SPECIAL FEATURE

The following article is the result of efforts by the Amino Acid Analysis and Education committees to provide ABRF members with information that will be useful as they strive to improve their own services. It was felt that one very good source for educational information would be those laboratories who did very well on the 1992 amino acid analysis sample. As a result, letters were sent to six participants from that study who did exceptionally well in providing above average data, asking if they would consider revealing their identity and prepare an article describing their protocol. This letter was sent by the same individual who assured anonymity during the study, and the identity of the participants was not known to anyone at ABRF until they revealed themselves. We hope that you will find the information contained in this article useful and would welcome any comments you might have.

Strategies For Performing High Quality Amino Acid Analysis

Jeffrey D. Caplan and P. Balasubramanian, Protein and Analytical Chemistry Department, Lederle-Praxis Biologicals, Rochester, New York

High quality amino acid analysis is a goal for many who use this very beneficial technique. Based on the results we obtained (site #33, Table 1) (24k) in the 1992 ABRF amino acid analysis study (1), we were asked to share information regarding our sample preparation and analysis procedures with other ABRF members. During normal, routine types of samples, our laboratory performs a standard analysis that consists of determining the common amino acids with the exception of cysteine and tryptophan. Having had additional experience with the determination of cysteine and tryptophan, we decided to perform each type of analysis requested in the amino acid analysis study.

The instrument we used for this study was a Beckman 6300. It is designed for ion exchange separation followed by postcolumn ninhydrin derivatization. This instrument uses Beckman's sodium based 4 mm x 12 cm High Performance column and sodium citrate buffers. A three step pH and three step temperature gradient are used in the normal run program. Beckman's System Gold version 4.03 is the data system that was used. Our experience has shown that the analyzer detection limit, at less than 10% error, is 50 pmoles per amino acid (2).

A standard amino acid mixture is used routinely and is prepared monthly (regardless of whether or not it has been completely used). The source of the mixture is Pierce (Amino Acid Standard H) and the concentration is diluted from 2.5 umoles/ml to 500 pmoles/50 ul using Beckman's sample buffer. Three, 500 pmoles/50 ul standards are run before the unknown samples and the third standard is usually selected for calibration. This standard calibration procedure is performed each time unknown samples are to be analyzed. S-beta-(4-pyridylethyl) L-cysteine (see paragraph on cysteine determination) is purchased from Sigma, prepared at 2.5 ,umoles/ml and diluted to the standard concentration of 500 pmoles/50 ul using Beckman's sample buffer. Tryptophan is obtained from Pierce (Amino Acid Standard Kit 22) and prepared as described in the previous sentence.

Instrument maintenance is performed at various intervals. Maintenance performed before every run includes checking the pressure, column temperature, and bubble traps. Other frequent maintenance includes cleaning the instrument, oiling both pumps (monthly), and checking the flow rate (weekly). Maintenance performed on a six month basis (or as needed depending on instrument use) includes changing the ninhydrin filter, cleaning the flow cell with 70deg.o ethanol, and changing all necessary pump components (i.e., seals, springs, check valves).

The method used for a normal, routine type of analysis begins with a sample that is in solution (water, dilute acetic acid, dilute trifluoroacetic acid). The hydrolysis ampules (Wheaton, 2 ml) are thoroughly washed with methanol and dried in an oven for 2 hours. Duplicates of each sample are always performed. To perform the standard analysis on the ABRF sample, the volume equivalent of 5 ug was transferred to an ampule and evaporated in a Savant Speed Vac system. Once dried, 100 ul of 6N HCl (Pierce, Sequanal Grade), 5 u1 of Phenol (Aldrich, 99+ %) and 1 ul of 100% 2-mercaptoethanol were added (3, 4). Prior to sealing the ampule, a freeze and thaw methodology was carried out (5). The sample was frozen in liquid nitrogen and evacuated to approximately 100 mtorr. A valve between the vacuum pump and the sample was closed temporarily while the sample was thawed (releasing any dissolved oxygen contained in the solution). Again, the sample was frozen and evacuated to approximately 75 mtorr (evacuation was accomplished by opening the valve to the vacuum pump). The valve was then closed and the sample was thawed to release any additional oxygen that may be dissolved in the sample. This freeze and thaw cycle was repeated once again, evacuating to a final 50 mtorr before the ampule was heat-sealed while still under vacuum (using a gas/oxygen burner)(4-6). The sealed ampule was placed in a glass bead bed, within a 110deg.C convection oven, for 22 hours (6). Our experience indicates that this freeze and thaw method, in general, increases the recovery of many amino acids. After hydrolysis, the ampule was opened and the sample was evaporated to dryness in a Savant Speed Vac system. Resolubilizing the sample was accomplished by adding 250 u1 of Beckman's sample buffer (1). This was then filtered through a 0.2 um filter and 50 u1 of the resulting amino acid solution was separated using the Beckman analyzer. Standard analysis conditions and chromatograms for the ABRF sample can be seen in Fig. 1.(24k)

