Administration of cocaine and amphetamine increases cocaine- and amphetamine-regulated transcript (CART) expression in the rat striatum (Douglass et al., 1995). CART mRNA is highly expressed in different parts of the human and rat brain, including hippocampus (Douglass et al., 1995; Couceyro et al., 1997; Kuhar and Yoho, 1999; Hurd and Fagergren, 2000). The presence of CART peptide 55–102 immunoreactivity in dense core vesicles of axon terminals suggests that the peptide may be released and may act as a neuromodulator (Smith et al., 1997) to induce neurophysiological and behavioral effects. Little is known, however, about CART peptide-responsive cells, receptor(s), or intracellular signaling mechanisms that mediate CART peptide action. Here we show that CART peptide 55–102 inhibits voltage-dependent intracellular Ca2+ signaling and attenuates cocaine enhancement of depolarization-induced Ca2+ influx in rat hippocampal neurons. The inhibitory effect of CART peptide 55–102 on Ca2+ signaling is likely mediated by an inhibition of L-type voltage-gated Ca2+ channel activity via a G-protein-dependent pathway. These results indicate that voltage-gated Ca2+ channels in hippocampal neurons are targets for CART peptide 55–102 and suggest that CART peptides may be important in physiology and behavior mediated by the hippocampus, such as certain forms of learning and memory.
Cocaine- and amphetamine-regulated transcript (CART) peptides are biologically active peptides that mediate feeding and are implicated in psychomotor stimulant drug abuse and stress response (Elias et al., 1998; Couceyro and Lambert, 1999;Kuhar and Dall Vechia, 1999; Kask et al., 2000; Kuhar et al., 2000). CART peptides are processed into shorter biologically active forms from either one or two propeptides in rat (Kristensen et al., 1998; Thim et al., 1998a). CART peptide 55–102 is biologically active and found in the rat brain (Kristensen et al., 1998; Thim et al., 1998a). The human CART cDNA sequence predicts only the shorter propeptide of 116 residues found in the rat (Douglass et al., 1995). (The peptide numbering used here is based on the longer rat propeptide of 102 residues excluding the leader sequence.) The three disulfide bridges (C68–C86, C88–C101, and C74–C94 using rat CART numbering) are present in the physiologically active CART peptide 55–102 (Thim et al., 1998a) and may stabilize its three-dimensional structure, suggesting that disruptions of these disulfide bridges impair biological activity (Kristensen et al., 1998).
CART expression increases in the striatum but particularly in the nucleus accumbens within 1 hr after administration of cocaine or amphetamine (Douglass et al., 1995; Kuhar and Yoho, 1999; Smith et al., 1999; Hurd and Fagergren, 2000). CART mRNA and CART peptides are normally expressed throughout the brain, especially in those areas involved in motivation, reward, and feeding, such as the nucleus accumbens, amygdala, and hypothalamus (Douglass et al., 1995; Couceyro et al., 1997; Koylu et al., 1997, 1998). Consistent with this expression pattern, exogenously applied CART peptides induce many marked behavioral changes, including inhibition of feeding (Lambert et al., 1998a,b; Thim et al., 1998a) and heightened anxiety (Kask et al., 2000).
In addition to the hypothalamus and amygdala, CART peptides are highly expressed in the hippocampus of both rodents and humans (Douglass et al., 1995; Koylu et al., 1998; Hurd and Fagergren, 2000), but their function and the effects of cocaine in the hippocampus are not known. CART mRNA is expressed in granule cells of the dentate gyrus, and CART peptides are found in the dentate gyrus CA3 and CA2 fields (Koylu et al., 1998; Hurd and Fagergren, 2000). The subcellular localization of CART peptides within dense core vesicles of axon terminals (Smith et al., 1997; Couceyro and Lambert, 1999) suggests that the peptides may be released from neurons into the extracellular space. Many neuropeptides, such as neuropeptide Y, which plays a significant role in the regulation of feeding (Flynn et al., 1999), interact with specific receptors on the plasma membrane that are coupled to second messenger systems (Qian et al., 1997). Intracellular Ca2+ signaling is intimately involved in synaptic activity and neurophysiological plasticity of hippocampal neurons (Alkon et al., 1998; Magee et al., 1998). Thus, we examined the effects of CART peptide 55–102, a biologically active form of rat CART (Thim et al., 1998a,b), on intracellular Ca2+ signals and voltage-gated Ca2+ channels in cultured rat hippocampal neurons.
