New results published in Nature from a team co-led by researchers who have received BBRF grant support mark an important step forward in the decades-long effort to connect genetic risk factors for autism spectrum disorders (ASD) with biological processes involved in causing the illness.

Many of the specific processes identified in the new study occur during the early phases of cortical development, before birth, and suggest how different genetic mutations involved in autism give rise to a variety of changes in developmental processes that are linked to changes in the regulation of gene expression.

The team’s senior members were Daniel H. Geschwind, M.D., Ph.D., of UCLA, and Sergiu P. Pasca, M.D., of Stanford, whose labs collaborated for over a decade in this research. Dr. Geschwind is a 2015 BBRF Distinguished Investigator, 2012 winner of the BBRF Ruane Prize, and a 1999 BBRF Young Investigator. Dr. Pasca is a 2017 BBRF Independent Investigator and 2012 Young Investigator. One of the paper’s three co-first authors was Aaron Gordon, Ph.D., a 2018 BBRF Young Investigator at UCLA, whose grant project was specifically devoted to the research now reported by the team. The other co-first authors were Drs. Se-Jin Yoon and Lucy K. Bicks.

Hundreds of so-called risk genes for autism have been identified over the last 20 or so years, with over 200 characterized as “high-confidence.” Mutations in these ASD risk genes are very rare. Collectively, they are thought to account for 10% to 15% of people diagnosed with ASD. It is hypothesized that combinations of commonly occurring mutations in concert with environmental factors are implicated in causation in the majority of ASD cases, but this leaves open the key question of mechanisms. Is there any way to narrow down the many potential biological paths that end in autism, whether the genetic culprits are rare or common?

Despite ASD’s genetic heterogeneity, note authors of the new study, postmortem analysis of gene activation patterns (“the transcriptome”) in brain cells of individuals who died with autism have revealed patterns of changes in gene expression relative to healthy brain cells. This, along with other evidence, has supported a theory that the diversity of involved genes in ASD may generate biological effects in the brain that can, if probed, reveal “a convergent molecular pathology,” i.e., a recurring and relatively consistent set of effects from a great diversity of distinct genetic causes. This theory, if it can be validated, supports the hope that a limited number of future interventions may be developed to treat a broad spectrum of ASD-related illnesses that have many different primary genetic causes.

One problem immediately facing this and other research teams is the fact that ASD risk genes, in many instances, are most active during the time of human fetal development, when researchers cannot directly access emerging brain structures to observe how things go wrong. For this reason, the researchers conducted their study in cultured brain cells grown from cells harmlessly sampled from ASD patients. Using stem-cell technology, they reprogrammed patient skin cells to redevelop in the lab as neurons and “helper” glial cells. They used these reprogrammed cells to form organoids, which have been pioneered by Dr. Pasca and colleagues over the past decade as models of human brain development in the lab. Because the cells composing the organoids are derived from patients, and bear their genomes (replete with mutations in ASD risk genes), they are potentially capable of reproducing pathological process during early brain development that are involved in autism’s causation.

The researchers assembled a large collection of pluripotent stem cells based on skin cells derived from patients. The 70 cell lines, used in over 100 different experiments in the project, were based on cells sampled from 55 individuals. These included individuals who carried eight distinct mutations linked to specific ASD syndromes, as well as from other ASD patients with no known genetic risk factors (“idiopathic ASD”); and a number of neurotypical individuals to serve as controls. The 70 cell lines were the basis for cellular reprogramming to generate human cortical organoids. These self-organizing clusters of cells are the equivalent of mini-test beds for cortical development, in which the cells “differentiate,” or mature, into cell types that populate the emerging fetal cerebral cortex. The organoids were profiled at four time points, 25, 50, 75 and 100 days following initial differentiation.

The method of profiling, called RNA-seq, enables researchers to capture, quantify, and analyze the entire transcriptome (all RNA molecules) of the cells being analyzed. In this project, when applied across the organoids generated from the 70 different cells lines, RNA-seq makes possible comparisons of the levels at which individual genes are expressed, i.e., activated, in the cells comprising the organoids. High-confidence autism genes with major mutations cause changes in gene expression; this is a starting point for inquiry into the effects upon developmental biology that these changes in gene expression cause.

Dr. Geschwind explains that while prior postmortem studies of the adult brain in autism patients “demonstrated convergent changes” in gene expression and in molecular factors regulating gene expression, the team now set out to show “how such convergence is related to diverse genetic risk factors, and how it emerges during development.”

Early in development, and specifically at 25 days following initial differentiation of cortical cells, the team found large changes in gene expression that were related to specific DNA mutations in the eight ASD syndromes from which some of the analyzed cells were generated. The key finding was that different mutations generated changes in gene expression that converged over time, as development progressed. The changes involved factors that initiate transcription, as well as those that regulate how multiple genes “co-express” at a given moment in time.

Reducing the immense detail that their many experiments generated to a key concept, Dr. Pasca said the study illustrates at least one important way in which the “risk” previously linked with particular gene mutations implicated in ASD is focused, through the process of altered transcriptional regulation, into signaling defects in emerging brain cells.

The biological pathways that are perturbed in the generation of these defects in some instances converge over the course of development. (In other cases, different mutations can generate divergent pathways, the data indicates.) It is the convergent pathways that are perhaps most intriguing. “We observe early mutation-specific effects, which converge on shared transcriptional patterns as development progresses,” Dr. Geschwind notes. “We view this study as an important first step and as strong motivation for broader analyses” of cellular and physiological changes in developing cortical cells derived from ASD patients.

In this study, the convergence noted by the team in transcriptional pathways was seen in cells grown from patients with known ASD risk genes. There was “no shared genetic signal” in cells grown from ‘idiopathic’ ASD patients with no known risk genes. The researchers said this was likely a function of sample size, with the small number of cell donors with idiopathic autism (11) not sufficient to generate a discernable signal (changes in gene expression and transcriptional regulation, for instance) in this series of experiments.

“This is just the beginning” said Dr. Pasca. “We will need larger cohorts and more advanced [cell] cultures that capture brain cell diversity and interactions.” He hopes to do this using a technology his team has previously introduced, in which organoids representing a wider range of cell types (in this case, those present in the developing brain) are combined to form larger structures called assembloids. These more complex structures can more faithfully reproduce the biological processes that occur in the brain as it actually develops, and potentially reveal, in the context of incipient ASD, how these are perturbed.