|"Beneath this mask there is an idea, Mr. Creedy. |
And ideas are bulletproof"
First off we have the paper from Nardone and colleagues  (open-access) which, following some [methylomic] analysis of brain tissues from individuals who had an autism spectrum disorder (ASD) concluded: "a possible role for epigenetic processes in the etiology of ASD". The second paper by Basil and colleagues  (open-access) suggested that: "exposure to inflammation during prenatal life is associated with epigenetic changes". I'll go through each paper separately and then perhaps highlight a few overlapping areas...
So, the Nardone paper:
- Following on from other reports detailing the analysis of methylation patterns in certain brain areas in samples from people with autism (see here), Nardone et al looked at "genome-wide DNA methylation patterns in autism brain and the significance of DNA methylation dysregulation in developing the disorder". DNA methylation by the way, is concerned with chemically altering gene function and has been explained in previous posts (see here for example).
- Brain tissues were the source material for analysis, obtained from a few tissue banks including from the UK Brain Bank for Autism for 13 people with autism and 12 non-autistic controls. The authors don't actually provide much information about the previous background of those donors aside from things like age, sex and post-mortem interval alongside a note that "clinical information is available upon request from the Autism Tissue Program". Also that autism samples were from those with "an ADI-R confirmed diagnosis of autism".
- Two brain areas, Brodmann Area 10 (BA10) and Brodmann Area 24 (BA24) were analysed for DNA methylation profiles.
- Results: "both brain regions have a profoundly distinctive epigenetic signature in the autistic brain" compared to asymptomatic controls. Further, the authors talk about "significant enrichment for genomic areas responsible for immune functions" in amongst hypomethylated methylation sites (CpGs) alongside enriched hypermethylation in CpG sites in genes "related to synaptic membrane". Hypermethylation (over-methylation?) is normally taken to mean gene silencing.
- A few further analyses, which are far and above my level of competence in this area, also suggested that "the autistic brain displays less region-specific identity, at the level of DNA methylation" and "a dysfunction in the developmental program that leads to less epigenetic distinction between cortical regions in the autistic compared with the control brain".
- Going back to the findings implicating immune system genes in the autism samples, when authors undertook some GO [gene ontology] enrichment analysis based on the genes which were differentially methylated in brain regions, they noted that 'immune response' was "the GO category that was most significantly represented in both transcriptome and our methylome analysis". This included processes such as leukocyte migration, cytokine-mediated signalling pathway(s) and "regulation of inflammatory response to antigens". Leukocyte migration and cytokine signalling... mmm.
- The authors conclude: "Our expression data strongly indicates the presence of an altered immune response in the autistic brain that correlates well with epigenetic modulation of genomic regions relevant to immune functions". Neuroinflammatory processes are also mentioned.
Then to the Basil paper:
- Before heading into some of the details, I note a familiar name on the authorship list of this paper - Chloe Wong - who first-authored that very interesting paper on the methylome and autism  covered in a previous post.
- Authors aimed to test the hypothesis "that prenatal exposure to MIA [maternal immune activation] in the mouse results in global methylation differences in the brain and specifically alters DNA methylation in the promotor of Mecp2". MIA, by the way, is something which has been talked about with reference to autism (see here), in particular through the work of the late Paul Patterson. Mecp2 has been most prominently discussed with reference to a condition called Rett syndrome.
- So, the brains of offspring of mice artificially stimulated to reproduce the MIA model were examined. This examination included looking at global methylation differences (compared with saline exposed control mice) using LINE1 methylation differences as "a proxy" across the groups. LINE1 (long interspersed element 1) fall into the category of retrotransposons. Not a million miles away from those HERVs which I keep banging on about (see here), they cover about 17% of the genome and according to some authors continue "to diversify human genomes" .
- Results: "We found significant DNA hypomethylation in the hypothalamus and a similar trend in the striatum of offspring exposed to MIA compared with saline-exposed controls". This hypomethylation also stretched to the "Mecp2 gene promotor region" (at least in the hypothalamus). This effect was especially prominent in female mouse offspring. Ergo, early life exposure to inflammation may be associated with epigenetic changes in certain brain areas.
Bearing in mind the Basil results were based on mice and the Nardone findings based on the examination of post-mortem neuronal tissue samples from people, I'm sure that you can perhaps see where there may be common threads between these studies. I should again point out that there isn't a complete clinical picture for all deceased persons included in the Nardone data, so one has to be a little cautious about extrapolating too much to just a diagnosis of autism. Supplementary table 9 (yes, I did go through the data) for example, lists 'seizure' and 'no seizure' alongside each participant which seems to be present for some of the data for participants with autism but none of the controls. If this is taken to mean that some participants with autism also had epilepsy or seizure-type disorders comorbid(?), straight away you've got a disparity in sample matching and another potential explanation for the some of the findings.
That being said, the interest in gene expression in autism being affected by epigenetic mechanisms does seem to be picking up some pace in recent times and these papers certainly add something to that growing body of research. The focus on immune function and inflammation (that word again) fits in with other studies on autism, including those looking at more traditional genomic issues turning up immune function as potentially being related to symptoms (see here). Without generalising to every person with autism, or should that be the autisms, I'd be minded to say that immune function (in its many guises) and autism is beginning a new research ascendancy? Now, about going that one stage further...
Music to close and George Ezra and Budapest.
 Nardone S. et al. DNA methylation analysis of the autistic brain reveals multiple dysregulated biological pathways. Transl Psychiatry. 2014; 4: e433.
 Basil P. et al. Prenatal maternal immune activation causes epigenetic differences in adolescent mouse brain. Transl Psychiatry. 2014; 4: e434.
 Wong CC. et al. Methylomic analysis of monozygotic twins discordant for autism spectrum disorder and related behavioural traits. Mol Psychiatry. 2014 Apr;19(4):495-503.
 Beck CR. et al. LINE-1 Elements in Structural Variation and Disease. Annu Rev Genomics Hum Genet. 2011; 12: 187–215.
Nardone, S., Sharan Sams, D., Reuveni, E., Getselter, D., Oron, O., Karpuj, M., & Elliott, E. (2014). DNA methylation analysis of the autistic brain reveals multiple dysregulated biological pathways Translational Psychiatry, 4 (9) DOI: 10.1038/tp.2014.70
Basil, P., Li, Q., Dempster, E., Mill, J., Sham, P., Wong, C., & McAlonan, G. (2014). Prenatal maternal immune activation causes epigenetic differences in adolescent mouse brain Translational Psychiatry, 4 (9) DOI: 10.1038/tp.2014.80