Abstract
The cerebral cortex is the principal region of the mammalian brain that controls complex cognitive behaviours. Across evolutionary time, a great expansion of the cortex – particularly in humans - has taken place with a concomitant increase in regulatory complexity of the genome. A consequence of these rapid expansive events could be a trade-off that has rendered the brain particularly susceptible to certain disorders. Disorders of neuronal mispositioning during brain development are phenotypically heterogeneous and characterised clinically by epilepsy and intellectual disability. Periventricular neuronal heterotopia (PH) is one such disorder where neurons fail to populate the outer cortex of the brain resulting in their heterotopic positioning along their sites of origin – the lateral ventricles. Currently, only 25% of sporadic instances of the disease have a definable molecular genetic cause, leaving the vast majority of patients without a genetic explanation for their condition. This gap in understanding also points towards a limitation on our current knowledge of normal human brain development.
In this study, I sought to characterise the contribution of rare genetic variants towards the causation of PH. Insights from such studies were hypothesised to hold the potential to indicate recent evolutionary dynamic events that could pose a vulnerability to such conditions. Given the phenotypic heterogeneity of PH, the predominantly sporadic nature of the disease and previous work on other neuro-developmental disorders, it was predicted that many genes might underlie its aetiology, limiting the use of genetic evidence alone to assess causality. Thus in this study, three main approaches were employed. First, an unbiased next-generation sequencing study, screening only the coding regions of the genome (the exome) for rare candidate variants in genes not currently implicated in the condition was performed in sporadic cases of the disease. Using this whole-exome sequencing approach on 65 probands and their parents, 50 rare, coding, de novo variants were identified. No two patients were mutated at the same locus across the entire cohort, confirming the extreme genetic heterogeneity underlying this condition. A likelihood analyses, accounting for gene size and mutability, did however, identify a significant excess of de novo variants in loci most intolerant to functional change (P = 1.28x10-12). Thus I was able to estimate that 28% of the de novo variants identified in this study contribute to the causation of the PH phenotype (95% CI: 16% – 42%). Specifically focusing on a set of genes enriched for activity within a specific human neural stem cell transcriptional network, critical for governing recent mammalian brain evolution, I identify a significant excess (P = 0.024) of de novo variants preferentially occurring in these loci.
Secondly, I undertook a range of in vitro and in vivo functional studies to investigate candidate loci that, although short for proof-of-pathogenicity on genetics grounds alone, warranted further investigation as a potential novel gene involved into mammalian (i.e MOB2), or primate (i.e PLEKHG6), neurogenesis.
Finally, I employed a framework that exploited both brain-specific transcriptional networks and current molecular knowledge of PH to organise the information generated from the whole-exome sequencing screen into co-regulated gene sets that present clues into novel avenues that potentially could mediate the pathogenesis of PH. The 14 genes prioritised using this approach are significantly more conserved and under greater purifying selection than the original set of 50 genes identified with de novo variants in the exome study (P = 0.019). These 14 candidates are also significantly enriched for RNA-processing factors (P = 0.011), pointing to relevant pathways for future study into the aetiology of PH. In addition, I focused on two genes that exhibited qualities that made them very promising loci for further consideration as genes that when mutated could lead to PH. One of these genes, NEDD4L, has recently been independently validated as a novel contributor to PH aetiology and underscores the validity of this analytical approach. Future work, relying on mutational recurrence is needed to further investigate the relevance of the second gene prioritised here – LRRK2 – to PH aetiology.
Results from this study outline the contribution of rare, coding, de novo variation into the underlying genetic architecture of PH. A model suggesting PH as a genetically heterogeneous condition, with an apparent aetiological role intrinsically related to recently evolved gene regulatory networks that govern neocortical expansion, is also proposed. Collectively, this study offers important lessons on the approach to understanding not only the contribution of genetics towards the causation of PH, but also the utility of exome-sequencing data derived from patients with disorders possibly resulting from evolutionary dynamic processes in identifying genes that have facilitated such events. This information is of importance for future studies aimed at modelling such conditions and identifying therapies.