Multiple T-type Ca²⁺ current subtypes in electrophysiologically characterized hamster dorsal horn neurons: possible role in spinal sensory integration

Abstract

Whole cell patch-clamp recordings were used to investigate the contribution of transient, low-threshold calcium currents (I_T) to firing properties of hamster spinal dorsal horn neurons. I_T was widely, though not uniformly, expressed by cells in Rexed's laminae I–IV and correlated with the pattern of action potential discharge evoked under current-clamp conditions: I_T in neurons responding to constant membrane depolarization with one or two action potentials was nearly threefold larger than I_T in cells responding to the same activation with continuous firing. I_T was evoked by depolarizing voltage ramps exceeding 46 mV/s and increased with ramp slope (240–2,400 mV/s). Bath application of 200 μM Ni²⁺ depressed ramp-activated I_T. Phasic firing recorded in current clamp could only be activated by membrane depolarizations exceeding ∼43–46 mV/s and was blocked by Ni²⁺ and mibefradil, suggesting I_T as an underlying mechanism. Two components of I_T, “fast” and “slow,” were isolated based on a difference in time constant of inactivation (12 ms and 177 ms, respectively). The amplitude of the fast subtype depended on the slope of membrane depolarization and was twice as great in burst-firing cells than in cells having a tonic discharge. Post hoc single-cell RT-PCR analyses suggested that the fast component is associated with the Ca_V3.1 channel subtype. I_T may enhance responses of phasic-firing dorsal horn neurons to rapid membrane depolarizations and contribute to an ability to discriminate between afferent sensory inputs that encode high- and low-frequency stimulus information.

the dorsal horn of the spinal cord is a structure of major importance for processing sensory information from the body. It receives synaptic terminations from primary sensory neurons that innervate a variety of tissues, including skin, muscle, tendons, joints, and viscera (Willis and Coggeshall 2004). In this region, diverse sensory information is integrated by local neural circuits, which are assembled from synaptic connections between multiple, physiologically distinct types of interneurons (Lu and Perl 2003, 2005; Santos et al. 2007; Schneider 2008) and then relayed to segmental reflex circuitry and ascending sensory pathways transmitted to the brain and brain stem.

Intrinsic electrophysiological properties of neurons can influence the integration of sensory information within dorsal horn circuits. At least three types of dorsal horn neurons have been identified in rodents based on discharge properties to current injection: tonic, phasic, and delayed firing (Graham et al. 2004; Hochman et al. 1997; Prescott and De Koninck 2002; Ruscheweyh and Sandkühler 2002; Schneider 2003; Walsh et al. 2009). Repetitive discharge of tonically firing cells has been reported to be regulated by a combination of voltage-gated Na⁺ and K⁺ channels and persistent Ca²⁺ currents (Melnick et al. 2004; Prescott and De Koninck 2005). Other evidence suggests that a transient, low-voltage-activated (T-type) Ca²⁺ conductance underlies the rapidly adapting discharge of phasic neurons (Russo and Hounsgaard 1996).

T-type Ca²⁺ channels are widely expressed in the central nervous system (CNS) (McKay et al. 2006; Nilius et al. 2006; Talley et al. 1999), where they promote burst firing and intrinsic oscillatory activity, and participate in synaptically evoked Ca²⁺ influx (Huguenard 1996). A Ca²⁺ current mediated by T-type channels has also been reported in spinal dorsal horn neurons of rat and mouse (Huang 1989; Ryu and Randic 1990; Walsh et al. 2009) and also may contribute to rhythm generation by ventral horn interneurons (Wilson et al. 2005). Dorsal horn T-channels have received renewed focus because of evidence that they enable calcium-dependent long-term potentiation of nociceptive transmission mediating hyperalgesia (Heinke et al. 2004; Ikeda et al. 2003). Our interest in the contribution of T-type Ca²⁺ channels to spinal somatosensory processing was prompted by two observations. First, fast depolarizing current ramps activate a transient membrane depolarization and discharge in many phasic-firing dorsal horn neurons, with the amplitude of depolarization and firing frequency being graded with the ramp slope (Schneider 2003). Second, phasic cells in the dorsal horn region where mechanically sensitive cutaneous afferents terminate respond selectively to stimuli that produce rapid deformations in the skin (Schneider 2005), suggesting that this class of neurons may be part of specialized spinal circuits for encoding velocity information. A similar observation was made by Prescott and De Koninck (2002), who found that phasic cells in lamina I are driven by trains of high-frequency stimuli and argued that they act as coincidence detectors responding to simultaneous occurrence of separate afferent inputs. In thalamic neurons, T-type currents are selectively activated by fast voltage ramps (Crunelli et al. 1989) and generate large, transient Ca²⁺ depolarizations that contribute to oscillatory behavior (Gutierrez et al. 2001). Given the importance of T-type channels to rhythmic firing behavior in thalamic neurons, we considered the possibility that these channels also function in dorsal horn sensory integration to shape responsiveness of neurons to dynamic membrane depolarizations induced by primary sensory afferents.

Three pore-forming α₁-subunits of T-type Ca²⁺ channels (Ca_V3.1, Ca_V3.2, and Ca_V3.3) have been cloned (Perez-Reyes 2003), and mRNAs of all three subunits are expressed in the spinal cord (Talley et al. 1999). Moreover, each T-channel subunit exhibits gating properties that may enable them to contribute differentially to neuronal excitability (Chemin et al. 2002). In this study, we investigated the distribution of the T-type Ca²⁺ current among dorsal horn neurons with the goal of determining whether T-type current expression is related to neuronal firing properties. Experiments were performed in an isolated spinal cord preparation from Syrian hamsters that we have used previously to investigate neuronal properties and connectivity in the spinal dorsal horn (Schneider 2003, 2005, 2008). Our observations indicate that dorsal horn neurons with phasic firing patterns express a higher T-type Ca²⁺ current density than tonically firing cells and provide evidence that a fast T-type current may help tune responses of phasic-firing neurons to rapid membrane depolarizations and stimulus movement or rate of change (Schneider 2005).

METHODS

All protocols involving the use of live animals were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at Michigan State University.

Slice preparation.

Spinal cord tissue was obtained from 9- to-14-day-old Syrian hamsters of both sexes under urethane anesthesia (1.5 mg/g ip) followed by exsanguination as described previously (Schneider 2003, 2008). After removal of vertebrae and meninges, a block of lumbosacral spinal cord was isolated and glued to the stage of a vibrating microtome (Vibratome 3000, St. Louis, MO) and 300-μm-thick transverse slices were cut. Dissection and slicing were carried out at 4–8°C in a solution containing (mM) 179 sucrose, 2.5 KCl, 0.2 CaCl₂, 10 MgCl₂, 1.25 NaH₂PO₄, 26 NaHCO₃, and 25 glucose (pH 7.35–7.45, 290–310 mosM) equilibrated with 95% O₂-5% CO₂. In some experiments, sucrose was replaced by equimolar glycerol in the dissection solution (Ye et al. 2006). (There were no differences in resting potential or amplitude and half-width of action potentials recorded from neurons in slices prepared with sucrose- or glycerol-based dissection solutions.) Slices were incubated in an oxygenated artificial cerebrospinal fluid (ACSF) containing (mM) 120 NaCl, 2.5 KCl, 2.5 CaCl₂, 1.5 MgSO₄, 1.25 NaH₂PO₄, 26 NaHCO₃, and 10 glucose (pH 7.35–7.45, 290–310 mosM) at room temperature for 1 h prior to electrophysiological recording.

