By Mattias Mattsson on Wednesday, October 22, 2003 - 04:38 am:
I am testing a method for amino acid analysis (pre-column OPA/FMOC-derivatization) for Agilent 1100 HPLC. I have experienced unstable respons (area between 9 and 22 mAu*s) for lysine when running 6 injections of an amino acid standard (250 pmol/ul). The respons for all other protein amino acids, except proline, are more stable (RSD~5%). The detection is made using a UV-detector on 338 nm. The separation are good and retention stable (RSD<0,2%). Has anyone else experienced the same problem or have an possible explanation?
By Anonymous on Thursday, October 23, 2003 - 03:49 am:
The use of OPA for pre-column derivatisation is strongly influenced by time factor.
The OPA derivatives are not stable and so the critical point is time interval between OPA adjunct and injection. In literature I found that exactly one minute is a good procedure.
Good Luck
Regards
By Rolf Hecker on Thursday, October 23, 2003 - 05:36 am:
Hi Mattias,
I´ve been working with OPA/FmocCl derivatization for amino acids for years and never encountered these problems for Lysine. But it is definitely true that products are not stable. Normally You should get and RSD less than 1% using standards. If not, there could be a problem with some sort of contamination on columns that have already seen many samples. These contaminations could catalyze a faster decay of Your Amino acid derivatives and make quantification difficult.
If You want, send me Your protocol for sample prep, derivatization and chromatography, I´ll send You a short description of the parameters I used.
Good luck
Rolf
By HW Mueller on Thursday, October 23, 2003 - 11:56 pm:
To me it´s amazing that there are people who can get anything useful using OPA...., but still: Why bother? Dansyl-Cl is another problem that doesn´t die. Why?
By Rolf Hecker on Thursday, October 30, 2003 - 04:04 am:
Hello HW Mueller,
I´ve been working with OPA and Dansyl-Cl (fluorescence detection) routinely for years, and those methods were quite simple and robust even using biological samples, so where`s the problem?
By HW Mueller on Friday, October 31, 2003 - 12:11 am:
It must be a matter of looking at things, robustness is a matter of conjecture, in our hands and minds: totally unacceptable. Are you working with animal/human samples?
What about the multitude of products? If you get single products I surely would like to review your protocols and dig up our old notes and references to other researchers for comparison.
FMOC derivatives turned out much better, that is they seemed useful to us, but even these were not what we would have wished.