MATERIALS AND METHODS
Rat hippocampal neurons (embryonic day 18) isolated as described previously (Brewer et al., 1993; Brewer and Price, 1996) were purchased from BrainBits (Springfield, IL). The neurons and undifferentiated PC12 cells were plated at a density of 4 × 105 cells/cm2on poly-l-lysine-coated rectangular glass coverslips and used on days 6–12 after plating.
The cells were incubated with 1 μm fura-2 AM (Molecular Probes, Eugene, OR) for 1 hr in the standard recording medium (see below), washed, and mounted in the chamber of the spectrofluorometer (FP-750; Jasco, Tokyo, Japan). The standard recording medium contained (in mm): 140 NaCl, 5 KCl, 1 MgCl2, 1.5 CaCl2, 5 glucose, and 10 HEPES, pH 7.4 (NaOH). High [K+] medium contained 30 mm KCl in place of NaCl. The chamber was continuously perfused at 7 ml/min. The media were changed, and the drugs were applied through the perfusion system, which took <0.5 sec. Measurements of cytosolic Ca2+concentration were performed as described previously (Grynkiewicz et al., 1985). Fura-2 AM was excited at 340 and 380 nm, and the emission was measured at 510 nm. The background fluorescence was determined after each experiment by quenching fura-2 fluorescence with 10 mm MnCl2 in nominally Ca2+-free medium in the presence of 5 μm ionomycin. The background fluorescence at 340 and 380 nm was subtracted from the respective experimental readings, and the ratio of signals at 340 and 380 was obtained. The data were digitized at 2 sec intervals.
The Ca2+ channel openings were recorded in the cell-attached configuration of the patch clamp with an AxoPatch 200A amplifier (Axon Instruments, Union City, CA). The pipette solution contained (in mm): 90 BaCl2, 10 TEA-Cl, and 10 HEPES, pH 7.3 (TEA-OH). The bath solution contained (in mm): 140 potassium gluconate, 10 EGTA, 10 HEPES, 15 NaCl, 3 MgCl2, and 10 glucose, pH 7.4 (KOH). The output of the amplifier was filtered through an eight-pole low-pass filter unit (Frequency Devices, Haverhill, MA) and digitized by an ITC-16 analog-to-digital/digital-to-analog interface (Instrutech, Port Washington, NY) attached to an Apple (Cupertino, Ca) Power Macintosh computer running Pulse (HEKA, Lambrecht, Germany). The data were analyzed using Patch Machine (www.hoshi.org) and IgorPro (WaveMetrics, Lake Oswego, OR).
Two independent sources of CART peptide 55–102 were used in this study: rat CART peptide 55–102 from Peptides International (Louisville, KY; lot 490129) and a recombinant rat CART peptide 55–102 with an N-terminal histidine (HIS) tag (Fritz et al., 2000). Afterin vitro reduction and refolding, the recombinant CART peptide 55–102 HIS tag was biologically active, because it reduced food intake in rats. Inhibition of food intake was not seen if the peptide was not refolded, and an improperly folded inactive peptide was used as a control.
Bay K 8644, staurosporine, genistein, cyclosporine A, and pertussis toxin were obtained from Calbiochem (La Jolla, CA). Other reagents were obtained from Sigma (St. Louis, MO). Experiments were performed at room temperature (20–23°C).
