K_v2.1 and silent K_v subunits underlie the delayed rectifier K⁺ current in cultured small mouse DRG neurons

Abstract

Silent voltage-gated K⁺ (K_v) subunits interact with K_v2 subunits and primarily modulate the voltage dependence of inactivation of these heterotetrameric channels. Both K_v2 and silent K_v subunits are expressed in the mammalian nervous system, but little is known about their expression and function in sensory neurons. This study reports the presence of K_v2.1, K_v2.2, and silent subunit K_v6.1, K_v8.1, K_v9.1, K_v9.2, and K_v9.3 mRNA in mouse dorsal root ganglia (DRG). Immunocytochemistry confirmed the protein expression of K_v2.x and K_v9.x subunits in cultured small DRG neurons. To investigate if K_v2 and silent K_v subunits are underlying the delayed rectifier K⁺ current (I_K) in these neurons, K_v2-mediated currents were isolated by the extracellular application of rStromatoxin-1 (ScTx) or by the intracellular application of K_v2 antibodies. Both ScTx- and anti-K_v2.1-sensitive currents displayed two components in their voltage dependence of inactivation. Together, both components accounted for approximately two-thirds of I_K. A comparison with results obtained in heterologous expression systems suggests that one component reflects homotetrameric K_v2.1 channels, whereas the other component represents heterotetrameric K_v2.1/silent K_v channels. These observations support a physiological role for silent K_v subunits in small DRG neurons.

voltage-gated k⁺ (K_v) channels are tetramers composed of four α-subunits arranged around a central pore. Each α-subunit consists of six transmembrane segments (S1–S6) with cytoplasmic NH₂- and COOH-termini. Members of the K_v1–K_v4 subfamilies generate functional K⁺ channels in a homotetrameric configuration. The diversity within these subfamilies is increased by the formation of heterotetramers and by interactions with auxiliary β-subunits. The K_v2 subfamily contains only two members (K_v2.1 and K_v2.2) that display very similar properties (2), but its diversity is increased by heterotetrameric interactions with α-subunits of the K_v5, K_v6, K_v8, and K_v9 subfamilies. These are designated as silent K_v subunits because they are not able to form functional homotetrameric channels despite sharing the typical topology of a K_v channel. The nonfunctionality is caused by the retention of these subunits in the endoplasmic reticulum (ER) (20, 28, 29, 31). Functionality is restored by forming heterotetrameric channels with K_v2.1 or K_v2.2 channels, thus relieving the ER retention. Therefore, these silent α-subunits act as modulating subunits for K_v2.x currents. The most profound effects consist of a reduction in amplitude, a slowing of inactivation and deactivation kinetics, and an alteration of the voltage dependence of inactivation of K_v2.1 currents (3, 11, 13, 20, 25, 28–30, 37, 42).

K_v channels play an important role in neurons by regulating the resting membrane potential, the repolarization, the frequency of firing, the shape of action potentials, and neurotransmitter release. Differences in spatial and temporal expression and the combination of several voltage-gated α- and β-subunits contribute to the heterogeneity in these neuronal electrophysiological properties. Members of the K_v2 subfamily are widely expressed in neuronal tissues, where they underlie the neuronal outward delayed rectifier K⁺ current (I_K). K_v2 subunits have also been reported in neurons in dorsal root ganglia (DRG) (10, 12), which contain the cell bodies of visceral and somatic sensory neurons. K_v current is composed of one or more fast inactivating A-type currents and several I_Ks depending on the subpopulation of DRG neurons (6–8, 26, 27).

The contribution of K_v2.1 to I_K in small mouse DRG neurons is unknown, and, because silent K_v subunits interact with K_v2 subunits, we also investigated the potential presence of these subunits in cultured small DRG neurons.

MATERIALS AND METHODS

Cell culture and transfection.

DRG neurons were obtained from 12- to 14-day-old mouse (FVB) embryos. Experiments were conducted in agreement with the European Communities Council Directive on the protection of animals used for experimental and other scientific purposes (86/609/EEC) and were approved by the Institutional Animal Care and Use Committee, University of Antwerp. The embryonic spinal cords with the attached DRG were mechanical dissociated. After dissociation, cells were kept in culture as previously described (23) and used for electrophysiological experiments after 4–6 wk in culture.

Mouse Ltk⁻ cells (CCL 1.3, American Type Culture Collection) were cultured in DMEM supplemented with 10% horse serum and 1% penicillin-streptomycin under a 5% CO₂ atmosphere. Ltk⁻ cells were transiently (co-)transfected with 50 ng K_v2.1, 10 μg K_v4.2, or 0.5 μg/5 μg K_v2.1/K_v9.3 channel cDNA in combination with 0.5 μg green fluorescent protein as a transfection marker using Lipofectamine 2000 reagent according to the manufacturer's instructions (Invitrogen, San Diego, CA). Cells were trypsinized 12–24 h after transfection and used for current recordings within 5 h.

