Increases in Acetylcholine Release Produced by Phorbol Esters Are Not Mediated by Protein Kinase C at Motor Nerve Endings 1

0022-3565/98/2851-0247$03.00/0 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 285, No. 1 Copyright 1998 by The American Society for Pharmacology and Experimental Therapeutics Printed in U.S.A. JPET 285:247 251, 1998 Increases in Acetylcholine Release Produced by Phorbol Esters Are Not Mediated by Protein Kinase C at Motor Nerve Endings 1 T. J. SEARL and E. M. SILINSKY Department of Molecular Pharmacology, Northwestern University Medical School, Chicago, Illinois Accepted for publication December 10, 1997 This paper is available online at http://www.jpet.org Biological responses evoked by active phorbol esters generally have been interpreted as indicative of a role for the enzyme PKC in that biological process (for reviews see Nishizuka, 1988; Newton, 1995; Jaken, 1996). Specifically, phorbol esters have been envisaged as cofactors for PKC that act by binding to the C1 domains (Zn-finger receptors) of the enzyme and increase the affinity of PKC for its target membranes (Newton, 1995). In this regard, phorbol esters have been shown to produce large increases in neurotransmitter release at a variety of synaptic loci in the peripheral and central nervous systems (Shapira et al., 1987; Sebastiao and Ribeiro, 1990; Wardell and Cunnane, 1994; Redman et al., 1997). Although these data may indicate a role for PKC in the process of neurotransmitter release, alternative interpretations are possible. For example, homologous phorbol-binding C1 domains have been found in proteins other than PKC (e.g., diacylglycerol kinase, chimaerin, and unc-13) (Ahmed et al., 1990, 1991, 1992; Maruyama and Brenner, 1991). These findings have led to the suggestion that some of the previously described effects of phorbol esters on neurotransmitter release might Received for publication July 15, 1997. 1 This work was supported by a research grant from the National Institutes of Health (NS12782) ABSTRACT Recent work from our laboratory has demonstrated that phorbol esters known to stimulate protein kinase C (PKC) also stimulate acetylcholine (ACh) secretion by an action at a strategic component of the secretory apparatus [Redman et al. (1997) J Physiol (Lond) 501:41 48]. In an attempt to determine whether the stimulatory effects of phorbols are mediated by PKC, we examined the effects of several PKC antagonists on ACh release promoted by phorbol 12,13-dibutyrate (PDBu) at the frog neuromuscular junction. PKC antagonists that act at the ATP binding site (C3 domain) were examined for their ability to antagonize the stimulatory action of PDBu. Neither the nonselective PKC inhibitor, staurosporine (at concentrations as high as 1 M), nor its more selective derivative, GF109203X (at concentrations as high as 10 M), attenuated the stimulatory effects of PDBu. PKC antagonists that act at the phorbol ester binding site (C1 domain) were examined for their ability to antagonize the stimulatory action of PDBu. Neither sphingosine (500 M) nor calphostin C (25 M) reduced the stimulatory actions of PDBu on ACh release. These results suggest that a presynaptic protein possessing a phorbol ester receptor and not the enzyme PKC is the target site for the stimulatory effects of phorbol esters at motor nerve endings. be caused by mechanisms other than activation of PKC (Brose et al., 1995; Redman et al., 1997). Of particular interest is unc-13, and its mammalian homologue Munc-13; Munc-13 has been implicated as an important component of the secretory apparatus and in the control of cholinergic transmission (Brose et al., 1995). Because preliminary experiments at the frog neuromuscular junction suggested that PKC inhibitors did not block the increases in neurotransmitter release produced by phorbol esters, we decided to explore the possibility that the effects of phorbol esters on transmitter release were in fact mediated through a non-pkc mechanism. To this end we investigated the ability of a range of PKC inhibitors to affect the actions of PDBu on the release of ACh at the frog neuromuscular junction. Materials and Methods Electrophysiological methods. Experiments were performed on the isolated cutaneous pectoris nerve-muscle preparation of the frog (Rana pipiens). Animals were anesthetized with 5% ether, followed by double pithing. The motor nerve was stimulated with supramaximal stimuli of 0.05 ms or less in duration. The electrophysiological correlates of evoked ACh release (EPPs) were recorded from the end-plate regions of skeletal muscle by use of microelectrodes filled with 3 M KCl (resistances ranging from 10 to 20 megohms). After a stable impalement of the skeletal muscle was made, we ABBREVIATIONS: PKC, protein kinase C; PDBu, phorbol-12,13-dibutyrate; EPP, end-plate potential; ACh, acetylcholine; HEPES, N-2-hydroxyethylpiperazine-N -2-ethanesulfonic acid. 247

