Dax Hoffman, PhD, Head, Molecular Neurophysiology and Biophysics Unit
Rebecca Hammond, PhD, Postdoctoral Fellow
Sung-Cherl Jung, PhD, _Postdoctoral Fellow-
Jinny Kim, PhD, Postdoctoral Fellow
Dongsheng Wei, PhD, Postdoctoral Fellow
Ann Clemens, BS, Postbaccalaureate Fellow

The central nervous system underlies all our experiences, actions, emotions, knowledge, and memories. With billions of neurons firing at frequencies of hundreds of hertz, the brain exhibits a complexity that is stunning. We study brain function by focusing on the workings of a single type of central neuron, the pyramidal neuron from the CA1 region of the hippocampus, a region of the brain important for learning and memory and among the first regions affected in Alzheimer's disease and epilepsy. In the dendrites of hippocampal CA1 pyramidal neurons, a nonuniform density of subthreshold, rapidly inactivating potassium channels regulates signal propagation. This nonuniform distribution (with higher expression in the dendrites than in the soma) means that the electrical properties of the dendrites differ markedly from those of the soma. The activity of these channels shapes incoming synaptic signals. Once initiated in the axon, action potentials progressively decrease in amplitude as they propagate back into the dendrites. By combining patch-clamp recording with molecular biology, we investigate the electrophysiological properties and molecular nature of the voltage-gated channels expressed in CA1 dendrites, regulation of their expression, and their role in learning and memory.

Kv4.2 control of firing patterns in hippocampal CA1 pyramidal neurons

Kim, Wei, Hoffman

Although recent molecular cloning studies have found that several families of voltage-gated K+ channel genes are expressed in the mammalian brain, information regarding the relationship between the protein products of these genes and their various neuronal functions is currently lacking. We have used a combination of molecular, electrophysiological, and imaging techniques to show that the voltage-gated potassium channel subunit Kv4.2 controls action potential (AP) half-width, frequency-dependent AP broadening, and dendritic AP propagation. To visualize Kv4.2, we fused the enhanced green fluorescence protein (EGFP) to the cytoplasmic C terminus (Kv4.2g). Using a modified Sindbis virus, which reduces cytotoxicity and improves neuro-specificity of infection, we expressed either Kv4.2g or an EGFP-tagged dominant negative mutant of Kv4.2 (Kv4.2g(W362F)) in dissociated hippocampal neurons and in CA1 pyramidal neurons of organotypic slice cultures. Previously, we showed that Kv4.2g(W362F) acts as a subfamily-specific dominant negative (i.e., blocking only Kv4.x or shal-containing channels after heterologous expression).

Nucleated and outside-out patch-clamp recordings showed a two-fold higher transient outward current density in CA1 hippocampal neurons overexpressing Kv4.2 than in non-expressing control neurons, although the sustained component density was not affected by the overexpression. Expressing the dominant negative Kv4 mutation knocked down the transient outward current density to 63 percent of control, again without affecting the sustained outward current density. Kv4.2g and Kv4.2g(W362F) expression did not alter transient current steady-state activation and inactivation. A comparison between Kv4.2g properties when expressed in HEK-293 cells versus in CA1 neurons suggests that an auxiliary subunit modulates Kv4.2g channel properties in neurons. In CA1 neurons, all measured kinetics (time to peak, inactivation rate, and rate of recovery from inactivation) are greater than those recorded in HEK-293 cells. These differences can be attributed to the auxiliary protein DPPX. Our group has begun to characterize the functional role of DPPX in CA1 neurons (see below).

To assess the contribution of Kv4 family subunits to the electrical properties and firing patterns of CA1 pyramidal neurons, we carried out whole-cell current-clamp experiments in each of three experimental groups (EGFP, Kv4.2g, and Kv4.2g(W362F)). Consistent with the window current found when plotting the transient current's steady-state activation and inactivation, subthreshold current injections indicated that Kv4.2 is open at rest in CA1 neurons. Initially, the window current acts against the current injection, determining the subsequent activation of other currents such as Ih.

Contrary to the established view that somatic A-type currents act solely subthreshold, suprathreshold current injections in Kv4.2g- and Kv4.2g(W362F)-expressing neurons showed that Kv4.2 is engaged throughout the physiologically relevant voltage range. In response to depolarization, Kv4.2 prevented AP initiation by delaying AP onset and increasing the threshold for initiation. We found that, subsequent to initiation, Kv4.2 aids in repolarization and contributes to the fast after-hyperpolarization potential. Single AP half-width was greatly reduced in Kv4.2g-expressing neurons, and somatic Ca2+ influx was barely half that of control neurons. Accumulating Kv4.2 inactivation during an AP train contributed to frequency-dependent AP broadening and subsequent somatodendritic Ca2+ influx. Finally, Kv4.2g limited backpropagation as measured with Ca2+ imaging, whereas Kv4.2g(W362F) enhanced propagation. These results demonstrate a prominent role for Kv4.2 in regulating AP shape and dendritic signaling. Given that Ca2+ influx occurs primarily during AP repolarization, Kv4.2 activity can regulate cellular processes involving Ca2+-dependent second messenger cascades such as gene expression and synaptic plasticity.

