CAMP-Independent Effects of 8-(4 ... - Semantic Scholar

JOURNALOF NEUROPHYSIOLOGY Vol. 72. No. 3. September 1994.

CAMP-Independent Effects of 8-(4=parachlorophenylthio)-cyclic AMP on Spike Duration and Membrane Currents in Pleural Sensory Neurons of Ap/“sia SHUZO SUGITA, DOUGLAS A. BAXTER, AND JOHN H. BYRNE Department ofNeurobiology and Anatomy, University of Texas Medical School at Houston, Houston, SUMMARY

AND

CONCLUSIONS

1. The serotonergic modulation of pleural sensory neurons in Aplysia is mediated via two second messenger systems: the adenosine cyclic monophosphate/protein kinase A (cAMP/PKA) and diacylglycerol/protein kinase C systems. Often membrane permeable derivatives of CAMP, such as 8-(4-parachlorophenylthio)CAMP (pcpt-CAMP), have been used to investigate the role of cAMP/PKA in modulating sensory neurons. In light of recent findings that pcpt-CAMP may have CAMP-independent actions, we have reexamined the effects of pcpt-CAMP on the action potential and membrane currents of the sensory neurons. 2. Although pcpt-CAMP (500 PM to 1 mM) and serotonin (5 HT; 10 PM) induced comparable measures of spike broadening (an average increase above baseline of 29 and 40%, respectively), the broadening produced by the two was qualitatively different. Serotonin-induced broadening developed slowly over 9- 12 min, was most prominent during later phases of the spike repolarization, and reduced the spike afterhyperpolarization. In contrast, pcpt-CAMP-induced broadening developed rapidly, was rather uniform throughout the repolarization phase of the spike, delayed the peak of the action potential, and increased the afterhyperpolarization. 3. Preexposure of sensory neurons to 5-HT did not occlude further spike broaden by subsequent application of pcpt-CAMP. Indeed the effects of the two were additive. In addition, the effects of pcpt-CAMP were not mimicked by another analogue of CAMP, 8-bromo-CAMP. Interestingly, most of the effects of pcpt-CAMP on the action potential were mimicked by 8-(4-parachlorophenylthio)-guanosine cyclic monophosphate (pcpt-cGMP), but not by 8-bromo-cGMP. 4. During voltage-clamp pulses to 20 mV, pcpt-CAMP reduced the membrane current throughout the voltage-clamp pulse, which was qualitatively different from the modulation of the membrane current by 5-HT. In addition, the pcpt-CAMP-induced reduction in the membrane current at the beginning of the pulse was much greater than that induced by 5-HT. Moreover, preexposure of sensory neurons to 5-HT did not occlude further reduction in the membrane current by subsequent application of pcpt-CAMP. 5. These results suggest that pcpt-CAMP has some mechanisms of action that are not shared by 5-HT or CAMP but are shared by pcpt-cGMP. In addition, these findings provide further evidence that results obtained with this compound should be interpreted with caution.

INTRODUCTION

Serotonin (5HT)-induced modulation of membrane currents in sensory neurons is believed to be a key mechanism underlying 5-HT-induced facilitation of undepressed synapses via changing the waveform of action potentials 1250

Texas 77225

and excitability of the sensory neurons (Gingrich et al. 1988; Hochner et al. 1986; Kandel and Schwartz 1982; see also Klein 1993). Recently it has become clear that the actions of 5-HT are mediated via two second messenger systems: the adenosine cyclic monophosphate/protein kinase A (cAMP/PKA) and diacylglycerol/protein kinase C (DAG/PKC) systems (Bacskai et al. 1993; Baxter and Byrne 1990a; Bernier et al. 1982; Braha et al. 1990, 1993; Ghirardi et al. 1992; Goldsmith and Abrams 1992; Hochner and Kandel 1992; Jarrard et al. 1993; Mercer et al. 199 1; Ocorr and Byrne 1985; Pollock et al. 1985; Sacktor and Schwartz 1990; Sossin and Schwartz 1992; Sugita et al. 1992, 1994; Walsh and Byrne 1989). The cAMP/PKA system is believed to modulate a number of membrane currents, including the S-K+ current (1&, a slow component of the Ca2+-activated K+ current (1k,&, and a voltagedependent K+ current (1kv). The DAG/PKC system is believed to modulate I, v and the nifedipine-sensitive component of the Ca2+ current (L-type &a>. The cAMP/PKAmediated effects are believed to contribute to the 5-HT-induced increases in excitability and a component of spike broadening that develops rapidly (within 3 min of exposing the sensory neurons to 5-HT), whereas the DAG/PKC mediated effects are believed to contribute to a component of spike broadening that develops more slowly (after about 3 min) (see Sugita et al. 1994 for discussion). Previous work often has relied on the use of derivatives of CAMP, such as the commonly used analogue 8-(4-parachlorophenylthio)-cyclic AMP (pcpt-CAMP), to investigate the role of the cAMP/PKA system in mediating the actions of 5-HT (e.g., Baxter and Byrne 1990a; Goldsmith and Abrams 1992; Noel et al. 199 1; Sweatt and Kandel 1989; see also Kramer et al. 1988a,b). This compound offers several experimental advantages. It is membrane permeable, relatively resistant to cyclic nucleotide phosphodiesterases, and it more readily actives PKA as compared with CAMP (Braumann et al. 1986; Meyers and Miller 1974; Miller et al. 1980; Simon et al. 1973). Thus in studies relating CAMP to the modulation of neuronal function, the effects of this analogue of CAMP have been assumed to result solely form activation of PKA. Several recent studies, however, indicate that analogues of CAMP, including pcpt-CAMP, can have effects that are not related to their ability to mimic CAMP and stimulate PKA activity. For example, several analogues of CAMP have been shown to decrease y-aminobutyric acid A (GABA,) receptor-mediated influx of Cl- (Lambert and Harri-