CART peptide 55–102 dose-dependently reduced the amplitude of cytosolic Ca2+ signals elicited by K+ depolarization in fura-2-loaded rat hippocampal cultured neurons (Fig.1 A,B). At 1 μm, CART peptide 55–102 decreased the total Ca2+ signal during 200 sec depolarization by 48 ± 14% (n = 4). The inhibitory effect of CART peptide 55–102 on Ca2+ signals elicited by depolarization was long-lasting and continued for at least 30 min after the peptide was washed out. The reduction of Ca2+ signals was noticeable at concentrations of CART peptide 55–102 as low as 250 nm and had an EC50 of ∼600 nm (Fig. 1 B).
Pretreatment of the neurons with fluorocitrate (0.1 mm for 30 min), a selective glial metabolic poison (Fonnum et al., 1997), altered neither the high K+-induced Ca2+ signals nor the ability of CART peptide 55–102 to inhibit the signals (data not shown;n = 6). This suggests that glial cells potentially present in the preparation made an insignificant contribution to the results obtained.
The specificity of the CART peptide 55–102 effects on the Ca2+ signal was tested in several ways. We used CART peptide 55–102 from two independent sources synthesized in different ways (see Materials and Methods), and they were equally effective in inhibiting the Ca2+ signal. Recombinant rat CART peptide 55–102 purified from bacteria (Fritz et al., 2000) inhibited high-K+-induced increases in Ca2+ signals in hippocampal cultures with a potency similar to that of CART peptide 55–102 made by solid-phase chemical synthesis. The biologically active CART peptide requires appropriate formation of three disulfide bonds (Thim et al., 1998a,b; Kuhar and Dall Vechia, 1999). Inappropriate formation of these bonds eliminates the ability of CART peptide to inhibit food intake (Fritz et al., 2000). Reduction of the disulfide links destabilizes the structure, and the peptide loses its ability to inhibit feeding. We reduced active CART peptide 55–102 with 5 mm dithiothreitol (DTT) for 1 hr at room temperature immediately before use. The DTT-treated CART peptide 55–102 had no effect on Ca2+ signals in hippocampal cultures (Fig. 1 C). The corresponding control solution with 5 mm DTT left at room temperature for 1 hr without CART peptide failed to affect K+-induced Ca2+ signals (data not shown;n = 4). Recombinant CART peptide 55–102 that did not undergo the refolding step (see Materials and Methods) and did not inhibit feeding (Fritz et al., 2000) was also much less effective in reducing the Ca2+ signal (15 ± 4%;n = 4) than the wild-type CART peptide. These results suggest that the observed inhibition of high-K+-induced Ca2+ signal requires a properly folded peptide. Furthermore, we found that the CART action was tissue-specific. CART peptide 55–102 (up to 1 μm) failed to alter the Ca2+ signals in undifferentiated rat pheochromocytoma (PC12) neurosecretory cells (Fig.1 D), illustrating its cell-specific action.
The total cytoplasmic Ca2+ transient elicited by depolarization results from a number of Ca2+ transport mechanisms that balance Ca2+ influx and extrusion. CART peptide 55–102 did not significantly alter basal Ca2+ levels in the absence of depolarization (Fig. 1 A). The rate of recovery of intracellular Ca2+ concentration after the depolarization-induced influx was also unaffected (Figs.1 A, 2 A). These observations suggest that the peptide may modulate depolarization-activated Ca2+ influx rather than affecting a resting Ca2+ leak into the cytoplasm or Ca2+ extrusion mechanisms.
To assess the contribution of Ca2+-induced Ca2+ release from endoplasmic reticulum (ER) Ca2+ stores, cells were treated with 5 μm thapsigargin (TG), which depleted intracellular Ca2+ stores by inhibiting ER ATPase (Thastrup et al., 1990), in nominally Ca2+-free medium for 15 min. The pretreatment with TG did not markedly alleviate the inhibitory effects of CART peptide 55–102 on depolarization-induced Ca2+ signals (Fig.2, compare A, B;n = 8). This observation argues against modulation of Ca2+-induced Ca2+ release by CART peptide 55–102. In TG-treated cells, the kinetics of Ca2+influx inactivation and the recovery of resting Ca2+ levels after depolarization in the presence of CART peptide 55–102 were unchanged (Fig.2 B). Thus, the peptide has a negligible effect on plasma membrane Ca2+ extrusion mechanisms.
