The action potential (AP) is a fundamental feature of excitable cells that serves as the basis for long-distance signaling in the nervous system. There is considerable diversity in the appearance of APs and the underlying repolarization mechanisms in different neuronal types (reviewed in Bean BP. Nat Rev Neurosci 8: 451–465, 2007), including among pyramidal cell subtypes. In the present work, we used specific pharmacological blockers to test for contributions of Kv1, Kv2, or Kv4 channels to repolarization of single APs in two genetically defined subpopulations of pyramidal cells in layer 5 of mouse somatosensory cortex (etv1 and glt) as well as pyramidal cells from layer 2/3. These three subtypes differ in AP properties (Groh A, Meyer HS, Schmidt EF, Heintz N, Sakmann B, Krieger P. Cereb Cortex 20: 826–836, 2010; Guan D, Armstrong WE, Foehring RC. J Neurophysiol 113: 2014–2032, 2015) as well as laminar position, morphology, and projection targets. We asked what the roles of Kv1, Kv2, and Kv4 channels are in AP repolarization and whether the underlying mechanisms are pyramidal cell subtype dependent. We found that Kv4 channels are critically involved in repolarizing neocortical pyramidal cells. There are also pyramidal cell subtype-specific differences in the role for Kv1 channels. Only Kv4 channels were involved in repolarizing the narrow APs of glt cells. In contrast, in etv1 cells and layer 2/3 cells, the broader APs are partially repolarized by Kv1 channels in addition to Kv4 channels. Consistent with their activation in the subthreshold range, Kv1 channels also regulate AP voltage threshold in all pyramidal cell subtypes.
- somatosensory cortex
- potassium channel
the action potential (AP) is a fundamental feature of neurons that serves as the basis for long-distance signaling in the nervous system. The number, frequency, and timing of APs can serve as a neural code (Larkum et al. 1999b; Mainen and Sejnowski 1995; Singer 2001; Williams and Stuart 2000). In neocortical pyramidal neurons, APs are initiated in the axonal initial segment and then propagate along the axon as well as backpropagate into the soma and dendrites (Kole and Stuart 2012; Martina et al. 2000; Palmer and Stuart 2006; Shu et al. 2007; Stuart et al. 1997). In general, the time course of the AP determines the temporal pattern of voltage changes that gate voltage-dependent ionic conductances and the shape of the AP regulates Ca2+ entry (e.g., Bean 2007; Stewart and Foehring 2001). In turn, intracellular Ca2+ concentration ([Ca2+]i) controls many cellular processes, including activation of Ca2+-dependent conductances (e.g., Andrade et al. 2012; Schwindt et al. 1988b), activation of enzymes and signaling pathways, and induction of synaptic plasticity. The basics of AP form are similar in most neurons (Bean 2007): the interplay between Na+ channels, leak channels, and voltage-gated K+ channels regulates voltage threshold (which is dynamic), the AP upstroke is primarily due to gating of voltage-gated Na+ channels, and repolarization of the AP is dominated by K+ conductances and inactivation of Na+ conductance. Repolarization mechanisms may also contribute to the fast afterhyperpolarization.
There is considerable diversity, however, in the appearance of APs and underlying repolarization mechanisms in different neuronal types (reviewed by Bean 2007). For example, fast-spiking cells (e.g., GABAergic interneurons in cortical regions: Rudy et al. 1999; neurons in auditory brain stem: Gan and Kaczmarek 1998; Wang et al. 1998; cerebellar Purkinje cells: Martina et al. 2007) are repolarized by high-threshold Kv3 channels with rapid activation/deactivation kinetics. Cortical pyramidal neurons typically do not express Kv3 channels (Martina et al. 1998; Massengill et al. 1997; Rudy et al. 1999) and thus must use different mechanisms to repolarize the AP. For example, Ca2+-dependent BK channels are important for AP repolarization in somas of CA1 pyramidal cells (Poolos and Johnston 1999; Storm 1987) but do not play such a role 1) in distal dendrites of CA1 pyramidal cells (Poolos and Johnston 1999) or 2) in neocortical pyramidal cell somas (Guan et al. 2015; Lorenzon and Foehring 1992; Pineda et al. 1998; Schwindt et al. 1988a, 1988b).
There is also considerable variability in AP properties among neocortical pyramidal cells. Large layer 5 pyramidal cells from the motor cortex of cats (Betz cells) have very narrow APs, and repolarization is due to multiple voltage-gated K+ channels (which have not been characterized as to their genetic or molecular basis; Schwindt et al. 1988a, 1988b). In rodent neocortex, pyramidal neurons have been classified on the basis of projection target as IT type (intratelencephalic: project within the telencephalon) or PT type (project beyond the telencephalon, including the pyramidal tract; Dembrow et al. 2010; Reiner et al. 2003; Shepherd 2013; Suter et al. 2013). A consistent finding is that PT-type cells have narrower APs with more rapid repolarization, whereas IT-type cells have broader APs with slower repolarization (reviewed in Guan et al. 2015; Shepherd 2013).
