Slice preparation and artificial cerebrospinal fluid

Brain slices were prepared from 14- to 30-day-old C57/Bl6 mice (Charles River Laboratories). Animals were decapitated following induction of deep anesthesia with isoflurane (Aerrane, Baxter Pharmaceuticals, inhalation). A block of the brain containing the DR or the LDT was rapidly removed and incubated in ice-cold artificial cerebrospinal fluid (ACSF, osmolarity = 295) that contained (in mM) 121 NaCl, 5 KCl, 1.2 NaH₂PO₄, 2.7 CaCl₂, 1.2 MgSO₄, 26 NaHCO₃, 20 dextrose, 4. 2 lactic acid, and oxygenated by bubbling with carbogen (95% O₂–5% CO₂). The brain stem was further sectioned at 250–350 μm in a coronal plane on a Leica vibrotome (VT1000S). Slices containing the DR or LDT were incubated at 35°C for 15 min and stored at room temperature. Recordings were obtained from slices submerged in a recording chamber that was perfused at 2–5 ml/min with the continuously oxygenated ACSF at room temperature (21°C). DR recordings were obtained in slices just rostral to the level of the LDT from cells in the central gray within ~50 μm of the midline. LDT recordings were obtained from the LDT cholinergic cell group, beginning rostral to the main locus coeruleus, in slices in which the dorsal tegmental nucleus was also present. Relatively large cells appearing in clusters were targeted in both nuclei, with large cells apposing the blood vessel within the LDT preferred (estimated long axis DR: 17.3 ± 0.7 μm, n = 19; LDT: 18.6 ± 1.0 μm, n = 19 from fura2 fluorescence in the slice). However, histochemical methods were used for final confirmation of neurotransmitter phenotype.

Drugs and experimental solutions

Orexin-A (Sigma; Phoenix Pharmaceuticals; American Peptides) was dissolved in water in 10 μM aliquots and frozen (−80°C). Aliquots were dissolved in ACSF to a final concentration of 300 nM or 1 μM immediately before use. Orexin was either bath applied or delivered more locally via a “Y” tube (modified from Wu et al. 2002). A “Y” tube was made by forming silicone tubing into a U and converting this construction to a Y tube by inserting a microfil tubing (ID, 100 μm; OD, 164 μm; CMF34GxxL, World Precision Instruments, Sarasota, FL) into the bend of the U with cyanoacrylic glue. During recordings, the microfil tubing was placed within 50 μm of the cell. The 15 s application time was controlled by a ALA-VM4 valve manifold (Scientific Instruments). Latency from closing of the valve (which started drug solution flow into tissue) and response of the neuron was typically <15 s. Rapid wash of orexin was obtained by switching superfusion solutions to control ACSF. Leakage of agonist was monitored and found to not be an issue. TTX (Alomone) was dissolved in ACSF to a final concentration of 500 nM to block voltage-gated sodium channels. In most of these experiments, DR or LDT neurons were recorded in an ACSF solution (DABST) containing the ionotropic receptor antagonists DNQX (15 μM, Sigma), APV (50 μM, Sigma), bicuculline (10 μM, Sigma), and strychnine (2.5 μM, Sigma), in addition to TTX (0.5 μM, Alamone). In some recordings, CsCl (2 or 3 mM) was added to the DABST to block H-current. In some experiments, the [Ca²⁺] of the ACSF was buffered to <20 μM by the addition of 2.7 mM EGTA (calculated with Patcher's Power Tools XOP for Igor Pro). In others, a low-Na⁺ ACSF was used in which NaCl in the ACSF was replaced with an equivalent concentration of N-methyl-d-glucamine Cl resulting in a solution containing 26 mM Na⁺. The L-type Ca²⁺ channel antagonist nifedipine and agonist, Bay-K-8644 (Bay-K, Sigma) were dissolved in DMSO to a stock concentration of 10 mM and delivered at the final concentration of 10 μM in ACSF. Bisindolylmaleimide I, HCl (Calbiochem, EMD Biosciences) was dissolved in DMSO to a stock concentration of 5 mM. On the day of experiments, it was diluted in TTX-ACSF to a final concentration of 1 μM and applied for 5 min before application of orexin. Final dilutions of these drugs were made immediately before application, and light exposure was minimized during all phases of their preparation and application.

Whole cell electrophysiological recording and imaging

Micropipettes (2–4 MOhms) used for patch-clamp recordings (Borosilicate, cat number 8050, AM systems) were pulled on a horizontal puller (Sutter Instruments, P87). Pipettes were filled with a recording solution of either the potassium salt of Fura-2 (bis-fura 2, 50 μM, Molecular Probes) dissolved in the recording solution (in mM) 144 K-gluconate, 3 MgCl₂, 10 HEPES, 0.3 NaGTP, and 4 Na₂ATP or, in those cases where calcium imaging was not being conducted, pipettes were filled with a recording solution containing (in mM) 144 K-gluconate, 0.2 EGTA, 3 MgCl₂, 10 HEPES, 0.3 NaGTP, and 4 Na₂ATP (osmolarity = 310). In some experiments, a Cs⁺-based patch solution containing 143 Cs⁺-gluconate, 0.2 EGTA, 5 KCl, 4.3 NaCl, 2 MgCl₂, and 10 HEPES was used. In most cases, biotinylated Alexa-594 (25 μM) was also included in the patch solution so cells could be histochemically identified. Neurons were visualized for whole cell recordings at ×160 magnification with visible-light, differential interference contrast optics, using a Nuvicon tube camera (Dage VE-1000) mounted on a fixed stage microscope (Olympus BX50WI). Cells for recording were chosen within the boundary of the DR or LDT nuclei which were identified using a ×4 objective.

Gigaseals were obtained under visual control using an Axopatch 200B amplifier (Axon Instruments) operated in voltage-clamp mode and filtered at 2 or 5 KHz with a four-pole Bessel filter at the amplifier output and sampled at 4, 10, or 20 KHz. In those cases where calcium imaging was performed, after establishing the whole cell recording configuration, cells were filled with both the calcium indicator and biotinylated Alexa-594 by either passive diffusion or brief hyperpolarizing pulses. Data were not collected until ≥20 min had passed following break-in to allow dye equilibration. Neurons were imaged through a ×40 objective using a cooled, CCD camera equipped with a back-illuminated EEV 57 frame transfer chip having an imaging area of 512 × 512 pixels (field size = 160 μm/side; MicroMax, Roper Scientific). Fura 2 was excited at 380 nm with light from a 75-W Xenon lamp that was shuttered to reduce light exposure to the tissue in between data acquisition. Recordings were either conducted in voltage clamp or “I-clamp fast” current-clamp mode, following appropriate compensation for the pipette capacitance; quality of the cells was assayed by monitoring the holding current and the input resistance as determined by the voltage or current response to a brief, negative going step. Recordings were uncorrected for liquid junction potentials, which were calculated to be ~12 mV. Current and voltage traces were digitized, and command pulses were generated by custom software (TIWB; Inoue et al. 1998) run on a Mac OS computer that controlled an ITC-18 interface (Instrutech). TIWB software also controlled the camera and shutter that allowed precise synchronization between electrophysiological and optical signals. The camera was read out through a 1-MHz, 14-bit A/D converter. Images were binned on the chip at 4 × 4 and acquired in two different modes: 1) “discontinuous,” a low-speed imaging mode used to monitor changes in calcium associated with depolarizations or inward currents in which an image was collected every 4 s or 2) “continuous,” a high-speed imaging mode in which an image was collected every 50 ms, a rate fast enough to monitor changes in calcium accompanying rapid alterations of the membrane potential. Changes in fluorescence (dF/F) were quantified by the average pixel values within regions of interest (ROIs) that were positioned on background- and baseline-subtracted fluorescent images. The background was determined from a ROI that was positioned at a location remote from the filled cell or its processes. Baseline fluorescence was determined as the fluorescence measured in the first few frames of each sequence before stimulation. bis-Fura 2 fluorescence decreases when calcium increases and is bound by the dye; however, for purposes of clarity, dF/F responses have been inverted in figures such that upward dF/F deflections indicate rises in calcium.

