Effect of General Microdialysis-Induced Depletion on Extracellular

estimation of the extracellular dopamine (DA) concentra- tion was evaluated for various degrees of depletion. The basal level of DA in the striatum of...
0 downloads 0 Views 130KB Size
Download PDF
Articles Anal. Chem. 1996, 68, 724-728

Effect of General Microdialysis-Induced Depletion on Extracellular Dopamine Patrick M. Sam and J. B. Justice, Jr.*

Department of Chemistry, Emory University, 1515 Pierce Drive, Atlanta, Georgia 30322

Continuous sampling by microdialysis induces a general depletion of all low molecular weight materials around the microdialysis probe. The effect of this depletion on the estimation of the extracellular dopamine (DA) concentration was evaluated for various degrees of depletion. The basal level of DA in the striatum of male Wistar rats was estimated by the zero net flux method at six different flow rates ranging from 0.3 to 1.6 µL/min. The extracellular DA concentrations in the striatum of male rats measured by this method were found not to differ significantly from one flow rate to another (p > 0.05) with a mean of 6.5 nM ( 0.11. These findings suggest that the estimation of extracellular dopamine by quantitative microdialysis is independent of the flow rate and that the depletion of other extracellular substances around the probe has no effect on the determination of the extracellular concentration of dopamine. Microdialysis has been widely used as a quantitative tool to sample the extracellular fluid of the brain for neurotransmitters and other neuroactive materials.1 It is also used in pharmacokinetics to determine the concentration of drugs in brain,2 blood,3 and subcutaneous tissue.4 As a sampling device, the probe removes materials from the surrounding extracellular fluid. This makes it difficult to directly determine the undisturbed extracellular concentration, since the very act of sampling alters the concentrations in the extracellular fluid. Recently, several mathematical models concerning quantitative aspects of microdialysis have been proposed. Benveniste et al.5 and Amberg and Lindefors6 addressed the diffusion of substances through the interstitial space. Bungay et al.7 and Morrison et al.8 also took into account metabolism, vascular transport, and intra(1) Robinson, T. E.; Justice, J. B., Jr. Microdialysis in the Neurosciences; Elsevier: New York, 1991. (2) Stahle, L.; Segersvard, S.; Ungerstedt, U. Eur. J. Pharmacol. 1990, 185, 187-193. (3) Scott, D. O.; Sorenson, L. R.; Steele, K. L.; Puckett, D. L.; Lunte, C. E. Pharm. Res. 1991, 8, 389-392. (4) Lo ¨nnroth, P.; Jansson, P. A.; Smith, U. Am. J. Physiol. 1987, 253, E228E231. (5) Benveniste, H. J. Neurochem. 1989, 52, 1667-1679. (6) Amberg, G.; Lindefors, N. J. Pharmacol. Methods 1989, 22, 157-183. (7) Bungay, P. M.; Morrison, P. F.; Dedrick, R. L. Life Sci. 1990, 46, 105-119. (8) Morrison, P. F.; Bungay, P. M.; Hsiao, J. K.; Mefford, I. N.; Dykstra, K. H.; Dedrick, R. L. Microdialysis in the Neurosciences; Elsevier: New York, 1991; pp 47-79.

