ANALYST, FEBRUARY 1993, VOL. 118 123 Tutorial Review Multi-electrode Array Detectors in High-performance Liquid Chromatography: A New Dimension in Electrochemical Analysis C. N. Svendsen" ESA Analytical Ltd., 7 Cromweli Mews, St. Ives, Huntingdon, Cambridgeshire, UK PE 17 4HJ Liquid chromatography has been evolving rapidly over recent years and now provides a fast, accurate and reliable method for determining the identity and concentration of a wide range of compounds. Recent advances in detector systems include the introduction of a multi-electrode high efficiency electrochemical sensor, which enables compounds to be resolved as a function of their individual oxidation potential. Combining such array sensors with gradient chromatography increases both the range of compounds in samples that can be identified in a short period of time and the confidence with which they can be matched to known standards.This paper introduces the field of three-dimensional chromatography using electrochemical array detectors and provides examples of where this new technology is currently being used. Keywords: Electrochemistry; array detector; coulometry; high-performance liquid chromatography Introduction High-performance liquid chromatography (HPLC) was estab- lished over 15 years ago and has now become the method of choice for measuring a broad spectrum of compounds in a wide variety of samples. This is a result of the speed of the technique, its versatility and the ease with which it lends itself to automation.' A schematic diagram of the basic HPLC system is shown in Fig.1 and consists of: (i) a pump which circulates buffer (mobile phase) under high pressure through the system, (ii) an injector, either automatic or manual, which is used to introduce samples into the mobile phase flow, (iii) an analytical chromatography column which is packed with a material that interacts with compounds in the sample and selectively affects the speed at which they pass through it and (iv) a detection system of a type that can quantify the separated compounds as they elute from the column. Known pure compounds of interest (standards) are injected into the HPLC system and elute from the column as peaks at a specific retention time that is a function of the column packing, the mobile phase and the size and structure of the compounds. Unknown compounds in samples injected into the HPLC system can then be identified and quantified based on matching their elution time and peak area with that of the standards.The most commonly used detector for HPLC is the ultraviolet (UV) photometer. Ultraviolet detectors work by passing light of a specific wavelength through a cell connected to the HPLC system and measuring the absorbance in real I Usample 1- - Solvent Recorder U t Detect0 n U Fig. 1 Schematic diagram of a typical HPLC system * Present address: Department of Experimental Psychology, Downing Street, Cambridge University, Cambridge, UK CB2 3EB. time .2 Compounds which absorb light (chromophores) will increase the absorbance measured by the detector as they pass through it, creating a peak that has an area directly related to the concentration of the chromophores [Fig.2(a)]. The amount of light absorbed is also dependent upon the wavelength used, with each compound having a specific wavelength at which it exhibits maximum absorption [Fig. 2(c)]. The popularity of UV detection stems from its ease of use (UV detectors are generally very stable) and the fact that many compounds absorb light at some wavelength. This Time - Time - 1, , I +80 225 325 Wavelengthtnm +0.40 0 -0.40 Pote ntia IN Fig. 2 A comparison between UV and EC detection. ( a ) Three Compounds detected using a UV photometer where absorbance peaks are measured in real time (0.05 a.u.f.s., 254 nm). Note the large level of noise and small peak height. (b) The same three compounds detected using EC (4 nA full scale, -0.85 V).Note the small noise level and large peak height due to the increased sensitivity of EC detection. (c) A UV spectrum for hydroquinone showing a specific wavelength