In those cases where calcium imaging was not being performed, membrane currents and voltages were controlled and recorded with PCLAMP8 software (Axon Instruments). Access resistance (R_acc) was estimated on-line by exponential fitting of the uncompensated capacitative transients resulting from ±2.5-mV voltage steps delivered at 3.3 Hz using the software's membrane test routine. The quality of recorded cells was assessed with this routine by simultaneously monitoring input resistance, holding current, and capacitance. These parameters were monitored continuously over the initial several minutes after establishing the whole cell configuration, as well as before and after each experimental protocol. In all cases, recordings were terminated if the estimated R_acc became unstable or changed by >20% between measurements. Recordings were also terminated if cell parameters became unstable.

Electrophysiological analysis

Current and voltage waveforms were analyzed using Igor Pro Software (Wavemetrics). Holding current was measured from 1 s averages of the holding current at –60 mV. I-V relations were obtained from voltage ramps from –100 to –30 mV or to +10 mV (Cs⁺-rich internal solution). Ramps (14 mV/s) began after holding V_m at –100 mV for 1 s. Reversal potentials were measured by a linear fit to the I-V curve from just negative to just positive to the zero current potential. In some experiments, a previously recorded Hcrt/Orx current (V_m = –60) was used as a command current while monitoring membrane potential. To examine the effect of Hcrt/Orx on the Ca²⁺ influx mediated by voltage-gated Ca²⁺ channels, Hcrt/Orx (300 nM) was applied while the voltage of the cell was periodically stepped from a holding potential of –60 to –30 or –40 mV for 5 s. To determine the contribution of L-type Ca²⁺ channels and the involvement of protein kinase C (PKC) on Hcrt/Orx-induced calcium increases, this protocol was also performed in the presence of nifedipine, a blocker, or Bay-K, an agonist, of this channel or bisindolylmaleimide I, HCl, a PKC inhibitor, respectively. Results were compared with Hcrt/Orx effects in the absence of these agents. Results are reported as means ± SE. Differences between means were compared using either a paired or unpaired two-tailed Student's t-test with α set at 0.05, as appropriate.

Hcrt/Orx depolarizes DR and LDT neurons and evokes a coordinated elevation of [Ca²⁺]_i

Application of 300 nM Hcrt/Orx produced membrane depolarization and action potentials in a majority of DR and LDT neurons (≈70%) as expected from previous reports (Brown et al. 2002; Burlet et al. 2002; Leonard et al. 2005; Liu et al. 2002). Correlated with these depolarizations, in many cells, were changes in dF/F measured at the soma and proximal dendrites, indicating a rise in intracellular calcium (Fig. 1, A and B). These findings confirm our prior observation in fura-2AM–loaded neurons that Hcrt/Orx elicits Ca²⁺ transients in both the soma and dendrites (Kohlmeier et al. 2004). Changes in dF/F were never elicited in the absence of depolarization; however, Hcrt/Orx could produce depolarization without detectable change in dF/F in cells from both nuclei—a point we examined in greater detail later. In the DR, the average depolarization was 3.8 ± 1.2 mV and the maximal dF/F increase at the soma was 5.4 ± 1.3% (n = 5). In the LDT, the average depolarization was 3.5 ± 0.8 mV and the maximal dF/F increase at the soma was 5.2 ± 1.8% (n = 10). To determine whether Hcrt/Orx induced these changes in the principal cells of the DR and LDT, we immunolabeled for TpH or bNOS, respectively, in slices containing recorded Alexa-594–injected neurons. Immunohistochemistry from a population of recovered cells showed that Hcrt/Orx-induced effects were present in serotonergic DR cells and cholinergic LDT neurons (Fig. 1C).

Open in a separate window

FIG. 1.

Hypocretin/orexin (Hcrt/Orx) depolarizes dorsal raphe (DR) and laterodorsal tegmental (LDT) neurons causing action potentials and rises in somatic and proximal dendritic [Ca²⁺]_i. A and B: (top right) images of bis-fura-2 fluorescence in a DR (A) and LDT (B) neuron during whole cell recordings from which changes in fluorescence and membrane potential were measured (left column). Boxes indicate regions of interest (ROIs) from which dF/F signals were computed. Scale equals 10 μm. In left panels, bottom trace indicates membrane potential, middle trace indicates somatic dF/F, and top trace indicates dendritic dF/F. Orexin-A (300 nM) induced a depolarization, firing, and rises in [Ca²⁺]_i in these DR and LDT neurons. Boxed insets are high-gain traces showing that the Ca²⁺ signals preceded the induction of spiking (arrows). Scales in insets equal 10% dF/F, 10 mV, and 20 s. C: immunocytochemistry verified that these actions occurred in serotonergic and cholinergic neurons of the DR (C1) and LDT (C2). Left image was taken at low power (×4 objective) and shows TpH-positive cells in the Raphe (C1) and bNOS-positive cells in the LDT (C2) labeled with FITC. Top image in right columns show a higher-power FITC image (×20 objective) of the field containing the recorded neuron. Bottom image in right columns shows the recorded neuron (arrow) in the DR (C1) and the LDT (C2) visualized with alexa-594, which was included in the pipette solution. These neurons showed an orexin-A (300 nM)–mediated depolarization and change in [Ca²⁺]_I and were immunopositive for TpH (C1) and hence serotonergic and bNOS (C2) and hence cholinergic. Scale equals 100 (×4) and 20 μm (×20). Image brightness and contrast were adjusted uniformly for each image to facilitate viewing fluorescent labels. Aq, aqueduct; MnR, median raphe; DTg, dorsal tegmental nucleus.

Although the depolarizations produced by Hcrt/Orx often resulted in spiking, it was clear from inspection that Ca²⁺ transients began before the onset of spiking and became larger after spiking onset (Fig. 1, insets). This suggests that part of the Ca²⁺ influx evoked by Hcrt/Orx arises from the subthreshold depolarization and part arises from spike-evoked calcium transients. Indeed, in our previous study of fura2-AM–loaded DR and LDT neurons, Hcrt/Orx produced calcium transients having different temporal signatures described as “plateau,” “spiker,” and combined “smooth/spiker” responses (Kohlmeier et al. 2004). The responses shown in Fig. 1 are reminiscent of the smooth/spiker responses described previously and suggest that “plateau” responses might be mediated by subthreshold depolarizations, whereas “spiker” responses might be mediated by action potentials. The Ca²⁺ influx during the subthreshold depolarization might be mediated by activation of voltage-gated Ca²⁺ channels during the depolarization and/or by Ca²⁺ influx though any Ca²⁺-permeable cation channels contributing to the depolarization.

