Abstract
Activation of cortical nicotinic receptors by cholinergic axons from the basal forebrain (BF) significantly impacts cortical function, and the loss of nicotinic receptors is a hallmark of aging and neurodegenerative disease. We have previously shown that stimulation of BF axons generates a fast α7 and a slow non-α7 receptor-dependent response in cortical interneurons. However, the synaptic mechanisms that underlie this dual-component nicotinic response remain unclear. Here, we report that fast α7 receptor-mediated EPSCs in the mouse cortex are highly variable and insensitive to perturbations of acetylcholinesterase (AChE), while slow non-α7 receptor-mediated EPSCs are reliable and highly sensitive to AChE activity. Based on these data, we propose that the fast and slow nicotinic responses reflect differences in synaptic structure between cholinergic varicosities activating α7 and non-α7 classes of nicotinic receptors.
Introduction
Cholinergic axons from the basal forebrain (BF) activate nicotinic acetylcholine receptors (nAChRs) throughout the cortex and play an important role in sensory processing (Metherate and Hsieh, 2004; Disney et al., 2007), attention (Howe et al., 2010), and learning (Gu and Yakel, 2011; Letzkus et al., 2011; for review see Levin, 2002). Moreover, loss of cholinergic function is associated with several neurodegenerative diseases including Alzheimer's disease, for which cholinergic cell death and the loss of nAChRs in cortex are hallmarks of disease progression (Martin-Ruiz et al., 1999).
Despite its functional importance, the synaptic properties of nicotinic receptor-mediated signaling in the cortex remain poorly understood. Anatomical studies have shown that a fraction of cholinergic varicosities in the cortex form synaptic structures, while others are not associated with any postsynaptic membrane (Mrzljak et al., 1993; Umbriaco et al., 1994; Turrini et al., 2001). Although the precise ratio of synaptic to nonsynaptic cholinergic varicosities has been disputed, recent findings suggest that the cholinergic system may operate primarily by volume transmission (Yamasaki et al., 2010; for review, see Lendvai and Vizi, 2008).
In the cortex, nAChRs are classified into two families: homomeric receptors composed of the α7 subunit and heteromeric receptors composed of the β2 subunit together with either the α4 or α5 subunit (Cordero-Erausquin et al., 2000). Homomeric nAChRs exhibit high calcium permeability, low ACh sensitivity, and rapid desensitization; whereas heteromeric receptors exhibit low calcium permeability, high ACh affinity, and relatively slow kinetics (Dani and Bertrand, 2007). These diverse properties may allow homomeric α7 and heteromeric non-α7 nicotinic receptors to respond to distinct spatiotemporal ACh profiles produced by activation of synaptic and nonsynaptic cholinergic varicosities. However, without a method allowing selective stimulation of cholinergic axons, the roles of specific nicotinic receptor subtypes in cholinergic signaling remain unclear.
We recently found that stimulation of BF cholinergic axons produces dual-component responses consisting of a fast α7 receptor-mediated response and a slow non-α7 receptor-mediated response (Arroyo et al., 2012). These findings gave rise to the hypothesis that the fast component is mediated by conventional synaptic contacts whereas the slow response is mediated by transmitter diffusing over some distance (i.e., volume transmission). Here we investigated mechanisms that may underlie the dual-component nicotinic response in cortical interneurons.
Materials and Methods
Animals.
Both a BAC transgenic (GENSAT GM24; Tamamaki et al., 2003) and a knock-in mouse line (Rossi et al., 2011) expressing Cre recombinase (Cre) under the choline acetyltransferase promoter were used to transduce cholinergic neurons in the BF with a channelrhodopsin-2-enhanced yellow fluorescent protein (ChR2-EYFP) construct. The expression of Cre in the BF was similar for both ChAT-Cre mouse lines, and subsequent data were pooled. All procedures were approved by the Administrative Panel on Laboratory Animal Care at Stanford University.
Viral transduction of the basal forebrain.
Both male and female mice aged P20–P60 were anesthetized and placed in a stereotaxic frame. AAV2/5 virus (1–2 μl) bearing a pAAV-EF1α-DIO-hChR2 (H134R)-EYFP-WPRE (Zhang et al., 2010) construct were pressure injected into the brain using stereotaxic coordinates for several basal forebrain nuclei, including the nucleus basalis, the horizontal diagonal band of Broca, and the substantia innominata. Typically, four sites were injected, with no more than 500 nl of virus in each location.
Slice preparation.