at which it maximally absorbs light. (d) An EC current voltage curve for hydroquinone which also has a specific voltage at which it oxidizes and produces a current124 Oxidation d B + e Reduction OH 0 Hydroqu i none Uuinone Fig. 3 The principle of EC detection is bascd on oxidation or reduction of compounds under study where, at a certain applied voltage, electrons are either donated or accepted thus producing a current which can be measured (4 Inlet w Reference electrode Outlet electrode Gasket Working electrode (b) 1 [Test electrodes Reference Counter electrodes electrodes Fig.4 The two major types of EC detector. (a) The flat-plate flow-over amperometric detector which oxidizes approximately 5% of the compounds flowing over it. ( h ) The porous flow-through coulometric electrode with a large surface area that makes it coulometrically efficient and enables 100% oxidation of compounds flowing through it review focuses on another type of detection system for HPLC known as electrochemical (EC) detection. Instead of passing light through a cell, EC detectors apply a voltage at an electrode surface over which the HPLC eluent flows. Elec- troactive compounds eluting from the column either donate electrons (oxidize) or acquire electrons (reduce) generating a current peak in real time [Fig. 2(6) and Fig. 31. Importantly, the amount of current generated depends on both the concentration of the analyte and the voltage applied, with each compound having a specific voltage at which it begins to oxidize [Fig.2(d)]. Although fewer compounds will oxidize at a working electrode surface than will absorb light, EC detectors are extremely sensitive, a feature often of primary importance to the analyst using HPLC. There are two types of EC detectors: (i) flat-plate ampero- metric detectors which generally have a glassy carbon surface which oxidizes or reduces only 5% (approximately) of compounds passing over it [Fig. 4(a)] and (ii) porous flow-through amperometric detectors which have a large surface area and oxidize close to 100% of the compound flowing through them [Fig. 4(b)]. When 100% of the compound is oxidized this is referred to as coulometry and in this paper amperometric detectors operating at close to 100% efficiency are termed to be coulometric.This difference in the amount of analyte oxidized at the electrode surface leads to a number of advantages with regard to sensitivity and selectiv- ity394 and is significant when considering EC array detectors as discussed below. ANALYST, FEBRUARY 1993, VOL. 118 Fig. 5 Three-dimensional representations of UV and EC array detector chromatograms. ( a ) A typical chromatogram from a UV diode array detector where peaks can be identified by both their retention time and absorbance characteristics at various wavelengths. Note that there is not true separation in the wavelength axis. ( h ) How the same hypothetical sample may look using an amperometric EC array detector.Note that once the oxidation potential for the compound is reached, all subsequent electrodes in the array give a similar peak thus obscuring resolution in the z-axis. ( c ) Finally how the same sample would look using a coulometric array EC detector. Note that due to the 100% conversion of compounds at their oxidation potentials there is now full resolution in the electrode axis, revealing some compounds which were previously obscured in both ( a ) and ( h ) Array Cell Concepts in HPLC When matching standard peaks with unknown peaks in complex sample matrices, there are often a large number of endogenous interfering substances present which elute at the same time as, or close to, the peaks of interest.Changing the mobile phase, column, flow rate or sample preparation technique may help to change the elution time or remove interfering compounds, but there is the consistent worry that a single peak may be contaminated by closely related sub- stances. For many years, the idea of using two different wavelengths and assessing the absorbance ratio between these for both standards and unknown cornpounds in samples has