Introduction

Major depressive disorder is a common debilitating psychiatric illness affecting more than 350 million people worldwide. It is also one of the significant contributors to the overall global burden of the disease, accounting for 8.2% of global years lived with disability, the highest among the brain diseases [1, 2]. According to the World Health Organization (WHO), approximately 8 million people end up their lives due to depression after failing to cope with the social and economic stress resulting from the mental condition. Only half of the patients treated with antidepressants are able to achieve remission upon completion with at least two different treatments [3], making the management of the condition even more challenging and a top public health priority. However, the development of novel drugs and interventions has been difficult, which could be attributed to our limited understanding of underlying biological pathways. Recently, screening of single nucleotide polymorphisms (SNP) at the genome level, also called genome-wide association studies (GWAS), in an epidemiological setup have evolved as a popular choice to understand the genetics of depression. It is hoped that the identification of genetic loci through such large scale population studies would help us identify potential susceptible genes, which would subsequently map to relevant biological pathways.

MDD is a complex disorder resulting from the interaction of genetic and environmental factors. Owing to this underlying complexity, the identification of the causal genetic loci through GWAS has been a significant roadblock [4]. A GWAS rely upon testing differences in allele frequencies of millions of SNPs between diseased individuals and healthy individuals from a sample belonging to a population with a shared ethnic background such as an Asian or a Caucasian population. Several large consortia comprising of worldwide MDD populations already exist, including 23andMe, China, Oxford and Virginia Commonwealth University Experimental Research on Genetic Epidemiology (CONVERGE), Psychiatric Genomics Consortium (PGC), Cohorts for Heart and Aging Research in Genomic Epidemiology (CHARGE), and UK Biobank (UKB)  [5-9]. Each consortium often comprises 1000s of MDD cases and healthy controls and have already made considerable progress by conducting meta-analyses of respective GWAS in each of their different population cohorts. However, despite their large pooled sample sizes, the findings of their meta-analyses have been highly heterogeneous, with limited replication across cohorts. This article discusses the reported findings and possible reasons for our continued need to identify the causal risk loci for MDD.

Major Genetic Findings

23andMe and CONVERGE

The study by Hyde et al. collected data on 75,607 self-reported cases and 231,745 controls of

European ancestry (n=307,354) [8]. Their analysis involving approximately 9 million autosomal SNPs lead to the identification of 17 significant genetic variants spanning 15 genetic loci. These loci included the MEF2C gene with a known role in the regulation of synaptic functioning. Notably, the study was able to replicate LHPP and SIRTI loci, previously reported in low coverage whole genome sequencing study employing 5303 Han Chinese women with recurrent MDD, also popularly known as the CONVERGE study. While SIRT1 loci is involved in mitochondrial dysfunction, the functional role of LHPP and its involvement in MDD is unknown [9].

            UKB

A later study by Howard et al. identified 17 genetic loci (15 novel) among 9.3 million autosomal variants genotyped in 121,380 cases and 338,301 controls of European ancestry from UK [10]. The cases comprised of individuals categorized into three different phenotypes, including broad-depression (anxiety, symptom or depression), probable MDD (self-identified) and MDD (hospital records). The study further mapped the discovered loci to pathways involved in excitatory neurotransmission, dendritic functions, mechanosensory behaviour, post synapse, and neuron spine functions.

PGC

The study by Wray et al. identified 44 loci (30 novel) associated with a broad spectrum of depression phenotypes among 130,664 cases and 330,470 controls of European ancestry [11]. The study-specific highlighted the role of their loci with gene expression in the prefrontal and anterior cingulate cortex. Furthermore, for the first time, the authors implicated development gene regulatory processes in the pathophysiology of depression. Lastly, the study also suggested an overlap of the shared mechanism of MDD with several psychiatric disorders.

Pooled PGC, 23andMe, and UKB

Most recently, study by Howard et al. conducted largest GWAS to date by pooling summary statistics provided by PGC, 23andMe, and UKB [12].  The pooling resulted in a total of 807,553 individuals in the discovery cohort (246,363 cases and 561,190 controls) and 1,306,354 individuals in the replication cohort (414,055 cases and 892,299 controls). The meta-analysis led to the identification of 102 genetic variants from 269 genes, mostly mapping to pathways involved in the regulation of synaptic structure and neurotransmission. The study further implicated the prefrontal brain region for involvement in the aetiology of MDD.

Major Challenges in the elucidation of causal variants

Polygenicity and Low heritability

Using the polygenic risk approach (PRS), the large pooled sample sizes of different consortia have also revealed a highly polygenic architecture of MDD, i.e. MDD results from the cumulative effect of several genetic variants, with each variant contributing a small effect. This polygenicity is reflected in the high SNP heritability reported by some of the recent reports. The SNP heritability, which is also defined as the proportion of variance in the trait, i.e. MDD phenotype, that could be attributed to common SNPs, was as low as 0.21 in a recent finding reported by PGC. Twin studies further demonstrated modest heritability estimates of 28-37% for MDD, suggesting that most risk alleles likely show small effect sizes.  The polygenic nature of MDD and low heritability necessitates a need for higher sample size studies in the existing consortia that would enable consortia to conduct well-powered meta-analyses of GWAS to detect association of genetic variants with small effect estimates at a highly conservative genome-wide significance level of 5 x 10^-8 [13].