0022-3077/94 $3.00 Copyright 0 1994 The American Physiological Society

CAMP-INDEPENDENT

EFFECTS

son 1990; Leidenheimer et al. 1990). This effect was not blocked by kinase inhibitors, was not mimicked by activators of adenylyl cyclase and was independent of protein phosphorylation. Similarly, pcpt-CAMP was found to inhibit a cyclic nucleotide phosphodiesterase, an action not mimicked by other analogues of CAMP (Connolly et al. 1992). In addition, many analogues of CAMP are subject to enzymatic degradation (e.g., via cyclic nucleotide phosphodiesterases) (Braumann et al. 1986; Connolly et al. 1992; Sandberg et al. 199 1) and the resulting metabolites have been found to have actions such as modulating gene expression (e.g., Hall 1992; Zhu and Chuang 1988) and inhibiting growth of several cell types (e.g., Lange-Carter et al. 1993; Zorn et al. 1993). Finally, pcpt-CAMP has been shown to be a potent activator of cGMP-dependent kinase (Francis et al. 1988; Sandberg et al. 199 1). Thus it would appear that the mechanisms of the physiological effects of analogues of CAMP should be interpreted with some caution. In light of these recent results, we have reexamined the effects of high concentrations of CAMP analogues (pcptCAMP and 8-bromo-CAMP) on the waveform of action potentials and membrane currents and compared the actions of the CAMP analogues with those of 5-HT. The results indicate that although high concentrations of pcpt-CAMP mimicked and occluded some effects of 5-HT, the overall effects of pcpt-CAMP were qualitatively different from those of 5-HT. In addition, the effects of high concentrations of pcpt-CAMP were not mimicked by another CAMP analogue (8-bromo-CAMP), even at the concentrations of 50 mM. Furthermore, most of the actions of pcpt-CAMP were mimicked by 8-(4-parachlorophenylthio)-cyclic GMP (pcpt-cGMP) but not 8-bromo-cGMP. Finally pcpt-CAMPinduced modulation of membrane currents during voltageclamp pulses to 20 mV is qualitatively and quantitatively different from those by 5-HT. These results suggest that high concentrations of pcpt-CAMP may have many actions that are not CAMP dependent. Our results also suggest that CAMP analogues are not sufficient to mimic fully the effects of 5-HT on spike duration and in particular the modulation of Zk,v by 5-HT. METHODS

Measurements

of spike duration

and membrane

current

The methods for preparing sensory neurons and for measuring spike durations and membrane currents were described in the preceding paper (Sugita et al. 1994).

Chemicals Stock solutions of serotonin creatine sulfate (5HT, 1 mM, Sigma) and 8-bromo-CAMP (1 M; Sigma) were dissolved in artificial seawater (ASW) and prepared daily. As described in the preceding paper, 4P-phorbol 12,13-diacetate (PDAc; Sigma) were prepared. Stock solutions of pcpt-CAMP (Sigma) were prepared in two different ways. In one method, pcpt-CAMP (50 or 100 mM) was dissolved in dimethyl sulfoxide (DMSO) and in the other method, pcpt-CAMP (20 mM) was dissolved in H,O. When DMSO was used, a final concentration of DMSO in the bath was 1% (v/v). This DMSO concentration was shown previously to have little effect on membrane currents (Baxter and Byrne 1990b). In addition, in a separate series of control experiments, 1% DMSO had little effect on the waveform of action potentials or resting

OF

A CAMP

ANALOGUE

1251

membrane potential. pcpt-cGMP (100 mM, Biolog Life Science Institute) was dissolved in DMSO, and 8-bromo-cGMP (50 mM, Sigma) was dissolved in 50% (v/v) of DMSO and H20. Final DMSO concentrations in the bath for these compounds were 0.51%. Stock solutions of pcpt-CAMP, pcpt-cGMP, and &bromocGMP were stored at -4OC. Small aliquots of stock solution were added directly into the recording chamber. RESULTS

pcpt-CAMP-induced spike broadening d@erent from that of 5-HT

is qualitatively

Recently, Goldsmith and Abrams ( 1992) found that high concentrations (500 PM to 1 mM) of pcpt-CAMP produced more spike broadening than originally reported by Baxter and Byrne ( 1990a). Using two-electrode current-clamp techniques, we compared the changes in the waveforms of action potentials produced by pcpt-CAMP with those produced by 5-HT and PDAc, an activator of PKC (Fig. 1). We prepared pcpt-CAMP at concentrations comparable with those used by Goldsmith and Abrams (1992) in two different ways (see METHODS). pcpt-CAMP prepared with DMSO (DMSO-pcpt-CAMP; 500 PM to 1 mM) produced more rapid effects than that prepared with Hz0 (H,O-pcptCAMP; 1 mM). DMSO-pcpt-CAMP (n = 9) produced rapid (within 3 min) depolarization (--3 mV) of resting membrane potentials, presumably because of the closure of S-K+ channels (Baxter and Byrne 1989; Pollock et al. 1985; Walsh and Byrne 1989; see also Siegelbaum et al. 1982). In addition, it produced rapid spike broadening (within 3 min, Fig. 1, A and E). In contrast to DMSO-pcpt-CAMP, H,Opcpt-CAMP (n = 6) slowly depolarized the resting potentials and broadened the spike (Fig. 1, B and E). Both depolarization and spike broadening induced by H,O-pcpt-CAMP began to be observed 9 min after the application and reached the maximum 24 min after application. The difference in the time course of the effects between DMSO- and H20pcpt-CAMP may be because of the difference in the vehicle: DMSO appeared to facilitate the membrane permeability of pcpt-CAMP. Regardless of the vehicle (DMSO or H,O), however, pcpt-CAMP produced very similar and consistent changes in the waveform of action potentials (compare Fig. 1, A and B). Thus changes in the waveforms of action potentials produced by DMSO- or H,O-pcpt-CAMP were not because of high concentrations of DMSO (I%, v/v), but because of the effects of pcpt-CAMP. In addition, in a separate series of experiments, 1% DMSO did not produce detectable changes in the waveform of action potentials (data not shown). The effects of DMSO-pcpt-CAMP were examined only until 9 min after application (Fig. 19, because 5-HT was added at this time in most experiments (n = 5, see below and Fig. 2). Although pcpt-CAMP depolarized the membrane potential like 5-HT, there were several differences between the effects of 5-HT and pcpt-CAMP on the action potential. First, the peak of action potential tended to be delayed in pcpt-CAMP (unlike 5-HT) and a larger stimulus current (lo-30% increase) was necessary to elicit an action potential whose peak occurred at the same point in time as the baseline spikes. Such increases in stimulus current were never necessary in 5-HT. Second, DMSOpcpt-CAMP produced its maximum broadening at the same

