Inactivity-induced increase in nAChRs upregulates Shal K(+) channels to stabilize synaptic potentials.

Long-term synaptic changes, which are essential for learning and memory, are dependent on homeostatic mechanisms that stabilize neural activity. Homeostatic responses have also been implicated in pathological conditions, including nicotine addiction. Although multiple homeostatic pathways have been described, little is known about how compensatory responses are tuned to prevent them from overshooting their optimal range of activity. We found that prolonged inhibition of nicotinic acetylcholine receptors (nAChRs), the major excitatory receptors in the Drosophila CNS, resulted in a homeostatic increase in the Drosophila α7 (Dα7)-nAChR. This response then induced an increase in the transient A-type K(+) current carried by Shaker cognate L (Shal; also known as voltage-gated K(+) channel 4, Kv4) channels. Although increasing Dα7-nAChRs boosted miniature excitatory postsynaptic currents, the ensuing increase in Shal channels served to stabilize postsynaptic potentials. These data identify a previously unknown mechanism for fine tuning the homeostatic response.

Introduction

Maintaining neural activity within an optimal range, while at the same time, allowing for the storage of long–term changes in synaptic strength is an important challenge of the nervous system[1]. Homeostatic synaptic plasticity has widely been received as a mechanism by which neuronal circuits preserve stability when presented with changes in activity. Synaptic inactivity has been shown to result in an increase in postsynaptic GluA1/GluA2 receptor number and/or presynaptic vesicle release in mammalian neurons, as well as at the neuromuscular junction (NMJ)[1-3]. Although multiple homeostatic feedback mechanisms exist for scaling up synaptic strength, maintaining activity within an optimal range must also require precise tuning of activity to prevent over–shooting the target range. Downstream control mechanisms are likely to exist, although no examples have been reported.

Most cell intrinsic responses to activity blockade have been reported to contribute to the homeostatic response[4-6]. For example, in cultured cortical pyramidal neurons, activity blockade results in an increased voltage–dependent Na+ current and a reduced delayed rectifier type K+ current, both predicted to increase excitability[4]. In contrast, however, deprivation of visual input during the critical period of development reduced intrinsic excitability of pyramidal neurons in the visual cortex[7]. In all cases, little is known about the signaling pathways inducing these intrinsic changes, how these changes are regulated, and their roles in synaptic homeostasis[1].

Homeostasis has also been implicated to underlie the up–regulation of neuronal nicotinic acetylcholine receptors (nAChRs) following prolonged exposure to nicotine[8]. Although nicotine is an agonist, extended exposure to low levels of nicotine leads to desensitization of nAChRs, which is thought to trigger homeostatic pathways[9,10]. The increased number of nAChRs is thought to contribute to the increased sensitivity to nicotine when receptors are available for activation, and conversely, tolerance to nicotine when receptors are desensitized[8,9]. A greater understanding of the homeostatic regulation of nAChRs is likely to provide insight into the pathogenesis of nicotine addiction.

Here, we block nAChRs, which mediate the vast majority of fast excitatory synaptic transmission in central Drosophila neurons, and reveal a homeostatic increase in mEPSC carried by newly translated Dα7 nAChRs. We show that this increase in Dα7 induces an increase in expression and function of the transient A–type Shal K+ channel, and this increase is triggered by increased Ca2+ influx through Dα7 receptors and CaMKII activation. While increasing Dα7 boosts mEPSCs, the ensuing increase in Shal K+ channels evokes a novel mechanism to stabilize synaptic potentials.

Results

Homeostatic Increase in mEPSCs in Excitatory Neurons

To examine homeostatic changes at inter–neuronal synapses in Drosophila, we began by using primary cultures from late–gastrula stage embryos. These cultures have been well studied, with respect to voltage–dependent currents and synaptic physiology[11-14], allow the labeling of identified neurons for analysis[15,16], and have been shown to exhibit stable, matured, electrical properties after three days[13]. Since nAChRs mediate the vast majority of fast excitatory synaptic events in the Drosophila central nervous system[13,16], we blocked synaptic activity with curare, which completely eliminates mEPSCs (). To analyze identified neurons, these specific GAL4 lines were used to drive expression of UAS–mCD8–EGFP: (1) RRa–GAL4 or RN2–GAL4 were used to drive expression in aCC and RP2 motoneurons (MNs), (2) GH146–GAL4 was used to drive expression in projection neurons (PNs), which receive input from olfactory neurons and project to higher centers in the brain, (3) EL–GAL4 was used to drive expression in the lateral cluster of even–skipped expressing cells (EL), which are reported to be exclusively interneurons[17].