High phenotypic heterogeneity

Another reason could be the highly heterogeneous nature of MDD. Although MDD comprises nine characteristic symptoms, more than 1000 underlying symptoms or sub-phenotypes, which are a part of diagnostic criteria, makes it highly unlikely to have patients with similar sub-phenotypes [14]. Furthermore, the possibility of different biological pathways or gene sets contributing to the symptom profile of each MDD patient cannot be ruled out.

Existence of confounding variables

A multivariable analysis adjusting for confounding traits such as age and gender provides an unbiased estimate of different genetic variants contributing to disease susceptibility. Most GWAS adjust for these variables along with principle components accounting for variability in genetic ancestry. The confounding traits such as gender often show opposite effect estimates in the association analysis with causal variants for MDD, making multivariable analysis a wrong approach in such a scenario. In such a scenario, it is critical to report stratified analysis by gender. However, the lack of adequate sample sizes to conduct a stratified analysis by traits such as gender poses a significant limitation in uncovering novel genetic loci underlying MDD.

Expensive downstream functional analysis

Another reason for our failure to identify causal variants could be that GWAS studies flag potential loci represented by a top variant. The top identified variant may not be directly causal to MDD. Identification of a causal variant would require sequencing of the given region followed by conduct of functional studies, both of which are highly expensive and time-consuming.

Existence of pleiotropic variants

It is also now well established that most of the genetic variants discovered through GWAS are pleiotropic, i.e. they influence several phenotypes at a time. In fact, several recent reports have shown a substantial overlap between MDD and various other psychiatric and somatic conditions [15]. Although new approaches such as Mendelian randomization have been applied to overcome the handicap of pleiotropic SNPs, the methodology itself has several inherent assumptions, which are challenging to implement in the real world [16].

Conclusion

Considerable progress has been made in recent years in identifying genetic loci increasing susceptibility to MDD. However, the challenge still remains to identify and functionally validate the causal genetic variants. It is unclear how and when the genetic findings would lead to pinpointing specific biological pathways that could be targeted to develop personalized treatment.

References

1.         Whiteford, H.A., et al., The global burden of mental, neurological and substance use disorders: an analysis from the Global Burden of Disease Study 2010. PLoS One, 2015. 10(2): p. e0116820.

2.         Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990-2015: a systematic analysis for the Global Burden of Disease Study 2015. Lancet, 2016. 388(10053): p. 1545-1602.

3.         Rush, A.J., et al., Acute and longer-term outcomes in depressed outpatients requiring one or several treatment steps: a STAR*D report. Am J Psychiatry, 2006. 163(11): p. 1905-17.

4.         Sullivan, P.F., M.C. Neale, and K.S. Kendler, Genetic epidemiology of major depression: review and meta-analysis. Am J Psychiatry, 2000. 157(10): p. 1552-62.

5.         de Moor, M.H., et al., Meta-analysis of Genome-wide Association Studies for Neuroticism, and the Polygenic Association With Major Depressive Disorder. JAMA Psychiatry, 2015. 72(7): p. 642-50.

6.         Hek, K., et al., A genome-wide association study of depressive symptoms. Biol Psychiatry, 2013. 73(7): p. 667-78.

7.         Fabbri, C., et al., Genetic and clinical characteristics of treatment-resistant depression using primary care records in two UK cohorts. Mol Psychiatry, 2021.

8.         Hyde, C.L., et al., Identification of 15 genetic loci associated with risk of major depression in individuals of European descent. Nat Genet, 2016. 48(9): p. 1031-6.

9.         consortium, C., Sparse whole-genome sequencing identifies two loci for major depressive disorder. Nature, 2015. 523(7562): p. 588-591.

10.       Howard, D.M., et al., Genome-wide association study of depression phenotypes in UK Biobank identifies variants in excitatory synaptic pathways. Nat Commun, 2018. 9(1): p. 1470.

11.       Wray, N.R., et al., Genome-wide association analyses identify 44 risk variants and refine the genetic architecture of major depression. Nat Genet, 2018. 50(5): p. 668-681.

12.       Howard, D.M., et al., Genome-wide meta-analysis of depression identifies 102 independent variants and highlights the importance of the prefrontal brain regions. Nat Neurosci, 2019. 22(3): p. 343-352.

13.       Nishino, J., et al., Sample Size for Successful Genome-Wide Association Study of Major Depressive Disorder. Front Genet, 2018. 9: p. 227.

14.       Ostergaard, S.D., S.O. Jensen, and P. Bech, The heterogeneity of the depressive syndrome: when numbers get serious. Acta Psychiatr Scand, 2011. 124(6): p. 495-6.

15.       Polushina, T., et al., Identification of pleiotropy at the gene level between psychiatric disorders and related traits. Transl Psychiatry, 2021. 11(1): p. 410.

16.       Yang, H., et al., Mendelian randomization integrating GWAS and eQTL data revealed genes pleiotropically associated with major depressive disorder. Transl Psychiatry, 2021. 11(1): p. 225.