The human brain is arguably the most complex material system in the universe, but an applied mathematician at Xi’an Jiaotong-Liverpool University is part of an international research team that is cracking the genetic code underlying its function.
Dr Pascal Grange (pictured below), a lecturer in the Department of Mathematical Sciences at XJTLU, carried out supporting computational analysis for a major study published in Nature Neuroscience, which sought to analyse data from the publicly available Allen Brain Atlas of ‘gene expressions’ in the brain.
Each cell in our body, including brain cells, acts like a small chemical plant that produces a certain cocktail of proteins, called its expression pattern. These proteins are encoded by genes.
Over the last 10 years the Allen Atlas has mapped gene expressions for the mouse brain and in six human brains, measuring expression patterns for about 20,000 genes. These are the first data sets to cover much of the brain in both the mouse and human, as well as the entire genome.
The study sought to discover whether the expression pattern of a gene across the brain is stable, when comparing mouse models to humans and whether it is stable between human specimens.
It follows on from Dr Grange’s work last year to propose a mathematical model to define brain regions through genetic data, also in collaboration with the Allen Institute of Brain Science.
Dr Grange, who also teaches on the BSc Applied Mathematics programme at XJTLU, explained: “We asked whether we can identify genes whose expression is stable among humans and stable from mouse models to human beings. Across 20,000 genes you can compute correlation coefficients between the data, thereby defining differential stability of gene expression, and rank genes according to the consistency of their expression patterns across specimens.”
In general the study found that gene expression patterns were not all highly reproduced between mouse models and humans, or between human specimens.
However, it did uncover 32 sets of genes whose expression is highly stable, which is relevant to clinical research. Indeed, the most stable gene expression patterns tend to be highly expressed in neurons, the cells in the brain that conduct electricity, which means that mouse models are probably valid for clinical research concerned with neurological diseases such as autism, Alzheimer’s or schizophrenia.
These stable gene expression patterns provide insights into what makes the human brain distinct from other systems and raise new opportunities to target therapeutics.
Additionally, the research illustrates how recent advances in biochemistry and image processing are increasing our detailed knowledge, allowing us to explore it on microscopic scales. For example, the paper identified one set of genes whose collective expression highlights a part of the brain that was first discovered 2,000 years ago (see image below for the expression pattern of these genes in the mouse brain).
“The cerebellum is a small folded structure at the back of the mouse brain which also exists in the human brain. It was first observed in dissected animals by Aristotle in the fourth century BC and we are still talking about the same structure, only at a more microscopic scale. Not only do we have sets of genes for which the expression is differentially stable from mouse to human, but these sets also define brain regions in a data-driven way,” explained Dr Grange.
Dr Grange’s work on the study was supported by the XJTLU Research Development Fund. Healthcare is a huge growth area in China and globally and this work is part of the Department of Mathematical Sciences’ approach to pursuing fundamental research relevant to other fields of science and technology.
“The study of mathematics goes well beyond the improvement of numeracy skills,” said Dr Grange. “Mathematical tools allow researchers to master complexity, engage in dialogue across disciplines and contribute to work such as this that is crucial to human development and supports clinical research.”
Nature Neuroscience is a monthly scientific journal published by Nature Publishing Group. It is a multidisciplinary journal that publishes papers of the highest significance in all areas of neuroscience.