1252

S. SUGITA,

D. A. BAXTER,

AND J. H. BYRNE

b. +l-40-pcpt-CAMP

b. +DMSO-pcpt-CAMP /

/

a. ASW

a. ASW

30mV a. ASW

a. ASW

. l+O-pcpt-CAMP n

5-HT

A PDAc x DMSO-pcpt-CAMP I I

1

I

I

I

-6

-3

0

3

6

I

9

I

I

1

I

1

12

15

18

21

24

+ 5-HT

Time (min) FIG. 1. High concentrations of 8-(4-parachlorophenylthio)-cyclic AMP (pcpt-CAMP) induced spike broadening, but this broadening was qualitatively different from that induced by serotonin (5-HT) or 4p-phorbol 12,13-diacetate (PDAc). Traces represent averaged data from all the sensory neurons examined. A: dimethyl sulfoxide (DMSO)-pcpt-CAMP (500 PM to 1 mM; n = 9) increased spike duration by 28% of baseline as well as induced a deeper afterhyperpolarization after 6 min exposure. No apparent difference was observed between spike broadening induced by either 500 PM or 1 mM of DMSOpcpt-CAMP. B: H,O-pcpt-CAMP (1 mM; n = 6) induced similar changes in the waveform of action potentials as DMSOpcpt-CAMP after 24 min of exposure. Because the peaks of the action potentials tended to be delayed after application of DMSO-pcpt-CAMP or H,O-pcpt-CAMP, a stronger current stimulus was necessary to elicit action potentials whose peaks occurred at the same point in time as the baseline spikes. C 5-HT (n = 7) increased spike duration by 40% of baseline after 12 min exposure. D: PDAc (3 PM; n = 5) increased spike duration by 25% of baseline after 2 1 min exposure. This datum was taken from Sugita et al. (1992). E: time course of changes in spike duration by DMSO-pcpt-CAMP, H,O-pcpt-CAMP, 5-HT, or PDAc. Time at which these agents were applied (+). Times at which traces in A-D are shown are the times at which maximal broadening was induced by each agent. Time at which 5-HT was added to the bath (j), which already contained DMSO-pcpt-CAMP. Error bars in this and subsequent illustrations represent mean k SE.

CAMP-INDEPENDENT

EFFECTS

time as the depolarization of the resting potential. As mentioned in a previous paper (Sugita et al. 1992), the maximum 5-HT-induced broadening developed slowly and followed, by - 10 min, the 3-6 mV depolarization of the resting potential (observed within l-2 min after 5-HT application). Third, pcpt-CAMP increased the spike afterhyperpolarization. This increase in the afterhyperpolarization was observed in all (n = 15) of the neurons treated with pcpt-CAMP but was never observed in the cells treated with 5-HT (n = 7) or PDAc (n = 5; Fig. 1, C and D). Fourth, pcpt-CAMP-induced broadening was observed rather uniformly throughout the repolarizing phase of the spikes. In contrast, 5-HT- and PDAc-induced broadening became more evident during the repolarizing phase at membrane potentials -0 mV (Fig. 1, C and D). In addition, the action of 5-HT and PDAc were not additive (Sugita et al. 1992). These results suggest that 5-HT- and PDAc-induced spike broadening share some common mechanisms, whereas the qualitative differences between pcpt-CAMP- and 5-HT (or PDAc)-induced spike broadening led us to suspect that these two types of broadening might be because of the different mechanisms. To compare these two types of broadening further, we applied 5-HT and pcpt-CAMP to the same cell (Fig. 2). That is, 5-HT was applied to a cell that was pretreated with pcptCAMP, or pcpt-CAMP was applied to a neuron that was pretreated with 5-HT. Addition of DMSO-pcpt-CAMP produced maximum spike broadening within -3 min (Figs. 1E and 2A). As described above, this broadening was uniform throughout the repolarizing phase of the spike and was followed by an enhanced afterhyperpolarization. Application of 5-HT to the bath 9 min after application of DMSO-pcpt-CAMP (n = 5) further depolarized the resting membrane potential by l-3 mV and produced additional broadening (Fig. 2, A and D). This additional broadening was similar to the broadening produced by 5-HT alone in that the broadening developed more slowly and was limited to later phases of the repolarization and that the afterhyperpolarization was not increased. Similar results were observed when 5-HT was added to the bath 24 min after the application of H20- pcpt-CAMP (n = 5; Fig. 2B). The longer preexposure time was necessary because of the slower time course of the H,O-pcpt-CAMP-induced broadening (Fig. lg. As before, the effects of H20-pcpt-CAMP and 5-HT were qualitatively different. Application of DMSO-pcpt-CAMP (n = 5) to the bath, which already contained 5-HT, did not produce further depolarization of the resting potential. Instead, it produced hyperpolarization of l-3 mV in 4 of 5 cells. Despite the hyperpolarization, addition of DMSO-pcpt-CAMP produced rapid spike broadening (Fig. 2, C and D). This effect was similar to the broadening produced by pcpt-CAMP alone, including a deeper afterhyperpolarization, and thus appears different from 5-HT-induced broadening. The additivity of pcpt-CAMP- and 5-HT-induced broadening suggests that some of the mechanisms underlying these two types of broadening may differ. An alternative interpretation is that the doses of 5-HT and pcpt-CAMP used in this study were not sufficient to saturate the CAMP cascade. Because the concentrations of pcpt- CAMP used in this study was the highest we were able

OF

A CAMP

ANALOGUE

1253

to obtain, we could not fully test this possibility. In five additional experiments, however, we increased the concentration of 5-HT to 50 PM (data not shown). Previous work (Jarrard et al. 1993; Ocorr and Byrne 1985) indicated that 50 PM is a saturating concentration for 5-HT-induced synthesis of CAMP. This higher concentration of 5-HT produced greater spike broadening than 10 PM of 5-HT [89.8 -+ 4.8% in 50 PM; 40.5 t 5.9% in 10 PM (means t SE)]. The characteristics of the broadening by the two concentrations of 5-HT were similar, however. Subsequent addition of DMSO-pcpt-CAMP to the bath containing 50 PM 5-HT produced additional broadening (from 89.8 t 4.8% to 13 1.3 t 14.1%), and once again this broadening was qualitatively different from that induced by 5-HT. The qualitative differences in the changes in the waveform of the spike produced by pcpt-CAMP and 5-HT suggest that some of the mechanisms underlying pcpt-CAMP-induced broadening differ from those of 5-HT. The effects of pcpt-CAMP and 5-HT were not additive at the all time points, however. Preincubation with DMSOpcpt-CAMP seemed to occlude partially the rapid component of 5-HT-induced spike broadening (3 min after application of 5-HT to the bath, which contained DMSO-pcptCAMP). Serotonin increased the duration of the spike by 16% of baseline 3 min after application in ASW, whereas 5-HT increased the duration of the spike by only 4% (from 123 to 127%) of baseline 3 min after application in the presence of DMSO-pcpt-CAMP (Fig. I,??). Previous work (Goldsmith and Abrams 1992; Hochner and Kandel 1992) has indicated that this rapid component of 5-HT-induced spike broadening is primarily CAMP dependent. Thus these data suggest that some of the effects of pcpt-CAMP are CAMP dependent and are nonadditive with the rapid component of 5-HT-induced spike broadening. High concentrations of 8-bromo-cAMI changes induced by pcpt-CAMP

did not mimic

the

Although pcpt-CAMP-induced changes of action potentials appear qualitatively different from those of 5-HT, the differences may simply be because of the fact that pcptCAMP-induced effects are purely CAMP dependent whereas the effects of 5-HT represent the combined actions of CAMP/PI&A and DAG/PKC. If pcpt-CAMP-induced changes are solely CAMP dependent, these changes should be seen in the cells treated with other analogues of CAMP. To test this hypothesis, we used high concentrations of 8bromo-CAMP (20-50 mM) and examined their effects on the waveform of action potentials (Fig. 3A). Membrane potentials were depolarized rapidly (within 1 min) by 5-9 mV by 8-bromo-CAMP (n = 6); this action was even more pronounced than the depolarization produced by 5-HT. In addition, 8-bromo-CAMP produced some broadening (7% increase after 9-min exposure) (Fig. 3A), which is similar to 5-HT-induced spike broadening at 3 min. Our results confirmed and extended the previous observation by Baxter and Byrne ( 1990a) that presumably saturated concentrations of this CAMP analogue produced modest broadening. The broadening induced by 8-bromo-CAMP, however, is qualitatively different from that induced by pcpt-CAMP. Unlike pcpt-CAMP, 8-bromo-CAMP did not increase the

1254

S. SUGITA,

D. A. BAXTER,

AND J. H. BYRNE

b. +DMSO-pcpt-CAMP

b. +HzO-pcpt-CAMP c. HzO-pcpt-CAMP + 5-HT

c. DMSO-pcpt-CAMP + 5-HT

a. ASW

a. ASW

C

D

-pcpt-CAMP

30 mV 3 ms

FIG. 2. pcpt-CAMP- and 5-HT-induced spike broadening were additive. Traces represent averaged data from all the sensory neurons examined. A: action potentials are shown before and 9 min after DMSO-pcpt-CAMP (1 mM; yt = 5) application and 12 min after addition of 5-HT ( 10 PM) to the bath, which still contained DMSO-pcpt-CAMP. Serotonin produced further broadening after preexposure to DMSO-pcpt-CAMP. B: action potentials before and 24 min after H,Opcpt-CAMP (1 PM; n = 5) application and 12 min after addition of 5-HT (10 PM) to the bath, which contained H,O-pcptCAMP. C action potentials before and 15 min after 5-HT (n = 5) application and 6 min after the addition of DMSO-pcptCAMP to the bath, which still contained 5-HT. DMSO-pcpt-CAMP produced additional spike broadening after preexposure to 5-HT, as well as increased the afterhyperpolarization. D: summary data of the experiments in A-C.

afterhyperpolarization tential.

or delay the peak of the action po-

Many actions of pcpt-CAMP were mimicked cGMP, but not Sbromo-cGMP

by pcpt-

To gain some insight into the mechanisms of the action of pcpt-CAMP, we examined the action of pcpt-cGMP on the waveform of action potentials (Fig. 3B). If the CAMPindependent actions of pcpt-CAMP are mediated by parachlorophenylthio moiety of this compound, we speculated that pcpt-cGMP (500 PM to 1 mM) may induce similar actions to those of pcpt-CAMP. As controls, the effects of 8-bromo-cGMP (500 PM to 1 mM) also were examined. The final concentrations of DMSO for pcpt-cGMP and 8-

bromo-cGMP were identical (0.5% v/v for 500 PM and 1% v/v for 1 mM). Like pcpt-CAMP, pcpt-cGMP (n = 8) delayed the peak of an action potential and a lo-30% larger stimulus current was necessary to elicit an action potential whose peak occurred at the same time point as the baseline spike (Fig. 3B). In addition, pcpt-cGMP-induced spike broadening was qualitatively similar to that induced by pcpt-CAMP, and the broadening was observed rather uniformly throughout the repolarizing phase of the spikes. One difference between pcpt-CAMP and pcpt-cGMP was that an increase in afterhyperpolarization was not consistently induced by pcptcGMP. The effects of pcpt-cGMP on the waveform of action potentials were not mimicked by 8-bromo-cGMP (n = 5, data not shown). Because most of the actions of pcpt-

CAMP-INDEPENDENT

1255

EFFECTS OF A CAMP ANALOGUE

30mV 3ms FIG. 3. Many actions of pcpt-CAMP were not mimicked by &bromo-CAMP, but were mimicked by 8-(4-parachlorophenylthio)-cyclic GMP (pcpt-cGMP). A: action potentials are shown before and 9 min after application of 8-bromo-CAMP (20-50 mM). Traces are averaged data from 6 sensory neurons. At 9 min 8-bromo-CAMP increased spike duration maximally by 7% of baseline, but did not increase the afterhyperpolarization or delay the peak of action potential. B: action potentials shown are before and 6 min after application of pcpt-cGMP (500 PM to 1 mM). Traces represent averaged data from 8 sensory neurons. Like pcpt-CAMP, pcpt-cGMP delayed the peak of the action potential and induced spike broadening of 30% above baseline.

CAMP including spike broadening are mimicked by pcptcGMP, but not &bromo-cGMP, these actions appear to be mediated by parachlorophenylthio moiety of pcpt-CAMP and pcpt-cGMP. Unexpectedly, we found that pcpt-cGMP and 8-bromocGMP had opposite effects on the resting membrane potential of the sensory neurons. pcpt-cGMP hyperpolarized the cell by -3 mV, whereas 8-bromo-cGMP depolarized the cell by -3 mV. Membrane

currents modulated

by pcpt-CAMP

Baxter and Byrne (1989, 1990b) examined the effects on the membrane currents by various agents that are known to elevate CAMP levels in the sensory neurons, including application of CAMP analogues (8.bromo-CAMP and pcptCAMP), a derivative of the adenylyl cyclase activator forskolin, and a neuromodulatory peptide transmitter (SCPB). Although all these agents mimicked and occluded the 5HT-induced reduction of IK,s, at the concentrations used, none of these mimicked or occluded 5-HT-induced modulation of 1k v. In contrast, Goldsmith and Abrams (1992) reported that higher concentrations of pcpt-CAMP modulated lkv in a manner similar to 5-HT. Our observation that pcpt-CAMP- and 5-HT-induced spike broadening is qualitatively different led us to reexamine the membrane currents modulated by pcpt-CAMP (Fig. 4). Like 5-HT, DMSO-pcpt-CAMP (1 mM; n = 6) reduced an outward current during voltage-clamp pulses to -20 mV (Fig. 4C). In addition, pcpt-CAMP reduced the outward current elicited at 20 mV (Fig. 4A). One critical difference between pcpt-CAMP-and 5-HT-induced modulation of the membrane currents at 20 mV was that pcpt-CAMP never increased the outward current at the end of the pulse. In contrast such an increase was consistently observed in cells treated with 5-HT (compare Fin. 4A with Fin. 5A and with

Fig. 1A of Sugita et al. 1994) (see also Baxter and Byrne 1989; 1990ab; White et al. 1994). Therefore pcpt-CAMP did not fully mimic 5-HT-induced modulation of the membrane currents at 20 mV, suggesting that some of the mechanisms underlying pcpt-CAMP and 5-HT-induced modulation of the membrane currents may differ. This result suggests, in particular, that 5-HT-induced increase in the membrane current at the end of the pulse is mediated by other second messenger/protein kinase systems, possibly PKC (Sugita et al. 1994). However, the pcpt-CAMP-induced decrease in the outward current at the beginning of the voltage-clamp pulse to 20 mV appeared to be similar to that by 5-HT, confirming part of the observation by Goldsmith and Abrams (1992). This result also seemed to support the involvement of CAMP in the selective modulation of activation kinetics of lkv (Hochner and Kandel 1992). Quantitative analysis, however, reveals some difference between pcpt-CAMP- and 5-HT-induced decrease in the currents at the beginning of the pulse. We found that pcpt-CAMP-induced decrease was twice as large as that by 5-HT (at 10 ms after the beginning of the pulse, pcpt-CAMP: 34.3 t 2.3%; 5-HT: 18.4 t 3.1%). We also found that addition of pcpt-CAMP (n = 3) to the bath that already contained 5-HT produced an additional, and rather profound reduction, of the outward currents at 20 mV (Fig. 5A). The pcpt-CAMP difference current in the presence of 5-HT is somewhat similar to the pcpt-CAMP difference current in the absence of 5-HT (compare Fig. 4A, 2, and Fig. 5A, 2). The reduction of outward currents at the end of pulses by additional application of pcpt-CAMP suggests that pcpt-CAMP antagonized the 5-HT-induced modulation of inactivation kinetics of 1k v. This reduction at the end of clamp-pulses cannot be explained by the additional reduction of lk,s by pcpt-CAMP, because the reduction of 1k s at -20 mV was completely occluded (Fig. 5B, I) by the preexposure to 5-HT. Therefore, the failure of 5-HT to occlude pcpt-CAMP-induced effects at 20 mV suggests that

1256

S. SUGITA,

Al

D. A. BAXTER,

A2

Step to +20 mV

AND J. H. BYRNE

pcpt-CAMP Difference

20 nA

40 nA

B1

Current (a - b)

Step to o mV a. ASW

Cl

Step to -20 mV

WL

a. ASW

b. +pcpt-CAMP

1 nA

’

30 ms 30 ms 4. Modulation of membrane currents by DMSO-pcpt-CAMP. Procedures were the same as those used for 5-HT (e.g., Fig. 1 of Sugita et al. 1994). Traces represent averaged data from 6 sensory neurons. At 20 mV, pcpt-CAMP reduced the outward current at both the beginning and the end of clamp pulses, which is qualitatively different from 5-HT-induced modulation (e.g., Fig. 54) (see also Baxter and Byrne 1989; Sugita et al. 1994; White et al. 1994). At -20 mV, pcpt-CAMP reduced an outward current with properties similar to IKs. FIG.

pcpt-CAMP has some mechanism not shared by 5-HT (or CAMP).

