NEW YORK (GenomeWeb) – Genetic variability within a certain metabolic signaling pathway gene may contribute to the abnormal brain development seen among some preterm infants.
About 11 percent of all births are preterm and more than 30 percent of survivors experience neurocognitive issues during their life, such as anxiety, social, and communication problems, and attention difficulties. Imaging studies have linked these adverse outcomes with changes in brain structure, connectivity, and function.
In a combined imaging and genetic analysis, researchers from King's College London took an unbiased machine-learning approach to link brain alterations observed in affected individuals to their SNP genotypes. As they reported in the Proceedings of the National Academy of Sciences today, the researchers connected variability in the peroxisome proliferation activated receptor (PPARG) gene to abnormal development of white matter in the brains of preterm infants.
"The present results are consistent with the hypothesis that changes in white-matter structure that predict adverse outcome are modulated by genetic variability in the PPAR signaling pathway," KCL's David Edwards and his colleagues wrote in their paper.
The researchers studied a cohort of 272 infants who were born at less than 33 weeks of gestational age and for whom diffuse MRI scans and genomic DNA were available. Within this cohort, they searched for genetic variations associated with the brain endophenotype.
Using the machine-learning Sparse Reduced Rank Regression (sRRR) model, the researchers ranked white-matter tracts within the imaging data to develop a consistent endophenotype present among the cohort. An sRRR approach allows the investigation of cohorts in which the number of individuals is smaller than the number of features examined, the researchers said.
The brain tracts linked to the phenotype are involved in information flow, the researchers said, and often are connected to the insula.
With the same sRRR approach, the researchers linked SNPs found among the preterm infants in the cohort to the image features they identified. They fitted a predictive model for the phenotype using all the SNPs and then ranked the SNPs based on how well they predicted the phenotype.
The top 100 SNPs they ranked could be traced to 47 genes. Though many of these genes were in linkage equilibrium with one another, the researchers noted three hotspots of linkage disequilibrium near the PPARG, ITGA6, and FXR1 genes.
SNPs within PPARG were overrepresented among the top SNPs linked to the brain endophenotype. The six PPARG SNPs they uncovered were largely located in introns or regulatory regions, the researchers reported.
A Gene Ontology-based analysis of those 47 genes found that processes like lipid metabolism, neuron projection regeneration, and response to nerve growth factor stimulation were overrepresented among that group. Additionally, the researchers noted that the genes linked to the top 100 SNPs were further overrepresented for a role in lipid metabolism.
These genes have also been linked in the literature to autism spectrum disorder, intellectual disability, and schizophrenia, conditions the researchers noted are more common among preterm individuals.
This suggested to the researchers that genetic variability within PPARG influences brain development in preterm infants. PPARG, they noted, is known to be involved in lipid metabolism, its expression is up-regulated in neurons following excitotoxicity and ischemia, and it modulates the microglial response to injury.
Drugs targeting PPARG are already available, and the researchers said they could be explored as a way to protect preterm infants from developing these brain changes.
They also noted that more research into the roles of the ITGA6 and FXR1 genes is needed.