Autism is a hard disorder to nail down genetically- single nucleotide polymorphisms (SNPs) or even multiple locus genetic effects are not able to account for the large genetic component to the disorder. In recent times, Copy number variations (CNVs) has come to the forefront of Autism research , suggesting that microdeletions, duplications etc may account for some cases. Another new , till now unsuspected mechanism that has recently been implicated in autism is epigenetic mechanism of increased methylation in promoter regions that has the effect of silencing/reducing the expression of genes involved to a certain extent. The recent study by Gregory et al, is just such a step in the right direction, which will hopefully bring us closer to the truth.
In a nutshell, the study authors used CNV determining methods to discover that a deletion of OXTR (oxytocin receptor) gene was presnet in an autistic subject and was not de novo , but the deletion was inherited from his mother. One of the affected siblings of the autistic subject, who too was autistic, on the other hand did not have a deletion, but had increased methylation of the OXTR. this led the study authors to revisit their genomics data and look at adta across all autistic subjects and controls and discover that indeed, in other autistics too the OXTR had increased methylation. Then they looked for expression of OXTR in peripheral blood cells and temporal cortex and found that indedd in autistics, as compared to controls, there was reduced expression of OXTR. This strongly suggest that the epigentic changes that lead to autism (the efffect of OXTR suppression) happen quite early in the devlopmenet and might happen in utero.
Before I elaborate on my take home from the study, there are some excerpts (as I know you didn’t read the originals)
Classic autism comprises a spectrum of behavioral and cognitive disturbances of childhood development. The core autism phenotype includes deficits in social interaction, language development and patterns of repetitive behaviors and/or restricted interests. The population prevalence of the spectrum of autism disorders is estimated to range between 1/300  to 1/100 (http://www.nschdata.org/), with a male: female ratio of 4:1 [2,3]. The disorder has been shown to be highly heritable with the relative risk for siblings being approximately 2% to 8%, much higher than that of the general population . To date, only a small percentage of autism cases (<10%) have been ascribed to single gene disorders such as fragile X syndrome, tuberous sclerosis  and Rett syndrome . Numerous approaches including genetic linkage, genome-wide association, candidate gene association and gene expression analysis have been used to identify the additional genes
implicated in the development of autism [7,8]. However, the heterogeneous nature of autism and autism spectrum disorders has limited their success.
An additional approach to identify genes involved in autism is to characterize copy number variants (CNVs), that is, chromosomal deletions and duplications, that are known to be present within at least 5% of individuals with idiopathic autism . Autism CNVs have been shown to involve almost all chromosomes [10,11], with the most frequently observed alteration localizing to chromosome 15q11-13 [12-23]. A number of different methods have been used to characterize autism related CNVs, including but not limited to, cytogenetic Gbanding [14,23,24], metaphase fluorescence in situ hybridization (FISH) , Southern blotting , loss of heterozygosity (LOH) analysis [15-17,19], quantitative polymerase chain reaction (PCR)  and, more recently, genotyping and representational oligonucleotide microarray analysis (ROMA) .
Here we describe the use of genome-wide tilepath microarrays and array comparative genomic hybridization (CGH) to identify CNVs in a dataset of 119 unrelated probands from multiplex autism families . The genomic profiles of our autism dataset were compared to the array CGH profiles of 54 phenotypically normal individuals, to previously published CNVs present within the database of genomic variants  and to the Autism Chromosome Rearrangement Database (http://projects.tcag.ca/autism/). The most significant finding thus far from our analysis is a heterozygous deletion of the oxytocin receptor gene (OXTR) (MIM accession no.: 167055) in an individual with autism and his mother with putative obsessive-compulsive disorder (OCD). We further investigated the relationship between OXTR and autism by carrying out epigenetic analysis of the promoter region of OXTR. We show that the gene is hypermethylated in independent cohorts with autism as compared to controls, in both peripheral blood mononuclear cells (PBMCs) and the temporal cortex. Additionally, our analysis of expression levels of OXTR in the temporal cortex shows decreased levels of expression in individuals with autism as compared to controls matched for age and sex.
These data suggest that OXTR and the oxytocin signaling pathway play an important role in the etiology of autism and autism spectrum disorders and implicate epigenetic misregulation of OXTR in this complex disease.
