Neuroscience 178 (2011) 68–81

ELECTROPHYSIOLOGICAL, PHARMACOLOGICAL AND MOLECULAR PROFILE OF THE TRANSIENT OUTWARD RECTIFYING CONDUCTANCE IN RAT SYMPATHETIC PREGANGLIONIC IN VITRO

A. D. WHYMENT,a,b E. CODERRE,c J. M. M. WILSON,c Key words: A-current, KChIP, DPP6, rtPCR, patch clamp, L. P. RENAUD,c E. O’HAREd AND D. SPANSWICKa,b* spinal cord. aWarwick Medical School, University of Warwick, Coventry, CV4 7AL, UK bNeuroSolutions Ltd., Coventry, CV4 7ZS, UK Outward currents through potassium (Kϩ) channels are cNeurosciences Program, Ottawa Hospital Research Institute and Depart- the primary means by which excitable cells oppose mem- ment of Medicine, University of Ottawa, Ottawa, Ontario, K1Y 4E9, Canada brane excitability. Several distinct types of Kϩ-mediated dSchool of Psychology, Queen’s University Belfast, Belfast, Northern conductances have been identified in sympathetic pregan- Ireland, UK glionic neurons (SPN; Polosa et al., 1988; Pickering et al., 1991; Sah and McLachlan, 1995; Miyazaki et al., 1996; Wilson et al., 2002). One such conductance, the transient Abstract—Transient outward rectifying conductances or A-like conductances in sympathetic preganglionic neurons (SPN) are outward rectifying conductance or current (ITR), is a hall- prolonged, lasting for hundreds of milliseconds to seconds and mark of SPN and is used as an identifiable, characteristic are thought to play a key role in the regulation of SPN firing electrophysiological feature of these neurons (Yoshimura frequency. Here, a multidisciplinary electrophysiological, pharma- et al., 1987; Pickering et al., 1991; Inokuchi et al., 1993). cological and molecular single-cell rt-PCR approach was used to This conductance has historically been referred to as the ؉ investigate the kinetics, pharmacological profile and putative K A-like conductance in SPN, based upon its similarity to channel subunits underlying the transient outward rectifying con- A-type Kϩ currents and conductances described in other ductance expressed in SPN. SPN expressed a 4-aminopyridine neurons (see Jerng et al., 2004a for review). The unusual (4-AP) sensitive transient outward rectification with significantly longer decay kinetics than reported for many other central neu- feature of the conductance in SPN is its variable and often rons. The conductance and corresponding current in voltage- prolonged time-course, lasting for hundreds of millisec- clamp conditions was also sensitive to the Kv4.2 and Kv4.3 onds to seconds. ϩ blocker phrixotoxin-2 (1–10 ␮M) and the blocker of rapidly inacti- Transient outward currents are voltage-dependent K vating Kv channels, pandinotoxin-K␣ (50 nM). The conductance currents that possess rapidly activating and inactivating and corresponding current was only weakly sensitive to the Kv1 kinetics. In many neurons, A-type channels are unavail- tityustoxin-K␣ and insensitive to dendrotoxin I able at resting membrane potentials due to pronounced ␮ (200 nM) and the Kv3.4 channel blocker BDS-II (1 M). Single-cell steady-state inactivation and only become available in RT-PCR revealed mRNA expression for the ␣-subunits Kv4.1 and these cells during afterhyperpolarization following action Kv4.3 in the majority and Kv1.5 in less than half of SPN. mRNA for accessory ␤-subunits was detected for Kv␤2 in all SPN with dif- potential discharge, when the membrane potential be- ferential expression of mRNA for KChIP1, Kv␤1 and Kv␤3 and the comes sufficiently negative to remove inactivation. A-type peptidase homologue DPP6. These data together suggest that the channels activate during the decay of the afterhyperpolar- transient outwardly rectifying conductance in SPN is mediated by ization and tend to delay depolarization. In this manner, members of the Kv4 subfamily (Kv4.1 and Kv4.3) in association they are believed to participate in the regulation of firing with the ␤-subunit Kv␤2. Differential expression of the accessory ␤ frequency (Connor and Stevens, 1971; Miyazaki et al., subunits, which may act to modulate channel density and kinetics 1996) and in some neurons, sustain spike in SPN, may underlie the prolonged and variable time-course of virtually alone (Storm, 1987). this conductance in these neurons. © 2011 IBRO. Published by ϩ At the molecular level, voltage-gated K channels (Kv) Elsevier Ltd. All rights reserved. mediate these rapidly inactivating outward currents. Kv *Correspondence to: D. Spanswick, Warwick Medical School, Univer- channels are composed of two classes of subunits; the sity of Warwick, Coventry, CV4 7AL, UK. Tel: ϩ44-2476-574868; fax: ␣ ␤ ϩ44-2476-523701. -subunit, which forms the channel pore and auxiliary ( ) E-mail address: [email protected] (D. Spanswick). subunits. The Kv ␣-subunit is a typical six transmembrane Abbreviations: aCSF, artificial cerebrospinal fluid; BDS-II, Blood de- spanning , with unusually large intracellular N- and pressing substance II; BSA, bovine serum albumin; DPP6, dipeptidyl C-termini loops. These confer control of the activity-depen- aminopeptidase-like protein; DTX-I, Dendrotoxin-I; IML, intermediolat- dent regulation of channel closure (channel inactivation) eral cell column; ITR, transient outward rectifying potassium conduc- tance; KChIP, Kv channel interacting ; Kv, voltage gated po- and are the site of protein–protein interactions (interaction tassium channel; NCS, neuronal calcium sensor; PaTx-2, phrixo- with cytoplasmic ␤-subunits) within the channel. At least 18 -2; PiTx-K␣, pandinotoxin-K␣; scRT-PCR, single-cell reverse transcriptase polymerase chain reaction; SPN, sympathetic pregangli- genes for Kv-channels are expressed in the CNS (Jan and onic ; TiTx-K␣, tityustoxin-K␣; 4-AP, 4-aminopyridine. Jan, 1997). They are grouped into four conserved subfam- 0306-4522/11 $ - see front matter © 2011 IBRO. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.neuroscience.2010.12.061

68 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 69 ilies; Kv1.1–1.8 (Shaker), Kv2.1, 2.2 (Shab), Kv3.1–3.4 chamber continuously perfused with aCSF at a rate of 4–10 ml Ϫ (Shaw) and Kv4.1–4.3 (Shal) (see Gutman et al., 2005). min 1, illuminated from below and viewed under a dissection Four mammalian ␤-subunits, (Kv␤1–4) have also been microscope. The aCSF was of the following composition (mM); NaCl, 127; KCl, 1.9; KH PO , 1.2; CaCl , 2.4; MgCl , 1.3; identified (Scott et al., 1994; Rettig et al., 1994; Heinemann 2 4 2 2 NaHCO , 26; D-glucose, 10; equilibrated with 95% O -5% CO . et al., 1995; Jan and Jan, 1997). Expression studies have 3 2 2 ␣ shown that six -subunits from three Kv subfamilies me- Recordings diate rapidly inactivating outward currents: Kv1.4, Kv3.3, Kv3.4, Kv4.1, Kv4.2 and Kv4.3 (Song et al., 1998; Jerng et Whole cell recordings were performed at room temperature (17– al., 2004a,b). Moreover, Kv1 subfamily channels that alone 21 °C) from neurons in the intermediolateral cell column (IML) with an Axopatch 1 D amplifier (MDS Analytical Technologies, Sunny- yield delayed rectifier currents, inactivate rapidly in the vale, CA, USA), using the blind version of the patch clamp tech- presence of specific ␤-subunits (Rettig et al., 1994; Heine- nique (Pickering et al., 1991). Patch pipettes were pulled from mann et al., 1995) for example, Kv1.5 and Kv␤3(Leicher et thin-walled borosilicate glass and had resistances of between 3 al., 1998). and8M⍀ when filled with intracellular solution of the following Although pore-forming ␣-subunits form rapidly inacti- composition (mM): Kgluconate, 130; KCl, 10; MgCl2, 2; CaCl2,1; EGTA-Na, 1; HEPES, 10; Na ATP, 2 and Lucifer Yellow, 2 (or vating outward currents in heterologous cells, these differ 2 Biocytin, 5); pH adjusted to 7.4 with KOH, osmolarity adjusted to considerably from native currents (An et al., 2000). How- Ϫ 310 mosmol 1 with sucrose. In voltage clamp recordings, tetro- ever, four Kv channel interacting proteins (KChIP 1–4) dotoxin (TTX; 500 nM) and carbenoxolone (100 ␮M) were added have been identified, which when expressed with Kv4 to the extracellular solution to block voltage-dependent sodium subunits reconstitute several features of native channels channels and gap-junctions, respectively. (An et al., 2000; Morohashi et al., 2002). All four KChIP Series resistance compensation of approximately 70–80% proteins are differentially expressed in the CNS (Xiong et was applied for whole-cell voltage-clamp experiments. Access ⍀ al., 2004) with KChIP1, KChIP3 (An et al., 2000) and resistance ranged between 5 and 25 M . Neuronal input resis- tance was measured by injecting small rectangular-wave hyper- KChIP4 (Rhodes et al., 2004) being highly expressed in polarizing current pulses (10–100 pA) and measuring the ampli- neuronal tissues. Other modulators of Kv channel kinetics tude of resulting electrotonic potentials. Recordings were moni- include the neuronal calcium sensor-1 (NCS-1, frequenin; tored on an oscilloscope (Gould 1602, Gould Instrument Systems) Nakamura et al., 2001a), expressed throughout the gray and displayed on a chart recorder (Gould, Easygraf TA240) along matter of the spinal cord (Averill et al., 2004), dipeptidyl with being stored on digital audio tapes (Biologic, DTR-1205) for aminopeptidase-like protein 6 (DPP6) (DPPX, BSPL; Kin subsequent analysis off-line. In addition, data were filtered at 2–5 kHz, (1 kHz for voltage clamp data), digitized at 2–10 kHz (Digi- et al., 2001; Nadal et al., 2003) and DPP10 (DPPY, Jerng data 1322, MDS Analytical Technologies) and stored on PC run- et al., 2004b; Zagha et al., 2005). As such, differential ning pCLAMP 9 data acquisition software. Analysis of electro- expression of these subunits may be a contributing factor physiological data was carried out using Clampfit 9 software (MDS prolonging the time-course of the transient conductance Analytical Technologies). observed in SPN. Thus, in the present study, using a multidisciplinary Cell identification whole-cell electrophysiological, pharmacological and mo- SPN were identified by their characteristic electrophysiological lecular approach, we have characterized further the elec- properties, a long duration (5–10 ms) with a shoul- trophysiological and pharmacological profiles of ITR in der on the repolarization phase, a large amplitude (18–30 mV) SPN. Furthermore, using single-cell RT-PCR we have be- and prolonged action potential afterhyperpolarization, and the gun to identify putative molecular constituents of the expression of inwardly rectifying and transient outwardly rectifying channels forming the substrates underlying this conduc- conductances (Pickering et al., 1991; Wilson et al., 2002). To further confirm recordings from SPN, the neuronal morphology tance and accessory subunits that may contribute to the was also routinely determined retrospectively with Lucifer Yellow prolonged time-course observed in SPN. (dipotassium salt, 1 mg mlϪ1) or Biocytin (5 mM) in the patch pipette solution. Methods for visualizing filled SPN have been EXPERIMENTAL PROCEDURES reported in detail previously, (see Pickering et al., 1991 for Lucifer Yellow and Spanswick et al., 1998 for Biocytin). Slice preparation Cell harvest and single cell RT-PCR Electrophysiological recordings were made from transverse tho- racolumbar spinal cord slices as described previously (Pickering For RT-PCR experiments, thoracic spinal cord sections were cut et al., 1991; Wilson et al., 2002). Protocols were in accordance into 200–300 ␮m thick slices using a Leica VT1000 S tissue slicer. with the UK. Animals (Scientific Procedures) Act (1986) and the Whole cell recordings were performed at room temperature (17– Canadian Council for Animal Care guidelines, and approved by 21 °C) from neurons in the IML with an Axopatch 1 D amplifier the Ottawa Hospital Research Institute Animal Care and Use (MDS Analytical Technologies, Sunnyvale, CA, USA), using the Committee. Care was taken to minimize the number of animals visualized version of the patch clamp technique. A Zeiss Axioskop used and to reduce their suffering. Briefly, Wistar Kyoto rats, aged FS2 microscope (Carl Zeiss Ltd., Welwyn Garden City, UK) fitted 6–14 days (either sex), were terminally anaesthetized using 4% with a 63ϫ water-immersion objective lens together with gradient enflurane in O2, cervically dislocated, decapitated, the spinal cord contrast optics (Luigs and Neumann, Ratingen, Germany) was removed and thoracic sections cut into 300–400 ␮m thick slices used to view slices. Light in the infrared range (Ͼ740 nm) was using a Leica VT1000 S tissue slicer. Slices were maintained in used in conjunction with a contrast-enhancing Newvicon camera artificial cerebrospinal fluid (aCSF) at room temperature for 1 h (Hamamatsu, Hamamatsu City, Japan) to resolve individual neu- after slicing before experimentation was performed. For recording, rons within slices (Stuart et al., 1993). Recording solutions were individual slices were held between two grids in a custom built as above. 70 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

Table 1. Gene-specific primers used for single-cell RT-PCR

Name Gene accession number Primer sequence (ϩ), sense; (Ϫ), antisense Product length (bp)

Kv 1.4 NM_012971 (ϩ)5=-GCACCAATCAGTTCAATGAC-3= 111 (Ϫ)5=-CAGGAATCGTGTGCTGTTAT-3= Kv 1.5 NM_145983 (ϩ)5=-GGCAGACTGGCACCAGTGAA-3= 161 (Ϫ)5=-ACGGCTCCTGCGTCATGGAA-3= Kv 3.4 X62841 (ϩ)5=-CAGAGGATGCGTGCAATAGC-3= 144 (Ϫ)5=-CTAGGCGTGCCTCAGACACA-3= Kv 4.1 U89873 (ϩ)5=-GCTGCTGTCATACCAATGTG-3= 168 (Ϫ)5=-GCTCTTGCTGGATGTTAGTC-3= Kv 4.2 NM_031730 (ϩ)5=-AATTGGCACAGGTCTGAGAG-3= 111 (Ϫ)5=-CCTTCTTGCTTGGCTGTCTT-3= Kv 4.3 NM_031739 (ϩ)5=-CTAGACCTGGCCTCAAGATT-3= 129 (Ϫ)5=-CTACTTGGCCTACCTGCTAA-3= Kv␤1 X70662 (ϩ)5=-CTCGCACCGCTCTGCTCATT-3= 175 (Ϫ)5=-GGCCAGCCTGGCTTCTATTG-3= Kv␤2 X76724 (ϩ)5=-GAGCCACTTAAGCTGCATGA-3= 143 (Ϫ)5=-GTGGTAGCAACAGCAGAGAC-3= Kv␤3 NM_031652 (ϩ)5=-CCGGCTTCGAGTAGCGATTC-3= 157 (Ϫ)5=-ATGAGCAGAGCGGTGCGAGA-3= KChIP1 NM_014592 (ϩ)5=-GGAGAGGCAGCAACTAACAT-3= 136 (Ϫ)5=-TGATCGTAGGTGCTAGTGTG-3= KChIP2 AB040031 (ϩ)5=-GCAGGAGACCTTCTGTCTAA-3= 163 (Ϫ)5=-GGCAAGAGTAGGCAAGCTAA-3= KChIP3 AB043892 (ϩ)5=-CTGAGAGTCCATGCGTGTTC-3= 250 (Ϫ)5=-GCCTTGGCTTCTCACAGGAT-3= NCS-1 NM_024366 (ϩ)5=-GTTGTTAACGAGGCCTGCTA-3= 187 (Ϫ)5=-CGGAGCACAGTATCTGAGAA-3= DPP6 NW047685 (ϩ)5=-ACAGGTCCATCATCGGATTC-3= 107 (Ϫ)5=-AGGTCTCAAGCGTGGTTAGT-3=

The neuronal harvest and single-cell RT-PCR from SPN was (72 °C). After the final cycle the reaction was held for 5 min at carried out as described in detail previously (van den Top et al., 72 °C. The PCR products were then separated on an Ethidium 2004). Briefly, the cytoplasm of a SPN was gently aspirated Bromide stained 2% agarose gel and photographed. Direct se- under visual control into a patch-clamp recording electrode. quencing was performed to confirm the identity of the amplified The contents of the electrode were subsequently expelled into products. All gene-specific primers are listed in Table 1. Only cells a microtube and reverse transcribed in a reaction volume of 10 expressing the housekeeping gene, ␤-actin were used in the study ␮l containing; 1ϫ first-strand buffer, 0.1 M DTT, 10 mM dNTP, and a negative control (the contents of an electrode which was 1.5 U RNasin (Promega, Southampton, UK), 200 U Superscript placed next to a cell without seal formation or harvesting of II reverse transcriptase (Invitrogen, Paisley, UK) and 0.5 ng cytoplasmic contents) was performed which showed negative ex- reverse transcription primer for 60 min at 42 °C. Three prime pression for all mRNA transcripts tested. end amplification (TPEA) was performed using methods de- scribed in detail previously (Dixon et al., 1998). Briefly, the RT Statistical analyses and curve fitting primer was composed of an anchored oligo(dT) primer with a specific 5= heel sequence : 5=-GACTGCCAGACCGCGCGCCTGA- All values given are meansϮSEM. Statistical significance was CGCGTAATACGACTCACTATAGGGTTTTTTTTTTTTTTTTTTTT- determined using student’s two-tailed t-tests, paired or unpaired 3=. Second-strand cDNA synthesis was initiated by incubation of as appropriate. PϽ0.05 was taken to indicate statistical signifi- the first-strand cDNA with 1 ng of a primer consisting of 5=-AAAA- cance. Steady-state inactivation and activation data were fit after CTGCCAGACCGCGCGCCTGAACGCGTCGTATTAACCCTCA- normalization to a single Boltzmann function: CTAAAGGGNNNNNNNNNNNNNNN-3= (where N represents C, ϭ ϩ ͓͑ Ϫ ͒ ͔Ϫ1 G, T or A) during PCR amplification for 29 cycles, 10 s annealing G ⁄ Gmax 1 exp V Vh ⁄ k (50 °C), 2.5 min extension (72 °C), 1 min denaturing (94 °C). After (where Gmax is the maximal conductance, Vh is the half-inactiva- the initial round further amplification was performed by the addi- tion potential and k the slope factor) by non-linear regression tion of 230 ng heel primer consisting 5=-ACTGCCAGAC- using Microcal Origin 7. The decay of the current was fitted by the CGCGCGCCTGA-3= Samples of amplified cDNA were diluted following equation: 1:10 and subjected to hot-lid PCR carried out in a total reaction volume of 25 ␮l. Reaction components were as follows; 2.5 ␮l I͑t͒ ϭ ͕IS͑exp͑Ϫt ⁄ ␶S͖͒͒ ϩ ͕If ͑exp͑t ⁄ ␶f͖͒͒ ϩ I0 10ϫ PCR buffer, 1 ␮l25␮M 5’primer, 1 ␮l25␮M 3’primer, 1 ␮l ␮ ␮ 25 mM MgCl2, 0.5 l 10 mM dNTP, 12.5 l 2.6 mM betaine/2.6% (where I(t) is the total current at time t, Is and If are the initial ␮ ␮ ␶ DMSO, 4.25 lH20, 0.25 l platinum Taq polymerase (Invitrogen, amplitudes of the two components relating to the slow ( s) and fast ␶ Paisley, UK). Amplifications were carried out on PTC-225 thermal ( f) decay time constants, and the non-inactivating residual current cyclers (Tetrad, MJ Research, MA, USA). Following an initial 4 (Io)). Current-clamp time constants were fit with a single exponen- min denaturing step (95 °C), each PCR cycle consisted of 30 s tial function of the first membrane-voltage response to hyperpo- denaturing (94 °C), 30 s annealing (60 °C) and 20 s extension larizing current pulses. A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 71

Fig. 1. Transient outward rectification in SPN. (A) Superimposed membrane responses (a) and a single membrane response in SPN (b) illustrating as a delayed return to resting membrane potential following hyperpolarizing current injection. (B) Superimposed (ء) transient outward rectification membrane responses to current injection showing transient outward rectification observed as a delay in reaching threshold for firing following depolarizing current injection from a relatively negative resting membrane potential (–68 mV). (C) Superimposed samples of a continuous whole-cell voltage clamp recording in a SPN, at a holding potential of –60 mV, showing currents evoked in response to depolarizing and hyperpolarizing voltage D) Family of currents evoked) .(ء) steps (lower panel). Note the transient outward current evoked in response to the largest depolarizing voltage step in response to depolarizing and hyperpolarizing step commands from a holding potential of –40 mV in a SPN. Note the transient outwardly rectifying evoked at the offset of the responses to hyperpolarizing voltage step commands (lower panel). Voltage clamp records are in the (ء) tail currents presence of TTX (500 nM) and carbenoxolone (100 ␮M) to block voltage-dependent sodium channels and electrical synapses, respectively.

Drugs was observed in 98% (75 of 77) of recorded SPN. This conductance was observed as a delay in the return to rest The following drugs were used at the final concentrations indi- ϳ cated: 4-aminopyridine (4-AP, 0.1–5 mM); BDS-II, 1 ␮M; dendro- of membrane response greater than 10 mV negative to toxin-I (DTX-I, 200 nM); pandinotoxin-K␣ (PiTx-K␣, 50 nM); tity- the RMP following hyperpolarizing current pulse injection ustoxin-K␣ (TiTx-K␣, 100 nM), all from Sigma and phrixotoxin-2 (Ϫ20 to Ϫ200 pA, 800 ms duration, 0.1 Hz; Fig. 1Aa, Ab). (PaTx-2, 1–10 ␮M); TTX (500 nM) from Alomone Laboratories, Alternatively the conductance was observed as a delay in Israel. The concentrations of employed here were based the membrane response to depolarizing current pulses upon previously published work. Drugs were prepared as a stock solution using distilled water (10–300 pA, 800 ms duration, 0.1 Hz) from hyperpolarized and diluted to the required concentration in aCSF immediately membrane potentials (Ͼ–65 mV, Fig. 1B), manifesting as prior to use with the exception of 4-AP which was made up as a a delay in reaching threshold for action potential firing (Fig. 100 mM stock in DMSO. All peptide toxins were prepared as a 1B). The mean time-constant of decay of the conductance stock solution using distilled water and 0.1% bovine serum albu- Ϯ min (BSA). A further 0.1% BSA was added to the bathing solution in current clamp recordings was 1983 478 ms (range immediately prior to use. Final BSA concentrations did not exceed 998–4145 ms, nϭ24) which is an order of magnitude 0.2% and appropriate vehicle controls were performed. The drugs longer than the membrane time-constant (104Ϯ36 ms were administered to the slice by perfusion from 50 ml syringes range 35 to 380 ms, nϭ48) suggesting activation of an arranged in line with the main aCSF reservoir by a series of active conductance. In voltage-clamp, current-voltage re- three-way valves. blockers were applied for at least 10 min prior to the addition of agonists to ensure complete equil- lations generated with a range of voltage step commands ibration within the recording chamber. from a holding potential of Ϫ60 mV (20 mV increments, Ϫ120 to Ϫ20 mV, 800 ms duration, 0.1 Hz) revealed a RESULTS large transient outwardly rectifying current in response to depolarizing voltage step commands (Fig. 1C). The tran- Whole-cell recordings from 77 neurons, identified as SPN, sient outwardly rectifying current had a mean peak ampli- were included in this study. The neurons had a mean tude of 256Ϯ35 pA (measured as the absolute current; Ϫ Ϯ resting membrane potential of 54.7 6.2 mV and a mean nϭ38; Fig. 1C). From a more depolarized holding potential Ϯ ⍀ resting input resistance of 456 49 M . (Ϫ40 mV), the transient outward current was observed as a tail current at the offset of the response to hyperpolariz- Identification of the transient outward rectifier ing voltage steps (Ϫ100 to Ϫ60 mV, 800 ms, 0.1 Hz, Fig. In accord with others (Pickering et al., 1991; Wilson et al., 1D) and had a mean peak amplitude of 157Ϯ26 pA ϭ 2002), a transient outwardly rectifying conductance (ITR) (n 24). 72 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

Kinetics of the transient outward rectifier observed as a delayed return to rest of the membrane response to hyperpolarizing current pulse injection, was The steady-state inactivation of the rectifying conductance also investigated in current-clamp recording conditions was investigated by applying 5 mV depolarizing voltage (Fig. 4, nϭ7). Similar to the results obtained in voltage- steps to Ϫ20 mV applied from a range of holding potentials clamp and in accord with previous data (Pickering et al., (Ϫ100 to Ϫ20 mV, 1.2 s, Fig. 2Aa). Fitting the normalized 1991; Wilson et al., 2002), 4-AP blocked I in a concen- data to a single Boltzmann function revealed a mean half- TR tration-dependent manner. maximal voltage of Ϫ67.2Ϯ3.7 mV and a mean slope of Ϫ1 6.1Ϯ1.0 mV (nϭ13, Fig. 2Ac). Furthermore, the curve Sensitivity of the transient outward rectifier to toxins revealed that current inactivation was observed at mem- ϩ brane potentials more positive than Ϫ75 mV and was In order to evaluate the K channel ␣-subunits mediating the complete at Ϫ35 mV. Thus, the current did not inactivate SPN ITR, a series of experiments were performed using pep- completely at resting membrane potential. The voltage- tide toxins specific for either channel subunits or subfamilies dependent activation of the conductance was evaluated by of subunits. Initially, the sensitivity of the current to phrixo- applying defined depolarizing voltage steps (20 mV incre- toxin-2 (PaTx-2), a highly specific blocker of Kv4.2 and Kv4.3 ments, Ϫ80 to 0 mV, 500 ms) from a holding potential of channels was investigated (Diochot et al., 1999). PaTx-2 Ϫ100 mV (Fig. 2Ab). The activation of the current was inhibited the mean peak transient outwardly rectifying current likewise well fitted with a single Boltzmann function with a in a concentration-dependent manner. Bath application of 1 ␮ Ϯ mean half-activation potential of Ϫ41.7Ϯ5.7 mV and a M PaTx-2, induced a 4.2 0.3% inhibition of the current Ϫ ϭ mean slope of Ϫ7.4Ϯ1.3 mV 1 (nϭ15, Fig. 2Ac). A small within 5 min of application (n 6). Subsequent exposure of window current exists for this conductance near the ob- the slice to progressively higher concentrations of PaTx-2 Ϯ ϭ Ͻ Ϯ ϭ served range of resting membrane potentials (Fig. 2Ac, induced a 30.9 6.2% (n 6, P 0.01) and 72.3 5.6% (n 5, Ͻ ␮ shaded area). The conductance exhibited a mean time to P 0.01) inhibition of the current in the presence of 5 M and ␮ peak ranging from 26.2Ϯ2.3 ms at Ϫ40 mV to Ϫ17.4Ϯ0.3 10 M PaTx-2, respectively (Fig. 5A, B). Higher concentra- msat0mV(nϭ17, Fig. 2Ad). tions of this toxin were without further effect and the effects of The time-course of recovery from steady-state inacti- the toxin could not be reversed upon wash. PaTx-2 inhibited ␮ vation was also determined by first fully inactivating the the current with an IC50 of approximately 5 M(Fig. 5C, ϭ conductance at 0 mV, then applying a hyperpolarizing n 5), suggesting the expression of and mediation by Kv4.2 voltage pre-pulse to Ϫ90 mV of varying duration (5–100 and/or Kv4.3 channels. ␣ ␣ ms) before stepping back to 0 mV (nϭ8, Fig. 2Ba). The Next, the actions of pandinotoxin-K (PiTx-K ), a scor- normalized mean peak currents were plotted as a function pion toxin which selectively binds rapidly inactivating Kv ␣ of the pre-pulse duration. The conductance recovered from channels (Klenk et al., 2000) was investigated. PiTx-K inactivation with a mono-exponential time-course and time (50 nM) reduced the mean peak transient outward current Ϯ Ϯ ϭ constant of 19.3Ϯ2.5 ms (nϭ8, Fig. 2Bb). The decay of the from a control value of 255 41 pA to 196 32 pA (n 4, Ͻ conductance was fit with a single exponential function and P 0.05, Fig. 6A) in the presence of the toxin within 5 min Ϯ had a time constant ranging from 113.9Ϯ8.0 ms at Ϫ60 of application, a 23.2 6.1% reduction of the current. The mV to 121.1Ϯ3.2 ms at Ϫ20 mV (nϭ17, Fig. 2Cb). effects of the toxin could not be reversed upon wash. Tityustoxin-K␣ (TsTx-K␣), another toxin which selectively blocks delayed rectifier voltage-gated, non-in- Sensitivity of the transient outward rectifier to ϩ 4-aminopyridine activating K currents (Kv1 channels) was also tested on the transient outward current. TsTx-K␣ (100 nM), reduced One of the characteristic features of A-type currents is their the mean peak current from a control level of 320Ϯ40 pA sensitivity to 4-AP, which blocks both Kv1 and Kv4 chan- to 250Ϯ34 pA in the presence of the toxin (nϭ5, Fig. 6B), nels with high affinity (see Coetzee et al., 1999). Thus, the although the effects of this toxin were not statistically sig- sensitivity of the transient outward current to 4-AP (100 ␮M nificant. to 5 mM) was investigated. Bath application of 100 ␮M BDS-II (1 ␮M), which specifically blocks Kv3.4 chan- 4-AP, blocked 7.1Ϯ2.1% of the control transient outward nels (Diochot et al., 1998) was without significant effect on current, reducing the control current from a mean peak the transient outward rectifying current (nϭ4). Likewise, amplitude of 279Ϯ23 pA to 260Ϯ19 pA (nϭ10, Fig. 3A). dendrotoxin-I (DnTx-I), a blocker of Kv1 channel subunits Subsequent application of 1 mM 4-AP blocked 35.8Ϯ4.6% (Katoh et al., 2000) was without significant effect on the of the outward current, reducing the mean peak amplitude peak current, (nϭ4). to 181Ϯ12 pA (nϭ10, PϽ0.01) and revealed a rebound These pharmacological data suggest the substrate depolarization at the offset of the membrane response to mediating the transient outward rectifying current ex- hyperpolarization (Fig. 3B). In the presence of 5 mM 4-AP pressed in SPN is composed of Kv4.2 and, or Kv4.3 chan- the transient outward current was reduced to 92Ϯ13 pA nels, but is unlikely to involve Kv3.4 channels. It would also (nϭ10, PϽ0.01), a 77.0Ϯ7.1% reduction compared to con- seem unlikely from this data that either Kv1.4 or Kv1.5 trol (Fig. 3A, B). Greater concentrations of 4-AP were channels contribute to the conductance, although they without further effect, thus the maximal block of the current cannot be conclusively ruled out. Similarly, the results ϳ ␣ was at a concentration of 5 mM (IC50 1 mM, Fig. 3C). The obtained from PiTx-K suggest a contribution of Kv4.1 sensitivity of the transient outward rectification to 4-AP, cannot be ruled out. Therefore to further clarify the nature A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 73

Fig. 2. Kinetics of the transient outward rectification in SPN. (A) Activation and steady-state inactivation of the transient outward rectifying current is voltage-dependent. (a) Steady-state inactivation was investigated with a voltage step command to –40 mV applied from holding potentials of –100 to –20 mV. Dashed line represents 0 mV. (b) Voltage-dependent activation was investigated by applying defined depolarizing voltage steps from a holding potential of –100 mV. (c) Plot of the activation (closed circles) and steady-state inactivation (open squares) of the transient outward rectifier expressed as normalized peak current (I/Imax) as a function of the conditioning potential and fitted with a Boltzmann function. Dashed lines mark the voltages of half-activation and inactivation. Note the shaded window current. (d) Plot of the time to peak of the transient current as a function of the test potential. (B) Recovery from inactivation of the transient current followed a mono-exponential time-constant. (a) Recovery from inactivation was assessed by first inactivating the transient current completely at 0 mV, then applying a hyperpolarizing voltage pre-pulse to –90 mV of varying duration

(2–100 ms) before stepping back to 0 mV. (b) Normalized peak currents (I/Imax) were plotted as a function of recovery interval. (C) Plot of the decay time course of the transient current. (a) Same trace as shown in (Aa) with superimposed single exponential best fitting curves. (b) Decay ␶ plotted as a function of the conditioning potential. All recordings are in the presence of TTX (500 nM) and carbenoxolone (100 ␮M). 74 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

Fig. 3. The transient outward rectification is sensitive to 4-AP. (A) Superimposed samples of a continuous whole-cell voltage-clamp recording from a SPN at a holding potential of Ϫ60 mV showing current responses to a series of depolarizing and hyperpolarizing rectangular-wave voltage-steps (lower panel), and the effects of 4-AP on the transient outward rectification. Transient outward rectification (a) was progressively reduced with increasing concentration of 4-AP (b, c, d). (B) Superimposed plots from records shown in (A) illustrating the concentration-dependent effect of 4-AP on the transient outward rectifier. (C) Bar chart illustrating the concentration-dependent inhibition of the peak transient outward current by 4-AP. Error bars represent SEM and are the mean values generated from 10 cells. of the candidates mediating the molecular composition of attempting to correlate these properties with target projec- these ion channels, single-cell RT-PCR experiments were tion of the SPN. employed. Single cell RT-PCR revealed ␣-subunit mRNA expres- sion for Kv4.1 and Kv4.3 in the majority of SPN tested ؉ Single-cell RT-PCR analysis of K channel subunit (86%, nϭ6/7). In less than half of SPN tested mRNA expression in SPN expression for Kv1.5 (43%, nϭ3/7) was also detected. Of ϩ ␤ Five K channel subunits are known to form homomeric the accessory subunits probed, mRNA expression for channels that have A-type-like potassium conductance Kv␤2 was detected in all SPN. Furthermore, mRNA for the properties, and in addition, Kv1 channel members that ␤-subunits KChIP1 (43%, nϭ3/7), Kv␤1 (14%, nϭ1/7) and normally yield delayed rectifier-type currents inactivate Kv␤3 (29%, nϭ2/7) was observed differentially expressed rapidly in the presence of ancillary ␤-subunits. Thus, to by SPN in addition to the peptidase homologue DPP6 generate as comprehensive a profile as possible of A-type (29%, nϭ2/7; Fig. 7A, B). No mRNA expression was de- ion channel subunit mRNA expression, SPN were simul- tected for Kv1.4, Kv3.4. Kv4.2, KChIP2, KChIP3 or the taneously analyzed for the expression of Kv1.4, Kv1.5, calcium binding protein NCS-1. Kv3.4, Kv4.1, Kv4.2 and Kv4.3 ␣-subunits, KChIP1, ␤ ␤ ␤ ␤ KChIP2, KChIP3, Kv 1, Kv 2 and Kv 3 -subunits, DPP6 DISCUSSION and NCS-1 mRNAs. In the present study we have made no attempt to correlate mRNA expression with the electro- Data presented here demonstrate, for the first time, a physiological properties of the conductance, although sub- detailed pharmacological and putative molecular profile of sequent studies will investigate this issue in addition to the Kϩ channels comprising the transient outward rectify- A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 75

Fig. 4. 4-AP inhibits the transient outwardly rectifying conductance. (A) Superimposed whole-cell current-clamp recordings showing membrane potential responses of a SPN, to a series of depolarizing and hyperpolarizing rectangular-wave current pulses, showing the concentration-dependent inhibition of the transient outward rectifying conductance, observed as the delayed return to rest of membrane responses to hyperpolarizing current injection, by 4-AP. Experiments were performed in the presence of CsCl2 (1 mM) to reduce anomalous inward rectification and TTX (500 nM) to eliminate indirect effects elicited by 4-AP. (B) Samples from same cell as (A) shown on a higher gain. Bottom superimposed traces show the progressive inhibition of the transient outward rectifier with increasing concentrations of 4-AP. ing conductance expressed in SPN. The major findings of density and kinetics, thereby underpinning the prolonged this study are that SPN express an unusual transient out- and variable time-course of this conductance in SPN. wardly rectifying conductance with significantly longer de- cay kinetics than reported for many other central neurons. Fundamental properties of the transient outward The conductance is similar in activation and inactivation rectifier in SPN kinetics to the A-type-like conductance described previ- ously in SPN (Dembowsky et al., 1986; Yoshimura et al., In agreement with other recent studies (Miyazaki et al., 1996; 1986; Pickering et al., 1991; Inokuchi et al., 1993; Miyazaki Wilson et al., 2002), it was found that 98% of neurons ex- et al., 1996; Wilson et al., 2002). Pharmacological evi- pressed a 4-AP-sensitive transient outwardly rectifying con- dence using selective -targeting toxins ductance, with activation and inactivation kinetics similar to and single-cell RT-PCR suggests that the transient out- those reported previously for SPN (Sah and McLachlan, ward current in SPN is attributable to the Kv4 subfamily 1995; Miyazaki et al., 1996). However, a mono-exponential members, specifically Kv4.1 and Kv4.3. Whereas the most time-course of recovery from steady-state inactivation was prominent accessory ␤ subunit expressed by SPN was recorded in the present study while others report a double Kv␤2, the differential expression of other accessory ␤ sub- exponential time course (Bordey et al., 1995). Bordey et al. units including KChIP1, Kv␤1 and Kv␤3, and the peptidase (1995) speculated that this reflected accumulation of calcium homologue DPP6, may act to modulate channel gating, in dendrites after repetitive depolarizations. Also, in adrenal 76 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

Fig. 5. Phrixotoxin-2 blocks the transient outward rectifying conductance in SPN. (A) Current-voltage relations of a SPN evoked from a holding potential of Ϫ60 mV showing the sensitivity of the transient outward rectification to increasing concentrations of the Kv4.2, Kv4.3-specific blocker Phrixotoxin-2 (a, control; b, 1 ␮M; c, 5 ␮M; d, 10 ␮M). (B) Superimposed records from cell shown in (A) illustrating the concentration-dependent effect of Phrixotoxin-2 on the transient outward rectifier. (C) Bar chart illustrating the concentration-dependent inhibition of the peak transient outward current by Phrixotoxin-2. Error bars represent SEM. medulla-innervating SPN, the transient outward rectification Lien et al., 2002), dopaminergic substantia nigra neurons could be pharmacologically split into two distinct Kϩ conduc- (ϳ25 ms, Liss et al., 2001), each of which are mediated by tances (Wilson et al., 2002). These included a fast compo- Kv4.3 channels (in combination with KChIP3 subunits in nent that was sensitive to 4-AP and a slower 4-AP-insensitive the substantia nigra) and hamster suprachiasmatic nu- component. The slower component was reminiscent of a cleus neurons (ϳ8 ms, Alvado and Allen, 2008), mediated D-type conductance, a conductance that inactivates slowly by Kv4.1 and Kv4.2 channels (Itri et al., 2010). below spike threshold over a time-course of seconds and The mean resting membrane potential of the SPN re- gives rise to a prolonged delay in the onset of action potential corded in this study and in others falls into a window for the firing (Storm, 1987). Very few neurons recorded in this study transient conductance. At these potentials, there is fractional expressed a D-like conductance and the nature of this 4-AP activation and incomplete inactivation of the conductance. insensitive component requires further investigation. A re- This is interesting as small variations of the resting mem- bound depolarization was uncovered upon block of the tran- brane potential will most likely have a considerable impact on sient outward rectifier conductance (by Csϩ and 4-AP) in properties of the neuron modulated by this conductance, these neurons. This is in accord with previous studies that among them action potential width, firing-frequency and firing attribute the rebound to a low voltage activated T-type cal- patterns (Rogawski, 1985). In relation to this, in SPN inner- cium conductance (Wilson et al., 2002). vating the adrenal gland, the transient outward rectification The hallmark feature of the transient outwardly rectify- has been suggested to play an important role both in regu- ing conductance in SPN is the significantly longer decay lating spike repolarization and contributing to the afterhyper- kinetics than reported for many other central neurons. The polarization (AHP) (Wilson et al., 2002). Furthermore, excit- decay time constant of the conductance at Ϫ20 mV was atory postsynaptic potentials (EPSPs), in particular large- 121 ms in SPN, significantly longer than the decay time amplitude EPSPs, are shortened in neurons expressing constant of the transient outward rectifying conductance A-currents, which may be important for temporal summation observed in hippocampal GABAergic interneurons (15 ms, of synaptic inputs (Cassell and McLachlan, 1987). As a gen- A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 77

Fig. 6. Effects of Pandinotoxin-K␣ and Tittyustoxin-K␣ on the transient outward rectification. (Aa) Current-voltage relations of a SPN evoked from a holding potential of Ϫ60 mV showing the sensitivity of the transient outward rectification to Pandinotoxin-K␣ (50 nM), a blocker of rapidly inactivating Kv channels. (b) Superimposed records of data shown in (Aa). (c) Pooled data showing the effects of Pandinotoxin-K␣ on the transient outward rectification in SPN. (Ba) Current-voltage relations of a SPN evoked from a holding potential of Ϫ60 mV showing the weak sensitivity of the transient outward rectification to Tityustoxin-K␣ (100 nM), a blocker of delayed rectifier Kv1 channels. (b) Superimposed records of data shown in (a). (c) Pooled data showing the effects of Tityustoxin-K␣ on the transient outward rectification in SPN. eral rule, the transient outward rectifier reduces the excitabil- Molecular and pharmacological characterization of ity of neurons (Banks et al., 1996), in particular when the the ion channel subunits underlying the transient membrane is already hyperpolarized by other factors such as outward conductance inhibitory synaptic input. The potential multitude of roles this conductance may have in fine-tuning information in SPN Molecular cloning and heterologous expression studies requires further study. have shown that Kv1.4, Kv3.4, Kv4.1, Kv4.2, and Kv4.3 78 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