of action on lkv, that is

DISCUSSION

Baxter and Byrne (1989, 1990a) examined the effects of a relatively low concentration (50 PM) of pcpt-CAMP on spike duration, excitability and membrane currents in the sensory neurons. Although this concentration of pcptCAMP failed to occlude most of the 5-HT-induced slowly developing spike broadening, it mimicked and occluded the 5-HT-induced enhancement of excitability. In addition, pcpt-CAMP mimicked and occluded the 5-HT-induced reduction of 1&S. These effects of pcpt-CAMP also were mimicked by another CAMP analogue, 8-bromo-CAMP, and a derivative of forskolin, an activator of adenylyl cyclase (Baxter and Byrne 1989, 1990a,b). Therefore 50 PM of pcpt-CAMP appeared to induce CAMP-dependent, physiological effects in the sensory neurons. Several lines of evidence suggest that at least some of the additional effects, which require much higher concentrations, prod uced b y pcpt-CAMP may not be ph .ysiological. First, although high concentrations (500 PM to 1 mM) of

pcpt-CAMP did produce more broadening than reported by Baxter and Byrne (1990a), our results indicated that some components of this broadening were qualitatively different from that produced by 5-HT and CAMP. Second, some aspects of pcpt-CAMP- and 5-HT-induced broadening were additive. Third, another CAMP analogue, 8-bromo-CAMP (20-50 mM), did not mimic all of the actions of pcptCAMP. Fourth, pcpt-CAMP-induced modulation of outward currents at 20 mV was different from that of 5-HT, and pcpt-CAMP reduced the currents at 20 mV rather profoundly, even after preexposure to 5-HT. These results strongly indicate that some of the actions of high concentrations of pcpt-CAMP are CAMP independent. Interestingly, most of the actions of pcpt-CAMP on the waveform of action potentials were mimicked by pcptcGMP but not 8-bromo-cGMP. These results suggest that CAMP-independent effects of pcpt-CAMP may be mediated by parachlorophenylthio moiety of the compound. Moreover, the observation that both of pcpt-CAMP and pcptcGMP delayed the peak of action potentials suggests these compounds also may affect N ?a+currents. Although CAMP analogues and a derivati .ve of forskoli n consistently induce spike broadening, the degree of

CAMP-INDEPENDENT

A,

1257

EFFECTS OF A CAMP ANALOGUE

A2

Step to +20 mv

Difference

Currents

a. ASW b. ASW + 5HT 1pcpt-CAMP

(b-c)

c. (ASW + 5HT) + pcpt-CAMP

40 nA

IOnA w

B1

B2

Step to -20 mv

c. (ASW + 5HT)

M

+ pcpt-CAMP

1 nA

2 nA 40 ms

40 ms

FIG. 5. DMSO-pcpt-CAMP reduced the membrane currents, which already were modulated by 5-HT. Membrane currents were elicited by voltage-clamp pulses from -70 mV to 20 mV (A) or -20 mV (B) in artificial seawater (ASW, a) after the addition of 5-HT ( 10 PM, b) and after the addition of DMSO-pcpt-CAMP to the bath, which still contained 5-HT (c). DMSO-pcpt-CAMP was applied after preexposure to 5-HT for 9- 13.5 min. Traces are averaged data from 3 sensory neurons. At 20 mV, DMSO-pcpt-CAMP produced an additional reduction of the outward current. Reduction of the outward current by pcpt-CAMP at the end of clamp pulses is not because of the further reduction of ZKs, because the reduction of this outward current at -20 mV was occluded by preexposure to 5-HT. The DMSO-pcpt-CAMP difference current in the presence of 5-HT was isolated by subtracting the current response elicited in the presence of DMSO-pcpt-CAMP (and 5-HT) from the current response elicited in ASW containing 5-HT.

broadening appears to vary among experiments (Baxter and Byrne 1990a; Goldsmith and Abrams 1992; Hochner and Kandel 1992; present study). One of the factors that may contribute to these quantitative differences is the temperature at which the experim .ents were performed. Our work an .d Baxter and Byrne ( 1990a) wesre performed at 15 “C, whereas Goldsmith and Abrams ( 1992) and Hochner and Kandel (1992) were performed at room temperature. Moderate shifts in temperature can result in dramatic changes in neuronal function (e.g., Hakozaki et al. 1989; Karpenter et al. 1990; Matsumoto et al. 1988; Poulain et al. 1992; Redman and Berry 199 1-1993; Tamazawa et al. 199 1; Treistman and Grant 1990). Other factors that may contribute to these quantitative differences are the relative potencies of the CAMP analogues or derivatives of forskolin as the activators of PKA and their relative membrane permeabilities. Although considerable progress has been made in elucidating the role of second messengers and protein kinases in mediating the actions of SHT, a complete quantitative description of their temporal and spatial domains as

well as their interactions and the effects that they have on various substrates is not yet available. Further experiments will be required that utilize an a more extensi ve array of specific activators and inhibi tors applied under va rious experimental conditions. This work was supported by National Institute of Health Awards K05 MH-00649 and RO 1 NS- 19895. Address reprint requests to J. H. Byrne. Received 4 March 1994; accepted in final form 28 March 1994. REFERENCES BACSKAI, B. J., HOCHNER, B., MAHOAUT-SMITH, M., ADAMS, S. R., KAANG, B.-K., KANDEL, E. R., AND TSIEN, R. Y. Spatially resolved dynamics of CAMP and protein kinase A subunits in Aplysia sensory neurons. Science Wash. DC 260: 222-226, 1993. BAXTER, D. A. AND BYRNE, J. H. Serotonergic modulation of two potassium currents in pleural sensory neurons ofAplysia. J. Neurophysiol. 62: 665-679, 1989. BAXTER, D. A. AND BYRNE, J. H. Differential effects of CAMP and serotonin on membrane current, action potential duration, and excitability in