Electrophysiology.

For electrophysiological recording, slices were transferred to a chamber (volume 1.5 ml) mounted on a fixed-stage microscope (Olympus BX51WI) and continuously perfused at 5–6 ml/min with ACSF bubbled with 95% O₂-5% CO₂. Patch pipette recording electrodes (4–7 MΩ) were fabricated from borosilicate glass (N-51A, Drummond Scientific, Broomall, PA). The internal pipette solution contained (mM) 130 K-gluconate, 5 NaCl, 1 CaCl₂, 1 MgCl₂, 11 EGTA, 10 HEPES, 2 Mg-ATP, and 0.1 Li-GTP (pH 7.3, 280–285 mosM). For isolating T-type Ca²⁺ currents, we switched to a modified extracellular solution with the following constituents (mM): 107 NaCl, 3 BaCl₂, 2 CsCl, 1.3 MgCl₂, 26 NaHCO₃, 10 glucose, 20 TEA, 0.5 μM TTX, and 10 μM CdCl₂ (pH 7.35–7.45, 290–310 mosM). In some experiments, the Ba²⁺ concentration was lowered to 2 mM and Cd²⁺ was raised to 20 μM or 40 μM. All recordings took place within 4–5 h after tissue dissection and were performed at 27°C to promote slice viability (Schneider 2003).

Recording sites were first identified at low power (×10) relative to the translucent band corresponding to the substantia gelatinosa (lamina II) visible under transillumination. Individual neurons spanning a range of sizes were then targeted for patch-clamping using a ×40 water immersion objective and infrared differential contrast optics. Whole cell current and membrane voltage were recorded with an Axopatch 1D amplifier (MDS Analytical Technologies, Toronto, ON, Canada). All recordings were initiated in voltage-clamp mode after nulling the input offset voltage and minimizing capacitance transients. After whole cell recording was established, cell capacitance [25 ± 8 pF (SD)] and series resistance [33 ± 9 MΩ (SD)] were compensated. Signals were amplified (bandwidth 0–5 kHz) and saved to disk with a Digidata 1320A data acquisition system running pCLAMP software (MDS Analytical Technologies). Membrane current exceeding −0.05 nA at a holding potential of −60 mV indicated a poor recording condition, and these cells were excluded from analysis. Voltage-activated currents were low-pass filtered at 500 Hz (Gaussian). P/N leak subtraction was used to correct for passive membrane current. In most cases, currents are expressed as current density (pA/pF) by dividing by cell capacitance, as estimated by measuring the transient current in response to 5-mV hyperpolarizing current pulses and observing the value of the whole cell capacitance adjustment on the amplifier. Voltage commands and membrane potentials were corrected post hoc for a −14-mV liquid junction potential calculated with pCLAMP (MDS Analytical Technologies).

Neuronal electrophysiological properties were studied via membrane potential recordings performed in current-clamp mode prior to isolation of T-type Ca²⁺ currents under voltage clamp. Neurons were stimulated to produce action potentials by application of depolarizing current pulses (3 s) through the recording pipette for analyses of spike-frequency adaptation patterns. Neuronal input resistance (R_in) was measured by recording passive membrane responses to pulses of hyperpolarizing current. The dynamics of spike activation were studied by presenting a series of ramp-hold waveforms (duration 4.5 s) with increasing ramp slopes and determining the minimum rate of depolarization required to activate action potentials (Schneider 2003).

T-type current was activated at −44 mV from a conditioning potential of −104 mV held for 500 ms, and averages of three to five trials were used for subsequent analyses. A small, sustained current at a conditioning potential of −54 mV was subtracted from the current activated from −104 mV prior to analysis. To estimate different kinetic components of the net T-current, a semilogarithmic plot of the current was produced. The slow phase was fitted with a linear function, and the slope of the regression was used to determine the decay time constant. The amplitude was estimated by extrapolating the linear function to the time to peak of the net current. The amplitude and decay time of a fast current were then approximated by subtracting the slow current amplitude from the peak current (Zhuravleva et al. 1999).

Single-cell RT-PCR.

In some experiments, intracellular material was harvested by applying negative pressure to the electrode pipette. The pipette contents (∼5–8 μl) were expelled into a PCR tube containing 5 μl of nuclease-free water (Fisher Scientific, Pittsburgh, PA), 1 μl of dithiothreitol (DTT, 0.1 M; Invitrogen, Carlsbad, CA), 1 μl of RNase inhibitor (40 U/μl; Promega, Madison, WI), and 1 μl of random hexanucleotides (0.5 μg/μl; Promega) and stored at −80°C. Reverse transcription was carried out as previously reported (Han et al. 2005) with minor modification. The mixture was heated to 70°C for 10 min and quickly chilled on ice. After centrifugation, the solution was mixed with 4 μl of 5× First-Strand Buffer (mM: 250 Tris·HCl, 375 KCl, 15 MgCl₂; Invitrogen), 1 μl DTT (0.1 M; Invitrogen), 1 μl of mixed deoxynucleotide triphosphates (dNTPs, 10 mM; Invitrogen), and 1 μl of SuperScript II reverse transcriptase (RT; Invitrogen). After 10-min incubation at room temperature, the solution was warmed to 42°C for 50 min for cDNA synthesis. The reaction was terminated by heating at 70°C for 15 min, and RNase H (2 U/μl; Invitrogen) was added to remove RNA from RNA-DNA hybrids. cDNA was stored at −20°C until being used in PCR. PCR conditions were optimized by using total hamster spinal cord RNA. The presence of glyceraldehyde-3-phosphate dehydrogenase (GAPDH) cDNA was examined to confirm successful cell harvesting and reverse transcription. For GAPDH-positive samples, a two-step PCR protocol was carried out to detect the three known Ca_V3 subunit cDNAs. The first amplification was performed on a TGradient thermocycler (Biometra, Göttingen, Germany). The product was then diluted 10 times and used in subsequent real-time PCR with the Stratagene Mx3000P QPCR system (Agilent Technologies, Santa Clara, CA). For the first-step of Ca_V3 PCR, one 50-μl reaction mixture was prepared for each sample and contained 25 μl of PCR Master Mix (Promega), 5 μl of single-cell cDNA template, and outer primers: Ca_V3 forward, 5′-CGCTCCAGCCGGAACAGC-3′; Ca_V3 reverse, 5′-GCAATGGTGATACAGTTGAGGA-3′. The reaction went through 40 cycles of 95°C for 1 min, 53°C for 1 min, and 72°C for 1 min. Second-step real-time PCR was performed with SYBR GreenER qPCR SuperMix (Invitrogen). Triple 25-μl reactions with 2 μl of single-cell cDNA template or the diluted first PCR product were prepared for each cDNA species of interest. The amplification proceeded with the following profile: 50°C for 2 min and 95°C for 10 min, followed by 45 cycles (50 cycles for GAPDH) of 95°C for 30 s, 60°C for 1 min, and 72°C for 30 s. The sequences of the primers are as follows: GAPDH forward, 5′-AGCAATGCATCCTGCACCACCA-3′; GAPDH reverse, 5′-ACAGCCTTGGCAGCACCAGT-3′; Ca_V3.1 forward, 5′-GCTTCCTGCCTGTTGCCGAG-3′; Ca_V3.2 forward, 5′-TGCTCGAACCCTATGCTCCCC-3′; Ca_V3.3 forward, 5′-GCAGAATGCCCAACATCGCCA-3′. The Ca_V3 reverse primer was paired with subunit-specific forward primers and used in the second-step semi-nested Ca_V3 PCR. All primers were designed by Primer3 (Rozen and Skaletsky 2000) with reported cDNA sequences [DDBJ/EMBL/GenBank accession no.: GAPDH, U10983; Ca_V3.1, AF027984 (rat); Ca_V3.2, AF290213 (rat); Ca_V3.3, AF086827 (rat)] and are intron spanning. The sizes of amplicons were (bp) 209 GAPDH, 153 Ca_V3.1, 166 Ca_V3.2, 277 Ca_V3.3. PCR-generated fragments were verified by electrophoresis and sequence analysis, and the melting temperature of the fragments was used for confirmation of specific amplification in real-time PCR. No amplification of GAPDH was detected after reverse transcription in controls in which RT had been omitted. Negative controls without cDNA template and positive controls using total hamster spinal cord cDNA template were included in each PCR. Detection of a transcript was confirmed if the linear phase of the amplification plot crossed the fluorescence intensity threshold before reaching the 40th (or 45th for GAPDH) amplification cycle [threshold cycle (C_t) < 40 or 45].

