This guy has an interesting theory

http://content.nejm.org/cgi/content/full...?query=TOC

A Hot Spot of Genetic Instability in Autism
Evan E. Eichler, Ph.D., and Andrew W. Zimmerman, M.D.

Sixty-five years after Leo Kanner first described autism,1 we are beginning to move beyond description of this clinically heterogeneous neurobehavioral syndrome toward a deeper understanding of its biologic complexity. In areas such as genetics and neuroscience, researchers have joined the search for objective measures to elucidate autism's pathogenesis.

It has become clear that the solutions to autism will be neither simple nor uniform among patients with various autistic syndromes. At least 60 different genetic, metabolic, and neurologic disorders have been associated with autism and involve approximately 10% of patients, whose clinical presentations frequently vary, even among those with known disorders.2 For example, some (but not all) children with Rett's syndrome, the fragile X syndrome, Down's syndrome, fetal valproate embryopathy, or congenital rubella may also present with autism.

Each new objective finding expands the number of forms, or "autisms," like layers of an onion. This expansion occurred with the discovery of mutations in MECP2 that cause Rett's syndrome.3 After this discovery, the disorder was recognized as a singular entity with varied manifestations, rather than a form of autism. Likewise, boys with autism and the fragile X syndrome have been characterized as a subgroup of the latter disorder. Another promising approach has been to characterize the dysmorphic phenotypes found in 20% of children with autism, which will probably lead to the discovery of other genetic disorders.4

In this issue of the Journal, Weiss et al.5 describe deletions and duplications at chromosome 16p11.2 that appear to be associated with approximately 1% of unexplained, idiopathic, and nonsyndromic autism. The phenotypic patterns and severity of autism varied among their 13 case subjects with deletions and 11 subjects with duplications and overlapped with findings of delays in motor and cognitive development and dysmorphic features. In the study subjects, there was a notable absence of developmental regression, which typically affects 40% of children with late-onset autism, usually between 14 and 24 months of age,6 indicating that the 16p11.2 deletion or duplication events are mostly associated with autism of early onset.

The study by Weiss et al. supports the general notion that large, spontaneous deletions and duplications contribute to the molecular causes of autism.7,8 That said, the only credible and novel association reported by Weiss et al. was that between autism and the 16p11.2 locus. Indeed, the authors report that they observed no additional de novo events in multiple unrelated cases, other than duplications at 15q13 (the region disrupted in the Prader–Willi and Angelman syndromes), which was described previously.9,10 Last year, Sebat and colleagues8 described a similar 16p11.2 deletion, along with many other candidate loci, on the basis of the screening of copy-number variants associated with autism.

What might explain the different results arising from these two screenings? Differences in the genotyping platforms, analytical methods, and study design seem to be likely contenders. However, Weiss et al. had a different burden of proof: they required that their mutation events be observed in multiple unrelated samples and be confirmed in replication studies.

The genomic region identified by Weiss et al. corresponds to 1 of approximately 150 regions of the human genome that are predicted to be "hot spots" for recurrent deletion and duplication.11,12 The presence of large, highly similar duplications flanking the 16p11.2 region predisposes this particular portion of the chromosome to unequal crossing over during meiosis (Figure 1). Consequently, although the parental DNA is normal, the unique sequence between these duplicated sequences becomes microduplicated or microdeleted in offspring.13 The critical genomic segment described by Weiss et al. seems to be identical to a previously described de novo microdeletion in male monozygotic twins with mild mental retardation, aortic-valve abnormalities, and seizure disorder; no evidence of autism spectrum disorder was presented.14 As in other diseases associated with genomic disorders (e.g., the velocardiofacial syndrome and schizophrenia), it is likely that the effect of the 16p11.2 deletion or duplication extends beyond autism and that variability in clinical manifestations depends on differences in genetic background. This theory is consistent with an observation made by Weiss et al. in two families: affected children inherited the 16p11.2 duplication from unaffected parents.