been of great interest as it would give more information on the purity of a single peak eluting from the column .s This idea led to the development of diode array UV detectors whichANALYST, FEBRUARY 1993, VOL. 118 125 perform multiwavelength scanning in real time6 resulting in a third chromatographic dimension; the wavelength (z) axis [Fig.5(a)]. Thus, in addition to using retention time, ratioing across different wavelengths in the z-axis and comparing standards with unknowns is a powerful tool for confirming the identity of a compound. Recent technology advances in the design of UV detectors have increased both their sensitivity7 and, by using algorithms8 or certain types of spectral suppression,9 the accuracy with which compounds can be identified. However, compounds with different absorbance wavelengths eluting from the column at the same time cannot generally be resolved in the z-axis and, therefore, the separation power of the column is not enhanced by using diode array systems [Fig. 5 ( a ) ] . A similar type of array detector has been developed using EC detection where voltage, rather than wavelength, is the third dimension.10 A typical chromatogram from an amper- ometric EC array detector is shown in Fig. S(b) where, following elution from the columm, compounds are intro- duced to a series of electrodes set at incrementally higher voltages. Although there is a specific potential, and thus electrode, where compounds begin to oxidize, they continue to be detected on all subsequent electrodes, because only 5% of the analyte is oxidized at each point of the array. This reduces the resolution of amperometric detectors and other methods of detection currently under development which use low efficiency sensors, such as microelectrode array systems. 11 An alternative approach is to use coulometric electrodes which, as mentioned previously, convert close to 100% of the analyte passing through them to the oxidized state.Therefore, in a serial array of coulometric electrodes set at incrementally higher voltages, compounds will be detected at a certain electrode depending on their individual oxidation potentials and once fully oxidized will be ‘invisible’ to electrodes further up the array. In contrast to the UV diode array and amperometric EC detectors described above, the coulometric array allows full resolution in the electrode axis enabling the separation of compounds co-eluting from the column but with different oxidation potentials [Fig. 5(c)]. Other EC technologies are also being developed for HPLC using a single electrode, which consistently performs sweeps through a specific voltage range (rapid scanning voltammetric detection), thus increasing the confidence of peak confirma- tion.” Very recently, a microelectrode array has been developed where up to 80 individual electrodes in a space of a few millimetres can be held at incremental voltages and the current measured at each in real time.13 Although rapid scanning voltammetry has exciting possibilities when used in situ to measure endogenous levels of electroactive com- pounds, both this method and the microelectrode array use amperometric detection and so are unable fully to resolve co-eluting compounds from the HPLC column in the electrode axis.Coulometric electrodes have a further advantage over amperometric electrodes with regard to their use in array systems. The flow-through porous-carbon graphite coulo- metric electrodes are contained in a sealed unit and require no maintenance for extended periods of time.As they have far more surface area than is required to oxidize normal levels of compounds passing through them, up to 95% of their surface can be contaminated with no loss in response. Should they eventually become contaminated by more than 95% (which may occur any time between one and three years of continual use provided multiple filters are used), they can be removed from the system and flushed with nitric acid to remove the contamination. In contrast, amperometric detectors are far less efficient and require frequent cleaning (sometimes