Neocortical pyramidal cells can also be classified by genetic differences. For example, consider two lines of mice that express enhanced green fluorescent protein (EGFP) in cells expressing the genes etv1 or glt (Doyle et al. 2008; Gong et al. 2007). In mouse somatosensory cortex, etv1 cells are a subset of IT-type neurons located primarily in superficial layer 5 and glt cells are a subset of PT-type neurons, primarily located in deep layer 5 (Bishop et al. 2015; Groh et al. 2010; Guan et al. 2015). glt cells have narrower and more rapidly repolarizing APs than etv1 or layer 2/3 cells (Groh et al. 2010; Guan et al. 2015). Morphology and projection targets also differ for these cell populations (Groh et al. 2010). Most previous work on the ionic basis for AP repolarization in rodent central neurons used relatively nonspecific pharmacological agents [e.g., 4-aminopyridine (4-AP), tetraethylammonium (TEA)] and thus could not definitively identify the roles of specific channel types. Pyramidal neurons express several types of Kv channels that could potentially shape AP properties. There are transient and persistent components to outward K+ currents in rodent pyramidal neurons (Bekkers 2000; Foehring and Surmeier 1993; Korngreen and Sakmann 2000; Locke and Nerbonne 1997). The persistent currents are due to strong expression of Kv1, Kv2 (the largest component), and Kv7 channels (Bekkers and Delaney 2001; Bishop et al. 2015; Guan et al. 2006, 2007a, 2007b, 2011a, 2011b, 2013, 2015; Murakoshi and Trimmer 1999). In addition, Nerbonne and colleagues used genetic approaches to reveal that most of the transient, A-type current in rat neocortical cells is due to Kv4.2 and Kv4.3 channels (Carrasquillo et al. 2012; Norris and Nerbonne 2010; Yuan et al. 2005; see also Guan et al. 2011b), with a contribution from Kv1.4 channels.
We previously found that in layer 2/3 pyramidal neurons from rats Kv1 channels did not affect AP width or repolarization (Guan et al. 2007a), and we used a dominant-negative approach to show that Kv2.1-containing channels did not play a role in AP repolarization in rat neocortical pyramidal neurons (Guan et al. 2013). We also showed that AP repolarization in mouse layer 2/3, etv1, or glt cells is not Ca2+ dependent (Guan et al. 2015), consistent with previous studies on neocortical pyramidal cells (Lorenzon and Foehring 1992; Pineda et al. 1999; Schwindt et al. 1988a, 1988b). In addition, the kinetics of Kv7 channels are too slow to influence single APs of neocortical pyramidal cells (layer 2/3: Guan et al. 2011a) and other pyramidal cell types (Aiken et al. 1996; Gu et al. 2007; Hu et al. 2007; Marrion 1997; Mateos-Aparicio et al. 2014; Prescott and Sejnowski 2008; Storm 1989; Yue and Yaari 2004). On the basis of these findings, we hypothesized that Kv4 channels may be the primary channels involved in repolarizing APs in neocortical pyramidal cells. We also asked whether the underlying mechanisms are pyramidal cell subtype dependent.
We studied layer 2/3 pyramidal neurons and layer 5 pyramidal neurons from two bacterial artificial chromosome lines of mice, each of which expresses EGFP in a different subpopulation of layer 5 pyramidal neurons (Gong et al. 2002, 2003, 2007; Guan et al. 2015). In the somatosensory cortex of Tg(Etvl-EGFP)BZ192Gsat/Mmucd (etv1) mice, EGFP is primarily expressed in pyramidal neurons from superficial layer 5 (Groh et al. 2010; Guan et al. 2015). The etv1 gene is a transcription factor that has been shown to be involved in neurogenesis in the olfactory bulb (Stenman et al. 2003) and circuit formation in the spinal cord (Arber et al. 2000). In somatosensory cortex of Tg(Glt25d2-EGFP)BN20Gsat/Mmnc (Glt) mice, glt-EGFP is primarily expressed in a subset of deep layer 5 pyramidal neurons (Groh et al. 2010; Guan et al. 2015), although some EGFP+ cells were also observed in superficial layer 5. The glt gene is a glycosyl transferase (Gong et al. 2003, 2007). We maintain breeding colonies of both mouse lines (Swiss-Webster background), which were originally obtained from the Mutant Mouse Regional Resource Centers (MMRRC) of the GENSAT project.
The present studies were performed on juvenile mice from 2 to 4 wk of age. All procedures were approved by the Animal Care and Use Committee, University of Tennessee Health Science Center. The animals were anesthetized with isoflurane until they were areflexive. Briefly, the animal was placed into a sealed plastic container into which gauze soaked with isoflurane was placed under a fiberglass screen floor. After anesthesia with isoflurane, the animal was decapitated and the brain was removed and dropped into ice-cold cutting solution bubbled with O2 for 30–60 s. This solution contained (in mM) 250 sucrose, 2.5 KCl, 1 NaH2PO4, 11 glucose, 4 MgSO4, 0.1 CaCl2, 0.4 ascorbate, 0.6 sodium pyruvate, and 15 HEPES (pH 7.3–7.4, 300 mosM). The brain was then sliced into 300-μm-thick coronal sections with a vibrating tissue slicer (Vibroslice, Campden Instruments).
Slices were placed in a recording chamber on the stage of an Olympus BX50WI upright microscope and bathed in artificial cerebrospinal fluid (aCSF) bubbled with 95% O2-5% CO2, delivered at 2 ml/min, and heated with an in-line heater (Warner Instruments, Hamden, CT) to 33 ± 1°C (measured with a thermistor in the bath adjacent to the slice). The aCSF contained (in mM) 125 NaCl, 3 KCl, 2 CaCl2, 2 MgCl2, 1.25 NaH2PO4, 26 NaHCO3, and 20 glucose (pH 7.4, 310 mosM). For most of the experiments with toxins, we used a recirculating bath system in which a peristaltic pump (Gilson Minipuls 3) was used for controlling both the inflow (∼2 ml/min) and the outflow of carbogenated aCSF.
