.  The modern disease traits had arisen in the last 50 years, so clearly seemed a response to modern living conditions, basically of post WWII origin, and applying more to those living a poorer existence compared to those of us in suburbia. Since Mexican Americans often worked as farm laborers, certainly not deprived of exercise, and because traits like gallstones appeared frequently in pubescent girls (almost unheard of in the 'Anglo' or African American populations), the syndrome seemed to be a kind of response to something dietary that  had changed--we couldn't identify it, despite some guesses.

In any case, it seemed genetic in the response-to-environment sense.  And if there were an evolutionary antecedent it might have been for some genotype that was good at storing food energy as body fat, during the settlement of the Americas in the cold and relatively barren Siberia-to-Alaska passage.

We published our hypothesis in various places at the time.  The papers were seen by a lot of people, but we moved at that time from our positions in Texas, up north and far enough away people doing the work that it was hard to keep on.  So we stopped working seriously on the problem.  There was interest from colleagues in various places in Texas and the Southwest but for various reasons we never could get a proper, large study going, one in which the syndrome itself was studied rather than the traits independently of each other, and with a clear admixture-based approach.

As a result, nothing has been found to date.  But perhaps a new report will provide that lead, even if the authors are unaware of the previous work and our long-standing hypothesis.   The paper, published online in Nature by a consortium of many investigators on December 25, claims that variants in the SLC16A11 gene may be responsible for adult diabetes in Mexico.  Ironically, if not perhaps even suspiciously, this is based on fossil data!  Could such a stretch be the answer to the NWS?

The new paper and SLC16A11
The huge SIGMA consortium responsible for this study examined various genotypes that might account for the high prevalence of diabetes in Mexico.  After a lot of data description, the bottom line is that there is evidence of known genes' contributing to diabetes in the Mexican data set, but it doesn't account for all of it.  Searching various data bases led to a consideration of the available Neanderthal sequences.  One east-Eurasian (Denisovan) specimen is homozygous for the candidate variant in SLC16A11, and it has high frequency in the Mexican study population and in some east Asians.

While the current study only tested this for diabetes risk, which was statistically substantial, they did not test other aspects of the hypothesis, not just that the susceptibility is of Native American origin, but that it involves other traits than diabetes.  Of course some diabetes in Mexico will be more or less strictly environmental and much due to European-derived alleles, because a large segment of the Mexican population is a European-Amerindian admixture.

Right now there are too many unanswered questions to be able to tell from this study whether the NWS idea is correct, or whether this newly discovered risk allele is responsible.  But the necessary data are available if investigators care to look at it in this way.  (We have contacted one of them to see if there would be interest).

If the story turns out to be true, it will not only have potential major health importance for a huge multi-continental population (because, if genetically rather simple, perhaps targeted therapy could be developed), but will tie evolutionary and biomedical genomics in a nice package.  We are eager to see.


  1. "One east-Eurasian (Denisovan) specimen is homozygous for the candidate variant in SLC16A11"

    Nature and other journals seem to be on a roll on Denisovans, but I guess you know how much material has been found on them so far - one finger bone and two teeth. Dan Graur is cautious about many of the findings related to 'ancient DNA'.

    1. I don't know what Dan's issues are, but I take his views seriously. In this case, the inference is entirely testable, in many direct and indirect ways. It could quickly be shown to be entirely false or, more likely, far from definitive.

      But this is a major health problem, a pandemic one, so it's worth checking, I think. I've written to one of the people in the study to see if there might be interest in that.

    2. The Denisovan connection is mostly a gee-whiz factor here, I think. If a particular variant of SLC16A11 is at high frequency in Mexican Americans it doesn't matter where it came from, if it in fact is associated with this pandemic.

    3. Yes, I wrote that in a follow-up comment that the blogger ate up :(