Hcrt/Orx depolarizes DR and LDT neurons by activating a noisy inward current that is Na⁺ dependent

Previous studies have implicated a cation current in mediating the Hcrt/Orx-evoked depolarization of both DR and LDT neurons, although no I-V data have been published for the orexin currents in LDT neurons. To extend these observations and compare the currents under identical recording conditions, we examined the Hcrt/Orx current from the somata of 45 DR neurons and 58 LDT neurons using whole cell voltage-clamp methods. DR neurons recorded with a K⁺-rich internal solution had an average input resistance of 511 ± 31 MΩ (range: 265–990 MΩ, n = 33) and LDT neurons recorded with a K⁺-rich internal solution had an average input resistance of 310 ± 25 MΩ (range: 106–771 MΩ, n = 36). Orexin-A (300 nM; 1.5-min superfusion; holding potential = –60 mV) produced a long-lasting inward current (–35.3 ± 5.9 pA, n = 5, DR and –22.7 ± 2.7 pA, n = 11, LDT; Fig. 2, A and H) that was accompanied by an increase in membrane current noise (Fig. 2, A and H) in neurons recorded with ACSF containing DABST to block action potentials and fast synaptic potentials and 2 mM Cs⁺ to block I_h. In this sample of neurons, the average membrane noise at –60 mV increased over control by 108.5 ± 15.9% (n = 5) in the DR and 55.6 ± 9.0% (n = 11) in the LDT. These effects were long-lasting and were typically only partly reversed by the end of the recording period (10–20 min). As previously reported, for both DR and LDT neurons, these effects resulted from a direct action on the recorded neurons because neither the inward current nor the membrane current noise increase were attenuated in low Ca²⁺ ACSF, which was shown previously to block evoked excitatory postsynaptic currents (EPSCs) in the LDT (Burlet et al. 2002). In DR neurons, the orexin-A current (at –60 mV) was larger, on average (−65.0 ± 9.6 pA, n = 7, P < 0.05) in low Ca²⁺ ACSF, whereas it was not significantly larger in LDT neurons (−27.9 ± 3.7 pA, n = 4, P > 0.05).

Open in a separate window

FIG. 2.

Orexin activates an Na⁺-dependent cation current that is similar in DR (A–G) and LDT neurons (H–L). A: bath application of orexin-A (300 nM) produces a long-lasting inward current (bottom trace) and increase in current noise (top trace) in DR neurons (holding potential = –60 mV). B: following activation of this current by orexin-A, superfusion with low-Na⁺ artificial cerebrospinal fluid (ACSF) rapidly attenuated the current and the noise increase, consistent with a large reduction in concentration of a permeant ion. Return to ACSF containing normal Na⁺ reinstated a residual noisy inward current. C: application of low Na⁺ ACSF also attenuated the baseline holding current and noise (−60 mV), showing that the basal inward current at −60 mV is also Na⁺ dependent. D: application of orexin-A (300 nM) under low Na⁺ conditions produced only a small inward current in DR neurons. E: Voltage ramps between –100 and –30 mV obtained before and after orexin application showed an inward current that was larger and noisier at –100 than at –30 mV. F: I-V relation of the orexin-evoked current (bottom) appeared quite linear over this range as indicated by the best fit line. The noise also decreased linearly between –100 and –30 mV (top). G: the orexin current appeared similar in recordings obtained with Cs⁺-rich intracellular solutions and reversed near 0 mV, suggesting the orexin-activated channels were permeable to both Cs⁺ and K⁺. H: in LDT neurons, bath application of orexin-A (300 nM) also produced a long-lasting inward current (bottom trace) and an increase in membrane current noise (top trace). I: both were rapidly attenuated by low-Na⁺ ACSF. J: voltage ramps (–100 to –30 mV) often showed a noisy current similar to that seen in DR neurons. K: in these LDT neurons, the I-V relation of the orexin-evoked current was linear over this range as indicated by the best-fit line (bottom). The membrane current noise also decreased between –100 and –30 mV (top). L: this orexin current appeared similar in LDT neurons recorded with Cs⁺-rich intracellular solutions and reversed near 0 mV, suggesting the underlying channels are permeable to Cs⁺ and K⁺ as in DR neurons. For experiments A–F and H–K, the internal solution contained K⁺ gluconate and the ACSF contained DABST and 2 mM Cs⁺. For experiments G and L, the internal solution contained Cs⁺ gluconate and the ACSF contained low Ca²⁺, DABST, and 2 mM Cs⁺.

The ability of Hcrt/Orx to evoke current and membrane noise changes in ACSF containing 2 mM Cs⁺ also indicated that the current was not mediated by Cs⁺-sensitive K⁺ channels or changes in I_h, which was effectively blocked by this treatment. To test for involvement of other Na⁺-dependent ionic currents, we superfused low-Na⁺ ACSF just after activating the Hcrt/Orx current. This reversibly attenuated the inward current and membrane noise evoked by orexin-A in both DR (Fig. 2B; n = 7) and LDT neurons (Fig. 2I; n = 12). Because lowering extracellular Na⁺ also typically reduced the baseline holding current and noise (Fig. 2C), we conducted additional experiments in which we switched to low-Na⁺ DABST ACSF before applying orexin-A. Under these conditions, the Hcrt/Orx-current and noise were nearly abolished in DR neurons (−8.9 ± 2.3 pA; 6.37 ± 8.5%; n = 3; Fig. 2D) and were undetectable in LDT neurons (change in holding current <6 pA; n = 3). Thus in both our DR and LDT recordings, the Hcrt/Orx current was dependent on extracellular Na⁺.

In some neurons, the Hcrt/Orx current seems to be sensitive to KB-R7943, an antagonist of the electrogenic Na⁺/Ca²⁺ exchanger (Burdakov et al. 2003; Eriksson et al. 2001; Wu et al. 2002, 2004). Because we previously found that the calcium influx evoked by Hcrt/Orx in both DR and LDT neurons was insensitive to KB-R7943 (Kohlmeier et al. 2004), a role for the electrogenic Na⁺/Ca²⁺ exchanger seemed unlikely. Nevertheless, we verified that KB-R7943 did not attenuate the current activated by orexin-A in DR neurons (data not shown). As expected, a low concentration of KB-R7943 (10 μM) had no effect on the inward current evoked by orexin-A (95.8 ± 21.9% of control, n = 2), whereas a higher concentration (80 μM), which is reported to have nonspecific effects (Iwamoto et al. 1996), trended toward only a modest attenuation (76.5 ± 6.9% of control, n = 2). These findings are in concordance with prior findings (Kohlmeier et al. 2004; Liu et al. 2002) and indicate that the Na⁺/Ca²⁺ exchanger does not make a significant contribution to the Hcrt/Orx-mediated depolarization and calcium influx in DR and LDT neurons.

Noisy inward current in DR and LDT neurons is a nonselective cation current

Using voltage-clamp ramps from –100 to –30 mV, we examined the current-voltage relationship of the orexin-A–induced current in DR (Fig. 2E) and LDT (Fig. 2J) neurons using a K⁺-rich pipette solution. These data showed that Hcrt/Orx produced an inward shift in the membrane current and an increase in noise that were both larger at –100 than –30 mV. Subtracting these curves and plotting them against V_m shows the I-V relation of the Hcrt/Orx-evoked current (I_orx), which was approximately linear over this range of voltages in DR (Fig. 2F) and LDT (Fig. 2K) neurons. By plotting the rms current versus V_m (computed every 5 mV), it was also evident that the current noise was largest at negative potentials and decreased as membrane potential moved more positively, as expected for channel noise associated with I_orx. We estimated a slope conductance by fitting a line to the I-V relation between –100 and –40 mV, yielding an average value of 1.65 ± 0.42 nS in DR neurons (n = 6) and 0.50 ± 0.10 nS in LDT neurons (n = 11). These data suggest that the current may reverse near 0 mV and are consistent with Hcrt/Orx activating a nonselective cation current in both DR and LDT neurons.