every week) of their glassy carbon surface to avoid a loss in response. Although this may be practical when a single electrode is used, the time taken to clean multiple electrodes is obviously prohibitive.Fig. 6 How ratioing across the electrode axis can confirm compound identity. (a) A typical standard EC array chromatogram with two compounds (1,2) showing different oxidation potentials and a specific response at three different voltages. (b) In an unknown sample run under the same conditions as the standard above, two peaks were found with similar retention times to 1 and 2 in the standard. Although peak 1 has an identical relative response at each electrode in the array to that of peak 1 in the standard, peak 2 shows a very different response in the electrode axis. Using this method the analyst is immediately notified that peak 2 in the sample represents either a different compound from that of the standard or a co-elution with another compound in the sample Coulometric electrodes are unique in possessing such selectivity, sensitivity and stability and are the only type of EC detector that have been fully developed for commercial use, currently available as a fully computer controlled package (The Couloehem Electrode Array System or CEAS; ESA, Bedford, MA, USA, and ESA Analytical, Cambridge, UK).The remainder of this review uses the CEAS to describe in more detail the advantages of E C array systems. Benefits of EC Array Detectors There are a number of advantages in using EC array detectors over conventional single-channel detectors relating mainly to their increased resolution and accuracy. Enhanced Resolving Power In HPLC the column and mobile phase should be chosen such that there is maximum separation of all compounds in the sample.However, there are often circumstances where even after optimizing the chromatographic conditions, some con- taminants in the sample co-elute with peaks of interest. Under, these conditions further resolution can often be obtained using coulometric E C array detectors provided the co-eluting compounds have different oxidation potentials. Confirmation of Compound Purity When compounds pass through the E C array they are normally detected on three contiguous electrodes. The first electrode will oxidize a small portion of the compound, the second or dominant electrode oxidizes a large portion of the126 ANALYST, FEBRUARY 1993, VOL. 118 Table 1 Electrode versus time (ET) map and final data for three compounds in a standard and brain sample.( a ) A typical ET map where the response in peak height from each electrode (El-E16) is listed in columns for peaks found at specific retention times (RT) in the standard and the concentration and peak identification (Peak ID) can be entered. Standard ET maps are then merged with sample ET maps (brain) and peaks in the sample can then be matched with known peaks in the standard based on both retention time and oxidation potential. ( b ) Final data reports are generated from the ET maps and include concentration, retention time, peak height and ratio across the dominant and two sub-dominant electrodes (see text for details on ratios) ( a ) Electrode versus time map for three compounds in a standard and brain sample- Sample ID Peak ID RT/min Concentratiodng E l E2 E3 E4 E5 E6 E7 E8 E9 El0 El1 El2 El3 El4 El5 El6 Standard 4.56 TYr 10 - - - - - - - - - 170 2 134 10 324 5 032 204 - - ( b ) Final data report for sample brain- Concentration/ Compound ng ml-1 Tyrosine 21.48 Dopamine 0.108 HVA 0.510 Brain Peak 1 4.52 21.49 - - - - - - - - - 356 4 865 22 190 4 890 398 - - RT/ min 4.52 6.33 12.44 Standard DA 6.33 1 - 20 1 320 4 700 2 234 269 - - - - - - - - - - Brain Peak 4 6.39 0.108 - 120 508 240 23 - Standard HVA 12.31 1 - - - - - - 45 2 320 9 870 2 500 120 - - - - - Brain Peak 7 12.44 0.501 - - - - - - 20 2 300 5 043 1 022 L30 - - - - - Ratio accuracy Height/ E13/E12 pixels Ell/E12 22 190 94% 91% E3/E4 EYE4 E8/E9 E10/E9 508 84% 99% 5 043 52% * 82% * Although the ratios for tyrosine and dopamine were close to loo%, one of the ratios for