RT-PCR analysis and cDNA amplification.

After the dissection of 10 spinal cords, DRG were pinched off from the spinal cords and collected in TRIzol Reagent (Invitrogen) for RNA precipitation. Before the RT-PCR, the RNA sample was made DNA free with deoxyribonuclease I (Fermentas, Burlington, ON, Canada) to exclude genomic DNA contamination. Total RNA (1 μg) was reversed transcribed using random hexamer primers with the Thermoscripts RT-PCR System (Invitrogen) according to the manufacturer's instructions. cDNA amplification was performed for 36 or 44 cycles using gene-specific primers that spanned intron boundaries (except for intronless K_v5.1). Amplified fragments were subcloned in the TOPO-TA vector (Invitrogen) and sequenced to confirm the identity of the PCR products. Each RT-PCR analysis was performed in duplicate on two different RNA isolations.

Confocal imaging.

DRG neurons were cultured on glass-bottom dishes (MatTek, Ashland, MA) as described above. Cell fixation for confocal imaging was performed with 4% paraformaldehyde. Cells were permeabilized by 0.1 M PBS supplemented with 0.1% BSA (Aurion, Wageningen, The Netherlands), 10% horse serum, and 1% Triton X-100 for 30 min at room temperature. After permeabilization, cells were incubated overnight with primary antibodies (1:50) and dissolved in a 0.1 M PBS solution containing 10% horse serum and 0.1% Triton X-100. The following primary antibodies were used: rabbit anti-K_v2.1 (Alomone, Jerusalem, Israel), rabbit anti-K_v2.2 (Alomone), goat anti-K_v9.1 (Santa Cruz Biotechnology, Santa Cruz, CA), goat anti-K_v9.2 (Santa Cruz Biotechnology), and chicken anti-K_v9.3 (Abcam, Cambridge, UK). FITC-labeled anti-rabbit (Jackson ImmunoResearch Laboratories, West Grove, PA), FITC-labeled anti-goat (Abcam), and phycoerythrin (PE)-labeled anti-chicken (Jackson ImmunoResearch Laboratories) antibodies were used as secondary antibodies (1:100) and were dissolved in a 0.1 M PBS solution supplemented with 1% horse serum. After samples had been incubated for 1 h with secondary antibodies, confocal images were obtained on a Zeiss CLSM 510 microscope equipped with an argon laser (excitation: 488 nm) and a helium-neon laser (excitation: 543 nm) for the visualization of FITC- and PE-labeled antibodies, respectively. The immunofluorescence detection of K_v2.x and silent K_v subunits was performed on six different isolations.

Electrophysiology.

DRG neurons with diameters of ∼20 μm were chosen for the neuronal current recordings. Whole cell current recordings were obtained with an Axoclamp-2A amplifier (Axon Instruments, Union City, CA) in the two-electrode voltage-clamp mode. The amplifier was interfaced with a TL1-Labmaster (Axon Instruments). Patch pipettes with a resistance of 3–5 MΩ were pulled from 1.7-mm-diameter glass capillaries with a Brown Flaming P-87 horizontal pipette puller and heat polished. DRG neurons were superfused continuously with an extracellular solution containing (in mM) 140 N-methyl-glucamine, 5 KCl, 1 MgCl₂, 1.8 CaCl₂, 10 glucose, and 5 HEPES with the pH adjusted to 7.4 with HCl. The pipette solution contained (in mM) 30 KCl, 2 MgCl₂, 1 CaCl₂, 110 K-aspartate, 5 Na₂ATP, 2 EGTA, 0.1 cAMP, and 10 HEPES with the pH adjusted to 7.4 with NaOH.

Current recordings of Ltk⁻ cells heterologously expressing K_v subunits were made with the Axopatch-200B amplifier (Axon Instruments) in the whole cell configuration and were lowpass filtered and sampled at 1–10 kHz with a Digidata 1200A data-acquisition system (Axon Instruments). pCLAMP8 software (Axon Instruments) was used to control command voltages and data storage. Patch pipettes were pulled with laser puller P2000 (Sutter Instruments, Novato, CA) from 1.2-mm borosilicate glass (World Precision Instruments, Sarasota, FL) and heat polished. Cells were superfused with an extracellular solution containing (in mM) 145 NaCl, 4 KCl, 1.8 CaCl₂, 1 MgCl₂, 10 HEPES, and 10 glucose with the pH adjusted to 7.35 with NaOH. Pipettes were filled with a solution containing (in mM) 110 KCl, 5 K₄BAPTA, 5 K₂ATP, 1 MgCl₂, and 10 HEPES with the pH adjusted to 7.2 using KOH. The differences in the solutions used for DRG or Ltk⁻ cells were to remain comparable with previous studies using either system (and to prevent Na⁺ current contamination in the DRG recordings). Junction potentials were zeroed with the filled pipette in the bath solution. Experiments were excluded from analysis if the voltage errors originating from the series resistance exceeded 5 mV.