1258

S. SUGITA,

D. A. BAXTER,

somata of pleural sensory neurons of ApZysia. J. Neurophysiol. 64: 97% 990, 1990a. BAXTER, D. A. AND BYRNE, J. H. Reduction of voltage-activated K+ currents by forskolin is not mediated by CAMP in pleural sensory currents in ApZysia. J. Neurophysiol. 64: 1474- 1483, 1990b. BERNIER, L., CASTELLUCI, V. F., KANDEL, E. R., AND SCHWARTZ, J. H. Facilitatory transmitter causes a selective and prolonged increase in adenosine 3’5’-monophosphate in sensory neurons mediating the gill and siphon withdrawal reflex in ApZysia. J. Neurosci. 2: 1682- 169 1, 1982. BRAHA, O., DALE, N., HOCHNER, B., KLEIN, M., ABFZAMS, T. W., AND KANDEL., E. R. Second messengers involved in the two processes of presynaptic facilitation that contribute to sensitization and dishabituation in Apl.ysia sensory neurons. Proc. NatZ. Acad. Sci. USA 87: 20402044, 1990. BRAHA, O., EDMONDS, B., SACTOR, T., KANDEL, E. R., AND KLEIN, M. The contributions of protein kinase A and protein kinase C to the actions of 5-HT on L-type Ca*+ current of the sensory neurons in Aplysia. J. Neurosci. 13: 1839-1851, 1993. BRAUMANN, T., EMEUX, C., PETRIDIS, G., STOHRER, W.-D., AND JASTORFF, B. Hydrolysis of cyclic nucleotides by a purified cGMP-stimulated phosphodiesterase: structural requirements for hydrolysis. Biothem. Biophys. Acta 87 1, 199-206, 1986. CONNOLLY, B. J., WILLITS, P. B., WARRINGTON, B. H., AND MURRAY, K. J. 8-(4-chlorophenyl)thio-cyclic AMP is a potent inhibitor of the cyclic GMP-specific phosphodiesterase (PDE VA). Biochem. Pharmacol. 44: 2303-2306, 1992. FRANCIS, S. H., NOBLETT, B. D., TODD, B. W., WELLS, J. N., AND CORBIN, J. D. Relaxation of vascular and tracheal smooth muscle by cyclic nucleotide analogs that preferentially activate purified cGMP-dependent kinase. Mol. Pharmacol. 34: 506-5 17, 1988. GHIRARDI, M., BRAHA, O., HOCHNER, B., MONTAROLO, P. G., KANDEL, E. R., AND DALE, N. Roles of PKA and PKC in facilitation of evoked and spontaneous transmitter release at depressed and nondepressed synapses in Aplysia sensory neurons. Neuron 9: 479-489, 1992. GINGRICH, K. J., BAXTER, D. A., AND BYRNE, J. H. Mathematical model of cellular mechanisms contributing to presynaptic facilitation. Brain Res. Bull. 2 1: 5 13-520, 1988. GOLDSMITH, B. A. AND ABRAMS, T. W. CAMP modulates multiple K+ currents, increasing spike duration and excitability in ApZysia sensory neurons. Proc. Natl. Acad. Sci. USA 89: 1148 l-l 1485, 1992. HAKOZAKI, S., MATSUMOTO, M., AND SASAKI, K. Temperature-sensitive activation of G-protein regulating the resting membrane conductance of Aplysia neurons. Jpn. J. Physiol. 39: 115- 130, 1989. HALL, A. K. Influence of cyclic AMP and serum factors upon expression of a retinoid-responsive gene in neuroblastoma cells. J. Mol. Neurosci. 3: 155-163, 1992. HOCHNER, B. AND KANDEL, E. R. Modulation of a transient K+ current in the pleural sensory neurons of ApZysia by serotonin and CAMP: implications for spike broadening. Proc. NatZ. Acad. Sci. USA 89: 11476- 11480, 1992. HOCHNER, B., KLEIN, M., SCHACHER, S., AND KANDEL, E. R. Action potential duration and the modulation of transmitter release from the sensory neurons of Aplysia in presynaptic facilitation and behavioral sensitization. Proc. Natl. Acad. Sci. USA 83: 84 lo-84 14, 1986. JARRARD, H. E., GOLDSMITH, B. A., AND ABRAMS, T. W. In ApZysia sensory neurons, the neuropeptide SCP, and serotonin differ in efficacy both in modulating cellular properties and in activating adenylyl cyclase: implications for mechanisms underlying presynaptic facilitation. Brain Res. 616: 188-199, 1993. KANDEL, E. R. AND SCHWARTZ, J. H. Molecular biology of learning: modulation of transmitter release. Science Wash. DC 2 18: 433-443, 1982. KARPENTER, D. O., KING, M. V., AND AIRAPETIAN, S. N. The cross desensitization and modulation of Cl- currents activated by gamma-aminobutyric acid and L- glutamate in the isolated neurons of Aplysia. BioZ. Nauki. Most. 6: 44-50, 1990. KLEIN, M. The role of action potential broadening in synaptic facilitation at soma-soma synapses. Sot. Neurosci. Abstr. 19: 18, 1993. KRAMER, R. H., LEVITAN, E. S., CARROW, G. M., AND LEVITAN, I. B. Modulation of a subthreshold calcium current by the neuropeptide FMRFamide in ApZysia neuron R15. J. Neurophysiol. 60: 1728-l 738, 1988a. KRAMER, R. H., LEVITAN, E. S., WILSON, M. P., AND LEVITAN, I. B. Mechanism of calcium-dependent inactivation of a potassium current in