Fig. 7. Putative transient outward rectifying Kϩ channel subunit mRNAs expressed by SPN. (A) Ethidium bromide stained gels of the scRT-PCR products amplified with primers specific for the transcripts indicated above the appropriate columns. Differential expression of Kv1.5, Kv4.1, Kv4.3, KChIP1, Kv␤1, Kv␤2, Kv␤3 and DPP6 in addition to the housekeeping gene ␤-actin mRNA were all detected in SPN. These neurons did not express mRNA for Kv1.4, Kv3.4, Kv4.2, KChIP2, KChIP3 or NCS-1. A negative (Ϫve) control was performed on the contents of an electrode which was placed next to a cell without seal formation or harvesting of cytoplasmic contents. Molecular weight markers are shown in the lanes on the far left and far right, together with corresponding number of base pairs (Int, Intronic; -RT, cell contents lacking reverse transcriptase in the PCR amplification process). All gene-specific primers are listed in Table 1 (Methods). (B) Summary histogram illustrating the expression of the mRNA transcripts expressed in SPN (data from seven SPN). subunits are capable of forming homomeric channels that specific Kv4.1 channel blocker negates the possibility of rapidly inactivate and are sensitive to 4-AP, the hallmarks clarifying whether both or one of these ␣-subunits are of A-type channels (Serodio et al., 1994, 1996). In studies involved in mediating transient outward rectification in carried out here, single-cell RT-PCR revealed expression SPN. In relation to this point, studies in other central neural of mRNA for Kv1.5, Kv4.1 and Kv4.3. No detectable levels areas using single-cell reverse transcriptase polymerase of mRNA for Kv1.4, Kv3.4, or Kv4.2 were detected in SPN. chain reaction (scRT-PCR) techniques have revealed mul- These data strongly suggest that the ␣-subunits compris- tiple genes coding for the pore-forming ␣-subunit of Kϩ ing the channels underlying the A-like conductance in SPN channels are co-expressed in mammalian neurons (Song are coded for by co-expression of members of the Kv4 et al., 1998; Martina et al., 1998). For example, in basal subfamily of Kϩ channels. More specifically, the lack of ganglia and basal forebrain neurons, all three members of expression of mRNA for Kv4.2 indicates channels contain- the Kv4 subfamily are co-expressed. The number of Kv4.2 ing Kv4.1 and/or Kv4.3 subunits mediate the current. In transcripts was found to be linearly related to the conduc- heterologous expression systems, Kv4.1 and Kv4.3 chan- tance of the A-type channels across the basal forebrain nels are potently blocked by 4-AP, with IC50 values in the and basal ganglia neurons, while those of Kv4.1 and Kv4.3 low millimolar range (Serodio et al., 1994, 1996), values in were not, suggesting that Kv4.2 is primarily responsible for the range of the IC50 recorded here. Furthermore, PaTx-2, the current in these neurons (Tkatch et al., 2000). Simi- a specific blocker of Kv4.2 and Kv4.3 channels, and PiTx- larly, a combined scRT-PCR and patch clamp study in the K␣, a selective blocker of rapidly inactivating Kv channels, hippocampus also suggested Kv4.2 and Kv4.3 contribute inhibited a significant proportion of the transient outwardly to an A-type current (Martina et al., 1998). rectifying current. In around 50% of SPN, mRNA expression for a third ␣ Taken together, the molecular and pharmacological subunit, Kv1.5 was detected. As subunits from Kv1 and data outlined here strongly suggest Kv4.1 and Kv4.3 chan- Kv4 gene families do not co-assemble to form heteromeric nels form one of the substrates mediating at least a con- channels (McCormack et al., 1990; Covarrubias et al., stituent part of the recorded conductance. The lack of a 1991), the precise role of Kv1.5 in SPN remains to be A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 79 clarified. One of the hallmarks of channels composed of Bahring et al., 2001; Beck et al., 2002). A range of different the Kv1 family ␣ subunits is the presence of a slow com- properties of these channels have been observed depen- ponent of inactivation recovery with a time constant in the dent upon the specific combination of Kv4 and KChIP range of seconds (Rettig et al., 1994). The time constant of proteins expressed. For Kv4.2, co-expression with KChIP1 inactivation recorded here was almost two orders of mag- slowed the rate of inactivation, accelerated the time course nitude below this (19.3 ms). Kv1.5 channels can co-as- of recovery from inactivation, and shifted the voltage of semble with Kv␤3.1 subunits to form a novel type of het- half-activation to more negative potentials (An et al., 2000; eromeric channel, which inactivates significantly faster Bahring et al., 2001; Nakamura et al., 2001b). On the other than homomeric Kv channels and recover from inactivation hand, co-expression of KChIP1 with Kv4.1 accelerated significantly faster, with double exponential time-courses inactivation and shifted the voltage of half-activation in the (Leicher et al., 1998). However, these are still substantially positive direction (Nakamura et al., 2001b). More strikingly, longer than the time constants recorded here for SPN. This co-expression of a novel KChIP4 isoform (KChIP4a), coupled with the lack of effect of either TsTx-K␣ or DnTx-I, which contains a K-channel inactivation suppressor specific Kv1 channel subunit blockers on the transient (KIS) domain, nearly abolished Kv4 channel inactivation outward current, suggest it unlikely that Kv1.5 channels (Holmqvist et al., 2002). Thus, the assembly of different contribute significantly to the transient outwardly rectifying KChIP and Kv4 combinations likely contributes to the di- current observed. Indeed, it would appear Kv1.5 contrib- versity and complexity of features of A-type currents in utes solely to the delayed rectifier observed in SPN. How- neurons and may underlie the variations in the time-course ever, it may be that adrenal medulla-innervating SPN ex- and properties of A-type conductances in SPN. press Kv1.5 channels which could underlie the slow com- Another auxiliary subunit differentially expressed in ponent of the transient outward current observed in this SPN was DPP6. DPP6 has widespread expression in the specific functional subtype of SPN (Wilson et al., 2002). CNS, particularly in areas of abundant Kv4 ␣-subunit ex- This notion is further supported by the lack of sensitivity to pression (Nadal et al., 2003). In rat cerebellum, native 4-AP of Kv1.5 and the slow 4-AP-insensitive conductance A-currents may be made of three proteins, Kv4 ␣-subunits, observed in adrenal medulla-innervating SPN. KChIP and DPP6 (Nadal et al., 2003). As some SPN can express mRNA for both KChIP1 and DPP6, it is possible Transient outward rectifier accessory subunits that a subpopulation of these neurons may express tran- expressed by SPN sient outwardly rectifying conductances comprised of Auxiliary (␤) subunits act to modulate the functional prop- these subunits. Interestingly, in contrast to the slowing of erties of A-type channels when expressed with ␣ subunits. channel kinetics by KChIP (An et al., 2000; Nakamura et For example, Kv␤1 subunits are capable of transforming al., 2001a), DPP6 accelerates the kinetics of inactivation delayed rectifier-like channels composed of Kv1 family considerably. Co-expression of both ␤-subunits produces subunits (Kv1.1, Kv1.2, and Kv1.5) into rapidly inactivating, currents with intermediary properties which depend upon A-like channels (Rettig et al., 1994; Heinemann et al., the relative concentrations of each protein (Nadal et al., 1996; Sewing et al., 1996). Furthermore, Kv4.2 channels 2003). Thus, differential expression of KChIPs and DPP6 have been shown to interact with PSD95 (Wong et al., leads to another possible mechanism by which SPN may 2002) and NCS-1 (frequenin; Nakamura et al., 2001a). modulate the transient outwardly rectifying current.. The NCS-1 expression is diffuse throughout the gray matter of involvement of other auxiliary subunits such as KChIP4 the spinal cord, with expression being localized in axons and the non-KChIP members of the neuronal calcium sen- and axon terminals (Averill et al., 2004). Of the auxiliary sor family, Kϩ channel-associated protein (KChAP, Pour- subunits probed in the present study, mRNA for Kv␤2 was rier et al., 2003), mink-related peptide 1 (MiRP; Deschenes found expressed by all SPN. Thus taken together, our data and Tomaselli, 2002; Pourrier et al., 2003) or MinK (Pour- suggest the transient outward rectification observed in rier et al., 2003) requires further investigation. SPN is likely comprised of Kv4.1 or Kv4.3 together with ␤ Kv 2 subunits. Physiological significance of the rectifying One interesting feature of the data described here was conductance that ␤ subunit mRNA’s were found differentially expressed in SPN. These included KChIP 1, Kv␤1, Kv␤3 and DPP6 One of the first cellular functions attributed to transient, (DPPX, BSPL; Kin et al., 2001; Nadal et al., 2003). KChIP outward, rapidly inactivating currents was the slowing of (1–4) when expressed with Kv4 subunits reconstitute sev- interspike depolarization, enabling neurons to discharge at eral features of native channels (An et al., 2000). KChIPs very slow rates (Connor and Stevens, 1971). To accom- possess calcium binding domains and belong to the su- plish this goal, the conductance needed to open at sub- perfamily of EF-hand-containing proteins (Braunewell and threshold membrane potentials and then inactivate, allow-

