ARTICLE IN BRIEF
Using a novel technique called laser microdissection, investigators observed that noncoding RNAs active in dopamine neurons are a surprising link between genetic risk, Parkinson's, and psychiatric disease. They contend that these risk variants might reguate the gene switches of dopamine-producing brain cells, explaining addiction, schizophrenia, and schizophrenia.
A team of scientists at Harvard Medical School set out to understand the relationship between the many genetic risk factors associated with Parkinson's disease (PD) and how they alter the life of a dopamine neuron. What they uncovered, however, was a vast galaxy of dark matter DNA — so called junk DNA — that had never been seen before, according to the study published in the September 17 issue of Nature Neuroscience.
These clusters of noncoding DNA elements were expressing enhancer RNAs that are linked to several genetic risk variants for PD, the researchers found. Many of the enhancer RNA switches in the dopamine neurons were also involved with addiction, schizophrenia, and sleep disorders, they reported.
The study opens up new explanations for these conditions and potential new treatments. “We think that these are key switches at the core of PD, as well as psychiatric conditions,” said the senior study author Clemens Scherzer, MD, director of the American Parkinson Disease Association Center for Advanced Parkinson's Disease Research.
“This work suggests that noncoding RNAs active in dopamine neurons are a primary link between genetic variation and neurodegenerative and neuropsychiatric diseases affecting dopamine neurons in our brains,” Dr. Scherzer said. “We think that these risk variants target the enhancer RNAs active in dopamine neurons.”
In PD, not enough dopamine is produced, he explained, whereas in psychiatric conditions, there is too much.
“The new findings could help explain complications faced by some PD patients, but not by others — not just movement problems but depression, anxiety, fatigue, memory, and sleep,” Dr. Scherzer said. “These are multisystem impairments that could be modulated by switches turned on or off in dopamine neurons.”
The Harvard scientists now have a complete encyclopedia of the entire RNA content in dopamine and pyramidal neurons in the human brain — comprising more than 70,000 new RNAs that have never been described. They are sharing this content with other scientists through the website, www.humanbraincode.org.
STUDY FINDINGS, METHODS
To do the analysis, Dr. Scherzer and his colleagues used laser capture microdissection to identify noncoding elements transcribed in dopamine neurons and pyramidal neurons of human brains. They analyzed dopamine neurons cut from the substantia nigra of 89 postmortem brains of people, ages 72 to 80, who were healthy at the time of death.
Next, they did high-throughput sequencing, identifying the entire RNA in dopamine neurons, pyramidal neurons, and non-neuronal cells. They sequenced RNA because that indicates which parts of the genome are “active,” Dr. Scherzer said.
“We found a big surprise: 64 percent of the human genome is active in the dopamine neurons. Most of the activity is in the non-coding dark matter,” he said. “These non-coding RNAs don't make protein but they do have regulatory responsibilities. There is a complex regulatory circuit that determines which cells become dopamine neurons and other cell types.”
The researchers found the risk variants for PD were clustered in the active dark matter RNAs. “These RNAs act as switches that enhance the expression of protein-coding genes,” Dr. Scherzer said. “Our finding suggests that these risk variants may fiddle with the genetic switches.”
“This gives us a new clue about how these genetic risk variants might function in dopamine neurons,” he added. “Now we can better understand function and figure out how to target the active genes in the dark matter of these cells. Maybe we can correct these gene switches.”
Dr. Scherzer said that there are gene-editing tools or drugs that target RNAs that may ultimately be helpful in fixing these switches. It is not an easy fix, he added. “These RNAs are active at low levels and they don't hang around. You could go upstream and target the DNA. We are planning to go downstream and identify the cellular networks affected in response to turning these switches on or off. Then we can screen for therapeutics to treat the cellular network nodes.”
This study was a collaborative effort between scientists at Harvard, Brigham & Women's Hospital, the University of Birmingham, UK, Southeast University in Nanjing, China, The University of Sydney in Australia, the German Center for Neurodegenerative Diseases, Mayo Clinic, the Harvard Brain Tissue Resource Center at McLean Hospital, University of Auckland, the Kubik Laboratory for Neuropathology at Massachusetts General Hospital, the University of Kentucky, and the Banner Sun Health Research Institute.