Drugs and chemicals.

TTX was purchased from MP Biomedicals (Irvine, CA). Mibefradil (Ro 40–5967, dihydrochloride hydrate) was purchased from Sigma-Aldrich (St. Louis, MO). All other reagents were obtained from Sigma-Aldrich or Mallinckrodt Baker (Phillipsburg, NJ).

Statistical analyses.

Numerical data are presented as means ± SE, unless noted otherwise. Statistical analyses were performed with InStat 3 (version 3.06; GraphPad Software). Statistical comparisons between two groups were made with the unpaired t-test with Welch correction or Mann-Whitney test as appropriate. Comparison among three or more groups was conducted with one-way ANOVA with Tukey-Kramer multiple comparisons or Kruskal-Wallis one-way ANOVA with Dunn's multiple comparisons. The criterion for statistical significance was a value of P < 0.05.

RESULTS

Whole cell recordings were obtained from 108 dorsal horn neurons in spinal slices prepared from 71 hamsters. Neurons had resting membrane potentials more negative than −64 mV (corrected for liquid junction potential) under current-clamp conditions. Membrane potential was held at −74 mV during voltage-clamp recordings, unless otherwise noted.

Isolation and characterization of T-type calcium currents in hamster dorsal horn neurons.

T-type Ca²⁺ currents (I_T) were characterized previously in rat dorsal horn neurons (Huang 1989; Ryu and Randic 1990); however, the expression of T-currents and their properties in hamster dorsal horn is not known. We therefore began our study by first examining I_T in hamster dorsal horn neurons under voltage clamp prior to making comparisons with spike discharge patterns. Voltage-clamp recordings were performed in the presence of TEA, TTX, and Ba²⁺ under ionic conditions appropriate for isolating Ca²⁺ currents. In the absence of Cd²⁺, depolarizing the membrane potential to −44 mV after a 500-ms conditioning potential of −104 mV activated an inward current with two components (Fig. 1A), an initial rapidly decaying phase followed by a late sustained phase. Addition of 10 μM Cd²⁺ reduced the amplitude of the initial phase and almost completely abolished the late sustained current within 3 min of application (Fig. 1B). Increasing the concentration to 20 μM had no additional effect on the current, suggesting that 10 μM Cd²⁺ was sufficient to block high-voltage-activated Ca²⁺ channels. Therefore, subsequent experiments were performed with the lower Cd²⁺ concentration.

Fig. 1.

Isolation and characteristics of low-voltage-activated Ca²⁺ currents. A: inward currents recorded from a representative dorsal horn neuron were activated every 20 s by a 200-ms voltage step to −44 mV from a conditioning potential of −104 mV held for 500 ms (top). Aa: each of the superimposed traces (black, before application of Cd²⁺; gray, 10 μM Cd²⁺ and 20 μM Cd²⁺) was averaged from 4 consecutive current responses (+ in B). Ab: inward current that was sensitive to 10 μM Cd²⁺ was obtained by subtraction. Ac: inward current that underwent additional blockade after increase of Cd²⁺ from 10 μM to 20 μM. Dotted lines indicate the baseline for current measurements. B: amplitude of peak and residual current was plotted against the time elapsed. Cursor positions for measuring peak (●) and residual (○) current are shown in Aa; 10 μM and 20 μM Cd²⁺ were added sequentially to the bath solution as indicated by shading. C: voltage-dependent inactivation and activation of T-type Ca²⁺ currents in spinal dorsal horn neurons. The current amplitude was normalized to the maximum level (I/I_max), averaged, and plotted with the holding potential (V_h) for the inactivation curve (□) or the test potential (V_test) for the activation curve (●). The curves were fitted with the Boltzmann equation. Error bars represent SD (n = 4). D1 and D2: effect of T-type Ca²⁺ channel blockers Ni²⁺ (D1) and mibefradil (D2) on I_T. I_T was activated every 20 s by stepping the membrane potential to −44 mV for 200 ms after a 500-ms conditioning potential of −104 mV. Peak current amplitude is plotted as a function of time elapsed. Cursor position used to measure peak current (●) is shown in a. Shading indicates when the drug was present in the bath. D1a and D2a: the black traces were averaged from 3 consecutive measurements (+) before application of Ni²⁺ or mibefradil (MIB), and the gray trace was averaged from 3 trials (+) near the end of the drug presentation. D1b and D2b: inward currents that were sensitive to either of the blockers, obtained by subtraction.

We next examined the voltage dependence of steady-state inactivation and activation of I_T. Inactivation kinetics were analyzed by stepping the membrane potential to −44 mV every 20 s in 10-mV increments from 500-ms conditioning potentials of −134 mV to −54 mV. I_T was largely inactivated at potentials positive to −74 mV, and this inactivation was removed as the membrane potential approached −114 mV. Averaged normalized currents (I/I_max) were plotted as a function of conditioning potential (n = 4) and fitted with the Boltzmann equation (I/I_max = 1/{1 + exp[(V − V_1/2)/k]}) having a slope factor (k) of 12 mV and voltage for half-maximal current (V_1/2) of −88 mV (Fig. 1C). Activation properties of I_T were studied by delivering 200-ms depolarizing voltage pulses after a conditioning potential of −104 mV to −24 mV in 10-mV increments. I_T appeared in the low-voltage range with an activation threshold near −74 mV and reached a maximum near −34 mV. Values for I/I_max were plotted against the membrane potential to generate an activation curve yielding k = 8 mV and V_1/2 = −61 mV when fitted with the Boltzmann equation. Overall, activation and inactivation parameters were consistent with I_T reported previously under similar conditions (Huang 1989; Huguenard and Prince 1992; Ryu and Randic 1990).

Having established that activation and inactivation characteristics of I_T were similar in rat and hamster dorsal horn cells, we then investigated the sensitivity of the current to T-channel antagonists. As shown in Fig. 1D1, 200 μM Ni²⁺ added to the bathing solution depressed I_T by 52 ± 6% (n = 6). Reduction in I_T was partially reversible, with recovery to 78 ± 10% of control after 4–16 min of washout. The I_T antagonist mibefradil (5 μM) resulted in a 47% ± 4% (n = 3) reduction in I_T amplitude after 6 to 10 min of perfusion (Fig. 1D2), and the effect persisted after drug washout. The results are consistent with previous pharmacological studies of T-current activation (Martin et al. 2000; Perez-Reyes 2003) but indicated that a portion of the transient inward Ca²⁺ current was resistant to these two T-channel antagonists.

T-type current in neurons with identified membrane and discharge properties.

Although nearly all hamster dorsal horn neurons expressed I_T (n = 41/42, 98%), the current density varied considerably from cell to cell [6 ± 4 pA/pF (SD)]. Despite this, we found no differences in I_T between neurons recorded in laminae I/II and III/IV (6 ± 1 pA/pF vs. 7 ± 1 pA/pF; P = 0.46), indicating that expression of I_T is similar in the superficial and deep dorsal horn. Previous studies have suggested that burst firing of turtle dorsal horn neurons at a hyperpolarized membrane potential is associated with activation of a low-voltage-activated Ca²⁺ conductance (Russo and Hounsgaard 1996). Hamster dorsal horn neurons exhibit a range of spike frequency adaptation (Schneider 2003, 2005) and also include a type with a rapidly adapting burst discharge. Thus we examined I_T expression in a population of neurons with characterized firing properties. Initial characterization of firing properties was performed in current-clamp mode in standard ACSF solution. Measurements of I_T were then made after switching to voltage clamp by depolarizing the membrane potential to −44 mV from −104 mV held for 500 ms and subtracting residual current activated from a conditioning potential of −54 mV to minimize contribution from any unblocked high-voltage-activated Ca²⁺ current.

Cells were classified as tonic (n = 16/46, 35%) or phasic (n = 26/46, 57%) according to their firing behavior in response to depolarizing current pulses (Fig. 2, Table 1). [Analyses do not include four neurons that responded to the current pulse with a marked delay to the first action potential (not shown).] Tonic cells fired action potentials continuously during the period of depolarization (Fig. 2A, left). Phasic cells generated a transient, rapidly adapting discharge when activated by a constant level of depolarizing current (Fig. 2A, center and right). Phasic cells had a more hyperpolarized resting membrane potential and lower R_in than tonic cells (Table 1), consistent with previous observations (Schneider 2003). They displayed a fairly wide range of spike frequency adaptation and were further divided into two subgroups (Fig. 2A, Table 1). Short burst (SB) phasic cells responded to depolarizing current injection with only one or two action potentials superimposed on a transient membrane depolarization (Fig. 2A, arrowheads). The remaining phasic cells (non-SB) typically responded to depolarizing current with three or more action potentials at rapidly diminishing frequency, and the number of spikes increased with stimulus amplitude (Fig. 2A, right). When depolarized by ramp-hold current commands, tonic cells discharged primarily during the steady-state depolarization and the discharge frequency was unrelated to the trajectory of the ramp phase (Fig. 2B, left). Typically, SB and non-SB-type phasic cells were activated only by the most rapid membrane depolarizations (Fig. 2B, center and right). The slope of membrane depolarizations needed to activate phasic cells was five times higher than those required for activating tonic-firing neurons (Table 1). The responses of SB cells to step and ramp-hold current injection were blocked by Ni²⁺ (Fig. 2C).

Fig. 2.

Firing patterns of hamster dorsal horn neurons. A: responses of a representative cell from each category (top 2 traces) to depolarizing current pulses applied through the recording pipette (bottom trace) (categories: SB, short burst phasic; non-SB, non-short burst phasic; see text for details). Resting membrane potential (V_m) for each neuron is indicated to left of the first response. Arrowheads indicate transient membrane depolarizations characteristic of SB neuron activation. B: responses of the same neurons to ramp-hold current commands that depolarized V_m at different rates (waveforms are superimposed on individual traces). Slope of membrane depolarization (dV/dt) for lower and middle traces: 5 mV/s and 10 mV/s for tonic cell; 18 mV/s and 37 mV/s for SB; 25 mV/s and 55 mV/s for non-SB. Top trace: rectangular current step to the same level of depolarization. Action potential responses of phasic-firing cells in B are indicated by arrows. C: effect of Ni²⁺ on SB-type phasic neurons. Top traces show current commands, and bottom traces show voltage responses. Ni²⁺ (200 μM) blocked the spike discharges generated by current steps (C1) and ramp-activated depolarization (C2). In C2, blockade of the transient depolarization response is indicated by arrows. The series of ramp commands depolarized V_m at rates of 6 mV/s to 48 mV/s, and a transient depolarization was activated by a voltage ramp of 48 mV/s. Records showing responses to current steps (C1, top traces) consist of 5 consecutive traces. Resting V_m for each condition is indicated to left of the traces. ACSF, artificial cerebrospinal fluid.

Table 1.

Comparison of I_T and intrinsic membrane properties of spinal dorsal horn neurons with different firing properties

		Phasic
	Tonic (n = 16)	Short burst (n = 7)	Non-short burst (n = 19)
Resting membrane potential, mV	−69 ± 2 (14)	−81 ± 1*†	−76 ± 2*
Input resistance, MΩ	654 ± 84	261 ± 66*	377 ± 63* (18)
Ramp slope threshold, mV/s	9 ± 3 (12)	43 ± 3*	46 ± 7* (11)
IT, pA/pF	5 ± 1	14 ± 4*†	5 ± 1

All values are listed as means ± SE. The number of measurements used to calculate each value is given in parentheses if different from the number of cells for the group. I_T, transient calcium current.

^*Significantly different from tonic cells;

^†significantly different from non-short burst cells.

As expected, I_T was found in the majority of neurons in this sample (n = 43/46, 93%) and showed considerable variation in amplitude [7 ± 6 pA/pF (SD); 1–36 pA/pF], consistent with the results of our initial voltage-clamp experiments. Figure 3A shows examples of I_T from tonic, SB-type phasic, and non-SB-type phasic cells, and data from all cells in the sample, expressed as current density, are summarized in Table 1 and Fig. 3B. Notably, I_T in SB-type phasic cells was almost threefold larger than that recorded from tonic-firing neurons. Thus I_T expression was greatest in dorsal horn neurons with signature rapidly adapting burst firing that are selectively activated by rapid membrane depolarizations.

