Cell division is a risky business. DNA damage unavoidably accompanies the enormous number of cell divisions required to generate the human body from a single fertilized egg. In most tissues, cell turnover and regeneration ameliorate the deleterious effects of somatic mutation. The nervous system, however, has a unique vulnerability—neurons generally don’t turn over. As a result, we are all cursed to live our entire lives with somatic mutations acquired during the embryonic development and differentiation of neural progenitor cells.
A conservative mutation rate of 5 x 10-7 mutations/cell/generation would mean that every adult human brain harbors around 5 x 104 somatic mutations in its neurons. Of course, the number of cells carrying each of these mutations depends on when the mutations arise during development, with early-arising mutations populating much larger portions of the brain than later-arising mutations. Based on this rough calculation, the potential exists for every gene in the genome to be mutated in at least some neurons in each of our brains—we are all walking repositories of neuronal genomic diversity.
What are the consequences of sporting a mosaic brain? The somatic patchwork of alterations scattered throughout our neural networks would certainly contribute to human diversity and might account for some of the differences in personality and behavior that often distinguish otherwise identical twins. Most of these mutations will not affect genes that influence the phenotype of a neuron, and some may precipitate cell death before the host cell is integrated into a functional circuit. In rare cases, however, some genetic changes could significantly impact the behavior of a cluster of neurons, or a neuronal circuit, or even an entire neuroarchitectonic structure.
How somatic mutations affect neuronal function
Unraveling the genetic underpinnings of major neurological and psychological disorders has defied traditional approaches. While rare cases of genomic lesions have been documented in familial epilepsy, schizophrenia, autism, and some other disorders, genetic and even newer, more powerful genomic technologies have failed to account for the majority of disease. Some argue that the genetics is complex, with no single gene contributing in a readily detectable manner to disease occurrence. Others contend that disease arises as a consequence of a single rare gene variant that is undetectable because many different single gene mutations could cause the same disease, and earlier studies have been woefully underpowered for detecting this genetic treasure trove. An examination of somatic mutations in neural precursor cells could provide an alternative explanation of the genetics underlying neurological and psychiatric diseases, opening a window on this genetic “dark matter.”
What kinds of neuronal mutations would impart profound changes to the function of the nervous system? Dominant mutations exhibit their consequences directly, whereas the more common recessive lesions must be complemented by a second genetic change in the remaining normal allele for their effects to be expressed. The probability that two independent somatic changes would be acquired in both alleles of a critical gene in a neuronal precursor cell is vanishingly low. However, based on the calculations outlined above, an individual who inherits a recessive allele of a critical gene would be almost certain to lose the second allele by somatic mutation in some subset of his or her neurons. The severity of the phenotype caused by this unfortunate two-hit genetic accident would depend on the type of neurons that acquire the second lesion as well as the number of neurons spawned by the affected precursor cell. For example, loss of the second allele in a majority of cerebellar granule neurons might manifest as an ataxic phenotype, whereas loss of function of the same gene in neurons confined to a single hypothalamic nucleus would have entirely different consequences. Thus, a neurogenetic syndrome associated with inheritance of a recessive allele would be characterized by variable penetrance and a range of phenotypic expression, because of the unpredictable nature and timing of the acquired somatic mutation required to unveil its presence. In this model, there would be relatively little phenotypic concordance between consanguineous twins, making genetic studies of such disorders difficult, if not impossible, to interpret.
We are all walking repositories of neuronal genomic diversity.How can we deconstruct the somatic brain to uncover the genetic alterations that may underlie the majority of devastating neurological, psychiatric, and perhaps even psychological dysfunctions? Modern DNA-sequencing technologies provide an avenue. Next-generation sequencing technologies, and their soon-to-market offspring, allow entire genomes to be deduced from relatively small amounts of DNA. The cost of this technology is dropping rapidly, and may eventually reach the point where funding agencies would support analysis of the neuronal somatic genome.
Pinpointing critical mutations
In anticipation of reaching that point, what sort of approach would reveal the potential extent and contribution of neuronal somatic mutations to human suffering?The human brain represents an enormously intricate puzzle, assembled from neurons derived from a series of committed progenitor cells. Lineage-tracing studies have shown that cohorts of neurons are often derived from the same ancestor, and that specific neuronal subtypes arise from common precursors. In certain brain regions, such as cortical columns, functional circuits are assembled from sibling cells that share a somatogenetic heritage. Thus, an entire neuronal circuit, or an anatomically defined structure, or the majority of cells with a specific neuronal phenotype, may exhibit the same gene alterations originating from a common precursor cell. Although the technology and skills exist today to dissect out such structures, and even to purify specific cell populations, from which DNA could be extracted, one would ultimately prefer to analyze the entire genome of individual neurons—but this is beyond current capability. Deep sequencing, which can generate many hundreds of reads through each gene, allows the application of statistical tools to assist in the identification of mutations that exist in only a relatively small number of neurons within a sampled brain structure.
All we need for this approach to work is a few good brains. While a normal brain would suffice, assuming it is no longer of use to its former owner, this would only allow detection of the general rate of somatic mutation in the nervous system without reference to disease. Ideally, it would be best to investigate both normal brain tissues and brain tissues from individuals with clearly documented diseases, akin to what is now being done with cancer genomes. Multiple samples from distinct brain regions, some representing specific cell lineages, would be best compared in parallel to constitutional DNAs from the same individuals, to document their personal neuronal somatic genome. In some disease states, certain neuroanatomic features have already been described that could serve as a focus for sequence analysis.
The information gleaned from deconstructing the somatic brain could be a “game changer” in our understanding of so many uniquely human, devastating conditions. All it takes is the will (and, of course, a substantial commitment of funding).
Tom Curran is deputy scientific director of the Children’s Hospital of Philadelphia Research Institute and professor of pathology and laboratory medicine at the Perelman School of Medicine at the University of Pennsylvania. Currently, the Curran laboratory studies brain development and pediatric brain tumors.