The rStromatoxin-1 (ScTx) experiments were performed by dissolving 100 nM ScTx (Alomone) in the external solution and applying it with a fast perfusion system. Alkaline phosphatase (AP)- and anti-K_v2-sensitive currents were obtained by dissolving 10 U/ml rAPID AP (Roche, Basel, Switzerland) or 10 μg/ml K_v2 antibody (Alomone) in the pipette solution, respectively. The patch pipette was dipped in the normal intracellular solution and then back filled with the AP- or anti-K_v2.1-containing solution. Anti- K_v2.1 block started 2 min after patch disruption, and steady-state block was reached after ∼8 min.

Pulse protocols and data analysis.

The pulse protocols are shown in the figures. The voltage dependence of channel activation or inactivation (activation and inactivation curves) was fitted with the following Boltzmann equation: y = 1/{1 + exp[−(E − V_1/2)/k]}, where E is the applied voltage, V_1/2 is the voltage at which 50% of the channels are activated or inactivated, and k is the slope factor. Results are presented as means ± SE. Statistical analysis was done using Student's t-test. P values of <0.05 were considered to be significantly different.

RESULTS

Presence of mRNA for K_v2.x and several silent K_v subunits in DRG neurons.

The presence of K_v2 and silent K_v subunits in mouse DRG neurons was demonstrated by RT-PCR analysis of RNA from freshly isolated DRG (as described in materials and methods). After 36 cycles of cDNA amplification with subunit-specific primers, K_v2.1 and K_v2.2 were readily observed (Fig. 1A). For each PCR analysis, two negative controls (reactions without cDNA or reverse transcriptase) and a positive control (reaction with cDNA of the K_v subunit under investigation) were performed (Supplemental Fig. 1).1 Silent subunits K_v8.1, K_v9.1, and K_v9.3 were also readily detected after 36 cycles of amplification (Fig. 1A). For the silent K_v subunits that were not detected after 36 cycles, 44 cDNA amplification cycles were performed. In the latter case, subunits K_v5.1, K_v6.1, K_v6.2, K_v6.3, and K_v9.2 were also detected, whereas K_v6.4 and K_v8.2 were not. The resulting PCR products were cloned in the TOPO-TA vector for sequencing to exclude the possibility of nonspecific amplification. This confirmed the presence of K_v2.1 and K_v2.2 subunits and silent subunits K_v6.1, K_v8.1, K_v9.1, K_v9.2, and K_v9.3. The signals obtained with K_v5.1, K_v6.2, and K_v6.3 gene-specific primers (after 44 cycles of cDNA amplification) represented nonspecific amplification (marked by astericks in Fig. 1B) of casein kinase 2, Tmem16H, and tubulin-β, respectively, which are all expressed ubiquitously in neuronal tissue.

Fig. 1.

Expression of K_v2 and silent K_v subunits in dorsal root ganglia (DRG). RT-PCR analysis was performed on freshly isolated mouse DRG. RNA was reverse transcribed with random primers, and the amplification was performed with gene-specific primers for the different K_v subunits. The silent K_v subunits that were not detected after 36 cycles of amplification (A) were subjected to an amplification reaction of 44 cycles (B). The subunits for which the expression was confirmed by sequencing are encircled. * Nonspecific amplification. The different controls are shown in Supplemental Fig. 1.

K_v2.x and K_v9.x subunits are expressed in cultured small DRG neurons.

After the detection of mRNA for K_v2.x and some silent K_v subunits in freshly isolated DRG, the presence of K_v2.x and K_v9.x subunits in cultured small DRG neurons was investigated by immunofluorescence. In our DRG cultures, the DRG neurons were morphological identified and selected based on their soma shape and number of neuronal offshoots. Using subunit-specific antibodies, we detected the presence of K_v2.1 (n = 62), K_v2.2 (n = 45), K_v9.1 (n = 58), K_v9.2 (n = 55), and K_v9.3 (n = 54) subunits (Fig. 2, A and B). The specificity of each antibody was tested by pretreatment of the antibodies with an excess of the appropriate antigens (Fig. 2A and Supplemental Fig. 2, A–D). We could not confirm the expression of K_v6.1 and K_v8.1 protein due to the lack of appropriate selective antibodies. To investigate if K_v2.x and silent K_v channel subunits were coexpressed in the same DRG neurons, we incubated DRG neurons with antibodies against both K_v2.1 and K_v9.x (Fig. 2C and Supplemental Fig. 2E). Since green and red fluorescence were detected in the same neurons, at least K_v2.1 and K_v9.1, K_v9.2, or K_v9.3 subunits were present in the same neurons (n = 19, 14, and 22, respectively).

Fig. 2.

