Journal of Chromatography B, 870 (2008) 222–232 Contents lists available at ScienceDirect Journal of Chromatography B journal homepage: www.elsevier.com/locate/chromb Automated GC–MS analysis of free amino acids in biological fluids Hannelore Kaspar 1 , Katja Dettmer ∗,1 , Wolfram Gronwald, Peter J. Oefner Institute of Functional Genomics, University of Regensburg, Regensburg, Germany article info Article history: Received 29 February 2008 Accepted 11 June 2008 Available online 18 June 2008 Keywords: Metabolomics Amino acids Gas chromatography–mass spectrometry Urine Plasma Alkyl chloroformate abstract A gas chromatography–mass spectrometry (GC–MS) method was developed for the quantitative analysis of free amino acids as their propyl chloroformate derivatives in biological fluids. Derivatization with propyl chloroformate is carried out directly in the biological samples without prior protein precipitation or solid- phase extraction of the amino acids, thereby allowing automation of the entire procedure, including addition of reagents, extraction and injection into the GC–MS. The total analysis time was 30min and 30 amino acids could be reliably quantified using 19 stable isotope-labeled amino acids as internal standards. Limits of detection (LOD) and lower limits of quantification (LLOQ) were in the range of 0.03–12 M and 0.3–30 M, respectively. The method was validated using a certified amino acid standard and reference plasma, and its applicability to different biological fluids was shown. Intra-day precision for the analysis of human urine, blood plasma, and cell culture medium was 2.0–8.8%, 0.9–8.3%, and 2.0–14.3%, respectively, while the inter-day precision for human urine was 1.5–14.1%. © 2008 Elsevier B.V. All rights reserved. 1. Introduction Metabolomics aims at the quantitative analysis of all metabo- lites in a given biological system [1,2]. In the absence of a single analytical technique that can cover the entire metabolome, anal- ysis is typically limited to the quantitative profiling of selected pathways or building blocks of the metabolome [3]. Important tar- gets for metabolic profiling are amino acids. Besides being the basic structural units of proteins, amino acids have several non-protein functions. They are a source of energy either through formation of keto acids from the ketogenic amino acids or through gluconeoge- nesis from glucogenic amino acids. Glutamate and -aminobutyric acid are neurotransmitters [4], while tryptophan and tyrosine are precursors of serotonin and catecholamines, respectively [5]. Glycine is a precursor of porphyrins, whereas ornithine is a precur- sor of polyamines [6] and arginine can be metabolized to form nitric oxide [7]. Elevated amino acid levels in blood plasma and urine are well-known markers for inborn errors of metabolism, such as phenylalanine in phenylketonuria (PKU) or branched-chain amino acids in maple syrup urine disease (MSUD) [8,9]. Amino acids also serve as markers for nutritional influences, e.g., urinary taurine lev- els are an indicator for fish intake [10], while the 1-methylhistidine level in urine correlates with meat protein intake [11]. ∗ Corresponding author. Tel.: +49 941 9435015; fax: +49 941 9435019. E-mail address: katja.dettmer@klinik.uni-regensburg.de (K. Dettmer). 1 These authors contributed equally to this work and therefore should be consid- ered equal first authors. Due to the important biological functions of amino acids, fast and reproducible analytical methods are needed for their quantita- tive analysis. There are several chromatographic methods available to quantify amino acids in biological samples. The most com- monly used method is cation-exchange chromatography followed by post-column derivatization with ninhydrin and UV detection [12–14]. Derivatization with o-phthalaldehyde (OPA) has been used both post-column after cation-exchange chromatography and pre- column coupled with reversed phase high-performance liquid chromatography (RP-HPLC) [15,16]. Reaction with phenylisothio- cyanate (PITC) produces phenylthiocarbamyl derivatives, which are separated by RP-HPLC and detected at 254 nm [17,18]. All these methods require manual sample preparation steps, including pro- tein precipitation, and analysis may last up to 2.5 h per sample. Another drawback is the UV absorbance detection: compared to mass spectrometry it lacks substance specificity and, therefore, co- eluting matrix components can cause over-quantification. GC–MS analysis of silylated amino acids is feasible [19], but not all derivatives are stable; arginine, for example, decomposes to ornithine [20], and glutamate rearranges to form pyro-glutamate. Another drawback is the sensitivity of the reagents and deriva- tives to moisture. Other derivatization procedures for GC analysis include reaction of the amino acids with pentafluoropropyl anhy- drid/isopropanol [21,22] or trifluoroacetic anhydrid/isopropanol [23]. However, these methods involve reagent removal and solvent exchange, which renders their automation difficult. Amino acids can be derivatized directly in aqueous solution using alkyl chloroformates [24–26]. The amino acids react very quickly, for instance, with propyl chloroformate and the derivates 1570-0232/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.jchromb.2008.06.018