We blocked synaptic activity with curare in the culture medium, then washed out antagonist for ~3 minutes, and allowed the cultures to recover for 30 minutes in fresh medium. Medium was then changed to extracellular recording solution and mEPSCs were recorded from EGFP–labeled neurons. This treatment protocol is referred to as Protocol #1 (see Methods). When synaptic activity was blocked for up to 12 hours, no changes in mEPSCs were observed (). With 24 hours of synaptic inhibition, however, there was a clear increase in mEPSC amplitude and frequency in the excitatory MNs (Ctr: 11.3 ± 1.1 pA, T: 18.5 ± 2.0 pA) and PNs (Ctr: 8.0 ± 0.9 pA, T: 15.1 ± 1.2 pA) (). In contrast, EL interneurons showed a slight decrease in frequency, and only immediately after washout of antagonist (). Enhancement of mEPSCs in excitatory neurons, however, was long–lasting, persisting for hours after antagonist washout (). Since no significant difference in dendritic branching, quantified at radial intervals from the soma of single neurons, were observed between treated and untreated controls (), changes in mEPSCs are not likely to be due to major structural changes induced by curare treatment. Increases in mEPSC amplitude and frequency in MNs and PNs in response to activity blockade are similar to homeostatic responses reported for glutamatergic mammalian neurons. Thus, although mediated by different receptors, excitatory synaptic homeostasis appears to be a conserved phenomenon in central Drosophila neurons.

Inactivity-Induced Up–Regulation of the Dα7 nAChRs

An increase in mEPSC amplitude suggests an enhancement in postsynaptic nAChR receptor function and/or number. Interestingly, mEPSCs recorded from curare treated MNs also appeared to displayed a faster rate of decay, compared to untreated MNs (). When mEPSC decays were fit with a double exponential function, we found a similar fast component present in both mock and curare treated MNs (Control τ(fast) = 0.32 ± 0.03 ms, Treated τ(fast) = 0.27 ± 0.03 ms). The contribution of this fast component, however, was significantly increased with synaptic blockade (; A1/(A1+A2)). This result suggested that the increased mEPSC amplitudes might be due to up–regulation of a particular subtype of nAChRs. Interestingly, vertebrate α7 nAChRs have been shown to display more rapid deactivation kinetics than other subtypes[18,19]. Although there is a paucity of specific inhibitors, memantine has previously been used to block mammalian α7 receptors[20]. Indeed, we found that the increase in mEPSC amplitude, following curare treatment, was blocked when memantine was used during recording (Ctr: 8.3 ± 0.9 pA, T: 8.7 ± 0.9 pA; ), suggesting that the increase is memantine– sensitive, and likely carried by the Drosophila α7 (Dα7) receptors. To examine whether the increase in mEPSC amplitude might be due to an increase in Dα7 receptor number, we used an antibody against Dα7 to immunostain cultures that were either mock or curare treated for 24 hours. Indeed, we observed a clear and reproducible increase in Dα7 signal in most neurons, including GFP–labeled MNs, from curare treated cultures ().

We next examined if the increase in Dα7, induced by synaptic blockade, can be observed in adult brains in vivo, and further, if this homeostatic response is specific to the Dα7 subunit. To do this, whole brains were mock or curare treated in culture; cell viability was confirmed by showing that protein synthesis could indeed be induced in cultured brains (). We incubated wild–type brains in culture medium either with, or without, curare for 15 minutes, 12 hours, or 24 hours; protein levels were analyzed by immunoblot analysis. Strikingly, we found that Dα7 protein levels were increased by more than 60% after 24 hours of curare treatment (), while no differences were observed with shorter incubations. In addition, similar to the enhancement in mEPSCs, the increase in Dα7 protein persisted for at least 5 hours (). In contrast, other nAChR subunits, including Dα1, Dα2, Dβ1, and Dβ2 showed no significant change following synaptic inactivity (). Altogether, our results show a selective increase in Dα7 receptors that likely underlies the increased amplitudes of mEPSCs following prolonged synaptic block.