To further test this idea, we made recordings using a cesium-based internal solution and low calcium ACSF containing DABST to improve the space clamp by blocking K⁺ currents and to limit current from voltage-gated Ca²⁺ currents, which produced Ca²⁺ spikes otherwise. Under these conditions, orexin-A also induced an inward current at –60 mV in both the DR (−74.0 ± 22.6 pA, n = 6) and LDT (−21.5 ± 5.0 pA, n = 8) that was accompanied by increased membrane current noise (DR: 223.13 ± 51.71%; LDT: 62.0 ± 20.7%).

This current was studied with slow voltage ramps between –100 and +10 mV, and its I-V relation was constructed from the difference between ramp currents measured at the peak of the Hcrt/Orx effect and from before Hcrt/Orx application. The resulting I-V curves were approximately linear below –40 mV, as observed with the K⁺-rich internal solution, with a slope conductance of 0.97 ± 0.25 nS for DR (Fig. 2G) and 0.25 ± 0.09 nS for LDT (Fig. 2L) neurons. These I-V curves flattened at more positive potentials and reversed near 0 mV for both DR (–10.2 ± 1.8 mV, n = 4) and LDT (–11.4 ± 2.5mV, n = 8) neurons. Moreover, the membrane current noise behaved similarly to that observed with the K⁺-rich intracellular solution, was largest at –100 mV, and decreased as the reversal potential was approached. These data suggest that channels mediating I_orx have similar permeabilities to Cs⁺ and K⁺. To better evaluate this point, we recorded from four DR neurons with the K⁺-rich internal solution and measured I_orx using voltage ramps to +10 mV. Under these conditions, the measured reversal potential was –3.6 ± 3.4 mV (n = 4), which was not different from that measured using the Cs⁺-rich internal solution (P > 0.05), consistent with equal Cs⁺ and K⁺ permeabilities and indicating that I_orx is mediated by nonselective cation channels.

In addition to Hcrt/Orx activating cation channels in LDT neurons (n = 11/16 responders recorded with the K⁺ internal solution, 69%), a subpopulation had orexin currents with I-V curves that were flat or inwardly sloping between –100 and –40 mV in recordings with a K⁺-rich internal solution and ACSF containing DABST (n = 5/16 responders, 31%). For these cells, the peak current evoked by 300 nM orexin-A at –60 mV was −10.0 ± 2.3pA (n = 5), and the membrane current noise only increased by 9.7 ± 8.4%. The Hcrt/Orx-evoked current and membrane noise in these cells were significantly smaller than those observed in LDT neurons with cationic currents (P < 0.01).

Collectively, these data indicate that Hcrt/Orx-induced depolarizations are mediated by the activation of a noisy NSCC in DR neurons and a majority of LDT neurons. A minority of LDT neurons responded to Hcrt/Orx with smaller inward currents that, based on the I–V curves and current noise, were mediated by one or more additional mechanisms that we did not further study.

Hcrt/Orx-evoked elevation of [Ca²⁺]_i requires depolarization and is not mediated by Ca²⁺ influx through Hcrt/Orx activated nonspecific cation channels

Given the potential space-clamp issues associated with clamping neurons with extended processes and the fact that Ca²⁺-spiking could occur at positive membrane potentials during voltage ramps in normal extracellular calcium, we did not attempt to estimate the Ca²⁺ permeability of the NSCCs by measuring the reversal potential shifts produced by changing extracellular Ca²⁺ concentration. Instead, we measured somatic intracellular Ca²⁺ transients while manipulating membrane potential. We first examined the Ca²⁺ transients produced by orexin-A (300 nM) depolarizations in neurons recorded in current clamp following blockade of Na⁺ spikes and ionotropic synaptic receptors. Under these conditions, the average change in dF/F was 5.1 ± 1.4% in the DR (n = 8) and 4.5 ± 1.8% in the LDT (n = 5). Interestingly, Hcrt/Orx sometimes induced small amplitude spikes in both DR (n = 3) and LDT (n = 4) neurons that were similar to the calcium spikes evoked by somatic current injection in the presence of TTX (Fig. 3, A and inset). However, changes in dF/F were also evident when these presumed Ca²⁺ spikes were not elicited (DR: n = 5; LDT: n = 1, Fig. 3B). Conversely, in some neurons, Hcrt/Orx elicited changes in membrane potential without corresponding changes in dF/F (Fig. 3C). Indeed, we found that the magnitude of the elicited Ca²⁺ transient was correlated with the membrane potential before orexin-A application. On average, orexin-A produced significantly smaller Ca²⁺ transients in neurons with membrane potentials negative to –46 mV (n = 16) than in neurons with membrane potentials positive to –45 mV (n = 22; Fig. 3, histogram). These data suggest that the Hcrt/Orx depolarization, per se, might be responsible for triggering the majority of the Ca²⁺ influx.

Open in a separate window

FIG. 3.

Orexin-A causes rises in [Ca²⁺]_i and Ca²⁺ spiking in some DR and LDT neurons, if they are sufficiently depolarized. A: following application of TTX and the blockade of ionotropic glutamate, GABA, and glycine receptors, orexin-A often elicited Ca²⁺ spikes in DR (left traces) and LDT (middle traces) neurons. Right traces show the Ca²⁺ spikes obtained by depolarization in the DABST-containing ACSF and the spikes obtained in normal ACSF. Note the difference in threshold and amplitude. B: changes in [Ca²⁺]_i were also elicited by orexin-A in some cells in which Ca²⁺ spikes were not induced. C: in some cells, the orexin-A–induced depolarization was not accompanied by changes in dF/F. Note: the small change in dF/F during the application is artifact, often seen with the “Y-tube” method of drug delivery, and was easily distinguished from specific peptide effects. The bar graph shows the maximal percent change in dF/F produced by orexin-A (300 nM) in cells that were held close to −45 mV and those that were held close to −55 mV. The holding potential influenced the change in [Ca²⁺]_i induced by the Hcrt/Orx depolarization. D: depolarization elicited by a simulated Hcrt/Orx current induced rises in [Ca²⁺]_i. The current command was derived from a previously recorded inward current induced by orexin-A. This current command produced a slow depolarization that was sufficient to induce an increase in dF/F in the absence of Hcrt/Orx in this DR cell at a baseline potential of –43 mV (D1, left). The same current command elicited a negligible change in dF/F when injected from a baseline potential of –60 mV in the same cell (D1, right). Similarly, at a given baseline potential, increasing the injected current, increased the resulting dF/F (D2; LDT neuron).

To test the plausibility of this idea, we simulated an orexin-A depolarization using a previously recorded Hcrt/Orx current for the current command. When the simulated Hcrt/Orx current was injected from a membrane potential of –43 mV, it elicited smooth Ca²⁺ transients similar to those seen in our previous fura2-AM study (Kohlmeier et al. 2004; Fig. 3D1, left), but when the same current was injected from a membrane potential of –60 mV, little change in dF/F was apparent (LDT; Fig. 3D1, right). Similarly, when the amplitude of the simulated current was increased to produce a greater depolarization, the change in dF/F also increased (Fig. 3D2). Thus simulated Hcrt/Orx depolarizations mimic the Ca²⁺ transients produced by authentic Hcrt/Orx. These data suggest that depolarization is necessary for Hcrt/Orx to evoke a Ca²⁺ transient in these neurons.

