CLIN.CHEM.31/11, 1849-1854 (1985) CLINICALCHEMISTRY, Vol. 31, No. 11, 1985 1849 AnalyticalConsiderations forQuantitativeDeterminationofSerotoninand Its MetabolicallyRelatedProductsinBiologicalMatrices Richard B. Mailman’ and Clinton D. Kilts2 We discussthe useof“high-performance”liquidchromatog- raphywith electrochemicaldetectionfor the quantificationof serotonin (5-hydroxytryptamine)and metabolically related products(i.e., 5-hydroxyindoleacetic acid and 5-hydroxytryp- tophan) in biological matrices. Two methods are descnbed: onefor routinequantificationof nanogramamountsor more; the other a two-column purification-separation(“trace en- richment”)techniquefordeterminingsubnanogramamounts of these compounds. Various factors affecting the liquid- chromatographic separationand quantificationof 5-hydroxy- indolesare detailed, includingselectionof appropriateinter- nal standard, sample purification,and effectsof changes in mobile phase composition,e.g., ion-pair reagent, organic modifiercontent, and pH. We describe in detail the trace- enrichmentmethod for determinationsin the range of 25 pg to 25 ng, and discuss the general factors to considerwhen developingor modifying an assay for specificapplications. Just over a decade ago, Kissinger et al. (1) described in this journal the use of electrochemical detection with “high- performance” liquid chromatography (HPLC-EC) for the quantification of selected organic compounds.3 Three years later, Sasa and Blank (2) described one of the first applica- tions of HPLC-EC for the quantitative determination of serotonin (5-hydroxytryptamine; 5-HT). In the intervening years, literally dozens of methods based on these principles have been published for quantification of indoles in a variety of biological matrices, the more recent ones empha- sizing simultaneous determination of several related indoles and catecholamines, and their metabolically related com- pounds. It would be redundant simply to detail one or another minor modifications of these methods, and, as noted, the literature abounds with such papers. Many differ little in substance, or offer little or no advantage over others. Here we intend to summarize how HPLC-EC can be used to quanti1’ serotonin and various indoles in various matrices, through a discussion of two specific methods that have proven useful in our research. Although the power of reversed-phase HPLC as a separa- tion technique certainly permits the simultaneous determi- nation of literally dozens of compounds (or even more if gradient elution is used), we believe that, while technically impressive, such an approach is of limited value. These methods usually entail chromatographic times of 45 to 60 mm per sample, which limits throughput markedly, often to ‘Departments of Psychiatry and Pharmacology and the Biologi- cal Sciences Research Center, University of North Carolina School of Medicine, Chapel Hill, NC 27514. Address correspondence to this author at the Biological Sciences Research Center. 2Depastments of Psychiatry and Pharmacology, Duke University Medical Center, Durham, NC 27701. 3Nonstandard abbreviations: HPLC-EC, “high-performance” liq- uid chromatography with electrochemical detection; 5-HT, 5-hy- droxytryptamine (serotonin); 5-HTP, 5-hydroxytryptophan; 5- HIAA, 5-hydroxy-3-indoleacetic acid; and NMHT, N-methyl-5-hy. droxytryptainine. Received June 24, 1985; accepted August 15, 1985. the extent that only a standard curve and no more than a few samples can be analyzed in a working day. Very few good research questions about the neurochemical responses to pharmacological, environmental, or other manipulations are so unfocused as to need the simultaneous separation and detection of a large number of neurotransmitters, their precursors, and metabolites. Rather, we recommend the use of a more practical approach: optimize the chromatographic and detector selection conditions to permit the rapid and simultaneous measurements of only a few (e.g., two or three) selected compounds. The advantages of this approach are several: higher signal-to-noise ratios (and therefore greater sensitivity), improved internal standardization, less chance of intra-assay variability, and, finally, a smaller requirement for absolutely consistent, optimal chromato- graphic performance. Thus, the ability to quanti1 dozens of compounds simultaneously, often with different degrees of accuracy, is seldom actually an advantage, despite its al- lure. We do not, however, minimize the importance of screening endogenous neurochemicals for potential chro- matographic interference. Here, our purpose is to outline two methods that we have used routinely, and to discuss several of the variables that permit these methods to be tailored to particular applications. This flexibility repre- sents the greatest strength of HPLC. The focus of this discussion is the neurotransmitter 5-HT, which, fortunately for the neuroscientist, involves a relatively simple synthesis and metabolism (see Figure 1). Materials and Methods Quantitative Determination of Suprananogram Amounts of 5-HT, 5-HIP, and 5-HIAA in Various Matrices (3) Chromatographic conditions. For these studies, we used a HG or ifi solvent-delivery pump (Laboratory Data Control, Riviera Beach, FL), a six-port rotary injection valve (Model 7125; Rheodyne, Berkeley, CA), and a flow rate of 1.5 mL/min for all mobile phases. A typical detection system would consist of a glassy-carbon working electrode and an LC-4 amperometric controller, both from Bioanalytical Systems, West Lafayette, IN, with detector potential maintained at 0.72 V against an Ag/AgC1 refer- ence electrode. The chromatographic separations reported here were per- formed with one of several steel analytical columns, 250 to 300 x 4.6 mm (i.d.), packed with octadecyl (C18) bonded microparticulate (5 pm) silica gel. The mobile phase was a mixture of a solution containing, per liter, 0.1 mol of citric acid, 75 mmol of Na2HPO4, and 0.75 mmol of sodium 1- heptanesulfonate, and methanol (100-140 mL’L). After adjusting the pH to 3.8-4.3 (depending on the column), we filtered the mobile phase through a ifiter of 0.45-pm av. pore size (Gelman Sciences, Ann Arbor, MI), and degassed it under reduced pressure and with ultrasonic agitation. Sample preparation. To the tissue sample (e.g., brain, uterus, lung), a fixed amount (e.g., 10 or 25 ng) of the internal standard, N-methyl-5-hydroxytryptamine