Inactivity-Induced Increase in Shal K+ Currents

Homeostasis can be achieved by synaptic and/or cell intrinsic changes. Indeed, other systems have demonstrated cell intrinsic changes after activity blockade[4-7,21]. How these intrinsic changes are triggered in by synaptic inactivity, however, has remained a mystery. We set out to investigate whether cell intrinsic changes also occur in our system, and planned to examine how they are triggered by synaptic blockade. Since all of the voltage–dependent K+ channels and currents have been genetically and electrophysiologically identified in these cultured neurons[11,12], we first examined whether IA currents carried by Shal/Kv4 channels, or delayed rectifier (DR) currents carried by Shab/Kv2 and Shaw/Kv3 channels, were changed in response to curare treatment. We blocked synaptic activity with curare for 24 hours using the same treatment Protocol #1 used for mEPSC recordings (see Methods), and performed voltage paradigms to isolate DR and IA currents. We found that IA currents carried by Shal K+ channels were dramatically increased in MNs (from 176 ± 22 to 438 ± 30 pA) and PNs (from 269 ± 26 to 413 ± 35 pA) (), with no change in cell size, as indicated by capacitance measurements (). No changes in IA and DR currents were observed in EL interneurons (). Interestingly, however, if antagonist was not washed out (Protocol #2, see Methods), no changes in IA or DR currents were observed in either MNs or PNs (). These results suggest that prolonged synaptic inhibition results in a selective increase in Shal K+ current density in excitatory neurons. Up–regulation of Shal required recovery of synaptic transmission, but was independent of AP firing, as it occurred even in the presence of TTX ().

To test whether the increase in IA is due to an increase in the number of Shal K+ channels, we examined whole brains following curare treatment. Levels of Shal protein from brains examined right after mock or curare treatment (Protocol #2) were not significantly different (, No wash). If, however, synaptic transmission was allowed to recover for 60 minutes (Protocol #1), there was ~80% increase in Shal protein (, Recovery). Like Dα7, this increase was seen only after 24 hours of curare treatment, and persisted for at least 5 hours (). Altogether, our results show that prolonged blockade of synaptic activity results in an increase in Shal K+ channels that requires recovery of synaptic activity. In contrast, treatment Protocols #1 and #2 both resulted in an up–regulation of Dα7, suggesting that the increase in Dα7 does not require a period of recovery following synaptic blockade ().

Since Shal K+ currents play an important role in neuronal firing[15], we speculated that the up–regulation of Shal channels should also affect the firing behavior of cells. We curare treated wild–type cultures for 24 hours, washed out antagonist with medium, and then monitored firing during the following 30 minutes (treatment Protocol #3, see Methods), when Shal channels should be actively up–regulated. We compared MN firing patterns, in response to a 500 ms current injection of 60 pA, at 5–10 minutes and then at 25–30 minutes. We found that the number of APs fired decreased during this recovery period (). Since Shal channels have been shown to affect both latency to AP firing and firing frequency in these neurons[15,22], these changes are consistent with an increase in Shal channels during this recovery period following synaptic blockade.

To gain insight into whether Shal channels were preferentially up–regulated in a particular subcellular region of neurons, we recorded Shal channel activity in cell–attached patches from the somas and dendrites of MNs. We compared Shal currents in patches from cells that were mock or curare treated; antagonist was washed out and synaptic transmission was allowed to recover for 30 minutes to induce the up–regulation of Shal (Protocol #1, see Methods). Cell–attached patches from dendrites were obtained from a branch point ~30 µm from the soma (). We found that Shal currents were increased with curare treatment in patches from both the cell body and dendrites (). Up–regulation of Shal currents, however, was 1.5–fold larger in dendrites than in the cell body ().