To directly test this possibility, we applied orexin-A (300 nM) and monitored dF/F while voltage clamping the soma to –60 mV. Under these conditions, Hcrt/Orx produced typical inward currents but without detectable changes in dF/F. In our sample of DR neurons, orexin-A (300 nM) produced an inward current of 12.3 ± 2.2 pA (n = 6/6 cells tested) but only an average change in dF/F of 0.75 ± 0.28% (n = 6/6; Fig. 4A, left). In the LDT, orexin-A (300 nM) produced an inward current of 15.4 ± 0.63 pA (n = 6/6 cells tested) but only an average change in dF/F of 0.4 ± 0.33% (Fig, 4A, right). These changes in dF/F were not significantly different from baseline (P > 0.05). Thus depolarization is necessary for Hcrt/Orx to stimulate Ca²⁺ transients in DR and LDT neurons.

Open in a separate window

FIG. 4.

The nonselective cation channel (NSCC) activated by Hcrt/Orx is not a Ca²⁺ influx pathway that contributes to the Ca²⁺ transients elicited by Hcrt/Orx. A: voltage-clamp recordings (holding = −60 mV) and corresponding changes in dF/F following Y-tube application of ACSF (1) and orexin-A in ACSF (2) to a DR (left traces) and LDT (right traces) neuron. Orexin-A induced an inward current but no rise in somatic or dendritic [Ca²⁺]_i when DR and LDT neurons were voltage clamped near rest (−60 mV). B: voltage jumps to –100 mV increased the driving force for Ca²⁺ and augmented the dF/F under baseline conditions, indicating a resting Ca²⁺ permeability. However, this increase in dF/F was not further augmented during activation of the NSCC by orexin-A. B1: membrane current (holding = –60 mV) recorded from a DR neurons while orexin-A (300 nM) was applied. The trace was “blanked” during voltage jumps before (B1a) and after (B1b) orexin-A application. B2a: dF/F (top trace) increased during voltage steps from –60 to –100 mV (bottom traces), indicating a resting Ca²⁺ permeability (3 sweeps superimposed; middle trace is membrane current). B2b: This slight increase in dF/F was not different when tested during the inward current elicited by orexin-A application (3 traces superimposed). B3: bar graph summarizes these changes in dF/F elicited by voltage jumps to −100 mV from before and after orexin-A application. These changes were not different (P > 0.05, n = 5), indicating that any Ca²⁺ influx that might occur via the NSCC does not contribute directly to the Hcrt/Orx-mediated Ca²⁺ transients.

These results also indicate that any Ca²⁺ influx through Hcrt/Orx-activated nonselective cation channels is too small to produce detectable changes in somatic [Ca²⁺]_i at –60 mV. Because many nonselective cation channels are Ca²⁺ permeable, we tested whether detectable changes in dF/F could be provoked by increasing the driving force for Ca²⁺. We therefore monitored dF/F with high-speed imaging (20 frames/s) while stepping the membrane potential from –60 to –100 mV for 10 s before and after application of orexin-A (300 nM). Control voltage jumps to –100 mV resulted in a slight increase of dF/F, consistent with an increased Ca²⁺ influx (Fig. 4B2a) carried by the baseline current shown in Fig. 2C. However, application of orexin-A, which evoked a noisy inward current (Fig. 4B1), did not further increase the dF/F produced by the voltage jump (Fig. 4B2b). Results from these experiments are summarized in Fig. 4B3 (n = 5, P > 0.05). Taken together, these data indicate 1) that the baseline somatic conductance at −60 mV in DR and LDT neurons is calcium-permeable and 2) that any Ca²⁺ influx through the Hcrt/Orx-activated cation channels is too small to contribute to Hcrt/Orx-mediated rises in [Ca²⁺]_i, even at negative membrane potentials when the cation current is increased. Thus somatic depolarization is necessary for Hcrt/Orx-mediated somatic rises in [Ca²⁺]_i.

Hcrt/Orx enhances Ca²⁺ transients mediated by L-type calcium channels

Presumably, this depolarization is necessary to trigger voltage-operated Ca²⁺ channels, like the L-type channels that we have already shown contribute to the Hcrt/Orx-mediated Ca²⁺ transients in DR and LDT neurons (Kohlmeier et al. 2004). However, is depolarization alone sufficient to account for the Hcrt/Orx-mediated Ca²⁺ transients or does Hcrt/Orx also enhance the Ca²⁺ transients evoked by a given depolarization? To address this question, we monitored the changes in dF/F resulting from voltage clamp steps from —60 to –30 mV in DR and LDT neurons before and after orexin-A (300 nM) application. We found that the Ca²⁺ transients produced by these voltage jumps were significantly enhanced by ~30% following Hcrt/Orx application (Fig. 5A). Interestingly, on average, this enhancement was the same regardless of whether orexin-A also activated a slow inward current in the recorded neuron (P > 0.05). In DR neurons, the Ca²⁺ transient was enhanced by 34.4 ± 3.1% in cells with an inward Hcrt/Orx current (n = 3/14 cells tested) and by 26.0 ± 4.84% in cells without inward current (n = 6/14 tested). In LDT neurons, the Ca²⁺ transient was enhanced by 33.3 ± 12.4% in cells with inward current (n = 4/13 tested) and by 30.0 ± 5.7% in cells without inward current (n = 5/13 tested). In five DR cells and four LDT neurons, step-evoked calcium transients were not enhanced regardless of whether orexin-A elicited an inward current (inward current evoked in 3/5 DR cells and 1/4 LDT cells). These data indicate that, in a large fraction of DR and LDT neurons, Hcrt/Orx enhances Ca²⁺ transients independently of the inward current that produces direct excitation. Although this enhancement is most likely caused by increased Ca²⁺ influx through voltage-gated Ca²⁺ channels, we saw only small and variable changes in the total whole cell current, which was monitored simultaneously during the voltage steps (Fig. 5A, middle trace). This probably reflects the fact that the Ca²⁺ current makes up only a small fraction of the total current that is dominated by outward currents, some of which are also Ca²⁺ dependent. Indeed, on average, this current was 11.1 ± 1.1 pA more outward (P < 0.05; n = 8) at –30 mV for the same sweeps in which the Ca²⁺ transients were enhanced by ~30%.

Open in a separate window

FIG. 5.

Hcrt/Orx enhances the Ca²⁺ transients produced by activation of L-type Ca²⁺ channels in DR and LDT neurons. A: Ca²⁺ transients elicited by voltage steps from –60 to –30 mV were reversibly enhanced by ~30% following application of orexin-A (300 nM) in DR and LDT neurons, even in the absence of an orexin-A–induced inward current. Top traces show the Ca²⁺ transients measured as dF/F superimposed from before orexin-A application (Con), after orexin-A application (Orx), and after recovery from the orexin-A application (Rec). The bottom and middle traces show the corresponding membrane voltage (bottom) and membrane current traces (middle). B: nifedipine (10 μM) nearly completely attenuated the Ca²⁺ transient enhancement produced by Hcrt/Orx. Top traces show the Ca²⁺ transients elicited by voltage steps from –60 to –30 mV in the presence of nifedipine (Nif) after orexin-A application (Orx/Nif) and after the expected orexin effect would have recovered (Rec Orx/Nif). C: bar graph summarizes the effect of nifedipine (10 μM) on the Ca²⁺ transient enhancement produced by Hcrt/Orx. Nifedipine significantly attenuated this effect in both the DR and LDT (*P < 0.05). D: application of Bay-K 8644 (10 μm; Bay-K), which enhances current through L-type Ca²⁺ channels, increased the control Ca²⁺ transients elicited by voltage steps from –60 to –30 mV (Con). Application of orexin-A (300 nM) further enhanced these Ca²⁺ transients (Orx/Bay-K), which recovered after wash-out of orexin-A (Rec Orx/Bay-K). E: bar graph comparing the enhancement of the voltage step–evoked Ca²⁺ transient produced by orexin-A in normal ACSF (Orx ACSF) and in ACSF containing Bay K (Orx + Bay). The Ca²⁺ transient enhancement was significantly larger in the presence of Bay-K than it was in the absence of Bay-K for both DR and LDT neurons (*P < 0.05).

