Free 3-Methoxy-4 ... - Clinical Chemistry

CLIN.CHEM.35/2, 202-205 (1989)

Free 3-Methoxy-4-hydroxyphenylglycolDetermined in Plasma by LiquidChromatography with CoulometricDetection M. Harlharan,1 Ted VanNoord,1

Oliver

G. Cameron,’

George

C. Curtis,’

We describe a simple “high-performance”liquid-chromatographic method for determiningfree 3-methoxy-4-hydroxyphenyigiycol(MHPG) in plasma, with coulometnc detection and with 4-methoxy-3-hydroxyphenylglycol (iso-MHPG) as the internal standard. MHPG and isO-MHPG are extracted from plasma into ethyl acetate and the extract is washed with a sodium bicarbonatesolution,evaporated, reconstituted,and injected into a 25 x 0.4 cm column packed with 3-j.m particiesof C18material.We used a mobilephase of acetate buffer(100 mmol/L, pH 5.0) and methanol(92:8 by vol) and an oxidationvoltageof 0.5 V. The detectionlimitof the assay is 0.1 g/L. The interassay CV for a sample with a mean MHPG concentrationof 3.84 g/L was 5.2%. The average absoluterecoveryfor the method was 37%. Additional Keyphrases: depression

.

monitoring therapy

mental disorders

and David

G. Ostrow2

conjugates). Both the CNS and sympathetic nervous system contribute to plasma ?v1C. The exact contribution of the CNS to the MHPG in plasma and the clinical and diagnostic use of r’sm’G determinations in biological fluids have been reviewed (21). The concentration of free MHPG in plasma is about 1.5 to 5.5 pgfL. This low concentration, the complex nature of its biological matrix, the unusual (hydrophilic) chemical nature of the analyte, the lack of a reliable commercial internal standard, and the limitations of amperometric detectors have all impeded development of a simple HPLC-EC method for the assay. Many of the published HPLC-EC methods have poor sensitivity (13, 16, 17), involve laborious extraction procedures (12, 13, 16, 17, 19), and (or) lack an internal standard (6, 13, 17, 19, 20). Here we describe a simple HPLC-EC (coulometric) method that overcomes these problems, detects as little as 0.1 tg of a&m’o per liter, and requires only 1-mL plasma samples. Materials and Methods

3-Methoxy-4-hydroxyphenylglycol (.srn’a) is a major metabolite of norepinephrine, which is a neurotransmitter in the central nervous system (CNS). Determination of MHPG in plasma is an important tool for assessing the noradrenergic function of the CNS. In the last two decades there has been increased interest in quantifying rssm’o in spinal fluid, urine, and plasma. This compound is of great interest to psychiatrists (a) as a possible marker for depression (1, 2); (b) for the diagnosis of anxiety disorder (3, 4); (c) as a predictor for the outcome of treatment with tricyclic antidepressants (5); (d) for monitoring the treatment effectiveness of monoamine oxidase inhibitors (6); and (e) for monitoring subjects’ response to challenge tests that change the sensitivity of the presynaptic a2 receptor. (7). Plasma is the most convenient specimen for the MHPG determination, being more easily obtained than either a 24-h urine specimen or spinal fluid. Various techniques used for this assay include gas chromatography, with either electron capture detection (8, 9) or mass spectrometric detection (10, 11), and “high performance” liquid chromatography (HPLC), with electrochemical (EC) detection (6, 12-20). Plasma contains free (20-30%) and conjugated s.wr (sulfate and glucuronide ‘Department of Psychiatry, University of Michigan Medical School, 1301 Catherine St., Dock II, MSRB-1, Room A526, Ann Arbor, Ml 48109-0656. 2V.A. Medical Center Mental Health Clinic, 2215 Fuller Rd., Ann Arbor, MI 48109. Presented in part at the 40th national meeting of the AACC, New Orleans, LA, July 1988 (Clin Chem 1988;34:1185, Abstract 158). 3Nonstandard abbreviations: ampo, 3-methoxy-4-hydroxyphenylglycol; iSO-MHPG, 4-methoxy-3-hydroxyphenylglycol; CNS, central nervous system; and EC, electrochemical.

Received October 10, 1988; accepted November 4, 1988.

202 CLINICALCHEMISTRY, Vol. 35, No. 2, 1989

Chemicals 3-Methoxy-4-hydroxyphenylglycol hemipiperazine salt was from Sigma Chemical Co., St. Louis, MO. Ethyl acetate and methanol were “HPLC grade,” sodium acetate and sodium bicarbonate were reagent grade (Fisher Scientific, Fair Lawn, NJ).