724

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

cellular and extracellular exchange. Together, these models help us to understand factors that govern the spatial and timedependent solute distribution. They also help to establish methods to accurately estimate the extracellular concentration of neurotransmitters and other materials in brain tissue. Several quantitative microdialysis methods have been proposed: It is well-known that depletion of material in the extracellular fluid surrounding the probe is a function of perfusion flow rate. As slower perfusions remove less material, Jacobson et al.9 described a nonlinear regression approach of measuring a substance at different perfusion flow rates and extrapolating the data to zero flow, at which there is no removal. Lo¨nnroth et al. proposed the concentration difference or zero net flux method (ZNF), whose principle has been described elsewhere.4,10 In the internal reference technique,11 a radioactive-labeled analogue of a solute of interest is added to the perfusate as an internal standard and its relative loss is observed and then compared to the relative recovery of that solute. The relative recovery of a solute is defined as the ratio of its dialysate concentration to the concentration in the external solution. If the relative loss of the internal standard is known with respect to the relative recovery of the solute, the external concentration can be estimated.11 Among these methods, the internal reference technique is less time consuming, but the labeled and unlabeled species need to be determined separately. Although use of an unlabeled species as an internal standard has been proposed, it requires further characterization.12 The ZNF method has been used for measurement of steady state concentrations.4,10 A series of concentrations of a substance of interest are perfused into the microdialysis probe. The range of concentrations brackets the expected concentration in vivo or exceeds it.13 When the concentration of the perfused substance is lower or higher than that in the extracellular fluid, there will be a net gain or loss from the tissue into the probe. To determine concentration at which there is no net flux of material, a plot is (9) Jacobson, I.; Sandberg, M.; Hamberger, A. J. Neurosci. Methods 1985, 15, 263-268. (10) Lo ¨nnroth, P.; Jansson, P. A.; Fredholm, B. B.; Smith, U. Am. J. Physiol. 1989, 256, E250-E255. (11) Kurosawa, M.; Hallstro ¨m, A.; Ungerstedt, U. Neurosci. Lett. 1991, 126, 123126. (12) Wang, Y.; Wong, S. L.; Sawchuk, R. J. Pharmacol. Res. 1993, 10, 14111419. (13) Thompson, A. C.; Justice, J. B., Jr.; McDonald, J. K. J. Neurosci. Methods 1995, 60, 189-198. 0003-2700/96/0368-0724$12.00/0

© 1996 American Chemical Society

made of the net gain or loss of the analytes (Cin - Cout, where Cin is the concentration of the analyte in the perfusate and Cout is that in the dialysate) versus the concentration of Cin. The Cin at which the regression line intercepts the line of Cin - Cout ) 0 equals the extracellular concentration in vivo.4,10 Although the ZNF method estimates the concentration in vivo by determining the concentration at which there is no depletion of a particular material, it does not take into account the effect of the depletion of other low molecular weight molecules by the microdialysis process. The removal of neurotransmitters, nutrients, and other materials may affect the measured concentration of the substance of interest either positively or negatively. In the present case, such effects may occur through presynaptic receptors for other transmitters on the dopamine nerve terminals or by other actions on dopamine (DA) synthesis, release, uptake, or metabolism. The purpose of this study is to investigate this question by estimating the concentration of DA in vivo as a function of degree of general depletion of material in the extracellular fluid surrounding the probe. If the neurotransmitter concentration is dependent on the removed materials, it should show up as a dependence on the flow rate, which determines the extent of depletion. At very slow flow rates, there is little depletion, but as the flow rate is increased, more material is removed per unit time, increasing the depletion. For all the flow rates, the greatest depletion will occur closest to the dialysis probe. Therefore, any effect will be primarily on the region of the dopamine gradient nearest the probe. By perfusing different concentrations of dopamine through the microdialysis probe at various flow rates and then comparing the observed extracellular DA concentrations as a function of flow rate, the extent that the depletion of other materials affects the extracellular DA concentration can be evaluated. METHODS Materials. All chemicals were obtained from Fisher Scientific except L-ascorbic acid, chloral hydrate, 3-hydroxytyramine hydrochloride (DA-HCl), and sodium octyl sulfate which were obtained from Sigma. All chemicals were used as obtained. The water was filtered and distilled (Corning, Mega-pure). Subjects. Male albino Wistar rats (275-300 g) were housed in a temperature-controlled room with a 12 h lighting cycle (lights were on from 7 a.m. to 7 p.m.). Rats were undisturbed in their home cages prior to the experiment and were housed in groups of three. Apparatus. The high-performance liquid chromatography (HPLC) system consisted of an amperometric detector (EG&G Princeton Applied Research, Model 400) with working electrode (Model MF-1000) and reference electrode (Model RE1) from Bioanalytical Systems Inc. The applied potential was +700 mV versus Ag/AgCl. The system had a sample injection loop of 0.5 µL, and the mobile phase was delivered by a syringe pump (Isco LC-5000) with flow rate of 33 µL/min. The HPLC was equipped with a microbore column (10 cm × 0.5 mm i.d.) with C-18 stationary phase. The mobile phase was composed of 27.2 mM sodium phosphate buffer with 10% methanol (v/v), 4.9 mM triethylamine, 0.13 mM Na2EDTA, and 0.86 mM sodium octyl sulfate at pH 5.4. Chromatograms were collected on chart recorders (Perkin Elmer, Model 56).