248 Searl and Silinsky Vol. 285 averaged a series of control EPPs (see below for statistical procedures). The preparation was then superfused with solutions containing PKC reagents (e.g., PDBu, PDBu PKC antagonist, etc.) and EPPs averaged again while maintaining the impalement during the experiment. The signal from the microelectrode was fed into a conventional high-input impedance microelectrode preamplifier (Axoclamp 2A, Axon Instruments Inc., Foster City, CA). Responses were averaged by use of an IBM AT-compatible microcomputer, TL-1 interface and pclamp software (Axon Instruments) or the Strathclyde Software Suite created by Dr. John Dempster. Generally, solutions were delivered by superfusion with a peristaltic pump (Watson-Marlow Inc, Wilmington, MA) at a flow rate 3 to 5 ml/min and removed by vacuum suction. In the experiments with sphingosine and calphostin C, however, the solutions were recirculated. The normal Ringer s solution contained (mm): NaCl, 115; KCl, 2; CaCl 2 1.8; HEPES, 2 (ph 7.2 7.4) and 6 to 7 mg/l tubocurarine chloride to reduce EPPs below threshold for the generation of muscle action potentials. In most experiments, the nerve was stimulated at frequencies ranging from 0.05 to 0.1 Hz. In some experiments, EPPs were measured in Mg -paralyzed preparations in which 3 mm Mg and 0.35 mm Ca were substituted for the normal calcium. In these experiments the nerve was stimulated at frequencies of 0.5 to 1 Hz. Data analysis. Statistical methods were similar to those described previously for quantifying evoked ACh release (see Silinsky, 1981, 1984). In most experiments, after making approximations of the mean number of ACh quanta release, the formula 800/m was used to determine the number of evoked responses to average to reduce the coefficient of variation (i.e., the standard deviation/mean) to 5%. This establishes the control condition and also generally makes statistically significant differences at P.01 (Rahamimoff, 1967; Silinsky, 1984). At normal levels of ACh release, the variability in EPPs is small and only four to eight responses need to be averaged to determine statistical significance. Indeed, the S.E.M. values were generally 5% of the mean in all of the experiments to be discussed. For example, in the experiment of fig. 2, A and B, S.E.M. values ranged from 1 to 2% of the mean. In some experiments, a more complete test of the method was performed as follows. First, the individual data traces were tested for normality and then comparisons between control and treated neuromuscular junctions made by either parametric statistics or nonparametric statistics In the experiments to be described here, an analysis of variance for the normally distributed data was followed, where necessary, by multiple comparisons with use of the Bonferroni inequality. This method is appropriate when no greater than four groups are compared. The Bonferroni method is the most conservative of the multiple comparisons procedures (see Glantz, 1992, page 93). For a detailed discussion of the use of these statistical methods in the construction of doseresponse curves (e.g., fig. 1) and for corrections necessary because of changes in resting potential, see Silinsky (1981). In the Mg -paralyzed preparations, quantal analysis was performed by either the direct method with blocks of 64 averaged EPPs or the method of failures on the individual EPPs (del Castillo and Katz, 1954). In the experiments with d-tubocurarine, changes in EPP amplitudes reflect changes in transmitter release, because active phorbol esters at the concentrations we used do not affect postjunctional sensitivity to ACh in the frog (Shapira et al., 1987; Redman et al., 1997). Solutions. All drugs used in this study were obtained from Research Biochemicals International (Natick, MA). Stock solutions of PDBu, calphostin C, sphingosine and GF 109 