The Journal of Neuroscience, October 1995, 75(10): 6586-6591 Inhibitors of Cyclic-GMP Phosphodiesterase Alter Excitation of Limulus Ventral Photoreceptors in Ca2+-Dependent Fashion Edwin C. Johnsonifa and Peter M. O’Day* ‘Department of Physiology, Marshall University School of Medicine, Huntington, West Virginia 257559340 and *Institute of Neuroscience and Department of Biology, University of Oregon, Eugene, Oregon 97403-1254 We have examined the hypothesis that Ca*+-dependent cy- 1994) consistent with the idea that the following events occur clic-GMP metabolism may play a role in visual transduction in phototransduction. Activation of visual pigment, rhodopsin, in Limulus photoreceptors. Although phosphoinositide hy- by light activates a G-protein, stimulating hydrolysis of phos- drolysis is central to phototransduction and phosphoinos- phatidylinositol 4,5-bisphosphate(PIP,) by phospholipaseC; itide-dependent Ca2+-mobilization seems to be required for this createscytosolic inositol 1,4,5trisphosphate(IP,), the trig- transduction, the subsequent steps leading to ion channel ger for mobilization of intracellular Ca2+ and elevation of gating (the immediate cause of excitation) are not under- [CaZ+,].In Limulus photoreceptors,Ca2+-mobilization causes ex- stood. Channels normally opened in response to light can citation by opening plasma membrane channels; however, the be opened in excised membrane patches by cGMP but not link betweenCa*+ elevation and channel gating remainsobscure. by Ca*+, suggesting that cGMP acts as a channel ligand in Electrophysiological evidence implicates cGMP involvement excitation. in Limulus transduction (Johnsonet al., 1986; Bacigalupo et al., Using phosphodiesterase inhibitors, we investigated 1991; Feng et al., 1991). The observation that light-activated whether changes in cGMP metabolism could affect exci- channels are gated selectively by cGMP and not by Ca2+(Ba- tation. We report that zaprinast and IBMX increased the am- cigalupo et al., 1991) points to cGMP as a channel ligand in plitudes and retarded the kinetics of physiological light re- transduction, supporting the notion that excitation is mediated sponses. These effects were maximal for brightest stimuli. by elevation of the cGMP concentration ([cGMP],) near the The effects were markedly enhanced in low Ca2+ condi- channels. Light-induced elevation of [cGMP& could arise from tions. In contrast, excitation induced by direct IP,-injection light-induced changes in the synthesis and/or degradation of and by direct Ca2+-injection were inhibited. These obser- cGMP; however, to date, characterization of light-dependent vations suggest that PI-induced excitation is dependent on changesin [cGMP], in any invertebrate photoreceptorshas been cGMP metabolism in a Ca2+-dependent manner, and they elusive (Brown et al., 1984a; Saibil, 1984; Johnsonet al., 1986; support the possibility that transduction involves modifi- Brown et al., 1992). Nonetheless,there is evidence for the pres- cation of cGMP metabolism by Ca*+-release resulting from ence of guanylate cyclase and of phosphodiesterasein inverte- phosphoinositide hydrolysis. brate photoreceptors (Robinson and Cote, 1989; Inoue et al., [Key words: phototransduction, zaprinast, M&B22948, 1992; Schraermeyer et al., 1993; Yoshikawa et al., 1993). Fur- cGMP phosphodiesterase, photoreceptor, Limulus, Ca2+, thermore, there is recent evidence for cGMP induced excitation phosphoinositide, IP, cyclic-GMP] (Bacigalupo and O’Day, unpublished observations) and a cGMP-gated channel in Drosophila retina (Baumann et al., Two intracellular secondmessenger systems, the phosphoinosi- 1994). tide (PI) pathway and the cyclic-GMP (cGMP) pathway, appear In the presentwork, we have usedan electrophysiologicaland to be important in light-induced excitation of Limulus photore- pharmacological approach to investigate the hypothesis that ceptors(Brown et al., 1984b, 1987; Fein et al., 1984; Bacigalupo cGMP metabolismis involved in light- and PI-induced excita- et al., 1990; Feng et al., 1991; Faddis and Brown, 1992, 1993; tion. Our results indicate that excitation can be modified by in- Inoue et al., 1992). The involvement of the PI pathway in in- hibitors of cGMP degradation in a Ca2+-dependentfashion. vertebrate excitation is supportedby several lines of evidence These results support the idea that excitation is dependenton (Payne et al., 1988; Ranganathanet al., 1991; Hardie and Minke, cGMP metabolism,modulated by CaZ+ releasedduring excita- tion. Some of theseresults have appearedin abstract form (O’Day Received Apr. 14, 1995; revised May 30, 1995; accepted June 1, 1995. et al., 1991; Johnsonand O’Day, 1992). We gratefully acknowledge Toufik Sadat, Jamie Crona, and Linda Lloyd for experimental assistance. We also thank Drs. John Lisman, Juan Bacigalupo, Judith Eisen, Phyllis Robinson, Mark Gray-Keller, and Mark Simmons for help- Materials and Methods ful discussions and readings of the manuscriot. Funding came from National Preparation. Limulus polyphemus were obtainedfrom Marine Biolog- Science Foundation BNS&12455 (to E.C.J.