All slice current-clamp recordings were done in the presence of 6,7-dinitroquinoxaline-2,3-dione (DNQX, 20 μM) to block α-amino-3-hydroxy-5-methylisoxazole-propionic acid (AMPA) receptors, d-(−)-2-amino-5-phosphonopentanoic acid (D-AP5, 50 μM) to block N-methyl-d-aspartate (NMDA) receptors, and picrotoxin [100 μM; to block γ-aminobutyric acid type A (GABAA) receptors] to minimize the influence of effects of the pharmacological agents on fast synaptic transmission presynaptic to the recorded cell. Pharmacological agents were added directly to the aCSF. Most reagents were purchased from Sigma-Aldrich (St. Louis, MO). In addition, α-dendrotoxin (DTX) and guangxitoxin-1E (GxTx) were purchased from Alomone Labs (Jerusalem, Israel), DNQX, D-AP5, and picrotoxin were purchased from Tocris Bioscience (Ellisville, MO), and AmmTx3 was purchased from Smartox Biotechnology (Saint-Martin-d'Hères, France). For experiments with DTX, GxTx, or AmmTx3, 0.2% bovine serum albumin was added to saturate nonspecific binding sites.
Pyramidal neurons in layer 2/3 and layer 5 were visualized with infrared-differential interference contrast (IR-DIC) videomicroscopy (Dodt and Zieglgänsberger 1990; Stuart et al. 1993) using a ×40 (0.8 NA) Olympus water-immersion objective and an IR-sensitive camera (Olympus OLY-150 or DAGE-MTI). etv1 or glt pyramidal cells were visually identified by the presence of EGFP epifluorescence with an FITC filter. Layer 5 recordings were directed within the main band of EGFP+ cells in layer 5 in each animal. In etv1 neurons, recordings were biased toward the most superficial EGFP+ cells, and for glt neurons recordings were biased toward deeper EGFP+ cells. We previously showed that in somatosensory cortex, etv1 neurons have a more superficial (and narrower) expression pattern in layer 5. glt neurons are found deeper in layer 5 (Guan et al. 2015), but there is overlap with the etv1 distribution (Bishop et al. 2015; Guan et al. 2015). We switched between IR-DIC and epifluorescence to determine cell type and to obtain a gigaohm seal. Electrode position was controlled with Sutter ROE-200 manipulators and PC-200 controller or Luigs-Neumann manipulators and controller. Whole cell patch-clamp recordings were acquired with an Axon Multiclamp 700A or Multiclamp 700B amplifier (Molecular Devices, Sunnyvale, CA) and pCLAMP 9 or 10 software. For current-clamp recordings, the data were digitized at 20–50 kHz and filtered at 10 kHz. We recorded with borosilicate electrodes (Warner G150F; 3–8 MΩ in the bath) produced with a horizontal electrode puller (Flaming-Brown P-87; Sutter Instruments, Novato, CA). For current- and voltage-clamp recordings, electrodes were filled with an internal solution containing (in mM) 130.5 K-methylsulfate (KMeSO4), 10 KCl, 7.5 NaCl, 2 MgCl2, 10 HEPES, 2 ATP, and 0.2 GTP; 100 μM EGTA was added to the intracellular solution for current-clamp recordings (5 mM EGTA for voltage clamp). Data were collected only from cells forming a 1-GΩ or tighter seal. For current-clamp recordings, we optimized bridge balance and capacitance compensation. All reported voltages were corrected by subtracting the measured liquid junction potential (8 mV). We measured AP amplitude [from resting membrane potential (RMP)], half-width [HW; width at (peak − RMP)/2], voltage threshold (Vth; voltage at the time when the rapid increase in positive dV/dt occurs), width at voltage threshold, and the maximal rate of AP polarization and repolarization (dV/dt up and down, respectively).
Voltage-clamp data were obtained from outside-out macropatches at 33 ± 1°C (see Bishop et al. 2015 for detailed methods). These patches were formed after break-in to whole cell mode by withdrawing the pipette without additional suction. The final patch capacitance was typically 3–4 pF, and series resistance was 5–9 MΩ. Data were digitized at 20 kHz and filtered at 10 kHz. Series resistance was not compensated for, and linear leak and capacitance were subtracted off-line. Tetrodotoxin (0.5 μM) was added to the extracellular solution to block voltage-gated Na+ conductance. Holding potential was −75 mV. The membrane was stepped to −100 mV for 1 s, and then blocker selectivity was tested with 500-ms voltage steps from −100 mV to +10 mV.
Prism (GraphPad Software, San Diego, CA) software was used to perform statistical tests. We used a one-way ANOVA to compare multiple experimental groups (etv1, glt, layer 2/3), with post hoc Tukey's multiple-comparisons tests to determine which individual means differed. Paired t-tests were used to compare control vs. drug effects. For all statistical tests, P values <0.05 were considered to be significantly different. Summary data are presented as means ± SE, unless noted otherwise. Some data are presented as box plots (Tukey 1977). Box plots indicate the upper and lower quartiles as edges of the box, with the median represented as a line crossing the box; the stems indicate the largest and smallest nonoutlying values.
We recorded from two genetically defined subpopulations of pyramidal cells in layer 5 of mouse somatosensory cortex (etv1 and glt) as well as pyramidal cells from layer 2/3. Pyramidal cells that fulfilled our criteria of RMP negative to −60 mV and an overshooting AP with amplitude ≥70 mV were included in the study (Table 1). The RMP in control vs. drug was matched to within 1 mV by DC injection in drug solution, and representative single APs were compared before drug application and after stabilization of drug effects. Consistent with our previous findings (Guan et al. 2015), we found that 1) glt pyramidal cells have narrower APs, significantly higher rates of AP polarization and repolarization (greater dV/dt up and down), and lower Vth vs. the other groups and 2) layer 2/3 pyramidal cells have more negative RMP and higher input resistance than either layer 5 cell type (and AP width and dV/dt similar to etv1 cells). etv1 and glt cells did not differ from one another in input resistance or RMP. etv1 cells had significantly smaller AP amplitude than glt cells (P < 0.001). glt cells had the narrowest APs, and layer 2/3 and etv1 cells were similar in HW (Table 1).