203X {structural formula: 3-[(1-dimethylamino)propyl]-[1H-indol-3-yl]-4-(1H-indo-3-yl)- 1H-pyrrol-e-2,5,-dione} were made up in dimethyl sulfoxide. Control EPPs were measured in Ringer s solution containing the same final concentration of dimethyl sulfoxide as in the drug solutions. Because the inhibitory effects of calphostin C are light-dependent (Bruns et al., 1991), solutions containing calphostin C were exposed to fluorescent light for at least 30 min before commencement of the experiment, and the experiments were carried out under fluorescent light. Results We began these experiments by determining a base-line concentration of PDBu to use in this study. Figure 1 shows rough concentration-response curves in which evoked ACh release (reflected as EPP amplitudes) is plotted as a function of PDBu concentration. These experiments were made both at normal levels of ACh release in d-tubocurarine-blocked preparations (fig. 1A) and at low levels of ACh release in Mg -blocked preparations (fig. 1B). In either condition, maximal increases in neurotransmitter release were evoked by PDBu concentrations of 50 nm and greater (n 8). We thus decided to use concentrations of 20 nm and 50 nm PDBu to determine the effectiveness of PKC inhibitors to attenuate the stimulatory action of PDBu on ACh release. We first investigated the ability of the nonselective protein kinase inhibitor, staurosporine, to inhibit the effects of PDBu (fig. 2, A and B). Staurosporine, which has been used in numerous studies as a PKC inhibitor, acts at the ATP binding site (C3 domain) in the catalytic region of this enzyme with an IC 50 of approximately 3 nm (Tamaoki et al., 1986). Because straurosporine produces complete inhibition of PKC when present at an extracellular concentration of 200 nm in isolated neuronal preparations (e.g., Considine et al., 1992), we chose to use 1 M staurosporine in all our experiments. Figure 2 shows an experiment in which the stimulatory effects of PDBu (20 nm) were compared in the absence (fig. 2A) and presence of staurosporine (fig. 2B). Note that the averaged control response (fig. 2A, lower trace, 5.5 0.1 mv, mean 1 S.E.M.) is increased reversibly to 8.5 0.1 mv during superfusion with PDBu in this experiment. After wash with drug-free solution, we then applied staurosporine to the same cell in the absence of PDBu and observed no Fig. 1. Concentration-response curves for phorbol dibutyrate-evoked increases in EPP amplitude. Semilog coordinates plot PDBu concentration against % change in EPP amplitudes. In (A), d-tubocurarine was used to paralyze the neuromuscular junction; in (B), elevated magnesium and reduced Ca was used to block neuromuscular transmission. EPP amplitudes recorded in the presence of PDBu were plotted as a percentage of the control EPP amplitude for each cell (n 4 cells in both A and B). Data show mean 1 S.E.M.

1998 Phorbol Esters and Neurosecretion 249 Fig. 2. Absence of effects of ATP binding site PKC antagonists on the increases in EPP amplitude produced by PDBu (20 nm). In (A), the effect of 20 nm PDBu on EPP amplitudes was recorded in the absence of PKC inhibitors. Traces in (B) show the effects of 20 nm PDBu on the same cell as shown in an ensuing 40-min wash and 20 min in the presence of 1 M staurosporine. The coefficients of variation for the averaged data traces (n 8) were control EPPs 5.1%; EPPs in PDBu 2.1%; EPPs in staurosporine 4.4%; EPPs in staurosporine PDBu 3.2%. Note the 55% increase in average EPP amplitude produced by PDBu in panel A and the 63% increase produced in panel B. Analysis of variance followed by multiple comparisons (see Materials and Methods ) revealed highly statistically significant differences between control EPPs and EPPs in PDBu (in the absence or presence of staurosporine, P.01). No statistically significant differences between control EPPs in the absence or in presence of staurosporine were observed, nor were differences observed between EPPs in PDBu in the presence or absence of staurosporine. When only four EPPs were averaged, similar levels of statistical significance were observed, although the coefficients of variation were larger than those observed when the responses to eight stimuli were averaged. Traces in (C) show the effect of 20 nm PDBu on EPPs in the presence of the selective PKC antagonist GF109203X (10 M). The differences between in EPPs in PDBu and those