Recessive mutations in autism – the return of hidden metabolic disorders

My wrong guesses of 2012. Two weeks ago during a presentation, I had to admit that there is little evidence for a large contribution of recessive or compound heterozygous mutations in epileptic encephalopathies. At the beginning of 2012, I had initially suggested that recessive or compound heterozygous mutation of known neurometabolic disorders could be identified through exome sequencing in sporadic epileptic encephalopathies. However, as of 2013, there is little evidence for this in our data or the data from other consortia. Now, two papers in Cell suggest a significant contribution of recessive mutations in autism including a revival of the “hidden neurometabolic hypothesis”.

The papers we have been waiting for. The relevance of de novo mutations in autism has been well established through a series of high-ranking papers in early 2012. I felt however, that the field had been relatively quiet since. The analysis of de novo mutations raised more questions than answers and other genetic mechanisms including recessive inheritance, compound heterozygous mutations and X-linked inheritance seemed out of reach due to difficulties in distinguishing pathogenic from random genetic variation. Every individual is a “knockout” for 2-3 genes and differentiating signal from noise would require stringency and larger sample sizes. These studies have now been performed and provide interesting insight into the genetic architecture of autism. The paper by Lim and collaborators investigates the role of gene knockouts in patients with autism versus controls, whereas the paper by Yu and colleagues looks at autism genes in families.

Knockouts. The paper by Lim and collaborators investigates the role of homozygous and compound heterozygous loss-of-function mutations in ~ 1000 patients with autism and ~1000 controls including a larger follow-up cohort. They find an excess of loss-of-function mutations in patients with autism, resulting in a 2-fold increase in patients. In addition to the autosomal variants, they find a similar enrichment of hemizygous knockouts in males on the X-chromosome. In total, autosomal and X-chromosomal knockouts explain ~5% of the risk for autism spectrum disorder. I have mentioned earlier that every individual is a knockout for 2-3 genes. This frequency, however, includes many “common” variants that are present in more than 5% of the population. Lim and collaborators specifically looked at rare variants. “Knockouts”, i.e. either homozygous or compound heterozygous variants were much less frequent in controls. The frequency of these “rare knockouts” is 3% in controls, i.e. the majority of individuals do not carry such variants. This frequency is almost doubled in patients with autism. The implicated genes, however, are highly heterogeneous. The list of genes found to be “knocked out” only in patients only included a single known gene, USH2A, the gene for Usher Syndrome 2A. In addition, out of the 81 genes only found in cases, only 10 appeared more than once. Again, this points towards a complex genetic architecture, but the identified genes provide an interesting start.

An overview of the explained genetic risk factors in autism. Despite the fact that two recent studies add autosomal recessive mutations and compound heterozygous mutations to the list of candidates, the majority of the genetic risk still remains to be explained. On the left, genes with "complete knockouts" in more than one patient not present in controls are shown.

An overview of the explained genetic risk factors in autism. Despite the fact that two recent studies add autosomal recessive mutations and compound heterozygous mutations to the list of candidates, the majority of the genetic risk still remains to be explained. On the left, genes with “complete knockouts” in more than one patient not present in controls are shown.

Hidden neurometabolic disorders. Earlier this week, we discussed that metabolic screening in patients with intellectual disability and autism might be skipped in the diagnostic workflow. There is a very low frequency of “hidden metabolic disorders” masquerading as intellectual disability or autism and it may be discussed whether they exist at all.  However, the paper by Yu et al. shows the opposite. The authors identified the causative genes in consanguineous autism pedigrees through linkage, exome sequencing and gene panel analysis and discovered mutations in PAH (phenylketonuria) and AMT (nonketotic hyperglycinemia) in some of their pedigrees in addition to other genes implicated in known genetic syndromes. Some of these mutations were categorized as “hypomorphic alleles”, i.e. mutations that are not as damaging as a complete knockouts. This provides some evidence that at least in familial autism, the “hidden neurometabolic hypothesis” has some merit. However, it is yet a different question whether these diseases would be picked up through metabolic tests or only through genetic analysis. Nevertheless, the study by Yu and collaborators suggests that neurometabolic disorders may have an unanticipated phenotypic width that may well expand into neurodevelopmental disorders including autism.

Relevance for EuroEPINOMICS. The recessive mutations are back in 2013 and we need to look for them. Even though we had prematurely announced an increasing role of de novo mutations in epileptic encephalopathies, let’s keep in mind that autism genetics has always been a good guideline for what happens in epilepsy. It is provocative to assume that up to 5% of epileptic encephalopathies are due to homozygous mutations and that some of those might be potentially treatable. Furthermore, the same frequencies might occur in milder epilepsies such as IGE/GGE or Benign Rolandic Epilepsy. Both are hypotheses that can be tested in the existing EuroEPINOMICS cohorts. Well aware that I made the same prediction as in early 2012, my guess would be 5%.

15 thoughts on “Recessive mutations in autism – the return of hidden metabolic disorders

  1. Pingback: The return of TBC1D24 | Beyond the Ion Channel

  2. Pingback: Rare Epilepsy Syndromes and the Congenital Disorders of Glycosylation | Beyond the Ion Channel

  3. Pingback: Cold fusion – joining exome datasets to identify autism genes | Beyond the Ion Channel

  4. Pingback: The eye of the tiger, brain iron and the beta-propeller | Beyond the Ion Channel

  5. Pingback: ST3GAL3 and exome sequencing in autosomal recessive West Syndrome | Beyond the Ion Channel

  6. Pingback: EuroEPINOMICS-RES reloaded – reinventing a consortium | Beyond the Ion Channel

  7. Pingback: A metabolic disorder masquerading as adult-onset focal epilepsy | Beyond the Ion Channel

  8. Pingback: Thalamus, timing and TSC1 deletions | Beyond the Ion Channel

  9. Pingback: Mutation intolerance – why some genes withstand mutations and others don’t | Beyond the Ion Channel

  10. Pingback: Chutzpah, Zen and Fahrvernügen – the 2013 Sde Boker workshop | Beyond the Ion Channel

  11. Pingback: CACNA2D2, the ducky mouse, and what it takes to be an epilepsy gene | Beyond the Ion Channel

  12. Pingback: TBC1D24, DOORS Syndrome, and the unexpected heterogeneity of recessive epilepsies | Beyond the Ion Channel

  13. Pingback: Treatable causes of intellectual disability and epilepsy that you don’t want to miss | Beyond the Ion Channel

  14. Pingback: How metabolism alters epigenetics | Beyond the Ion Channel

  15. Pingback: Microcephaly, WDR62, and how to analyze recessive epilepsy families | Beyond the Ion Channel

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s