Welcome

This website summarizes the main results from our preprint available on BioRxiv.

It was written by Chenfu Shi as an addition to the poster presented at the Wellcome Epigenomics of Common Diseases 2020 online conference to make our results clearer and more accessible.

Functional genomics of common diseases

Genome Wide Association Studies (GWAS) have linked a large number of genetic variants to psoriasis and other complex diseases. These variants however do not directly affect the coding region of genes but instead often lie in regulatory regions such as enhancers. These can that act through long range mechanisms and it may therefore be difficult to assign them to a specific gene. Moreover, regulatory regions are highly dynamic and context specific, requiring us to study them in disease relevant tissues and samples.

The objective in our lab is to use functional genomics techniques to uncover the underlying biological mechanisms that can contribute to disease risk.

In the figure below you can see an example of our general workflow:

  • An example GWAS loci contains multiple SNPs far away from genes making it difficult understanding their role in disease.
  • We apply a combination of techniques such as ChIP-seq, eQTL and chromatin conformation.
  • This results in a better understanding of the functional aspects underlying this loci such as identifying which genes are affected.

chromatin looping figure

GWAS SNPs affect elements that are highly cell type specific

The majority of GWAS SNPs has been associated with enhancer regions that are highly cell type and condition specific. This requires the study of these loci in cell types that are relevant to the traits studied.

For example dermtalogical conditions such as psoriasis are known to enrich in cell types such as keratinocytes.

SNP enrichment figure

Despite this most previous studies have focused on immune cell types derived from blood as these are simpler to isolate and study.

Here we used H3K27ac HiChIP, Hi-C and RNA-seq to study the global chromatin conformation and activity in cells derived from Human Keratinocytes (HaCaT) and a CD8 T cell line derived from a cancerous skin-plaque (MyLa).

Graphical Abstract

We used our datasets to link genes to disease associated loci from 5 dermatological conditions. This results in between 84 and 399 potentially linked genes identified for each disease studied. Below you can see that these genes strongly enriched for disease relevant pathways.

pathways figure

Using these datasets to investigate the underlying mechanisms in disease associated loci

We first looked at the ETS1 locus in particular. Here psoriasis and atopic dermatitis have distinct associations. However, whilst the psoriasis associated variants overlap the promoter of ETS1 directly, the association with atopic dermatitis is located 130 kb downstream of the gene.

Using our HiChIP interactions we can link the atopic dermatitis associated variants with the ETS1 promoter. This suggests a putative mechanism in which these two disting associations are mediated by a single gene.

ETS1 figure

Linking genes with psoriasis loci

We then focused on 4 psoriasis loci which show previously unknown candidate genes.

The first one is the SATB1 locus. Here we link the psoriasis SNP rs73178598 to SATB1 which is located 240kb away. This gene is an important regulator of regulatory T cells and autoimmunity.

SATB1 figure

The Ps loci indexed by SNP rs9504361 is intronic to EXOC2 and is typically associated with this gene. Here we show long-range interactions between this SNP and the promoters of IRF4 and DUSP22.

IRF4 figure

Another example of a novel gene target can be found at the 1p36 locus (rs10794648), which to-date has been associated with IFNLR1. However, our HiChIP data showed long-range interactions between the Ps SNPs at 1p36 and the distal gene GRHL3 primarily in HaCaT keratinocytes. GRHL3 encodes a transcription factor which stimulates keratinocyte migration and has previously been independently linked to psoriasis.

GRHL3 figure

Finally, the locus indexed by the Ps SNP rs73183592 with previously undescribed mechanism was linked via a long-range interaction spanning about 500 kb to FOXO1 a gene with important functions in regulatory T cells.

FOXO1 figure

Conclusion

Chromatin conformation and functional genomics studies have the potential to uncover the underlying mechanisms that drive the disease susceptibility of many complex traits. Here we used H3K27ac HiChIP, a novel technique that allows combined analysis of both chromatin conformation and chromatin activity, to create the first global study of promoter-enhancer interactions in keratinocyte and CD8+ T cell lines.

We use this data to explore disease associated SNPs for Psoriatic Arthritis, Psoriasis, Atopic Dermatitis, Melanoma and Systemic Sclerosis, and identify all the genes that are linked by chromatin interactions to these variants. All of these results are available on the bioRxiv preprint and further updates will come when our article is published.

The HiChIP gene targets have the potential to be used as therapeutic targets in drug repurposing and discovery, as recently applied for other diseases. However, whilst the techniques used in this study are an important strategy to identify potential candidate genes, they require further individual experimental validation (such as genetic perturbation techniques) before they can be used for drug targeting.

Regardless, we hope that these genes and mechanisms identified here represent a useful resource for further research aimed at characterising how genetic variation impacts on disease susceptibility for these and other complex diseases.

Thank you for reading!

This website was created by Chenfu Shi to display the poster presented at the Wellcome Virtual Conference “Epigenomics of Common Diseases” 2020.

For more information contact:  chenfu.shi@postgrad.manchester.ac.uk