Apparatus The HPLC system consisted of a pump (Model L-6000; EM Science, Cherry Hill, NJ), a 25 x 0.4 cm analytical column packed with 3-tm particles of C18 material (also from EM Science), a guard column (from Upchurch Scientific, Oak Harbor, WA) that we packed, a coulometric detector (Model 5100A; Environmental Science Associates (ESA) Inc., Bedford, MA), a high-sensitivity analytical cell (Model 5011; ESA), an autosampler (“WISP,” Model 710B; Waters Associates, Milford, MA), and an integrator-plotter (Model D2000; EM Science). The mobile phase, a 92:8 (by vol) mixture of sodium acetate buffer (100 mmol/L, pH 5.0) and methanol, was run at a flow rate of 0.7 mL/min and a pressure of 21.4 MPa. The analytical and pre-analytical cell voltages of the EC detector were set at + 0.5 and + 0.25 V, respectively.

Standards Prepare the r&su’ stock standard by dissolving 30.8 mg (equivalent to 25 mg of free assn’G) of the hemipiperazine salt in HPLC-grade water, final volume 1 L. Prepare working standards at concentrations of 1,3,6, 12, 15, and 18 pg/L in charcoal-stripped plasma (Biocell Laboratories, Carson, CA). Repeated analyses of the stripped plasma (n = 6) showed it to be free of 1fflPG. The primary stock solution is

to use iso-MRPG as the internal standard, so they quantified stable in the refrigerator (4#{176}C) for a year. The working MHPG by the less-desirable method of external standardizastandards without preservatives are also stable for at least a tion. Huber-Smith et al. (13) also reported that they could year at -70 #{176}C. 3-Hydroxy-4-methoxyphenylglycol (iso-Mm’G), the internot use iso-po as the internal standard because of interference from unknown compounds. This indicates that their nal standard, was synthesized by reducing 3-hydroxy4. methoxymandelic acid with lithium aluminum hydride as extraction procedure did not clean the MHPG extracts suffidescribed by Shipe et al. (22). An iso-rnn’G solution of ciently well and (or) they did not get good resolution with unknown concentration was obtained and stored at -70 #{176}Cthe column (10-m j.Bondapak) and mobile phase (acetate until needed. We diluted this internal standard solution as buffer, pH 4.15, containing 136 mL of methanol per liter) that they used. necessary so that injection of 100 j.tL of it produced a peak height of about 10 cm. This solution is stable for a year at MHPG has three neutral groups (two alcohol groups, one 4#{176}C. methoxy group) and one wealdy acidic (pKa -10) phenolic group. Unlike many common acidic and basic analytes, Patients and ControlSubjects these functional groups of MHPG preclude simple acid/base extraction followed by purification with back-extraction. Depressed patients were taken off all medications for at However, ethyl acetate is an excellent solvent for extracting least two weeks before blood was sampled for rv!m’GdetermiMHPG and the internal standard, iso-r,.un’, from plasma. nation. All the patients had Hamilton depression rating (23) Addition of solid sodium chloride to the plasma enhances scores of 25 or greater and were diagnosed as having ethyl acetate’s extraction efficiency. The extract is further unipolar major affective illness by “SADS-L” interview (24) purified by washing it once with a weak base (sodium and DSM-ffl-R criteria (25) bicarbonate solution). Compared with other methods, such Control subjects were healthy individuals undergoing as those involving solid-phase extraction (6, 13, 15, 20) or employee health-screening examinations and were deterdeproteinization of plasma (12, 16, 20), this extraction mined to be free of any psychiatric illness, past or current, method is simple, cost effective, and gives relatively interby interview. AU subjects, both controls and patients, were ference-free chromatograms, and facilitates the use of the kept on a low-monoamine diet for four days, had fasted for 8 internal standard. This last factor improves the quality of h, and had been supine for at least 45 miii before blood was the work by ensuring reproducibility and compensating for sampled at 0800 hours. The heparinized or EDTA-treated extraction losses. plasma, separated from the blood cells soon after collection, was stored in polypropylene tubes at -70 #{176}C until analysis. Iso-MIIPG, 3-ethoxy-4-hydroxyphenylglycol, and vanillyl alcohol have all been used as internal standards in MHPG Patients’ samples, so stored, were stable for at least a year. assays. The first two compounds must be synthesized and Extraction are not commercially available. In only one publication (18) is vanillyl alcohol used as the internal standard in the Pipet 1 mL of standard, unknown, or quality-control assay, and it is a solid-phase extraction method. Our atspecimen into an 8-mL screw-capped polypropylene tube tempts to use it in ethyl acetate extractions were unsuccess(transport mailing tubes, cat. no. 127-0916-350; Elkay, ful. Many authors have used solid-phase extraction to obtain Shrewsbury, MA), then add 100 L of the iso-inwG solution cleaner chromatograms, but in some of these methods use of and 6 mL of ethyl acetate. Add 400 mg of solid sodium isO-MHPG as the internal standard is precluded (6, 13). chloride to each tube, cap, and tumble-mix the tubes’ However, Semba et al. (15) were successful in using Bond contents by gentle inversion for 20 miii. Centrifuge the Elut (PH) columns for extraction along with iso-rmi’ as the tubes at 2000 x g for 10 mm. Transfer the organic (top) internal standard. Apparently, the nature of the solid-phase layer into a second set of screw-capped polypropylene tubes. extraction material used is critical. Add 1 mL of 100 mmol/L sodium bicarbonate reagent to In most of the previous HPLC-EC work (12,13,15-20) on each tube, cap, and shake the tubes in an Eberbach shaker plasma i,mpo, amperometric detectors were used. These at high speed for 10 mm. Centrifuge the samples for 10 mm detectors are a serious handicap in the assay of picogram’ and transfer the organic (top) layer into 15-mL polypropylnanogram amounts of analytes because only 5% to 10% of ene conical centrifuge tubes (cat. no. 000-0000-15 1; Elkay). analytes are oxidized or reduced by the electrodes of the Evaporate the organic solvent under nitrogen in a water analytical cell. However, in a coulometric detector, because bath at 45 #{176}C. Reconstitute the residue in 200 ,uL of mobile of its inherent design, nearly 100% of the analyte is oxidized phase and inject 40 tL into the HPLC unit with the or reduced (26), and such a detector requires less-frequent autosampler. maintenance. To compensate for the poor oxidation efficienResults and Discussion cy of the amperometric cell, larger sample volumes (16, 19) In developing this assay we concentrated on the following and laborious multiple extractions (12, 13, 16, 17,20) were four aspects: (a) extraction and purification of the analyte resorted to, and other attempts to save the analyte by from its matrix; (b) the use of a good internal standard; (c) silanizing the glassware (12) were also used in the past. use of the more sensitive coulometric electrochemical detecWith a 1-mL sample of plasma and the coulometric detector tor rather than an amperometric detector; and (d) use of a (with the analytical cell set at the optimum +0.5 V and precolumn and mobile phase that provide good resolution and analytical cell set at +0.25 V), the detection limit of our sharp peaks for the analytes. assay is 0.1 gfL, matching or exceeding the sensitivity Recent publications (6, 13) clearly illustrate the practical limits of other HPLC-EC assays of rmpo. For most routine difficulties in duplicating simple HPLC-EC methods (14,15) samples, 0.3 mL of plasma suffices when our extraction for quantifying MHPG in plasma. Gupta et al. (6) developed a method and the coulometric detector are used. The recent simple solid-phase extraction method for extracting and publication (27) of an ultrasensitive HPLC-EC method for purifying i.izu’o from plasma. Analytical recovery of isothe quantification of haloperidol and its reduced metabolite MHPG in their method was very low, making it impractical with use of a coulometric detector rather than an amperCLINICAL