CART peptide 55–102 was unable to inhibit Ca2+ signals in hippocampal neurons in the presence of nifedipine, which reduces the opening frequency and duration of voltage-gated L-type Ca2+channels (Hess et al., 1984; Miller, 1987). Application of nifedipine reduced the amplitude of the high-K+-induced Ca2+ signal by ∼55% (Fig.2 C), indicating that L-type voltage-gated Ca2+ channels represent a major contributor to the Ca2+ signal measured. In the presence of nifedipine, the Ca2+signals recorded before and after CART peptide 55–102 application were essentially indistinguishable (Fig. 2 C; n = 4). The ineffectiveness of CART peptide 55–102 in the presence of nifedipine suggests that the peptide may inhibit L-type Ca2+ channels, although the possibility that the channels are not modulated in the presence of the antagonist cannot be totally ruled out (Dolphin, 1999). In contrast, CART peptide 55–102 was still effective in the presence of Bay K 8644, an L-type Ca2+ channel agonist (Fig. 2 D, compare Bay K 8644 trace, Bay K 8644 + CART trace; n = 5). These results obtained with nifedipine and Bay K 8644 suggest that CART peptide 55–102 preferentially inhibits L-type Ca2+ channels.
We directly examined the action of CART peptide 55–102 on voltage-gated Ca2+ channels using the patch-clamp method. Voltage-gated Ca2+channel openings were recorded in the cell-attached configuration using Ba2+ as the charge carrier, and CART peptide 55–102 was applied to the bath solution outside the recording pipette. Hippocampal neurons in culture (>6 d) express a high density of L-type Ca2+ channels (Porter et al., 1997). We recorded openings with a unitary current amplitude of ∼2 pA at −5 mV (Fig. 3 A,Control), which we identified as L-type Ca2+ channel openings on the basis of their sensitivity to nifedipine and Bay K 8644, their voltage dependence of activation, and their inactivation kinetics (data not shown). Application of CART peptide 55–102 to the bath markedly reduced the number of openings elicited by depolarization (Fig.3 A, CART). The unitary current amplitudes were unchanged before and after application of CART peptide 55–102, and the ensemble averages showed that CART peptide 55–102 (1 μm) decreased the mean current amplitude by 80–90% (Fig. 3 A, bottom traces) without noticeably altering the channel kinetics. The inhibitory effect of CART peptide 55–102 was typically observed within 1 min of application (Fig. 3 B). Similar results were observed in all seven patches examined (Fig. 3 D). Bay K 8644, which promotes L-type Ca2+ channel openings, increased the frequency and the mean open duration of channel openings in the cell-attached configuration (Fig. 3 C, Control+Bay K 8644), confirming that the openings indeed represented L-type Ca2+ channel openings. CART peptide 55–102 applied to the bath produced a dramatic reduction in the number of channel openings recorded in the presence of Bay K 8644 (Fig.3 C, CART+Bay K 8644). These effects were observed in all five patches examined (Fig. 3 E). Thus, CART peptide modulates the same channels that are upregulated by Bay K 8644, consistent with the idea that CART peptide 55–102 inhibits L-type Ca2+ channels. Because in the patch clamp experiments the bath solution contained the Ca2+ chelator EGTA, and the recording solution contained Ba2+, it is unlikely that the CART peptide caused noticeable Ca2+ influx across the entire cell membrane, which in turn inhibited the Ca2+channel activity via a Ca2+-dependent inactivation mechanism. The finding that CART peptide 55–102 added to the bath outside the recording pipette reduced the channel activity suggests that the action of CART peptide 55–102 is not mediated by the direct blocking or membrane-delimited effect on the channels but is likely to involve diffusible second messengers.