recorded in it absence were highly statistically significant (P.01). Similar results were found in three other experiments. statistically significant effect on ACh release as compared with control (mean EPP amplitude in staurosporine 5.4 0.1 mv, n 8 stimuli, fig. 2B). PDBu added in the presence of staurosporine increased the EPP amplitude to the level observed in the absence of staurosporine in the same cell (8.8 0.1 mv, fig. 2B). The failure of staurosporine to attenuate the stimulatory effects of PDBu was reproduced in all four experiments. It seemed important to examine the effects of a more selective staurosporine derivative that acts at the same C3 domain of PKC (Newton, 1995). One such agent, GF109203X, at concentrations up to 10 M, failed to block the potentiation of EPP amplitudes by PDBu (20 50 nm) (fig. 2C, n 4). GF109203X has a reported IC 50 of approximately 10 nm (Gordge and Ryves, 1994). Next, we looked at two inhibitors of PKC that act at the zinc-finger C1 binding site, the site also activated by the physiological modulator, diacylglycerol (see Gordge and Ryves, 1994). The effects of these agents, calphostin C and sphingosine, are depicted in figure 3. Figure 3A shows that 25 M calphostin C had no significant effect on the average EPP amplitude (see figure legend) and also did not inhibit the effect of PDBu (20 nm) on ACh release (n 4 experiments). The IC 50 for calphostin C as an inhibitor of PKC was approximately 50 nm (Kobayashi et al., 1989) and produced 80 to 90% inhibition of neuronal PKC when applied at an extracellular concentration of 2 M (Considine et al., 1992). Figure 3B depicts the effect of sphingosine (500 M) on EPP amplitudes. First, sphingosine produced a reduction in both EPP amplitudes and the amplitudes of spontaneous miniature EPPs (not shown) from control levels, presumably through actions on the postjunctional nicotinic receptors. Despite these postjunctional effects, 500 M sphingosine failed to block the stimulatory action of 20 nm PDBu on ACh release (n 4). This concentration of extracellular sphingosine is 500-fold higher than the concentration required to produce 80 to 90% inhibition of phorbol-stimulated PKC when added to the extracellular fluid of other neuronal cells (Considine et al., 1992; see also table 2 in Gordge and Ryves, 1994 for further details). Fig. 3. Absence of effects of phorbol ester/diacylglycerol PKC inhibitors on the increases in EPP produced by 20 nm PDBu. As shown in (A), application of calphostin C (25 M) had no statistically significant effect on EPP amplitudes and failed to attenuate the statistically significant increase in EPP amplitudes produced by PDBu (20 nm, n 4 experiments). As shown in (B), sphingosine (500 M) produced a highly significant reduction in EPP amplitude to approximately 25% of the control EPP amplitude through a postjunctional mechanism (see text for details). Application of PDBu (20 nm) resulted in an approximate doubling of the EPP amplitude compared with those recorded in sphingosine alone (n 4 experiments).

250 Searl and Silinsky Vol. 285 Figure 4 provides a graphical summary of the experimental results with the various PKC antagonists. The data in figure 4 show that no statistically significant effects of any of the PKC antagonists on PDBu-evoked increases in ACh release were observed. Discussion The seminal study describing the effects of phorbol esters on the electrophysiological correlates of neurotransmitter release was made at frog neuromuscular junctions (Shapira et al., 1987); in that study, an active phorbol ester increased ACh release but an inactive congener did not. Results such as these generally are interpreted as evidence for the involvement of the enzyme PKC in the process. However the results presented here, which demonstrate the absence of effects of a variety of PKC antagonists on the stimulatory action of phorbol esters, suggest that a presynaptic protein other than PKC is responsible for these effects of phorbol esters at the neuromuscular junction (see also Redman et al., 1997). The absence of effects of PKC antagonists is not likely to be caused by the lack of access of the inhibitor to the PKC enzyme at the concentration used. For example, considerably lower concentrations of the PKC inhibitors staurosporine, calphostin C and sphingosine than we used in this study, when applied in the extracellular bathing