Ternary Kv4.2 channels recapitulate voltage-dependent inactivation kinetics of A-type K+ channels in cerebellar granule neurons
- Yimy Amarillo1,
- Jose A. De Santiago-Castillo2,
- Kevin Dougherty2,
- Jonathon Maffie1,
- Elaine Kwon1,
- Manuel Covarrubias2 and
- Bernardo Rudy1
- 1Department of Physiology & Neuroscience and Department of Biochemistry, Smilow Neuroscience Program, New York University School of Medicine, Smilow Research Center, 522 First Avenue, 6th Floor, New York, NY 10016, USA2Department of Pathology, Anatomy and Cell Biology, Jefferson Medical College of Thomas Jefferson University, 1020 Locust Street, Philadelphia, PA 19107, USA
- Corresponding author B. Rudy: Smilow Neuroscience Program, Smilow Research Center, New York University School of Medicine, 522 First Avenue, 6th Floor, New York, NY 10016, USA. Email: rudyb01{at}med.nyu.edu
Abstract
Kv4 channels mediate most of the somatodendritic subthreshold operating A-type current (ISA) in neurons. This current plays essential roles in the regulation of spike timing, repetitive firing, dendritic integration and plasticity. Neuronal Kv4 channels are thought to be ternary complexes of Kv4 pore-forming subunits and two types of accessory proteins, Kv channel interacting proteins (KChIPs) and the dipeptidyl-peptidase-like proteins (DPPLs) DPPX (DPP6) and DPP10. In heterologous cells, ternary Kv4 channels exhibit inactivation that slows down with increasing depolarization. Here, we compared the voltage dependence of the inactivation rate of channels expressed in heterologous mammalian cells by Kv4.2 proteins with that of channels containing Kv4.2 and KChIP1, Kv4.2 and DPPX-S, or Kv4.2, KChIP1 and DPPX-S, and found that the relation between inactivation rate and membrane potential is distinct for these four conditions. Moreover, recordings from native neurons showed that the inactivation kinetics of the ISA in cerebellar granule neurons has voltage dependence that is remarkably similar to that of ternary Kv4 channels containing KChIP1 and DPPX-S proteins in heterologous cells. The fact that this complex and unique behaviour (among A-type K+ currents) is observed in both the native current and the current expressed in heterologous cells by the ternary complex containing Kv4, DPPX and KChIP proteins supports the hypothesis that somatically recorded native Kv4 channels in neurons include both types of accessory protein. Furthermore, quantitative global kinetic modelling showed that preferential closed-state inactivation and a weakly voltage-dependent opening step can explain the slowing of the inactivation rate with increasing depolarization. Therefore, it is likely that preferential closed-state inactivation is the physiological mechanism that regulates the activity of both ternary Kv4 channel complexes and native ISA-mediating channels.
K+ channels containing Kv4 pore-forming subunits (Kv4 channels) mediate most of the subthreshold-operating somatodendritic transient or A-type K+ current in neurons (also known as ISA) (Serodio et al. 1994; reviewed in Jerng et al. 2004a). This current is fundamental to neuronal function. It can contribute to spike repolarization and has critical roles in the modulation of the frequency of repetitive firing, signal processing in dendrites and spike timing-dependent plasticity (Connor & Stevens, 1971; Hoffman et al. 1997; Schoppa & Westbrook, 1999; Adams et al. 2000; Johnston et al. 2000, 2003; Hille, 2001; Liss et al. 2001; Ramakers & Storm, 2002; Kim et al. 2005, 2007; Chen et al. 2006; Hu et al. 2006; Thompson, 2007). These functions rely on the precise voltage dependence and kinetic properties of the underlying K+ channels.
Studies in heterologous expression systems have shown that Kv4 channels may exist as ternary complexes composed of a pore-forming subunit and at least two distinct auxiliary subunits: Kv channel interacting proteins (KChIPs) and the dipeptidyl-peptidase-like proteins (DPPLs) DPPX and DPP10 (An et al. 2000; Nadal et al. 2001, 2003; Jerng et al. 2004b, 2005; Ren et al. 2005; Zagha et al. 2005). These associations control the trafficking of channel complexes to the plasma membrane and have major effects on the voltage dependence and kinetics of the channels. However, the significance of these auxiliary proteins in governing the patterns of expression and functional properties of the native ISA channels in neurons has yet to be demonstrated.
The kinetics of inactivation of ternary Kv4 channels expressed in heterologous expression systems has unusual voltage dependence (G. Wang et al. 2005; Jerng et al. 2007). With increasing step depolarizations, the rate of inactivation slows down. This trend is apparent over a wide range of membrane potentials (−20 to +100 mV; G. Wang et al. 2005) and is also observed with binary combinations of Kv4.3 and KChIP1 (Kaulin et al. 2008). In contrast, Kv4 channels expressed in the absence of the auxiliary proteins (Serodio et al. 1996; Jerng & Covarrubias, 1997; Franqueza et al. 1999; Bahring et al. 2001; Beck & Covarrubias, 2001; G. Wang et al. 2005; S. Wang et al. 2005) as well as inactivating channels from other Kv subfamilies (Tseng-Crank et al. 1990; Rudy et al. 1991; Schroter et al. 1991; Vega-Saenz de Miera et al. 1992; Riazanski et al. 2001) do not exhibit such a voltage dependence.
This unusual behaviour is potentially significant because it has been observed to one degree or another in recordings of A-type K+ currents from hippocampal neurons (Klee et al. 1995; Hoffman et al. 1997; Martina et al. 1998; Lien et al. 2002); however, it has not received much attention and its mechanism is not well understood. Moreover, the ISA is usually only one of several components of the total current observed in whole cell neuronal recordings. Therefore, the ISA has to be properly isolated from other current components to determine its properties accurately. For example, the inactivating current in CA1 hippocampal pyramidal neurons appears to include a contribution from channels that are not members of the Kv4 subfamily (and are possibly mediated by Kv1 subunits) (Chen et al. 2006). This is a minor component of the total inactivating current. However, since it has a different voltage dependence and kinetics than the Kv4 component, it would influence the properties of the total inactivating current differently at different membrane potentials.
Previously, we proposed that the unusual voltage dependence of inactivation kinetics is a hallmark of preferential closed-state inactivation in Kv4 channel complexes, and is determined by auxiliary subunits (Kaulin et al. 2008). Therefore, this behaviour in neuronal Kv4 channels would suggest that they include a specific combination of Kv4 auxiliary subunits, which confer preferential closed-state inactivation. To test this hypothesis systematically and under stringent conditions, we recorded the ISA from cerebellar granule cells in acute slices and compared the voltage dependence of inactivation kinetics with that of Kv4 channels heterologously expressed in mammalian cells. The latter experiments included the Kv4.2 subunit alone and in binary or ternary complexes with the accessory proteins KChIP1 and DPPX-S. Our results show that the kinetics of inactivation of the ISA in cerebellar granule neurons has a voltage dependence that is remarkably similar to that observed for ternary Kv4 channels containing KChIP and DPPX proteins in heterologous cells. These results support the view that KChIP and DPPX auxiliary proteins are components of the native ISA channels. Furthermore, using kinetic modelling we show that preferential closed-state inactivation and a weakly voltage-dependent opening step can explain the slowing of the inactivation rate with increasing depolarization.
Methods
Electrophysiological recording from heterologous mammalian cells
Electrophysiological recording from heterologous mammalian cells was carried out as previously described (Dougherty & Covarrubias, 2006). Briefly, tsA-201 cells (provided by Dr R. Horn, Thomas Jefferson University, Philadelphia, PA, USA) were transfected with the appropriate plasmids (Kv4.2, KChIP1 and/or DPPX-S) at a 1 : 1 mass ratio using the calcium phosphate method. A plasmid containing a CD8 cDNA (5 μg) was included in the cotransfection to allow the identification of individual transfected cells by decorating them with beads bearing anti-CD8 antibody (Dynal Biotech, Brown Deer, WI, USA) (Kaulin et al. 2008). Kv4.2 currents were recorded using the tight-seal whole-cell configuration of the patch-clamp method with a pipette (intracellular) solution containing (mm): 120 KF, 1 CaCl2, 2 MgCl2, 11 EGTA and 10 Hepes, pH 7.2, adjusted with KOH; and the external bath solution containing (mm): 130 NaCl, 2 KCl, 1.5 CaCl2, 1 MgCl2, 20 TEA (tetraethylammonium chloride) and 10 Hepes, pH 7.4, adjusted with NaOH. The calculated free [Ca2+] and [Mg2+] in the intracellular solution were ∼40 nm and ∼1.4 mm, respectively (MaxChelator, WEBMAXC v2.10; http://www.stanford.edu/~cpatton/maxc.html). Extracellular TEA was necessary to eliminate a small but significant endogenous delayed-rectifier K+ current. The experiment began once current kinetics and amplitude became stable. Series resistance (Rs = 2–5 MΩ) was compensated to yield a total voltage error of < 3 mV. A P/−4 leak subtraction protocol consisting of four subpulses from a subsweep holding potential of −110 mV was used to subtract passive components of the total current. Currents were filtered at 2 kHz and sampled at 10 kHz. Analyses and graphical displays were produced with pCLAMP 9.0 (Axon Instruments, Inc., Union City, CA, USA) and Origin 7.5 (OriginLab Corp., Northhampton, MA, USA) software. Recordings in tsA-201 cells were most stable in the intracellular solution indicated above. To confirm that the voltage-dependent behaviour observed under these conditions was independent of the intracellular solution used, we obtained recordings from CHO cells coexpressing Kv4.2 and KChIP1 with and without DPPX-S with the same intracellular pipette solution as that used to record from cerebellar granule cells (online supplemental material, Supplementary Fig. 1). All measurements were taken at room temperature (22–24°C) and results are expressed as means ± s.e.m.
Slice preparation
All experiments were carried out in accordance with the NIH Guide for the Care and Use of Laboratory Animals and were approved by the New York University School of Medicine Animal Care and Use Committee.
Acute brain slices were prepared using standard techniques (Stuart et al. 1993). Mice (10–30 days old) were deeply anaesthetized via intraperitoneal injection of pentobarbital (100 mg (kg body weight)−1) and decapitated. The brain was rapidly removed to oxygenated, ice-cold artificial cerebrospinal fluid (ACSF) that contained (mm): 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 26 NaHCO3, 10 glucose, 2 CaCl2 and 1 MgSO4. Cutting solution was continuously bubbled with 95% O2 and 5% CO2 to maintain a pH of approximately 7.40. Slices (350 μm thick) were cut on a Vibratome 3000 Plus (Vibratome, St Louis, MO, USA) and incubated in a holding chamber at 32–35°C for approximately 30 min followed by continued incubation at room temperature prior to electrophysiological recording, at which point slices were transferred to a submersion-type recording chamber attached to the microscope stage and perfused with ACSF.
Patch-clamp recordings from cerebellar granule cells
Whole-cell voltage clamp recordings were performed from cerebellar granule cells using the whole-cell configuration of the patch-clamp technique (Hamill et al. 1981). Cells were visualized using a 40×, 0.8 NA water-immersion objective (Olympus) on an Olympus BX-50 upright microscope equipped with IR-DIC optics (Stuart et al. 1993). Patch electrodes (3–6 MΩ) were fashioned from borosilicate glass using an horizontal puller (Model P-97, Sutter Instrument Co., Novato, CA, USA) and filled with a solution containing (mm): 130 CH3KO3S (potassium methanesulphonate), 10 NaCl, 10 Hepes, 10 phosphocreatine (Tris), 2 Mg-ATP, 0.4 Na-GTP, 0.4 EGTA, 2 MgCl2 and 0.16 CaCl2, pH adjusted to 7.40 with KOH, and the osmolarity adjusted to 290 mosmol l−1 with sucrose. Membrane potentials were corrected for a 10 mV junction potential. Currents were recorded using an Axopatch 200B amplifier (Axon Instruments), lowpass filtered at 5 kHz, digitized at 16-bit resolution (Digidata 1322A; Axon Instruments) and sampled at 20 kHz. Cells with a series resistance (Rs) larger than 20 MΩ upon break-in were rejected. Rs was compensated to the extent possible (typically, 50–85% correction) and data obtained from a given cell were rejected if Rs changed by > 20% during the course of the experiment. pCLAMP 9 software (Axon Instuments) was used for data acquisition, and analysis was performed using the Clampfit module of pCLAMP.
Immunohistochemistry
For localization of Kv4.2 and Kv4.3 protein in the cerebellar cortex, mice were anaesthetized with an intraperitoneal injection of pentobarbital and transcardially perfused with 0.9% saline containing heparin (1 U ml−1), followed by 30–50 ml of 0.1 m sodium phosphate buffer (NaPB; pH 7.4) containing 4% paraformaldehyde. Brains were dissected out and postfixed in the same fixative solution for 1 h at room temperature and then placed in a 30% sucrose solution at 4°C for 24 h. Frozen sagittal sections through the cerebellum, 40 μm thick, were cut on a sliding microtome and collected in phosphate-buffered saline (PBS). Sections were washed in PBS and then incubated in a blocking solution (10% normal goat serum, 1% BSA, 0.2% cold fish gelatin, and 0.2% Triton X-100 in PBS) for 1 h at room temperature to minimize non-specific binding. Sections were then incubated with antibodies directed against Kv4.2 (10 μg ml−1; mouse monoclonal anti-Kv4.2; K57/1; Neuromab; http://www.neuromab.org; or rabbit polyclonal directed against Kv4.2; Nadal et al. 2003; at 1 : 1000), or Kv4.3 (1 μg ml−1; mouse monoclonal anti-Kv4.3; K75/41; Neuromab) in diluted (1 : 10) blocking solution overnight at 4°C. Following incubation with primary antibodies, sections were washed with PBS and incubated for 1 h in fluorescent Cy3-conjugated secondary antibody (goat anti-mouse or goat anti-rabbit, Jackson ImmunoResearch Laboratories, Inc., West Grove, PA, USA) at room temperature. Sections were then rinsed and finally mounted using Vectashield (Vector Laboratories, Inc., Burlingame, CA, USA). Immunofluorescence images were acquired with an Olympus MVX10 microscope equipped with a Leica DFC-340FX digital camera and Leica Firecam 3.0 software. Images were transferred to a graphics program (Adobe Photoshop CS2), in which brightness and contrast were adjusted. The monoclonal and polyconal antibodies for Kv4.2 both produced the same pattern of staining in the cerebellum. Figure 2 depicts staining with the monoclonal antibody against Kv4.2.
Quantitative global kinetic modelling
Previously, we applied global kinetic modelling and a model discrimination method to explain gating of Kv4 channels quantitatively (Kaulin et al. 2008). Using a Chi-squared minimization function, appropriately weighted macroscopic measurements (activation, inactivation and deactivation) were evaluated simultaneously to constrain the analysis and estimate the best global-fit parameters for specific kinetic models. Here, we implemented this strategy in the new program IChMASCOT (Ion-Channel-Markov-Scheme-Optimizer; http://www.ichmascot.org) and assumed the kinetic scheme shown in Fig. 6A. This scheme is similar to those models that successfully described Kv4 gating in the study of Kaulin et al. (2008). In particular, these models explained the slowing of inactivation with progressive depolarization. For the sake of simplification, however, we did not include an unstable closed state connected to the open state, which reduces the open probability. Also, the scope of the kinetic analysis was limited to explain the voltage dependence of the rate of inactivation semiquantitatively. For all subunit combinations, we evaluated the following measurements simultaneously to constrain global kinetic modelling: (1) complete families of currents evoked by step depolarizations (Fig. 1); (2) the steady-state inactivation curve (not shown); and (3) the recovery from inactivation (Fig. 1F). For display, model simulations were performed in the program IChSimlab (http://www.ichmascot.org).
Results
Voltage dependence of inactivation of Kv4.2 channels in heterologous expression systems
Previous studies of the influence of KChIP and DPPL proteins on the properties of Kv4 channels have primarily used two-microelectrode voltage-clamp in the Xenopus oocyte expression system (Nadal et al. 2003, 2006; Jerng et al. 2005). To minimize the variability that methodological differences may introduce and facilitate comparison with the properties of native neuronal channels, we chose to express the Kv4 channel complexes in a mammalian heterologous expression system and obtain the currents with the same recording method (whole-cell patch-clamp).