Gundelfinger, 1999). They interact with the long NH2 ter- ing spike threshold to be reached. On repolarization of the minus of Kv4 proteins (Bahring et al., 2001) enhancing spike, channels must de-inactivate quickly, so that the surface expression, thereby increasing current magnitude ensuing depolarization is capable of opening them once (An et al., 2000; Decher et al., 2001; Liss et al., 2001). In again. These three features, subthreshold activation, rapid addition to enhancing surface expression, KChIPs also inactivation and rapid recovery from inactivation, are all modulate Kv4 channel kinetic behavior (An et al., 2000; features of the transient outward current recorded in SPN. 80 A. D. Whyment et al. / Neuroscience 178 (2011) 68–81

The membrane potential range for activation of the Cassell JF, McLachlan EM (1987) Two calcium-activated potassium transient conductance in SPN suggests a pivotal role in conductances in a subpopulation of coeliac neurones of guinea-pig regulating the frequency of firing. The repolarizing phase of and rabbit. J Physiol 394:331–349. the action potential and activation of the afterhyperpolar- Coetzee WA, Amarillo Y, Chiu J, Chow A, Lau D, McCormack T, Moreno H, Nadal MS, Ozaita A, Pountney D, Saganich M, Vega- izing potential may indeed be sufficient for de-inactivation ϩ Saenz de Miera E, Rudy B (1999) Molecular diversity of K chan- of the channels and subsequently activate this conduc- nels. Ann N Y Acad Sci 868:233–285. tance, thus prolonging the return to resting potentials (Yo- Connor JA, Stevens CF (1971) Prediction of repetitive firing behavior shimura et al., 1986; Spanswick and Logan, 1990; Sah and from voltage clamp data on an isolated neurone soma. J Physiol McLachlan, 1995). However, as the channels are partially 213:31–53. inactivated at rest, especially in neurons with depolarized Covarrubias M, Wei AA, Salkoff L (1991) Shaker, Shal, Shab, resting membrane potentials, hyperpolarization due to in- and Shaw express independent Kϩ current systems. Neuron hibitory synaptic input or synaptic release of modulatory 7:763–773. Decher N, Uyguner O, Scherer CR, Karaman B, Yuksel-Apak M, substances may be required to increase the number of Busch AE, Steinmeyer K, Wollnik B (2001) hKChIP2 is a functional available Kv4 channels (Talukder and Harrison, 1995). modifier of hKv4.3 potassium channels: cloning and expression of a short hKChIP2 splice variant. Cardiovasc Res 52:255–264. CONCLUSION Dembowsky K, Czachurski J, Seller H (1986) Three types of sympa- thetic preganglionic neurones with different electrophysiological These observations provide the first evidence that the properties are identified by intracellular recordings in the cat. transient outward rectifying current in SPN may be coded Pflugers Arch 406:112–120. for by co-expression of members of the Kv4 subfamily. Deschenes I, Tomaselli GF (2002) Modulation of Kv4.3 current by Differential expression of accessory ␤ subunits including accessory subunits. FEBS Lett 528:183–188. KChIP 1; Kv␤1 and Kv␤3 and the peptidase homologue Diochot S, Drici MD, Moinier D, Fink M, Lazdunski M (1999) Effects of phrixotoxins on the Kv4 family of potassium channels and implica- DPP6, which may act to modulate channel gating, density tions for the role of Ito1 in cardiac electrogenesis. Br J Pharmacol and kinetics, could underpin the prolonged and variable 126:251–263. time-course of this conductance in SPN. Further work is Diochot S, Schweitz H, Beress L, Lazdunski M (1998) Sea anemone now required to correlate differential expression patterns of peptides with a specific blocking activity against the fast inactivat- channel subunits with functional properties of the A-like ing potassium channel Kv3.4. J Biol Chem 273:6744–6749. conductance in SPN. These differential expression profiles Dixon AK, Richardson PJ, Lee K, Carter NP, Freeman TC (1998) may contribute a level of plasticity to SPN information Expression profiling of single cells using 3 prime end amplification processing, not previously recognized. (TPEA) PCR. Nucleic Acids Res 26:4426–4431. Gutman GA, Chandy KG, Grissmer S, Lazdunski M, McKinnon D, Pardo LA, Robertson GA, Rudy B, Sanguinetti MC, Stuhmer W, Acknowledgments—The financial support of the BHF and the Wang X (2005) International Union of . LIII. Nomen- BBSRC (to DS) and the Canadian Institutes of Health Research clature and molecular relationships of voltage-gated potassium (to LPR; grant MOP 97803) are gratefully acknowledged. channels. Pharmacol Rev 57:473–508. Heinemann SH, Rettig J, Graack HR, Pongs O (1996) Functional REFERENCES characterization of Kv channel beta-subunits from rat brain. J Physiol 493:625–633. Alvado L, Allen CN (2008) (TEA) increases the Heinemann SH, Rettig J, Wunder F, Pongs O (1995) Molecular and inactivation time constant of the transient Kϩ current in supracha- functional characterization of a rat brain Kv␤3 potassium channel ismatic nucleus neurons. Brain Res 1221:24–29. subunit. FEBS Lett 377:383–389. An WF, Bowlby MR, Betty M, Cao J, Ling HP, Mendoza G, Hinson JW, Holmqvist MH, Cao J, Hernandez-Pineda R, Jacobson MD, Carroll KI, Mattsson KI, Strassle BW, Trimmer JS, Rhodes KJ (2000) Modu- Sung MA, Betty M, Ge P, Gilbride KJ, Brown ME, Jurman ME, lation of A-type potassium channels by a family of calcium sensors. Lawson D, Silos-Santiago I, Xie Y, Covarrubias M, Rhodes KJ, Nature 403:553–556. Distefano PS, An WF (2002) Elimination of fast inactivation in Kv4 Averill S, Robson LG, Jeromin A, Priestley JV (2004) Neuronal calcium A-type potassium channels by an auxiliary subunit domain. Proc sensor-1 is expressed by dorsal root ganglion cells, is axonally Natl Acad SciUSA99:1035–1040. transported to central and peripheral terminals, and is concen- trated at nodes. Neuroscience 123:419–427. Inokuchi H, Masuko S, Chiba T, Yoshimura M, Polosa C, Nishi S Bahring R, Dannenberg J, Peters HC, Leicher T, Pongs O, Isbrandt D (1993) Membrane properties and dendritic arborization of the in- (2001) Conserved Kv4 N-terminal domain critical for effects of Kv termediolateral nucleus neurons in the guinea-pig thoracic spinal channel-interacting protein 2.2 on channel expression and gating. cord in vitro. J Auton Nerv Syst 43:97–106. J Biol Chem 276:23888–23894. Itri JN, Vosko AM, Schroeder A, Dragich JM, Michel S, Colwell CS Banks MI, Haberly LB, Jackson MB (1996) Layer-specific properties of (2010) Circadian regulation of A-type potassium currents in the the transient K current (IA) in piriform cortex. J Neurosci suprachiasmatic nucleus. J Neurophysiol 103:632–640. 16:3862–3876. Jan LY, Jan YN (1997) Cloned potassium channels from eukaryotes Beck EJ, Bowlby M, An WF, Rhodes KJ, Covarrubias M (2002) Re- and prokaryotes. Annu Rev Neurosci 20:91–123. modelling inactivation gating of Kv4 channels by KChIP1, a small- Jerng HH, Pfaffinger PJ, Covarrubais M (2004a) Molecular physiology molecular-weight calcium-binding protein. J Physiol 538:691–706. and modulation of somatodendritic A-type potassium channels. Bordey A, Feltz P, Trouslard J (1995) Kinetics of A-currents in sym- Mol Cell Neurosci 27:343–369. pathetic preganglionic neurones and glial cells. Neuroreport Jerng HH, Qian Y, Pfaffinger PJ (2004b) Modulation of Kv4.2 channel 7:37–40. expression and gating by dipeptidyl peptidase 10 (DPP10). Bio- Braunewell KH, Gundelfinger ED (1999) Intracellular neuronal calcium phys J 87:2380–2396. sensor proteins: a family of EF-hand calcium-binding proteins in Katoh E, Nishio H, Inui T, Nishiuchi Y, Kimura T, Sakakibara S, search of a function. Cell Tissue Res 295:1–12. Yamazaki T (2000) Structural basis for the biological activity of A. D. Whyment et al. / Neuroscience 178 (2011) 68–81 81