Kv4.2 trafficking in CA1 pyramidal neuron dendrites

Kim, Jung, Clemens, Hoffman; in collaboration with Petralia

In cultured hippocampal neurons, Kv4.2g mimicked the somatodendritic expression pattern found for endogenous Kv4.2. Kv4.2-positive spines were apposed to the presynaptic marker synaptophysin. EGFP fluorescence in spines of Kv4.2g-expressing neurons appeared brighter than that from the adjacent dendritic shaft. To quantify this observation, we compared the relative fluorescence intensity in spines with that of the adjacent dendritic shaft. In addition, electronmicroscopy data obtained in collaboration with Ron Petralia show that Kv4.2 is expressed in CA1 spines.

We found that AMPA stimulation results in activity-dependent redistribution of Kv4.2g away from synaptic sites to the dendritic shaft; we also observed punctate accumulation of Kv4.2g within the soma. This AMPA-induced redistribution of Kv4.2g occurred within 15 minutes of AMPA application and was reversible, indicating that the treatment was not excitotoxic (six-hour washout). We also observed reversible redistribution of Kv4.2g with KCl depolarization and glutamate treatment. Co-expression with pre- and post-synaptic markers (synaptophysin and NR1) showed that Kv4.2 undergoes activity-induced redistribution without a gross change in synaptic architecture or number. Using a biotinylation assay, we found that the surface expression level of Kv4.2g was markedly reduced after stimulation, without a significant change in total protein level. We confirmed these findings with live imaging of Kv4.2g egress from the spine in response to AMPA stimulation. The large number of synapses stimulated under these conditions enabled us to measure internalization directly as a decline in endogenous whole-cell transient K+ current from uninfected hippocampal neurons. The decrease occurred without change in sustained or non-inactivating delayed rectifier-type voltage-gated K+ current amplitudes. Thus, activity-dependent Kv4.2 internalization occurs natively and is not an artifact of overexpression.

FIGURE 8.1

Kv4.2g internalization is mediated through clathrin-dependent endocytosis. Imaging and electrophysiological experiments show that Kv4.2g and/or endogenous A-current internalization is blocked by high sucrose and by a synthetic dynamin-derived peptide that inhibits clathrin-mediated endocytosis by blocking the recruitment of dynamin to clathrin-coated pits. Scrambled peptide had no effect on internalization. Interestingly, we found that activity-dependent Kv4.2 internalization shares common requirements with LTP induction. Pre-incubation with either the NMDAR antagonist APV or the membrane-permeant intracellular Ca2+ buffer BAPTA-AM nearly abolishes AMPA-induced Kv4.2 internalization.

Spine enrichment of Kv4.2 suggested a role for Kv4.2 in shaping single synaptic events at the site of input. Accordingly, we have shown that the average miniature excitatory post-synaptic current (mEPSC) size is dependent on the functional Kv4.2 expression level. The effect of Kv4.2 on mEPSCs suggests a mechanism by which active synapses could increase their efficacy by locally decreasing Kv4.2 surface expression. To test such a possibility, we monitored mEPSCs before and after bath-applied AMPA. The results suggest that AMPA stimulation leads to Kv4.2 internalization. Pre-application of APV blocked this effect.

Given these results, removal of Kv4.2 from the spine seemed a possible mechanism for LTP, which we investigated with a brief application of glycine to induce chemical long-term potentiation (cLTP). The application caused not only GluR1 insertion but also Kv4.2g internalization. We observed a significant, decremental decline in endogenous transient (but not sustained) K+ current amplitude during cLTP (control neurons showed an approximate 20 percent reduction over the first five minutes after glycine application and a 50 percent reduction by 30 minutes). Thus, activity-dependent trafficking of the Kv4.2 in hippocampal neurons presents a novel mechanism for post-transmission synaptic integration and plasticity by activity-dependent regulation of Kv4.2 channel surface expression levels.

Role of voltage-gated potassium channels in synaptic plasticity

Jung, Kim, Hoffman

Using the Sindbis virus system to infect organotypic slice cultures with Kv4.2g and Kv4.2g(W362F), we have begun following a depolarization pairing protocol to investigate the role of Kv4.2 in LTP. For the first 10 minutes after pairing, potentiation is similar in all three groups, achieving an approximately 100 percent increase in EPSC size. After this period, however, Kv4.2-overexpressing neurons fail to maintain potentiation, such that EPSC size returns to baseline after about 25 minutes. Conversely, expression of Kv4.2g(W362F) results in a potentiation that, compared with controls, reaches a greater level 40-50 minutes after initiation. The data indicate that the functional expression level of Kv4.2 influences the degree of potentiation (see Figure 8.1).

Creation and characterization of Kv4.2 transgenic mice

Hammond, Hoffman; in collaboration with Anderson

We are currently characterizing a transgenic mouse expressing a dominant negative pore mutation in the voltage-gated potassium channel subunit Kv4.2, which is the likely molecular identity of transient currents recorded in CA1 dendrites. The mouse expresses the mutant Kv4.2 channel along with green fluorescent protein under control of a tetracycline transactivator (tTA)-responsive promoter. Expression is spatially controlled by a new line of tTA-expressing mice that limit tTA activity to the CA1 and dentate gyrus regions of the hippocampus. Expression can be controlled temporally by administration of doxycycline. We will perform experiments in acute hippocampal slices from the mice to investigate Kv4.2's role in regulating AP backpropagation into CA1 dendrites and synaptic integration and plasticity. In collaboration with Anne Anderson, we are investigating seizure susceptibility in the mice. In addition, we are using the mice to investigate Kv4.2's role in hippocampus-dependent learning and memory in the Morris water maze, a well-established task used to examine hippocampus-dependent learning and memory in rodents.

Role of auxiliary proteins in regulating Kv4.2 expression and function

Clemens, Kim, Hoffman; in collaboration with Rudy

A-type K+ currents have unique kinetic and voltage-dependent properties that allow them to fine-tune synaptic events, action potentials, and neuronal firing. However, in heterologous cells, Kv4 currents display slower kinetics of inactivation and recovery from inactivation than that typically recorded in neurons. Moreover, Kv4.2 channels likely contribute the large share of A-type current in a number of cell types, including hippocampal CA1 pyramidal neurons, layer 5 cortical neurons, and granule cells of the olfactory bulb, yet A-current properties and distributions vary between these cell types. Reconciliation of these results came with the finding of neuron-specific auxiliary subunit expression of KChIPs (Kv channel-interacting proteins) and DPLs (dipeptidyl-peptidase-like proteins). In hippocampal CA1 pyramidal neurons, DPPX (also called DPP6) is the prominent DPL family member while two KChIPs (KChIP2 and KChIP4) are expressed. When Kv4 subunits are co-expressed with DPPX, many properties of native CA1 A-type currents are restored, including a leftward shift of steady-state inactivation and activation curves, acceleration of recovery from inactivation, time-to-peak, and decay times.

To investigate the physiological role of DPPX in CA1 neurons, we developed, in collaboration with Bernardo Rudy's laboratory, short-interfering RNAs (siRNAs) to suppress the expression of all DPPX variants. We tested the efficacy of the siRNAs in CHO cells co-expressing Kv4.2 and DPPX-S. The reduction of DPPX proteins in CHO cells transfected with DPPX siRNA (siDPPX) was more than 95 percent complete as quantified by immunoblotting. To investigate whether DPPX alters the kinetics of A-type currents in a native system, we conducted voltage-clamp experiments in outside-out patches from CA1 pyramidal neurons in hippocampal organotypic slices infected with siDPPX using the Sindbis virus system. EGFP was co-expressed under control of a second promoter, allowing us to target expressing neurons for visualized patching. After allowing two to three days post-infection for DPPX knockdown, we found, in accordance with heterologous studies, that siDPPX results in a delayed recovery from inactivation, slowed time-to-peak, and a rightward shift in the steady-state inactivation and activation curves for A-type currents. We found no significant changes in A-type current density, perhaps owing to the presence of endogenous KChIP proteins, which alone are capable of augmenting current density to levels seen with the DPPX-KChIP-Kv4 triple complex. To determine the physiological effect of A-type current kinetic modifications by siDPPX, we carried out current-clamp experiments in siDPPX-expressing cells. Compared with negative control siRNA neurons, siDPPX-infected neurons exhibited less input resistance, delayed time to AP onset, a higher AP threshold, lower firing frequency, a higher AP half-width, and lower fast AHP amplitudes. Thus, siDPPX had contrasting effects, decreasing excitability subthreshold and increasing excitability suprathreshold. We are using computer modeling to determine whether the shift in inactivation and activation curves can explain, respectively, these sub- and suprathreshold effects.

COLLABORATORS

Anne Anderson, MD, Baylor College of Medicine, Houston, TX
Ron Petralia, PhD, Laboratory of Neurochemistry, NIDCD, Bethesda, MD
Bernardo Rudy, MD, PhD, New York University Medical Center, New York, NY

For further information, contact hoffmand@mail.nih.gov.