CAPTION

For decades, Cold Spring Harbor Laboratory scientists and collaborators have invested millions of dollars into deciphering the genetic causes of autism Cold Spring Harbor Laboratory

Cold Spring Harbor Laboratory (CSHL) researchers have flipped the script on autism ) genetics.

Scientists long thought that siblings born with autism share more of their mother’s genome than their father’s. But CSHL Associate Professor Ivan Iossifov and Professor Michael Wigler have now shown that, in many cases, it’s dad who might be playing a bigger genetic role.

Autism spectrum disorders cover a range of neurological and developmental conditions. They can affect how a person communicates, socializes, learns, and behaves. Autism may also manifest as repetitive behaviors or restricted interests. In the United States, it affects around one in 36 children. 

“There are children diagnosed with autism who are high functioning,” Iossifov says. “They have a completely productive life, although they have some minor troubles in social interactions, as most of us do. But also, there are children diagnosed with autism who never learn to speak, and they have definitely a difficult life.”

Over the last two decades, CSHL scientists have led a multimillion-dollar effort to uncover the genetic origins of autism. They discovered thousands of genes that, when damaged, may cause a child to be born with autism. But their work was not able to account for all cases of autism. So Iossifov and Wigler set out to find the missing sources.

The duo analyzed the genomes of over 6,000 volunteer families. They found that in families that have two or more children with autism, the siblings shared more of their father’s genome. Meanwhile, in families where only one sibling had autism , the children shared less of their father’s genome. While the discovery reveals a new potential source of autism, it also poses a provocative question. Could other disorders play by the same genetic rules?

No one is sure how dad’s genome makes its mark on children with autism. But Iossifov has a couple interesting ideas. He thinks some fathers may carry protective mutations that fail to get passed on. Or fathers may pass down mutations that trigger the mother’s immune system to attack the developing embryo. Both theories offer hope for parents of children with autism .

“Our future research is exciting,” Iossifov says. “If one of those theories or two of them prove to be true, then it opens different treatment strategies, which can, in the future, affect quite a lot of families.”

In addition, this research offers helpful tools for educators and therapists. It may allow for earlier diagnoses and a better overall understanding of autism.

Study finds that major vault protein is needed for homeostatic plasticity.David Orenstein

The MVP protein appears green in these stained neurons (red denotes neuronal marker SATB2). Credits: Image courtesy of the Sur Lab

In a new study of one of the most common genetic causes of autism, neuroscientists at MIT’s Picower Institute for Learning and Memory have identified a molecular mechanism that appears to undermine the ability of neurons in affected mice to properly incorporate changes driven by experience. The findings, published in the Journal of Neuroscience, suggest that a particular gene, MVP, is likely consequential in people with 16p11.2 deletion syndrome.

Accounting for up to 1 percent of autism cases, 16p11.2 deletion occurs in people who are missing a small region of DNA near the center of one copy of chromosome 16. For years, scientists have been working to determine exactly how the reduced presence of 29 protein-encoding genes leads to clinical symptoms of the syndrome, such as autism-like behaviors, developmental delay, and intellectual disability.

“This has been a major problem for the field,” says senior author Mriganka Sur, the Newton Professor of Neuroscience at the Picower Institute and director of the Simons Center for the Social Brain at MIT. “People have looked at the entire region in mice. Our strategy was different. We thought, can we make a hypothesis about a critical gene that should play an important role?”

The team, led by postdoc Jacque Pak Kan Ip, focused on MVP, which is known for encoding a protein that shuttles RNAs and proteins from the nucleus to the rest of cells — a generally vital function. It is particularly important in the immune system and is known to regulate other genes as well. Although MVP is also known to be among the 29 genes affected in people with 16p11.2 deletion syndrome, its specific function in neurons has barely been investigated. To change that, the researchers devised a series of tests of MVP in a well-understood region in the visual cortex, where the brain processes sight from both eyes.

Keeping an eye open for change

Ip and the co-authors employed the time-honored protocol of monocular deprivation, or temporarily shutting one eye for a week. Normally, responses in visual cortex neurons related to the closed eye become weaker, but responses related to the remaining open eye become stronger, as if to compensate for the change in ability. This neural circuit adjustment to experience is called homeostatic plasticity.

In the normal mice, closing an eye for a week had the expected effect. But in mice with one copy of MVP missing, the researchers observed a telling difference. Responses related to the closed eye still weakened, but responses from the open eye did not get stronger. Homeostatic plasticity was disrupted.

“After seeing that MVP is responsible, we asked how MVP does it,” Ip says.

In those further tests, the team found that neurons in MVP-reduced mice experienced less electrical current across their excitatory synapse connections with other neurons, suggesting reduced functional excitatory synapses. They found other differences, too. MVP-reduced mice overexpressed the gene STAT1, an immune-system gene known to be regulated by MVP.

In 2014, Sur’s lab had found that mice without STAT1 have unusually strong open-eye responses after monocular deprivation. Now in the new study, with less MVP than usual and too much STAT1, mice experienced the opposite. Sure enough, when the researchers knocked down STAT1 as well as MVP, they were able to bring back a near-normal open-eye response to monocular deprivation, suggesting that loss of MVP disrupts homeostatic plasticity by allowing for an overabundance of STAT1.

