Patricia F. Fitzpatrick Dimond Ph.D. Technical Editor of Clinical OMICs President of BioInsight Communications
As research unravels more genetic clues, it also reveals how tough drug R&D is going to be.
Autism and autism spectrum disorders (ASDs) affect an estimated 3.4 in every 1,000 children ages 3–10. According to an article published this June in Nature, ASDs vary greatly in severity. By including children in regular education who received no special help, an epidemiological study found these disorders to be up to three times more prevalent than previously thought.
Few diseases have provided the opportunity for rampant quackery and exploitation of despairing parents. Real science, though, is rapidly replacing pseudoscience. Evidence is cropping up for the highly complex nature of these diseases and contributing to clarity, if not a cure, and potential therapeutic strategies.
From a biological perspective, autism presents a “profoundly complicated” array of disorders with significant genetic components and genomic alterations thought to “organize around a central theme of neural network infirmities and neuroimmunodysregulations,” according to Daniel J. Guerra, writing in the March issue of Autism Research and Treatment.
Autism is linked to different genes in different people, and multiple genes could be involved in each affected person. These genetic factors, in turn, may interact with as yet unidentified environmental factors. It is now known that new mutations show up in children whose parents do not carry the mutation. Additionally, it turns out that in cases where underlying genetic mutations have been identified, the gene abnormalities don’t necessarily predict the disorder.
“Do we really know that every time you make that mutation or you delete one copy of that gene, you cause autism?” asked John Constantino, M.D., a pediatric psychiatrist at Washington University in St. Louis, in an interview with Los Angeles Times. “We have no idea.”
Demonstrating Complexity
A paper published last March in the Journal of Molecular Psychiatry confirmed the complexity of the genetics of these diseases, but also supported the concept that common biological themes “underlie this complexity.” Study results suggested several new candidate genes and genomic variants as contributors to autism, and the research team concluded that many more remain to be discovered.
Further evidence for the role of genomic aberrations in autism is found in the fact that the disorder occurs concurrently with diseases associated with known cytogenetic etiologies such as Fragile X syndrome. These account for <10% of cases, though. The remainder, often referred to as idiopathic autism, are considered highly heritable with a 5–10% recurrence rate in siblings and a 60–90% concordance rate in monozygotic twins.
Genome-wide studies of autistic individuals have implicated numerous minor risk alleles but few common variants, according to the researchers. This suggests a complex genetic model with many contributing loci. Genetic linkage and association studies have reportedly had limited success with pinpointing risk loci.
Studies of genome copy-number variations (CNVs), on the other hand, have identified several candidate loci, the scientists added. The researchers focused their analysis on rare inherited structural variants in autistic individuals and looked for potential enrichment of any functional categories assigned to genes overlapping the CNVs.
“We identified a set of genes that were over-represented for structural variations in our autism cases, and then we asked whether these genes were preferentially associated with biological functions such as brain development, or with similar phenotypes observed in mouse models of these genes,” Peter White, Ph.D., director of the Center for Biomedical Informatics at The Children’s Hospital of Philadelphia and lead author of the paper, explained to GEN. “It turned out that these genes were enriched in processes consistent with brain development, including regulation of synaptic function, synapse development, and neurotransmission.”
Dr. White’s analysis was based on the AGRE (autism genetic resource exchange) cohort of 5,431 affected and parental samples from 1,000 families, grouped into four sets based on the time when they were recruited. After applying various exclusion criteria, the final discovery cohort, which comprised set 4, included 1,793 subjects, and the replication cohort, composed of sets 1–3, included 1,702 subjects.
Of the genes they identified as enriched for inherited CNVs in ASD subjects, the ionotropic glutamate receptor GRIN2A and CNVs spanning both the islet cell autoantigen (ICA1) and the adjacent α-neurexin ligand NXPH1 were of particular interest. These genes were disrupted by CNVs in subjects from both autism cohorts and had no evidence of structural variation in healthy controls.
Their results, according to the authors, were consistent with the hypothesis that inherited autism risk is genetically highly heterogeneous, both from “our failure to find even a moderate frequency of autism-specific CNVs overlapping any single gene in our analysis and the lack of overlap between gene sets represented by the same autism-enriched functional terms in our two cohorts.”
Animal Model
Understanding the underlying pathophysiology of autism, and ultimately the development of treatments, relies on robust animal models. Scientists working in the laboratory of Daniel H. Geschwind, M.D., professor of neurology at the David Geffen School of Medicine, UCLA, characterized a mouse knockout of the Cntnap2 gene, which is strongly associated with autism and related disorders. Common and rare variants of Cntnap2 are associated with increased autism risk in the general population. In addition, recessive mutations in this gene have been reported in a syndromic form of autism called cortical dysplasia-focal epilepsy syndrome (CDFE).
The Cntnap2(-/-) mice showed deficits in hyperactivity and epileptic seizures like patients with CDFE as well as the three core ASD behavioral domains as have been reported in humans with Cntnap2 mutations, including reduced vocal communication, reduced social interactions, and increased repetitive behaviors.
The animals’ brains also showed abnormalities in development of the neural circuitry. This included abnormal migration of neurons, abnormal activity of the neural network, and fewer interneurons that connect neurons carrying impulses to the central nervous system with those sending them to the rest of the body.
“We believe that the Cntnap2 knockout mouse parallels the CDFE human phenotype very well,” Olga Peñagarikano, Ph.D., the paper’s first author, told GEN. “We see a similar behavioral and neurological phenotype as human patients as well as similar abnormalities in neuronal migration.”
She also explained that she and her team saw reduced numbers of GABAergic neurons in the mice, a finding not previously reported in humans with mutations in this gene. “This is very important because if confirmed in humans with the CDFE phenotype it could be a target pathway for pharmacological treatment.”
The End Game
In 2007, Michael Wigler, Ph.D., of Columbia University, and colleagues reported research results indicating that spontaneous duplications or deletions of at least 130 sites in the genome could contribute to the risk of autism. Dr. Wigler believes that in total there are closer to 400 such sites.
“It is a large number,” he acknowledged. “Given the number of genes that might cause autism, one shouldn’t expect that one treatment is going to cure them all.”
Dr. White added, “theoretically, you could devise a therapeutic regimen that addresses a specific function even though the function is impacted by multiple genes, but it’s too soon to tell how effective this approach might be. The alternative would be an approach that would be highly individualized.
“I think as geneticists and researchers we are always hoping for a simple answer,” Dr. White continued. “But a compelling amount of the evidence says autism is highly complex, and it’s not going to be a quick fix in most cases.”
“The good news is that the gene sets appear to be impacting the same biological processes across the AGRE autism population,” he added. Further R&D in autism “will likely require more focus on systems biology and personalized genomics approaches for success.” The aim would be for research to provide more focused drug targets and the means to test them.
Patricia F. Dimond, Ph.D. ([email protected]), is a principal at BioInsight Consulting.