CHEMISTRY,

Vol.35, No.2, 1989 203

ometric detector method is added proof of the superiority of coulometric detection. Figure 1 shows representative chromatograms. The 3-pm C18 column and the mobile phase of pH 5.0 acetate buffer containing 80 mL of methanol per liter provided wellresolved peaks of interest. No endogenous or exogenous compounds interfered in the assay. Extraction of plasma with ethyl acetate and the subsequent washing of the extract with sodium bicarbonate eliminates many possible interferences in the assay. These observations concur with the findings of Semba et al. (15), and Molyneux and Franklin (14), who reported that ethyl acetate extraction eliminates interferences from catecholamines, indoleamunes, and their precursors and metabolites. The use of methanol rather than acetonitrile in the mobile phase and the EM Science C18 column rather than other manufacturers’ columns was crucial for the separation of the unidentified peaks eluting at 11.6 and 19.5 mm from the mpc and isori&iip& peaks. The standard curve is linear from 0 to 20 pg/L. We found it unnecessary to standardize the instrument with a series of standards every time an assay was done. A single 4 pg/L standard per run provided reproducible and accurate quantification. We found that it was not possible to use aqueous standards in our ethyl acetate extraction procedure, as was also noted by earlier workers. A few standardized their assays by direct injection of standards (12, 13), a few prepared standards with dialyzed serum as the matrix (14, 15), a few used additive standardization (16, 18-20), and Gupta et al. used albumin-based standards. We prepared plasma-based standards in charcoal-stripped plasma. Indeed, we routinely use stripped plasma in our laboratory B

A

to prepare standards in (e.g.) assays of nicotine-cotinine, steroids, catecholammnes, haloperidol, and navane, and to perform analytical recovery studies. This plasma, on repeated analyses (a = 6), was free of any MHPG. Table 1 gives details of precision data for two patients’ samples. The average absolute extraction efficiency of our method is 37% (as determined by comparing peak heights of unextracted and extracted samples). Table 2 gives details of analytical recovery for mp-supplemented samples. We have assayed >1000 samples during the past year and can say unequivocally that the assay is simple and reproducible. As many as 40 samples can be assayed in a day with use of an autosampler. We determined the MHPG in plasma of 20 depressed patients and 20 age- and sex-matched controls (Table 3). The mean value for free rm for depressed patients was lower than the mean value for the matched controls, both males and females (t-test, P