solution of isolated neuronal preparations, inhibit phorbol-stimulated PKC activity in these cells (see e.g., Considine et al., 1992). Indeed, the concentrations of sphingosine we used were so high as to produce nonspecific depressant effects on postjunctional sensitivity to ACh (e.g., fig. 3), yet sphingosine did not inhibit the stimulatory effects of PDBu. Finally, both published results and preliminary studies from this laboratory demonstrate that nonselective kinase inhibitors are capable of blocking Fig. 4. Graphical summary showing the absence effects of PKC antagonists on the stimulatory action of PDBu. The ordinate shows EPP amplitude presented as a percentage of the control EPP in the absence of PDBu. Data show mean 1 S.E.M. (n 11 experiments for PDBu, in which data from fig. 1 and from other PDBu experiments were pooled, and n 4 experiments for each of PKC antagonists in the presence of PDBu). One-way analysis of variance revealed no significant differences between the groups. the effects of cyclic AMP analogs on protein kinase A (Hirsh et al., 1990) yet are ineffective in attenuating the phorbol ester effects at neuromuscular junctions (Sebastiao and Ribeiro, 1990) at concentrations found in other preparations to inhibit PKC. These results thus raise the possibility that a presynaptic protein possessing the appropriate phorbol ester receptor, i.e., the C1 zinc-finger binding domain, could be the target site for the action of phorbol esters. One such protein could be an isoform of Munc-13, the mammalian homologue of unc-13. Although the function of Munc-13 is still unknown, this protein has been found to be concentrated in the plasma-membrane fraction of synaptosomes and also to be implicated in Ca -dependent docking of synaptic vesicles (Brose et al., 1995). Our earlier results suggest that phorbol esters increase the number of ACh release sites or the efficiency by which Ca is able to increase ACh release (Redman et al., 1997); indeed, active phorbols esters have also been shown to decrease the concentration of Ca required for binding to the C2 domain and subsequently activating PKC (Newton, 1995). As Munc-13 possesses both phorbol ester receptors and Ca -binding C2 domains (Brose et al, 1995), it would not appear too presumptuous to suggest that this protein or a related component of the secretory apparatus might be involved in mediating the effects of phorbol esters. Furthermore, mutation of the unc-13 gene in Caenorhabditis elegans is known to produce uncoordinated movements and the accumulation of ACh in the nervous system of this nematode; phorbol ester treatment of C. elegans likewise causes uncoordinated movements (Miwa et al., 1982). An alternative phorbol ester binding site is n-chimaeran (Ahmed et al, 1990, 1991), which also possesses C1 domains and may play a role in regulating GTPase activity apart from activating PKC (Ahmed at al, 1993). Unfortunately, we still do not have any tools other than the absence of effects of PKC antagonists to distinguish between PKC and non-pkc-mediated effects of active phorbol esters. For example, whereas high concentrations of calphostin C have been found to inhibit PDBu binding to the cysteine rich region of Unc-13 (EC 50 approximately 9 M, see Kazanietz et al., 1995), 25 M calphostin C failed to block the effects of PDBu in our experiments. This may indicate differences in binding characteristics of the cysteine-rich region of unc-13 to the full unc-13 protein or the possibility that a different form of unc-13 is present in the frog. The ability of calphostin C to reach its specific target site on the secretory apparatus may be impaired in this intact preparation. It should also be noted that several isoforms of PKC exist (Newton, 1995). Hence, it is remotely possible that a particular isoform of PKC, insensitive to normal PKC inhibitors, is involved in the action of phorbol esters. This seems unlikely, however, because both staurosporine and GF109203X act on the ATP binding site which is well conserved in all PKC isoforms (see Gordge and Ryves, 1994, for review). In conclusion, the inability of PKC inhibitors to diminish PDBu-induced increases in transmitter release at the frog neuromuscular junction suggests that a strategic component of the secretory apparatus other than PKC is the target site for the stimulatory action of phorbol esters at motor nerve endings.

