Post by Former NIMH Director Thomas Insel: Autism Spring
Looking back over NIMH related events these past few months, one might wonder if this has been Autism Spring. It has certainly been a busy season for autism spectrum disorder (ASD): a White House meeting, unprecedented press coverage, and the largest International Meeting for Autism Research (IMFAR) to date. But perhaps most exciting has been the early scientific harvest evident in a series of high-profile papers published over the past two months. Some of these discoveries with autism have implications for mental disorders like schizophrenia and mood disorders, which increasingly are being addressed as neurodevelopmental disorders.
While the new findings range from epidemiology to new diagnostic tests, here I will focus on new insights into the molecular basis of autism. Three studies, written up in the June issue of Neuron, based their findings on data from the Simons Simplex Collection.1,2,3 The Simons Foundation funded the collection of careful clinical descriptions and DNA from over 1,000 children with autism who were the only affected member of their families; such families are referred to as simplex, in contrast to multiplex families, which have multiple affected members. The new reports look specifically at copy number variations (CNVs). These are rare, structural changes in the genome leading to a deletion or duplication of a segment of DNA. Many of these are de novo, meaning that the duplication or deletion is not found in the genome of either parent but develop in the DNA of germ cells (egg or sperm) over the life of one parent. Small de novo changes in DNA sequence, which occur in all of us, demonstrate that effects can be genetic without necessarily being inherited. And, of course, these germ cell changes may be the result of environmental factors, increasing with parental age.
The results are both intriguing and frustrating. Intriguing, in that children with ASD were found to have many more CNVs. These CNVs were more likely to be larger and more frequently involved specific genes than those found in unaffected siblings. But only 1 in 38 affected children had a recurrent CNV, meaning a CNV that appeared in any of the other children in the study. Of these recurrent CNVs, six genomic regions were discovered to be associated with ASD, including four duplications of the chromosome 7q11.23 region. This region is deleted in Williams-Buren syndrome, a disorder with hyper-social, hyper-verbal behavior that, in some ways, appears as the inverse of the autism phenotype. Unfortunately, the papers did not describe whether these recurrent CNVs were associated with distinct clinical characteristics.
The frustration comes from the relative rarity and complexity of these de novo CNVs. In two separate studies of this same sample using different techniques, only about 8 percent of ASD children in simplex families had CNVs. Add this number to the 8 percent with a mutation known to cause autism, such as Fragile X or tuberous sclerosis, and that still leaves more than 80 percent of ASD children with no evident genomic cause for their disorder. Traditional estimates of the heritability of ASD range as high as 90 percent. It is quite possible that these heritability estimates were too high, but even if the heritability were less, as Scharff and Zoghbi noted in an essay that accompanied these three Neuron papers, “the results are humbling.”4
Of course, there are more genomic risk factors to be found, given that the CNVs identified in these studies are large (100,000 bases or greater). As the technology for genomic research improves, smaller CNVs are likely to be identified. These papers estimate that there may be 200-300 CNVs in the genome contributing to ASD. The next step in this journey will involve sequencing all the coding regions of the genome, no doubt with even more variations emerging.
But there are more questions than answers in these projections. Many CNVs are incidental (2 percent occur in unaffected siblings), many different genes may be affected by the known CNVs, and the biological significance of any of these mutations remains to be determined. Indeed, an independent paper looking at the interaction of proteins from genes implicated in ASD found networks centered on two synaptic proteins: Shank 3 and PSD95.5 A separate analysis of the genes thought to be implicated in ASD identified a network that included genes involved in synapse formation, axon targeting, and neuronal motility.6 All of this suggests that, from a genomic perspective, autism is a synaptic disease.
Why is this important? If nothing else, these humbling results beg for more exploration of the brain. And one of the most exciting studies , just out, reports that RNA expression patterns in post-mortem brains yield some surprising clues. RNA is the key intermediary for translating DNA into protein. Patterns of RNA expression define which proteins will be expressed, determining the function of each cell. In ASD brains, the expected differences in expression between different regions of the brain are less distinct, as if mature cortical patterns have not developed. While the differences in expression are complex, they converge around a few key pathways and may reflect differences in RNA splicing. For instance, with brain maturation, genes are spliced in different regions to yield different fragments of RNA and different protein products. We need much more study of this process in ASD, but this initial project of frontal and temporal cortex suggests that whatever the DNA variations in ASD may be, the RNA fragments are strikingly abnormal.