homovanillic acid (HVA) was 52%, suggesting that either this was not pure HVA or there was co-elution with another compound using this particular method.Table 2 Areas of study currently under investigation using EC array detectors Area of study ( a ) Group I . Compound related areas- Analysis of neurotransmitters and Analysis of pharmaceuticals Explosive analysis Analysis of phenolic compounds in Analysis of isocyanates in polymers Amino acid analysis Analysis of microdialysates Alzheimer’s disease Huntington’s disease Parkinson’s disease Stroke and excitotoxicity Alcohol abuse Arthritis Epilepsy Motion sickness (c) Group 3. General- Invertebrate biology Circadian rhythms Pediatric neurology modulators beverages ( b ) Group 2.Disease related areas- References 14,15,16,17,18,19 20 21 22 23 24 25,26 27,28,29,30 31,32,33 34,35,36,37 38,39,40,41 42 43 44,45 46 47,48,49 50 51.52 compound and the third electrode oxidizes the remainder of the compound. Pure standards eluting at a given retention time will give a predictable response at all three electrodes and the ratio across these three electrodes remains constant and is independent of concentration [Fig. 6(a)]. In real samples, compounds eluting at a retention time matching that of the standard can also be ‘ratioed’ across three electrodes. If the ratio of the unknown matches the ratio of the standard this can help to confirm the identity of the compound as well as that of the standard [Fig.6(b), peak 11. Any difference in the ratio signifies that either there is some contamination (co-elution) of this compound with another unknown compound or that this compound is not the same as the standard compound [Fig. 6(b), peak 21. Using ratioing methods, accurate peak identifi- cation is possible which gives the analyst a higher degree of confidence in the identity of the compound under study. Increased Stability at High Voltages Working at high oxidation potentials often results in high background currents and decreased sensitivities. Unfortu- nately, some important compounds will only oxidize at high potentials. However, with an EC array, the electrodes set at high voltages are ‘buffered’ by the lower electrodes in the series which results in reduced background current which in turn increases sensitivity and stability. Simplified Sample Preparation Certain samples ( e .g . , plasma and urine samples) require extensive and time consuming clean-up procedures t o be carried out before being introduced into the HPLC system in order to remove interfering compounds. Although it is always important to remove protein and other possible contaminants, which may affect the HPLC system, from samples before injecting them, the added resolving power of EC array detectors can often separate interfering compounds from the compound of interest. Hence, in many instances sampleANALYST, FEBRUARY 1993, VOL. 118 127 clean-up procedures can be simplified and the possibility of artificially changing the concentration of compounds under investigation can be minimized.Data Processing With EC Array Detectors Although having a large number of detectors in series dramatically increases the amount of information about a compound of interest, it also produces escalating amounts of data which have to be simultaneously monitored and inter- preted. In single-electrode systems the information from the detector is normally fed into an integration system, often PC based, using a hard disk for permanent storage. Once stored on the hard disk chromatograms can then be analysed later using peak detection algorithms. The CEAS system allows between 8 and 16 electrodes to be connected in series. The signal from each electrode is stored simultaneously in real time on a computer hard disk and is later integrated using standard peak detection algorithms.Finally, each peak is converted into an electrode versus time (ET) format where its position is plotted with regard to time and oxidation potential. An ET map from a standard and unknown sample can then be merged Fig. 7 gradient from 1% mcthanol to 40% methanol in a