Then Ip, Sur, and co-authors dug even deeper. They found that in MVP-reduced mice, neurons weren’t producing the expected open-eye responses because they weren’t expressing a key receptor, the GluA1 AMPA receptor, on the surface of dendritic spines.

Even though the findings were made, by design, in the visual cortex, the authors said they expect the disruption of homeostatic plasticity to occur elsewhere as well.

“This gene’s presence is reduced everywhere in the brain,” Sur notes.

The research adds to what scientists know about the connection between copy number variants (repeats or deletions of genes) and autism spectrum disorder, says Mustafa Sahin, a professor of neurology at Harvard University and director of the Translational Neuroscience Center and the Translational Research Program at Boston Children’s Hospital.

“Copy number variations in our genome are commonly associated with intellectual disability and autism spectrum disorder,” says Sahin, who holds the Rosamund Stone Zander Chair. “While 16p11.2 microdeletion was identified as a genetic cause of autism 10 years ago, identifying which gene or genes in this region play a crucial role in brain development and function has been a holy grail of research in this field. Sur lab has made major inroads into this question by demonstrating the MVP+/- mice show deficits similar to 16p11.2 mutant mice and even identifying potential therapeutic targets downstream of the MVP gene.” 

Relationship with immune system

In addition to illustrating how a specific gene may contribute to the symptoms of 16p11.2 deletion syndrome, Sur said, the findings also raise an intriguing, broader question about the central nervous system: Is it perhaps not a coincidence that some of its ability to adjust to and incorporate experience comes from genes that are also active in the immune system?

“The immune system is really a system of learning and memory,” Sur says. “You get infected and the body makes antibodies and the next time there is a memory of the infection. It is conceptually a very similar idea.”

Workflow of M-DATA CREDIT Yuhan Xie, CC-BY 4.0 (https://creativecommons.org/licenses/by/4.0/)

Researchers identified almost two dozen genes that contribute to heart defects by studying genetic data from people born with congenital heart disease or autism. Hongyu Zhao of Yale University and colleagues developed a new algorithm to analyze genetic data from related conditions, which they describe in a new paper publishing November 4^th in the journal PLOS Genetics.

Multiple diseases that start early in life appear to be linked to mutations in the same genes.  Recent research looking at de novo mutations—new mutations that pop up in children that were not present in the parents—has demonstrated a connection between congenital heart defects and autism. However, sequencing de novo mutations is expensive, so small studies of individual diseases have limited power to identify genes that increase a person’s risk of the disease.

In the new study, researchers developed an algorithm called M-DATA (Multi-trait De novo mutation Association Test with Annotations) that combines sequencing data from people with related conditions to identify genes that contribute to disease. They applied the new method to genetic data from people with congenital heart disease or autism and successfully identified 23 genes for congenital heart disease, including 12 that were previously unknown.

The researchers conclude that M-DATA is more effective at identifying genes that increase a person’s risk than analyses focusing on a single disease. This is because instead of analyzing a small number of genomes from affected individuals, M-DATA analyzes a larger number of combined genomes from multiple groups of people. The new method may help researchers identify previously unknown genes linked to disease and improve our understanding of the cause and potential treatment for different conditions.

Zhao adds, “By jointly analyzing de novo mutations from congenital heart disease (CHD) and autism, we identified novel genes that may play an important role in explaining the shared genetic etiology of CHD and autism.”

Yuhan Xie, the lead student of the research, says, “As a biostatistics student, it’s very motivating to find what could be meaningful to the patients and their families.”

New Study in Nature Genetics uses SPARK database to investigate variants that are less damaging than de novo mutations, but contribute almost as much risk

Researchers have identified a rare class of genetic differences transmitted from parents without autism to their affected children with autism and determined that they are most prominent in “multiplex” families with more than one family member on the spectrum. These findings are reported in Recent ultra-rare inherited variants implicate new autism candidate risk genes, a new study published in Nature Genetics.

The hunt is on in earnest for the genes involved in autism, now that technology and vastly lower costs allow the aggregation of thousands of genomes of people with autism and their family members. Knowing precisely which genes are at play will enable greater understanding of the condition known as autism and may ultimately lead to treatments for those who desire them.

This new study is notable because the majority of autism genes discovered to date have been identified through studies of de novo mutations, genetic differences that first arise in the person with autism but are not present in either of their parents. The findings indicate that researchers should not assume that the set of autism genes altered by de novo mutations are the same genes as these newly identified inherited rare variants.

According to lead author Amy B. Wilfert, Ph.D., of the University of Washington, in an analysis of 10,905 people with autism, researchers identified and replicated a rare class of genetic variants that are passed (over-transmitted) from parents without autism to children with autism.

“While most autism studies focus on de novo mutations, this study focuses on rare inherited mutations, which are often understudied in autism,” says Dr. Wilfert. “We find that these variants are individually less damaging than de novo mutations but have the potential to contribute almost as much risk and impact the same molecular pathways, through a distinct set of genes. These variants, however, are only able to persist in the general population for a few generations before being selected out by evolution.”

