Weird. I can't seem to embed this image in a decent resolution at all. Fortunately this awesome new paper is open access!
Ever heard of a HERV? Me either. It stands for human endogenous retrovirus and these things make up almost 8% of the HUMAN GENOME. It actually starts to ring a bell now. I know that we have loads of viral insertion sites all over that we've accumulated throughout the 3,4,5 billion years of our evolution (new evidence I saw at an astronomy conference recently kept pushing the number back and I forget where it is right now. the whole thing made the speaker really enthusiastic because the earlier life started on our planet the more it affects this thing
Oh yeah! Okay so the human genome is huge and it is really convenient to ignore most of it. My friend Amol has been using his absurdly powerful Cloud computing platform (Bolt) to look at larger and larger parts of the genome and he's found protein and peptide evidence all over to support the fact that "noncoding" DNA/RNA is probably an overall dumb concept. We've been under a lot of evolutionary pressure these billion years why would we haul around billions of spare nucleic acids in every single one of our cells if they didn't do anything?
THIS group in THIS paper that I am typing about specifically THIS morning went directly after these viral HERV sequences to see if they might explain things we see with LCMS based approaches that we can't identify. The paper is huge and I strongly suggest you take a look at it, and some of the things they did were:
-Using machine learning to mine The Cancer Genome Atlas for HERVs that were most likely protein coding
-Searched CPTAC for these sequences!
-Took an absolutely brilliant to me (but maybe straight-forward if you're an immunologist?) and specifically looked for sequences that could be leveraged for CD8+ T-cells!
-Found evidence of translation of some of those?!?!?
-Went to The Cancer Cell Line Encyclopedia and separated out cells into different groups based on their clinical characteristics -- and found patterns of interest, suggesting that the expression of these HERVs could be used to differentiate different cancer cells(!!!)
-Went to the bench and validated that this was the case! They used flow cytometry and then they went to the best lab in the world for validating cell surface neo-antigens by mass spectrometry which synthesized heavy peptides and did targeted mass spec on them and demonstrated that these things were absolutely on the surface of some cells and would make great targets for immunotherapies!
That's a lot of exclamation points, right? But this group started with what seemed like a somewhat bizarre hypothesis, that ancient viral artifacts in our DNA that we carry around in every cell might have some use. Used publicly available resources and super smart approaches to find that they absolutely do something in cells, explain some data we've never been able to expain before AND figured out a way to leverage them!