\, Nationa? Institutes of Health EY09388 (to P.M.O.), and an American Heart Association, Oregon Affiliate ical Labs,Woods Hole, MA. Ventralnerves were removed, desheathed, grant (to P.M.0.). and treatedwith pronase.The dissectionand preparationof the ventral Correspondence should be addressed to Peter M. O’Day, Institute of Neu- photoreceptorswere similarto thosedescribed previously (Lisman et roscience, University of Oregon, Eugene, OR 97403.1254. al., 1982)as were the electricalrecording, extracellular superfusion, ion “Present address: Salk Institute Biotechnology/Industrial Associates, La Jolla, substitution,current-clamp, and voltage-clampprocedures. CA. Solutions. Duringdissection, desheathing, and impalement, the tissue Copyright 0 1995 Society for Neuroscience 0270-6474/95/156586-06$05.00/O wasbathed in artificial seawater(ASW), in mu: 425NaCl, 10 KCl, 22 The Journal of Neuroscience, October 1995, 15(10) 6567 MgCl,, 26 MgSO,, 10 CaCl,, 15 TrisCl, pH 7.8. “1 mM Ca2+ saline” indicates that only 1 mM CaCl, was added and MgCl, was substituted B mole-for-mole for CaCl,. The Ca2+ injection microelectrodes contained Normal saline Low-Co saline 10 mM Ca-aspartate, 100 mM K-aspartate, 10 mu HEPES, pH 7.0. The IP,-injection _ electrodes contained 100 p,M IP, [D-myo-inositol- (1.4.5)trisohosohate (IP,) (Calbiochem)l. 100 mM K-asoartate, 10 mM Control Control Zoprinast HEPES, pH 7:O. Microelectrodes from’which CaZ+-BAPTA mixtures Zaprinast were injected contained 100 mu K-aspartate, 10 mM HEPES, 75 mu A A BAPTA, pH 7.0; Ca-aspartate was added to attain desired levels of free [Caz+J. A Ca-aspartate/BAPTA ratio, R,,,,,, of 0.085 yields lo-* M JL Ca*,,; RcdBA, = 0.48 yields lo-’ M CaZ,; R,,,, = 0.625 yields LM 2.3*10-’ M Ca*,,; R,,,,, = 0.90 yields 10m6 M Ca*,,; R,,,,, = Ii I5mV I OmV 0.99 yields 10m5 M Ca*,,,). Ca*+/BAPTA mixtures were injected until I- the cell no longer adapted to light at its new level of sensitivity. At 0.1s 0.2s 10-8 M Ca*,,,, response kinetics were very slow, amplitudes were great- ly decreased, and the effects of IBMX were inconsistent. IBMX (Sigma) was solubilized in ethanol before addition to experimental salines; the final ethanol concentration in IBMX and control salines was 0.1% in C D these experiments. Zaprinast (M&B22948) was dissolved in triethano- lamine before addition to experimental salines; the final triethanolamine concentration in zaprinast test saline and control salines was 0.1% in these exoeriments. Concentrations of zanrinast (60 LM) and IBMX (l- 2.5 mMj were chosen to strike a balance between maximal effective PDE-inhibition and minimal effects of solvents. The selectivity and po- tency (K, - 150 nM) of zaprinast is substantially greater than other PDE inhibitors (IBMX K, - 100 JLM) against vertebrate photoreceptor PDE and other PDEs (Gillespie and Beavo, 1989). In all experiments re- ported, the concentration of PDE inhibitor in the vicinity of the cells’ transduction apparatus is unknown. Under the conditions of our exper- LM 1 OOpA iments, physiological responses recorded in ethanol and triethanolamine IOmV control solutions were identical to those in normal bathing saline. Optics. Illumination intensities shown in this paper are given in neu- 50ms 50ms tral density units (ND); each ND corresponds to a tenfold change in light intensity with 0 ND corresponding to the maximum available in- F&m 1. Effects of cGMP-PDE inhibitor, zaprinast, on electrical re- tensity, 1 .O mW/cm2; white light was used (Osram Xenophot hlx64625 sponses to brief test light flashes. A, Cells current clamped to dark V, filament bulb). In dark-adapted cells, approximately one quantum bump of -70 mV were illuminated with a dim test flash (-4.3 ND, 25 msec) per second was evoked with dim illumination of intensity -6.9 ND. every 10 sec. Responses to two such flashes are shown superimposed: one before and one 15 min after introduction of 60 PM zaprinast to Results normal CaZ+ (10 mM) saline. Zaprinast had only a small apparent effect on response waveforms and amplitudes. The bars (LM, light monitor) PDE inhibitors induced small depolarization of dark membrane adjacent to the traces indicate the period of illumination. Similar results potential and induced small dark currents were observed in 6 of 6 cells. B, Zaprinast had a similar effect on dim PDE inhibitors have small effects on resting potential and/or test flash responses in 1 mM Caz+ saline in 6 of 6 cells. C, Responses dark currents