Thursday, January 29, 2015

PROMIS-Quan! 3,000 plasma proteins quantified in a single run?!?!

I stole the picture above from this paper from Michal Harel et al., that is currently in press at MCP.  Technically, this is an incredibly sophisticated paper, but it might be the solution for frustrated researchers out there who just can't get beyond that few hundred protein limit in their plasma proteomics study.

I stole that pic cause it sums it up.  PROMIS-Quan is short for: – PROteomics of MIcroparticles with Super-SILAC Quantification.  And if that doesn't suggest the difficulty of setting up this study, nothing will.

The plasma proteomics is kind of broken up into two steps, the normal stuff, and then plasma microparticles (which are really freaking cool in their own right...just ridiculously low abundance).

If I have this all straight in my head, this is the general idea.  The SILAC supermix is going to have representative proteins from just about protein/peptide humans can produce.  We can put in as much of that stuff as we want, including levels that will be easily fragmented and identified.  To get around the insane difference in dynamic range of proteins in plasma, we don't actually go after fragmenting the same peptides/proteins in the plasma.  We know they are there because of the SILAC pairing effect.

Nice, right?  Some people set their instruments so that when they do SILAC they only fragment either the heavy or the light pair anyway.  Since the SILAC pairs are clear as day in high res, you don't need to waste time fragmenting both the heavy and the light.  This study exploits this fact to give us a completely unprecedented level of plasma peptide coverage.

Now, if I could find one fault with this study it is this:  That it would be hard for me to replicate these results myself because my SuperSILAC mix would be different than the one they used.  Now, if some enterprising soul would make up, I don't know, 10,000 L of SILAC cells and we could all use the same supermix for the next 5 years or something, we'd be in business.

If you are doing plasma proteomics, thinking about it, or did it for years and got nothing out of it, I strongly suggest you check this paper out.


  1. There's an active patent on the production of Super-SILAC mixes, this might be the reason we don't have any of these supermixes on the market...

  2. Hi Ben
    Are there any references for the statement "Some people set their instruments so that when they do SILAC they only fragment either the heavy or the light pair anyway." I am interested in the methods they followed on an orbitrap and it would be great to have a reference for selective fragmentation of either the light or heavy peptides.


  3. Prahlad,
    I'm looking for the old paper. Its pretty easy to set up, but different on each instrument. The function your looking for is "mass tags". On the hybrids there are a couple steps required. Description is in the iORBI 2012 talk called Bottom up top down.
    Now, there is a paper from 2008, I think, where they set up this experiment on an Orbi XL and found that the best coverage came from just doing SILAC without the mass tag function. This discouraged people from further working with this feature. Logically, though, I've never been able to figure out why this was the case. The mass tags should be better, right? I've always wondered if their results were due to instrument control issues on their particular Orbi.

    1. Woah! These slides are incredibly useful! Thanks for pointing out to these. Yeah, I think with the mass tags feature turned on the number of identified proteins should be higher. The comparisons shown on the slides are counterintuitive. We are experimenting with SILAC methods and will give this a try and see if an opposite effect is found. Will comment either ways on the results here.