dendrotoxin-I, a potent potassium channel blocker. Biopolymers Sah P, McLachlan EM (1995) Membrane properties and synaptic 54:44–57. potentials in rat sympathetic preganglionic neurons studied in hor- Kin Y, Misumi Y, Ikehara Y (2001) Biosynthesis and characterization of izontal spinal cord slices in vitro. J Auton Nerv Syst 53:1–15. the brain-specific membrane protein DPPX, a dipeptidyl peptidase Scott VE, Rettig J, Parcej DN, Keen JN, Findlay JB, Pongs O, Dolly JO IV-related protein. J Biochem 129:289–295. (1994) Primary structure of a beta subunit of alpha-dendrotoxin- Klenk KC, Tenenholz TC, Matteson DR, Rogowski RS, Blaustein MP, sensitive Kϩ channels from bovine brain. Proc Natl Acad SciUSA Weber DJ (2000) Structural and functional differences of two toxins 91:1637–1641. from the scorpion Pandinus imperator. Proteins 38:441–449. Serodio P, Kentros C, Rudy B (1994) Identification of molecular com- Leicher T, Bahring R, Isbrandt D, Pongs O (1998) Coexpression of the ponents of A-type channels activating at subthreshold potentials. KCNA3B gene product with Kv1.5 leads to a novel A-type potas- J Neurophysiol 72:1516–1529. sium channel. J Biol Chem 273:35095–35101. Serodio P, Vega-Saenz de Miera M, Rudy B (1996) Cloning of a novel Lien CC, Martina M, Schultz JH, Ehmke H, Jonas P (2002) Gating, component of A-type Kϩ channels operating at subthreshold po- modulation and subunit composition of voltage-gated K(ϩ) chan- tentials with unique expression in heart and brain. J Neurophysiol nels in dendritic inhibitory interneurones of rat hippocampus. 75:2174–2179. J Physiol 538 (Pt 2):405–419. Sewing S, Roeper J, Pongs O (1996) Kv beta 1 subunit binding Liss B, Franz O, Sewing S, Bruns R, Neuhoff H, Roeper J (2001) specific for shaker-related potassium channel alpha subunits. Neu- Tuning pacemaker frequency of individual dopaminergic neurons ron 16:455–463. by Kv4.3L and KChip3.1 transcription. EMBO J 20:5715–5724. Song WJ, Tkatch T, Baranauskas G, Ichinohe N, Kitai ST, Surmeier Martina M, Schultz JH, Ehmke H, Monyer H, Jonas P (1998) Func- ϩ DJ (1998) Somatodendritic depolarization-activated potassium tional and molecular differences between voltage-gated K chan- currents in rat neostriatal cholinergic interneurons are predomi- nels of fast-spiking interneurons and pyramidal neurons of rat nantly of the A type and attributable to coexpression of Kv4.2 and hippocampus. J Neurosci 18:8111–8125. Kv4.1 subunits. J Neurosci 18:3124–3137. ϩ McCormack K, Lin JW, Iverson LE, Rudy B (1990) Shaker K channel Spanswick D, Logan SD (1990) Sympathetic preganglionic neurones subunits form heteromultimeric channels with novel functional in neonatal rat spinal cord in vitro: electrophysiological character- properties. Biochem Biophys Res Commun 171:1361–1371. istics and the effects of selective excitatory receptor Miyazaki T, Dun NJ, Kobayashi H, Tosaka T (1996) Voltage-depen- agonists. Brain Res 525:181–188. dent potassium currents of sympathetic preganglionic neurons in Spanswick D, Renaud LP, Logan SD (1998) Bilaterally evoked mono- neonatal rat spinal cord thin slices. Brain Res 743:1–10. synaptic EPSPs, NMDA receptors and potentiation in rat sympa- Morohashi Y, Hatano N, Ohya S, Takikawa R, Watabiki T, Takasugi N, thetic preganglionic neurones in vitro. J Physiol 509:195–209. Imaizumi Y, Tomita T, Iwatsubo T (2002) Molecular cloning and Storm JF (1987) Potassium currents in hippocampal pyramidal cells. characterization of CALP/KChIP4, a novel EF-hand protein inter- Prog Brain Res 83:161–187. acting with presenilin 2 and voltage-gated potassium channel sub- Stuart GJ, Dodt HU, Sakmann B (1993) Patch-clamp recordings from unit Kv4. J Biol Chem 277:14965–14975. the soma and dendrites of neurons in brain slices using infrared Nadal MS, Ozaita A, Amarillo Y, Vega-Saenz de Miera E, Ma Y, Mo W, video microscopy. Pflugers Arch 423:511–518. Goldberg EM, Misumi Y, Ikehara Y, Neubert TA, Rudy B (2003) Talukder G, Harrison NL (1995) On the mechanism of modulation of The CD26-related dipeptidyl aminopeptidase-like protein DPPX is transient outward current in cultured rat hippocampal neurons by a critical component of neuronal A-type Kϩ channels. Neuron di- and trivalent cations. J Neurophysiol 73:73–79. 37:449–461. Tkatch T, Baranauskas G, Surmeier DJ (2000) Kv4.2 mRNA abun- Nakamura TY, Pountey DJ, Ozaita A, Nandi S, Ueda S, Rudy B, ϩ dance and A-type K(ϩ) current amplitude are linearly related Coetzee WA (2001a) A role for frequenin, a Ca2 -binding protein, as a regulator of Kv4ϩ-currents. Proc Natl Acad SciUSA in basal ganglia and basal forebrain neurons. J Neurosci 98:12808–12813. 20:579–588. Nakamura TY, Nandi S, Pountney DJ, Artman M, Rudy B, Coetzee van den Top M, Lee K, Whyment AD, Blanks AM, Spanswick D (2004) WA (2001b) Different effects of the Ca(2ϩ)-binding protein, Orexigen-sensitive NPY/AgRP pacemaker neurons in the hypotha- KChIP1, on two Kv4 subfamily members, Kv4.1 and Kv4.2. FEBS lamic arcuate nucleus. Nat Neurosci 7:493–494. Lett 499:205–209. Wilson JMM, Coderre E, Renaud LP, Spanswick D (2002) Active and Pickering AE, Spanswick D, Logan SD (1991) Whole-cell recordings passive membrane properties of rat sympathetic preganglionic from sympathetic preganglionic neurons in rat spinal cord slices. neurones innervating the adrenal medulla. J Physiol 545:945–960. Neurosci Lett 130:237–242. Wong W, Newell EW, Jugloff DG, Jones OT, Schlichter LC (2002) Cell Polosa C, Yoshimura M, Nishi S (1988) Electrophysiological proper- surface targeting and clustering interactions between heterolo- ties of sympathetic preganglionic neurons. Annu Rev Physiol gously expressed PSD-95 and the Shal voltage-gated potassium 50:541–551. channel, Kv4.2. J Biol Chem 277:20423–20430. Pourrier M, Schram G, Nattel S (2003) Properties, expression and Xiong H, Kovacs I, Zhang Z (2004) Differential distribution of KChIPs potential roles of cardiac Kϩ channel accessory subunits: MinK, mRNAs in adult mouse brain. Brain Res Mol Brain Res 128: MiRPs, KChIP, and KChAP. J Membr Biol 194:141–152. 103–111. Rettig J, Heinemann SH, Wunder F, Lorra C, Parcej DN, Dolly JO, Yoshimura M, Polosa C, Nishi S (1986) Afterhyperpolarization mech- Pongs O (1994) Inactivation properties of voltage-gated Kϩ chan- anisms in cat sympathetic preganglionic neurones in vitro. J Neu- nels altered by presence of beta-subunit. Nature 369:289–294. rophysiol 55:1234–1246. Rhodes KJ, Carroll KI, Sung MA, Doliveira LC, Monaghan MM, Burke Yoshimura M, Polosa C, Nishi S (1987) A transient outward rectifica- SL, Strassle BW, Buchwalder L, Menegola M, Cao J, An WF, tion in the cat sympathetic preganglionic neuron. Pflugers Arch Trimmer JS (2004) KChIPs and Kv4 alpha subunits as integral 408:207–208. components of A-type potassium channels in mammalian brain. Zagha E, Ozaita A, Chang SY, Nadal MS, Lin U, Saganich MJ, J Neurosci 24:7903–7915. McCormack T, Akinsanya KO, Qi SY, Rudy B (2005) DPP10 Rogawski MA (1985) The A-current: how ubiquitous a feature of modulates Kv4-mediated A-type potassium channels. J Biol Chem excitable cells is it? Trends Neurosci 8:214–219. 280:18853–18861.

(Accepted 30 December 2010) (Available online 4 January 2011)