AND J. H. BYRNE

Aplysia neuron R15: interaction between calcium and cyclic AMP. J. Neurosci. 8: 1804- 18 13, 1988b. LAMBERT, N. A. AND HARRISON, N. L. Analogs of cyclic AMP decrease gamma-aminobutyric acid A receptor-mediated chloride current in cultured rat hippocampal neurons via an extracellular site. J. Pharmacol. Exp. Ther. 255: 90-94, 1990. LANGE-CARTER, C. A., VUILLEQUEZ, J. J., AND MALKINSON, A. M. 8chloroadenosine mediates 8-chloro-cyclic AMP-induced down-regulation of cyclic AMP-dependent protein kinase in normal and neoplastic mouse lung epithelial cells by a cyclic AMP-independent mechanism. Cancer Res. 53: 393-400, 1993. LEIDENHEIMER, N. J., BROWNING, M. D., DUNWIDDIE, T. V, HAHNER, L. D., AND HARRIS, R. A. Phosphorylation-independent effects of second messenger system modulators on gamma-aminobutyric acid A receptor complex function. Mol. Pharmacol. 38: 823-828, 1990. MATSUMOTO, M., HAKOZAKI, S., TAKAHASHI, J., SASAKI, K., AND SETO, M. Temperature effects on dopamine-induced inhibitory and excitatory responses recorded from Aplysia ganglion cells. Jpn. J. Physiol. 38: 109113, 1988. MERCER, A. R., EMPTAGE, N. J., AND CAREW, T. J. Pharmacological dissociation of modulatory effects of serotonin in Aplysia sensory neurons. Science Wash. DC254: 181 l-1813, 1991. MEYERS, R. B., JR. AND MILLER, J. P. Analogs of cyclic AMP and cyclic GMP: general methods of synthesis and the relationship of structure to enzyme activity. L(fe Sci. 14: 10 19- 1040, 1974. MILLER, J. P., BOSWELL, K. H., MEYER, R. B, CHRISTENSON, L. F., AND ROBINS, R. K. Synthesis and enzymatic and ionotropic activity of some new 8-substituted and 6,8-disubstituted derivatives of adenosine cyclic 3’:5’-monophosphate. J. Med. Chem. 23: 242-25 1, 1980. NOEL, F., SCHOLZ, K. P., ESIUN, A., AND BYRNE, J. H. Common set of proteins in ApZysia sensory neurons affected by an in vitro analogue of long-term sensitization training, 5-HT and CAMP. Brain Res. 568: 6775, 1991. OCORR, K. A. AND BYRNE, J. H. Membrane responses and changes in CAMP levels in ApZysia sensory neurons by serotonin, tryptamine, FMRFamide and small cardioactive peptide B (SCP,). Neurosci. Lett. 55: 113-118, 1985. POLLOCK, J. P., BERNIER, L., AND CARMARDO, J. S. Serotonin and cyclic adenosine 3’5’-monophosphate modulate the S potassium current in tail sensory neurons in pleural ganglion of Aplysia. J. Neurosci. 5: 18621871, 1985. POULAIN, B., DE PAIVA, A., DOLLY, J. O., WELLER, U., AND TAUC, L. Differences in the temperature dependencies of uptake of botulinurn and tetanus toxins in Apl.ysia neurons. Neurosci Lett. 139: 289-292, 1992. REDMAN, R. S. AND BERRY, R. W. Temperature-dependent peptidergic feedback: potential role in seasonal egg laying in Aplysia. J. Neurosci. 11: 1780-1785, 1991. REDMAN, R. S. AND BERRY, R. W. ProELH-related peptides: influence on bag cell CAMP levels. Mol. Brain Res. 15: 2 16-220, 1992. REDMAN, R. S. AND BERRY, R. W. Temperature-dependent stimulation and inhibition of adenylate cyclase by ApZysia bag cell peptides. Mol. Brain Res. 17: 245-250, 1993. SACKTOR, T. C. AND SCHWARTZ, J. H. Sensitizing stimuli cause translocation of protein kinase C in Aplysia sensory neurons. Proc. NatZ. Acad. Sci. USA 87: 2036-2039, 1990. SANDBERG, M., BUTT, E., NOLTE, C., FISCHER, L., HALBR~~GGE, M., BELTMAN, J., JAHNSEN, T., GENIESER, H.-G., JASTORFF, B., AND WALTER, U. Characterization of Sp-5,6-dichloro- 1+d-ribofuranosylbenzimidazole-3’,5’-monophosphate (Sp-5,6-DCl-cBIMPS) as a potent and specific activator of cyclic AMP-dependent protein kinase in cell extracts and intact cells. Biochem. J. 279, 52 l-527, 199 1. SIEGELBAUM, S. A., CAMARDO, J. S., AND KANDEL E. R. Serotonin and cyclic AMP close single K+ channels in Aplysia sensory neurons. Nature Land. 299: 4 13-4 17, 1982. SIMON, L. N., SHUMAN, D. A., AND ROBINS, R. K. The chemistry and biological properties of nucleotides related to nucleoside 3’:5’-cyclic phosphate. Adv. Cyclic Nucleotide Res. 3: 225-353, 1973. SOSSIN, W. S. AND SCHWARTZ, J. H. Selective activation of Ca*+-activated PKCs in Aplysia neurons by 5-HT. J. Neurosci. 12: 1160-l 168, 1992. SUGITA, S., GOLDSMITH, J. R., BAXTER, D. A., AND BYRNE, J. H. Involvement of protein kinase C in serotonin-induced spike broadening and synaptic facilitation of sensorimotor connections in Aplysia. J. Neurophysiol. 68: 643-65 1, 1992.

CAMP-INDEPENDENT

EFFECTS OF A CAMP ANALOGUE

SUGITA, S., BAXTER, D. A., AND BYRNE, J. H. Activators of protein kinase C mimic serotonin-induced modulation of a voltage-dependent potassium current in pleural sensory neurons ofApZy.sia. J. Neurophysiol. 72: 1240-1249, 1994. SWEATT, J. D. AND KANDEL, E. R. Persistent and transcriptionally dependent increase in protein phosphorylation in long-term facilitation of ApZysia sensory neurons. Nature Lond. 339: 5 l-54, 1989. TAMAZAWA,Y.,MATSUMOTO, M., KUDO, A., ANDSASAKI, KPotassium ion channels operated by receptor stimulation can be activated simply by raising temperature. Jpn. J. Physiol. 4 1: 117- 127, 199 1. TREISTMAN, S. N. AND GRANT, A. J. Attributes of an alcohol-sensitive and an alcohol-insensitive transient potassium current in ApZysia neurons. AZcohoZ.CZin. Exp. Res. 14: 595-599, 1990.

1259

WALSH, J. P. AND BYRNE, J. H. Modulation of a steady-state Ca2+-activated, K+ current in tail sensory neurons of ApZysia: role of serotonin and CAMP. J. Neurophysiol. 6 1: 32-44, 1989. WHITE, J. A., BAXTER, D. A., AND BYRNE, J. H. Analysis of the modulation by serotonin of a voltage-dependent potassium current in sensory neurons of Aplysia. Biophy. J. 66: 7 10-7 18, 1994. ZORN, M., MARONDE, E., JASTROFF, B., AND RICHTER-LANDSBERG,~. Differential effectsof two structurally related N6-substituted CAMP analogues on C6 glioma cells. Eur. J. Cell BioZ. 60: 35 l-357, 1993. ZHU, X. 2. AND CHUANG, D. M. Differential regulation by butyrate and dibutyryl cyclic AMP of delta-opioid, alpha 2-adrenergic, and muscarinic cholinergic receptors in NCB-20 cells. J. Neurochem. 50: 17-26, 1988.