phosphate (0.1 mol I - l ) buffer with ion pairing (pH 3.43. For abbreviations see Table 3 EC array chromatogram of a 30 component standard used to measurc simultaneously a wide ran e of neuroactivc substances using a Table 3 List of compounds and abbreviations for the standard chromatogram shown in Fig. 7 Compound name Compound Oxidation abbreviation pontential/mV Dihydroxyphenylacetic acid DOPAC 150 Dihydroxyphenylethyleneglycol DOPEG 180 L-Dopa LD 150 Dopamine DA 150 Epinephrine E 180 Guanine GAN 700 Guanosine GSN 840 Homovanillic acid HVA 450 Hydroxyindoleacetic acid HIAA 180 Hydroxyphenylacetic acid HPAC 650 Hydroxyphcnyllactic acid HPLA 650 Hydroxybenzoic acid HBAC 700 Hydroxytryptophan HTP 300 Kynurenine KYN 800 Melatonin MEL 600 Metenephrine MN 480 Mcthoxyhydroxyphenyl glycol MHPG 450 Methoxyt yramine MT 450 N-Mct h y lserotonin NMET 300 Norepinephrine NE 180 Normetanephrine NMN 480 Salsolinol SA 180 Octopamine OCT 620 Serotonin HT 180 Tryptophan TRP 600 Tyrosine TYR 650 Uric acid UA 300 Vanillic acid VA 480 Xanthine X 700 Vanillylmandelic acid VMA 300 Retention time/min 8.91 1.95 5.61 12.13 8.35 5.65 5.89 12.97 9.95 10.75 13.32 9.32 11.42 13.82 17.36 10.98 4.95 16.48 12.13 5.16 12.87 12.98 7.83 15.46 19.11 8.94 1.62 12.03 2.97 2.02128 ANALYST, FEBRUARY 1993, VOL.118 and peaks matched based on both retention time and oxidation potential [Table l(a)]. The concentration and confidence of matched peaks can then be established and reported in a final format as shown in Table l(6) where peak purity is defined as the ratio of the standard peak divided by the ratio of the unknown sample peak multiplied by 100%. A ratio of 100% indicates that the standard and unknown sample peaks matched perfectly across the electrodes (identical ratios) and any lower percentage gives an indication of the degree of confidence in the match. Practical Applications for EC Array Detectors Any HPLC application with EC detection would benefit from using an array of coulometric cells, but in particular those requiring either a high level of confidence in peak purity or the simultaneous separation of a large number of compounds are most suitable.It would be beyond the scope of this review to cover extensively all the areas of analytical science currently using EC array cell techniques. Therefore, some of the major fields are summarized in Table 2 with relevant references. This final section summarizes the use of the E C array in the neurosciences and the pharmaceutical industry where it is often necessary to measure accurately trace amounts of neuroactive compounds or drugs in various tissues (for an extensive review see ref. 53). Neurosciences The brain is a chemical factory constantly manufacturing, amongst many other things, neurotransmitters, neuromodula- tors and hormones.Views into this world of neurochemistry using HPLC are limited by the resolution of current instru- ments which are often only able to separate and detect one or two compounds simultaneously. Once separated, the only criterion on which the identitity of the compound is based is its retention time which may be identical with another closely related compound eluting from the column at the same time. Fig. 7 shows a chromatogram generated using the CEAS array system which resolves and provides ratio values for 30 compounds of general interest to the neuroscientist (see Table 3) within 35 min and provides oxidation potentials for each. Using this standard method, tissue samples, including brain, cerebrospinal fluid and microdialysates, can be rapidly analy- sed for all of these compounds where they are above the sensitivity limits of the detector (normally around 2 pg on- column).Note that many of the primary metabolites of dopamine and noradrenaline have oxidation potentials higher than their parent compounds and can, therefore, be separated in the z-axis even when co-eluting in time (for example noradrenaline and methoxyhydroxyphenyl glycol). Recently, more advanced techniques have been used, which permit the concurrent measurement of monoamines, their metabolites and derivatized amino acids using column switching coulome- tric array te~hnology.