K. Ümit Yüksel
CryoLife, Inc.
Amino acid analysis (AAA) is an important technology for characterizing proteins (natural or expressed) and peptides (synthetic or derived by cleavage). In contrast to its more common uses, three aspects of AAA are generally neglected. First, AAA is an excellent tool for accurately measuring protein concentrations. Second, it is useful for both qualitatively and quantitatively evaluating the presence of free amino acids in physiological samples. Finally, physiological samples often contain more than the "common" twenty amino acids--for example, taurine and post-translationally modified amino acids (e.g., phosphorylated or deamidated)--and these unusual amino acids can be detected by AAA. This workshop was aimed to introduce and to provide practical assistance on the use of these three aspects of AAA.
Steven A. Cohen, Waters Corp., "Quantitative Amino Acid Analysis: Applications, Problems, and Solutions"
Quantitative measurements of protein concentration are undertaken to calculate the amount of protein present, to support protein sequencing results, to measure specific activity, or to calculate protein nutritional quality. Protein and peptide concentrations are often measured with general dye-binding techniques such as the Lowry, Bradford, or bicinchoninic acid procedures, ultraviolet spectroscopy, or total Kjehldahl nitrogen. All these are relative procedures that depend on inference to other "standard" proteins and can be affected by the amino acid composition of the protein under investigation (dye-binding methods), buffer components (Kjehldahl and UV), or detergents. These methods are adequate for solutions with high protein concentrations (e.g., 0.1-10.0 mg/ml), but accuracy is often reduced when more dilute samples are analyzed. In addition, the dynamic range of the dye techniques is limited, often requiring analysis of several dilutions to provide useful data. The linearity of these methods should be evaluated by examining plots of (response/protein concentration) vs. protein concentration in addition to plots of response vs. protein concentration.
Quantitative AAA can also be used and provides marked improvement at low sample concentrations. It is usually the preferred method for determining the absolute concentration of proteins, independent of reagents or methodology (6-amino-quinolyl-N-hydroxysuccinimidyl carbamate (AQC), phenyl isothiocyante (PITC), N-(9-fluorenylmethoxycarbonyl) (Fmoc), ninhydrin, or o-phtaldialdehyde (OPA)). The useful linear dynamic range of the procedure can span three orders of magnitude (5 ug/ml to 5 mg/ml). Historically, the ion-exchange (post-column detection) methods are slower than the reversed-phase (pre-column detection) methods. The quality of these analyses will depend on the choice of the derivatization method, the attention paid to sample preparation, and the precision of the individual conducting the analyses.
The data can be analyzed in several ways. Single residue normalization, where the amount of protein is calculated from the yield of only one amino acid residue, is the simplest method. For this method a residue whose yield is not particularly susceptible to hydrolysis and analysis conditions (e.g., Phe, Ala, or Leu) should be used for normalization. Best fit analysis is the most complex method. For best fit analysis, an initial average yield is calculated using the data for all amino acids. Then a refined average yield is calculated using only those amino acids whose yields deviate less than 15% from the initial average yield. Thus, values containing large errors are eliminated from calculations, and this makes the calculation more reliable. Best fit analysis is the method used by the ABRF Amino Acid Analysis Committee when evaluating the collaborative study results. Both these methods of data analysis require prior knowledge of the sample's amino acid composition. That is not necessary for total amino acid recovery calculations, allowing the calculation of concentrations of unknown samples. However, total amino acid recovery calculations are prone to errors from reduced recovery (underestimation) or interference.
Amino acid analysis is not always the preferred method for determining the absolute concentration of proteins because it can be affected by low yields of labile amino acids (Ser, Thr, Met, Tyr), by slow bond hydrolysis (Ile, Val, Leu), or by other chemistry factors (reaction of Pro with ninhydrin and of Asp, Glu, His, Lys with PITC).
In summary, to obtain the best results from AAA one should choose a method with good linearity, good reproducibility, good compositional accuracy, and ruggedness. The use of internal standards and replicate analyses is encouraged.
Peggy R. Borum, University of Florida, "Amino Acid Analysis of Physiological Samples"
Physiological AAA is conducted for basic research (e.g., biosynthesis, transport, metabolism, or function) or for patient care (e.g., metabolic disorders, nutrition, or pharmacology). Monitoring the status of patients with genetic deficiencies in amino acid metabolism such as maple syrup urine disease (MSUD) or phenylketonuria (PKU) is routine in clinical laboratories. Free amino acids can be obtained from cell cultures, tissue cultures, tissue biopsies, etc. Physiological fluids present challenges because of their complex matrices. In addition, platelets in plasma, contamination of plasma from hemolysis, contamination of cerebrospinal fluid with blood, and drug metabolites in urine make analysis of physiological samples difficult.
Whatever the source, sample preparation is the most important factor in physiological AAA and starts at the point of sample collection. For example, the tubes used for blood collection may include heparin or EDTA as anticoagulants, and these may interfere with analysis. All samples must be deproteinized either with picric acid (least favored because clean-up is required), trichloroacetic acid, sulfosalicylic acid (most common), or filtration (gaining acceptance) before freezing them for storage. Some amino acids might bind tightly to proteins (e.g., Trp to albumin), affecting analysis results. Sample storage can be at 4[[ring]]C (short-term), -20[[ring]]C, and -70[[ring]]C (long-term). Amino acids can differ in the stability during storage. Some are more stable at -70[[ring]]C than at -20[[ring]]C even for short periods; for example, Gln is rapidly degraded to Glu if the samples are kept overnight at 4[[ring]]C.
The quality of buffers used for analysis is more important for physiological samples than for hydrolyzed samples. Their purity and consistency are critical, and all precautions must be taken to prevent mold growth. Prepackaged buffers may be preferred for ease of use and standardization. The availability of standards for quantification and detection of physiological samples is a problem. Because synthetic mixtures are poor mimics of real physiological samples, biological reference standards should be used.
Physiological AAA requires standard operating procedures backed by validations (of the instrument, method, and operator), the presence of internal and external quality control programs, and good record keeping. The physiological values for the amino acids measured should be known in order to draw meaningful conclusions from the results. The results will also improve in quality and usefulness if the appropriate methods of analysis is selected. For this, one should know if it is more important to determine (a) the exact concentration of each amino acid or the profiles of the amino acids (i.e., absolute vs. relative concentrations) and (b) the concentration at a given time or over a time span (spot vs. trend).
What are the recent developments in physiological AAA? Instrumentation is being introduced to use microdialysis, capillary electrophoresis, and mass spectrometry for sample collection and analysis. With the changing instrumentation, in situ measurements of extracellular free amino acids for the evaluation of transgenic animals and of the expression of recombinant proteins are becoming possible.
Lowell Ericsson, University of Washington, "Unusual Amino Acids"
Once eukaryotic proteins are synthesized, they can undergo post-translational modifications that affect their function (e.g., phosphorylation, disulfide formation, glycosylation) or their degradation process [e.g., deamidation, glycation (non-enzymatic mono-glycosylation, racemization, and isomerization]. During routine DNA or protein sequence analysis and routine amino acid composition analysis, these changes will often be overlooked. Furthermore, physiological fluids and prokaryotes contain more than the twenty common amino acids. Detecting uncommon amino acids is complicated by their lability and rarity. For example, a single phosphoserine among 20 serines (5%) will go completely undetected if the protein sample is hydrolyzed for 22 hours at 110[[ring]]C, as is done routinely. Detecting modified or unusual amino acids requires several changes in sample preparation (e.g., shorter hydrolysis times). In addition, it is sometimes necessary to modify chromatographic conditions, using much longer analysis times and modified buffers to achieve adequate resolution.
Amino acid analysis on ion-exchange columns with post-column ninhydrin detection has been used for this purpose for a long time, resulting in the accumulation of a wealth of information on the elution times of several hundreds of ninhydrin positive compounds, including modified amino acids. Information about the elution positions of phenyl isothiocyanate (PITC)- and 6-aminoquinolyl-N-hydroxysuccinimidyl carbamate (AQC)-reactive compounds is extensive, and some data on diethylaminobenzene (DABS)-reactive compounds exists.
Chromatographic separation followed by detection based on light absorbance has been the benchmark AAA technology for the past several decades. Recently, mass spectrometric detection has begun to gain acceptance. A useful aid for this is Delta Mass, a list of the masses of several hundred compounds and their breakdown products that was compiled and is maintained by Ken Mitchelhill (St. Vincent's Institute of Medical Research, Melbourne, Australia). Delta Mass can be viewed from Ken Mitchelhill's World Wide Web site (http://www.medstv.uni-melb.edu.au/wwwdocs/svimrdocs/massspec/deltamassV2.html) or from the ABRF homepage (http://www.medstv.uni-melb.edu.au/ABRF.html).
During the workshop a booklet entitled "Tables of Amino Acids" was distributed. It contains tables listing amino acid retention times for the post-column ninhydrin system and for pre-column PTC-, DABSYL- and AQC- systems with references and the Delta Mass table. The booklet was assembled from data obtained from Beckman Instruments, Waters Corporation, and the literature. This booklet can be requested by sending electronic-mail to Lowell Ericsson (E-mail: ericssonlh@u.washington.edu).
Return to the The ABRF Home Page