The purpose of the 1992 amino acid analysis study was to poll the current techniques being used for amino acid analysis as well as to determine the methodologies and accuracy of cysteine and tryptophan identification and quantitation. Bovine pancreatic chymotrypsin (Sigma Chemical Co., 56 ug, 2.2 nmol) was sent to 231 ABRF facility directors as an unknown. Facilities were requested to perform amino acid analysis on the sample, quantitate the amount and report the composition, including values for cysteine and tryptophan. The participants were also requested to indicate their method of analysis with regard to instrumentation, calibration and quantification .
A total of 59 facilities participated in the study. The use of pre-column analysis techniques and post-column techniques were nearly equal with the majority of both the pre-column and post-column techniques using PITC or ninhydrin for derivatization or detection respectively. One interesting observation is that the use of automated derivatization and hydrolysis has increased.
For instrument calibration most users reported using unhydrolyzed free amino acids for standards. A small number reported hydrolyzing free amino acid standards prior to analysis and some users also noted using hydrolyzed peptide or protein standards for instrument calibration. Regardless of standard type, most users either carried out a single analysis or averaged several analyses to generate the amino acid response factors. Only a few users produced response factors generated from a concentration/absorbance curve of the standards. Most users stated that calibration of the instrumentation was performed either every day or for each "batch" of analyses.
6N HCl was used for standard hydrolyses at nearly all of the sites with most sites including additives. Phenol was the most common additive, typically present at a 1 % concentration. An increase in the use of vapor phase hydrolysis was observed this year compared to the past with an approximately 2: l ratio of users preferring this form of hydrolysis. As usual, there was a variety of times and temperatures reported for hydrolysis but essentially they fell into two groups: 110 + 2.7deg.C for 22 + 2.4 h at 34 sites and 160 t 15deg.C for 1.4 + 0.4 h at 19 sites. No distinct correlation was noted for hydrolysis conditions and the quality of the resultant data. Norleucine was the most commonly used internal standard (30 sites) being added either before (20 of the 30 sites) or after (4 of 30) hydrolysis. No definitive difference in quality of analysis or quantitation was noted in sites which used an internal standard.
The results from the standard analysis of chymotrypsin were good with an average % error for all the sites of 10.5%. Laboratories which performed standard analyses with less than 10% error were nearly equally divided between the post-column and precolumn methodologies. The range of sample amount used for the standard analyses generally fell into two groups: 1 - 3 ug and 3 - 10 ug. Most post-column methodologies used the higher amount of sample. The average amount of protein quantified was 43.4 + 17.4 ug which represents a 77 + 40% yield. This value was lower and more variable than the yields reported in the 1991 and 1990 ABRF surveys and thus may be sample dependent.
Seven different methods for determining Cys were reported. The average % error for Cys was 24.1 % which is higher than that reported for the standard analyses (10.5%). It is of interest to note that the majority (75%) of those sites which obtained < 10% Cys error also reported c 10% error for the standard analysis. This suggests a correlation between accuracy in amino acid analysis and the technical skills of the chemist performing the analyses. The most common technique for Cys analysis was an oxidative approach either with performic acid prior to hydrolysis or during hydrolysis in the presence of dimethylsulfoxide. Of the three Cys analysis methods used (oxidation, alkylation, disulfide exchange) no one method appeared superior; however, direct analysis of Cys was clearly inferior to any of the above techniques .
For Trp determination, 40 sites reported using one of seven different methods. As is typically anticipated in Trp determinations, the average % error was high (86.1 %) with only 10 of the 40 sites reporting less than 10% error. Five of the sites with low % Trp error used HCl hydrolysis in the presence of dodecanethiol, suggesting this to be a superior technique for Trp analysis.
This year's study was supported by more sites than in any previous year and provided significant information regarding the status and quality of Cys and Trp analyses. This study demonstrated that three Cys analysis methods, namely oxidation, alkylation and disulfide exchange, can yield satisfactory results if the chemist performing the analyses has good bench skills. Although the results from the Trp analyses were indicative of the greater difficulty in obtaining excellent results for this residue, a significant number of successful Trp analyses were obtained using vapor phase hydrolysis with added dodecanethiol. Most successful methods for Cys and Trp analysis require only a modest additional effort and if carried out correctly will provide satisfactory data for these residues.
Return to the The ABRF Home Page