The above data suggest that the CART peptide receptors may be present in rat hippocampal neurons and that diffusible second messengers are likely to mediate signaling of CART peptides. L-type Ca2+ channel activity is subject to modulation by G-proteins, cAMP-dependent protein phosphorylation mechanisms, or both (Dolphin, 1998, 1999; Jones, 1998). We pharmacologically examined which putative second messengers might mediate the action of CART peptide 55–102 on voltage-gated Ca2+ channels to reduce the Ca2+ signals elicited by depolarization. CART peptide 55–102 inhibited the Ca2+signal in the presence of 100 nm staurosporine, a nonselective serine-threonine protein kinase inhibitor that blocks Ca2+-calmodulin kinase, PKA, PKC, and PKG (Ruegg and Burgess, 1989; Fig.4 A; n = 4). CART peptide 55–102 also reduced the Ca2+ signal in the presence of 5 μm genistein (Akiyama et al., 1987), an inhibitor of protein-tyrosine kinases (Fig. 4 B;n = 4). The inhibitory effect of CART peptide 55–102 was also unaltered by the presence of 5 μmcyclosporine A, which inhibits protein phosphatase 2B, or calcineurin (Fig. 4 C; n = 4). In contrast, incubation of the neurons for 18 hr with 50 ng/ml pertussis toxin, an inhibitor of Gi, Go, and Gt regulatory proteins (Ui and Katada, 1990;Dolphin, 1998), completely abolished the inhibitory effect of CART peptide 55–102 (Fig. 4 D; n = 4). This sensitivity to pertussis toxin suggests that inhibition of the Ca2+ signal by CART peptide 55–102 may be exerted via Gi and Go proteins, which are known to inhibit of L-type Ca2+ channels in some cells (Degtiar et al., 1997; Farrugia, 1997; Zeng et al., 1999).
The ability of cocaine and amphetamine to selectively upregulate CART mRNA in the rat brain (Douglass et al., 1995) suggested that CART peptides may modulate some of the physiological and behavioral actions of stimulant drugs of abuse (Cooper and van der Hoek, 1993; Berke and Hyman, 2000). Cocaine has been shown to increase neuronal excitability and to enhance Na+ channel activity (Zhai et al., 1997) and cytosolic Ca2+ levels (Onaivi et al., 1996) in hippocampus neurons. The drug also potentiates L-type Ca2+ channel activity in cardiac myocytes (Premkumar, 1999). We found that cocaine dose-dependently potentiated intracellular Ca2+ signals produced by K+ depolarization in cultured hippocampal neurons (Fig.5 A,B). CART peptide 55–102 (1 μm) attenuated the enhancement of the Ca2+ signal caused by cocaine and shifted the dose–response curve to the right, increasing the EC50 for cocaine from 0.23 to 2.6 μm (Fig. 5 B). Moreover, application of 5 μm cocaine partially reversed inhibition of Ca2+ transients observed in the presence of CART peptide 55–102 (Fig. 5 C; n= 5).
