NIH-Oxford Gene Atlas of Mouse Cortex Offers Key Insights

Nancy A. Melville

August 30, 2011

August 30, 2011 — A new map of gene expression in the cerebral cortex of the mouse brain offers abundant insights into the molecular patterns of brain activity that could help significantly advance the understanding of neuropsychiatric disease.

The cerebral cortex, in humans, is the region responsible for higher cognitive function, including episodic memory, sensory perception, and language, and as 90% of mouse genes are shared by humans, the new findings hold relevance in terms of potential mouse models for human diseases.

In a study published in the August 25 issue of Neuron, the researchers, from the National Human Genome Research Institute (NHGRI), which is part of the National Institutes of Health (NIH), and Oxford University in the United Kingdom, note that the atlas shows the activity of more than 11,000 genes in 6 neocortical layers in the neurons of the normal adult mouse brain.

The international collaborators have made the new atlas freely available online.

Significant Roles

The gene activity was mapped with an approach that started with microdissecting the brains of 8 adult mice, separating the cortex layers, and then purifying processed RNAs, including messenger RNA, from each cortical layer.

The messenger RNA indicated whether a gene was switched on, and the amount of messenger RNA demonstrated the extent of the gene's activity.

The researchers were further able to determine which genes were turned on, and to what extent, with the relatively new sequencing technology RNA-seq, which involves copying processed RNA into a form of DNA and then sequencing the resulting DNA on a second-generation DNA.

"We collect RNA, copy that RNA into DNA, chop it into many tiny pieces, and then sequence tens of millions of those pieces at a time," explained T. Grant Belgard, lead author of the paper and an NIH–Oxford fellow in NHGRI's Genome Technology Branch.

The process revealed that more than half of the genes in the mouse cerebral cortex had varying levels of activity in different layers, suggesting that there are areas where certain genes appear to have significant roles.

Genes previously associated with diseases such as Alzheimer's and Parkinson's disease, for instance, were preferentially concentrated in specific layers. "For example, we detected genes previously associated with Parkinson's disease in layer 5, and Alzheimer's disease in layers 2 and 3," he noted. "These are correlations, not necessarily causal, but they do suggest directions for future research."

The researchers also discovered more than 1000 new noncoding RNAs that had not appeared before in existing gene databases.

"Our computational analyses of the molecular evolution of these RNAs suggest that at least some of these are likely to be doing something important in mouse and/or human," Belgard said.

"Previous studies have shown that long noncoding RNAs also exist and have important molecular functions in humans, and we are keen to determine the similarities and differences in how these noncoding RNAs are used in the brains of both species."

Dimension and Detail

Before the introduction of the new atlas, the most comprehensive account of gene expression across cortical layers was from the Allen Mouse Brain Atlas. For that resource, neuroscientists used in situ hybridization, which involved slicing the brain and staining the slices to visualize the activity of individual genes.

The Allen Institute for Brain Science, founded by Microsoft cofounder Paul Allen, published the atlas and also has a comprehensive gene map of the human brain that was published in April this year, as reported by Medscape Medical News at that time.

Allan Jones, PhD, the institute's chief executive officer, suggests that some may debate the designation of the new study as an "atlas," in that it details only 6 of more than 1000 structures in the brain.

"For perspective, the Allen Institute has publicly released microarray data for its Allen Human Brain Atlas that covers approximately 1000 samples across the entire brain for each of 2 human brains, and the earlier Allen Mouse Brain Atlas provides in situ hybridization-based gene expression data for all known genes throughout the entire mouse brain (all structures)," he told Medscape Medical News.

"That said, this new study really nicely shows the power of the new sequence-based techniques for open discovery (splicing, noncoding, new gene discovery, noncoding RNAs, etc)," he added.

However, Belgard says the new atlas in fact provides substantially more dimension and detail than the Allen Mouse Brain Atlas.

"Our approach expands the list of genes for which we know the activity across layers by several-fold. Even more importantly, it allows us to see the structure of genes as they are used in layers of cortex," he told Medscape Medical News.

In contrast, the in situ hybridization used in the Allen Mouse Brain Atlas relies on a molecular probe that only provides information about a small part of a gene, Belgard said.

"This additional dimension of information [with the new atlas] is important because genes can be used in different ways; for example, stitching together exons (fragments of the gene) in different ways can alter the function of the gene," Belgard explained.

"And because we still have a lot to learn about noncoding RNAs and about gene structure, one cannot design a probe for in situ hybridization if one does not know the gene exists. So this 'discovery approach' lets us see in far greater detail what was previously inaccessible."

The findings are as illuminating as they are complex, yet Belgard says neuroscientists are only just getting started.

"We do indeed believe that this is the tip of the iceberg, and that drilling down to individual cell types and larger neural circuits will help us to understand the broader context of where these genes that can modulate disease risk are active."

The next step is a logical one: Belgard and colleagues will attempt to replicate the mouse brain atlas for parts of the human brain next year, and the new approach should accelerate progress toward a better understanding of the genetic influences on disease risk, Belgard said.

"We're not there yet, but this type of information will help us get there."

Power of New Methods

Daniel Geschwind, MD, the Gordon and Virginia MacDonald Distinguished chair in human genetics at the University of California–Los Angeles School of Medicine, and professor of neurology and psychiatry, agreed that the findings provide valuable new insights on genetic activity in the brain and should expand areas of research.

"The paper demonstrates the power of using new genomic methods, such as RNA-seq and strong bioinformatic approaches, to understand the molecular basis of brain circuit structure," he said.

"By careful analysis, they are able to show that a new class of long noncoding RNA molecules actually [is] expressed in different layers of the mammalian brain. The biological function of these RNAs is not known, but this works opens up a new area of study."

The research was funded in the United States by the NHGRI, and in Britain by the Medical Research Council, Wellcome Trust, the Biotechnology and Biological Sciences Research Council, and the Marshall Scholarships. Belgard and Dr. Geschwind have disclosed no relevant financial relationships.

Neuron. 2011;71:605-616. Abstract


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.
Post as: