Accurate Determination of Amino Acids in High-Carbohydrate-Containing Samples

In a high-sensitivity amino acid analysis (AAA) technique that does not require sample derivatization (AAA-Direct™, Dionex Corp., Sunnyvale, CA), both amino acids and carbohydrates are separated by high-performance anion-exchange chromatography and detected by integrated pulsed amperometry (IPAD). The highly sensitive direct detection capability of amperometry used in the technique eliminates the need for pre- or postcolumn derivatization. Chemical derivatization techniques complicate analysis, increase cost for expensive reagents, introduce safety hazards to laboratory personnel exposed to toxic solvents, and add a hazardous waste stream that must be safely disposed of. These obstacles are eliminated with the AAA technique.

Unfortunately, high concentrations of carbohydrates in some samples may coelute with amino acids and complicate their determination. Published techniques1–5 describe the resolution of carbohydrates from amino acids by changing the eluent hydroxide concentration and extending the initial isocratic elution phase. Although this technique works well for injected samples having carbohydrate concentrations up to about 100 μM, decreasing hydroxide eluent reduces detector response for the early-eluting amino acids. When amino acid concentrations are exceptionally low and the carbohydrate concentrations are exceptionally high (e.g., fruit juices), it is impossible to accurately determine amino acids (Figure 1). These samples require the removal of carbohydrates prior to amino acid analysis. Jandik et al.6 published an automated in-line sample preparation procedure that eliminated carbohydrates from amino acid samples prior to their analysis by AAA-Direct, and Genzel et al.7 reported the successful use of the procedure for cell culture media applications. When carbohydrate-containing amino acid samples are loaded on a Carbohydrate Removal Cartridge (CRC, Dionex) under acidic conditions, the amino acids bind to the CRC while the unbound carbohydrates are washed to waste. After suitable carbohydrate clearance from the cartridge, the valve switches position and elutes amino acids from the CRC onto the AminoPac® column (Dionex), where they are separated by anion-exchange chromatography. This technique was difficult to install and program, required an additional pump and two additional high-pressure valves, and the carbohydrate elimination feature could not easily be turned on or off to switch the system back to its typical AAA-Direct setup.

Figure 1 - High carbohydrate concentrations in samples interfere with the determination of amino acids using AAA-Direct. The figure shows that grape juice has a high carbohydrate concentration that precludes amino acid quantification.

This paper describes a simpler procedure to eliminate carbohydrates prior to AAA-Direct analysis of amino acid using the Carbohydrate Removal Accessory (CRA) and features of the ICS-3000 (both from Dionex). Using a CRA, carbohydrate-containing amino acid samples are pumped onto the CRC trap using the AS autosampler sample preparation syringe (Dionex) instead of a separate accessory pump. The CRC replaces the typical injection loop on the inject valve, eliminating the need for additional accessory valves. All controls are automated through the software and can be turned either on or off. The ICS-3000 can be set up in a dual-system format that allows two CRA AAA-Direct applications to run at the same time and thereby double sample throughput. The performance of this method is demonstrated, and its application to challenging samples, including cell culture media and juices, is shown.


Equipment and conditions

Equipment included the ICS-3000 ion chromatography system configured for AAA-Direct, consisting of a dual gradient pump, a detector compartment with an electrochemical detector outfitted with an AAA-Certified™ disposable gold working electrode, an autosampler with sample preparation and the CRA, and Chromeleon® software for instrument control and data management (all from Dionex). The AAA-Direct gradient conditions and CRA setup are described elsewhere.8

Standards and samples

The following standards were used: arginine, asparagine, citrulline, cysteic acid, cysteine, fructose, fucose, galactosamine, galactose, glucosamine, glucose, glutamine, glycine, hydroxylysine, inositol (myo-), lactose, lysine, maltose, mannose, methionine sulfoxide, norleucine, ornithine, propylene glycol, sucrose, taurine, threonine, tryptophan, tyrosine, urocanic acid, and valine. These were obtained from Sigma-Aldrich (St. Louis, MO), Pfanstiehl Laboratories (Waukegan, IL), or Fisher Scientific (Pittsburgh, PA); hydroxyproline was from Calbiochem (San Diego, CA); and sucralose was from McNeil Nutritionals (New Brunswick, NJ). Standard Reference Material 2389 (amino acids in 0.1 M hydrochloric acid) was obtained from NIST (Gaithersburg, MD). Italian Muscat grapes were California-grown fresh produce manually pressed, and the juice decanted. Decanted grape juice, Bacto® Yeast Extract-Peptone-Dextrose (YPD) broth (BD Diagnostic Systems, Sparks, MD), McCoy’s 5A mammalian cell culture medium (Sigma-Aldrich), and 100% carrot juice were centrifuged at 16,000 × g for 10 min, and the supernatant was diluted in water.

Results and discussion

CRA performance

Figure 2 - The elimination of carbohydrates and the separation of amino acids present in YPD broth (400-fold dilution) (a) and in McCoy’s 5A medium (50-fold dilution) (b) using 25 μL of diluted samples. AAA-Direct separations are shown with (A) and without (B) the CRA.

The percent recovery of amino acid standards using the AAA-Direct system with the CRA ranged from 102 to 110%, except asparagine (79%), hydroxyproline (16%), and taurine (0%). Lower recovery of amino acids (except taurine) did not prevent accurate quantification of these amino acids because the slopes of the resulting calibration curves reflected the low recovery.

The capability of the CRA to eliminate different types of carbohydrates was investigated. Using 10 μM amino acid standards with 1 mM glucose, fructose, sucrose, maltose, inositol, or lactose, high amino acid recovery was determined for each carbohydrate matrix. Slightly elevated recovery values were observed for amino acids coeluting with each trace carbohydrate breakthrough peak. Accurate recovery of amino acids was possible up to at least 56 mM glucose, except for threonine, which coelutes with the glucose breakthrough peak. Varying either the amino acid concentration (1–10 μM range) or carbohydrate concentration (10 μM–56 mM) did not change amino acid recovery from a carbohydrate matrix. The CRA eliminated fucose, galactose, mannose, propylene glycol, and sucralose, but not the amino sugars (galactosamine, glucosamine), which as expected bind to the cation exchange cartridge. Except for asparagine and hydroxyproline, the CRA did not change amino acid calibration curves. Retention time RSDs ranged from 0.02% to 0.40%, and the peak area and peak height RSDs ranged from 1.2% to 4.1% and 0.7% to 9.1%, respectively, for 15 injections of amino acid standards (10 μM) over one day.

Application of the CRA

Figure 3 - The elimination of carbohydrates and separation of trace amino acids present in freshly pressed Italian Muscat grape juice (100-fold dilution) (a) and in commercial carrot juice (100-fold dilution) (b) using 25 μL of diluted samples.

Different high-carbohydrate sample matrices were investigated using products from the pharmaceutical, biotechnology, and food and beverage industries. The CRA allowed successful amino acid determinations for all samples investigated. Figure 2 shows the separation of amino acids present in a fermentation broth medium (YPD broth, 400-fold dilution) and in a mammalian cell culture medium (McCoy’s 5A medium, 50-fold dilution), with and without the CRA. Figure 3 shows the separation of amino acids present in 100-fold dilutions of grape juice and carrot juice, with and without the CRA. For each of these samples, amino acid determinations at these dilutions were impossible to perform without the use of the CRA due to large concentrations of glucose, sucrose, or fructose. Using the CRA, carbohydrates were removed prior to chromatography, thus allowing accurate amino acid determinations. Table 1 shows the amino acid spike (5 μM) recovery from these matrices, where recovery from YPD broth ranged from 75% (Asp) to 105%, with an average of 91%; recovery from McCoy’s 5A medium ranged from 45% (Asp) to 112%, with an average of 93%. The spike recovery from grape juice ranged from 25% (Asp) to 105%, with an average recovery of 86%. The recovery from carrot juice ranged from 56% (Asp) to 112%, with an average recovery of 90%.

This method was also installed on a dual ICS-3000 that allowed two AAA-Direct applications to operate in parallel, using only one autosampler. With this installation, throughput was nearly doubled from 18 to 35 injections per day, and the second system had the same high performance as the first.


Amino acids were accurately quantified in samples containing high concentrations of carbohydrates using automated elimination of carbohydrates. The method requires no sample derivatization for amino acid detection and was successfully applied to cell culture media and fruit and vegetable juices.


  1. Dionex Corp. Application note 150. Determination of amino acids in cell cultures and fermentation broths. Literature product no. (LPN) 1538, July 2003.
  2. Hong, Y.; Ding, Y.-S.; Mou, S.; Jandik, P.; Cheng, J. Simultaneous determination of amino acids and carbohydrates by anion-exchange chromatography with integrated pulsed amperometric detection. J. Chromatogr. A2002, 966, 89–97.
  3. Ding, Y.; Yu, H.; Mou, S. Direct determination of free amino acids and sugars in green tea by anion-exchange chromatography with integrated pulsed amperometric detection. J. Chromatogr. A2002, 982, 237–44.
  4. Hanko, V.P.; Rohrer, J.S. Determination of amino acids in cell culture and fermentation broth media using anion-exchange chromatography with integrated pulsed amperometric detection. Anal. Biochem. 2003, 324, 29–38.
  5. Hanko, V.P.; Heckenberg, A.; Rohrer, J.S. Determination of amino acids in cell culture and fermentation broth media using anion-exchange chromatography with integrated pulsed amperometric detection. J. Biomol. Tech.2004, 15, 315–22.
  6. Jandik, P.; Cheng, J.; Jensen, D.; Manz, S.; Avdalovic, N. Simplified in-line sample preparation for amino acid analysis in carbohydrate containing samples. J. Chromatogr.B2001, 758, 189–96.
  7. Genzel, Y; König, S.; Reichl, U. Amino acid analysis in mammalian cell culture media containing serum and high glucose concentrations by anion exchange chromatography and integrated pulsed amperometric detection. Anal. Biochem. 2004, 335, 119–25.
  8. Dionex Corp. Technical note 69. Carbohydrate Removal Accessory for the Determination of Amino Acids in High Carbohydrate-Containing Samples using AAA-Direct, 2006.

Mr. Hanko is Staff Chemist, and Dr. Rohrer is Director, Applications Development, Dionex Corp., 1228 Titan Way, Sunnyvale, CA 94086, U.S.A.; tel.: 408-737-0700; fax: 408-739-4398; e-mail: