Abstract
Many neurons in the nervous systems express afterhyperpolarizations that are mediated by a slow calcium-activated potassium current. This current shapes neuronal firing and is inhibited by neuromodulators, suggesting an important role in the regulation of neuronal function. Surprisingly, very little is currently known about the molecular basis for this current or how it is gated by calcium. Recently, the neuronal calcium sensor protein hippocalcin was identified as a calcium sensor for the slow afterhyperpolarizing current in the hippocampus. However, while hippocalcin is very strongly expressed in the hippocampus, this protein shows a relatively restricted distribution in the brain. Furthermore, the genetic deletion of this protein only partly reduces the slow hyperpolarizing current in hippocampus. These considerations question whether hippocalcin can be the sole calcium sensor for the slow afterhyperpolarizing current. Here we use loss of function and overexpression strategies to show that hippocalcin functions as a calcium sensor for the slow afterhyperpolarizing current in the cerebral cortex, an area where hippocalcin is expressed at much lower levels than in hippocampus. In addition we show that neurocalcin δ, but not VILIP-2, can also act as a calcium sensor for the slow afterhyperpolarizing current. Finally we show that hippocalcin and neurocalcin δ both increase the calcium sensitivity of the afterhyperpolarizing current but do not alter its sensitivity to inhibition by carbachol acting through the Gαq-11-PLCβ signaling cascade. These results point to a general role for a subgroup of visinin-like neuronal calcium sensor proteins in the activation of the slow calcium-activated afterhyperpolarizing current.
Introduction
Repetitive spiking by pyramidal cells of the hippocampus and cortex results in the development of a slow afterhyperpolarization (sAHP) that controls spike frequency accommodation (Madison and Nicoll, 1984; Schwindt et al., 1988b) and regulates neuronal gain (Higgs et al., 2006). Similar slow hyperpolarizing afterpotentials have also been reported in many other cell types, suggesting that sAHPs are ubiquitous regulators of neuronal excitability (Faber and Sah, 2003; Vogalis et al., 2003). A general property of these sAHPs is that they are regulated by neuromodulators signaling via Gαq-11 or through the Gαs/adenylate cyclase/cAMP/protein kinase A signaling cascade (Madison and Nicoll, 1982; Schwindt et al., 1988a; Pedarzani and Storm, 1993; Krause et al., 2002). Thus the sAHP is an important determinant of cellular excitability in a variety of cell types in the CNS and an important target for many neuromodulators.
There is now widespread consensus that the sAHP is mediated by a calcium-activated potassium current generally referred to as I sAHP (Faber and Sah, 2003; Vogalis et al., 2003). Surprisingly, however, neither the ion channel(s) carrying I sAHP nor the mechanism through which this current is gated by calcium are well understood. A recent breakthrough occurred with the discovery that the genetic deletion of the neuronal calcium sensor (NCS) hippocalcin results in the inhibition of I sAHP in hippocampal pyramidal cells. This observation has led to the idea that hippocalcin is the calcium sensor for I sAHP (Tzingounis et al., 2007). However, while I sAHP is a broadly expressed cellular property in many regions of the brain, the expression of hippocalcin is much more restricted (Saitoh et al., 1994; Paterlini et al., 2000). Furthermore, hippocalcin deletion inhibits but does not abolish I sAHP in hippocampal pyramidal cells. These considerations suggest that hippocalcin may not be the only NCS protein capable of functioning as a calcium sensor for I sAHP. The cerebral cortex offers the opportunity to address this issue since hippocalcin is expressed in this area at levels considerably lower than in hippocampus (Saitoh et al., 1994), but cortical pyramidal cells display a robust and well characterized I sAHP. In the present work, we have taken advantage of these properties to readdress the role of hippocalcin in the regulation of I sAHP.