Six to 20 weeks after surgery, mice were deeply anesthetized with isoflurane. Brains were removed in ice-cold, carbogenated sucrose composed of the following (in mm): 76 NaCl, 25 NaHCO3, 25 glucose, 75 sucrose, 2.5 KCl, 1.75 NaHPO4, 0.5 CaCl2, 7 MgSO4 2 pyruvic acid, 4 lactic acid, 4 β-hydroxybutyric acid. Sagittal slices (300-μm-thick) were generated (Integraslice 7550 MM; Campden Instruments) and transferred to a chamber with the same solution maintained at 32–35°C. After 30 min, the slices were transferred to artificial CSF (ACSF) composed of the following (in mm): 125 NaCl, 2.5 KCl, 1.25 NaH2PO4, 1 MgSO4 2 CaCl2, 26 NaHCO3, 20 glucose, 4 lactic acid, 2 pyruvic acid, 0.4 ascorbic acid, and 4 β-hydroxybutyric acid at 32–35°C. Slices were allowed to equilibrate to room temperature before being transferred to the microscope chamber.
Electrophysiological recordings.
Glass electrodes (3–7 MΩ) were filled with an internal solution composed of the following (in mm)2.7 KCl, 120 potassium methyl sulfate, 9 HEPES, 0.18 EGTA, 4 MgATP, 0.3 NaGTP, 20 sodium phosphocreatine, pH 7.3, 295 mOsm/L. To test for the contribution of electrical coupling to the slow response, the above solution was modified by replacing KCl with CsCl (3.3 mm) and potassium methyl sulfate with cesium methanesulfonate (133 mm). Whole-cell patch-clamp recordings were obtained by using two Axopatch 200B or 700A patch amplifiers (Molecular Devices) in current-clamp or voltage-clamp modes. Data acquisition and offline analysis were performed using custom software written in IGOR Pro (Wavemetrics). For photostimulation, blue light was emitted either from an LED attached to a fiber optic cable (920-μm-diameter) or a xenon lamp with a mechanical shutter (Uniblitz). The spot used to illuminate the slice ranged from 200 to 450 μm in diameter and the light intensity ranged from 10 to 400 mW/mm2. Photostimulations were typically 3–10 ms long, and 2–5 min were allowed between stimulations, except for paired pulse experiments for which the interstimulus interval was varied systematically from 0.5–16 s. For pharmacological experiments, TTX, DNQX, methyllycaconitine (MLA), and dihydro-β-erythroidine (DHβE;), and ambenonium dichloride were diluted in ACSF to 0.5 μm, 10 μm, 5 nm, 500 nm, and 100 nm, respectively, and bath applied. For the acetylcholinesterase (AChE) experiment, AChE (0.4 U/μl; Sigma-Aldrich C2888) was pressure ejected continuously from a patch pipette at low pressure (∼2 psi).
Nucleated patches and rapid perfusion.
Nucleated patches were pulled as described previously (Ford et al., 2009). Briefly, after obtaining whole-cell recordings, gentle suction was applied while slowly retracting the pipette to extract the nucleus from the cell. The nucleated patch was then positioned beside the agonist stream flowing from an application pipette, ∼500 μm above the slice surface. The application pipette was attached to a commercial piezoelectric device (Burleigh PZ-150M) or a bimorph strip. A brief voltage step was applied to induce a small deflection of the application pipette, moving the agonist stream over the nucleated patch. The voltage step applied to both actuators was low-pass filtered at 100 Hz to reduce oscillations.
Analysis.
To quantify the effect of ambenonium dichloride on the time course of nicotinic EPSCs, the interval between the peak current and subsequent half-maximum amplitude (thalf max) was calculated as follows. Traces were filtered at 1 kHz, and the response decay was fit by a double exponential function, f(t). thalf max was defined as the time interval from the peak to t1/2, where t1/2 was the time at which f(t) = (1/2) × (peak − baseline) + baseline.
To calculate the coefficient of variation, the amplitudes of the fast and slow components were found independently. For the fast component, a 2 ms window was averaged around the minimum between 0 and 20 ms after photostimulation. For the slow component, 20 ms was averaged around the minimum between 20 and 1000 ms after photostimulation. The amplitudes of both components were detrended before variance analysis. Two noise amplitudes, one generated using the parameters for the fast component and one for the slow component, were obtained over baseline intervals and noise variances were calculated. These were subtracted from the respective response variances to correct for noise variance. All values are reported as mean ± SE unless otherwise indicated. The nonparametric Wilcoxon signed-rank test was used for all paired statistical comparisons. The Wilcoxon rank-sum test was used for all nonpaired statistical comparisons.