-~4 Pharmaceuticals Fig. 8(a) shows a drug standard which had previously been analysed using both a radioimmunoassay (RIA) and a single-channel HPLC assay with differing results.The HPLC data suggested that the concentration of the drug in patient plasma was twice as high as suggested in the same plasma using an RIA technique. The pure drug standard had a very clear spread across three of the EC electrodes which enabled a ratio to be generated [Fig. 8(a)]. When the patient plasma was measured there was only one peak at approximately the correct retention time but the ratio was very different to that of the standard, suggesting that there may be co-elution. Using a sharp gradient (where the mobile phase constituents are changed during the run) it was possible to separate a 0 2 4 6 b) D I,/ I/ 0 2 4 6 8 Time/m in Fig. 8 EC arra detectors and drug analysis in complex matrices. ( a ) Drug standard [D) showing a characteristic and specific distribution over three electrodes in the z-axis (1.2,3).(b) Drug eak in a patient’s plasma showing the separation of a metabolite (My from the parent compound (D). Note how the ratio across the electrode axis is identical with that seen for the standard confirming that no other metabolites were co-eluting under these chromatographic conditions (see text for details). The response (peak height) is measured as currentANALYST, FEBRUARY 1993, VOL. 118 hidden metabolite (M) away from the parent compound (D), which now had the correct ratio when compared with the standard [Fig. 8(b)]. The ratioing feature of coulometric EC array systems can immediately identify situations where drug metabolites may be interfering with the accurate monitoring of the parent compound.Furthermore, complex mixtures of drugs with different oxidation potentials can be resolved in a short period of time using EC array systems and in some cases both drug and neuroactive compound can be measured simultaneously, significantly enhancing the quality of data in pharmacokinetic studies of drug metabolism. Conclusions Electrochemical array detectors provide a new analytical tool for resolving and accurately detecting trace amounts of any electroactive compound in a wide range of samples. Coulo- metrically efficient electrodes allow full resolution in the third dimension and using a simple electrode versus time mapping technique large amounts of data can be analysed rapidly with compounds of interest quantified and verified simultaneously.This provides a significant advantage over current single- channel detection methods and should soon become the method of choice for a wide range of analysts using HPLC. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 References Snyder, L. R., and Kirkland, J . J., Introduction to Liquid Chromatography, Wiley, New York, 1974. Henry, R. A., and Sivorinovsky, G., in Clinical Chromato- graphy in Clinical Analysis, eds. Kabra, P. M., and Marton, L. J., The Humana Press, Clifton, NJ, 1981, p. 21. Roe, D . K., Anal. Lett., 1983, 16, 613. Schieffer, G. W., Anal. Chem., 1985, 57, 968. Yost, R., Stoveken, J., and McLean, W., J. Chromatogr., 1977, 134, 73. Saitoh, R., and Suzuki, H., Anal. Chem., 1979, 51, 1683. Dong, M. W., Passalacqua, P.V., and Choudhury, D. R., J. Liq. Chromatogr., 1990, 13, 2135. Keller, H. R., Massart, D. L., Liang, Y. Z . , and Kvalheim, 0. M., Anal. Chim. Acta, 1992, 267, 63. Fabre, H., and Fell, A. F., J. Liq. Chromatogr., 1992,15,3031. Matson, W. R., Gamache, P. H., Bcal, M. F., andBird, E. D., Life Sci.. 1987, 41, 905. Atsushi, A., Masue, T., and Uchida. I., Anal. Chem., 1992,64, 44. Gunasingham, H., and Fleet, B., in Electrochemical Chemistry, ed. Bard, A. J., Marcel Dekker, New York, 1989, vol. 16, pp. 89-180. Atsushi, A., Matsue, T., and Uchida, I., Anal. Chem., 1992,64, 44. Matson, W. R., Langlais, P., Volicer, L., Gamache, P. H., Bird, E., and Mark, K. A., Clin. Chem., 1984, 30, 1477. Ikeda, R., Hikoya, S., Hikiba, S . , Ikeuchi, N., Tezuka, T., Sakaue, N., Shimizu. M., and