In the present study, we focused on effects of Kv1, Kv2, and Kv4 channels, using several specific pharmacological agents to test for functional roles of these channels in AP repolarization. These include DTX (at 100 nM) for Kv1 channels (channels containing Kv1.1, Kv1.2, or Kv1.6 subunits; Harvey and Robertson 2004), GxTx (at 100 nM) for Kv2 channels (Herrington et al. 2006; Liu and Bean 2014), and the peptide AmmTx3, a highly selective blocker of Kv4 channels in association with dipeptidyl peptidase-like (DPP) proteins (Maffie et al. 2013). We also tested the commonly used putative A-type current blockers, 4 mM 4-AP and 150 μM BaCl2.
Single APs were elicited with a 5-ms duration, just-suprathreshold DC current injection (adjusted so the AP comes at or near the termination of the current injection). The bridge was balanced and electrode capacitance compensated with the digital interface for the Multiclamp amplifier. These parameters were monitored and adjusted throughout the experiment. We used DC current to keep the RMP at −73 ± 1 mV for all cell types (we readjusted DC current after application of pharmacological agents to maintain this RMP). All recordings were performed in the presence of synaptic blockers (see methods) to ensure that the effects were not due to altering presynaptic fast synaptic transmission. We observed no effects of the combined synaptic blockers alone on AP parameters in neocortical pyramidal cells (4 layer 2/3 cells, 12 etv1 cells, and 11 glt cells; data not shown).
Role for Kv1 channels.
To test for a role for Kv1 channels in spike repolarization, we applied 100 nM DTX in the bath solution. DTX had no effects on RMP [layer 2/3 (10 cells): −77 ± 2 mV control vs. −74 ± 3 mV DTX; etv1 (9 cells): −70 ± 2 mV control vs. −70 ± 2 mV DTX; glt (11 cells): −70 ± 1 mV control vs. −71 ± 1 mV DTX] or input resistance [layer 2/3 (9 cells): 137 ± 24 MΩ control vs. 140 ± 30 MΩ DTX; etv1 (7 cells): 77 ± 10 MΩ control vs. 71 ± 6 MΩ DTX; glt (10 cells): 88 ± 10 MΩ control vs. 93 ± 11 MΩ DTX] for any cell type. In mouse layer 2/3 pyramidal cells, DTX caused significant broadening of the AP (AP HW: P < 0.0001, Figure 1, A and B) width at threshold (P < 0.008, Table 2) and slowed the rate of AP repolarization (dV/dt down: P < 0.007; Table 2). The average change in AP HW was 21 ± 3% (n = 10 cells). DTX also had a small but significant hyperpolarizing effect on voltage threshold in layer 2/3 pyramidal cells (Fig. 1C, P < 0.04). AP amplitude and dV/dt for the AP upstroke were not affected in layer 2/3 cells by DTX (Table 2).
In etv1 pyramidal cells, there was a similar significant increase in AP HW (25 ± 6% change, P < 0.004, n = 8) and width at threshold (P < 0.008) and a significant decrease in dV/dt for repolarization (P < 0.008; Fig. 1, D and E, Table 2). There was also a small but significant hyperpolarizing shift in voltage threshold for the AP in etv1 cells (P < 0.02; Table 2, Fig. 1F). AP amplitude was unchanged after DTX (Fig. 1D). In glt cells there was a significant negative shift in voltage threshold for the AP in DTX (P < 0.03, Fig. 1I), but, in contrast to etv1 and layer 2/3 cells, no significant effects on AP HW, width at threshold, or dV/dt for repolarization were observed in glt cells (Fig. 1, G and H, Table 2).
Role for Kv2 channels.
GxTx was found to be a relatively selective blocker of all Kv2 subtypes by Herrington et al. (2006). They found that in pancreatic beta cells 40–100 nM GxTx caused spike broadening and increased intracellular Ca2+ levels. In that preparation the IC50 for Kv2 channels was 10–50 times lower than for Kv4. Similarly, Liu and Bean (2014) observed <10% block of the Kv4-mediated A current in CA1 pyramidal neurons with 100 nM GxTx (the A current is Kv4 mediated in CA1 pyramidal cells; Kim et al. 2005). We found that 100 nM GxTx had no effects on RMP [layer 2/3 (10 cells): −71 ± 3 mV control vs. −70 ± 3 mV GxTx; etv1 (13 cells): −69 ± 2 mV control vs. −68 ± 2 mV GxTx; glt (11 cells): −69 ± 2 mV control vs. −70 ± 2 mV GxTx] or input resistance [layer 2/3 (9 cells): 101 ± 19 MΩ control vs. 103 ± 19 MΩ GxTx; etv1 (11 cells): 110 ± 15 MΩ control vs. 124 ± 17 MΩ GxTx; glt (11 cells): 94 ± 15 MΩ control vs. 100 ± 16 MΩ GxTx] for any of the cell types we tested. We also found that 100 nM GxTx had no effect on single APs in any pyramidal cell subtype (Fig. 2, Table 3). We previously showed that GxTx did block Kv2-mediated currents in mouse neocortical pyramidal neurons at this dose (Bishop et al. 2015). These data indicate little to no role for Kv2 channels in AP repolarization in neocortical pyramidal cells (at least for a single AP in response to a brief current injection; the function of Kv2 channels becomes more important with high-frequency firing: Du et al. 2000; Guan et al. 2013; Johnston et al. 2008; Liu and Bean 2014). The lack of effect of GxTx is consistent with our previous conclusion, based on use of genetic manipulation of Kv2.1 channels in rat pyramidal cells, that Kv2 channels play little role in AP repolarization (Guan et al. 2013).
Role for Kv4 channels.