“It is widely understood that de novo mutations cannot and do not explain all of the genetic causes of autism, a phenomenon sometimes referred to as ‘missing heritability,’ ” says Pamela Feliciano, Ph.D., Scientific Director, SPARK (Simons Powering Autism Research). The SPARK Consortium contributed more than 50 percent of the genetic data analyzed in this study, including exomes from 21,331 SPARK participants — 6,539 of them individuals with autism spectrum disorder (ASD). The number of genomes accessible to scientists at this time enables the search for certain categories of genetic changes — such as de novo changes and ultra-rare inherited variants — but not all of them. As more genomes come online, larger categories of variants will be accessible for analysis.

“Interestingly, the vast majority of those variants (95%) are not found in genes already known to be autism genes, indicating that there is much more to be learned about autism genetics,” says Dr. Feliciano, noting that this study is the first step in a much larger investigation. “While the current study is not large enough to confidently identify individual genes that have these rare inherited variants, we are learning more about these genes. Future research that focuses on multiplex families is increasingly important to yield novel insights.”

The researchers also said the study confirmed their expectations that this class of rare inherited variants is more prominent in families with multiple members with autism than in families with only one affected individual. Consistent with this finding, children with ASD in these families are more likely to carry two of these variants as compared with their unaffected siblings.

The study also highlighted the need for greater diversity when conducting research of this kind, because investigators were less able to detect rare variants in people who belong to ancestral groups that are underrepresented in genomic research, including people of African, East Asian and South Asian descent.

H3K9 methylation levels in the cerebellum were lower in the Suv39h2-deficient mice than in control mice controls. CREDIT RIKEN

New research from the RIKEN Center for Brain Science (CBS) in Japan shows that a deficit in histone methylation could lead to the development of autism spectrum disorders (ASD). A human variant of the SUV39H2 gene led researchers to examine its absence in mice. Published in Molecular Psychiatry, the study found that when absent, adult mice exhibited cognitive inflexibility similar to what occurs in autism, and embryonic mice showed misregulated expression of genes related to brain development. These findings represent the first direct link between the SUV39H2 gene and ASD.

Genes are turned on and off throughout our development. But genetic variation means that what is turned off in some people remains turned on in others. This is why, for example, some adults can digest dairy products and others are lactose intolerant; the gene for making the enzyme lactase is turned off when some people become adults, but not others. One way that genes can be turned on and off is through a process called histone methylation in which special enzymes transfer methyl groups to histone proteins that are wrapped around DNA.

Variations in genes related to methylation during brain development can lead to serious problems. One such variation occurs in a rare disorder called Kleefstra Syndrome, in which a mutation prevents methylation of H3K9–a specific location on histone H3. Because Kleefstra Syndrome resembles autism in some ways, RIKEN CBS researchers led by Takeo Yoshikawa looked for autism-specific variations in genes that can modify H3K9. Among nine such genes, they found one variant in an H3K9 methyltransferase gene–SUV39H2–that was present in autism, and the mutated SUV39H2 prevented methylation when tested in the lab. Similar loss-of-function results were found for the mouse version of the variant.

The next step was to see what happens in mice that lack the Suv39h2 gene. Behaviorally, the researchers found that the mice could learn a simple cognitive task, but had difficultly when the task required cognitive flexibility. In the simple task, mice learned to get a reward by poking a door at alternating diagonal corners of a cage. After they could do this well, the possible reward locations switched to the other two diagonal corners. The genetically modified mice did this as well as wild-type mice. In another task, after learning to alternate between the two diagonal corners, only the location of one reward was switched. When the mice were challenged to alternate randomly between these two tasks, wild-type mice could adapt quickly, but the Suv39h2-deficient mice took much longer. “This serial reversal-learning task was essential,” says first author Shabeesh Balan. “Cognitive inflexibility is a core symptom of ASD, and our new task was able to address this behavioral feature in ways that previous mouse studies could not.”

When the researchers examined what happened in the mouse brain when H3K9 methylation failed to occur, they found that important genes that are usually silenced in early development were turned on in the experimental mice. “Suv39h2 is known to be expressed in early neurodevelopment and to methylate H3K9,” explains Yoshikawa. “This keeps a check on genes that should be switched-off. But without it, genes in the protocadherin β cluster were abnormally expressed at high levels in embryonic mice.” Because protocadherins are critical for the formation of neural circuits, the researchers believe they have found an important biological pathway that could be central to several neurodevelopmental disorders.

The team then verified the importance of SUV39H2 in human ASD by finding that its expression was lower in the postmortem brains of people with ASD than of controls. “What began with a loss-of-function mutation in only one person with ASD,” says Yoshikawa, “has led to a general causal landscape for ASD that culminates in brain circuit abnormality.”

Protocadherins have already been proposed to be related to a broad range of mental disorders. This study shows that activating the SUV39H2 gene is a potential therapy for mental disorders–including ASD–that should be investigated more thoroughly in future studies.