Figure 1. A Hot Spot of Genomic Instability Associated with Autism.
Interspersed duplication blocks (12 and 13) on 16p11.2 promote unequal crossing over during meiosis (two of four chromosomes are shown). Gametes are produced that either lack or carry a double dose of the critical interval. Dosage-sensitive differences of genes in the critical interval (A, B, C) probably increase the susceptibility to disease. There are more than 25 genes or transcripts in the critical interval (e.g., DOC2A, QPRT, and TBX6), as well as rapidly evolving genes in the flanking duplications.



The short arm of chromosome 16 is exceptional from an evolutionary perspective because it is populated by an excess of duplicated segments that emerged relatively recently during evolution (less than 15 million years ago).15 More than 16 blocks of segmental duplication are interspersed across the chromosome, and rearrangements among these blocks have been associated with various genomic disorders involving mental retardation, multiple congenital abnormalities, and autism. It is interesting that most of the duplicated sequences on chromosome 16 also carry copies of one of the most rapidly evolving gene families in the human species.16 Both the gene family and the genomic architecture are specific to apes and humans, which is consistent with other reports17 that from an evolutionary standpoint, autism may be a relatively "young" disease.

Will other hot spots of recurrent deletion and duplication be identified that are associated with autism? The fact that deletion and duplication at 16p11.2 and duplication at 15q11.2 together account for approximately 2 to 3% of cases of autism warrants further study of spontaneous copy-number variation as a cause of this disease. More important, these examples highlight a different paradigm for the genetic basis of autism. Rather than being an inherited disease, autism may be the result of many independent loci that rarely delete or duplicate during gamete production. Collectively, such de novo events might contribute significantly to the disease and explain why few genetic loci have been confirmed with the use of traditional linkage-based approaches. A key factor in the search for additional large, highly penetrant structural changes will be the frequency of the new mutation event.13 Regions such as 16p11.2 and 15q11.2 may be the tip of an iceberg, discovered first because the frequency of their de novo mutations is much higher than that of other autism-associated regions of de novo copy-number variation. It is possible that after much larger case–control groups have undergone genotyping, some of the 50 other large de novo events observed by Weiss et al. and other events described in recent studies will turn out to be specifically associated with autism. The discovery of significant associations for the rarer loci may require the screening of tens of thousands of DNA samples from patients rather than a few thousand samples. Deeper sample collection and new cost-effective genomic techniques may be needed to peel away the remaining layers of the onion.


Dr. Eichler reports receiving lecture fees from Applied Biosystems. No other potential conflict of interest relevant to this article was reported.


Source Information

From the Howard Hughes Medical Institute and the Department of Genome Sciences, University of Washington, Seattle (E.E.E.); and the Kennedy Krieger Institute and Johns Hopkins University, Baltimore (A.W.Z.).

This article (10.1056/NEJMe0708756) was published at http://www.nejm.org on January 9, 2008. It will appear in the February 14 issue of the Journal.

