Shared and distinct genetic architecture of autoimmune and hormonal alopecias

Hair loss can be very debilitating for patient wellbeing and most hair loss disorders are neither well understood, nor well-treated.

This research study is investigating autoimmune, scarring and non-scarring hair loss:
- frontal fibrosing alopecia (FFA), an increasingly common scarring and inflammatory autoimmune condition affecting almost exclusively women.
- alopecia areata (AA), which is the commonest human autoimmune disease and a non-scarring form of hair loss affecting men, women and children; and
- androgenetic alopecia (AGA), which is due to hormones and genetics and commonly affects men and women.
In FFA, skin inflammation starts from the scalp, which is affected by intense burning and itching. The inflammation is very difficult to manage and the disease will forge ahead to affect the entire scalp, especially if not treated with strong drugs that suppress one's immune system. Women suffering from this condition are usually very distressed psychologically. Treatments for FFA on the whole are inadequate and very little was known about what causes the disease, prior to our original large-scale research study, where our team has discovered a number of genes that are strongly associated with FFA.

In AA, there is inflammation in deeper skin layers, which is not evident on the skin surface. The condition commonly causes patches of hair loss, which can often enlarge and coalesce, leading to complete scalp, eyebrow or universal scalp and body hair loss. Stress is a common attributing factor of AA, which is known to trigger and exacerbate the condition. A lot is known about the genetic mechanisms underpinning AA, which, in molecular terms, is also thought to be similar in many respects to FFA. Hormones are thought to play a role in the pathogenesis of both AA and FFA.

AGA is very common in men and women but also in adolescents. Widely thought of as of genetic and hormonal cause, the condition may present with thinning and hair margin recession in female and male pattern baldness. AGA often co-exists with AA and FFA and there is increasing evidence that hormonal factors play a role in both.
The proposed project will use modern statistical genetic approaches in a multi-disease analysis to try and resolve how these forms of alopecia correlate genetically. To maximise the ability of this approach, we will also evaluate sharing of genetic pathways with related conditions for which large scale genetic association studies are available, including systemic lupus erythematosus (SLE).

The proposed project promises to improve statistical power for the discovery of new genetic associations, improve our ability to predict risk of developing these diseases whilst also helpings us identify shared and distinct biological pathways of susceptibility. The latter has the potential to inform about the most suited treatment strategy for each, which could be by either repurposing existing drugs or discovering new ones.

Our team of researchers is based at King's College London and the academic partner in this research is Prof Michael Simpson.

We strongly believe that this research study will enhance our understanding of several basic disease processes and direct us towards developing better treatments, which alopecia sufferers are in so desparate need for.

Background: Genome-wide association studies (GWAS) have begun to reveal specific loci that contribute to the genetic susceptibility of common hair loss disorders, including androgenetic alopecia (AGA), alopecia areata (AA) and frontal fibrosing alopecia (FFA). Whilst there are analogous pathomechanisms underpinning these disorders, the exact nature of their shared and distinct biological processes remains unclear.

Research aim: The proposed project will utilize contemporary statistical genetic approaches in a multi-trait analysis to resolve the global and local genetic correlations between AGA, AA and FFA.

Methodological aspects
- The study will utilise GWAS summary statistics from hair loss traits including FFA, AA and AGA
- Genetic correlations and sample overlap estimates will be performed using LD Score regression analysis.
- Cross trait meta-analysis will be performed using the multi-trait analysis of GWAS (MTAG) framework that leverages cross trait correlations and enables correction procedures in the presence of overlapping samples.
- Loci will be partitioned into those with evidence of an effect on single traits or multiple traits, where effects are present in multiple traits the effect size, and directions across traits will be used to further discriminate.
- Fine-mapping of association signal will be performed using established Bayesian procedures, annotation of credible sets will be performed using the variant effect predictor.
- Co-localisation with eQTLs will be performed using Coloc on eQTLs and single cell deconvoluted eQTLs from hair follicles (in house), skin (GTEx and TWINSUK), blood (GTEx and TWINSUK) and endocrine tissues (GTEx).
- Relationships between implicated genes at associated loci will be established using databases of gene-gene relationships including stringDB; significance testing will be performed using a permutation framework.