“This is a very interesting paper that really advances how we think about genetic regulation in neurological and psychiatric diseases,” said Jeremy J. Day, PhD, assistant professor in the department of neurobiology at the Evelyn F. McKnight Brain Institute at the University of Alabama at Birmingham.
“We've known for a while that the majority of genetic variants linked to human disease from genome-wide association studies [GWAS] fall outside of protein coding regions of the genome. Some of this non-coding genomic space encodes what we refer to as ‘enhancers,’ or regions of DNA that help to regulate which genes are turned on or off in specific types of cells. These enhancers are highly relevant in the brain, where there is enormous diversity in gene expression patterns even within brain structures that we know a lot about — like dopamine circuitry.
“This has led many people in this field to wonder whether GWAS hits linked to neuropsychiatric disease might be altering the function of key enhancers in specific cell types,” Dr. Day continued. “However, this is difficult to test because enhancers are highly cell-type specific; an enhancer that is active in one cell type might not be active in another cell type. To identify active enhancers in dopamine neurons, this paper took advantage of the fact that active enhancers are actively transcribed by RNA polymerases, generating long-noncoding RNAs called enhancer RNAs, or eRNAs.”
“Intriguingly, as compared to other cell types, enhancers that were active in dopamine neurons were highly enriched for GWAS variants that confer risk for things like schizophrenia, addiction, and Parkinson's,” Dr. Day said. “This result is exciting because it ascribes a functional role to these GWAS hits for the first time, and opens the door for future discovery and characterization of what these enhancers do — which genes they regulate, how they work at the mechanistic level, and how they are dysregulated in disease states. Overall, these results also tell us that in order to understand the link between genetics and complex neuropsychiatric disease, we must understand how the genome is uniquely regulated in different cell populations that give rise to disease-linked behavioral outcomes.”
Joshua M. Shulman MD, PhD, FAAN, principal investigator of the laboratory for integrative functional genomics, and associate professor in the departments of neurology, neuroscience, and molecular and human genetics at Baylor College of Medicine, agreed.
“Dr. Scherzer and colleagues have generated an atlas of gene regulation within human dopamine neurons. By focusing efforts on genomic regulatory elements (non-coding sequences that control expression of other genes) they provide insight into the underlying circuitry and logic for gene expression in dopaminergic neurons — and its potential relevance for disease,” Dr. Shulman said.
“Importantly, the researchers have found that a significant proportion of genetic variants affecting risk for Parkinson's disease and other neuropsychiatric disorders fall within these regulatory elements, informing their potential mechanism. Thus, this work creates a powerful resource for further functional genomic dissection of Parkinson's and likely many other neurologic disorders.”
Mark R. Cookson, PhD, senior investigator and cell biologist in the laboratory of neurogenetics at the National Institutes of Health, agreed that the paper provided an important contribution to the literature. “Resource papers [like these] are very helpful in outlining gene expression that the community can use,” Dr. Cookson said. “The availability of a simple interface is also useful, especially in getting the data out to the broader community. It is clear that a huge amount of work went into generating the data and analysis, which is very impressive.”
But he added: “I'm not surprised that non-coding regions contain information about genetic risk for diseases. Nearly all genome-wide association studies, whether in neuropsychiatric conditions or anything else, show that coding variation accounts for a tiny amount of the signal, therefore genomic regions like enhancers are good candidates for the regulation of genes and expression regulation.”
“We, and many others, have published quite a lot over the years showing that in many loci for PD or other conditions that non-coding (not junk) variants are associated with altered gene expression. I would also be cautious about the assumption that dopamine neurons are ‘the’ important cell type in PD or the psychiatric conditions listed here,” Dr. Cookson said.
“It has been reported previously that some genes for PD are expressed only at low conditions in dopamine neurons but are expressed in inflammatory cells instead,” he continued. “In the longer term, solving GWAS loci will take thoughtful approaches integrating expression in the context of linkage disequilibrium, which is a fiendishly difficult problem, as the confusion around the chromosome 17 locus that includes MAPT demonstrates.”