Materials and Methods
Experiments were conducted on cortical brain slices either acutely prepared or maintained in organotypic culture (Villalobos et al., 2004; Yan et al., 2009). Briefly, rats or mice during the second (organotypic) or third to fifth (acute) postnatal weeks were anesthetized with halothane or isoflurane and killed by decapitation. The brain was removed and cooled in ice-cold Ringer's solution of standard composition (in mm: 119 NaCl, 2.5 KCl, 1.3 MgSO4, 2.5 CaCl2, 1 NaH2PO4, 26.2 NaHCO3, and 11 glucose), supplemented with 10 mm HEPES and bubbled to saturation with 95%O2, 5% CO2. The brain was blocked and coronal brain slices of a nominal thickness of 300 μm were cut using a vibratome. For organotypic culture, slices were collected in sterile Ringer's solution and placed in culture essentially as described by Stoppini et al. (1991).
Electrophysiological recordings were obtained from pyramidal cells of layer II–III (shRNA experiment) or layer V (NCS expression experiments) of the medial prefrontal/anterior cingulate cortices as previously described (Villalobos et al., 2004, 2005; Yan et al., 2009). Slices were transferred one at a time to a recording chamber on the stage of a Nikon E600 microscope, where they were held in place by a platinum and nylon grid. Slices in the recording chamber were perfused continuously with Ringer's solution at 30 ± 1°C bubbled to saturation with 95% O2 and 5% CO2. Whole-cell recordings were obtained using a potassium-based intracellular solution (composition in mm: KMeSO4 120, KCl 5, NaCl 5, EGTA 0.02, HEPES 10, MgCl2 1, Na phosphocreatine 10, ATP magnesium salt 4, GTP Na salt 0.3, myo-inositol 30). Electrodes filled with this solution exhibited resistances ranging from 2 to 4 MΩ. Because holding currents and I sAHP amplitudes were relatively small, series resistance was left uncompensated. All recording were conducted using a Multiclamp amplifier under the control of pClamp 9 (Molecular Devices).
Pyramidal neurons were transfected using either in utero electroporation (shRNA experiments) or particle mediate gene transfer (NCS expression experiments). In utero electroporation was conducted on embryonic mice at E14.5 as described by Saito (2006) and used a pool of five shRNAs targeting hippocalcin to maximize the suppression of hippocalcin expression (Parsons et al., 2009). These shRNAs were developed by the RNAi Consortium, and their expression was driven by the U6 promoter in the pLKO.1-puro vector (Mission RNAi, Sigma) (supplemental Fig. S1, available at www.jneurosci.org as supplemental material). YFP in pCAG (Addgene) at a 4:1 ratio was electroporated with the shRNA to serve as a transfection marker. Particle-mediated gene transfer was conducted as previously described (Villalobos et al., 2004; Yan et al., 2009). Gold particles were coated with plasmids encoding either hippocalcin (in pCMV-Sport6, ATCC Cat. No. 10470251/MGC Image ID 5360894), neurocalcin δ (in pCMV-Sport6, ATCC Cat. No. MGC-36496/MGC Image ID 5365560), or VILIP-2 (in pcDNA3.1, a kind gift from Dr. W. A. Catterall, University of Washington, Seattle, WA) (Lautermilch et al., 2005) and EGFP (pEGFP-N1, Clontech) at a 4:1 ratio. In all these vectors, expression was driven by the CMV promoter, which effectively drives gene expression in this preparation (Béïque and Andrade, 2003). The PLCβ1 knock-out mouse (Kim et al., 1997) was a kind gift from Drs. R. K. Wong (SUNY Downstate Medical Center, Brooklyn, NY) and H. S. Shin (Korea Institute of Science and Technology, Seoul, Korea).