The microdialysis probes were constructed in this laboratory as described previously.14 Briefly, the probe had a 2 mm active area and the membrane had a molecular cutoff of 13 000 (Spectrum, Houston, TX). During the perfusion, the inlet was connected to a Hamilton syringe (500 µL) whose flow rates were controlled by an infusion pump (Kd Scientific, Model Kds 100). The dialysates were collected from the outlet line into polyethylene microcentrifuge tubes (250 µL, Baxter Scientific). Surgical Procedure. Rats were anesthetized with 2.42 M chloral hydrate (0.1 µL/100 g of body weight) and placed in a stereotaxic instrument. The probe was lowered to the striatum using the coordinates A +3.4, L +1.5 (from bregma), and V -6.0 (from the skull)14 with the incisor bar set at +5.0 mm. Guide cannula, skull screw, and cranioplastic cement were not used for the surgery. The probe was first immobilized on the vertical bar of the stereotaxic instrument by adhesive tape and held in a straight and vertical position. The tip of the probe was used as a guide to locate the coordinate so that a hole could be drilled and into which the probe was inserted and extended an extra 2 mm. When needed, additional chloral hydrate was injected (0.1 µL) to maintain the animals in a state of anesthesia. Atropine was only administered when rats exhibited difficulty in breathing. All surgical procedures were performed according to the recommendations of Emory University Institutional Animal Use and Care Committee. Histology. At the end of the surgery, rats were deeply anesthetized with chloral hydrate and perfused intracardially with a saline solution, followed by formalin (10%, v/v). Rats were then decapitated and the brains were removed and stored in 10% formalin for later histological analysis. In order to examine the placement of the probe, the brain was mounted on a cryostat, frozen, and sliced into 50 µm sections. The sliced tissues were mounted on slides and stained with thionine in order to be examined. The process of verification was done by comparing the slides under the microscope with a stereotaxic atlas.15 EXPERIMENTAL PROCEDURE (a) In Vitro Experiment. A known amount of dopamine was dissolved into distilled water to make a standard solution (50 µM). From this solution, 50, 25, and 12.5 nM DA solutions were made by serial dilution in artificial cerebrospinal fluid (CSF). The CSF consisted of sodium chloride (149 mM), potassium chloride (2.8 mM), calcium chloride (1.2 mM), magnesium chloride (1.2 mM), ascorbic acid (0.25 mM), and D-glucose (5.4 mM), at pH 7.2-7.4. A 2 mm probe was placed into a stirred 25 nM DA solution which was maintained at 37 °C in a water bath (Haake Instruments, Model E12). The probe was perfused with a standard DA solution (either 0, 12.5, 25, or 50 nM, where 0 nM was the CSF solution without DA). After an equilibration time of 30 min, dialysate collection began. For each concentration, dialysate samples were collected at eight different flow rates (2.2, 2.0, 1.6, 1.2, 0.8, 0.6, 0.4, and 0.3 µL/min) in a random order so that each concentration at each flow rate yielded two samples. The collection intervals corresponding to these flow rates were 3, 3, 4, 4, 5, 8, and 10 min to yield at least 3 µL of sample. For every change of the flow rate, the first dialysate was discarded so that the residual effect of the previous flow rate would not carry over (14) Parsons, L. H.; Justice, J. B., Jr. J. Neurochem. 1992, 58, 212-218. (15) Pellegrino, L.; Pellegrino, A.; Cushman, A. A Stereotaxic Atlas of the Rat Brain; Plenum Press: New York, 1979; p 17.