We next examined the role of L-type Ca²⁺ channels in these Hcrt/Orx-enhanced Ca²⁺ transients. We first verified that nifedipine, a blocker of L-type calcium channels, antagonized part of the Ca²⁺ transient produced by our voltage-clamp steps from –60 to –30mV. Nifedipine (10 μm) significantly attenuated the step-evoked change in dF/F by 38.3 ± 2.3% in the DR (n = 5) and by 24.6 ± 1.8% in the LDT (n = 4; P < 0.05), indicating substantial L-channel–mediated Ca²⁺ influx was evoked with this protocol. We determined whether nifedipine attenuated the Ca²⁺ transient enhancement produced by Hcrt/Orx. Because we determined in pilot experiments that the enhancing effect of Hcrt/Orx declined with each Hcrt/Orx application, we were unable to compare the effect of nifedipine on this enhancement in the same cells. We therefore compared the Ca²⁺ transient enhancement between two populations of neurons: one with, and one without, nifedipine present. In the presence of nifedipine (10 μm), orexin-A (300 nM) failed to significantly enhance the step-evoked change in dF/F in cells from either nucleus (6.0 ± 4.8% increase in DR neurons, n = 6; 1.3 ± 3.3% increase in LDT neurons, n = 9; P > 0.05; Fig. 5B). These increases were significantly smaller than those obtained from a control population of DR and LDT neurons and are compared in Fig. 5C. These data suggest that most, if not all, of the increase in the step-evoked Ca²⁺ transient produce by Hcrt/Orx was mediated by increased Ca²⁺ influx via L-type channels.

To further test the role of L-channels, we examined the effect of Bay-K, an L-type Ca²⁺ channel agonist. Application of Bay-K (10 μm) alone significantly enhanced the step-evoked Ca²⁺ transient by 87.3 ± 40.3% in the DR (n = 5) and by 126.9 ± 46.0% in the LDT (n = 5). Application of orexin-A (300 nM) in the presence of Bay-K produced a further enhancement of the step-evoked Ca²⁺ transient in both nuclei (DR: 159.4 ± 42.6%, n = 5; LDT: 303.5 ± 96%, n = 5; compared with the baseline transient before Bay-K and orexin-A), suggesting that the Hcrt/Orx enhancement was augmented by potentiating L-channel activity with Bay-K.

Indeed, the enhancement in the step-evoked Ca²⁺ transient produced by Hcrt/Orx in the presence of Bay-K was significantly greater than that produced in the control population without Bay-K (Fig. 5E). In the DR, orexin-A application increased the step-evoked Ca²⁺ transient measured in Bay-K by 45.7 ± 10.8% (n = 5) compared with a 28.6 ± 3.5% increase measured in the absence of Bay-K (n = 9; P < 0.05). In the LDT, a similar relation was seen, with orexin-A increasing the step-evoked Ca²⁺ transient measured in Bay-K by 71.7 ± 21.5% (n = 5) compared with the 31.6 ± 5.9% increase in the absence of Bay-K (n = 9; P < 0.05; Fig. 5E). This is the outcome expected if Hcrt/Orx is specifically enhancing influx through L-type Ca²⁺ channels, because a larger fraction of the total Ca²⁺ transient is mediated by L-channels following Bay-K action. Taken together, these data strongly indicate that Hcrt/Orx enhances the depolarization-evoked Ca²⁺ transients mediated by L-type Ca²⁺ channels in DR and LDT neurons.

PKC mediates the Hcrt/Orx enhancement of Ca²⁺ transients

Previously, we found that the PKC inhibitor Bis I attenuated, but did not abolish, the orexin-induced calcium transients in Fura AM–loaded DR and LDT neurons. Because our findings here indicate that these Ca²⁺ transients are produced by both depolarization and enhanced L-type Ca²⁺ channel transients, we determined whether Bis I antagonized one or both of these actions. To do so, we examined the effect of applying Hcrt/Orx on voltage step-evoked Ca²⁺ transients following pretreatment of the slices with Bis I (1 μM). We found that Bis I application alone sometimes resulted in a slight, slow outward drift in the holding current at −60 mV and a progressive increase in the outward current produced by the voltage step (–60 to −30 mV). Despite these subtle changes in whole cell current, Bis I did not have an effect on the Ca²⁺ transients evoked by voltage steps. The average dF/F produced by these steps in the DR (6.7 ± 0.8%, n = 7) and LDT (5.5 ± 9.2%, n = 11) after Bis I application was not different from the values before Bis I application (DR: 6.4 ± 1.9%; LDT: 5.8 ± 0 0.9%; P > 0.05) for the same cells. Moreover, in the presence of Bis I, Hcrt/Orx failed to increase the step-evoked Ca²⁺ transients above control levels (DR: 5.7 ± 07%, n = 7/7; LDT: 5.6 ± 0.9%, n = 11/11; P > 0.05, Fig. 6A). Because Hcrt/Orx enhanced these Ca²⁺ transients in ~70% of recorded cells in the absence of Bis I, these data indicate that activation of PKC is required for the enhancement of L-type Ca²⁺ channel function by Hcrt/Orx in DR and LDT neurons.

Open in a separate window

FIG. 6.

Inhibition of protein kinase C (PKC) with bisindolylmaleimide I abolishes enhancement of voltage-step evoked Ca²⁺ transients by Hcrt/Orx but does not completely block the Hcrt/Orx-evoked inward currents. A: Ca²⁺ transients elicited in DR and LDT neurons by voltage-steps from –60 to –30 mV following preincubation with bisindolylmaleimide I (Bis I; top trace labeled bis) were not enhanced by orexin-A (300 nM; top trace labeled Orx/bis). Corresponding voltage (bottom) and total membrane current (middle) traces are also shown. B: following preincubation with Bis I, orexin-A (300 nM) retained the ability to evoke an inward current in some neurons even though these neurons did not show an enhancement in Ca²⁺ transients evoked by voltage steps. Trace in B is taken from 1 such neuron recorded in the DR. Note that the vertical deflections are currents associated with test voltage steps and are truncated. C: left bar graph: effect of Hcrt/Orx on Ca²⁺ transients following preincubation with Bis I. The average Ca²⁺ transient amplitude was not different following orexin-A application. Right graph: magnitude of the current evoked by orexin-A following preincubation of the slice with Bis I in the same neurons that were imaged. There was no detectable current in 5/7 DR neurons and 9/11 LDT neurons. The average (±SE) Hcrt/Orx current for these neurons is plotted as solid symbols for DR and LDT neurons (I_Orx). Two neurons from each nucleus, however, retained an I_Orx of apparently normal magnitude. The amplitudes of these currents are plotted as open symbols for each DR and LDT neuron.

We also found that, although most of these neurons recorded in the presence of Bis I did not show an inward current in response to Hcrt/Orx, two cells in each nucleus displayed normal Hcrt/Orx-mediated inward currents (DR: n = 2/7; LDT: n = 2/11; Fig. 6, B and C). The presence of these currents is consistent with a PKC-independent mechanism, as previously suggested for DR currents (Brown et al. 2002) and DR and LDT Ca²⁺ transients (Kohlmeier et al. 2004).