Results
Perturbations of AChE activity exclusively impact the non-α7 response
To study the mechanisms underlying dual-component nicotinic responses in the cortex, we recorded from L1 interneurons while photostimulating ChR2-expressing cholinergic BF axons (see Materials and Methods, above). We recently demonstrated that this interneuron subtype exhibits dual-component nicotinic responses to photostimulation (Arroyo et al., 2012), an example of which is shown in Figure 1, A and B. These EPSCs were abolished by application of the sodium channel blocker TTX (0.5 μm), indicating that both the fast and slow components were dependent on action potentials in cholinergic axons (slow EPSC control amplitude: −8.4 ± 2.8 pA; slow EPSC TTX amplitude: −0.2 ± 0.1 pA; n = 9, p < 0.01, Wilcoxon signed-rank test; fast EPSC control amplitude: −9.9 ± 3.6 pA; fast EPSC TTX amplitude: −0.3 ± 0.1 pA; n = 6, p < 0.05, Wilcoxon signed-rank test; Fig. 1C,D). The rise and decay of the slow nicotinic EPSC are >10-fold slower than the rise and decay of the fast EPSC (fast rise time: 2.6 ± 0.5 ms; fast decay τ: 4.9 ± 0.6 ms; slow rise time: 35 ± 5 ms; slow decay τ: 190 ± 17 ms) (Arroyo et al., 2012). We hypothesized that the difference in the kinetics of the α7 and non-α7 components results from different localization of these receptors with respect to ACh release sites; i.e., the slow kinetics of the non-α7 EPSC reflect diffusion of ACh from distant varicosities to activate nonsynaptic receptors, while the fast kinetics of the α7-mediated component reflect conventional synaptic transmission (Vizi, 2000).
If non-α7 receptors are indeed located farther from release sites, then the slow non-α7 receptor-mediated response should be more susceptible to hydrolysis of ACh by AChE. We found that application of the AChE inhibitor ambenonium dichloride (100 nm) (Hodge et al., 1992) dramatically prolonged the decay of the slow component (Fig. 2A). To determine the effect of AChE blockade on the fast component, we first bath applied the selective non-α7 receptor antagonist DHβE (500 nm) to block the slow non-α7 response. Subsequent addition of ambenonium had no effect on the kinetics of the fast response (Fig. 2B). To quantify the effect of AChE inhibition, we measured the interval between the peak of the nicotinic EPSC and the subsequent half-maximum, t1/2, before and after the addition of ambenonium (slow control t1/2: 119 ± 9 ms; n = 8; slow ambenonium t1/2: 1050 ± 250 ms; n = 7, p < 0.001, Wilcoxon rank-sum test; fast control t1/2:5.8 ± 1.1 ms, n = 9; fast ambenonium t1/2: 5.3 ± 1.1 ms; n = 5, p > 0.05, Wilcoxon rank-sum test; Fig. 2C). These data suggest that the time course of the slow component, but not the fast component, depends on the hydrolysis of ACh by AChE.
If the slow component relies on the diffusion of ACh from distant varicosities, then enhancing AChE activity should diminish the slow response. To test this hypothesis, we obtained whole-cell recordings from L1 interneurons and applied exogenous AChE via pressure injection from a nearby patch pipette (Fig. 3A). AChE was ejected continuously at low pressure for the duration of drug application trials. As predicted, application of AChE significantly reduced the amplitude of the slow component (control: 18 ± 2.6 pA; AChE application: 6.3 ± 1.2 pA; wash: 9.0 ± 1.4 pA; n = 9, p < 0.005, Wilcoxon signed-rank test; Fig. 3B,D) but did not significantly affect the amplitude of the fast component (control: 24 ± 5.3 pA; AChE application: 18 ± 5.5 pA; wash: 20 ± 5.1 pA; n = 6, p > 0.05, Wilcoxon signed-rank test; Fig. 3B, inset, D). As a control experiment, we showed that pressure application of ACSF did not affect the slow component (control: 19 ± 3.4 pA; ACSF application: 15 ± 2.6 pA; wash: 14 ± 2.1 pA; n = 7, p > 0.05, Wilcoxon signed-rank test; Fig. 3C,D) or the fast component (control: 18 ± 8.6 pA; ACSF application: 19 ± 9.9 pA; wash: 19.4 ± 8.9 pA; n = 3; Fig. 3C, inset, D).Together, these data indicate that the slow response, but not the fast response, is sensitive to perturbations of AChE activity.