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

725

to the next sample. At the end of the eighth flow rate, a different concentration of DA solution was randomly chosen and the cycle was repeated for the three remaining concentrations. The dialysates and the standard DA solutions were stored in dry ice until analysis by HPLC. To quantify the concentration of the dialysates, a standard curve consisting of 0, 12.5, 25, and 50 nM DA was constructed. Standards and samples were treated identically and analyzed on the same day. During the analysis, the samples were thawed and 2 µL of each sample was withdrawn into a syringe (Hamilton, gas type, 10 µL) and loaded into the 0.5 µL injection loop of the HPLC. (b) In Vivo Experiment. DA solutions of 30, 20, 15, 10, 7.5, 5, and 2.5 nM were prepared as described in the in vitro experiment. Each rat received only two of these concentrations of DA selected randomly. After inserting the probe into the striatum, the probe was perfused with the DA solution and allowed to equilibrate for at least 1.5 h at a fixed flow rate (either 1.6, 1.2, 0.8, 0.6, 0.4, or 0.3 µL/min). The collection interval and the procedure were the same as described in the in vitro experiment. At the end of the sixth flow rate, the concentration of the perfusate was changed and allowed 15-20 min to equilibrate. Then the procedure was repeated for the remaining concentrations. Data Analysis. The data from both the in vitro and the in vivo experiment were analyzed by linear regression (Origin, 3.0, MicroCal Inc.). The data analysis was performed by graphing [DA]in - [DA]out, the difference between the concentration of DA perfused through the probe ([DA]in) and that obtained from the dialysate ([DA]out), versus [DA]in (see Figure 1). Linear regression of these data yielded a slope (recovery) and an intercept (-[DA]out at 0 [DA]in). To calculate the extracellular concentration, the linear function, y ) ax + b was solved for x at the point of no net flux, that is, when y ) 0. Therefore, [DA]ext is equal to xy)0 ) -b/a, where [DA]ext is the extracellular concentration and b is -[DA]out at 0 [DA]in. Error for the probe recovery was obtained directly from the regression statistics. The equation SEMext ) [(Sb/b)2 + (Sa/a)2]1/2 was used to calculate the error for the extracellular DA concentration.16 SEMext is the standard error of the mean of the extracellular DA concentration, Sb is the standard deviation of the y intercept, and Sa is the standard deviation of the slope (recovery). In all cases, error was reported as the standard error of the mean (SEM). The estimated extracellular DA concentrations (zero points) were then plotted against the flow rates, and a linear regression of these data was performed. The slope of this regression line was tested for a significant difference from zero. RESULTS AND DISCUSSION In Vitro Experiment. Figure 1 shows the in vitro estimation of DA concentration in a stirred 37 °C solution. Data points above the x intercept represent a region where the concentrations of the perfused DA exceed that in the standard solution and diffuse into the solution. Similarly, data points below the x intercept indicate a region where the concentrations of the perfused DA are lower than that in the medium and result in a diffusion of the external DA into the dialysate. The estimated concentration of DA was determined at eight different flow rates. Linear regression yielded a straight line for (16) Young, H. D. Statistical Treatment of Experimental Data; McGraw-Hill: New York, 1962; Chapter 13.

726 Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

Figure 1. Effect of perfusion flow rates on the estimation of dopamine concentration in vitro. A 2 mm probe was placed in a stirred, 37 °C solution containing 25 nM DA. Four concentrations of DA (0, 12.5, 25, and 50 nM) were perfused into the probe. Each concentration was perfused at eight flow rates: 2.2 (0), 2.0 (b), 1.6 (2), 1.2 (1), 0.8 ([), 0.6 (]), 0.4 (O), and 0.3 µL/min (4). The extraction fractions of these eight flow rates were 10.2 ( 0.01%, 14.4 ( 0.01%, 17.4 ( 0.02%, 24.5 (0.03%, 29.0 ( 0.02%, 34.8 ( 0.03%, 46.5 ( 0.02%, and 57.5 ( 0.02%, respectively; n ) 5-7. Inset: zero point concentrations as a function of flow rate. The linear regression of the zero point concentrations indicates that there is no relation between the zero point and the flow rate (slope, -0.07 ( 0.42; r2 ) 0.005). The mean of the estimated concentrations is 25.6 ( 0.3 nM (n ) 8). The concentration of DA in the solution was determined to be 27.1 ( 0.6 nM (n ) 6) by direct HPLC injection. Table 1. Recovery as a Function of Flow Rate in in Vivo and in Vitro Experiments flow rate (µL/min) 2.2 2.0 1.6 1.2 0.8 0.6 0.4 0.3

in vitro recovery % 10.2 14.4 17.4 24.5 29.0 34.8 46.5 57.5

SEM 0.01 0.01 0.02 0.03 0.02 0.03 0.02 0.02

in vivo recovery %

29.7 35.6 40.9 49.9 61.5 69.8

SEM

0.01 0.02 0.02 0.03 0.02 0.03

each flow rate (r2 ) 0.99 for all eight flow rates). The linearity of the regressions indicates that the diffusion into and out of the probe is symmetric. The slopes (extraction fraction or recovery) of the regression lines are listed in Table 1. The data clearly indicate that the extraction fraction decreases as the flow rate increases. This is consistent with previous findings.9 Each of these regression lines produced a similar DA concentration (see Figure 1). The mean concentration was 25.6 ( 0.3 nM. At the end of each experiment, 6 µL of solution was withdrawn and analyzed. The mean of the DA concentration in the medium was