If the CNVs discovered in genomic DNA reflect a fundamental genomic instability in ASD, could there be somatic mutations (mutations found in neurons but not in blood cells) in ASD brains? Perhaps the biology of cancer, with mutations in oncogenes and tumor suppressor genes found only in the tumor, will be a useful model for the biology of ASD, with mutations found only in the cortex.
Oliver Wendell Holmes once said, “I wouldn’t give a fig for the simplicity on this side of complexity; I’d give my right arm for the simplicity on the far side of complexity.” We are, unfortunately, not near the far side of complexity of autism. These recent studies raise questions about the limits of genetics, even with the enormous power of our current techniques. Genetic signals will be complex and may not converge as we would hope around a simple developmental mechanism or pathway. Post-mortem brain analysis may be highly informative, but we have little tissue from ASD children, and comparisons with age-matched tissue continue to be a challenge.
The great uncharted territory of environmental factors remains, which might begin to explain the infrequent mutation rate and apparent increase in autism prevalence. Here, we are stymied by a different kind of complexity. Most evidence points to environmental factors acting in the second trimester, two or more years before a diagnosis of ASD. Several studies funded by NIH are looking for differences in the gestational environment of children later diagnosed with ASD. The answers — and there will be answers — will no doubt merge genetic risk and environmental exposure to help us reach the far side of the complexity of ASD.
- Levy D, Ronemus M, Yamrom B, Lee YH, Leotta A, Kendall J, Marks S, Lakshmi B, Pai D, Ye K, Buja A, Krieger A, Yoon S, Troge J, Rodgers L, Iossifov I, Wigler M. Rare de novo and transmitted copy-number variation in autistic spectrum disorders. Neuron. 2011 Jun 9;70(5):886-97. PubMed PMID: 21658582.
- Sanders SJ, Ercan-Sencicek AG, Hus V, Luo R, Murtha MT, Moreno-De-Luca D, Chu SH, Moreau MP, Gupta AR, Thomson SA, Mason CE, Bilguvar K, Celestino-Soper PB, Choi M, Crawford EL, Davis L, Davis Wright NR, Dhodapkar RM, Dicola M, Dilullo NM, Fernandez TV, Fielding-Singh V, Fishman DO, Frahm S, Garagaloyan R, Goh GS, Kammela S, Klei L, Lowe JK, Lund SC, McGrew AD, Meyer KA, Moffat WJ, Murdoch JD, O'Roak BJ, Ober GT, Pottenger RS, Raubeson MJ, Song Y, Wang Q, Yaspan BL, Yu TW, Yurkiewicz IR, Beaudet AL, Cantor RM, Curland M, Grice DE, Günel M, Lifton RP, Mane SM, Martin DM, Shaw CA, Sheldon M, Tischfield JA, Walsh CA, Morrow EM, Ledbetter DH, Fombonne E, Lord C, Martin CL, Brooks AI, Sutcliffe JS, Cook EH Jr, Geschwind D, Roeder K, Devlin B, State MW. Multiple Recurrent De Novo CNVs, Including Duplications of the 7q11.23 Williams Syndrome Region, Are Strongly Associated with Autism. Neuron. 2011 Jun 9;70(5):863-85. PubMed PMID: 21658581.
- Gilman SR, Iossifov I, Levy D, Ronemus M, Wigler M, Vitkup D. Rare de novo variants associated with autism implicate a large functional network of genes involved in formation and function of synapses. Neuron. 2011 Jun 9;70(5):898-907. PubMed PMID: 21658583.
- Schaaf CP, Zoghbi HY. Solving the autism puzzle a few pieces at a time. Neuron. 2011 Jun 9;70(5):806-8. PubMed PMID: 21658575.
- Sakai Y, Shaw CA, Dawson BC, Dugas DV, Al-Mohtaseb Z, Hill DE, Zoghbi HY. Protein interactome reveals converging molecular pathways among autism disorders. Sci Transl Med. 2011 Jun 8;3(86):86ra49. PubMed PMID: 21653829.
- Voineagu I, Wang X, Johnston P, Lowe JK, Tian Y, Horvath S, Mill J, Cantor RM, Blencowe MJ, Geschwind DH. Transcriptomic analysis of autistic brain reveals convergent molecular pathology. Nature. 2011 Jun 16;474: 380-86. PMID: 21614001.