Detection of the members of the K_v2 and K_v9 subfamilies with confocal microscopy in cultured DRG neurons. A: immunofluorescent detection of K_v2.1 by incubation with a K_v2.1-specific antibody followed by a FITC-labeled secondary antibody. Left, light transmission image; middle, FITC fluorescence image; right, overlay of the left and middle images. In the top images, the presence of Kv2.1 is determined, whereas in the bottom images, the specificity of the K_v2.1 antibody is verified by pretreatment with the epitope. B: immunofluorescent detection of K_v2.2, K_v9.1, K_v9.2, and K_v9.3. K_v subunits were detected by incubation with a subunit-specific antibody followed by FITC- or phycoerythrin (PE)-labeled secondary antibodies. Images were obtained as in the overlay image in A, right. The different controls are shown in Supplemental Fig. 2. C: coexpression of K_v2.1 and K_v9.3 subunits. K_v2.1 subunits were detected by incubation with K_v2.1 antibodies followed by a FITC-labeled secondary antibody, whereas K_v9.3 subunits were detected by incubation with a K_v9.3-specific antibody followed by a PE-labeled secondary antibody. FITC and PE fluorescence are shown in the left and middle images, respectively. The right image is an overlay of the left and middle images. Scale bars = 10 μm in A–C. Similar data for K_v9.1 and K_v9.2 are shown in Supplemental Fig. 2E.

The major outward K⁺ current is sensitive to ScTx.

To investigate the potential contribution of K_v2.x subunits (alone and/or in combination with silent subunits) to the I_K of small DRG neurons, we tested for the sensitivity of the observed total I_K to ScTx. ScTx is a potent “blocker” of both K_v2.1 and K_v2.2 (K_d ≈ 13 and 21 nM, respectively) (5). It acts as a gating modifier by strongly shifting the voltage dependence of activation toward depolarized potentials, and it also blocks at least the heterotetrameric K_v2.1/ K_v9.3 channel (K_d ≈ 7 nM) (5).

Typical current recordings of DRG neurons before and after the application of 100 nM ScTx are shown in Fig. 3. Voltage-dependent activation and inactivation of ScTx-sensitive current was examined with the voltage protocols shown in Fig. 3, A and E, respectively. The addition of 100 nM ScTx caused a large reduction of the outward K⁺ current (compare Fig. 3, B and F with A and E, respectively): ScTx blocked 62.0 ± 4.6% of the outward K⁺ current at 0 mV, a potential where 100 nM ScTx blocked K_v2.1 and K_v2.1/K_v9.3 currents almost completely (>90%, data not shown). The ScTx-sensitive current displayed a voltage dependence of activation (V_1/2 = 8.5 ± 1.5 mV, k = 8.0 ± 1.2, n = 5) that was similar to that of K_v2.1 currents, as determined in heterologous expression (V_1/2 = 7.4 ± 1.4 mV, k = 8.5 ± 0.3, n = 5; Fig. 3D).

Fig. 3.

Voltage-dependent activation and inactivation of the rStromatoxin-1 (ScTx)-sensitive current in DRG neurons. A: outward K⁺ current recordings elicited by the voltage protocol shown in the inset. The scale bar applies to A–C. B: representative current tracings elicited by the same voltage protocol as in A after the extracellular application of 100 nM ScTx. C: ScTx-sensitive current, which was obtained by subtracting the current recordings of B from those of A. The red tracings in A–C represent the current recordings at 0 mV. D: activation curve of the ScTx-sensitive current obtained from the normalized tail current amplitude at −40 mV from current recordings as in C. A single Boltzmann function (solid line) was fitted to the data. The dashed line represents the activation curve of wild-type K_v2.1 expressed in Ltk⁻ cells. G/G_max, conductance/maximal conductance; V_m, membrane potential. E: raw current recordings elicited by the voltage protocol shown in the inset to determine the voltage dependence of inactivation in control solution. The scale bar applies to E–G. F: representative current traces (same voltage protocol as in E) after the addition of 100 nM ScTx. G: ScTx-sensitive current. H: voltage dependence of inactivation of the ScTx-sensitive current. The inactivation curve was obtained by plotting the normalized peak currents (I/I_max) recorded during the 250-ms test pulse to 60 mV as a function of the voltage of the 5-s prepulse. Data were fitted with a sum of two Boltzmann functions (solid lines). For comparison, the inactivation curves of K_v2.1, K_v4.2, and K_v2.1/K_v9.3, obtained from heterologous expression, are also shown. The midpoints of inactivation and slope factors are shown in Table 1.

The voltage dependence of inactivation clearly displayed two components and was therefore fitted with a sum of two Boltzmann functions (Fig. 3H). The component of the voltage dependence of inactivation with a midpoint of −4.1 mV compared well with that of the homotetrameric K_v2.1 current in heterologous expression (Fig. 3H and Table 1). The second component represented 29.9 ± 3.7% of the total ScTx-sensitive current and was characterized by a midpoint of inactivation of −62.2 mV (Table 1). At least three possibilities exist for this second component: a population of A-type K_v4.2 channels, heterotetrameric K_v2/silent K_v channels, or differently phosphorylated K_v2 channels. However, the subtraction tracings, as shown in Fig. 3G, gave no indication of an early rapidly inactivating component that would correspond to K_v4.2 channels, excluding the first possibility.

Table 1.

Overview of the biophysical parameters for the voltage dependence of inactivation

	n	First Component		Second Component
V1/2, mV		k	V1/2, mV	k
Control
Kv2.1	5	−18.4±2.8	6.2±0.4	NA	NA
Kv2.1/Kv9.3	6	NA	NA	−47.2±0.7	6.9±0.6
Kv4.2	8	NA	NA	−52.8±1.0	5.2±0.3
ScTx sensitive
ScTx sensitive	10	−4.1±2.0	7.1±0.8	−62.2±1.8	7.8±1.0
AP application
Kv2.1	5	−28.9±2.6	5.7±0.8	NA	NA
Kv2.1/Kv9.3	6	NA	NA	−59.0±1.1	8.5±1.1
DRG
Before	5	−10.1±2.6	5.5±0.8	−47.3±6.4	12.9±3.4
After	5	−22.0±3.2	7.3±1.4	−65.2±4.9	8.3±1.5
Anti-Kv2.1 sensitive
Kv2.1	9	−19.8±2.9	4.2±1.0	NA	NA
Kv2.1/Kv9.3	5	NA	NA	−42.2±1.4	4.7±0.8
DRG	8	−7.0±2.1	6.4±0.6	−62.3±3.1	5.8±1.1

Values are means ± SE; n, number of experiments. The midpoints of inactivation (V_1/2) and slope factors (k) were obtained from a single or double Boltzmann fit (see materials and methods). ScTx, rStromatoxin-1; AP, alkaline phosphatase; DRG, dorsal root ganglia; NA, not applicable. For comparison, K_v9.3 parameters are shown under “second component.”

The ScTx-sensitive current principally represents channels containing K_v2.1 subunits.

In a complementary approach, we used K_v2 antibodies to examine whether the second component of the ScTx-sensitive current originated from channels containing K_v2 subunits. K_v2 antibodies have been shown to block K_v2 currents in hippocampal and neocortical pyramidal neurons (9, 19). The K_v2 antibody-sensitive current was obtained by the subtraction of the current recordings after K_v2 antibodies had diffused into the cell (10–20 min) from those recorded immediately after patch rupture.

The K_v2.1 antibody blocked 30.1 ± 8.0% (during a 500-ms step to 0 mV) of the outward current in cells heterologously expressing K_v2.1, similar to previously obtained results (9, 19). The inactivation curve of the anti-K_v2.1-sensitive current was characterized by a V_1/2 of −19.8 mV, which compares well with the voltage dependence of inactivation in control (Fig. 4A and Table 1). Next, we tested, in this heterologous expression system, whether these K_v2.1 antibodies also blocked K_v2.1 in a heterotetrameric configuration with silent K_v subunits. For this purpose, we used silent subunit K_v9.3 as a model. As shown in Fig. 4B, K_v2.1 antibodies blocked 29.2 ± 4.1% of the K_v2.1/K_v9.3 current. The anti-K_v2.1-sensitive current from channels in the heterotetrameric K_v2.1/K_v9.3 configuration displayed a voltage dependence of inactivation characterized by a V_1/2 of −42.2 mV (Fig. 4A and Table 1). Thus, the voltage dependence of inactivation of the subtracted current was very similar to that of K_v2.1/K_v9.3 current as determined in the control. This indicates that the antibodies are not selective for homo- or heterotetrameric channels but do block heterotetrameric K_v2/silent K_v channels equally well.

Fig. 4.

Voltage dependent activation and inactivation of the anti-K_v2.1-sensitive current. A: voltage-dependent inactivation of the anti-K_v2.1-sensitive current of K_v2.1 or K_v2.1/K_v9.3 heterologous expressed. The inactivation curve was obtained as in Fig. 3H. For comparison, the inactivation curves of K_v2.1 and K_v2.1/K_v9.3 obtained from heterologous expression are also shown. B: outward K⁺ current traces of K_v2.1/K_v9.3 expressed in Ltk⁻ cells before (solid line) and after (dotted lines) diffusion of K_v2.1 antibodies into the cell. The voltage protocol is shown in the inset. C: representative current tracings of the outward K⁺ current of DRG neurons before (solid line) and after (dotted lines) intracellular diffusion of K_v2.1 antibodies. The same voltage protocol as in B was used. D: inactivation curve of the anti-K_v2.1-sensitive current was obtained as in Fig. 3H. Data were fitted with a double Boltzmann function (solid line). The inactivation curves of wild-type K_v2.1 and wild-type K_v2.1/K_v9.3 expressed in Ltk⁻ cells are also shown. The midpoints of inactivation and slope factors are shown in Table 1.