Figure 3. In vivo concentration of DA as a function of flow rate. The data indicate that the extracellular DA concentration is independent of the flow rate of the perfusion (r2 ) 0.22). The slope of the regression line was -0.12 ( 0.3. The DA concentrations were 6.7 ( 0.04, 6.3 ( 0.05, 6.1 ( 0.06, 6.5 ( 0.07, 6.7 ( 0.05, and 6.8 ( 0.06 nM for 1.6, 1.2, 0.8, 0.6, 0.4, and 0.3 µL/min, respectively, with n ) 4-6. Figure 2. Effects of different flow rates on the estimated extracellular DA concentration in vivo. One group of rats received 2.5, 5, 10, and 20 nM DA in the perfusate. Another group of rats received 7.5, 15, and 30 nM DA in the perfusate. Each concentration of DA was perfused at six different flow rates in random order: 1.6, 1.2, 0.8, 0.6, 0.4, and 0.3 µL/min. The extraction fractions of flow rates 1.6 (9), 1.2 (O), 0.8 (2), 0.6 (4), 0.4 ([) and 0.3 µL/min (0) were 29.7 ( 0.01% (mean ( SEM), 35.6 ( 0.02%, 40.9 ( 0.02%, 49.9 ( 0.03%, 61.5 ( 0.02%, and 69.8 ( 0.03%, respectively (n ) 4 except with 7.5, 15, and 30 nM, n ) 6).

27.1 ( 0.6 nM (n ) 5), which was not significantly different from the mean estimated concentration. In the inset to Figure 1, the in vitro zero point concentrations were plotted against the flow rates. Linear regression analysis yielded a slope of -0.07 ( 0.42 with coefficient of determination (r2) of 0.01. This finding demonstrates that the flow rate by itself does not affect the estimated zero points. In Vivo Experiment. The in vitro experiment illustrated that the concentration could be reproducibly estimated in vitro regardless of the flow rate. For the in vivo experiment, six flow rates with seven [DA]in each were used. The correlation coefficient for the data at each flow rate was 0.99 (see Figure 2). The extraction fractions are listed in Table 1. In Figure 2, all regression lines are linear and intercept at similar points. The estimated extracellular DA concentrations have a mean of 6.5 ( 0.11 nM (n ) 6) and agree with earlier estimates. For DA in the striatum, this laboratory previously reported 5.9 ( 0.1 nM DA.17 Figure 3 shows that regression of the concentrations against the flow rate has a slope of -0.12 ( 0.3 and r2 of 0.05. The slope was not significantly different from zero. The low r2 value indicates that there is no dependence on the flow rate, suggesting that the depletion of low molecular weight materials does not affect the concentration of DA. These in vivo results are discussed first with respect to dependence of the zero point concentration on flow rate and then with respect to the linearity of the regressions at each of the flow rates. The zero point concentration of DA could decrease with increasing depletion if precursor availability affected it. The (17) Olson, R. J.; Justice, J. B., Jr. Anal. Chem. 1993, 65, 1017-1022.

continuous microdialysis depletes the precursors tyrosine and L-dihydroxyphenylalanine (L-DOPA), so that synthesis and release