New Synthesis of Dα7 and Shal are Differentially Regulated

The increase in total Dα7 and Shal protein levels, following prolonged synaptic blockade (), suggests the involvement of new protein synthesis. To directly test this, and to examine whether regulation occurs at the transcriptional or translational level, we used Actinomycin D (Actino) to inhibit gene transcription, and either Anisomycin (Aniso) or cycloheximide (CXM) to block protein translation. Cultured brains were incubated with Actino (50 μM), CXM (100 μM), or Aniso (40 μM ) throughout mock and curare treatment Protocol #1 (). As expected, the increase in Dα7 and Shal proteins following curare treatment were both blocked by Aniso and CXM (), confirming that new protein synthesis is required in both cases. Actino, however, only blocked the increase in Shal K+ channels, but not Dα7 receptors (), suggesting transcriptional regulation of Shal.

To further test the differential regulation of Dα7 and Shal K+, we examined mEPSCs and IA currents in identified MNs when either transcription or translation was inhibited. We used Actino (50 μM) to block transcription, and Aniso (40 μM) to block protein translation. Synaptic blockade induced an increase in mEPSC and IA amplitudes, as expected (). When cultures were incubated with Aniso, the increase in both mEPSC and IA amplitudes were blocked (), confirming again the involvement of new protein synthesis. Inhibition of transcription by Actino, in contrast, only blocked the up–regulation of Shal K+ currents, and not mEPSC amplitudes (). Thus, immunoblot analyses and electrophysiological recordings both show that synaptic inactivity induces an increase in translation of Dα7 receptors, and an increase in transcription of the Shal gene.

Given that the increase in Shal channels occurs, not during the 24 hours of synaptic inhibition, but during the shorter 30–minute period of recovery of synaptic transmission (see above; ), we were surprised that regulation occurred at the transcriptional level. We further tested this finding by comparing Shal expression from the endogenous Shal gene, to that from an epitope (HA)–tagged Shal transgene (UAS–HA–Shal) lacking transcriptional regulatory domains. We generated a transgenic line in which UAS–HA–Shal expression is driven by RRa–GAL4. After 24 hours of either mock or curare treatment (Protocol #1), cultures were immunostained using an antibody against either HA or Shal. We found that synaptic blockade did not induce an increase in anti–HA signal, but did induce a clear increase in anti–Shal signal in both soma and dendrites (). The increase in Shal protein from the endogenous Shal gene, but not an exogenous transgene, is consistent with regulation at the transcriptional level.

Increase in Shal is Dependent on Dα7 nAChRs

We next examined the relationship between the increase in Dα7 receptors and the increase in Shal K+ channels that both follow prolonged synaptic inactivity. The increase in Dα7 receptors occurs first, since it is evident immediately after curare treatment (). In contrast, the increase in Shal channels and current required a subsequent 30–60 minute recovery period of synaptic transmission (). These results suggest that the up–regulation of Dα7 is triggered more “directly” by synaptic inactivity, and that up–regulation of Shal channels might occur in response to this increase in Dα7 receptors. To test this possibility, we first examined whether Shal currents are up–regulated in the absence of Dα7. We used RRa–GAL4 to drive expression of UAS–mCD8–GFP in a Dα7 null mutant background (Dα7), and recorded mEPSCs and Shal K+ currents from identified MNs before and after curare treatment (Protocol #1, see Methods); note that in this protocol, cultures were given a 30 minute recovery period to allow for Shal channel synthesis. In contrast to wild–type MNs, we found that Shal K+ currents in MNs lacking Dα7 were not increased following curare treatment (). Interestingly, mEPSCs were increased with synaptic blockade (Ctr: 8.71 ± 0.55 pA, n = 12; T: 10.98 ± 0.72 pA, n = 11; ), although to a lesser extent than in wild–type (), suggesting that either that there is some redundancy or compensation by other nAChR subunits that is revealed or developed in the absence of Dα7.

In addition, we compared Shal protein levels from wild–type and Dα7 mutant brains, similarly mock or curare treated. Immunoblot analyses showed no up–regulation of Shal protein in the absence of Dα7 (). Since Dα7 subunits are among three nAChR subunits in Drosophila (Dα5, Dα6, Dα7) which are ~60% identical to vertebrate α7 nAChR subunits[23], we also examined a null mutant of Dα6 (Dα6). Prolonged inactivity followed by recovery of synaptic transmission, however, still increased Shal protein levels in the absence of Dα6 (). Altogether, these results indicate that the increase in Shal K+ channels following synaptic blockade has a specific requirement for Dα7 receptors.

Ca2+ is Essential for the Increase in Shal K+ Channels

If Shal channel expression is up–regulated by an increase in Dα7 receptors, what is the signaling pathway from Dα7 to Shal transcription? We have shown that the increase in Dα7, combined with recovery of synaptic transmission, are key events in this pathway. Since α7 receptors have been reported to display greater Ca2+ permeability than other nAChRs[18,19], we tested whether Ca2+ is an essential secondary messenger in this pathway. To do this, we used the Ca2+ chealator BAPTA in our intracellular recording solution. Cultures were mock or curare treated, antagonist was washed out for three minutes, then individual MNs were monitored by whole–cell recording for the next 20–30 minutes, during which time Shal currents are actively up–regulated (Protocol #3, see Methods). shows scatter plots representing Shal K+ current amplitudes at 25–30 minutes versus at 5–10 minutes. Using normal intracellular solution in the pipet, synaptic blockade increased the ratio of (IA at 25–30 minutes)/(IA at 5–10 minutes) from 208 ± 17 to 378 ± 24 pA, as expected (). To further confirm that Shal K+ currents were indeed acutely up–regulated, we used Actino or Aniso to block transcription or translation, respectively, and monitored Shal currents following antagonist washout. Indeed, ratios of (IA at 25–30 minutes)/(IA at 5–10 minutes) from single cells were no longer increased (), demonstrating that the increase in Shal currents is a result of transcriptionally up–regulated channels. We then monitored Shal currents after curare treatment with BAPTA included in our intracellular solution. We found that curare treatment no longer increased Shal K+ currents, and (IA at 25–30 minutes)/(IA at 5–10 minutes) ratios were close to 1:1 (197 ± 11 pA / 205 ± 13 pA; ), as seen for untreated cells ().

Since nAChRs can also induce Ca2+ release from intracellular Ca2+ stores, we tested the contribution of Ca2+ from internal stores in signaling Shal expression. We monitored Shal K+ currents in individual MNs during the ~25–30 minutes of recovery following synaptic blockade. During this time, we used thapsigargin to deplete intracellular Ca2+ stores. Inactivity–induced increases in (IA at 25–30 minutes)/(IA at 5–10 minutes) ratios, however, were not blocked by thapsigargin (). Together, our results suggest that the rise in intracellular Ca2+, possibly through new Dα7 receptors, and not intracellular stores, is required for signaling the increase in Shal channels.

CaMKII is Essential for Up–Regulation of Shal

Since Ca2+–calmodulin dependent kinases are well known Ca2+ targets in a variety of synaptic plasticity and homeostatic pathways, we investigated their potential role in signaling the up–regulation of Shal. Cultures were mock and curare treated by Protocol #1 (see Methods). During the 30 minute recovery period of synaptic transmission following antagonist washout, we used Myr–CaMKIINtide and STO–609 to inhibit CaMKII and CaMKK activity, respectively. While synaptic blockade still increased Shal K+ current amplitudes in the presence of STO–609, the increase in Shal currents was blocked in the presence of Myr–CaMKIINtide (), suggesting that CaMKII is essential for the up–regulation of Shal channels. These results were further confirmed in whole brains. Curare treatment induced an increase in Shal protein in the presence of STO–609, but this increase was completely blocked with Myr–CaMKIINtide (). Thus, CaMKII is likely to be the target of the Ca2+ influx through new Dα7 receptors, and a key component in the signaling pathway triggering up–regulation of Shal channels.

We next tested whether activation of CaMKII was sufficient to up–regulate Shal K+ currents. To do this, we used the UAS–CaMKII.T287D transgene which encodes a constitutively active form of CaMKII (CaMKIIT287D)[24]. We used RRa–GAL4 to drive expression of UAS–mCD8–EGFP and UAS–CaMKII.T287D in MNs. We found that Shal currents were indeed increased, nearly two–fold, in these MNs compared to wild–type MNs (), suggesting that active CaMKII alone can up–regulate Shal channels.

The Increase in Shal Stabilizes Synaptic Potentials

Why does the homeostatic up–regulation of Dα7 receptors trigger an increase in a hyperpolarizing current? Most reported cell intrinsic changes to synaptic inactivity increase excitability, thereby contributing to the homeostatic response. An increase in Shal K+ currents, however, would most likely decrease excitability. One possibility is that the increase in Shal K+ channels serves as a mechanism to regulate the homeostatic response, keeping potentiation “in check” and preventing over–excitation. To investigate this possibility, we examined the effects of the increase in Shal K+ currents on synaptic potentials (using treatment Protocol #1, see Methods). Strikingly, we found that while amplitudes of mEPSCs were significantly larger following curare treatment (), mEPSPs were stable (Ctr: 1.78 ± 0.12 mV, T: 1.80 ± 0.14 mV; ); this was not due to a difference in resting membrane potential since control and curare treated neurons displayed similar passive membrane properties ().

To test this hypothesis more directly, we recorded from MNs expressing a dominant– negative Shal subunit (DNKv4) that completely blocks Shal K+ channel function[15]. In the absence of Shal K+ channel function, mEPSCs are increased after curare treatment (), similar to wild–type MNs. Thus, homeostatic up–regulation of Dα7 receptors still occurs, independent of Shal K+ channel function. We next examined mEPSPs from control and curare treated DNKv4 MNs. Indeed, synaptic potentials were not stabilized as they were in wild–type MNs. mEPSP amplitudies were significantly larger after synaptic inhibition, in the absence of Shal function (Ctr: 2.5 ± 0.21 mV, T: 4.51 ± 0.30 mV; ), indicating that Shal channels are indeed required for stabilizing synaptic potentials. To test whether this function was specific to Shal channels, we performed the same experiments using a Shab/K null mutant (Shab). Although loss of Shab removes nearly all of the delayed–rectifier K+ current present in these neurons[11], mEPSPs remained stabilized after curare treatment ().

Finally, we also monitored mEPSPs in single neurons during the 30 minutes following antagonist washout (Protocol #3, see Methods). mEPSPs from curare treated neurons were increased at 5–10 minutes, presumably due to the increase in Dα7 receptors, then stabilized to control amplitudes after 25–30 minutes after antagonist washout (). To confirm that the acute mEPSP up–regulation at 5–10 minutes, then stabilization at 25–30 minutes was regulated by translation of Dα7 and transcription of Shal, we repeated these experiments, using Actino or Aniso, to block transcription or translation, respectively. When transcription was blocked, mEPSP amplitudes were increased compared to untreated cells, but were not stabilized after 25–30 minutes (). When translation was inhibited, the initial increase in mEPSP amplitude (at 5–10 minutes) was unchanged compared to untreated cells, and remained constant during the 30 minutes of monitoring (). Altogether, these results suggest that following curare treatment, Dα7 receptors are translationally up–regulated, resulting larger mEPSPs, then Shal is transcriptionally up–regulated, resulting in the stabilization of mEPSPs. When the mEPSPs were monitored in MNs expressing DNKv4, increased mEPSP amplitudes were no longer stabilized (). Together, our studies show that the increase in Dα7 receptors following synaptic inactivity triggers an increase in Shal K+ channels that serves to stabilize synaptic potentials.

Discussion

Synaptic homeostatic mechanisms serve to stabilize neural circuits in response to changes in activity. Studies of synaptic homeostasis in Drosophila have largely focused on the NMJ and underlying presynaptic mechanisms[2]. Reports of homeostatic changes at inter–neuronal synapses in Drosophila have mostly been limited to structural changes[25-27]. In this study, we show that inhibition of nAChRs resulted in mEPSCs that were increased in amplitude and frequency, indicating both pre and postsynaptic effects. Focusing on postsynaptic changes, we found that the Dα7 receptor was preferentially increased, mediating mEPSCs with larger amplitudes, faster decay rates, and likely increased Ca2+ influx. This homeostatic response, however, does not require the Dα7 receptor, since mEPSCs are increased in the Dα7 null mutant, suggesting that other nAChR(s) are up–regulated in the absence of Dα7. Thus, excitatory neurons exhibit a resilient homeostatic increase in nAChRs to compensate for conditions of inactivity. It is interesting to note the lack of this same plasticity in EL interneurons. Since interneurons are generally thought to be inhibitory (see[28]), this would also contribute to the homeostatic response. Our results parallel the homeostatic increase in GluA1/GluA2 receptors widely observed, and restricted to excitatory mammalian neurons[1,3].

Activity–dependent plasticity of muscular nAChRs has long been studied during development of the vertebrate NMJ[29]. Perhaps more relevant to our findings, however, is the increase in neuronal nAChRs that occurs following prolonged nicotine exposure[18]. Studies have suggested that desensitization of the receptors to nicotine exposure triggers a homeostatic response that up–regulates nAChRs[8]. In our study, direct antagonist exposure results in an increase in the Dα7 nAChR. Dα7 is highly homologous to the vertebrate α7 subunit, sharing ~60% amino acid identity. In the mammalian CNS, homomeric α7 receptors are one of the most abundant and widely expressed classes of nAChRs[19]. Homeostatic up–regulation of α7 has, interestingly, also been speculated to contribute to the progression of Alzheimer's disease[30]. Thus, Drosophila central neurons may provide a useful model for studying mechanisms of neuronal nAChR–mediated homeostasis that contribute to pathological conditions, such as nicotine addiction and Alzheimer's disease.

We show that the inactivity–induced increase in Dα7 is followed by an increase in the voltage–dependent Shal K+ channel. Shal K+ channels are highly conserved and underlie the somato–dendritic A–type K+ current in most neurons[31]. Shal currents have been shown to play important roles in regulating dendritic excitability, backpropagating action potentials, and postsynaptic potentials[32-34]. Our data suggests that inactivity induces a translational increase in Dα7 receptors that mediates larger mEPSCs. While the homeostatic increase in mEPSCs is not dependent on Dα7, we show that the up–regulation of Shal channels requires Dα7. Dα7 receptors likely carry a greater influx of Ca2+, which in turn activates CaMKII, leading to rapid (<30 minutes) up–regulation of Shal channels. Since CaMKII has previously been implicated in receptor homeostasis and Shal channel regulation[35-37], the localization and regulation of CaMKII will likely be critical to its roles in these different pathways.

What is the function of this up–regulation of Shal K+ channels observed after synaptic blockade and an increase in Dα7 receptors? Previously, cell intrinsic changes have mostly been suggested to increase excitability and contribute to homeostasis[1,4-6,38]. The up–regulation of Shal K+ current, however, is intriguing because it is likely to counter homeostatic mechanisms. mEPSPs recorded in cell bodies are likely to represent larger mEPSPs in dendrites[39] that would activate local Shal channels, which in turn, would modulate PSPs. EPSPs, both large and small, have been shown to activate Shal K+ channels in mammalian neurons, similarly reducing EPSPs[34,40]. Indeed, we found that the increase in Shal current stabilized synaptic potentials that would otherwise be increased by homeostatic pathways. Modulation of synaptic currents to match a fixed, or stable, synaptic potential has been proposed to underlie a homeostatic solution to the difference in dendritic arbor sizes, and input resistances, of Drosophila PNs[41]. A more generalized function may be that when activity is lowered/blocked, nAChRs are increased to boost activity, then when circuits run the risk of becoming over–active, Shal K+ channels are up– regulated to temper synaptic potentials. This represents a novel mechanism for fine–tuning the homeostatic response and preventing over–excitation.

The increase in Shal channel transcription in response to increased translation of Dα7 brings up many intriguing questions. For example, why is the up– and down–regulation of activity controlled at two different loci? That is, why not up– and down–regulate Dα7 for needed increases, or decreases in activity? Since up–regulation of Dα7/nAChRs is rather slow (eg. requiring 24 hours of synaptic blockade to up–regulate), one possibility is that it might not be beneficial for the cell to down–regulate the receptors, given that replacement would likely be slow and leave the cell unable to up–regulate activity rapidly if needed. Rapid up–regulation of Shal channels has perhaps evolved to be a better solution. It will also be important to understand if there is any specificity, or coordination with sites of activity/inactivity, to the trafficking and subcellular localization of Shal K+ channels.