References


Kanner L. Autistic disturbances of affective contact. Nerv Child 1943;2:217-250.
Coleman M. The neurology of autism. New York: Oxford University Press, 2005:3-39.
Amir RE, Van den Veyver IB, Wan M, Tran CQ, Francke U, Zoghbi HY. Rett syndrome is caused by mutations in X-linked MECP2, encoding methyl-CpG-binding protein 2. Nat Genet 1999;23:185-188. [CrossRef][ISI][Medline]
Miles JH, Takahashi TN, Bagby S, et al. Essential versus complex autism: definition of fundamental prognostic subtypes. Am J Med Genet A 2005;135:171-180. [Medline]
Weiss LA, Shen Y, Korn JM, et al. Association between microdeletion and microduplication at 16p11.2 and autism. N Engl J Med 2008;358. DOI: 10.1056/NEJMoa075974.
Landa RJ, Holman KC, Garrett-Mayer E. Social and communication development in toddlers with early and later diagnosis of autism spectrum disorders. Arch Gen Psychiatry 2007;64:853-864. [Free Full Text]
Szatmari P, Paterson AD, Zwaigenbaum L, et al. Mapping autism risk loci using genetic linkage and chromosomal rearrangements. Nat Genet 2007;39:319-328. [CrossRef][ISI][Medline]
Sebat J, Lakshmi B, Malhotra D, et al. Strong association of de novo copy number mutations with autism. Science 2007;316:445-449. [Free Full Text]
Baker P, Piven J, Schwartz S, Patil S. Duplication of chromosome 15q11-13 in two individuals with autistic disorder. J Autism Dev Disord 1994;24:529-535. [CrossRef][ISI][Medline]
Cook EH Jr, Lindgren V, Leventhal BL, et al. Autism or atypical autism in maternally but not paternally derived proximal 15q duplication. Am J Hum Genet 1997;60:928-934. [ISI][Medline]
Bailey JA, Gu Z, Clark RA, et al. Recent segmental duplications in the human genome. Science 2002;297:1003-1007. [Free Full Text]
Sharp AJ, Hansen S, Selzer RR, et al. Discovery of previously unidentified genomic disorders from the duplication architecture of the human genome. Nat Genet 2006;38:1038-1042. [CrossRef][ISI][Medline]
Lupski JR. Genomic rearrangements and sporadic disease. Nat Genet 2007;39:S43-S47. [CrossRef][Medline]
Ghebranious N, Giampietro PF, Wesbrook FP, Rezkalla SH. A novel microdeletion at 16p11.2 harbors candidate genes for aortic valve development, seizure disorder, and mild mental retardation. Am J Med Genet A 2007;143:1462-1471. [Medline]
Johnson ME, Program NCS, Cheng Z, et al. Recurrent duplication-driven transposition of DNA during hominoid evolution. Proc Natl Acad Sci U S A 2006;103:17626-17631. [Free Full Text]
Johnson ME, Viggiano L, Bailey JA, et al. Positive selection of a gene family during the emergence of humans and African apes. Nature 2001;413:514-519. [CrossRef][Medline]
Parker-Katiraee L, Carson AR, Yamada T, et al. Identification of the imprinted KLF14 transcription factor undergoing human-specific accelerated evolution. PLoS Genet 2007;3:e65-e65. [CrossRef][Medline]

Transposons. McClintock believed these "jumping genes" respond to some type of stress that a cell's existing structure could not manage. It has been noted that these "jumps" tend to target certain genes more than others; hence, a pattern of mutation ... and one that is heritable.  Not only humans, but all of life had been doing this for as long as life has existed.

Mutations occur over time in every species, some advantageous and some not.  The fact that there are so many of us, and that AutSpec is highly heritable, leads me to think there must be some advantages to our neurotype.

In any event--doesn't mutation occur from events such as deletions/duplication in the chromosome, anyhow?  I wish I was more knowledgeable about this.

Is this considered a mutation, or an "error" ?

Okay, here's an attempt to address the variation/mutation/error issue.

Variations in DNA sequence (genome) fall into several classes.  
(Note:  a "nucleotide" or a "base" is a single building block in the DNA polymer.  A chromosome is made up of millions of DNA building blocks.  Sequencing is the process of determining the order of the nucleotide building blocks in a chromosome.  The genome is the sum total of all the chromosomes. Each person has two copies of the genome, inherited from biological parents.  Any two copies of the genome vary with respect to each other by about 0.1%.  The genome can be conceptualized as a code--a set of instructions for building a cell and making it work right.  Variations in the code are like variations in the wording of the instructions--some of these variations result in screwed-up instructions.)

Imagine completely sequencing the genomes of 1000 people (that'd give 2000 genomes, since each person has two copies).  One could align the sequences to each other and look for where variability occurs.  Some variations will be common (e.g., 40% of the sequences say "A" and 60% say "T") and some will be extremely rare (e.g. 0.5% say "A" and 99.5% say "T").  The possible types of variations are:

Single nucleotide polymorphisms, called "SNP"s.
TTGCTTGAATAAAGGTGTGCCCTT
TTGTTTGAATAAAGCTGTGCCCTT
If you look at the pair of sequences above, you'll see two positions where they vary.
Some variations have functional effects on the code, but most are "neutral",  meaning no obvious effect.