Fig. 3.

I_T in neurons with characterized firing patterns. A: dorsal horn neurons were voltage clamped at a V_h of −104 mV, and I_T was evoked by applying depolarizing voltage steps to −44 mV. External bathing solution for isolating T-type Ca²⁺ currents contained 3 mM Ba²⁺ and 10 μM Cd²⁺ and was applied after firing pattern characterization under current clamp in normal ACSF. Currents for each firing pattern classification are computed from averages of 5 traces. I_T density: 6 pA/pF (tonic); 19 pA/pF (SB); 7 pA/pF (non-SB). Dotted line indicates the baseline for inward current measurements. B: quantitation of I_T current density for tonic (n = 16), SB (n = 7), and non-SB (n = 19) phasic cells. Each symbol represents the measurement from 1 neuron. Height of histogram bars reflects the mean of measured values for each group, and error bars signify SE.

Characterization of T-current components.

Previous studies in thalamic neurons (Shcheglovitov et al. 2005; Tarasenko et al. 1997; Zhuravleva et al. 1999, 2001) suggested that multiple channel subtypes may contribute to the net T-current in dorsal horn neurons when the high-voltage-activated Ca²⁺ current is blocked. We found that for the majority of cells (67%, 29/43), I_T could be dissected into two components based on time constants of decay (Fig. 4, A and B1). A rapidly inactivating component (“fast”) had a time constant (τ_f) of 12.0 ± 1.1 ms (n = 48, Fig. 4C), and a slower component (“slow”) inactivated with a significantly longer time constant (τ_s) of 177 ± 12 ms (n = 32; vs. τ_f, P < 0.0001; Fig. 4C). The amplitudes of fast and slow currents were similar across cells (4.6 ± 0.7 vs. 3.5 ± 0.4 pA/pF, P = 0.2823). However, 10–90% rise time for the fast subtype was shorter than for the slow current (5.2 ± 0.8 vs. 14.3 ± 3.0 ms, P = 0.0167), suggesting a difference in activation kinetics. Two fast components, τ_f,1 and τ_f,2 (Fig. 4B2), having significantly different decay τ (5.7 ± 0.9 vs. 21.1 ± 3.9 ms, P = 0.0018) but similar current densities (3.2 ± 0.5 vs. 3.4 ± 0.5 pA/pF, P = 0.7374) could be distinguished in eight neurons (19%, 8/43). Therefore, the statistical sample size reflects the number of current components, rather than that of recorded neurons. Bath application of Ni²⁺ produced greater reduction of the fast current than the slow component (51 ± 5% vs. 15 ± 11%, P = 0.0144; Fig. 4D). Furthermore, the fast current in SB-type cells was two- and threefold larger than in tonic and non-SB neurons, respectively, while the slow subtype was uniformly expressed across the different firing pattern groups (Table 2).

Fig. 4.

Estimating subtypes of T-type Ca²⁺ currents in dorsal horn neurons. A: an inward current (lower) was activated from a representative dorsal horn neuron by a voltage step (upper) under conditions used for isolating T-current described in methods. A single exponential was fitted to the slow decay phase and extended to the beginning of the current trace. The amplitude of a “slow” current component is estimated by extrapolating the fitted exponential to the time of the peak current, indicated by the vertical dotted line. The amplitude of a “fast” component is obtained by subtracting the slow component from the peak current. B1: semilogarithmic plot of the current trace in A (●) with the slow phase fitted with a linear function, the slope of which provides an estimate of the time constant for exponential decay (τ_s = 175 ms), and semilogarithmic plot of a “fast” current component (○) resulting after subtraction of the “slow” component and fitted with a line having a slope indicating the decay time constant (τ_f = 10 ms). B2: example from a cell in which two “fast” components were isolated; one with τ_f,1 = 4 ms (○) and another with τ_f,2 = 14 ms (▲). C: histogram showing the distribution of decay time constants estimated for fast and slow current subtypes (n = 80) from 43 dorsal horn neurons. D: effect of 200 μM Ni²⁺ on the fast and slow T-current subtypes in spinal dorsal horn neurons (n = 5 cells). Horizontal bars indicate mean % inhibition of the peak current. The difference in inhibition is significant (P = 0.0144).

Table 2.

Comparison of I_T subtypes in spinal dorsal horn neurons with different firing properties

		Phasic
	Tonic (n = 14)	Short burst (n = 7)	Non-short burst (n = 16)
“Fast” current, pA/pF	4.2 ± 0.9 (16)	9.0 ± 2.7 (9)*†	3.0 ± 0.3 (17)
τf, ms	15.3 ± 2.6 (16)	9.3 ± 1.2 (9)	10.3 ± 1.0 (17)
“Slow” current, pA/pF	3.8 ± 0.9 (8)	3.3 ± 0.8 (7)	3.6 ± 0.8 (13)
τs, ms	174 ± 18 (8)	167 ± 26 (7)	187 ± 25 (13)

Data are given as means ± SE. τ_f, time constant of fast component; τ_s, time constant of slow component. The number of measurements used to calculate each value is given in parentheses if different from the number of cells for the group.

^*Significantly different from tonic cells;

^†significantly different from non-short burst cells.

We next examined the voltage dependence of steady-state inactivation and activation of the fast and slow T-currents. Inactivation kinetics were assessed by stepping the membrane potential to −44 mV every 20 s in 10-mV increments from conditioning potentials of −134 mV to −84 mV held for 500 ms. Activation properties were studied by delivering 200-ms depolarizing voltage steps in 10-mV increments from −64 mV to −24 mV after a conditioning potential of −104 mV. Currents generated by voltage steps to values below −64 mV were too small to be reliably analyzed by the stripping procedure, perhaps because of the small size of hamster dorsal horn neurons. Average normalized currents (I/I_max) were plotted as a function of conditioning potentials for inactivation or testing steps for activation (n = 3, “fast”; n = 2, “slow”) and fitted with the Boltzmann equation (I/I_max = 1/{1 + exp[(V − V_1/2)/k]}) (Fig. 5). The half-inactivation potentials for the fast and slow components were −96 mV and −80 mV, respectively, while the voltage required for half-maximal activation was −59 mV for the fast subtype and −65 mV for the slow subtype. The overlap region between the activation and inactivation curves (“window” current) is greater for the slow current component.

Fig. 5.

Steady-state activation and inactivation of fast and slow T-currents in spinal dorsal horn neurons. A and B: inactivation (□) and activation (●) curves for the fast (A) and slow (B) subtypes were fitted with the Boltzmann equation (I/I_max = 1/{1 + exp[(V − V_1/2)/k]}). V_1/2 and k derived from the fits are shown next to each curve. Error bars represent SE (n = 3, A and n = 2, B). r = 0.99 for all fits in A and B.