The currents produced by expression of Kv4.2 proteins alone, were compared to those resulting from coexpression of Kv4.2 with the accessory subunits KChIP1 and DPPX-S (Fig. 1). Records of the currents recorded during step depolarizations from tsA-201 cells transfected with Kv4.2 alone, Kv4.2 plus KChIP1, Kv4.2 plus DPPX-S and Kv4.2 with both KChIP1 and DPPX-S are shown in Fig. 1A–D. As previously reported, KChIP1 slows down current inactivation and induces more complete inactivation by the end of the voltage pulse, and DPPX accelerates the overall channel kinetics (An et al. 2000; Beck et al. 2002; Nadal et al. 2003; reviewed in Jerng et al. 2004a). In Fig. 1E we compare the rate of inactivation as a function of voltage for the four conditions from a number of cells. As reported before, the currents mediated by Kv4 channels decay with a complex multiexponential time course (Jerng & Covarrubias, 1997; Bahring et al. 2001; Jerng et al. 2004b). Since the number of exponentials that may describe the decay varies depending on subunit composition, we used the time at which half of the peak current is inactivated (half-inactivation time or t1/2) to compare inactivation kinetics under different conditions. This model-independent strategy has been previously used by us and others to compare the inactivation kinetics of Kv4 currents (Nadal et al. 2001, 2003; Jerng et al. 2007).
In cells transfected with Kv4.2 alone, the rate of inactivation does not slow down during large depolarizations (Fig. 1E). In sharp contrast, the voltage dependence of the rate of inactivation of Kv4.2 channels cotransfected with KChIP1 exhibits a transition from an accelerating rate of inactivation at negative voltages to a decelerating rate at positive membrane potentials (Fig. 1E). In the presence of KChIP1, the t1/2 reaches a minimum between −40 and −10 mV; then, it increases with a slope of approximately 2.3 ms/10 mV (0 to +60 mV; Fig. 1E). As previously reported, DPPX-S alone speeds up the inactivation rate significantly. The differences between the t1/2 of Kv4.2 currents in cells coexpressing DPPX and cells expressing Kv4.2 channels alone are most prominent at membrane potentials more negative than ∼0 mV (Fig. 1E). Furthermore, the t1/2 of the currents in the presence of DPPX (without KChIP) does not increase with large depolarizations. However, inactivation slows down with increasing depolarization in cells transfected with both KChIP1 and DPPX-S with a slope of 2.9 ms/10 mV (measured between 0 and +60 mV). This slope is slightly steeper than that observed in cells cotransfected with KChIP1 only, but the main difference between the two conditions is in the rate of inactivation in the most negative voltage range. In cells expressing both auxiliary subunits, DPPX-S clearly accelerates the rate of inactivation between −60 and −10 mV, and thereby significantly broadens the voltage range over which the rate of inactivation decreases with increasing depolarization. Similar voltage dependence profiles were obtained in transfected CHO cells (Supplementary Fig. 1). The curves in Fig. 1E demonstrate that the relation between inactivation rate and membrane potential is distinct for the four conditions tested in these experiments. Therefore, subunit composition dictates Kv4 channel inactivation kinetics. The specific contributions of DPPX-S to Kv4.2 gating are also apparent when comparing the recoveries from inactivation (Fig. 1F). The ternary complex exhibits fast recovery from inactivation (τ = 45 ms) that is already set by the interaction between Kv4.2 and DPPX-S in the binary complex (τ = 40 ms).
Kv4.2 and Kv4.3 proteins in mouse cerebellar granule cells
Cerebellar granule neurons have a large ISA, consistent with prominent expression of Kv4.2 and Kv4.3 mRNAs and proteins (Serodio & Rudy, 1998; Rhodes et al. 2004; Strassle et al. 2005). They are also among the few neuronal populations where the patterns of expression of the various molecular components of Kv4 channels have been analysed with good cellular resolution (Nadal et al. 2003; Rhodes et al. 2004; Strassle et al. 2005; Zagha et al. 2005; Jerng et al. 2007). Moreover, compared to other neurons, cerebellar granule cells are elecrotonically compact due to the small size of these cells (∼5 μm) and the short length of dendritic processes, allowing good voltage control for whole cell recording (Cull-Candy et al. 1989). We therefore selected these neurons to investigate the properties of native neuronal ISA for comparison with the currents heterologously expressed in mammalian cells.
In situ hybridization and immunohistochemical studies in rat brain showed reciprocal gradients of expression of Kv4 mRNA transcripts and protein products, with Kv4.2 being expressed most prominently in anterior cerebellar lobules and Kv4.3 in posterior lobules and the flocculonodular lobe (Serodio & Rudy, 1998; Hsu et al. 2003; Strassle et al. 2005). We used immunohistochemistry with Kv4.2- and Kv4.3-specific antibodies to explore whether this is also the case in mouse cerebellar cortex. Immunohistochemical analysis of Kv4.2 and Kv4.3 protein expression shows a similar pattern of expression for the channel proteins in the mouse cerebellar cortex (Fig. 2). Since there are small but consistent differences in the properties of Kv4.2 and Kv4.3 channels in heterologous expression systems (Serodio et al. 1996), we targeted for electrophysiological recording and analysed separately the currents in cerebellar granule cells from anterior and posterior lobules to characterize either Kv4.2- or Kv4.3-dominated native currents, respectively.
Isolation and characterization of the ISA in cerebellar granule cells
Classical strategies for the isolation of native A-type currents from neurons have used prepulse protocols to inactivate the A-type K+ current, which is then obtained by subtraction of the current remaining after the prepulse from the total current (e.g. Connor & Stevens, 1971; Rudy et al. 1988; Cull-Candy et al. 1989). The outward current in granule cells includes a slowly inactivating component, which increased in amplitude with age and can be large relative to the transient component, particularly in anterior lobules (Fig. 3). We found that in slices from mice older than P12, the classical protocol was not useful to isolate the ISA from cerebellar granule cells, except during relatively small depolarizations, because the prepulse also inactivated a significant amount of the slowly inactivating components of the total outward current. Contamination of the ISA with slowly inactivating components suppressed by the prepulse was particularly large (but not exclusive) for cells in anterior lobules. Similar difficulties were found in previous attempts at isolating the ISA from cerebellar granule cells in rat, and past studies have often focused on characterizing the current during small depolarizations (Cull-Candy et al. 1989).
Cerebellar granule cells prominently express the Kv3 subunit Kv3.1 (Perney et al. 1992; Weiser et al. 1994; Weiser et al. 1995; Sekirnjak et al. 1997). Kv3 channels activate in a voltage range positively shifted compared to Kv4 channels, as is the case with the slowly inactivating component in cerebellar granule cells. It is likely that much of this component of the outward current in these neurons is mediated by Kv3 channels. We therefore utilized relatively low concentrations of TEA (5 mm), a drug that does not block Kv4 channels but completely blocks Kv3 channels (Coetzee et al. 1999), to block the slow inactivating component of the current in cerebellar granule cells and facilitate the isolation of the ISA. This concentration of TEA blocked a major portion of the total outward current in cerebellar granule cells, leaving a highly enriched ISA (Fig. 3C and D). A higher concentration of TEA (20 mm) gave identical results. Since TEA was used for many of the recordings in heterologous cells in order to block a current intrinsic to the tsA-201 cells, recording in the presence of TEA in cerebellar granule cells also allowed a better comparison of the native and the reconstituted currents.
Voltage dependence of the rate of inactivation in cerebellar granule cells
In the presence of TEA a prepulse protocol could be applied to isolate the ISA (Fig. 4). A 1–5 s prepulse to −40 mV inactivated the entire transient component. The remaining current may represent the ‘standing outward current’, a current mediated by two pore K+ channels (K2P), and/or could be mediated by TEA-insensitive non-inactivating Kv1 channels (Millar et al. 2000; Mathie et al. 2003). The current suppressed by the prepulse, obtained by subtraction, inactivated nearly completely (Fig. 4A and B, lower traces), and was used to characterize the properties of the ISA in cerebellar granule cells. This ISA was nearly eliminated by 10 mm 4-AP (Supplementary Fig. 2), a characteristic inhibitor of Kv4-mediated A-type currents at millimolar concentrations (Coetzee et al. 1999). Figure 4E compares the conductance–voltage relation and the voltage dependence of inactivation for the currents recorded in anterior versus posterior lobules, and Fig. 4F, the recovery from inactivation at −120 mV for the two cell types. The properties of the currents in the two cell types are very similar, but showed minor differences, consistent with the differences reported between Kv4.2 and Kv4.3 currents in heterologous cells (Serodio et al. 1996).
We measured the inactivation t1/2 against voltage from a number of neurons in anterior and posterior cerebellar lobules and the results are plotted in Fig. 4C and D. The plots show that the rate of inactivation of the ISA isolated through these procedures has voltage dependence similar to that seen in heterologously expressed Kv4 channels in the presence of KChIP-1 and DPPX-S. This is observed for the currents in both Kv4.2- and Kv4.3-containing granule neurons. The slope of the t1/2–voltage relation (2.6 ms/10 mV and 2.5 ms/10 mV for anterior and posterior lobules, respectively) and the overall profile of the voltage dependence of the inactivation rates are remarkably similar to those observed for the currents expressed by the ternary Kv4 channel containing Kv4.2, KChIP-1 and DPPX-S in heterologous cells.
To confirm that the observed behaviour is not the result of isolating a current containing two or more components with different voltage dependencies inactivating at different rates, we took advantage of the fast rate of recovery of Kv4-mediated transient currents compared to other fast inactivating K+ channels. If the values of the time constants of inactivation as a function of voltage are a feature of a single population of independent channels that recover rapidly from inactivation, we would expect their values to be the same for a partially or a fully recovered set of channels. We applied depolarizing pulses to different membrane potentials (test pulses) following a prepulse to +10 mV applied at different times before the test pulse. Between the two pulses the membrane was held at -110 mV. The t1/2–voltage relations of the currents obtained during the test pulses following different recovery times are shown in Fig. 5C and D, along with those of the currents elicited by the same depolarizations in the absence of a prepulse. The t1/2–voltage relations of fully or partially recovered currents overlap at all membrane potentials and recovery periods, indicating that they reflect a homogeneous population of channels with identical recovery time course.
Kinetic modelling of the voltage dependence of inactivation kinetics in Kv4.2 channels
The unusual voltage dependence of inactivation described here can be explained by assuming preferential closed-state inactivation (Klemic et al. 1998; Kaulin et al. 2008) and a weakly voltage-dependent opening step, a feature proposed for several types of voltage-gated K+ channels (Zagotta et al. 1994; Klemic et al. 1998; Schoppa & Sigworth, 1998; Smith-Maxwell et al. 1998). Thus, if inactivation takes place preferentially from the preopen closed but activated state, further membrane depolarization acting on a weakly voltage-dependent opening step would gradually increase the open probability and hence decrease the probability of residing in the inactivation-permissive preopen closed state. Consequently, the observed rate of macroscopic inactivation decreases at strongly depolarized membrane potentials.