Insertion-deletion (indel) polymorphisms
AAGTGTGGGCCGCAAATGTCAT
AAGTGTGG--CGCAAATGTCAT
In this pair of sequences, one of them is missing two bases compared to the other.
Indels can vary in size from one base to millions of bases.  Depending on what's in the DNA that is deleted, a person can have "fewer" copies of specific genes, and this can have functional effects. In the paper about chromosome 16 being discussed here, it was found that several hundred bases on chromosome 16 were missing from one copy of the chromosome in about 1% of the genomes of the people with autism that they looked at.

Copy number polymorphisms (CNP) (also called duplications, segmental duplications)
TTTACCACATTGCGCGGGAAATTGTGTCCCAA
TTTACCACATTACCACATTGCGCGGGAAATTGTGTCCAA
The sequence ACCACATT occurs once in the top sequence and twice in the bottom.
CNPs can involve from 2 bases up to millions of bases.  Depending on what's in the region of DNA that is copies, a person can have "extra" copies of specific genes, and this can have functional effects.

Rearrangements
AGAGG-TCAATTACATTA-GGCGGCACC
AGAGG-ATTACATTAACT-GGCGGCACC
In the two sequences above, there's a portion of the bottom string that's inverted relative to the top string.
You can have inversions of varying sizes, where, say, several million bases on a chromosome is flipped.  Or pieces of DNA can be moved to a different region or the chromosome or even to a different chromosome. Depending on where the break points for these types of rearrangement land, there can be functional effects (e.g. if the sequence of a key gene is disrupted by the rearrangement).

So okay--as I said above, there's different types of variation that naturally occurs in the sequences of chromosomes.  Because so few human genomes have been completely sequenced, it's not known really, what the extent of variation actually is.  Most of the attention so far has been directed to SNPs, because these are the easiest to detect and you don't have to do whole genome sequencing to find them.  As more sequencing is done (and it will be!), it's likely that all sorts of baroque chromosomal variations will turn up in people who appear to be in perfectly good health.  This is why I used the term "variation."  It doesn't imply anything about "good" or "bad", the  way that a term like mutation or error does.

On to mutation and error.  When cells divide, their genome is copied so that each cell gets one.  The enzymes that do the copying have evolved with error correction capabilities. Generally speaking, the copying process is extremely accurate.  But the occasional mistake is made.  Where there should be a "G", say, a "T" might get placed instead, thereby changing the code. This, then, would be an error or mutation, and it would propagate through each successive cell division. If the nucleotides in the genome being copied have been structurally damaged, say through UV radiation, then the probability of copying errors increases, and you'll get more mutations in the genome.

Most mutations are neutral--that is, they don't appear to have deleterious effects.  So over time, the human species has accumulated lots of them---they're called variations, as described above.  (The genome-based ancestry projects track the inheritance of specific variations.)

But some variations in the code's instructions do have deleterious effects. The structures and functional efficiencies of the proteins, the molecular machines that do the actual work of maintaining a cell can be altered.  Or the signals to say when a gene should be turned on or off can be altered. Or missing/extra copies of the gene can result in having the wrong amount of a protein for optimal functioning.

There can be multiple subtle effects of genome variation.  Trying to figure out the causal connections between sequence variations and complex diseases or conditions is an area of active research.  And it's slow-going.  I think that much more will be known in about 20 years because there will be a deluge of genome sequencing.  But the sequencing isn't enough.  One has to have good bases for any correlations between sequence variation and effect.  This will require different types of measurements and assessments.  And it'll require cooperation from millions of people to share their clinical results and histories.

Thanks for the explanation, energeia.  I believe I comprehend the original post at least a little better.

You're welcome, Batman.

err,bump, interesting energia