Given the effects of Kv1 blockers on AP width and repolarization (in layers 2/3 and etv1 only) and the lack of effect of Kv2 blockers, we hypothesized that the dominant outward current involved in AP repolarization in mouse neocortical pyramidal cells would be due to Kv4 channels. A major role for Kv4 channels in AP repolarization would confirm the findings of Nerbonne and colleagues for unidentified layer 5 pyramidal cells from visual cortex of Kv4.2- and Kv4.3-knockout mice (Carrasquillo et al. 2012; Nerbonne et al. 2008; Norris and Nerbonne 2010).
Our first approach was to examine APs before and after application of 4 mM 4-AP, since this readily available agent has been shown to block Kv4 channels and A-type current in many cell types (e.g., reviewed in Coetzee et al. 1999; Hille 2001). Application of 4 mM 4-AP resulted in a significant depolarization from RMP in layer 2/3 (from −73 ± 2 mV to −66 ± 1 mV, n = 8, P < 0.004), etv1 (from −69 ± 2 mV to −65 ± 2 mV, n = 11, P < 0.01), and glt (from −67 ± 1 mV to −64 ± 2 mV, n = 9, P < 0.02) pyramidal cells. 4-AP had no effects on input resistance in layer 2/3 (8 cells: 137 ± 29 MΩ control vs. 135 ± 42 MΩ 4-AP) or glt (8 cells: 110 ± 21 MΩ control vs. 98 ± 20 MΩ 4-AP) cells but significantly increased input resistance in etv1 cells (from 77 ± 15 MΩ control to 98 ± 14 MΩ 4-AP, n = 7, P < 0.03).
Consistent with a role for Kv4 channels in spike repolarization, AP HW was broadened by 4-AP in all three pyramidal cell types (Fig. 3; layer 2/3: P < 0.0001, etv1: P < 0.0001, glt: P < 0.003). A small but significant hyperpolarizing shift in AP voltage threshold was observed for glt cells (P < 0.03) but not layer 2/3 or etv1 cells (Fig. 3). The average increase in AP HW varied from 36 ± 5% in etv1 cells (n = 11) to 42 ± 7% in glt (n = 9) and 48 ± 6% in layer 2/3 pyramidal cells (n = 9). Width at threshold was significantly increased (P < 0.0001 for layer 2/3; P < 0.003 for etv1; P < 0.02 for glt) and dV/dt for repolarization was significantly decreased (P < 0.0001 for layer 2/3 and glt; P < 0.001 for etv1) in all three subtypes. These effects of 4-AP (broader APs and reduced dV/dt for repolarization) are consistent with a role for Kv4 channels in AP repolarization in all cell types, although the effects on input resistance and RMP suggest additional actions of 4-AP on channels other than Kv4. There was also a significant decrease of dV/dt for the AP upstroke with 4-AP in all cell types (P < 0.001 for layer 2/3; P < 0.007 for etv1 and glt), suggesting a possible influence of a 4-AP-sensitive current on the rate of AP depolarization, nonspecific effects of 4-AP on Na+ currents, or Na+ current rundown under our recording conditions.
We next tested the effects of BaCl2, another agent reported to block A-type current relatively selectively in hippocampal pyramidal cell dendrites (Harnett et al. 2013; Losonczy and Magee 2006) and heart (Li et al. 1998, 2000) at submillimolar doses. We found that 150 μM BaCl2 caused a large depolarization in layer 2/3 (from −70 ± 1 mV to −61 ± 2 mV, n = 11, P < 0.0001), etv1 (from −73 ± 2 mV to −66 ± 1 mV, n = 8, P < 0.002), and glt (from −65 ± 1 mV to −57 ± 3 mV, n = 7, P < 0.02) pyramidal cells. Input resistance was also significantly increased by BaCl2 in etv1 (from 137 ± 24 MΩ to 213 ± 38 MΩ, n = 7, P < 0.04) and glt (from 57 ± 8 MΩ to 115 ± 20 MΩ, n = 7, P < 0.02) cells but not layer 2/3 pyramidal cells (from 148 ± 14 MΩ to 187 ± 22 MΩ, n = 11, P < 0.08). We also found that BaCl2 dramatically broadened the AP. This was true for width at threshold (layer 2/3 P < 0.0001; etv1 P < 0.001; glt P < 0.0001) and AP HW (layer 2/3 P < 0.0001; etv1 P < 0.02; glt P < 0.007) (Fig. 4). BaCl2 also reduced the repolarization dV/dt (Fig. 4) in all three pyramidal cell subtypes (layer 2/3 and etv1 P < 0.0001; glt P < 0.0003). Spike broadening was especially dramatic with BaCl2 [64 ± 8% in layer 2/3 (n = 11), 123 ± 33% in etv1 (n = 8), 338 ± 89% in glt (n = 7)].
Finally, we compared APs of the three pyramidal cell subtypes before and after application of the peptide blocker AmmTx3 (200 nM). Rudy and colleagues showed that this agent is highly selective for channels containing Kv4 α-subunits in association with DPP6 or DPP10 auxiliary subunits (Maffie et al. 2013). We found that unlike 4-AP and BaCl2 AmmTx3 had no effects on RMP [etv1 (10 cells): control −73 ± 0.1 mV vs. −73 ± 0.4 mV AmmTx3; glt (10 cells): control −73 ± 0.4 mV vs. −74 ± 0.3 mV AmmTx3; layer 2/3 (5 cells): control −71 ± 0.9 mV vs. −72 ± 0.5 mV AmmTx3] or input resistance [etv1 (10 cells): control 83 ± 9 MΩ vs. 81 ± 9 MΩ AmmTx3; glt (10 cells): control 91 ± 11 MΩ vs. 103 ± 17 MΩ AmmTx3; layer 2/3 (5 cells): control 273 ± 51 MΩ and 201 ± 42 MΩ AmmTx3] in any pyramidal cell type. AmmTx3 application resulted in significant AP broadening (HW and width at threshold) and significant reduction of dV/dt for repolarization in all pyramidal cell subtypes tested, with no changes in voltage threshold (Fig. 5, Table 4). The percent change in AP HW was 22 ± 5% in layer 2/3 cells (n = 11), 10 ± 3% in etv1 cells (n = 10), and 18 ± 4% in glt cells (n = 10). There was a trend for decrease of dV/dt for the AP upstroke with time in all cell types that was significant in glt cells (P < 0.004), suggesting a role for Kv4 channels in regulating the AP upstroke, possible nonspecific effects on Na+ channels, or rundown of Na+ channels under our recording conditions.
