Sunday, March 26, 2017

Phosphoproteomics of irradiated cancer cells!

This awesome new paper in press at MCP takes a swing at filling in some of the blanks in our understanding of DNA repair when cells are hit with different kinds of radiation!

We know lots and lots about our various DNA repair pathways. Every big school has a radiation oncology department and that's what they do. Is there still more to learn here?  SILAC phosphoproteomics says yes!

The figure above describes what they did. They chose A549 because (unlike many of our normal lab cell lines) it has a reasonably normal DNA repair pathway (p53 works right), so it will be more like what we'd see with normal human cells than others where they'll just keep happily dividing, pushing broken DNA right into the next generation of cells.

The used an Orbi XL for the global proteomics/phosphoproteomics running in high/low mode (CID-ion trap MS/MS). The protein was divided by 1D gels and the phosphoproteome was enriched thoroughly with both TiO2 and IMAC.  All the data was processed in MaxQuant with thorough downstream analysis with some R packages, cytoscape and PhosphoSite.

Cool stuff they found was validated by PRM with heavy peptide standards, but I'm not clear on the details and this paper has way too much supplemental info for a Sunday morning paper ;)

What did they find? With these treatments, little to no changes in the global protein level -- but phosphorylation changes all over the place!

I'd like to mention that this isn't a "peak bagging" paper where they goal was "How many phosphopeptides can we detect" (which is great, we do need those to test new methodologies, no criticism!)  This is a purely biology centric study. If the confidence level of the phosphopeptide ID wasn't awesome -- they toss it early in the analysis. If it wasn't clearly differentially regulated in response to treatment (with statistical validity) -- they toss it. They're looking for big hits that they can (and do) validate.

You start with getting amazing confidence in their methodology -- because they find the normal players (brazenly stolen table...don't sue me and I'll take it down if it is a problem! I'm just excited about these awesome results!)

BRCA1/RAD50/53BP1? Check! (They mention in the paper ATM/ATR, but it didn't make this figure.)

They find about 200 things that pass their stringent thresholds and about 1/3 of them are the normal stuff we know about. And then...the rest...IS ALL NEW STUFF (meaning it isn't listed in PhosphoSite anywhere!)

The Supplemental Info is seriously old school (and I love it!) -- page after page of manually annotated phosphopeptide spectra, their validation for each phosphopeptide with quan and stats for each radiation type.

They did make some poor person do a bunch of westerns to prove the PRMs were okay on the ones they could find antibodies know....drives me a little crazy, but may never really go away because a western is so much easier to explain to people outside of mass spectrometry.  This team were experts from beginning to end -- the westerns look great and (surprise!) match the PRMs exactly.

No comments:

Post a Comment