Previous studies have elucidated a role for CART peptides in a variety of physiological effects, including inhibition of feeding and increasing anxiety. The cellular mechanism of the CART action was, however, for the most part unknown. We show here that CART peptide 55–102 reduces the depolarization-induced intracellular Ca2+ signal by inhibiting voltage-gated Ca2+ channels in cultured rat hippocampal neurons. The results presented here indicate that CART peptide 55–102 is a neuropeptide modulator of voltage-gated Ca2+ channels in hippocampal neurons. Several lines of evidence indicate that this observation is of biological relevance. First, the observed inhibition of the Ca2+ signal requires a biologically active form of CART. Rat CART peptide 55–102 contains three disulfide links to stabilize the biologically active structure (Thim et al., 1998a; Kuhar and Dall Vechia, 1999). Disruption of these disulfide links by reducing agents, such as DTT, is known to destabilize the structure and to render the peptide ineffective in inhibiting feeding (Fritz et al., 2000). Our results show that application of the reducing agent DTT to solid-phase synthesized CART peptide 55–102 or omitting the refolding step in the synthesis of recombinant CART virtually eliminates the ability of CART peptide 55–102 to reduce the depolarization-induced Ca2+ signal in hippocampal neurons. Second, the effect of CART peptide 55–102 is dose-dependent (Fig.1 B), and the concentration of CART 55–102 required to inhibit the Ca2+ signal is within a range expected for a peptide neuromodulator. We estimated that the EC50 value of the CART action was ∼600 nm. Other neuromodulator peptides have similar effective concentration ranges. For example, the inhibition of presynaptic Ca2+ entry in rat hippocampus by neuropeptide Y has an EC50 value of ∼1 μm (Qian et al., 1997). Opioid peptides, such as enkephalin, inhibit voltage-gated Ca2+channels effectively within a concentration of range of 1–10 μM (Hernandez-Guijo et al., 1999). Third, the action of CART peptide 55–102 is tissue-specific, consistent with the idea that some cells contain specific receptors designed for CART peptides. We observed robust inhibition of depolarization-activated Ca2+ signals in cultured rat hippocampal neurons, even in the presence of a glial-specific poison. In contrast, CART peptide 55–102 did not alter the Ca2+ signal in PC12 cells derived from adrenal chromaffin cells, even though they do contain a variety of voltage-gated Ca2+ channels, including L-type Ca2+ channels (McCullough et al., 1998). These Ca2+ measurements, therefore, suggest that the rat hippocampal neurons may contain specific CART peptide receptors that mediate the inhibitory action of the peptide on Ca2+ signaling. The rat hippocampus may represent a good source of CART peptide receptors for biochemical and molecular studies.
Both cytosolic Ca2+ measurements and electrophysiological results indicate that CART peptide 55–102 inhibits voltage-gated Ca2+ channel openings. Although CART peptide 55–102 inhibits the depolarization-induced Ca2+ signal, it alters neither the basal Ca2+ level nor the recovery time course after depolarization. These observations show that Ca2+ influx through voltage-gated Ca2+ channels is inhibited by CART peptide 55–102. The single-channel data corroborate this idea; CART peptide 55–102 applied to the bath decreases the frequency and the mean open duration of the Ca2+ channel openings. We suggest that CART peptide 55–102 preferentially inhibits L-type Ca2+ channels for the following reasons. First, the CART-induced inhibition of the Ca2+ signal is abolished by nifedipine, an L-type Ca2+ channel inhibitor. If L-type channels are already inhibited by nifedipine, CART peptide 55–102 should not have any effect. However, this observation alone does not definitively argue that CART inhibits L-type Ca2+ channels, because some modulatory pathways of voltage-gated Ca2+ channels may not operate normally in the presence of dihydropyridines (Dolphin, 1999). Second, our single-channel recordings show that the channel openings were prolonged by Bay K 8644, and those long openings were in turn inhibited by CART peptide 55–102. These results strongly argue that CART peptide 55–102 inhibits voltage-gated L-type Ca2+ channels, and their inhibition contributes to the reduced Ca2+ signal observed with CART peptide 55–102. It is possible that other Ca2+ channel types are also modulated by CART peptide 55–102; however, our experiment did not address this possibility.