Concentration dependence of α7 receptor-mediated response kinetics
The selective effect of perturbing AChE suggests that the slow component depends on transmitter diffusing over relatively long distances, whereas the fast component depends on a short diffusion pathway (Figs. 2, 3). However, since α7 receptors are known to desensitize rapidly (Quick and Lester, 2002), it is also possible that the fast component reflects desensitization of nonsynaptic α7 receptors activated by ACh released from distant varicosities. In this scenario, the ACh transient mediating the fast α7 response would be relatively slow. Moreover, due to dilution and hydrolysis of ACh by AChE, the concentration of transmitter would be many fold lower than the concentration at synaptic contacts (Eccles and Jaeger, 1958). Thus, this hypothesis predicts that prolonged activation of α7 receptors by low concentrations of ACh would mimic the fast component of the nicotinic response.
To test this possibility, we obtained nucleated patches from L1 interneurons and used a piezoelectric device to rapidly apply ACh (Fig. 4A). Patches were exposed to prolonged application of ACh, sufficient to completely desensitize receptors. Under these conditions, application of ACh produced an inward current that was abolished by MLA (5 nm; Fig. 4B) and displayed inward rectification (Fig. 4C), consistent with properties of α7 receptors. We did not observe non-α7 responses in nucleated patches, indicating that these receptors either are washed out from the soma or are not natively present in somatic membrane (Buisson et al., 1996). Both the rise time and the decay time constant of the response to rapid application varied with agonist concentration (20–80% rise time ranged from 2.9 ± 2.2 ms at 35 μm ACh to 0.12 ± 0.02 ms at 800 μm ACh; n = 15; decay time constant ranged from 36.2 ± 6.3 ms at 20 μm ACh to 0.5 ± 0.1 ms at 800 μm ACh; n = 19; Fig. 4D). From these data, we estimated the concentration of ACh that would evoke a nucleated patch response with the same decay kinetics as the α7-mediated EPSC (Fig. 4E). The average decay time constant of the synaptic α7 response was 4.9 ± 0.6 ms with a range of 2.8–6.9 ms (n = 6; Fig. 4D). Thus, if the α7 EPSC is generated by a prolonged ACh transient, the concentration of transmitter required to produce the average synaptic response would be between 50 and 200 μm (Fig. 4D). This estimate is more than tenfold greater than the concentration of ACh that has been proposed to mediate nonsynaptic transmission (Descarries et al., 1997). Moreover, since the decay time constants of somatically recorded EPSCs are likely overestimates of the unfiltered EPSC kinetics, the ACh concentration required to mimic synaptic responses probably exceeds our estimate. These data suggest that the kinetics of the fast component are unlikely to result from rapid desensitization of α7 receptors to a prolonged transient of dilute ACh. Thus, our results support the hypothesis that these receptors are located at synaptic contacts, where the peak concentration of transmitter has been estimated to range from 100 to 1000 μm (Eccles and Jaeger, 1958; Clements, 1996).
Paired-pulse depression is similar for fast and slow EPSCs
To investigate whether the varicosities mediating the fast and slow components exhibit different release probabilities, we measured the recovery from paired-pulse depression [paired pulse ratio (PPR)] for the fast and slow EPSCs. We found that the two components recovered from paired-pulse depression at similar rates, as measured by single exponential fits to population PPR data (τfast PPR = 2426 ± 460 ms; τslow PPR = 3260 ± 606 ms; uncertainties represent 1 SD; Fig. 5A).
To ensure that this paired-pulse recovery did not simply reflect recovery from receptor desensitization, we compared the PPR time course for the fast synaptic EPSC with the recovery of α7 receptors from desensitization in nucleated patches (Fig. 5B). We found that α7 receptors recovered from desensitization on a much faster timescale than the synaptic paired-pulse recovery (τpatch PPR = 180 ± 30 ms), indicating that receptor desensitization does not cause the paired-pulse depression observed in α7 receptor-mediated synaptic responses. Together, these data suggest that the fast and slow components are mediated by varicosities with similar release probabilities.
Fast and slow nicotinic EPSCs exhibit different response variability
The variability of synaptic responses depends on both the release probability of varicosities contributing to the postsynaptic response and the number of release sites. Given that the varicosities mediating the fast and slow components have similar release probabilities, nicotinic responses generated by few release sites should be more variable than responses mediated by many release sites (Del Castillo and Katz, 1954). Thus, if the non-α7 response reflects release from multiple nonsynaptic varicosities, but the α7 response reflects conventional synaptic transmission, then the slow component should exhibit lower variability than the fast component.