For experiments examining the effect of the shRNA targeting hippocalcin, we used cells in the untransfected side of the brain as controls, rather than non-YFP-expressing cells in the transfected side, because preliminary experiments indicated that detectable YFP expression in a cell in vivo did not unambiguously identify untransfected neurons in the transfected side (supplemental Fig. S2, available at www.jneurosci.org as supplemental material). Experiments aimed at examining the effects of expressing NCS proteins used paired recordings from neighboring layer V transfected and untransfected pyramidal neurons located within ∼150 μm of each other. This approach was preferred to using mock transfections in separate slices for the controls because it allows for a more stringent cell match with respect to cell location and slice history. We have previously shown that transfection with EGFP has no detectable effect on I sAHP (Yan et al., 2009) and the very low transfection efficiency of the gene gun assured that the overwhelming fraction of non-GFP-expressing cells correspond to untransfected neurons. All organotypic slice recordings were obtained 2–3 d after transfection.
Pyramidal cells were held at −60 mV and I sAHP was triggered using a 100-ms-long depolarizing step in voltage clamp. In most experiments, we administered apamin (300 nm) to block I mAHP. This allowed us to estimate the amplitude of I mAHP for most cells and also to record I sAHP in isolation. I sAHP amplitude was determined ∼300 ms after the end of the triggering step. The voltages reported are not corrected for the liquid junction potential. Most drugs were obtained from Sigma or Fisher. Tetrodotoxin was obtained from EMD Biosciences, while apamin was obtained from Tocris Bioscience. All drugs were administered to the slice dissolved in the bath. Statistical comparisons used t tests.
Results
In a first experiment, we used in utero electroporation to express shRNAs targeting hippocalcin in layer II–III pyramidal cells of cingulate cortex to ask whether hippocalcin played a role regulating I sAHP in this area. Previous studies have shown the effectiveness of this shRNA approach to knockdown the expression of specific proteins (e.g., Bai et al., 2003; Parsons et al., 2009), and we confirmed the effectiveness of this strategy in the current experiments using immunohistochemistry to confirm the knockdown of hippocalcin expression in transfected cortical neurons of the upper layers of cortex (supplemental Fig. S3, available at www.jneurosci.org as supplemental material). As illustrated in Figure 1 A, transfection of shRNAs targeting hippocalcin resulted in a significant reduction in the amplitude of I sAHP in YFP-expressing pyramidal cells when compared to the amplitude of I sAHP in pyramidal cells matched for location on the (nontransfected) contralateral cortex.
Hippocalcin expression regulates I sAHP. A1 , Superimposed traces illustrating I sAHP recorded from shRNA/YFP-transfected (green) and untransfected (black) pyramidal cells. Inset, Fluorescence image depicting transfected and untransfected cingulate cortices in a living brain slice at the approximate position illustrated in the diagram. Calibration bar = 100 μm. A2 , Graph summarizing the amplitude of I sAHP in 5 shRNA/YFP-transfected and 9 control untransfected cells. *p < 0.05. B1 , Superimposed traces depicting I sAHP recorded in a pair of hippocalcin-transfected (green) and neighboring untransfected (black) pyramidal cells in organotypic slices. Lower panel illustrates the same slices scaled to emphasize the change in the time course of I sAHP. Inset, Effect of hippocalcin on the amplitude of I sAHP. *p < 0.001. B2 , Graphs illustrating the effects of hippocalcin expression on the amplitude (upper plot) and decay (lower plot) of I sAHP for each cell pair tested in this experiment. The half decay for only 11 cell pairs is depicted in the lower graph because it was not possible to estimate the half decay for I sAHP in one of the control cells.
To test further the possible role of hippocalcin in mediating I sAHP in cortex, we transfected this NCS into pyramidal cells in cortical slices maintained in organotypic culture and assessed its effect on I sAHP using paired recordings from neighboring transfected and untransfected pyramidal cells (Béïque et al., 2007; Yan et al., 2009). As illustrated in Figure 1 B, expression of hippocalcin resulted in a large increase in the amplitude of I sAHP (from 54 ± 8 to 110 ± 12 pA, n = 12 cell pairs, p < 0.01) that was accompanied by a significant slowing of its decay kinetics (p < 0.01). Since expression of hippocalcin had no significant effect on cell resting membrane potential, input resistance, or the amplitude of the apamin-sensitive calcium-activated potassium current I mAHP (supplemental Table S1, available at www.jneurosci.org as supplemental material), we interpret these loss of function and overexpression experiments to indicate that hippocalcin regulates the amplitude of I sAHP in cortical pyramidal cells.
