Autism Genome Project

Hi Connie;

Several months ago the Autism Genome Project Consortium reported on two findings from their genome wide scan involving app 3,000 persons from multiple incidence families. The findings were massively hyped by the authors in the media as 'breakthroughs'. Reading some of what I could find without access to the total report, apparently the findings were located in chromosome 11p12-p13 and a deletion of a neurexin gene.

http://www.nature.com/ng/journal/v39/n3/abs/ng1985.html

Apparently the neurexin gene was in found in exactly two cases (out of some 3,000) and the cases were from the same family and they were sisters. The chromosome 11p12-p13 was a 'hot spot that contains some 160 genes and the findings were merely suggestive.

Unusual chromosome arrangements in autism have been reported for decades usually involving a handful of cases in each case and they are almost exclusively found in mental retardation syndromes with those afflicted carrying a second diagnosis within the PDD's.

There are other single gene disorders (Tuberous Sclerosis, PKU, Fragile 'X', Rhett Syndrome) that are mental retardation syndromes with a subset of those afflicted meeting criteria for a diagnosis of autism secondary to the mental retardation.

My question , which one of your scientific advisors might be able to answer, is does that report segregate the genetic findings by IQ? If not how can anyone segregate the confounding variable of mental retardation from autism and it is possible that these genome wide scans may identify a previously unrecognized mental retardation syndrome rather than an autism susceptibility gene.

Hi RAJ!

I have the article you refer to in hand, but realize that I am going to need some help answering your question, as I am not a geneticist and these topics get very complex, very quickly.

I'll investigate and get back to you!

Connie (IAN Staff) wrote:

Hi RAJ!

I have the article you refer to in hand, but realize that I am going to need some help answering your question, as I am not a geneticist and these topics get very complex, very quickly. I'll investigate and get back to you!

Thanks, all I am looking for is a simple yes or no. As parents, we are entitled to get the basic information of what was actually found in these studies without the inevitable conjecturing and extrapolating a finding to explain all of autism.

When this study was published the Consortium's PR machine went into a full court press with appearances on Good Morning America, CNN and the evening TV newscasts with the inevitable breathless 'Autism Gene Discovered!

Another study published a few months later also reported variants in the same two chromosomes ( 2 and 11) reported in the consortium's study published in Nature. The authors of that study didn't hold worldwide press conferences in fact the study was barely noticed anywhere.

That paper reported genetic variations on chromosome 2 and chromosome 11 were identified as susceptibility genes for adverse vaccine reactions.

http://news.biocompare.com/newsstory.asp?id=186110

Hi RAJ. Looking over your post again, I realize you were not looking at the complete article when you framed your question. You asked: "Does that report segregate the genetic findings by IQ?" and requested a simple yes or no answer.

Looking carefully at the article, I would say the answer is "no". I do not see that they conducted a specific analysis taking into account IQ level, although they did look at "broad" vs. "narrow" diagnoses when analyzing families' genetic data.

The families placed in the "narrow diagnosis" category had to have two or more affected individuals who met the criteria for autism on both the ADI-R and the ADOS. The families placed in the "broad diagnosis" category had to have at least one individual meeting the ADI-R criteria for autism and the ADOS criteria for autism or ASD. At least one other family member had to meet criteria for impairment on the social or communication domains of the ADI-R and meet criteria for at least ASD on the ADOS. (One could guess, then, that you might get less mental retardation in the broader category which could include, for example, individuals with Asperger's or PDD-NOS.)

What they were doing was using a new, large genetic database, and several new approaches, to try to identify which genes may be linked to ASD. Because there is so much heterogeneity (that is, so much variation) in people with various types of ASD, they were trying techniques that might remove some of the confusion from genetic data. They weren't claiming a major finding, as in "We've found the gene that causes autism!" They were mostly reporting, in great detail, how they collected genetic data, and how they analyzed it in several different ways to perhaps narrow the search.

One technique they tried was looking at the data with and without families who had individuals with "copy number variations" (CNVs). (The Wellcome Trust explained, for us non-geneticists, the new understanding that not just single letters, but sentences, paragraphs and even whole pages of our genetic code, can be missing or duplicated, and that's what "CNVs" are. See: http://www.wellcome.ac.uk/doc_WTX034711.html

Another technique was to look at whether a family was only "female-containing" (because those are thought to be more severe than families containing only males with ASD, and may provide a sharper picture, genetics-wise). The two sisters you mentioned represented one of the families with a "copy number variation" --their case was interesting, and discussed because of the link between the gene that was affected and how it interacts with "neuroligins, for which rare mutations apparently generate risk for ASDs and mental retardation."

That family, however, was not part of the main linkage analysis because the strongest linkage --which gave a graph with spikes for a certain suspect chromosomal area (11p12-p13)-- was the one in which they looked at only female-containing families and removed anybody with a known CNV. They said:

"We believe these explorations motivate thorough fine-mapping of the 11p12-p13 region. Modest peaks for linkage have been observed previously for this region, but 11p12-p13 has not been a major focus for discovery of autism risk loci." (pg 324-325 of the article)

In effect, they are winnowing down the genetic haystack to try to get closer to finding the needle they hypothesize is buried somewhere inside.

Thanks for wading through the complexity of the report. I have my answer, no, the study did not control for the confounding variable of mental retardation.

One of the disappointing findings from the study is that they were not able to replicate the consistent findings of regions of chromosome 7 as a candidate gene for autism.

There is now troubling evidence that many of the authors of this study may have an inherent bias on funding in this area since many of principal investigators have in fact taken out patents on various 'autism susceptibility' genes.

Joseph Buxbaum, one of the principle investigators is listed as the 'inventor' of an autism susceptibility gene:

http://patents1.ic.gc.ca/details?patent_number=2548590

I never realized that anyone could actually patent a gene but apparently that is the fact. I guess there is nothing wrong for them to take out patents since they should be rewarded for drug treatments that could be developed if the suspect gene eventually turns out to have a risk for autism, no matter how small the risk, but it certainly suggests an inherent bias.

Thanks again for providing me with the answer I requested.