Epigenetic control of autism susceptibility is a recent concept and most certainly a topic of great interest in the field. Over the past decade, researchers have uncovered suggestive links between epigenetics and autism, for example, autism is associated with duplications of 15q11-13 (especially maternally inherited), an imprinted region in the genome where DNA methylation status has been linked to Prader-Willi syndrome (PWS) and Angelman syndrome (AS) ; mutation within a gene that encodes a methyl-DNA-binding protein (MECP2, (MIM accession no.: 300005)) is the causative agent of Rett syndrome ; and mutation of this same gene has been associated with both autism and AS populations . Nagarajan et al. have shown that 79% of autism cases have a decrease in MECP2 expression in the frontal cortex and that an increase in aberrant DNA methylation correlates with this decrease in MECP2 expression . These data implicate epigenetic dysregulation as a mechanism for the development of autism and justify the examination of DNA methylation of autism candidate genes, such as OXTR identified in this study.
Now from the accompanying (more accessible) commentary:
The article by Gregory et al. published this month in BMC Medicine, reports on genomic and epigenetic alterations of OXTR, the gene encoding the receptor for oxytocin. The involvement of this gene was suggested by its deletion in an autistic patient. The subsequent analysis of a group of unrelated autistic subjects did not show an OXTR deletion, but rather hypermethylation of the gene promoter, with a reduced mRNA expression.
These findings address two major points of the current debate on the etiology and pathogenesis of autism: the role of oxytocin, known to be involved in modeling human behavior, and the possible involvement of epigenetic mechanisms. The nature of this epigenetic dysregulation is unknown but, if proved to be true, might explain the failure to identify sequence alterations in a host of candidate genes. Practical implications of these findings may be forthcoming, however not before extension and validation on a larger scale have confirmed their value.
The second issue raised by Gregory et al. deals with the epigenetic inhibition of OXTR expression in ASD. Such epigenetic modification, at least as reported so far, does not seem to be sequence based but rather of a different, as yet unknown nature. This might explain why researchers have been looking for decades for genetic mutations in ASD and yet have found almost none. An epigenetic mechanism would justify the ‘unusual’, non-Mendelian familial aggregations of ASD. In this respect, even the family with OXTR deficiency reported by Gregory et al. shows an unusual genotype-phenotype correlation, in that the same phenotype is caused by alterations of the same gene but due to different molecular defects (deletion versus hypermethylation).
Also, the possibility that in most ASD patients there might be an epigenomic instability is of interest in consideration of the fact that it has been shown that the epigenetic status in early fetal development can be reprogrammed by maternal behavior in a reversible way . Therefore, other environmental factors, yet to be discovered, might also be able to reprogram the epigenotype of the embryo.
I hope the fact that epigenetic changes may happen during pregnancy line of reasoning does not lead to the harmful and without-any-basis vaccination is cause of autism arguments. On the other hand I had covered earlier how Autism is more likely if mother was exposed to valproate during pregnancy or the child soon after birth. What if valproate is instrumental in an epigenetic fashion in leading to more or less methylation and gene expression. It is well known that valproate and valporic acid is given as treatment for psychosis/bipolar. In a similar vein, I am inclined to stick my neck out and claim that in schizophrenics/psychotics , the OXTR should be more expressed : perhaps more methylation, duplications etc . However I am checked in my musings by these studies that claim that negative symptoms of schizophrneia may be associated with reduced oxytocin activity in the brain. Yet, all said and done I would like to see a study that analyzes for epigenetic mechanisms in schizophrneia especially at the OXTR locus. Although the negative symptoms like social withdrawal of schizophrenia may lead to the opposite hypothesis regarding schizophrenia and oxytocin correlation, I am inclined to believe that schizophrenics (at least those suffering from positive symptoms predominantly) are too much oxytocin guided , trusting and socially too much involved in others type of people.
Gregory, S., Connelly, J., Towers, A., Johnson, J., Biscocho, D., Markunas, C., Lintas, C., Abramson, R., Wright, H., Ellis, P., Langford, C., Worley, G., Delong, G., Murphy, S., Cuccaro, M., Persico, A., & Pericak-Vance, M. (2009). Genomic and epigenetic evidence for oxytocin receptor deficiency in autism BMC Medicine, 7 (1) DOI: 10.1186/1741-7015-7-62
Gurrieri, F., & Neri, G. (2009). Defective oxytocin function: a clue to understanding the cause of autism? BMC Medicine, 7 (1) DOI: 10.1186/1741-7015-7-63
Hat tip to @Boraz for tweeting about this study.