Miura, S ., Jpn. J. Psychiatry Neurol., 1988,42, 675. Ikeda, R., Svendsen, C. N.. Shimizu, M., and Miura, S . , Jpn. J. Psychiatry Neurol., 1989, 43, 731. Takeda, H., Shibuya. T., and Matsumiya, T., J. Pharmacobio- Dyn., 1989, 12, 3. Volicer, L.. Matson. W. R., Schnepper, P., Gamache, P., Datz, D. I., and Wolf, N., Brain Res., 1990, 526, 169. Rizzo, V.. Melzi D’Eril. G., Achilli, G., and Cellerino, G. P., J. Chromatogr., 1991, 536, 229. Gamache, P., CEAS Report 1050, 1992, Available from ESA. Achilli, G., and Cellerino, G. P., CEAS Report 1037, 1991, available from ESA, Bedford, MA. Achilli, G., and Cellerino, G. P., CEAS Report 2033. 1992, available from ESA, Bedford, MA. 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 129 Achilli, G., and Cellerino, G.P., CEAS Report 1012, 1991, available from ESA, Bedford, MA. Langlais, P. J., Mair, R., Anderson, C. D., and McEntee, W. J., Neurochem. Res., 1988, 13, 43. Acworth, I. N., Biol. Psychiatry, 1991, 2, 329. Maruyama, W., Nakahara, D., Ota, M., Takahashi, T., Takahashi, A., Nagatsu, T., and Naoi, M., J. Neurochem., 1992, 59, 395. Volicer, L., Langlais, P. J., Matson, W. R., Mark, K. A., and Gamache. P. H., Arch. Neurol., 1985, 42, 1158. Matson, W. R., Bouckoms, A., Svendsen, C. N., Beal, M. F., and Bird, E. D., in Basic Clinical and Therapeutic Aspects of Alzheimer’s and Parkinson’s Disease, Plenum Press, New York, 1990, vol. 1, p. 513. Conner, D. J., Dietez, S . , Langlais, P. J., andThal, L. J., Exp. Neurol., 1992, 116, 69.Beal, M. F., MacGarvey, U., and Swartz, K. J., Ann. Neurol., 1990, 28, 157. Beal, M. F., Matson, W. R., Swartz, K. J., Gamache, P. H., and Bird, E. D., J. Neurochem., 1990, 55, 1327. Swartz, K. J., Matson, W. R., MacGarvey, U., Ryan, E. A., and Beal, M. F., Anal. Biochem., 1990, 185, 363. Beal, M. F., Matson, W. R., Storey, E., Milbury, P., Ryan, E. A., Ogawa, T., and Bird, E. D., J. Neurol. Sci., 1992, 108, 80. Wolf, J. A., Fisher, L. J., Xu, L., Jinnah, H. A., Langlais, P. J . , Iuvone, M., O’Malley, K. L., Rosenberg, M. B., Shimohama, S . , Friedmann, T., and Gage, F. H . , Proc. Natl. Acad. Sci, USA, 1989, 86, 9011. Ogawa, T., Saso, S . , Beal, M. F., Swartz, K., Matson, W. R., and Bird, E. D., in Basic Clinical and Therapeutic Aspects of Alzheimer’s and Parkinson’s Disease, Plenum Press, New York, 1990, vol. 1, p. 456. Tohgi, H., Abe, T., Kikuchi, T., Takahashi, S., and Nozaki, Y . , Neurosci. Lett., 1991, 132, 19. Ogawa, T., Matson, W. R., Beal, M. F., Myers, R. H., Bird, E. D., Milbury, P., and Saso, S . , Neurology, 1992, 32, 127. Uemura, Y., Miller, J., Matson, W. R., and Beal, M. F., Stroke, 1991, 22, 12. Langlais, P. J., and Mair, R. G., J . Neurosci., 1990, 10, 1664. Conncr, D. J., Langlais, P. J., and Thal, L. J., Brain Res., 1991, 555, 84. Beal, M. F., Swartz, K. J., Finn, S. F., Mazurek, M. P.. and Kowell, N. W., J. Neurosci.. 1991, 11, 147. Makino, Y., Ohta, S . , Tasaki, Y., Tachikawa, O., Kashi- wasakc, M., and Hirobe, M., J. Neurochern., 1990, 55, 963. Godefroy, F., Matson, W. R., Gamache, P. H., and Weil- Fugazza, J., Brain Res, 1990, 526, 169. Mori, A., Suzuki, S . , and Kabuto, H., Med. Sci. Res., 1991,19, 445. Suzuki, S., and Mori, A., Neurochem. Res., 1992, 17, 693. Lucot, J. B., Campton, G. H., Matson, W. R., and Gamache, P. H., Life Sci., 1989, 44, 1239. Shimizu, T., and Takeda, N., 2. Naturforsch., 1991,46, 127. Takeda, N., 2nd Svendsen, C. N., Hydrobiologica, 1991, 216, 554. Shimizu, T., Mihara, M., and Takeda, N., J. Chromatogr., 1991,539, 193. Siuciak, J. A., Gamache, P. H., and Dubocovich, M. L., J. Neurochem., 1992, 58, 722. Langlais, P. J., Wardlow, M. L., and Yamamoto, H., Pediatr. Neurol., 1991,7, 6. Yamamoto, H., Pediatr. Neurol., 1991, 7, 406. Gamache, P., Ryan, E., Svendsen. C. N., Murayama, K . , and Acworth, I. N., J. Chromatogr., in the press. Naoi, M., Murayama, W., Acworth, I. N., Nakahara, D., and Parvez, H., in Techniques in the Behavioral and Neural Sciences, Elsevier, Amsterdam, vol. 10, ch. 1, in the press. Paper 21061 1 OD Received November 17, 1992 Accepted January 8, I993