To find out what channel subtypes might contribute to I_T in hamster dorsal horn neurons, we performed single-cell RT-PCR (scRT-PCR) following patch-clamp recordings. Among 29 cells in which GAPDH mRNA was detected in the pipette aspirates, 6 were in LI/LII, 23 were from LIII–IV, and nearly all of them (93%, 27/29) evidenced I_T (7 ± 1 pA/pF). Ca_V3 mRNAs were detected in 12 neurons (Table 3), all except 1 recorded from LIII–IV. The majority (64%, 7/11) had phasic firing patterns, with the remainder (35%, 4/11) being tonic firers, not different from non-Ca_V3-expressing neurons. Similarly, Ca_V3-expressing cells evidenced I_T similar to their non-Ca_V3-expressing counterparts (8 ± 3 pA/pF vs. 6 ± 1 pA/pF). Analyses showed a higher occurrence of the Ca_V3.1 subunit (67%, 8/12) than Ca_V3.2 (8%, 1/12) and Ca_V3.3 (25%, 3/12). An example of a Ca_V3.1 amplification plot resulting from a second-step real-time PCR, along with a voltage-clamp recording of I_T from the same cell, is shown in Fig. 6. In contrast to our electrophysiological analyses, there was no evidence from scRT-PCR that single neurons express multiple T-channel subtypes.

Table 3.

Summary of T-type calcium channel subtype expression in characterized hamster dorsal horn neurons

		CaV Isotype			T-Current Component
Cell	Firing pattern	CaV3.1	CaV3.2	CaV3.3	“Fast”	“Slow”
080707-1	P	+	−	−	+	−
080708-1	T	+	−	−	+	−
080709-2	T	+	−	−	+	−
080523-1	T	+	−	−	+	+
080530-1	P	+	−	−	+	+
080619-1	P	+	−	−	+	+
080620-2	P	+	−	−	+	+
080718-1	T	+	−	−	+	+
080716-1	P	−	+	−	+	+
080605-3	P	−	−	+	+	+
080709-1	na	−	−	+	+	+
090212-4	P	−	−	+	+	+

Cell 080620-2 was located in Rexed's lamina (L) II; all other cells were in LIII–V. Expression of Ca_V3 subunit mRNAs was examined by single-cell RT-PCR (scRT-PCR). I_T was recorded under voltage clamp. “Fast” and “slow” current components were identified post hoc by fitting the net T-current with exponential functions (see methods). +, Detection of a channel isotype or current component; −, channel isotype or current absent; P, phasic; T, tonic; na, not analyzed.

Fig. 6.

Detection of Ca_V3.1 subunit mRNA from a dorsal horn neuron with a “fast” I_T component (see text). Amplification plots of Ca_V3.1 subunit fragments were generated from a real-time PCR experiment. Each measurement was averaged from 3 replicate reactions. The fluorescence signal was normalized to an internal passive reference dye (dRn), and plots were base-lined by Stratagene software. Amplified product from cell 080707-1 (●) and hamster spinal cord cDNA with (■, positive control 1) and without (▲, positive control 2) a first-step Ca_V3 PCR amplification are shown. For a negative control, cDNA template was replaced with water (○). The threshold fluorescence value (indicated by dashed line) was determined by the software using an amplification-based algorithm, resulting in threshold cycles (C_t) of 23, 31, and 35 for positive control 1, amplified cell product, and positive control 2, respectively. Inset: the decay of I_T (bottom) recorded from the same dorsal horn neuron can be fitted with a single exponential (τ = 20 ms). The voltage command used to activate the current is shown above the trace.

Activation of T-type calcium currents in dorsal horn neurons by voltage ramps.

T-type calcium currents in dorsal horn neurons could have important functional consequences for spinal sensory processing. I_T in thalamic neurons exhibits a strong rate dependence; activation requires membrane depolarization exceeding 30 mV/s (Crunelli et al. 1989). Therefore, nonuniformity of I_T expression by dorsal horn neurons could contribute to their differential responses to rate of membrane depolarization and time-varying sensory stimuli (Schneider 2003, 2005). Hence, we investigated whether I_T in dorsal horn cells is related to the slope of membrane depolarization (dV/dt), similar to the characteristic reported for thalamic neurons (Crunelli et al. 1989; Gutierrez et al. 2001).

I_T activation dynamics were investigated by using voltage ramps to depolarize neurons to −44 mV at different rates after a 500-ms hyperpolarizing prepulse to −104 mV. As can be seen in the example in Fig. 7A, the amplitude of I_T increased and time to peak decreased along with dV/dt between ramp slopes of 400 mV/s and 2,400 mV/s. Bath application of Ni²⁺ depressed the ramp-activated currents (Fig. 7A, gray) in a manner similar to those activated by step depolarizations. Best-fit regression lines calculated from plots of I_T amplitude versus ramp slope for seven cells are shown in Fig. 7B. Both the x-intercept and slope varied widely within our sample. The threshold for I_T activation, as estimated from the x-intercept values, averaged 46 ± 21 mV/s (2–138 mV/s). The sensitivity of I_T to ramp dV/dt (defined as slope of the regression lines, pA·s/mV) varied nearly fivefold (51–242 pA·s/mV) for the cells studied. Taken together, the data suggest that dorsal horn neurons are heterogeneous with respect to activation by voltage ramps. Four additional neurons were examined with a similar protocol ramping the membrane voltage from a conditioning potential of −114 mV to −104 mV to a potential of −44 mV to −34 mV, but with the Ba²⁺ concentration lowered to 2 mM Ba²⁺ and Cd²⁺ increased to 40 μM. The maximum amplitude, threshold, and rate sensitivity of I_T under these circumstances were not significantly different from data obtained with the standard conditions described above.

Fig. 7.

Activation of I_T in dorsal horn neurons by ramp depolarizations. A: I_T in a representative cell was activated by using voltage ramps (bottom) to depolarize the membrane potential to −44 mV from a 500-ms conditioning potential of −104 mV at different rates (indicated to left of each current trace). I_T before (control, black traces) and during (Ni²⁺, gray traces) presentation of 200 μM Ni²⁺ is shown. B: linear regression analysis between I_T amplitude and ramp slope is shown for 7 dorsal horn neurons recorded in media containing 2 or 3 mM Ba²⁺. Individual responses to ramp depolarizations for regression line a are shown in A. Inset: representative linear fit to data points for regression line b. Solid lines denote tonic cells, and dashed lines indicate phasic cells (without regard to burst duration). The coefficients of determination (r²) from regression analyses (0.72–0.99) indicate that I_T amplitude and ramp slope are highly correlated. C: relationship between the amplitude of I_T current components and the ramp slope is described by a linear function for 3 dorsal horn neurons. C1: rate sensitivity of the fast component. C2: rate sensitivity of the slow component. The coefficient of determination (r²) from linear regression analysis indicates that the amplitude of the fast subtype and the slope are highly correlated, but the amplitude of the slow component does not parallel changes in ramp slope.

Despite the cell-to-cell variability, regression analyses showed that 87 ± 4% of the current amplitude (72–99%) was determined by dV/dt. The question then became whether the fast and slow T-channel subtypes contribute to the ability of phasic-firing neurons to signal rapid membrane depolarization. I_T of three dorsal horn neurons possessing both subtypes were activated by ramping the membrane potential from a 500-ms hyperpolarizing prepulse of −104 mV to −44 mV at different rates ranging from 400 to 2,400 mV/s. After isolation of the T-current components, the fast subtype showed rate-dependent activation while the amplitude of the slow component appeared independent of dV/dt over the range of voltage ramps used (Fig. 7, C1 and C2). Linear regression analysis indicated that 90 ± 3% of the fast current amplitude was determined by dV/dt, and the activation threshold estimated from the x-intercepts was 533 ± 83 mV/s, significantly greater than the value observed for the net T-current (46 ± 21 mV/s).

