Chad Weldy, MD, PhD

Super excited we made the cover of @natcardiovascres.nature.com! Work represents ADAR1 RNA editing within the vascular wall

Excited for a major milestone in our efforts to map enhancers and interpret variants in the human genome:

The E2G Portal! e2g.stanford.edu

This collates our predictions of enhancer-gene regulatory interactions across >1,600 cell types and tissues.

Uses cases 👇

1/

Hey thanks! Hope you’re doing well!

There's a lot here and a lot more in the paper. But I get excited thinking about the potential. From rare to complex disease to novel mechanisms with real potential for a precision guided approach to therapy. A lot to do! @stanfordmedicine.bsky.social @stanforddeptmed.bsky.social

To then connect this back to humans, entirely grateful for collaborators @clintomics.bsky.social & Sander van der Laan where we investigated ISG activation and SMC modulation and plaque phenotype in the Athero-Express cohort, showing distinct relationships between ISG induction and calcification

Importantly, we define the cellular trajectory of MDA5 activation leading to vascular calcification and disease progression, an effect that can be entirely inhibited with simply haploinsufficiency of MDA5.

This MDA5 activation leads to increased plaque size due to increased SMC migration into the plaque with markedly increased vascular calcification.

But importantly, we show that with SMC ADAR1 haploinsufficiency, atherosclerosis studies reveal that MDA5 activation occurs in a cell type and context specific mechanism. MDA5 activation drives a distinct SMC cell state change.

In atherosclerosis - we show that SMCs appear to be enriched for these immunogenic RNA, and that as SMC undergo phenotypic modulation in both human and mouse there is significant activation of ISG genes, potentially suggestive of MDA5 activation

With homozygous deletion of ADAR1 in SMC, there is a loss of vascular integrity. Further single cell RNA sequencing reveals distinct ISG activation and cellular infiltration with critical receptor ligand interaction

Our work here gets at this mechanism.

We reveal a fundamental observation, that vascular SMC have a unique requirement for ADAR1 editing to prevent MDA5 activation.

SMC deletion of ADAR1 leads to severe phenotype within days and is entirely blocked with deletion of MDA5

In addition, rare LOF variants in MDA5 (IFIH1) have been found to be protective against CAD as well as other inflammatory disorders. This provides quite strong human genetic evidence to support ADAR1-dsRNA-MDA5 axis in CAD, but through what mechanism?

The big finding in 2022 by my colleagues Qin Li (now @upenn.edu and Billy Li @stanforduniversity.bsky.social)
was that beyond rare disease, common variants appear to regulate RNA editing (edQTLs), and these edQTLs predict numerous common inflammatory disorders, including CAD! t.co/t1i47lPlGG

Amazingly, mice that are deficient in ADAR1 are embryonic lethal, but dual knock out of ADAR1 and MDA5 essentially rescues the phenotype. In this case, the role of ADAR1 seems to be nearly entirely based on preventing MDA5 activation, less so the actual edit of the transcript

In rare disease, loss of ADAR1 causes a severe interferonopathy due to the build up of dsRNA and activation of the dsRNA receptor MDA5 (gene symbol IFIH1). Similarly, gain of function variants in MDA5 (IFIH1) cause the same disorders, including severe vascular calcification

RNA has the peculiar pattern of having long repetitive elements on either end, where these strands fold over on each other to make double strand RNA structures -> turns out this looks a lot like a dsRNA virus!

So why doesn't this dsRNA induce an antiviral response? ADAR1!

When ADAR editing occurs in the coding region of a transcript, it serves as an A -> G edit and can change protein function.

Even in coral and octopus in response to temperature changes of the ocean, whoa!

Although amazingly, the majority of editing sites are non-coding (hmm)

What is RNA editing and how does this relate to coronary artery disease??

There's a lot here but it's fascinating.

A to I editing is an under appreciated area of biology, where ADAR enzymes deaminate adenosine to inosine. Thousands of RNA molecules are edited all the time!

Smooth muscle expression of RNA editing enzyme ADAR1 controls activation of the RNA sensor MDA5 in atherosclerosis Nature Cardiovascular Research - Weldy et al. show that smooth muscle expression of the RNA editing enzyme ADAR1 regulates activation of the double-stranded RNA sensor MDA5 in a novel mechanism of...

Hard to understate how wonderful it is to see our manuscript in print today @natcardiovascres.nature.com. We discover ADAR1 to control dsRNA sensor MDA5 in atherosclerosis, creating a new paradigm of endogenous dsRNA sensing as a causal mechanism of disease. Let's get into it 👇 #RNAsky rdcu.be/eGEyu

This work is exciting in that it defines an important area of vascular biology with key relevance to understanding genetic drivers of disease risk, couldn't have been done with out the amazing support of Tom Quertermous and all our amazing collaborators and team @stanfordmedicine.bsky.social

Through ChromBPNet analysis, by identifying the variants that affect chromatin accessibility in a vascular site specific manner, we identified that many of these variants land in key developmental TF motifs such as MEF2A, HAND2, as well as other regulatory TFs important in disease risk such as SMAD3

Not only can we reveal and predict variant effect on chromatin accessibility, but we define that effect varies by vascular site even within cell type

But how does this relate to human disease?? Through an awesome collaboration with the @anshulkundaje.bsky.social lab, we trained ChromBPNet models with scATACseq datasets for each cell type and vascular site, and predict human variant effect on a cell type/site basis @soumyakundu.bsky.social

Gene regulatory network analysis through integrated RNA and ATAC datasets across cell types and vascular sites reveal cell type and vascular site specific GRNs, this highlighted ascending fibroblast specific MEOX1

Vascular site specific epigenomic patterns are distinct for SMCs, fibroblasts, as well as endothelial cells, but importantly not macrophage cells. While developmental TFs a enriched there are thousands of distinct enhancer elements across vascular sites

Many enhancers correspond to developmental origin and highlight specific developmental transcription factors such as HAND2, GATA4, and HOX family members, suggestive of an epigenetic 'memory' of developmental origin

In our work, by performing single cell RNA and ATAC sequencing across different vascular sites in mice, we reveal that the epigenomic landscape is distinct to not only cell type, but vascular site, defining vascular site specific enhancers

A critical element of complex genetics is that the majority of GWAS SNPs that influence disease regulate non-coding enhancer elements in the genome — where variants can influence disease risk through regulating cell type specific enhancers, but what about for the vasculature??

However we know that the genetic drivers of complex vascular traits vary based on vascular site, a beautiful example is the exceptional work by @jamespirruccello.com in 2022 showing the different genetic variants that impact ascending versus descending aortic dimension

Beautiful work in developmental biology going back decades has revealed that vascular diversity has a developmental basis, and that these vascular territories have distinct biology