Detection/quantitation of aminoglycosides in pure proteins

[gckiv]

We're using LC/ESI to detect small molecules (kanamycin, methotrexate, chloramphenicol, etc.) in purified protein as a contaminant release assay for cGMP pharmaceutical production. The goal is to have a simple, robust sample prep and analysis method for many of these small molecules that are used during fermentation/cell culture. We've looked at SPE and precipitation methods with mixed results.....LC/MS is pretty standard, using SIM on the ion of interest, plotting BP height.

I was wondering if anyone had any thoughts/experience on 1) a simple precip method that doesn't effect target analyte stability (mixed results with organic precip, how about NaOH or TCA?) and 2) a simple, reliable mobile phase composition (we've tried low TFA and FA, but dimerization becomes a problem; tried low conc. ammon. acetate in MeOH, but resolution and background are the problem, how about HFBA?). Any input is very much appreciated.

Thanks,

George

[Uwe Neue]

Nearly all of my experience is with SPE of blood plasma samples, and I know that one can get a very clean sample prep from plasma with the right approach.

I am surprised that you think that the addition of organic solvent affects analyte stability. Acid and base addition, yes, but organic? Could it be related to protein binding rahter than stability?

Resolution and backgound is less of a problem with good sample prep. I would work on this instead of thinking about different mobile phases. FA should be fine. I do not understand what you mean by dimerization?

If you clarify, maybe we can suggest a few things.

[Mary Carson]

There are numerous papers out there on these drugs. Check out the veterinary drug residue literature for methods with fairly low detection limits. Some issues you may have with LC-ESI-MS are
1. Chloramphenicol is a negative mode analysis. Depending on your instrument, formic acid may suppress the signal. Otherwise, it is "nice" compound to chromatograph in reverse phase (doesn't really even need a modifier...). Extracts nicely with EtOAc or ACN.
2. Kanamycin is a bit trickier to analyze. I've done it with HFBA in the mobile phase using a RP column. You pretty much have to dedicate your mass spec though--HFBA ruins it for negative ion work and can suppress positive ion analysis, too. HFBA is nearly impossible to get rid of, though I've heard replacing diaphrams on LC degasser helps (in addition to replacing all other plumbing and doing a really thorough source clean). If I had it to do over again, I do HILIC for aminoglycoside analysis (that ought to make at least one regular participant in this forum happy ). Aqueous TCA extracts aminoglycosides nicely, but you do want to minimize TCA in the MS--same problems as HFBA.

I'm not sure you'd be able to pull off residual analysis for both of these compounds in one assay....