The inhibitory modulation of L-type Ca2+channels by CART peptide 55–102 is likely to involve a diffusible intracellular second messenger and not likely to be solely mediated by a membrane-delimited system. Application of the peptide to the bath inhibited the channel activity recorded within the patch-clamp pipette (Fig. 3 A). The results obtained using pharmacological inhibitors of various intracellular signaling cascades implicate a pertussis toxin-dependent G-protein system in the CART action. Staurosporine, genistein, and cyclosporine A, which have been shown to alter serine-threonine kinase-, tyrosine kinase-, and phosphatase 2B-dependent signaling pathways, did not alter the inhibitory action of CART peptide 55–102 on the Ca2+ signal (Fig. 4 A–C). L-type Ca2+ channels are often upregulated by peptide neuromodulators via G-protein-dependent pathways (Dolphin, 1999). Our results indicate, however, that CART peptide 55–102 inhibits L-type Ca2+ channels in a pertussis toxin-sensitive manner (Fig. 4 D). This sensitivity to pertussis toxin suggests that inhibition of the Ca2+ signal by CART peptide 55–102 may be exerted via Gi and Goproteins (Ui and Katada, 1990; Dolphin, 1998). Although less common than upregulation, G-protein-dependent inhibition of L-type Ca2+ channels is not totally unprecedented. A similar mechanism of L-type Ca2+ channel inhibition has been demonstrated for the action of neuropeptide Y, somatostatin, galanin (Degtiar et al., 1997; Zeng et al., 1999), prostaglandins (Yamamoto et al., 1999), opioids, and catecholamines (Hernandez-Guijo et al., 1999). The long-lasting nature of the CART action may be surprising, considering the putative G-protein-dependent mechanism. One possible explanation is that CART peptide 55–102 may be difficult to wash out and that the observed long-lasting effect may not reflect the molecular nature of the underlying intracellular signaling cascade.
The exact physiological processes and behaviors modulated by CART peptide 55–102 effects on Ca2+homeostasis in the hippocampal neurons remain to be investigated. Our studies with cocaine do offer some clues. Cocaine increases the total amount of depolarization-induced Ca2+influx into the cultured hippocampal neurons, and this effect is antagonized by application of CART peptide 55–102 (Fig.5 A,B). The hippocampus has been implicated in the association of environmental cues and the affective states produced by drugs (White, 1996; Vorel et al., 2001). Electrical stimulation of the hippocampus elicits cocaine seeking in rats after extinction of cocaine self-administration behavior (Vorel et al., 2001), suggesting that hippocampal neurons are deeply involved in the learning and memory mechanisms associated with cocaine and possibly CART peptide action. Because CART peptides are rapidly produced in response to cocaine application at least in some cells (Douglass et al., 1995; Hurd and Fagergren, 2000), it is possible that the CART-mediated inhibition of depolarization-induced Ca2+ flux represents a homeostatic feedback response to counteract the enhanced Ca2+ entry induced by cocaine. The CART-mediated inhibition of voltage-gated Ca2+ channels described here may contribute to the drug tolerance observed with repeated use and also to the withdrawal symptoms and conditioned responses acquired during drug use (Gawin, 1991; Hyman, 1996; Breiter et al., 1997). Although cocaine does upregulate L-type Ca2+ channels in cardiac myocytes (Premkumar, 1999), it should be noted that our results do not directly address whether cocaine increases the total influx of Ca2+ in hippocampal neurons by modulating the hippocampal L-type Ca2+ channels.
The cytosolic Ca2+ measurements and single-channel measurements suggest that hippocampal neurons may possess membrane receptors for CART and that application of CART peptide 55–102 inhibits voltage-gated Ca2+ channels in a G-protein-dependent manner. CART peptides inhibit food intake and may mediate the central action of leptin (Elias et al., 1998; Thim et al., 1998a). Our observations further expand the role of CART peptides in the brain, but the exact behaviors affected remain unknown. It is quite possible that CART peptides help shape the neuronal plasticity associated with learning and memory mediated by the hippocampus.
This work was supported in part by National Institutes of Health Grants GM57654 (T.H.) and HL14388 (T.H.), the Pharmaceutical Research and Manufacturers Association Foundation (P.R.C.), the Schweppe Foundation (P.R.C.), and the National Alliance for Research on Schizophrenia and Affective Disorders (P.R.C.). We thank Heather Daggett for culturing hippocampal cells.
O.Y. and J.C. contributed equally to this work.
Correspondence should be addressed to Toshinori Hoshi, Department of Physiology and Biophysics, University of Iowa, Bowen Science Building, Iowa City, IA 52242. E-mail:.