To test these ideas, we applied the Kv4 kinetic model proposed by Kaulin et al. (2008), and assumed that Kv4.2 and Kv4.2:DPPX-S channels undergo open-state inactivation, whereas the presence of KChIP-1 precludes open-state inactivation in Kv4.2:KChIP-1 and ternary Kv4.2 channels (Methods; Fig. 6A). Thus, Kv4.2 channel complexes that include KChIP-1 undergo preferential closed-state inactivation. Under these conditions, the kinetic model accounted for the observations from all subunit combinations closely (Fig. 6B and Table 1). In particular, this model produced excellent descriptions of the t1/2–voltage relations (derived from the best-fit currents; Fig. 7A). As expected, the rate of inactivation increases with depolarization and levels off when there is open-state inactivation strictly coupled to activation; and conversely, the rate of inactivation decreases with depolarization when channels undergo preferential closed-state inactivation. In the latter case, the weakly voltage-dependent opening equilibrium is critical because otherwise the inactivation rate exhibits little or no voltage dependence (Fig. 7B). The theoretical t1/2–voltage relation of the ternary Kv4.2 complex with a slope of 2.9 ms/10 mV recapitulates the behaviour of native ISA in cerebellar granule neurons (2.6 ms/10 mV). Moreover, the model produced an excellent description of the recoveries from inactivation for all subunit combinations (Fig. 7C).
Discussion
Here, we showed that the rate of inactivation of the ISA in mouse cerebellar granule neurons has unusual voltage dependence that appears to be a property of a homogeneous population of channels and resembles the voltage dependence of the rate of inactivation of Kv4.2 channels expressed in heterologous mammalian cells containing the accessory subunits KChIP-1 and DPPX-S.
ISA channels in mouse cerebellar granule cells are likely ternary complexes containing accessory subunits KChIP and DPPX
Using biochemical methods, KChIPs and DPPLs were initially discovered as proteins that associate with Kv4 proteins, the pore-forming subunits mediating most of the somatodendritic ISA in neurons. KChIPs were isolated in a yeast two-hybrid screen that used the amino terminus of Kv4.2 as bait (An et al. 2000). Association of KChIPs with native Kv4 channel complexes in brain was subsequently demonstrated by co-immunoprecipitation. Association of DPPX with Kv4 channels in brain tissue was discovered by immunopurification of Kv4 channel complexes from rat cerebellar membranes using Kv4.2 antibodies (Nadal et al. 2003). These channel complexes were also shown to contain KChIP proteins (Nadal et al. 2003).
KChIPs and DPPX (or its homologue DPP10) have powerful effects on the expression and biophysical properties of Kv4 channels expressed in heterologous cells (An et al. 2000; Nadal et al. 2003, 2006; Jerng et al. 2005; Ren et al. 2005; Zagha et al. 2005; Dougherty & Covarrubias, 2006; Jerng et al. 2007). KChIPs and DPPLs facilitate the transport of Kv4 channels to the plasma membrane, which are otherwise retained in intracellular compartments. These proteins also have major effects on the biophysical properties of the channels. KChIPs slow down and DPPLs accelerate the rate of inactivation, while both accessory subunits speed up recovery from inactivation. In addition, DPPLs produce large shifts in the voltage dependence of activation and inactivation. The currents recorded in heterologous cells expressing the ternary complex containing Kv4 pore forming subunits as well as KChIPs and DPPLs more closely resemble several properties of the ISA in various neuronal populations (Nadal et al. 2003; Jerng et al. 2005, 2007; Zagha et al. 2005), supporting the notion that the native channels mediating the somatically recorded ISA contain these accessory proteins. In addition, KChIPs and DPPX are expressed prominently in neuronal brain populations that also express significant levels of Kv4.2 or Kv4.3 proteins, and with similar subcellular distributions (Nadal et al. 2003; Rhodes et al. 2004; Strassle et al. 2005; Zagha et al. 2005) (B. Clark, E. Kwon & B. Rudy, unpublished observations). Furthermore, in Kv4.2 knock-out mouse there is a region- and cell-specific down-regulation of individual KChIP auxiliary subunits (Menegola & Trimmer, 2006) and DPPX in brain (J. Trimmer, personal communication).
Here, we show that the voltage dependence of the rate of inactivation in cerebellar granule neurons closely matches the voltage dependence of the inactivation rate of the currents observed in heterologous mammalian cells expressing the ternary complex containing Kv4.2, KChIP-1 and DPPX-S. KChIP-1 is largely responsible for the large decrease in the rate of inactivation during large depolarizations. However, the studies in heterologous cells demonstrate that the voltage dependence of the inactivation rate is different in Kv4.2-KChIP channels in the presence or absence of DPPX-S (Fig. 1). The t1/2–voltage relation for the ISA in cerebellar granule neurons is very similar to that obtained in heterologous cells expressing both accessory proteins (compare Figs 4 and 5 with Fig. 1). The reconstitution of this complex and unique behaviour (among A-type K+ currents) by the ternary complex containing Kv4, DPPX and KChIP proteins provides additional strong evidence in favour of somatically recorded ISA being mediated by Kv4 channels that contain both types of accessory proteins.