The magnitude of the effects of 4-AP and BaCl2 (as well as effects on RMP and input resistance not seen with AmmTx3) led us to directly test whether BaCl2 or 4 mM 4-AP selectively blocked the A-type component of the current in neocortical pyramidal cells or if these agents also have effects on the sustained current (Kv1, Kv2, Kv7 mediated). We compared the effects of the blockers on the early peak current (dominated by A-type current in most cells) vs. current at the end of a 500-ms step to +10 mV (operationally defined as “steady-state” current). We tested outside-out patches from both etv1 and glt cells. The results were similar for both cell types, so the data were pooled for the purpose of testing blocker specificity. For 4 mM 4-AP, the block of the peak current was 56 ± 16% (n = 6 cells) and the steady-state current (at 500 ms) was blocked by 8 ± 3% (n = 6) (Fig. 6, A and D). BaCl2 (150 μM) blocked 56 ± 6% (n = 12) of the peak current and 48 ± 9% (n = 12) of the steady state current (Fig. 6, B and D). AmmTx3 blocked 42 ± 9% (n = 12) of the peak and 7 ± 2% (n = 12) of the steady-state current (Fig. 6, C and D).
These findings indicate that in neocortical pyramidal cells 150 μM BaCl2 is not at all selective for A-type or Kv4-mediated currents but also substantially blocks other components of the K+ current. Besides voltage-gated currents, BaCl2 blocked leak currents underlying RMP and input resistance (see above). 4-AP is considered a relatively nonselective drug in that in addition to blocking Kv4 channels at millimolar doses it is known to block Kv1 and Kv3 channels at micromolar doses (Grissmer et al. 1994; Kirsch and Drewe 1993). In our experiments, 4 mM 4-AP would be expected to block any Kv1.4-mediated component of A-type current (Carrasquillo et al. 2012). In addition, 4-AP is known to affect other channels at millimolar doses (including Kv2; Kirsch and Drewe 1993). We found that 4-AP (4 mM) affected a conductance active at RMP in our cells but surprisingly was relatively selective for the peak current vs. steady state, suggesting a greater degree of selectivity for A current than BaCl2. One caveat is the known effects of 4-AP on slowing the inactivation kinetics of Kv4 currents (Jackson and Bean 2007; Thompson 1982). Thus slowing of remaining Kv4 current may make less apparent the block of Kv currents that normally inactivate slowly.
We found that the recently characterized peptide blocker AmmTx3 (200 nM; Maffie et al. 2013) was selective for the peak transient A-type current over the steady-state current (and did not affect input resistance or RMP) in neocortical pyramidal cells. These voltage-clamp data with AmmTx3 also indicate that dipeptidyl peptidase-like-proteins (DPP6 or DPP10) are associated with at least some of the Kv4 subunits in neocortical pyramidal cells of mice.
It has been confirmed numerous times since the pioneering studies of Hodgkin and Huxley (1945) that the AP upstroke in various types of neurons is due to influx of Na+ through voltage-gated Na+ channels and that spike repolarization is due to Na+ channel inactivation and the activation of K+ channels (Bean 2007; Hille 2001). Which specific types of K+ channels underlie AP repolarization in different types of neurons is less well understood. In this study, we tested potential roles for Kv1, Kv2, and Kv4 channels in the repolarization of single APs recorded from the soma in three classes of pyramidal cells in mouse neocortex: two types of layer 5 cells genetically defined by expression of either etv1 or glt and layer 2/3 pyramidal cells. We blocked Kv1 channels with DTX and Kv2 channels with GxTx at doses that we had previously determined to be selective with voltage-clamp experiments on rodent pyramidal cells. Additionally, we tested the selectivity of three putative Kv4 blockers (4-AP, BaCl2, and AmmTx3) for A-type currents in voltage-clamp experiments using outside-out macropatches from pyramidal cell somas. We determined that only AmmTx3 was selective for the transient, A-type current in these cells.
On the basis of studies by the Nerbonne lab (Carrasquillo et al. 2012; Norris and Nerbonne 2010; Yuan et al. 2005), we hypothesized a major role for Kv4 channels in AP repolarization in neocortical pyramidal cells. We also hypothesized that the mechanisms underlying AP repolarization would be pyramidal cell subtype specific. In all three pyramidal cell subtypes tested, we found that the width of the AP and rate of AP repolarization were controlled by Kv4 channels. Blockade of Kv2 channels with GxTx did not affect any parameters of single APs in any cell type. The roles of Kv1 channels in shaping single APs were cell type dependent. In etv1 cells and layer 2/3 cells (but not glt cells), block of Kv1 channels resulted in broader APs and reduced the rate of repolarization. AP voltage threshold was hyperpolarized by the Kv1 channel blocker DTX in all three subtypes of pyramidal neurons.
