words by Eve Jacobs / photograph by Andrew Hanenberg
im Millonig’s laid-back demeanor belies the complexity of the tasks he has shouldered and the impact of his findings to date. The prevalence of autism spectrum disorder (ASD) seems to be rising precipitously, with one in every 91 children in the U.S. ages 3 to 17 affected, according to a report released in October 2009 in the journal Pediatrics. Its impact on families is immeasurable.
Genes are strongly implicated, but unlike Huntington’s disease and Down syndrome, no one gene is responsible. Millonig has located a gene on a stretch of chromosome that seems to raise the risk of inheriting this disorder.\
Lest you think this researcher has spent all his years puzzling out science in a laboratory, know that he is a Jersey boy from Lawrenceville and Monmouth Beach who spent his teen summers jumping the waves. He didn’t discover his affinity and talent for research until his college years at the University of Rochester. After earning his bachelor’s in three years, he pursued studies in microbiology and genetics at Oxford, graduating with a Master’s one year later. His undergraduate years coincided with the “exciting infancy” of molecular biology and the newfound ability to manipulate DNA. When he returned to the U.S, it was to join the lab of Shirley Tilghman, professor of molecular biology — now President of Princeton University — to work on his doctorate in mouse molecular genetics with a focus on how to use the mouse as a developmental system for studying humans.
In 1990, after earning his PhD, he moved to Rockefeller University to join a lab looking at brain development in the mouse as a model for human CNS development. For him, it was a welcome giant leap from the limitations of test tube science. It was here he learned about the pivotal role of the cerebellum. “If you damage the cerebellum, you get a mouse that’s uncoordinated, a mouse that falls over,” he comments.
He joined the faculty of RWJMS in the fall of 1999, and shortly thereafter, he remembers, the Governor’s Council on Autism was founded. “I was intrigued by research showing that the cerebellum is small in humans affected by autism, so I applied for a grant,” he recalls.
He was successful. “It was the time of the Human Genome Project and there was a long history of using the mouse to study human disease,” Millonig says. Research had been published linking damage to the mouse gene Engrailed 2 with a smaller-than-normal mouse cerebellum. “You identify a mutated gene in a mouse,” he explains, “then you look for the gene mutation in humans.”
The grant from the Governor’s Council gave momentum to the scientist’s search for human genes linked to autism susceptibility. Millonig hypothesized that ENGRAILED 2 should be a prime suspect. Together with Linda Brzustowicz, MD, a Rutgers professor, he followed his instinct and was able to pinpoint two small areas on the chromosome that are inherited far more frequently by individuals with autism than those exhibiting no symptoms. The groundbreaking results of their study — involving 167 families with at least two children diagnosed with autism — were published in 2004. Within several years, they replicated their findings in a total of 548 families, with an error rate of only one in a million, as near to a sure thing as one gets in science.
In 2008, their finding was licensed to a European diagnostic firm, which will soon launch an autism diagnostic kit, the first-ever test for the disorder. Its benefits are obvious. The earlier autism is diagnosed, the quicker interventions — that have proven beneficial for many children — can be started. Otherwise, diagnosis is based on observed behaviors, and can be missed until a child is 2 and older.
Millonig is investigating whether nongenetic environmental factors play a role in autism risk. “Many genes contribute to an increased risk for autism,” Millonig explains, “but many people with the gene mutations do not have autism. It seems to me you need a combination of genetic variants and environmental factors.”
There are still many differing views of what causes autism, he continues. “It’s probably caused by a combination of different risk factors for different individuals.”
The new diagnostic kits will test for several different common gene variants, some from discoveries made at other universities. While Millonig takes pride in the role his research has played in the launch of this first diagnostic tool, the commercial viability of his work is not high on his list of priorities.
“I learned early on from watching Shirley [Tilghman ] —who was an amazing mentor — that you just do the best science you can in a focused way and relate it to the big picture. You design experiments to see if your hypotheses are true or not.”
