A multi-omic approach to investigating biological mechanisms related to the aetiology and treatment of major depressive disorder.

Major depression is a brain disorder that affects 350 million people worldwide. Depressed patients experience extreme periods of low mood, loss of pleasure, problems with their appetite and difficulties sleeping. Currently, antidepressants are the first line of treatment for depression but only one third of patients respond to the first prescribed antidepressant, and one third will not respond to multiple forms of treatment. Consequently, a better understanding of what causes major depression and what affects response to antidepressants is desperately needed.

At least three biological mechanisms have been proposed to be important for causing major depression. Firstly, depressed patients tend to have shorter telomeres in their blood cells than non-depressed subjects. Telomeres are stretches of DNA at the end of each chromosome that shorten in length with our normal aging. It has been hypothesized that depressed patients have some cells which demonstrate advanced 'cellular aging' and die more readily. This could have subsequent impacts on the viability of neuronal cells and brain function.

Secondly, depressed patients have higher levels of a protein called interleukin-6 in their blood. Interleukin-6 is classically involved in inflammation, but recent research suggests it interferes with brain communication and function.

Thirdly, depression has been linked with abnormal function in the hippocampus, a structure in the brain which is important in regulating memory and mood. Throughout our lifetime stem cells in the hippocampus grow and generate new neurons in a continuous manner. However, it has been suggested that there is a reduction in the generation of these new neurons in depressed cases, based on evidence from animal models. Complementing this theory, antidepressants stimulate the generation of new neurons in the hippocampus. However, it remains unknown whether this is the true mechanism behind antidepressant action.

Using new statistical techniques on large clinical genetic datasets we aim to test whether each of our three biological mechanisms are really causal to major depression, or whether they instead, represent an effect of having the disease. Broadly, we will achieve this using a three-step plan. Firstly, we will investigate which genes are involved in each mechanism (e.g. which genes affect telomere length). Secondly, we will combine the most significant genes affecting each mechanism into a single genetic signature. Thirdly, we will test whether genetic signatures for each mechanism predict an increased risk of major depression. A genetic signature associated with a mechanism, which additionally predicts major depression, would support the notion that this mechanism is causal. We will also test whether the genetic signature that favours generation of new neurons in the hippocampus additionally predicts patient response to antidepressants. Such an association would provide pivotal evidence towards understanding antidepressant mode of action.

The two main benefits from this research include: (i) improving our understanding of which mechanisms are truly causal to major depression, which would allow for more focused research in the future and the development of preventative strategies; (ii) improving our understanding of what factors moderate antidepressant response, which may allow for the stratification of patient treatment, and the development of "add on" drugs for those genetically predisposed to respond poorly to antidepressants. Improvements in each of these two areas could save the lives of depressed patients who are at risk of suicide by providing interventions or effective treatment faster, Moreover, it could save the UK millions of pounds wasted on ineffective treatments and loss of working hours.

We will investigate three biological processes associated with major depression disorder (MDD), to discern whether they are causally related, these include: (i) shorter leukocyte telomere lengths; (ii) increased circulating interleukin-6 (IL-6) levels; and (iii) reduced hippocampal neurogenesis (HN). Furthermore, we plan to investigate biological mechanisms mediating antidepressant response. We will achieve this by considering the following aims:

(A) The heritability of (i) telomere length and (ii) interleukin-6 levels, and its genetic correlation with MDD. By performing a GWAS and applying genomic-relatedness-matrix restricted maximum likelihood analysis, we will determine: (a) heritability estimates for each biomarker; (b) whether genes responsible for moderating (i) telomere length, or (ii) IL-6 levels, correlate with genes predicting MDD case/control status.

(B) Generation of robust polygenic risk scores for (i) leukocyte telomere length and (ii) IL-6 levels. This will involve generation of polygenic risk scores in large GWAS data 'training cohorts' and validating them in other independent 'test cohorts'.

(C) To test whether genes predictive of (i) leukocyte telomere length and (ii) IL-6 levels are causally related to MDD. We will apply polygenic risk scores generated in (B), and use Mendelian randomization methods to test if these mechanisms are causally related to MDD.

(D) To determine if hippocampal neurogenesis is involved in the pathophysiology of MDD. Through the application of WGCNA to expression data we will determine gene sets driving hippocampal neurogenesis. We will then use genetic pathway analysis to test if genetic differences within this gene set are enriched for associative signal within MDD GWAS data.

(E) To determine mechanisms underlying antidepressant response. We will apply genetic pathway analysis to antidepressant pharmacogenetic GWAS data, and test for enriched GO terms, and gene sets relating to antidepressant-induced HN.

Impact Summary
I foresee the research performed in this fellowship benefiting four sectors of society: scientists, clinicians, drug companies, and major depressed patients. Initially, those working closest to the coalface will benefit, i.e. the geneticists. Psychiatric geneticists will benefit because results will directly influence future work exploring which mechanisms underlie the pathophysiology of major depression. For instance, a high genetic correlation between IL-6 levels and major depression case/control status obtained from bivariate GCTA, will encourage future work exploring inflammation-based mechanisms. The generation of polygenic risk scores for telomere length and IL-6 will have widespread applications, as these biomarkers have been implicated in other disease states. Therefore, scientists outside of the field of psychiatry can use these results to explore whether genes predictive of these biomarkers are also predictive of their disease states of interest.

In the future (5+ years), I believe clinicians and drug companies may benefit from the downstream research results from this fellowship. Specifically, I think results from genetic pathway analyses investigating which biological mechanisms moderate antidepressant response will be beneficial. Clinician's could use genetic testing on major depressed patients to stratify treatment, e.g. by determining those patients with an excess of "non-response" polymorphisms within pathways known to moderate antidepressant outcome, and providing them with an alternative therapy. Furthermore, elucidation of mechanisms governing antidepressant response (e.g. hippocampal neurogenesis) will help drug companies to develop mechanistically more precise treatments, and adjuvant therapies designed for patients genetically predisposed to poor response.

The eventual goal is that this work will benefit major depressed patients (5+ years). As described in my Case of Support, current treatment for major depression is quite strikingly poor. Only one third of patients will respond to the first antidepressant they are prescribed, and one third will fail to respond to multiple forms of treatment. So the outcomes this fellowship will have on clinicians and drug companies will ultimately benefit patients, allowing them to receive more effective treatment from the outset. This would subsequently save lives through reducing suicidal ideation faster and more effectively; and save the UK millions of pounds each year, wasted on ineffective treatment and loss of working hours.