Hcrt/Orx enhances the Ca²⁺ transients in identified serotonergic DR neurons and cholinergic LDT neurons

To determine whether Hcrt/Orx enhances the depolarization-evoked Ca²⁺ transients in the serotonergic and cholinergic neurons of these nuclei, we immunolabeled for TpH or bNOS in slices containing Alexa-594–positive neurons. In a small sample of recovered neurons, we verified that Hcrt/Orx enhanced voltage step–evoked Ca²⁺ transients in both serotonergic DR (n = 3) and cholinergic LDT neurons (n = 4). This also occurred in cells that were not TpH+ (n = 2) or bNOS+ (n = 3). We confirmed that the Hcrt/Orx-mediated enhancement was augmented after Bay-K application in both serotonergic (n = 3) and cholinergic neurons (n = 2). Thus Hcrt/Orx enhances Ca²⁺ transients from L-channels in both DR and LDT principal neurons.

Mechanisms mediating Hcrt/Orx-induced depolarization in DR and LDT neurons

The Hcrt/Orx-mediated depolarization in rat DR neurons was initially ascribed to blockade of a leak potassium current (Brown et al. 2001), but in a later publication (Brown et al. 2002), the same group suggested a more important involvement of a NSCC. The findings of Liu et al. (2002) in rat, our preliminary findings (Leonard et al. 2005; Tyler et al. 2001), and our findings here, in mouse, support this view. Our data also confirmed that this current substantially increases membrane noise, is very attenuated at –60 mV in a low-Na⁺ extracellular solution, and is insensitive to low concentrations of an antagonist of the electrogenic Na⁺/Ca²⁺ exchanger (Brown et al. 2002; Liu et al. 2002). Our new data show that the evoked membrane current noise co-varies with current magnitude between –100 and +10 mV as expected, if it results from channels opened by Hcrt/Orx. We also found that the current is not blocked by extracellular Cs⁺ at concentrations that block H-current and some K⁺ channels. Moreover, the Hcrt/Orx current was similar in neurons recorded with a Cs⁺-rich intracellular patch solution and reversed near 0 mV with both internal solutions.

In LDT neurons, the current activated by Hcrt/Orx was very similar. For example, membrane noise co-varied with current magnitude between –100 and +10 mV, the current was very attenuated at –60 mV in low Na⁺ ACSF, it was insensitive to extracellular Cs⁺, it reversed near 0 mV, and it was of a similar magnitude in neurons recorded with intracellular K⁺ or Cs⁺. Collectively, these data suggest that Hcrt/Orx activates cation channels that are permeable to Na⁺, K⁺, and Cs⁺, consistent with an NSCC in LDT and DR neurons. It is noteworthy that, on average, the Hcrt/Orx slope conductance measured with the K⁺ internal solution was ~3 times larger in DR neurons than in LDT neurons. This may indicate that DR neurons are more powerfully recruited to fire than are LDT neurons during periods of high Hcrt/Orx release.

It is also of interest to note that, in DR and LDT neurons recorded with the Cs⁺-based internal solution and low extracellular Ca²⁺, the I-V curve of the Hcrt/Orx current showed a decrease in slope conductance positive to –40 mV and outward rectification again as the membrane potential passed through the reversal potential. The tendency for the I-V to flatten at potentials positive to –40 mV was also observed in neurons recorded with a K⁺-rich intracellular solution and suggests that the Hcrt/Orx-activated cation current is slightly voltage dependent in DR and LDT neurons. This similarity further suggests that Hcrt/Orx activates a similar class of cation channels in both types of neurons. Several issues concerning this current remain unresolved. For example, it is not clear what makes this current so noisy and we did not determine the relative permeabilites to Na⁺ and K⁺ because we could only make slow changes in extracellular Na⁺ and did not use voltage ramps over a sufficiently large range of potentials to delineate the I-V and reversal potential shifts. These points should be studied in future experiments using isolated neurons, where rapid solution changes and better space clamp can be achieved.

Hcrt/Orx-activated cation currents have also been reported in area postrema neurons (Yang and Ferguson 2002), nucleus tractus solitarius (Yang and Ferguson 2003), paraventricular hypothalamic neurons (Yang and Ferguson 2002), and locus coruleus (LC) neurons (Murai and Akaike 2005). However, the current reported in those neurons seems different in several respects. First, there was no evidence of increased membrane current noise at negative membrane potentials in the published accounts, and second, the I–V curves where essentially linear and reversed without rectification near –40 mV. Cs⁺ permeability was not tested, but based on the lack of current noise and the nearly linear I-V relation, those currents are likely mediated by different NSCC channels. Thus it is likely that Hcrt/Orx activates more than one type of NSCC in the CNS.

Mechanisms mediating Hcrt/Orx-induced calcium transients

Based on the findings that most NSCCs carry calcium ions and the results of our previous study, which showed that Hcrt/Orx stimulates a somatic Ca²⁺ influx in neurons of the DR and LDT, we postulated that the NSCC activated by Hcrt/Orx contributed to the Hcrt/Orx-stimulated Ca²⁺ influx in these neurons. Contrary to this idea, however, concurrent monitoring of membrane potential/holding current and [Ca²⁺]_i showed that the NSCC activated by Hcrt/Orx was not a measurable source for these transients at −60 mV or even after increasing the Ca²⁺ driving force by stepping the membrane to −100 mV. These data, along with the observations in unclamped cells that a simulated Hcrt/Orx depolarization mimics the Ca²⁺ transients evoked by Hcrt/Orx, indicate that the activation of voltage-operated Ca²⁺-permeable channels is necessary. It is worth noting that the NSCC activated by orexin during our imaging experiments appeared somewhat smaller than we previously observed under standard whole cell recording conditions. Initially, we thought this might result from some rundown of the NSCC because of the longer waiting period before Hcrt/Orx application that was required for dye-filling in the imaging experiments. Indeed, under standard recording conditions, most of our Hcrt/Orx applications started by 11 min after breakthrough. Because we also had numerous recordings where orexin was applied between 20 and 35 min after breakthrough, which is comparable to the times used in the imaging experiments, we compared the average magnitude of orexin current evoked by early (n = 22) and late applications (n = 6). We found they were not significantly different (P > 0.2), suggesting that delayed application of Hcrt/Orx is unlikely to importantly alter the nature of the NSCC current. Another possibility is that the signaling pathways required to activate the NSCC may be somewhat reduced because of increased intracellular Ca²⁺ buffering expected from bis-Fura2. Although this might account for somewhat smaller currents, this too is unlikely to have altered the channels underlying this current, because a noisy inward current was still clearly present in our imaging experiments. Thus our combined whole cell and imaging data argue strongly against involvement of the Hcrt/Orx-activated NSCC in mediating the Hcrt/Orx-evoked Ca²⁺ transients in DR and LDT neurons.

