Thread by @tuuliel

Happy to share our new preprint, led by @silvakasela & UCSF's Stephanie Christenson! We analyzed genetic and non-genetic factors affecting the expression of COVID-19 relevant genes in the large airway epithelium, leveraging respiratory disease cohorts.🧵: medrxiv.org/content/10.110…

Back in March, we realized with lung disease collaborators that we have very relevant data on our hands: bronchial epithelium RNA-seq from lots of pre-COVID samples from 3 cohorts. Could this inform how COVID-19 risk factors affect molecular and cellular processes of infection?

We discovered that expression of ACE2 increases with obesity & hypertension - known COVID risk factors - in addition to previously known smoking. Reassuringly, we confirm that ACE2 increase with inflammation is driven by a truncated, non-binding isoform, as recently reported.

We find little evidence that the risk of age and male sex have anything to do with ACE2 expression. This makes sense: there are a lot of genes and biological processes involved in SARS-CoV-2 infection, so just focusing on ACE2 isn't enough. Thus, we turned to pathway analyses.

Earlier work has shown that SARS-CoV-2 infection, compared to other viruses, has a suppressed airway immune response. We were stunned to see how strikingly similar that was to the expression pattern linked to obesity, hypertension & cardiovascular disease in our data!

This suggests a hypothesis where these risk factors may create a COVID-19-susceptible airway environment that is stunting early anti-viral host responses and allows SARS-CoV-2 to gain a foothold before the human immune system can respond. More research on this is needed!

Finally, we mapped regulatory genetic variants in bronchial epithelium for COVID-19-related genes in humans, in order to find potential genetic risk factors and also identify if e.g. important drug target genes are affected by genetic variants.

We find many genes where the genetic effect on gene expression (x-axis) is ~similar to expression change induced by SARS-CoV-2 infection (y). E.g. the MERS receptor DPP4 has a regulatory variant that decreases its expression by 3.3-fold. So don't forget genetics!

We pinpoint several loci with genetic variants that affect the expression of SARS-CoV-2-interacting genes and also immunological/respiratory traits in pheWAS. This is interesting, although COVID-19 GWAS data is still too sparse to say if these affect COVID-19 risk.

My favorite gene of these is probably ERMP1: it’s known to interact with the SARS-CoV-2 protein Orf9cm and we show that a regulatory variant affecting the expression ERMP1 in bronch epithelium is associated to asthma. An interesting gene and variant to look into.

So that’s (one of) our contributions to COVID research, leveraging the tools, skills and data sets that we had. Being a basic researcher, it was also very inspiring to work closely with clinicians during this time. Many thanks to all co-authors, especially Silva and Steph!