ISA channels in mouse cerebellar granule neurons inactivate preferentially from closed states
The mechanisms of inactivation of Kv4-mediated ISA channels remain poorly understood and seem to be mechanistically different from those in other inactivating voltage-gated K+ channels (Jerng et al. 2004a). Previous studies have proposed closed-state inactivation as a significant pathway of inactivation in Kv4 channels (Jerng et al. 1999; Bahring et al. 2001; Beck & Covarrubias, 2001; Beck et al. 2002; Shahidullah & Covarrubias, 2003; Kaulin et al. 2007). Although Kv4 channels expressed without KChIPs exhibit a fast open-state N-type-like mechanism of inactivation (Gebauer et al. 2004), KChIPs preclude this mechanism because they sequester the N-terminal inactivation gate (Beck et al. 2002; Pioletti et al. 2006; Wang et al. 2007). It is therefore believed that Kv4 channel complexes that include KChIPs inactivate preferentially from closed states (Callsen et al. 2005; Kaulin et al. 2008). However, the importance of this pathway in Kv4 channels has remained controversial (Patel & Campbell, 2005; S. Wang et al. 2005).
The voltage dependence of the inactivation rate described here is predicted by kinetic models of closed-state inactivation (Klemic et al. 1998, 2001; Kaulin et al. 2007). The rate of inactivation decreases with depolarization once voltage-dependent activation is complete because further depolarization promotes the weakly voltage-dependent opening step, which decreases the probability of channels residing in the inactivation-permissive preopen closed state. A weakly voltage-dependent opening step is therefore critical in this mechanism. The observation that the inactivation rates of the ISA in cerebellar granule neurons decrease with depolarization at positive membrane potentials further supports the view that native Kv4 channels also inactivate preferentially from closed-states.
Closed-state inactivation in Kv4 channels is probably the mechanism that regulates the availability of the channels at subthreshold membrane potentials, and is optimally fast in this range of membrane potentials. Although at typical neuronal resting membrane potentials (−60 mV) these channels are mostly inactivated at steady state, a brief conditioning hyperpolarization will recruit them because their recovery from inactivation is fast. Then, activation of Kv4 channels by a subsequent depolarization will delay firing of the action potential. The characteristically fast voltage-dependent recovery from inactivation of native Kv4 channel complexes distinguishes them from other Kv channels that inactivate primarily via classical N-type and P/C-type mechanisms. This hallmark of Kv4 channel complexes might be a direct consequence of the mechanism responsible for closed-state inactivation, which remains unknown. The slowing of inactivation with depolarization may confer another role. By inactivating more slowly at depolarized membrane potentials, Kv4 channels could have a greater impact on spike repolarization.
Acknowledgements
We wish to thank Brian Clark and Eddie Zagha for their critical review of this manuscript. This work was supported by National Institutes of Health grants NS045217 and NS30989 and NSF grant IBN-0314645 to B.R., National Institutes of Health grant NS032337 and Research Enhancement Award (REA) from Thomas Jefferson University to M.C.; and NIH training grant T32 AA007463 to K.D.
Footnotes
-
(Resubmitted 26 December 2007; accepted after revision 12 February 2008; first published online 14 February 2008)
-
This paper has online supplemental material.
References
Figure 1. Modulation of the inactivation kinetics of Kv4.2 channels by associated proteins KChIP1 and DPPX-S A–D, whole-cell K+ currents from tsA-201 cells transfected with Kv4.2 (A), Kv4.2 and DPPX-S (B), Kv4.2 and KChIP1 (C) or Kv4.2 + KChIP1 and DPPX-S (D), evoked by step depolarizations from −109 to +61 mV in 10 mV increments from a holding potential of −149 mV. Note that DPPX accelerates and KChIP1 slows down inactivation kinetics. E, voltage dependence of the rate of inactivation of Kv4.2 channels. Plots of the half-inactivation time (t1/2) against membrane potential from the currents recorded in tsA-201 cells expressing Kv4.2 alone (•), Kv4.2 and DPPX-S (○), Kv4.2 and KChIP1 (▵) and Kv4.2 plus KChIP1 and DPPX-S (grey triangles). Shown are means s.e.m. (n = 4, 4, 10 and 10, respectively). Note that, for currents recorded in the presence of KChIP1 and KChIP1 plus DPPX-S, the rate of inactivation increases with strong depolarizations. However, DPPX accelerates the rate of inactivation between −60 and −10 mV, increasing the voltage range over which the t1/2 increases with progressive depolarization. F, recoveries from inactivation of Kv4.2 alone (•), Kv4.2 and DPPX-S (○), Kv4.2 and KChIP1 (▵) and Kv4.2 plus KChIP1 and DPPX-S (grey triangles); the recovery voltages were −115, −128, −140 and −140 mV, respectively. These voltages were adjusted to account for shifts in the voltage dependence of steady-state inactivation (not shown), and therefore are 40 mV more negative than the midpoint voltage of the corresponding steady-state inactivation curve. The curves shown are means s.e.m. (n = 3, 3, 8 and 9, respectively) and the lines are best-fit single exponentials with the following time constants: 165, 40, 96 and 45 ms, for Kv4.2, Kv4.2:DPPX-S, Kv4.2:KChIP-1 and Kv4.2:KChIP-1:DPPX-S, respectively.