of DA could be affected. Tyrosine is converted into L-DOPA by tyrosine hydroxylase (TH) in the rate-limiting step in the pathway of DA synthesis. Depleting extracellular tyrosine could affect this step. However, under normal conditions, the tissue levels of tyrosine are well above the Km for TH.18 Therefore, it is not surprising that its removal from the extracellular fluid has no observable effect on extracellular DA. The release of DA is not only triggered by presynaptic firing but is also regulated by presynaptic receptors of other neurotransmitters such as acetylcholine (ACh),19 glutamate (GLU),20 glycine,21 γ-aminobutyric acid (GABA),22 norepinephrine,23 serotonin (5-HT),24 opiates25 and other peptides such as substance P, prolactin, somatostatin, and angiotensin.26 If the depletion of these neurotransmitters in the extracellular fluid affected the release of DA, then the extracellular DA concentration would be affected. However, this effect was not observed. The heterogeneity of the brain microenvironment may contribute to the reason that microdialysis-induced depletions have no effect and high concentrations of transmitters are required to induce an effect. Application of these transmitters, as well as depletion of them, occurs in the extracellular fluid, where the concentrations are relatively low compared to the high concentrations in the synapse during release. Receptors located in the (18) Cooper, J. R.; Bloom, F. E.; Roth, R. H. The Biochemical Basis of Neuropharmacology; Oxford University Press: New York, 1991; Chapter 10. (19) Besson, M. J.; Cheramy, A.; Feltz, P.; Glowinski, J. Proc. Natl. Acad. Sci. U.S.A. 1969, 62, 741-748. (20) Moghaddam, B.; Gruen, R. J.; Roth, R. H.; Bunney, B. S.; Adams, R. N. Brain Res. 1990, 518, 55-60. (21) Giorguieff-Chesselet, M. F.; Kemel, M. L.; Wandscheer, D.; Glowinski, J. Eur. J. Pharmacol. 1979, 60, 101-104. (22) Giorguieff, M. F.; Kemel, M. L.; Glowinski, J.; Besson, M. J. Brain Res. 1978, 139, 115-130. (23) Reisine, T. D.; Chesselet, M. F.; Lubetzki, C.; Cheramy, A.; Glowinski, J. Brain Res. 1982, 241, 123-130. (24) Benloucif, S.; Keegan, M. J.; Galloway, M. P. J. Pharmacol. Exp. Ther. 1993, 265, 373-377. (25) Pollard, H.; Llorens, C.; Schwartz, J. C. Nature 1977, 268, 745-747. (26) Starr, M. S. Neurochem. Int. 1982, 4, 233-240.

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

727

synapse thus experience very high concentrations when transmitter release occurs. The extracellular concentration is much lower, resulting from diffusion of the transmitters out of the very small volume of the synapse into the relatively large volume of the extracellular space, where substantial dilution occurs.27 Thus, the extracellular concentration is relatively small compared to what is required for synaptic receptor mediated enhancement or inhibition of DA release. Local application of these neurotransmitters elicits an increase in the release of striatal DA, but high concentrations are required. For example, application of GLU lower than 10 mM had no effect on the release of striatal DA in vivo.21 In vivo application of 5-HT below 0.1 µM had no effect on dialysate DA.28 The GABA concentrations required to increase DA release were in the range of 10-5-10-3 M.22 For a majority of opiate agonists, micromolar concentrations were required.29 For Ach, 0.1 mM was needed.27 These concentrations are far above the low nanomolar extracellular concentrations found for these transmitters. To affect receptors within the synapse apparently requires concentrations much higher than that found in the extracellular fluid. Similarly, (27) Garris, P. A.; Ciolkowski, E. L.; Pastore, P.; Wightman, R. M. J. Neurosci. 1994, 14, 6084-6093. (28) Parsons, L. H.; Justice, J. B., Jr. Brain Res. 1993, 606, 195-199. (29) De Belleroche, J., Ed. Presynaptic Receptors: Mechanism and Function; Ellis Horwood Ltd.: Chichester, England, 1982; Chapter 7.

728

Analytical Chemistry, Vol. 68, No. 5, March 1, 1996

decreasing the extracellular concentration by microdialysis sampling is likely to have little effect on synaptic concentrations since such a decrease has little effect on the difference between the high concentration in the synapse and the already very low concentration in the extracellular fluid. CONCLUSION The results of the present study demonstrate that the estimation of the extracellular concentration of DA is not affected by the microdialysis-induced depletion of other materials. It is not known how general the result is beyond this particular neurotransmitter, but it is likely to hold for similar systems such as norepinephrine and serotonin. ACKNOWLEDGMENT Financial support of this work was provided by NSF IBN 9412703, J.B.J. was supported by Research Scientist Development Award 1-K02-DA00179-01. Received for review July 31, 1995. Accepted October 16, 1995.X AC950754+ X

Abstract published in Advance ACS Abstracts, November 15, 1995.