We next considered the possibility that other members of the NCS protein family could act as calcium sensors for I sAHP. Hippocalcin shares considerable sequence homology with the NCS proteins neurocalcin δ and VILIP1–3, and previous studies have suggested that these proteins may form a functionally distinct subfamily of visinin-like proteins (Burgoyne, 2007; Braunewell and Klein-Szanto, 2009). Therefore, we next transfected neurocalcin δ, a close homolog of hippocalcin (91% amino acid identity), and examined its effect on I sAHP. As illustrated in Figure 2 A, expression of neurocalcin δ, like hippocalcin, again resulted in a large increase in I sAHP amplitude (from 58 ± 10 to 98 ± 10 pA, n = 8 cell pairs, p < 0.01) and slowing of I sAHP decay kinetics (p < 0.01). In contrast, transfection of VILIP-2, a more distant member of the hippocalcin subfamily of NCS proteins (<70% amino acid identity), had no effect on either the amplitude or decay kinetics of I sAHP (Fig. 2 B) (control: 63 ± 13 pA, VILIP-2 59 ± 5 pA, n = 5 cell pairs, p = 0.79). Combined, these results indicate that I sAHP can be regulated by at least two members of the visinin-like subfamily of NCS proteins.
Neurocalcin δ, but not VILIP-2, increases the amplitude and prolongs the duration of I sAHP. A1 , I sAHP recorded in a pyramidal cell transfected with neurocalcin δ (green) and in a neighboring untransfected control pyramidal cell (black) in the same slice. Inset, Graph summarizing the effect of neurocalcin δ on the amplitude of I sAHP in 8 cell pairs. *p = 0.012. A2 , A3 , Graphs illustrating the effects of neurocalcin δ expression on the amplitude ( A2 ) and decay ( A3 ) of I sAHP for each cell pair. The half decay for only 7 cell pairs is depicted in A3 because it was not possible to estimate the half decay for I sAHP in one of the control cells. A3 , Inset, Traces in A1 scaled to illustrate the slower time course of I sAHP induced by neurocalcin δ. B1 , I sAHP recorded in a cortical pyramidal cell transfected with VILIP-2 (green) and a neighboring control untransfected cell (black). Inset, Graph summarizing the amplitude of I sAHP in control cells and in cells transfected with VILIP-2. B2 , B3 , Graphs illustrating the effects of VILIP-2 on the amplitude ( B2 ) and decay ( B3 ) of I sAHP for each cell pair tested in this experiment. N = 5 cell pairs tested. p = 0.49.
Hippocalcin and neurocalcin δ are expressed in the cerebral cortex in a roughly overlapping pattern, suggesting coexpression at the cellular level (supplemental Fig. S4, available at www.jneurosci.org as supplemental material). Previous work has suggested that two or more NCS proteins exhibiting different calcium affinities may work in tandem to extend the dynamic range of calcium sensing (reviewed by Burgoyne, 2007). Therefore we considered the possibility that a hippocalcin and neurocalcin δ could sense different calcium concentrations to activate I sAHP. If that was the case, it could be expected that the hippocalcin- and neurocalcin δ-enhanced I sAHPs should differ in their calcium sensitivity. To test this idea, we examined the apparent calcium sensitivity of I sAHP under control conditions and after transfection with hippocalcin or neurocalcin δ. As illustrated in Figure 3, both hippocalcin and neurocalcin δ, but not VILIP-2, shifted the apparent calcium sensitivity of I sAHP. However, the calcium-sensitivity curves for hippocalcin-enhanced and neurocalcin δ-enhanced I sAHPs are superimposable (Fig. 3 B, inset). These results do not support the idea that hippocalcin and neurocalcin δ sense different concentrations of calcium.