DISCUSSION

The present results supplement earlier studies showing that rodent dorsal horn neurons possess a low-threshold “T-type” Ca²⁺ current (Huang 1989; Russo and Hounsgaard 1996; Ryu and Randic 1990; Walsh et al. 2009). Our new findings point out that, although I_T appears to be prevalent among spinal LI–IV cells, expression is nonuniform, being highest in neurons responding to depolarizing current pulses with a rapidly adapting discharge of one or two action potentials. Here, we show that I_T in the dorsal horn is dependent on the rate of membrane depolarization. In light of our observations that both I_T and action potential discharges of phasic-firing dorsal horn neurons are activated by similar rates of membrane depolarization and are blocked by T-type Ca²⁺ channel antagonists, I_T appears to contribute to selective excitation of these cells by rapid membrane depolarizations (Schneider 2003, 2005).

I_T expressed in hamster dorsal horn resembles a low-voltage-activated, transient Ca²⁺ current reported elsewhere in the CNS. Steady-state activation and inactivation kinetics are broadly similar to T-type Ca²⁺ currents in thalamus (Crunelli et al. 1989), hypothalamus (Niespodziany et al. 1999), suprachiasmatic nucleus (Kim et al. 2005), and cerebellum (Mouginot et al. 1997). As in these studies, our internal pipette solution contained the Ca²⁺ buffer EGTA. EGTA has been reported to slow the time course of recombinant Ca_V3.1 channel activation and inactivation but have no effect on the current density or voltage dependence of steady-state inactivation (Lacinová et al. 2006). Therefore, as with similar studies using a Ca²⁺ buffer in the internal recording solution, we cannot rule out possible effects of EGTA on T-channel function and on neuronal firing properties. We found that Ni²⁺ was an incomplete antagonist of I_T when applied at a concentration of 200 μM. This might be explained by an observation that 100 μM Ni²⁺ blocked only 30–50% of the current through Ca_V3.1 and Ca_V 3.3 T-type currents recorded from cloned channels in oocytes and HEK-293 cells (Lee et al. 1999). We also found that mibefradil potency was lower than what has been reported in other studies (Kim et al. 2005; Lee et al. 2002; McDonough and Bean 1998; Viana et al. 1997). This may be due to the fact that mibefradil is a less effective T-channel antagonist when Ba²⁺ replaces Ca²⁺ as the charge carrier (Martin et al. 2000). As discussed below, incomplete antagonism of I_T by Ni²⁺ and mibefradil may also indicate that other ionic conductances contributed to the inward current in our experiments.

Our experiments were performed on hamster spinal cord at an age when organization and electrical properties of dorsal horn neurons undergo developmental changes in rodents (Beal et al. 1988; Bicknell and Beal 1984; Walsh et al. 2009). For example, it has been reported that A-type potassium current in mouse spinal LI–II neurons decreases during the first 3 wk of life (Walsh et al. 2009). Although the discharge properties of tonic and phasic firing neurons in 9- to 14-day-old hamsters were broadly similar to those recorded from animals at 3–4 wk of age (Schneider 2003), the extent to which dorsal horn T-type Ca²⁺ channel expression changes postnatally and how these changes influence discharge properties is not known. In the present study, we found that nearly all dorsal horn neurons expressed I_T, with no difference between neurons recorded in the superficial and deep laminae. However, Walsh et al. (2009) reported that a T-current was present in only 25% of superficial dorsal horn neurons in mouse at ages comparable to the hamsters used our study. Therefore, it is possible that T-type current expression in dorsal horn neurons may exhibit species, as well as developmental, differences.

Ionic basis of responses to time-varying membrane depolarizations.

The activation of I_T in dorsal horn neurons by rapid ramp depolarizations resembles a property of T-type Ca²⁺ currents described in thalamic neurons (Crunelli et al. 1989), adrenal glomerulosa cells (Várnai et al. 1995), and olfactory receptor neurons (Kawai and Miyachi 2001). In particular, depolarizations exceeding 30 mV/s are required to activate an I_T in cat lateral geniculate neurons (Crunelli et al. 1989). In thalamic cells, voltage ramps of lower velocity appear to reverse the removal of T-channel inactivation by a prior hyperpolarization, so that there are fewer channels available for activation, resulting in diminishing current amplitude at slower membrane depolarizations (Crunelli et al. 1989; Gutierrez et al. 2001). It should be noted that cell-to-cell variability in the threshold for ramp activation for I_T in hamster dorsal horn neurons (46 ± 21 mV/s) is greater than that reported by Crunelli et al. (1989) for thalamic cells (30 ± 2 mV/s). Furthermore, the sensitivity of dorsal horn neurons to rate of depolarization varied considerably between cells (Fig. 7B). These differences could reflect expression of T-channel subunits differing in activation kinetics (Chemin et al. 2002; Talley et al. 1999).

Several observations suggest that I_T contributes to an ability of phasic-firing dorsal horn neurons to selectively signal rapid membrane depolarizations and rapidly moving cutaneous stimuli (Schneider 2003, 2005). First, I_T activated when dV/dt exceeded ∼46 mV/s and increased with the rate of depolarization. Second, phasic-firing cells expressed the highest I_T density in the dorsal horn. Third, Ni²⁺ blocked rate-sensitive activation of I_T under voltage-clamp conditions and also a transient depolarization with associated action potential discharge under current clamp. Although these findings support an involvement of I_T in shaping the ability of certain dorsal horn neurons to signal dynamic membrane depolarizations, the present experiments do not rule out the participation of other ionic mechanisms. For example, we found that the rate of membrane depolarization for activating SB and non-SB type phasic cells was significantly higher than for activating tonic cells. However, I_T was elevated only in SB cells, suggesting that another mechanism may underlie rate sensitivity in non-SB cells (Table 1). Safronov et al. (1997) described a slowly inactivating, TTX-sensitive Na⁺ current accounting for ∼5–20% of total Na⁺ current in rat dorsal horn neurons that broadly resembles a persistent Na⁺ current (I_NaP) activated near −65 mV (Magistretti et al. 2006; Magistretti and Alonso 1999; Parri and Crunelli 1998). Recent studies have also reported that I_NaP in spinal lamina I neurons and ventral horn motoneurons is activated by voltage ramps (Kuo et al. 2006; Prescott and De Koninck 2005). This current has a slower threshold (∼10 mV/s) for ramp activation but could still contribute to neuronal excitation over the range of ramp depolarizations used in the present study.

Expression of multiple T-channel subtypes in dorsal horn neurons.

