Tuesday, November 25, 2008

Utterly confusing gremlin problems

So, it turns out that the problem that ruined my week is something much different than I thought it was, and maybe not the same kind of experiment-killing gremlins I thought at first.

See: we make peptides here, and the way you characterize peptides is by HPLC-mass spec. It tells you 1. how many components are in your crude material (hopefully less than 2!) and 2. the molecular weights (and thus identities) of those components. When you make your own peptides, you generally know which amino acids you added and in what order, and you also know what reagents you used that might have given strange byproducts.

The most common problems are a) deletions: missing an amino acid here and there for some % of your crude material; and b) terminations: chemical side reactions that stopped your peptide dead after some particular amino acid. Both of these give predictable products, so you can do fairly simple detective work to determine the source of a problem (made wrong concentration of solution, reagent was going bad, etc.). Thusly, uninterpretable LC/MS data, that gives you a mass or masses that have nothing to do with what you tried to make, is extraordinarily confusing because unless gremlins changed your amino acid solutions around, THERE IS NOTHING ELSE THAT IT COULD BE.

Our result had ONE peak in the UV trace from LC/MS--that pretty much means it's one compound (unless other components coelute, in which case they'd have to be pretty damn similar because our LC gets fantastic peak resolution)... and the mass spectra associated with that peak ARE COMPLETELY WEIRD. I have plumbed the depths of my knowledge of molecules, ions and fragmentations and have been completely and utterly stumped. I've also found that this strange behavior is exhibited by ONLY, but ALL, peptides with a certain sequence at their C-terminus, through a number of analogs I looked at. I also found that even my CONTROL peptide from my previous lab, which had NEVER looked like this before, is showing the same wacky thing. Given that this control peptide was stored lyophilized at -20 for the whole time, there is next to NO chance that it actually chemically degraded, and besides, the MALDI-MS (which only shows whole peptides in linear mode) is NOT showing the same thing, it looks fine there.

All of our other peptides that don't have this certain sequence are not showing this behavior, they look totally normal. So whatever I am seeing has something to do with this certain sequence, and my new-lab mass spec (which is a different brand than my old-lab one). I checked everything like the source voltages etc. and matched up as many of them as seem possible to match up to the settings on old-lab MS. Still no change in the spectra. SO WEIRD. There are only certain relatively predictable fragmentations one would expect here, and none of them explain my crazy ion series.

I ended up calling in the experts on this one, and so hopefully we'll get to the bottom of this crazy gas-phase ion chemistry mystery. I am at my wits end with trying to puzzle it out, and my poor lab staff is all new to peptides and mass spec so they are completely confused.

2 comments:

ScientistMother said...

When I've had microscopes do that (completely weird results), it usually ends up being a equipment problem. If you're control, which also behaves, its wacko I would say its the new mass spec.

Arlenna said...

The weirdest part, though, is that most peptides look just fine on this mass spec. It's only ones with this certain type of sequence at one end that do this, and of course, that sequence is essential to what we do here so that's a real PITA.