The addition of 10 μg/ml K_v2 antibody to small DRG neurons caused a reduction of the outward K⁺ current (Fig. 4C). To examine whether this reduction was due to K_v2 antibody block or due to run down, we investigated with standard pipette and bath solutions the degree of run down during a period of 20 min. After 20 min, the outward K⁺ current decreased 6.3 ± 0.8% (n = 4). In contrast, diffusion of K_v2.1 antibodies reduced the total outward K⁺ current by 24.2 ± 3.6% (n = 8). This block was comparable with the level of anti-K_v2.1 block in K_v2.1- or K_v2.1/K_v9.3-expressing cells. On the other hand, a similar application of K_v2.2 antibodies into the cells reduced the total outward K⁺ current by 6.5 ± 4.5% (n = 5). Since the latter was similar to run down, this argues against a significant K_v2.2-containing component in I_K of the soma of cultured small DRG neurons. The voltage dependence of inactivation of the anti-K_v2.1-sensitive K⁺ current was again composed of two components (Fig. 4D), and the second component constituted 31.2 ± 5.2% of the total anti-K_v2.1-sensitive current. This compares well with the contribution of the second component to the ScTx-sensitive current. The first and second components of inactivation of anti-K_v2.1-sensitive current were characterized by V_1/2 of −7.0 and −62.3 mV, respectively (Table 1).

Taken together, the biophysical properties of the first and second components of K_v2.1 antibody-sensitive current were comparable with those of ScTx-sensitive current. This suggests that the ScTx-sensitive current originates primarily from a population of homotetrameric K_v2.1 channels and a population of heterotetrameric K_v2/silent K_v channels, suggesting a physiological relevant function of silent K_v subunits.

The ScTx-sensitive current does not represent differentially phosphorylated K_v2.x channels.

Because phosphorylation of K_v2 channels modifies the voltage dependence of inactivation, the second component of the ScTx-sensitive current could derive from a population of channels composed of K_v2 subunits with altered phosphorylation status. Specifically, hypophosphorylation of K_v2.1 at specific residues shifts the voltage dependence of inactivation in the hyperpolarized direction (17, 21). To investigate the possibility that the second component of the ScTx-sensitive current reflected K_v2 subunits in such a hypophosphorylated state, AP was added to the intracellular solution.

As a control, the activity of AP was first tested on homotetrameric K_v2.1 channels expressed in Ltk⁻ cells. The voltage dependence of inactivation was determined immediately after the patch was ruptured and after intracellular diffusion of AP was allowed for 10–20 min. The application of AP shifted the inactivation curve by 10 mV into the hyperpolarized direction, whereas k values did not differ significantly (Fig. 5A and Table 1). Since it is not known if hypophosphorylation might also affect the voltage dependence of inactivation of heterotetrameric K_v2/silent K_v channels, we tested the effect of AP on heterologously expressed K_v2.1/K_v9.3 channels (using K_v9.3 as a model for the group of silent K_v subunits). The results shown in Fig. 5A and Supplemental Fig. 3A demonstrate that the inactivation curve was indeed shifted (∼12 mV; Table 1) in the hyperpolarized direction after AP had diffused into the cell.

Fig. 5.

Effect of alkaline phosphatase (AP) on the voltage dependence of inactivation of the outward K⁺ current of K_v2.1-containing channels and DRG neurons. A: voltage dependence of inactivation of K_v2.1 and K_v2.1/K_v9.3 currents before and after the diffusion of AP into the cell. Data were fitted with a single Boltzmann function (solid lines). The application of AP shifted the voltage dependence of inactivation of both K_v2.1 as K_v2.1/K_v9.3 channels in the hyperpolarized direction to a similar extent. B: voltage dependence of inactivation of DRG outward K⁺ current before and after the diffusion of AP into the cell. The inactivation curve was obtained as in Fig. 3H. Data were fitted with a double Boltzmann function (solid lines). After AP diffusion, the voltage dependence of inactivation of both components was shifted in the hyperpolarized direction to a similar extent.

After the effects of AP had been characterized on heterologously expressed K_v2.1 and K_v2.1/ K_v9.3 channels, the voltage dependence of inactivation of the total outward K⁺ current of DRG neurons was determined before and after AP diffusion into the cell. After equilibration with AP, the outward K⁺ current was still composed of two components. The contribution of the second component to the total outward K⁺ current was not modified by the diffusion of AP into the cell (35.1 ± 3.0% before and 35.6 ± 5.6% after AP, respectively; Fig. 5B). Furthermore, AP caused a significant negative shift of the voltage dependence of inactivation for both components (Table 1 and Supplemental Fig. 3B). Before AP, the first and second component displayed midpoints of −10.1 and −47.3 mV, respectively, whereas after AP had diffused into the cell, V_1/2 of the inactivation curves of the first and second component was −22.0 and −65.2 mV, respectively (Table 1). k values of both inactivation curves were not significantly different compared with those immediately after the patch had been ruptured (Table 1).