DTX is a selective blocker of channels containing the Kv1.1, Kv1.2, or Kv1.6 subunits (Harvey and Robertson 2004). Consistent with a subthreshold activation range and relatively rapid activation kinetics (Guan et al. 2006) for Kv1 channels, we found that block of Kv1 channels with DTX resulted in a more hyperpolarized voltage threshold for single APs generated by brief current injections in both layer 5 pyramidal cell types and layer 2/3 cells. Previously, DTX-sensitive Kv1 channels were implicated in voltage threshold determination in rat layer 5 (Bekkers and Delaney 2001) and layer 2/3 (Guan et al. 2007b; Higgs and Spain 2011) neurons, as well as CA1 pyramidal neurons (Giglio and Storm 2014). It is interesting that Kv1 channels were more effective in regulating Vth in our experiments than Kv4 channels, which activate even more rapidly and at more negative potentials (Carrasquillo et al. 2012; Guan et al. 2011b; Norris and Nerbonne 2010). Perhaps the strong expression of Kv1 channels at the axonal initial segment (Kole et al. 2007; Shu et al. 2007) grants these channels a greater role in regulating the voltage threshold in the soma. We found that in glt neurons block of Kv1 channels with DTX did not result in changes in AP width or rate of repolarization, consistent with results for somatic APs in rat layer 2/3 (Guan et al. 2007b) and rat and ferret layer 5 (Kole et al. 2007; Shu et al. 2007). In contrast, block of Kv1 channels broadened AP width and reduced repolarization rate in mouse etv1 cells and layer 2/3 cells. In rat pyramidal neurons from layer 2/3 and layer 5, Kv1 channels were found to have little impact on AP repolarization in the soma (Guan et al. 2007b; Kole et al. 2007; Shu et al. 2007) but play a major role in the initial segment and axon (Kole et al. 2007; Shu et al. 2007). These findings suggest species differences in the role of somatic Kv1 channels or the influence of axonal Kv1 channels on somatic APs.
As a caveat, it should be noted that Kv1.4 channels have been shown to be a contributor to A-type current in pyramidal cells from rat visual cortex (Carrasquillo et al. 2012) and DTX would not block Kv1.4 subunits (unless they were expressed in heteromeric channels with Kv1.1, 1.2, or 1.6 α-subunits). Kv1.4-containing channels would be 4-AP sensitive. Interestingly, Nerbonne and colleagues showed that genetic knockdown of Kv1.4 channels paradoxically increased the rate of AP repolarization (Carrasquillo et al. 2012) due to upregulation of Kv4 channels.
We found that block of Kv2 channels with 100 nM GxTx did not affect AP parameters in any of the three mouse pyramidal cell subtypes, consistent with our previous findings in rat layer 2/3 and layer 5 cells with genetic manipulation of Kv2 channel expression and function (Guan et al. 2013). In concurrently conducted experiments, we showed that this dose of GxTx blocks substantial Kv2-mediated currents in both etv1 and glt neurons (Bishop et al. 2015). These data are also consistent with the relatively slow activation kinetics and depolarized activation voltage range for Kv2.1-containing channels (Guan et al. 2007a; Murakoshi and Trimmer 1999). The lack of a GxTx effect on the AP also suggests minimal block of Kv4 channels by 100 nM GxTx in our hands. Liu and Bean (2014) observed modest spike broadening in CA1 pyramidal cells with GxTx. Honigsperger et al. (2013) also observed that GxTx blocked IK but slowed AP repolarization only after the first AP during repetitive firing in ERC stellate cells (layer 2). Both CA1 pyramidal cells and ERC pyramidal cells typically have broader APs than neocortical pyramidal cells, which may allow for the relatively kinetically slow Kv2 channels to contribute to repolarization.
The rapid activation kinetics and relatively hyperpolarized activation range of Kv4 channels suggest that they could contribute to repolarization of the AP in neocortical pyramidal cells (Carrasquillo et al. 2012; Guan et al. 2011b; Norris and Nerbonne 2010). Our data with the selective Kv4 blocker AmmTx3 (200 nM), which slowed AP repolarization (and depolarization) and increased AP width, suggest that K+ channels from the Kv4 family contribute to AP repolarization in layer 2/3 and both etv1 and glt layer 5 neocortical pyramidal cells. AmmTx3 had no effect on RMP or input resistance, and our voltage-clamp experiments showed that AmmTx3 was selective for the initial transient, A-type current vs. the persistent current.
Our voltage-clamp data also show that 150 μM BaCl2 is not selective for the transient, A-type current. In addition, 4-AP (mM) and BaCl2 also had effects on RMP and input resistance that likely reflect block of other K+ channels besides Kv4. At the doses we used, 4-AP (and especially BaCl2) had larger effects on APs than AmmTx3. This may be partly due to quantitatively larger block of the A current by 4-AP and BaCl2 at these doses but is also likely due to additional effects of 4-AP and BaCl2 on other Kv channels (e.g., Kv1, Kv2) and non-Kv K+ channels (e.g., BaCl2 is known to block inwardly rectifying GIRK and IRK channels; Coetzee et al. 1999). Interestingly, blocking other potential Kv targets of 4-AP and BaCl2 individually (e.g., Kv1 and Kv2) with more specific blockers did not cause significant AP broadening or slowing of repolarization. We interpret these data as follows: Kv4 channels provide a primary AP repolarizing conductance for all neocortical pyramidal cells. When Kv4 channels are blocked, this results in a broader AP. With the doses we used, 4-AP and BaCl2 blocked more A-type current than AmmTx3 and thus had greater effects of AP repolarization and width. Additionally, part of the effect of 4-AP and especially BaCl2 may reflect that when APs are broadened by Kv4 block, Kv1 and Kv2 channels will then be able to contribute to repolarization of the broader APs. Consistent with this interpretation, TEA and BaCl2 (400 μM) had much greater effects on AP repolarization in Kv4.2-knockout animals (Carrasquillo et al. 2012).