Effect of hippocalcin, neurocalcin δ and VILIP-2 on the apparent calcium sensitivity of I sAHP. Inset, Voltage steps to −60 to +20 mV were used to produce graded increases in calcium influx and hence graded activations of I sAHP (Gerlach et al., 2004). Calibration: 40 pA, 500 ms. Expression of hippocalcin and neurocalcin δ, but not of VILIP-2, resulted in an increase in the apparent calcium sensitivity of I sAHP. However, the apparent calcium sensitivity curves for cells transfected with hippocalcin or neurocalcin δ were superimposable (lower right). Hippocalcin, n = 15 cell pairs; neurocalcin δ, n = 6 cell pairs, VILIP-2, n = 4 cell pairs.
Little is known at the present time about the molecular identity of the channels carrying I sAHP. This makes it difficult to unambiguously determine whether hippocalcin and neurocalcin δ facilitate the ion channels carrying the native current. As noted in the Introduction, one of the defining characteristics of I sAHP is its inhibition by neuromodulators acting through G-protein-coupled receptors signaling through Gαs or Gαq-11. Previous studies have shown that carbachol inhibits I sAHP in pyramidal cells of the cerebral cortex by acting on muscarinic M1/M3 receptors (Gulledge et al., 2009). At a biochemical level, these receptors activate Gαq-11 and PLCβ, leading to the breakdown of PtdIns(4,5)P2. Therefore we first tested whether carbachol indeed inhibited I sAHP by activating Gαq-11 and PLCβ in these cells. As illustrated in Figure 4 A, the ability of carbachol to inhibit I sAHP was blocked by expression of the carboxy tail of PLCβ1 fused to GFP (PLCβ-ct), a dominant-negative construct that blocks Gαq-11 signaling in central and peripheral neurons (Kammermeier and Ikeda, 1999; Yan et al., 2009). In addition, the ability of carbachol to inhibit I sAHP was greatly reduced in animals in which PLCβ1, the predominant PLCβ isoform expressed in the cerebral cortex, had been deleted (Fig. 4 A3) (Kim et al., 1997; Watanabe et al., 1998). These results indicated that carbachol inhibits I sAHP in cortex through Gαq-11 and, at least in part, PLCβ1. To examine the sensitivity of the NCS-enhanced I sAHP to inhibition by Gαq-11 signaling, we transfected pyramidal cells with hippocalcin or neurocalcin δ and again examined the effects of carbachol. As illustrated in Figure 4 B, administration of carbachol (30 μm) inhibited I sAHP in these cells in a manner that was not statistically different from control (p = 0.94), thus suggesting that the NCS-enhanced I sAHP retains its ability to be inhibited by neuromodulators acting via Gαq-11 and PLCβ.
Effect of carbachol on I sAHP. A1 , Superimposed traces depicting the effect of carbachol (30 μm) on I sAHP in a pyramidal cell transfected with the Gαq-11 dominant-negative PLCβ1-ct and a neighboring control cell. Control, Black; Carb, gray. A2 , Graph summarizing the effect of carbachol on I sAHP in control and PLCβ1-ct-transfected pyramidal cells. *p < 0.001. A3 , Summary plot illustrating the effect of carbachol on I sAHP in slices derived from wild-type and PLCβ1 knock-out mice. *p < 0.001. B1 , Effect of carbachol (30 μm) on a pyramidal cell transfected with hippocalcin and on a neighboring untransfected pyramidal cell. B2 , Graph plotting the carbachol (30 μm)-induced inhibition of I sAHP in 7 control cells, 7 cells transfected with hippocalcin, and 5 cells transfected with neurocalcin δ. Please note that even in these slices derived from young animals, carbachol induced a slow afterdepolarization in a few cells (Yan et al., 2009), which resulted in a fractional inhibition of I sAHP bigger than 1.
Discussion
The identification of hippocalcin as a calcium sensor for I sAHP in hippocampus, while making much-needed fresh inroads into our understanding of this current, raised a number of important mechanistic questions (Brown et al., 2007; Tzingounis et al., 2007). Most notably, the restricted expression of hippocalcin in the brain, and the only partial suppression of I sAHP in the hippocalcin knock-out mouse, questioned whether hippocalcin alone could account for the calcium sensitivity of I sAHP (Brown et al., 2007). In the current work, we used loss of function and overexpression approaches to show that hippocalcin regulates I sAHP in the cerebral cortex, and show that neurocalcin δ, a closely related visinin-like NCS protein similarly expressed in this region, can also regulate I sAHP. These results solve inconsistencies surrounding the identification of hippocalcin as the calcium sensor for I sAHP and point to a broader role for NCS proteins in the regulation of this current.
The original suspicion that hippocalcin may not be the sole calcium sensor for I sAHP was based not only on the presence of a residual I sAHP in the hippocalcin knock-out, but also on the observation that this residual I sAHP is relatively resistant to inhibition by norepinephrine acting through β-adrenergic receptors (Tzingounis et al., 2007). This led Brown et al. (2007) to suggest that I sAHP may be gated through two parallel branches exhibiting differential sensitivity to inhibition by norepinephrine. In the current work, we show that carbachol inhibits I sAHP in cortical pyramidal neurons by acting through Gαq-11 and (at least in part) PLCβ1. Furthermore, we show that the native, the hippocalcin-enhanced, and the neurocalcin δ-enhanced I sAHPs are all inhibited to a comparable extent by this signaling cascade. These results suggest that the Gαq-11/PLCβ signaling cascade does not distinguish between hippocalcin- and neurocalcin δ-gated I sAHPs. However, they do support the idea that hippocalcin and neurocalcin δ act on the native I sAHP.
The main finding of the current report is that at least two distinct NCS proteins can act as calcium sensors for I sAHP. But why should that be the case, especially since it seems likely that both these proteins are coexpressed in at least some cortical neurons? One possibility is suggested by the observation that NCS proteins with different calcium affinities can regulate a single molecular target to allow for an increase in the dynamic range of calcium sensing (reviewed by Burgoyne, 2007). In the current work, we find that while hippocalcin and neurocalcin δ both increase the apparent calcium sensitivity of I sAHP, within the limits of this assay, the calcium sensitivities of these NCS-enhanced I sAHPs are indistinguishable. As such, these results are inconsistent with the idea that hippocalcin and neurocalcin δ sense different calcium concentrations when they regulate I sAHP. Further studies are needed to clarify and understand this apparent redundancy.
The observation that the genetic deletion of hippocalcin results in a reduction in the amplitude of I sAHP has been interpreted to suggest that hippocalcin functions as calcium sensor for this current (Tzingounis et al., 2007). The results of the present experiments support this idea, while also extending such a role to neurocalcin δ. Specifically, we now show that hippocalcin and neurocalcin δ increase not only the amplitude of I sAHP but also its calcium sensitivity. Since such an increase in calcium sensitivity would be predicted for the overexpression of a calcium sensor, these results support the idea that these NCS proteins sense calcium as part of the gating of I sAHP. Such an interpretation is consistent with previous studies showing that I sAHP follows the time course of cytoplasmic calcium (Abel et al., 2004) and that hippocalcin shuttles from the cytoplasm to the membrane in response to rises in intracellular calcium, and does so with a time course compatible with a role in signaling I sAHP (Markova et al., 2008). However, exactly how these visinin-like NCS proteins function in the context of gating I sAHP is unclear. Clearly a more detailed understanding of the molecular mechanisms underlying I sAHP will be required to understand how these visinin-like NCS proteins regulate this enigmatic current.
Footnotes
-
This work was supported by National Institutes of Health Grant MH43985. We thank Elaine Weber for technical assistance.
- Correspondence should be addressed to Rodrigo Andrade, Department of Pharmacology, 3108 Scott Hall, Wayne State University School of Medicine, 540 East Canfield Street, Detroit, MI 48230. randrade{at}med.wayne.edu