Figure 2. Immunolocalization of Kv4.2 and Kv4.3 proteins in the mouse cerebellar cortex Immunostaining of sagittal sections of the mouse cerebellum with antibodies to Kv4.2 and Kv4.3. The figure shows the Cy3 fluorescence signal. Kv4.2 is expressed predominantly in the granule cell layer (GCL); with prominent staining in anterior, but weak staining in posterior lobules. Kv4.3 is strongly expressed in the molecular layer (MoL), where the dendrites of Purkinje cells are brightly stained. In the granule cell layer (GCL) Kv4.3 immunostaining is prominent in posterior lobules and extremely weak in anterior lobules. Insets show higher magnification images of the GCL illustrating Kv4.2 and Kv4.3 labelling in the periphery of granule cell somata and in the glomeruli containing the dendritic processes of the granule cells. Scale bar: 800 μm; insets: 30 μm.
Figure 3. K+ currents in cerebellar granule cells Whole-cell ionic currents recorded in a representative cerebellar granule cell from an anterior (A and C) or a posterior (B and D) lobule, before (A and B) or after (C and D) the application of 5 mm TEA. Currents were recorded in the presence of 500 nm TTX and were elicited by step depolarizations from −90 to +60 mV in 10 mV increments from a holding potential of −130 mV.
Figure 4. Inactivation kinetics of ISA in cerebellar granule cells A and B, whole-cell K+ currents recorded in a representative cerebellar granule cell from an anterior (A) or a posterior (B) lobule recorded in the presence of 5 mm TEA. Shown are the currents elicited by step depolarizations from −90 to +60 mV in 10 mV increments, preceded by a prepulse to −100 mV from a holding potential of −130 mV (upper traces) or a prepulse to −40 mV (middle traces). The lower traces show the transient currents (ISA) obtained by subtracting the currents recorded during depolarizing steps preceded by a prepulse to −40 mV from the currents obtained during depolarizing steps preceded by a prepulse to −100 mV. C and D, half-inactivation time (t1/2) as a function of membrane potential of the ISA obtained in cerebellar granule cells from anterior (C) and posterior (D) lobules. Shown are means s.e.m. (n = 17 for anterior and n = 6 for posterior lobules). E, normalized peak conductance–voltage relations (G/Gmax) and steady-state inactivation curves (I/Imax) for granule cells recorded in anterior (squares) and posterior (circles) cerebellar lobules. Peak conductance (G) was calculated as G = Ip/(Vm − Veq), where Ip, Vm and Veq are the peak current, the test potential and the K+ equilibrium potential, respectively. Shown are means s.e.m. (n = 21 for anterior and n = 6 for posterior lobules). The continuous lines across the data points are the best-fits to Boltzmann functions with a V1/2 = −74 and −82 mV a slope factor k = 6.9 and 5.9 mV for the inactivation curves in anterior and posterior lobules, respectively; and a V1/2 = −15 and −21 mV a slope factor k = 22 and 19 mV for the conductance–voltage curves in anterior and posterior lobules, respectively. F, recovery from inactivation of ISA at −130 mV in cells recorded from anterior (squares) and posterior (circles) lobules. Shown are means s.e.m. (n = 11 for anterior and n = 6 for posterior lobules). The traces are single exponential fits through the data.
Figure 5. The voltage dependence of the rate of inactivation of the ISA in cerebellar granule cells is a property of a homogeneous population of channels A and B, representative records of the currents used to determine inactivation rates. Shown are the currents recorded during a test pulse to −30 mV (A) or +20 mV (B) preceded by a prepulse to +10 mV applied 10, 50 or 500 ms before the test pulse. Between the prepulse and the test pulse the cell was held at −110 mV. C and D, half-inactivation time (t1/2) as a function of membrane potential of the ISA obtained in cerebellar granule cells from anterior (C) and posterior (D) lobules during test pulses to the indicated voltages after a 10 ms (diamonds), 50 ms (triangles), or 500 ms (circles) recovery period at −110 mV from the inactivation produced by a prepulse to +10 mV. Black squares are the t1/2 for the currents recorded during identical test pulses in the absence of a prepulse. Shown are means s.e.m. (n = 4 for anterior and n = 3 for posterior lobules).
Figure 6. Kinetic modelling of Kv4.2 currents A, kinetic scheme for Kv4.2 channel gating based on the models proposed by Kaulin et al. (2008). To obtain the model's best-fit for each condition, the analysis was constrained simultaneously by all traces in the average family of currents, the average steady-state inactivation curve (not shown) and the average recovery from inactivation (Fig. 1F and Methods). The open-state inactivation pathway is enclosed in a box (dashed lines). B and C, families of observed (B) and best-fit (C) currents elicited by step depolarizations from −99 to +41 mV in 20 mV increments from a holding potential of −149 mV. The number of averaged families of observed currents is indicated in the plots. The best-fit parameters are shown in Table 1.
Figure 7. Kinetic modelling of Kv4.2 inactivation A, plots of half-inactivation time against voltage (t1/2–voltage relation) for simulated currents mediated by Kv4.2 channels with different subunit compositions. The t1/2 was obtained from the best-fit currents shown in Fig. 6C. B, effect of voltage dependence of the opening equilibrium on the t1/2–voltage relation. In one case (•), the rate constants ε and φ are voltage-dependent (Fig. 6A; Table 1), and in the other (continuous line), these rate constants are assumed to be voltage-independent. All other best-fit parameters are kept as depicted in Table 1. C, best-fit recoveries from inactivation of Kv4.2 channels with different subunit compositions. Continuous lines are best single exponential fits with the following time constants: 156 ms (Kv4.2); 47 ms (Kv4.2:DPPX-S); 99 ms (Kv4.2:DPPX-S); and 49 ms (Ternary). These values are in excellent agreement with the experimental observations (Fig. 1 legend).
Table 1. Best-fit parameters of the kinetic scheme for Kv4 channel complexes