Our findings are consistent with studies from Nerbonne and colleagues (Carrasquillo et al. 2012; Norris and Nerbonne 2010) in which genetic manipulation of Kv4 expression/function significantly broadened APs from unidentified rat and mouse layer 5 pyramidal cells. They found that more prolonged reduction in Kv4 expression may also lead to remodeling of K+ currents and compensatory changes in AP repolarization (Nerbonne et al. 2008). Our data are also consistent with single-channel studies of Kang et al. (2000), which indicate that A-type channels are active during AP repolarization in neocortical pyramidal cells. Kv4 channels also play a prominent role in repolarization of APs in CA1 pyramidal cells (Andrásfalvy et al. 2008; Chen et al. 2006; Kim et al. 2005) and cerebellar granule cells (Shibata et al. 2000). A 4-AP-sensitive A-type current has also been shown to be the primary repolarizer of the AP in neonatal auditory spiral ganglion neurons (Jagger and Housley 2002), nodose ganglion type 1C neurons (Ducreux and Puizillout 1995), and rat vagal motoneurons (Sah and McLachlan 1992).
AP amplitude and width are important variables regulating potential maximum firing rates, the effectiveness of AP backpropagation into the dendrites, and the amount of Ca2+ entry into cells (Bean 2007). While this is obviously true in nerve terminals for regulating neurotransmitter release (Jackson et al. 1991), [Ca2+]i is also regulated by somatic and dendritic APs in neocortical pyramidal neurons (Abel et al. 2004; Helmchen et al. 1996). Importantly, since most Ca2+ entry occurs during the repolarization of the AP and afterwards in these cells, the rate of repolarization strongly regulates Ca2+ entry (Stewart and Foehring 2001). Somatic [Ca2+]i is important for the activation of afterhyperpolarization currents, which regulate the temporal structure of pyramidal cell firing and somatic integration of inputs (Abel et al. 2004) as well as gene regulation and activation of enzymatic activity (Ghosh et al. 1994). Excessive Ca2+ entry and [Ca2+]i may lead to cell death after seizures or stroke or during neurodegenerative diseases (Roselli and Caroni 2015).
Recent work from our lab (Guan et al. 2015) and others (Avesar and Gulledge 2012; Brown and Hestrin 2009; Dembrow et al. 2010; Groh et al. 2010; Hattox and Nelson 2007; Larkman and Mason 1990; Le Bé et al. 2007; Mason and Larkman 1990; Schwindt et al. 1997; Sheets et al. 2011; Suter et al. 2013) has revealed differences in AP properties between pyramidal cells in different layers or with different genetic markers, anatomy, or projections. Most relevant to the present work, layer 5 pyramidal cells expressing the glt gene have much narrower APs and more rapid repolarization compared with etv1 cells or layer 2/3 pyramidal cells (Groh et al. 2010; Guan et al. 2015). Groh et al. (2010) showed that etv1 cells are a subset of IT-type neurons (Reiner et al. 2003; Suter et al. 2013) and glt cells a subset of PT-type neurons (Reiner et al. 2003; Suter et al. 2013). glt and PT-type pyramidal cells have narrower APs and exhibit fast, nonadapting firing patterns, and etv1 and IT-type neurons have broader APs, fire slower, and exhibit strong spike frequency adaptation (Avesar and Gulledge 2012; Brown and Hestrin 2009; Dembrow et al. 2010; Groh et al. 2010; Guan et al. 2015; Hattox and Nelson 2007; Larkman and Mason 1990; Le Bé et al. 2007; Mason and Larkman 1990; Sheets et al. 2011; Suter et al. 2013).
We have shown that AmmTx3-sensitive (Kv4 α-subunits associated with DPP6 or DPP10) channels play a critical role in repolarizing the APs of all of the pyramidal cell groups tested and that there are subtype-dependent differences in the effectiveness of Kv1 channels for AP repolarization. PT-type glt cells have narrower APs than IT-type etv1 cells or layer 2/3 pyramidal cells (Groh et al. 2010; Guan et al. 2015). The relatively exclusive role for Kv4 channels in repolarization of glt cells likely reflects the rapid kinetics and relatively hyperpolarized activation range of Kv4 channels vs. Kv1 and Kv2 channels. The narrower APs in glt neurons might reflect greater expression or different biophysical properties of Kv4 channels in those cells vs. IT-type etv1 or layer 2/3 cells. The role of Kv1 channels in etv1 cells and layer 2/3 cells may be permitted by their relatively broader APs compared with glt cells. One might expect greater expression of Kv1 current in etv1 and layer 2/3 cells. Another prediction is that glt cells may show more rapid spike broadening during repetitive activity than etv1 or layer 2/3 cells because of the reliance of glt cells on the rapidly inactivating Kv4 channels. Differential mechanisms for AP repolarization also provide a substrate for differential modulation of APs by transmitters. We are currently testing these predictions.
We found that Kv channels of the Kv4 type repolarize the AP in all tested neocortical pyramidal cell subtypes but there are also subtype-specific differences in AP repolarization mechanisms. While only Kv4 channels are involved in AP repolarization in the rapidly repolarizing glt cells, the broader APs in layer 2/3 and etv1 cells are also partially repolarized by Kv1 channels. Because of their activation in the subthreshold range, Kv1 channels play an important role in regulating AP voltage threshold in all pyramidal cell subtypes.
We gratefully acknowledge our funding from National Institute of Neurological Disorders and Stroke Grant NS-044163 (to R. C. Foehring).
No conflicts of interest, financial or otherwise, are declared by the author(s).
Author contributions: D.P. and D.G. performed experiments; D.P., D.G., and R.C.F. analyzed data; D.P., D.G., and R.C.F. interpreted results of experiments; D.P., D.G., and R.C.F. prepared figures; D.P. and R.C.F. drafted manuscript; D.P., D.G., and R.C.F. edited and revised manuscript; R.C.F. conception and design of research; R.C.F. approved final version of manuscript.
The authors express thanks to William E. Armstrong for critical reading of an earlier version of this manuscript.
- Copyright © 2016 the American Physiological Society