We used an indirect method in the present study to distinguish subtypes of T-currents in dorsal horn neurons based on differences in kinetics of inactivation. The results suggest that the net T-current is composed of “fast” and “slow” components, having decay time constants similar to two subtypes of low-voltage-activated Ca²⁺ currents reported for thalamic neurons (Shcheglovitov et al. 2005; Zhuravleva et al. 1999, 2001). The current densities for the dorsal horn subtypes (∼3–4 pA/pF) are similar to neurons recorded in thalamic slices (Tarasenko et al. 1997), and the two subtypes appear to be present in approximately equal ratios in the spinal cord and thalamus. Besides kinetics, the two subtypes also appear to have other distinguishing features. First, the activation and inactivation curves for the slow component exhibit a larger overlap region (“window” current) than the fast subtype, similar to the slowly inactivating low-voltage Ca²⁺ current in isolated thalamic neurons (Zhuravleva et al. 2001). Second, the fast subtype was more effectively blocked by Ni²⁺ than the slow subtype, consistent with thalamic neurons recorded in transverse brain slices (Joksovic et al. 2005; Shcheglovitov et al. 2005). Third, and perhaps most interesting from a functional standpoint, we found that the amplitude of the fast current component was correlated to the slope of depolarizing voltage ramps, whereas the slow subtype appeared to be independent of depolarization rate over the testing range. Although rate sensitivity of the T-current subtypes was determined in only a few neurons, it is nonetheless tempting to speculate that the responses of phasic cells in the spinal dorsal horn to rapid membrane depolarizations as reported here and elsewhere (Schneider 2003, 2005) may partly arise from expression of the fast channel subtype. A result that bears mentioning is that the average threshold for activation of the fast current by voltage ramps was >10-fold higher that the activation threshold of I_T. One possible explanation for this difference is that activation of the slow channels during ramp depolarizations might cause partial inactivation of the rate-sensitive fast channels, thereby shifting the activation threshold of the net I_T to slower voltage rates.

Our results showing that a subpopulation of spinal LI–IV neurons express Ca_V3 mRNA are consistent with a previous report of T-type channel transcripts in the spinal cord. The finding that Ca_V3.1 was the most abundant of the three known T-channel subtypes is in agreement with Talley et al. (1999), who reported relatively higher signal for α1G (Ca_V3.1) mRNA than for α1H (Ca_V3.2) and α1I (Ca_V3.3) mRNA in the rat spinal cord. We found that the proportion of dorsal horn neurons expressing Ca_V3 mRNA (44%) was much lower than the proportion of cells exhibiting a macroscopic T-current (>90%). In performing analyses only on samples in which GAPDH mRNA was detected, we hoped to reduce the impact of incompletely harvesting cellular mRNA in the patch pipette. Still, we may not have collected mRNA from dendrites, where some Ca_V3 gene products may function (Shcheglovitov et al. 2005; Tarasenko et al. 1997; Zhuravleva et al. 2001). We also took measures to limit recording time and protect the targeted Ca_V3 transcripts from degradation. However, it is possible that our procedures were not optimal for the detection of small amounts of Ca_V3 mRNA. In fact, other studies using scRT-PCR have also reported that only a fraction of cells with functional neurotransmitter receptors express the corresponding mRNA (e.g., Férézou et al. 2006; Gallopin et al. 2005). It is therefore likely that the proportion of dorsal horn neurons expressing Ca_V3 isotypes is higher than what is indicated by our scRT-PCR results. Nonetheless, our results suggest that Ca_V3.1 is the most common subtype in the hamster dorsal horn and is expressed by neurons having a wide range of firing properties.

Determining the identity of Ca_V3 subunits that underlie kinetically distinct fast and slow T-currents in dorsal horn cells was beyond the scope of the present study. However, comparisons of the properties of endogenous T-current components with studies on recombinant T-type channel expression may offer clues about what molecular subtypes contribute to neuronal excitability in the dorsal horn. First, the decay τ we estimated for the slow current is similar to the inactivation τ reported for recombinant Ca_V3.3 subunits (Chemin et al. 2002; Klöckner et al. 1999; McRory et al. 2001). Second, decay τ of the fast current is in the range reported for inactivation of cloned Ca_V3.1 and Ca_V3.2 subunits (Chemin et al. 2002; McRory et al. 2001). Finally, our isolation of two fast currents having different decay τ also imply contributions from Ca_V3.1 and Ca_V3.2 subunits. These findings suggest that Ca_V3.1, and possibly Ca_V3.2, underlie the fast T-current in dorsal horn cells and that Ca_V3.3 contributes to the slow component, in general agreement with conclusions reached for thalamic neurons (Shcheglovitov et al. 2005).

Functional significance of dorsal horn T-current.

Our results suggest a role for T-type Ca²⁺ channels in spinal sensory function besides spinal nociception (Heinke et al. 2004; Ikeda et al. 2003). I_T is largest in phasic-firing dorsal horn neurons that are selectively activated by a steep voltage trajectory. The dependence of I_T on high rates of membrane depolarization may boost responses to fast excitatory postsynaptic potentials (EPSPs), relative to slow EPSPs, generated by nociceptive and tactile inputs terminating in laminae I–II and III–IV, respectively. Almost all phasic cells in the dorsal horn are interneurons that form connections with neighboring neurons (Schneider 2003, 2008), suggesting that I_T contributes to signaling related to rapid stimulus movement or rate of change within local networks (Schneider 2003, 2005). Other mechanisms may selectively influence responses to sustained or slowly changing inputs. A persistent Na⁺ current (I_NaP) appears to enhance responses of spinal neurons to sustained excitatory inputs (Kuo et al. 2006). I_NaP is strongly expressed in tonic-firing dorsal horn cells and prolongs responses to transient inputs, leading to the proposal that it facilitates encoding of stimulus intensity (Prescott and De Koninck 2005). Thus differential expression of transient and persistent inward currents may contribute to processing of stimulus intensity and velocity information in functionally defined spinal circuits. T-channels may further tune responses to afferent inputs by modulating hyperpolarizing K⁺ currents. Calcium influx through T-channels modulates gating of small-conductance Ca²⁺-activated potassium (SK) channels (Cueni et al. 2009), which are strongly expressed in the dorsal horn (Mongan et al. 2005). Activation of SK channels decreases neuronal excitability by increasing firing adaptation to membrane depolarization (Stoker 2004) and may represent another means for sharpening dorsal horn neuronal responses to transient stimuli. Moreover, calcium entry from T-channels can also regulate the function of low-voltage-activated A-type K⁺ channels (Anderson et al. 2010). A-type K⁺ channels are also widely expressed in dorsal horn neurons (Grudt and Perl 2002; Hu and Gereau 2011; Huang et al. 2005; Walsh et al. 2009), suggesting that I_T may control firing frequency and timing by this mechanism as well. The influence of T-type channels on neuronal firing via modulation of K⁺ channel function would appear to be strongest for SB-type phasic cells, which express the largest I_T. Taken as a whole, we believe our results expand a role for low-voltage, T-type Ca²⁺ channels in spinal sensory processing and suggest that these channels are particularly important for sharpening responses of dorsal horn neurons to afferent sensory stimuli that vary rapidly in the time domain.

GRANTS

This research was supported by National Institute of Neurological Disorders and Stroke Grant NS-25771 (S. P. Schneider).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the author(s).