Replication, fine-mapping and causal inference of genetic loci associated with major depressive disorder among participants of the Uganda genome resource.

Abstract

Background: Major depressive disorder (MDD) is the leading cause of disability worldwide, with significant consequences like poor health outcomes, premature death, and reduced productivity. Twin-based studies have shown that the heritability of MDD is approximately 50% meaning that 50% of the MDD onset is due to genetic factors. Genome-wide association studies (GWAS) have been conducted to understand the genetic architecture of MDD, but most of these have focused on participants of European ancestry. Some recent studies have included participants of African ancestry who are either African Americans or Africans in the United States and the United Kingdom. It is however not clear whether the genetic variants identified in these studies are relevant to African population since these participants represent only a small fraction of the African population. This has limited the generalizability of findings to Africa, Uganda inclusive. This study aimed at identifying and characterizing the genetic underpinnings of MDD among participants in the Uganda Genome Resource (UGR). Materials and Methods: A total of 980 participants (227 MDD cases and 753 controls) from the UGR cohort with genetic data available were randomly selected and assessed for MDD using the Mini International Neuropsychiatric Interview (version 5.0.0). The collected phenotype data for MDD was linked to the pre-existing genotype data. A GWAS was performed using genome-wide efficient mixed model association (GEMMA) software to find genomic loci that are associated with MDD. Fine mapping using the Bayesian approach was then performed to pinpoint the most likely causal variants for MDD. Replication of GWAS findings was done to validate findings and check for reproducibility. A two-sample bidirectional Mendelian randomization (MR) was then performed to determine any bidirectional causal relationship between six cardiometabolic traits (fasting glucose, low-density lipoprotein, triglycerides, high-density lipoprotein, systolic blood pressure and body mass index) and MDD. Results: The GWAS discovered 5 novel intronic genome-wide significant single nucleotide polymorphisms (SNPs) at P-value < 5x10-8 of which 4 were lead SNPs. The lead SNPs; rs1403411845, 16:10768301, 5:7561685 and rs1513848 were found near genes NHPH3-ASI, TEKT5, ADCY2 and FHOD3 respectively. The SNP rs1513848 (β = 0.11, se =0.02, P-value = 3.92E-08) was replicated in the European ancestry (β = 0.0164, se =0.0076, P-value = 0.03003). Fine mapping analysis identified SNPs rs1403411845 and 5:7561685 as having a posterior probability ≥ 99%. The Two-sample bidirectional MR revealed no bidirectional causal relationship between cardiometabolic traits and MDD. Conclusion: This study identified novel genetic variants / SNPs associated with MDD in a well-characterized general population cohort from Uganda. These SNPs have been mapped to genes which might be linked to biological processes like cilia dysregulation, cAMP signaling pathway dysregulation and cytoskeleton modifications. These findings provide insights into the biological mechanisms involved in the development of MDD among the UGR participants.