These findings may also provide some insight into the molecular identity of the channels underlying this NSCC in DR and LDT. Transient receptor potential canonical (TRPC) channel subunits (TRPC I-VIIs) form classic, nonselective cation channels (for review, see Ramsey et al. 2006) and are expressed throughout the nervous system and in serotonergic DR neurons, with TRPC5 and TRPC6 mRNAs occurring most frequently (Sergeeva et al. 2003). TRPC I-VII mRNAs also appear present within the LDT (Allen Brain Map, www.brain-map.org). TRPCs mediate cation currents activated by Gq-coupled receptors and have recently been shown to mediate a Ca²⁺ influx activated by Hcrt/Orx in mammalian expression systems. Orexin receptors expressed in CHO cells elevate [Ca²⁺]_i via activation of both a store-independent Ca²⁺ influx and a store-operated Ca²⁺ influx triggered by Ca²⁺ release from thapsigargin-sensitive intracellular stores (Kukkonen and Akerman 2001; Lund et al. 2000). Using a dominant negative approach, TRPC1 and TRPC3 channels were elegantly shown to be involved in the store-independent Ca²⁺ influx (Larsson et al. 2005). Hcrt/Orx also activates a calcium influx pathway in excitable neuroblastoma cells with involvement of TRPC6 channels (Nasman et al. 2006). It therefore seems reasonable to speculate that native orexin receptors activate TRPCs in DR and LDT neurons because these channels can be operated by orexin receptors and are expressed in DR and LDT neurons. On the other hand, following Hcrt/Orx activation, a Ca²⁺ influx was undetectable at subthreshold membrane potentials in these neurons. Thus despite the relatively higher abundance of TRPC5 and 6 mRNA in DR, perhaps Hcrt/Orx activates channels formed by TRPC1-4 subunits because they have a lower Ca²⁺ selectivity (PCa/PNa = 1–3) than do channels formed from TRPC5 and 6 subunits (Pca/PNa = 5–9; see Table 1 of Owsianik et al. 2006). Of course, it is also possible that the Ca²⁺ influx mediated by these channels is too low, too restricted, or too remote from our measurement sites to be detected in our preparation. Unfortunately, there are currently no pharmacological agents that can distinguish between TRPC channels, so their particular roles in mediating Hcrt/Orx actions in DR and LDT neurons remains a topic for future study.

Hcrt/Orx likely modulates L-type Ca²⁺ channels in DR and LDT neurons

Our findings indicated that Hcrt/Orx produces a membrane depolarization that activates voltage-operated Ca²⁺ channels and produces Ca²⁺ spiking in some DR and LDT neurons. Evidence from Ca²⁺ imaging studies in dissociated hypothalamic and spinal neurons (van den Pol 1999; van den Pol et al. 1998) and VTA neurons (Uramura et al. 2001) found that Hcrt/Orx-stimulated Ca²⁺ transients were sensitive to L- or L- and N- type channel antagonists. Because patch-clamp studies of hypothalamic neurons failed to show an Hcrt/Orx-mediated depolarization, a direct modulatory role of Hcrt/Orx on L-type channels was proposed. However, in the VTA, Hcrt/Orx strongly depolarizes dopamine and GABA neurons (Korotkova et al. 2003), so activation of Ca²⁺ channels might be secondary to this depolarization. Similarly, because Hcrt/Orx depolarizes DR and LDT neurons, it was unclear from our previous fura-AM measurements whether L-type Ca²⁺ channel involvement was merely a result of activation by depolarization, although indirect evidence suggested more than a depolarization was involved (Kohlmeier et al. 2004). To test whether depolarization was sufficient, in this study, we examined the effect of orexin-A on a voltage-clamp step-evoked Ca²⁺ transient. This showed that Hcrt/Orx significantly enhances these Ca²⁺ transients. The enhancement was prevented by pretreating with nifedipine and was augmented by pretreatment with Bay-K, indicating the enhancement was largely, if not exclusively, mediated by enhancement of L-channel Ca²⁺ transients in both DR and LDT neurons. Because orexin-A enhanced these Ca²⁺ transients both in cells with and without an orexin-A–evoked inward current, these data indicate Hcrt/Orx can independently enhance the Ca²⁺ transients generated by L-type Ca²⁺ channels in these cells. The precise mechanism by which this happens has not been resolved, but our data are consistent with Hcrt/Orx increasing the L-type calcium current, which is present in both DR (Penington et al. 1991) and LDT neurons (Kohlmeier and Leonard 2006). Indeed, Hcrt/Orx has been shown to increase L-current in cultured ovine somatotropes (Xu et al. 2002), where is does so without altering voltage dependence and depends on PKC signaling. PKC inhibition also attenuates Hcrt/Orx-mediated Ca²⁺ transients in hypothalamic neurons (van den Pol et al. 1998) and VTA neurons (Uramura et al. 2001). Previously, we found PKC inhibition partly blocked these transients in DR and LDT neurons (Kohlmeier et al. 2004). In this study, we extended this observation and showed that the orexin-A–mediated enhancement of the voltage step–evoked Ca²⁺ transient was completely abolished by pretreatment with Bis I, a widely used inhibitor of PKC. Presumably, the partial block we observed previously results from residual depolarization and the activation of non–L-type Ca²⁺ channels because some DR and LDT neurons showed a PKC inhibitor–resistant inward current. Thus our data are consistent with a PKC-dependent enhancement of the L-type Ca²⁺ current in DR and LDT neurons, although the particular pore-forming subunit involved and the nature of the modulation by Hcrt/Orx remains to be determined. It is noteworthy, in this regard, that although the Ca²⁺ transients produced by three to five spikes in LDT neurons are partly sensitive to L-channel antagonists, these spike-evoked transients were not enhanced by orexin-A (Kohlmeier and Leonard 2006). This implies that the Hcrt/Orx enhancement of L-type Ca²⁺ transients might require particular stimulus conditions like prolonged depolarization to be manifest.

Ca²⁺ response profiles and transmitter phenotype of DR and LDT neurons

In our previous study of fura-AM–loaded neurons of the DR and LDT, we found that Hcrt/Orx elicited three types of calcium response profiles: plateau responses that exhibited a slow rise time, a stable plateau, and were slow to recover; spiker responses, which showed rapid, transient increases of calcium from a stable baseline; and smooth/spikers, which were a combination of these other two responses (Kohlmeier et al. 2004). Because of the recording method, it was not possible to determine whether these response profiles co-varied with transmitter phenotype. In this study, transmitter phenotype was identified for recorded cells, and we found no particular relation between Ca²⁺ response profile and transmitter phenotype. All three response types were observed in transmitter identified cells of the DR and LDT and in other cell types. Rather, the response profile depended on the magnitude of the Hcrt/Orx depolarization and the ability of the recorded neuron to fire action potentials or Ca²⁺ spikes. Plateau responses were generated by an Hcrt/Orx-mediated depolarization and activation of voltage-gated Ca²⁺ channels (VGCCs) in neurons with insufficient current density to generate action potentials or Ca²⁺ spikes. Spiker responses were generated when the Hcrt/Orx-mediated depolarization triggered action potentials or Ca²⁺ spikes but was insufficient to maintain steady activation of voltage-gated Ca²⁺ channels. Smooth/spikers were generated by an Hcrt/Orx-mediated depolarization sufficient to activate VGCCs and to generate action potentials or Ca²⁺ spikes (in TTX). These latter responses were readily observed in current-clamped cells following orexin-A application (Fig. 1). In our previous study, we found that nifedipine reduced, but did not abolish, the Hcrt/Orx-evoked Ca²⁺ transients. Presumably, this reflects the ability of the Hcrt/Orx mediated depolarization to also activate other types of VGCCs. The effect of Hcrt/Orx to enhance the L-type Ca²⁺ channel-mediated Ca²⁺ transient would be expected to facilitate all of these response types. It should also be noted that in slices loaded with fura-AM, it was not possible to determine the resting membrane potential of the imaged neurons, and some cells might have been quite depolarized. Indeed, such neurons might be induced to fire Ca²⁺ spikes following enhancement of L-type Ca²⁺ channels by Hcrt/Orx even in the absence of an Hcrt/Orx-mediated depolarization. This is supported by our prior observations that Bay-K alone sometimes induced “spiker” responses and that “spiker” responses were more likely to be abolished by nifedipine than other types of responses (Kohlmeier et al. 2004). Thus the range of Ca²⁺ responses previously observed by Ca²⁺ imaging alone in the DR and LDT can be explained by the dual actions of Hcrt/Orx elucidated here.

Present address of S. Watanabe: Hamamatsu Corporation, 812 Joko-cho, Higashi-ku, Hamamatsu-city, 431-3196 Japan.