Cysteine determination tends to be a difficult and time consuming procedure. This is due to the alternative methods that are required as a result of cysteine destruction during normal HCl hydrolysis. Some of the alternative methods that exist are S-pyridylethylation, performic acid oxidation, carboxymethylation with iodoacetate, dithiodiglycolic acid hydrolysis, and dithio-diproprionic acid hydrolysis (6, 7). S-Pyridylethylation is our choice for this procedure, as the S-beta-(4-pyridylethyl) L-cysteine derivatized product is highly resistant to HCl destruction and is easily separated using our analyzer. To accomplish this derivatization for the ABRF sample, the volume equivalent of 30 ug was transferred to a small glass vial, dried, and dissolved in 100 u1 of 6 M guanidine-HCl (buffered with 0.25 M Tris/l mM EDTA pH 8.5). The disulfide bonds were reduced by adding 2.5 ul of 10% 2-mercaptoethanol and incubating at room temperature under nitrogen for 2 hours. After incubation, 3 ul of 4-vinylpyridine was added and the sample was again incubated at room temperature under nitrogen for 2 hours. At this point, the derivatized sample was desalted using gel filtration (Bio-Rad, P6-DG, 7 g dry resin swelled in 5deg.70 acetic acid, 14 mm x 30 cm column, 5% acetic acid eluent, 280 nm detection). Generally, desalting is accomplished using the technique most applicable to the size and nature of the protein or peptide (8-10). After desalting the sample, the resulting solution was lyophilized and hydrolyzed as described above. Separate analyzer run conditions are required to resolve PE cysteine which coelutes with lysine under standard run conditions. Refer to Fig. 2 for analyzer conditions and a chromatogram of the PE cysteine run.(24k) To determine the other amino acids, the remaining PE cysteine sample was run using the standard conditions. Lysine was calculated by subtracting the number of PE cysteine pmoles from the total pmoles stated for lysine (which actually represents lysine + PE cysteine).

Tryptophan is significantly destroyed during normal HCl hydrolysis and therefore, an alternative hydrolysis agent should be used. The more common alternative hydrolysis agents are methanesulfonic acid, p-toluenesulfonic acid, thioglycolic acid, sodium hydroxide, and mercaptoethanesulfonic acid (7). Our experience using sodium hydroxide and mercaptoethanesulfonic acid suggests that, of the two, mercaptoethanesulfonic acid is the preferred method. To accomplish this alternative acid hydrolysis for the ABRF sample, the volume equivalent of 5 ug was transferred to a Wheaton 2 ml ampule followed by evaporating the sample in a Savant Speed Vac. Once dried, 100 u1 of 3 M mercaptoethanesulfonic acid was added. Next, the freeze and thaw procedure stated for normal analysis was performed. This was followed by heating the sample in a 110deg.C oven for 22 hours. The ampule was then opened and the pH of the resulting amino acid solution was adjusted to 2.2. This was accomplished by adding 75 ul of 1.4 M sodium citrate and 75 u1 of Beckman's sample buffer (11). Separate analyzer run conditions (isothermal and isocratic - using the highest pH buffer) are required to resolve tryptophan. This is due to tryptophan coeluting with ammonia when the normal analyzer run conditions are used. Refer to Fig. 3 for analyzer conditions and a chromatogram of the tryptophan run. (16k) To determine the other amino acids, the remaining mercaptoethanesulfonic acid hydrolyzed sample was run using the standard conditions.

In conclusion, we have determined that the above methodologies work very well for our amino acid analysis needs. The results of the 1992 ABRF sample indicated an average percent error of 3.3% (Table 1). We constantly strive to improve various aspects of our amino acid analysis.

1. Strydom, D. J., Anderson, T. T., Apostol, 1., Fox, J. W., Paxton, R. J. and Crabb, J. W., in "Techniques in Protein Chemistry IV" (Angeletti, R. H. ed), Academic Press, New York, 1993.

2. Arrizon-Lopez, V., Biehler, R., Cummings, J., and Harbaugh, J., Beckman System 6300 High-Performance Amino Acid Analyzer, in "High-Performance Liquid Chromatography of Peptides and Proteins: Separation, Analysis, and Conformation" (Mant, C.T., and Hodges, R.S., eds), CRC Press, Boca Raton, 1991, 859.

3. Smillie, L. B., and Nattriss, M., Amino Acid Analysis of Proteins and Peptides: An Overview, in "High-Performance Liquid Chromatography of Peptides: Separation, Analysis, and Conformation" (Mant, C.T. and Hodges, R.S., eds), CRC Press, Boca Raton, 1991, 847.

4. Apostol, I., (1992) Biomol. Newsletter Vol. I, No. 2 (23-25).

5. Moore, S., and Stein, W. H., Chromatographic Determination of Amino Acids by the Use of Automatic Recording Equipment, in "Methods in Enzymology" (Colowick, S. P., Kaplan, N. O., eds.) Vol. 6, Academic Press, New York, 1963, 819.

6. Benson, J. R., Louie, P. C., and Bradshaw, R. A., Amino Acid Analysis of Peptides, in "The Peptides" (Gross, E., and Meienhofer, J. eds) Vol. 4, Academic Press, New York, 1981, 217.

7. Grant, G. A., Evaluation of the Finished Product, in "Synthetic Peptides" (Grant, G.A., ed) W. H. Freeman and Company, New York, 1992, 185.

8. Hawke, D., Yuan, P., (1987) Applied Biosystems User Bulletin No. 28.

9. Wilson, K. J., and Yuan, P. M., Protein and Peptide Purification, in "Protein Sequencing" (Findlay, J.B. C., and Geisow, M.J. eds) IRL Press at Oxford University Press, Oxford, 1989, 1. 10. Charbonneau, H., Stratagies for Obtaining Partial Amino Acid Sequence Data from Small Quantities ( < 5 moles) of Pure or Partially Purified Protein, in "A Practical Guide to Protein and Peptide Purification for Microsequencing" (Matsudaira, P.T., ed) Academic Press, San Diego, 1989, 15.

11. Creamer, L. K. and Matheson, A. R., (1976) N.Z. Jl. Dairy Sci. Technol.. 11. 211-212.

Return to the The ABRF Home Page