Wednesday, 6 March 2013

Epigenetics, EN-2 and the 'autism brain'

A paper by Jill James and colleagues* (open-access) caught my eye recently. Centred on Engrailed-2 (EN-2), a gene with more than a passing relationship to cases of autism (see here), James et al report results based on analysis of a small number of post-mortem cerebellar samples with a particular focus on an epigenetic evaluation.
Cerebellum @ Wikipedia  

What is epigenetics? Well, I've written before about some of the basic concepts involved (see here) and how despite not everyone being enamoured with the rise and rise of the science, the discipline of epigenomics adds quite a distinctive layer to the functioning of a persons genome.

The basic tenet: your DNA might not necessarily be your destiny and that subtle changes to the epigenome can influence the expression of certain genes or not. Certainly in areas such as cancer medicine, epigenetics is starting to make some real waves (see here).

With autism in mind, epigenetics is also starting to make an impact on the scientific literature and promises so much more. I'm taken for example, back to some previous work looking at prefrontal cortex neurons** with autism in mind which concluded that there might be more to see in this area at least for some cases of autism.


  • James and colleagues focused on cerebellar samples because (a) the cerebellum has been a real area of interest to autism research, and (b) EN-2 is "highly expressed in Purkinje cells"; reaffirming some interesting observations noted about Purkinje cells in the cerebellum of people with autism***.
  • They analysed 26 samples from 13 people with autism and 13 asymptomatic controls. Details of how participants died and other details are provided in the paper, bearing in mind the various discussions on how post-mortem brain samples from those deceased who had autism are subject to various confounders including how they died and the role of any comorbidity. Incidentally, some of the autism samples originated from the same place which had that very unfortunate freezer malfunction last year (see here).
  • Various methods and techniques were used to assess the details of epigenetic functions focused on methylation. I can't and won't pretend to understand all of them but interestingly as well as looking at EN-2 promoter region methylation, global methylation and "the methylation status of histones H3K27 (associated with gene silencing) and histone H3 lysine 4 (H3K4; associated with gene activation)" was also included (see here), part of the histone code.
  • Results: some interesting ones such as the finding of hypermethylation of DNA extracted from autism cerebellum samples, which contrasts sharply with the DNA hypomethylation of immune cells noted by some of the authorship group on another occasion****. The authors speculate that this could be indicative of "tissue-specific" DNA methylation in autism; also noting that short of looking at brain samples - which is neither desirable or feasible for the living - we can't conclude too much from "peripheral cell DNA methylation patterns". This should make for some interesting future discussions I reckon.
  • Alongside this global hypermethylation, James reports hypermethylation of the EN-2 promoter region. This, alongside sustained gene expression of EN-2 and greater levels of EN-2 protein in the autism samples. Similarly when looking at the methylation of histones (H3K27 and H3K4), the histone H3K27 which is linked to gene suppression was decreased and the histone H3K4 linked to gene activation, was increased (albeit not significantly).
  • Assuming that I've understood this all correctly, the suggestion is that epigenetic issues with the histones involved in gene suppression or gene activation (via methylation) were congruent with a pattern of "sustained EN-2" gene over-expression which might tie into the loss of Purkinje cells***** noted in the cerebellum of some people with autism. At least I think so.

It all makes for some really rather interesting findings. That for example, the modification of histones ties into the levels of gene expression and importantly gene protein levels is really exciting and perhaps a valuable addition to the notion that mutation in the form of SNPs are the only influencing variable on gene function. Indeed that an epigenetic process might affect the timing of gene activation/suppression at critical periods of development is also an important point bearing in mind that we don't all walk around with all our genes permanently stuck in the 'on' position.

One also starts to wonder about not just the availability of methyl groups in this process but also the functioning of things like the DNA methyltransferase enzyme family (adding methyl groups) and indeed the demethylase enzymes (removing methyl groups) and the circumstances of their control at certain periods of development. Indeed methylation is only one facet of histone modification, as per the acetlyation and deacteylation of histone which potentially brings us back to things like the valproate connection being made to cases of autism (see here). It's all quite complicated.

Perhaps just as important are the implications of hypermethylation and those histone modifications to other genes tied into things like neuronal development and immune function in conditions like autism. Noting for example the Saxena paper (covered here) and their linking quite a few of the autism-related genes to things like immune function, James and colleagues make mention of one demethylase, JMJD3 (see here) and its potential link to the "IL-6 gene promoter" with regards to processes such as neuroinflammation. Certainly one has to ponder how deep the rabbit hole goes.

OK, coming back down to earth, caution is required in that this was a relatively small scale study which is again, always going to be confounded by the use of post-mortem brain samples and factors such as cause of death and the important point that autism is a behavioural label and that link of possible heightened comorbidity. Added to the fact that the focus was on one particular gene - one of quite a few - with some apparent connection to autism, the results should be viewed as preliminary at best.

That being said, we have a template now for expanding this area of work to cover other candidate genes in different tissues, to start working on those all-important rodent models. Then, with some degree of caution and assuming a strong connection is made, looking at the various factors which might potentially influence and moderate such epigenetic issues - including sex differences****** (open-access) - bearing mind the golden concept of phenotypes.

This could be something quite big...

Speaking of big (famous), they don't come much bigger than this lady....


* James SJ. et al. Complex epigenetic regulation of Engrailed-2 (EN-2) homeobox gene in the autism cerebellum. Translational Psychiatry. 2013: 3; e232.

** Shulha HP. et al. Epigenetic signatures of autism: trimethylated H3K4 landscapes in prefrontal neurons. Arch Gen Psychiatry. 2012; 69: 314-324.

*** Fatemi SH. et al. Purkinje cell size is reduced in cerebellum of patients with autism. Cell Mol Neurobiol. 2002; 22: 171-175.

**** Melnyk S. et al. Metabolic imbalance associated with methylation dysregulation and oxidative damage in children with autism. J Autism Dev Disord. 2012; 42: 367-377.

***** Baader SL. et al. Ectopic overexpression of engrailed-2 in cerebellar Purkinje cells causes restricted cell loss and retarded external germinal layer development at lobule junctions. J Neurosci. 1998; 18: 1763-1773.

****** McCarthy MM. et al. The epigenetics of sex differences in the brain. J Neurosci. 2009; 29: 12815-12823.

---------- James SJ, Shpyleva S, Melnyk S, Pavliv O, & Pogribny IP (2013). Complex epigenetic regulation of Engrailed-2 (EN-2) homeobox gene in the autism cerebellum. Translational psychiatry, 3 PMID: 23423141