These results suggest that the second component of the ScTx-sensitive current does not represent another phosphorylation state of channels composed of K_v2 subunits. Furthermore, the AP-induced shifts in the voltage dependence of inactivation of both components were comparable with those observed for heterologous expressed K_v2.1 and K_v2.1/K_v9.3 channels, respectively. These results exclude the phosphorylation hypothesis and strengthen the idea that both current components correspond to a homotetrameric K_v2 and heterotetrameric K_v2/silent K_v channel population, respectively.

DISCUSSION

K_v2 subunits, especially K_v2.1 subunits, are highly and ubiquitously expressed throughout the mammalian brain (36), but K_v2.x subunits are also present in other parts of the central and peripheral nervous systems (1, 10, 32, 33). It has been demonstrated that K_v2 subunits interact with silent subunits and that the latter are also highly expressed in the mammalian brain (3, 29, 30, 37, 38). The heterotetramerization of K_v2 with silent subunits increases the functional diversity, which may reflect specific functions for these heterotetrameric channel complexes. For example, heterotetrameric K_v2.1 channels containing K_v9.3 subunits are involved in hypoxic pulmonary artery vasoconstriction (22), coassembly of K_v2.1 with K_v5.1 and K_v6 subunits shapes I_K in smooth muscle of the mouse urinary bladder (34), and K_v8.2 subunits are involved in photoreception of the retina (41). In rat DRG neurons, both K_v2.1 and K_v2.2 have been shown to be well expressed (10, 12). However, the presence of silent K_v subunits that could modulate K_v2 properties in these sensory neurons has not been investigated to date. In this study, we confirm the presence of both K_v2 subunits, but the novel finding is that mRNA for several silent subunits (K_v6.1, K_v8.1, K_v9.1, K_v9.2, and K_v9.3) is also expressed in mouse DRG. Furthermore, we confirmed the protein expression of K_v2 and K_v9 subunits and the coexpression of K_v2.1 with K_v9.x in small DRG neurons through immunocytochemistry (Fig. 2 and Supplemental Fig. 2).

The composition of the outward K_v current of DRG neurons depends on the subpopulation. Based on the diameter of their soma, DRG neurons are divided into small (<30 μM), medium (30–50 μM), and large (>50 μM) neurons. In large cutaneous afferent DRG neurons, two transient and one sustained K⁺ current have been characterized (6). Based on the kinetic and pharmacological properties, two transient and one sustained K⁺ current have also been identified in small- and medium-sized neurons (26), whereas another report (8) has suggested the presence of four to six components. These fast transient outward K⁺ currents and I_K probably modulate the excitability of these small-sized DRG neurons, and it has been proposed that the fast transient outward K⁺ current originates from K_v1.4 channels (24, 39). In addition to K_v1.4 channels, K_v3.4, K_v4.2, and K_v4.3 channels could underlie the fast transient outward K⁺ current in small DRG neurons (4, 7, 40). In several neurons, K_v2.1 contributes to I_K, and, in some, it even represents the major outward K⁺ current (9, 14, 19), but the channel subunits that contribute to the I_K of small DRG neurons had not yet been identified.

Using ScTx and Kv2 antibodies, we now report that K_v2.1-containing channels represent a major component of the I_K in small DRG neurons. The application of 100 nM ScTx blocked ∼62% (at 0 mV) of I_K in our small cultured DRG neurons, suggesting that K_v2.1-containing channels represent two-thirds of the I_K in these neurons. Since ScTx is a gating modifier (voltage sensor toxin), “block” by 100 nM ScTx is diminished at voltages positive to +40 mV (5). Thus, the contribution of K_v2 and K_v2/silent K_v subunits to the I_K may even slightly be underestimated.

The comparison of the properties of the anti-K_v2.1-sensitive current with those of the ScTx-sensitive current indicated that both currents consisted of similar components. The voltage dependence of activation of the ScTx-sensitive current was very similar to that of the K_v2.1 currents determined in heterologous expression (Fig. 3D and Table 1). However, based on the activation properties, it is difficult to discriminate between homo- or heterotetrameric channels composed of K_v2.1 and silent K_v subunits because of the limited modulating effects of the silent K_v subunits on the voltage dependence of activation of K_v2.1. In contrast, most silent subunits markedly affect the inactivation. Thus, heterotetrameric K_v2/silent K_v currents can be distinguished from homotetrameric K_v2 currents based on the voltage dependence of inactivation. In most cases, inactivation is slowed, and the voltage dependence is shifted into hyperpolarized direction, by up to 40 mV (3, 11, 13, 20, 25, 28, 29, 37, 42). In our case, both the ScTx- as well as anti-Kv2.1-sensitive current clearly displayed two components in their voltage dependence of inactivation (Figs. 3H and 4D). One component likely represents a population of homotetrameric K_v2.1 channels as judged by its V_1/2 for the voltage dependence of inactivation, which corresponds with that of K_v2.1 in heterologous expression (Figs. 3H and 4D and Table. 1). Based on the negative position of the second component of the inactivation curve of the ScTx- and anti-K_v2.1-sensitive current, we propose that the latter is caused by channels with a heterotetrameric (K_v2/silent K_v subunit) composition. This hypothesis is supported by the fact that silent subunits were detected in DRG neurons and that the anti-K_v2.1-sensitive current still displayed two components. In the case that one of these components was not related to K_v2.1, only one component should have been blocked with K_v2.1 antibodies. To confirm that a silent subunit could be involved, we demonstrated in a heterologous expression system that K_v2.1 antibodies could indeed block heterotetrameric ion channels composed of K_v2.1/silent K_v subunits (Fig. 4, A and B), whereas the epitope (residues 841–857 of mouse K_v2.1) was only present on K_v2.1 subunits.

