Intelligence genetics: Motorways

 

It is a curious twist of fate that just as the  M3 motorway out of London to the South West is being converted from its formerly perfectly functional state to a “Smart Motorway”, an international team have identified two gene motorways, including an 150 gene long M3 which seems to have an important role to play in cognitive ability.

The long list of authors seems to include one or two whose deep knowlegde of the actual motorway matches my own, and their calling the convergent gene network the M3 is ironic, since the smarter the constructors try to make the motorway, the more impassable it gets. Of course, in the case of the brain,nature got there first, and did a much better job, so the gene network M3 seems to run very well.

Systems genetics identifies a convergent gene network for cognition and neurodevelopmental disease

Michael R Johnson, Kirill Shkura, Sarah R Langley, Andree Delahaye-Duriez, Prashant Srivastava,  W David Hill, Owen J L Rackham, Gail Davies, Sarah E Harris, Aida Moreno-Moral,  Maxime Rotival, Doug Speed, Slavé Petrovski, Anaïs Katz, Caroline Hayward, David J Porteous, Blair H Smith, Sandosh Padmanabhan, Lynne J Hocking, John M Starr, David C Liewald,  Alessia Visconti, Mario Falchi, Leonardo Bottolo, Tiziana Rossetti, Bénédicte Danis,  Manuela Mazzuferi, Patrik Foerch, Alexander Grote, Christoph Helmstaedter, Albert J Becker,  Rafal M Kaminski, Ian J Deary & Enrico Petretto

https://drive.google.com/file/d/0B3c4TxciNeJZMGZyNFRjVW1PRTQ/view?usp=sharing

Genetic determinants of cognition are poorly characterized, and their relationship to genes that confer risk for neurodevelopmental disease is unclear. Here we performed a systems-level analysis of genome-wide gene expression data to infer gene-regulatory networks conserved across species and brain regions. Two of these networks, M1 and M3, showed replicable enrichment for common genetic variants underlying healthy human cognitive abilities, including memory. Using exome sequence data from 6,871 trios, we found that M3 genes were also enriched for mutations ascertained from patients with neurodevelopmental  disease generally, and intellectual disability and epileptic encephalopathy in particular. M3 consists of 150 genes whose expression is tightly developmentally regulated, but which are collectively poorly annotated for known functional pathways.  These results illustrate how systems-level analyses can reveal previously unappreciated relationships between neurodevelopmental disease–associated genes in the developed human brain, and provide empirical support for a convergent gene-regulatory network influencing cognition and neurodevelopmental disease.

Unusually, I am going to quote the opening paragraphs verbatim, because they serve as a very brief introduction to the entire field.

Cognition refers to human mental abilities such as memory, attention, processing speed, reasoning and executive function. Performance on cognitive tasks varies between individuals, and is highly heritable1 and polygenic. To date, however, progress in identifying molecular genetic contributions to healthy human cognitive abilities has been limited. A distinction can be made between cognitive domains such as the ability to apply acquired knowledge and learned skills (so-called crystallized abilities), and fluid cognitive abilities such as the capacity to establish new memories, reason in novel situations or perform cognitive tasks accurately and quickly. Within individuals, performance on different measures of cognitive ability tends to be positively correlated such that people who do well in one domain, such as memory, tend to do well in other domains. Seemingly disparate domains of cognitive ability also show high levels of genetic correlation in twin studies, typically in excess of 0.6, and analyses using genomewide similarity between unrelated individuals (genome-wide complex trait analysis) has also demonstrated substantial genetic correlation between diverse cognitive and learning abilities. These studies suggest genes that influence human cognition may exert pleiotropic effects across diverse cognitive domains, such that genes regulating one cognitive ability might influence other cognitive abilities. As impairment of cognitive function is a core clinical feature of many neurodevelopmental diseases including schizophrenia, autism, epilepsy and intellectual disability (by definition), we sought to investigate gene-regulatory networks for human cognition and to determine their relationship to neurodevelopmental disease.

The authors have used a fascinating, and to me very complex technique: Gene coexpression network analysis. This is based on taking, in this case, 122 frozen swhole human hippocampus surgical samples and showing the clustering of coexpressed genes (modules). To use my motorway example, it would be like looking at the way in which bulldozers and cement mixers had been built and utilised to construct many motorways, and to distinguish that from the way those tools had been utilized in building other objects. In that way one determines the plan for the building of motorways, necessarily involving the network of associations which are the essence of the construction plan.

Having made corrections for those patients with epilepsy, of the four modules conserved in healthy hippocampi across species (M1, M3, M11 and M19), only M1 (n = 1,148 genes) and M3 (n = 150 genes) were enriched for functional categories explicitly related to synaptic processes.

These 4 cross-species “motorways” were then tested on cognitive data from human subjects in two cohorts of cognitively healthy subjects, using 6732 subjects to discover any associations, and another 1003 subjects from anther cohort to test them. In other words, they have not followed the usual psychological publication pattern, in which the authors discover an association, publish it, and ask others to replicate it (knowing that they probably will find it difficult to do so). In Psychology this gives authors a publication which become famous, is frequently quoted, and takes on a life of its own even when years later an eventual repication fails to confirm the original results. Here we have psychologists and geneticists measuring the replicability of their finding in one paper.

We observed that module M3 was strongly and specifically enriched for genes that, when mutated, are associated with intellectual disability and epileptic encephalopathy, and that this enrichment holds true for both pdDNM intellectual disability; epileptic encephalopathy; and all nsDNM intellectual disability.

Using a stepwise procedure we prioritized gene networks whose gene coexpression relationships were significantly reproducible across brain regions and species to facilitate the identification of functionally conserved and replicable networks. We demonstrated replicable association between two of these coexpression networks (M1 and M3) and healthy human cognitive abilities. As M1 is functionally enriched for genes involved in synaptic processes, these findings provide systems-level evidence for a relationship between long-term potentiation and postsynaptic processes and human cognition, as previously suggested by an analysis of known postsynaptic signaling complexes5. In contrast to the functional specialization of M1, M3 is relatively poorly annotated for known functional categories or canonical pathways and our study revealed previously unappreciated coexpression relationships between genes influencing cognitive abilities. The finding that M1 and M3 influence cognitive abilities generally (as opposed to influencing specific cognitive domains such as memory) is in agreement with the evidence from twin and genome-wide complex trait analysis demonstrating high genetic correlation between diverse cognitive and learning abilities9,10,48. The widespread expression and coexpression of M1 and M3 genes across the human cortex, and their tight developmental regulation, is also consistent with these modules playing a role across cognitive domains.

The genes in M1 and M3 are stable through the life course.  When considering common risk variants (that is, SNPs) for neuropsychiatric disease, we observed an association between M3 and schizophrenia but not with common forms of epilepsy.

In sum, this is a novel approach, and one which is likely to launch further searches for convergent gene networks. I had imagined that the new approaches would be looking at some theory of protein formation, and had not understood that particular brain regions might offer important clues. Looking at the revealed associations, in terms of clinical implications, it confirms the link between schizophrenia and cognitive decline, and may eventually explain how problems with networks are an important part of the cause of such decline.

And now, back to the actual M3 motorway, which will continue to be blocked for most of this year, the most noxious experiment in raising the IQ of a motorway ever suffered by motorists.