His joy is in striving to be a focused scientist and a great teacher-mentor. His small lab — six total, four of them graduate students — is exactly right, he says. “I meet with each student at least once a week to go over their data. If the lab were larger, I couldn’t do that. I love being a mentor.”
While the primary focus of the lab is autism, the team is also studying schizophrenia, and spina bifida and other neural tube defects. “We are trying to understand human neurodevelopmental processes,” Millonig explains. A giant undertaking for such a small operation — but this scientist has never shied away from daunting tasks. And, who knows? Another pivotal discovery could be just around the (genetic) corner.
Genetics and Increased Autism Risk
Autism Spectrum Disorder (ASD) is a human neurodevelopmental disease with an incidence of ~ 1:150. ASD includes autism and two milder forms: Asperger’s Syndrome and Pervasive Developmental Disorder-Not Otherwise Specified. Individuals with ASD display defects in language and emotional reciprocity, along with repetitive movements and behaviors. ASD is typically diagnosed between 2-3 years of age.
Risk for ASD has a genetic basis. For example identical twins have ~92% concordance for ASD compared to ~10% for dizygotic twins. In addition when parents have one child with ASD, they are 50-100x more likely than the general population to have another child with ASD. If genetic variants that increase ASD risk can be identified, then treatments can be started prior to the development of symptoms, which may decrease the incidence and severity of the disorder.
Individuals with autism display several neuroanatomical abnormalities including a smaller cerebellum and a decrease in the number of Purkinje cells, a type of cerebellar neuron. Rodent studies have determined the cerebellum coordinates motor movement. However fMRI analysis indicates the human cerebellum coordinates multiple tasks including attention, and language, which are affected in ASD.
Because developmental programs are typically conserved between rodents and humans, we decided to test genes important for cerebellar development for their involvement in ASD. Approximately 100 candidate genes were selected. These genes were then prioritized based upon whether the human gene mapped to a genomic region linked to ASD and whether mouse mutants caused similar cerebellar anatomical defects as observed in humans with autism. The homeobox transcription factor, ENGRAILED 2 (EN2), was chosen for further analysis because the gene mapped to distal chromosome 7 near markers that displayed some linkage to ASD and two different mouse mutants exhibited autistic-like cerebellar phenotypes, including hypoplasia and a reduction in Purkinje cell number.
EN2 is encoded by two exons and a single intron in ~8.0kb. In collaboration with Linda Brzustowicz’s group (Department of Genetics, Rutgers University), we tested whether genetic variants in EN2 were correlated with increased ASD risk. For this analysis, pedigrees obtained from the Autism Genetic Resource Exchange (AGRE) were used that have at least 2 children diagnosed with ASD, suggesting these families had increased genetic risk for ASD. Two intronic genetic variants were consistently and significantly inherited more often in affected children compared to unaffected siblings. These results were first obtained in an initial 167 families (750 individuals, P=.001), and then replicated in two additional datasets (518 families total; P=.00000035). Other groups have also reported genetic association of EN2 with ASD.
These studies suggested a genetic variant existed in or around EN2 that increased risk for ASD. Subsequent human genetic studies determined the genetically associated intronic EN2 variants were the best candidates for increasing ASD risk. We then tested whether these genetic variants were functional. Numerous in vitro and in vivo studies have determined these ASD associated variants affect the regulation of EN2 and result in increased expression.
In sum our experiments have demonstrated that certain genetic variants in EN2 are inherited more often in individuals with ASD and these same genetic variants are functional, resulting in increased levels of EN2.
Thus EN2 may be useful as a diagnostic or therapeutic tool for ASD. In 2004, a patent was filed to cover these claims. In 2007, IntegraGen licensed the patent to develop a diagnostic test for ASD. In combination with variants in other genes, the goal of the diagnostic test is to identify children at risk for the ASD before they develop symptoms.
Ongoing research in my lab is employing a series of mouse models to determine the downstream molecular and cell biological effects of the EN2 genetic variants. Our long-term goal is to identify the pathways disrupted by EN2 during CNS development, which may lead to the identification of new targets for pharmaceutical and behavioral treatments.