The phosphorylation state of K_v2.1 apparently varies between different native and heterologous expression cells (for a review, see Ref. 16), and K_v2.1 subunits are highly phosphorylated in mammalian neurons. Since the hypophosphorylation of K_v2.1 shifts the voltage-dependent inactivation in the hyperpolarized direction (17, 21), the ∼10-mV difference between the voltage dependence of inactivation of homotetrameric K_v2.1 channels recorded in heterologous cells and the voltage dependence of inactivation of the first component recorded in DRG neurons (Table 1) is probably due to a different phosphorylation state of K_v2.1 in expression cells and DRG neurons.

Since ScTx also blocks K_v2.2 channels and since the currents of homotetrameric K_v2.1 and K_v2.2 channels and heterotetrameric K_v2.1/K_v2.2 channels do not significantly differ (2), the ScTx-sensitive current could also represent K_v2.2-containing channels. However, it seems unlikely that the ScTx-sensitive current in these neurons represents homotetrameric K_v2.2 or heterotetrameric K_v2.1/K_v2.2 channels since the current reduction with K_v2.2 antibodies was small and similar to the run down current. This implies that K_v2.2 subunits play, at best, a minor role in the I_K of these cultured small DRG neurons. Furthermore, K_v2.2 has a divergent, nonoverlapping distribution compared with that of K_v2.1 (for a review, see Ref. 36).

ScTx (K_d ≈ 1 nM) also blocks K_v4.2 (5), and since K_v4.2 inactivates at negative potentials, the second component of inactivation might also represent a K_v4.2 channel population. Moreover, the presence of K_v4.2 and its contribution to A-type K⁺ current in small rat DRG neurons have been reported (40). However, it seems unlikely that a K_v4.2 channel population reflects the second component of the ScTx-sensitive current because no clear A-type current was noticeable in the ScTx-sensitive current (Fig. 3G) and because the anti-K_v2.1-sensitive current displayed the same two components as the ScTx-sensitive current. This strengthens the idea that the ScTx-sensitive current represents only K_v2.x-containing channels in the cells that we studied.

In the case of neocortical pyramidal neurons of the rat, it has been reported that a ScTx- and anti-K_v2.1-sensitive current displayed a voltage dependence of inactivation with a midpoint of about −62 mV (9). This result is comparable with that of the second population of our ScTx- and anti-K_v2.1-sensitive current. The authors suggested that the difference compared with results of K_v2.1 in a heterologous expression system could be due to the interaction of K_v2.1 with silent K_v subunits or to a different phosphorylation state of K_v2.1 in cultured neurons compared with heterologous expression. The degree of phosphorylation influences the voltage dependent properties of K_v2.1, and different degrees of K_v2.1 phosphorylation have been reported in different cells (15, 17, 18, 21, 35). To exclude the hypothesis that differentially phosphorylated K_v2.1 channels caused the two components of the inactivation curve, we investigated the effect of dephosphorylation. While the application of AP also shifted the voltage dependence of inactivation of heterotetrameric K_v2/silent K_v currents (Fig. 5A and Table 1), it did not alter the relative size of either component in DRG neurons but shifted both equally (Fig. 5B and Table 1).

In conclusion, we propose that the two components observed in both the ScTx- and anti-K_v2.1-sensitive current reflect populations of homotetrameric K_v2.1 and heterotetrameric K_v2.1/silent K_v channels. Furthermore, since K_v4.2 does not noticeably contribute to the ScTx-sensitive current, the K_v2.1-containing channels represent the major component (∼2/3) of the I_K in small mouse DRG neurons. The substantial contribution of the heterotetrameric channels supports an important physiological role of silent K_v subunits in these neurons and raises the question of how these neurons regulate the expression of homo- and heterotetrameric channels in vivo.

GRANTS

This research was funded by a PhD grant of the Institute for the Promotion of Innovation through Science and Technology in Flanders (to E. Bocksteins), by the Interuniversity Attraction Poles Program P6/31 of the Belgian Federal Science Policy Office, and by Grants G.0152.06 and G.0257.08 from the Fonds voor Wetenschappelijk Onderzoek Vlaanderen and GOA50.4426 (concerted action fund of the University of Antwerp).