FAME analysis problem - blank has same peaks as plasma

[and]

Dear all,

I am having huge problems with the finishing of a method development for fatty acid analysis in human plasma/serum. I am using a GC/MS with a HP-88 100m column, detection is with SIM. I get a nice separation of all the peaks of interest. The derivatisation is with 1.2 % (w/v) HCl in methanol (similar as http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2817593/), the samples are derivatised for 3h at 100 C. Then I do a liquid-liquid extraction with hexane, which I dry under N2 to concentrate it.

I have two types of problems, so I hope you can help me as I am stuck:
1. I read that I should use odd-chain fatty acids as internal standards. But as it seems all the odd-chain fatty acids are present in human plasma (I checked C15, C17, C19, C21, C23), albeit in small concentration. But as I am trying also to quantify fatty acid C26 that is present in very small concentration this bothers me. Do you think that it would be ok to add C21 in such a high concentration (20-100x times the endogenous concentration in plasma) so the endogenous peaks does not infuence it much? It would mean that the internal standard would be in much higher concentration than some of the fatty acids, but I do not know what else I can do. Deuterised internal standards are too expensive.

2. Last week I saw that when I do a whole derivatisation procedure with an empty tube (so no sample, everything else is the same) and then analyse it I see the same chromatogram as when I analyse plasma, only the peaks are much lower.
-I first thought that it was I contamination in the GC/MS. I tried a no injection run and there were no peaks present. I injected hexane, which also did not show anything in the chromatogram. So it is probably not the GC/MS.
-I prepared new solvents, did the same derivatisation with no sample and the peaks are there. I had a few vials of methanolic HCl (3N) from Sigma which I also tried and got the same results.
-So I thought the glass tubes were not washed properly. I poured a few mL of hexane in 10 of them, dried the hexane and analysed them. There were no peaks in the chromatogram. So I though the caps of the tube for derivatisation must be the problem - tried new tube, also different tubes from a different manufacturer, the no sample derivatisation again showed the "plasma" chromatogram.
-I have yet to try if the plastic pippete tips are the problem, but I do not believe I can get all the fatty acids that are present in plasma from a plastic pippete tip.

I would really like to finish the method and start validating so I will be really grateful for all your suggestions.

Thank you!

[MSCHemist]

That's unusual. Fatty acids are broken down and synthesized 2 carbons at a time from acetyl-CoA which is why fatty acids are even chained. You could perhaps use a branched chain fatty acid which should definitely not be endogenous.

I do see side products when I do the reaction. I use the same method for transesterfication though for TCA cycle acid (citric malic tartaric succinic fumaric and lactic) I use isobutanol HCl and quench the reaction with NaHCO3/water before injection. I see ethers and other side products. Perhaps try the 45 degree gently conditions instead.

[GOM]

Hi

Unfortunately I only have experience with oils and fats (including fish oils that go up to C26)

I may well be wrong but my initial thoughts are

Problem 1

a) I did see traces of odd numbered fatty acids but they were so low that using an internal standard of (say) tridecanoic acid would not affect your quantification.

b) I would only feel comfortable if the internal standard was at a level of the largest peak.

Problem 2

If you do repeated injections of "Last week I saw that when I do a whole derivatisation procedure with an empty tube (so no sample, " do the peak areas diminish with multiple injections? If so this would indicate an inlet problem. If not then it is something in the procedure.

As I say I may well be wrong

Regards

Ralph