1998 Phorbol Esters and Neurosecretion 251 Acknowledgment We wish to thank Dr. Jody Hirsh for discussion, for reading the manuscript, and for interpreting EMS s hideous handwriting. References Ahmed S, Kozma R, Lee J, Monfries C, Harden N and Lim L (1991) The cysteine-rich domain of human proteins, neuronal chimaerin, protein kinase C and diacylglycerol kinase binds zinc. Evidence for the involvement of a zinc-dependent structure in phorbol binding. Biochem J 280:233 241. Ahmed S, Kozma R, Monfries C, Hall C, Lim HH, Smith P and Lim L (1990) Human brain n-chimaerin cdna encodes a novel phorbol ester receptor. Biochem J 272: 767 773. Ahmed S, Lee J, Kozma R, Best A, Monfries C and Lim L (1993) A novel functional target for tumour-promoting phorbol esters and lysophosphatidic acid. The p21rac GTPase activating protein n-chimaerin. J Biol Chem 268:10709 10712. Ahmed S, Maruyama IN, Kozma R, Lee J, Brenner S and Lim L (1992) The Caenorhabditis elegans unc-13 gene product is a phospholipid-dependent high affinity phorbol ester receptor. Biochem J 287:995 999. Brose N, Hofmann K, Hata Y and Südhof TC (1995) Mammalian homologues of Caenorhabditis elegans unc-13 gene define a novel family of C 2 -domain proteins. J Biol Chem 270:25273 25280. Bruns RF, Miller FD, Merriman RL, Howbert JJ, Heath WF, Kobayashi E, Takahashi I, Tamaoki T and Hirofumi N (1991) Inhibition of protein kinase C by calphostin C is light dependent. Biochem Biophys Res Commun 176:288 293. Considine RV, Sherwin JS and Simpson LL (1992) Role of protein kinase C in short-term transmission at the mammalian neuromuscular junction. J Pharmacol Exp Ther 263:1269 1274. Del Castillo J and Katz B (1954) Quantal components of the end-plate potential. J Physiol (Lond) 124:560 573. Glantz SA (1992) Primer of Biostatistics, McGraw-Hill, Inc., New York. Gordge PC and Ryves WJ (1994) Inhibitors of protein kinase C. Cell Signal 6:871 882. Hirsh JK, Silinsky EM and Solsona CS (1990) The role of cyclic AMP and its protein kinase in mediating acetylcholine release and the action of adenosine at frog motor nerve endings. Br J Pharmacol 101:311 318. Jaken S (1996) Protein kinase C isozymes and substrates. Curr Opin Cell Biol 8:168 173. Kazanietz MG, Lewin NE, Bruns JD and Blumberg PM (1995) Characterization of the cysteine-rich region of the Caenorhabditis elegans protein Unc-13 as a high affinity phorbol ester receptor. Analysis of ligand binding interactions, lipid cofactor requirements and inhibitor sensitivity. J Biol Chem 270:10777 10783. Kobayashi E, Nakano H, Morimoto M and Tamaoki T (1989) Calphostin C (UCN- 1028C). a novel microbial compound, is a highly potent and specific inhibitor of protein kinase C. Biochem Biophys Res Commun 159:548 553. Maruyama IN and Brenner S (1991) A phorbol ester/diacylglycerol-binding protein encoded by the unc13 gene of Caenorhabditis elegans. Proc Natl Acad Sci USA 88:5729 5733. Miwa J, Tabuse Y, Furusawa M and Yamasaki H (1982) Tumour promoters specifically and reversibly disturb development and behaviour of Caenorhabditis elegans. J Cancer Res Clin Oncol 104:81 87. Newton AC (1995) Protein kinase C: Structure function and regulation. J Biol Chem 270:28495 28498. Nishizuka Y (1988) The molecular heterogeneity of protein kinase C and its implications for cellular regulation. Nature (Lond) 334:661 665. Rahimimoff R (1967) The use of the Biomac 500 computer for estimating facilitation at single end-plates. J Physiol (Lond) 191:12P 13P. Redman RS, Searl TJ, Hirsh JK and Silinsky EM (1997) Opposing effects of phorbol esters on transmitter release and calcium currents at frog motor nerve endings. J Physiol (Lond) 501.1:41 48. Sebastiao AM and Ribeiro JA (1990) Interactions between adenosine and phorbol esters or lithium at the frog neuromuscular junction. Br J Pharmacol 100:55 62. Shapira R, Silberberg SD, Ginsburg S and Rahamimoff R (1987) Activation of protein kinase C augments evoked transmitter release. Nature (Lond) 325:58 60. Silinsky EM (1981) On the calcium receptor that mediates depolarization-secretion coupling at cholinergic motor nerve terminals. Br J Pharmacol 73:413 429. Silinsky EM (1984) On the mechanism by which adenosine receptor activation inhibits the release of acetylcholine from motor nerve endings. J Physiol (Lond) 346:243 256. Tamaoki T, Nomoto H, Takahashi I, Kato Y, Morimoto M and Tomita F (1986) Staurosporine, a potent inhibitor of phospholipid/ca 2 dependent protein kinase. Biochem Biophys Res Commun 135:397 402. Wardell CF and Cunnane TC (1994) Biochemical machinery involved in the release of ATP from sympathetic nerve terminals. Br J Pharmacol 111:975 977. Send reprint requests to: Dr. Eugene M. Silinsky, Professor and Chairman, Department of Molecular Pharmacology, Northwestern University Medical School, 303 East Chicago Avenue, Chicago, IL 60611.