To study the response variability, we recorded from L1 interneurons under voltage clamp, allowing clear identification of the fast and slow components (Fig. 6A). To determine how the two components varied across the population of L1 interneurons, we plotted the amplitude of the slow EPSC against the amplitude of the fast EPSC for 39 cells (Fig. 6A,B). We found no correlation between the fast and slow response amplitudes (r = 0.14, linear correlation coefficient). Moreover, while several cells exhibited only a slow component (n = 14), we did not observe any responses consisting of only a fast component, suggesting that fewer ACh release sites contribute to the fast response.
To further examine response variability, we investigated how the two response components varied over multiple stimulations in single cells. Photostimulation of cholinergic fibers at 0.1–0.2 Hz elicited responses in L1 interneurons that varied in amplitude from trial to trial (Fig. 6C). We found no correlation between the amplitude of the fast and slow components across multiple stimulations (r = 0.15 ± 0.03, average linear correlation coefficient for 6 cells). The fast component exhibited greater response variability, reflected by a significantly larger coefficient of variation (CV) (CVfast: 0.49 ± 0.08; CVslow: 0.18 ± 0.04; n = 6, p < 0.01, Wilcoxon rank-sum test; Fig. 6D). Although we did not observe failures in the slow response, in several cells the fast component failed in a fraction of the trials (Fig. 6E).
Could the low variability of the slow response be explained by electrical coupling among L1 interneurons? Since these cells are known to be interconnected by gap junctions (Chu et al., 2003), the reliability of the slow component could reflect input onto many electrically coupled cells (d'Incamps et al., 2012). To test this possibility, we recorded nicotinic EPSCs while voltage clamping the postsynaptic neuron at −70 and 0 mV (see Materials and Methods, above). If the slow response is primarily mediated by current through gap junctions, then the response amplitude should be invariant to the holding potential (d'Incamps et al., 2012). However, in every cell tested, we found that the slow EPSC was dramatically diminished at 0 mV (amplitude0 mV/amplitude−70 mV: 0.13 ± 0.02; n = 3; data not shown), suggesting that the majority of the response was mediated directly by nicotinic receptors on the recorded cell.
Selectively averaging trials in which the fast component failed or did not fail produced identical slow responses (amplitudeslow: 13.2 ± 9.0 pA; amplitudefast + slow: 13.5 ± 8.1 pA; decay τslow: 218 ± 21 ms; decay τfast + slow: 212 ± 18 ms; n = 3 cells; Fig. 6F), indicating negligible overlap between the varicosities mediating the two components. Since the release probability is similar across all varicosities (Fig. 5), these data suggest that more release sites contribute to the slow component than the fast component.
Discussion
We have investigated the mechanisms that underlie dual-component nicotinic signaling in L1 cortical interneurons. Previously, we have shown that excitation of cholinergic basal forebrain axons produces a fast α7 receptor-dependent response and a slow non-α7 receptor-dependent response (Arroyo et al., 2012). We propose that the fast α7 receptor-dependent response is mediated by conventional synaptic contacts whereas the slow non-α7 receptor-dependent response is mediated by nonsynaptic varicosities (Fig. 7).
Mechanisms underlying the fast component
Several lines of evidence suggest that the fast α7 receptor-dependent component is mediated by a conventional synapse. First, we found that inhibition of AChE, which dramatically prolonged the waveform of the slow component (∼10-fold), did not affect the fast component (Fig. 2). Moreover, in a complementary experiment, we found that addition of exogenous AChE significantly attenuated the slow component but had no significant effect on the amplitude or time course of the fast component (Fig. 3). Finally, the fast component exhibited high trial-to-trial response variability and occasional failures. These data suggest that the fast component is mediated by a limited number of release sites operating over a short diffusion pathway, consistent with transmission at a conventional synapse.
The α7 receptors mediating the fast component exhibit exceptionally fast rates of desensitization. Thus, an alternative explanation for our data is that the rapid decay of the fast component reflects fast desensitization of α7 receptors in response to a prolonged transient of ACh. By rapidly applying ACh onto nucleated patches, we determined that a relatively high concentration of ACh is required to mimic the time course of the fast component (Fig. 4). Thus, we conclude that conventional synaptic transmission, where the concentration of transmitter at the postsynaptic receptor is high, best explains the kinetics of the fast nicotinic EPSC. Notably, perisynaptic α7 receptors, located near release sites but outside the postsynaptic density, have been shown to generate fast nicotinic responses in the chick ciliary ganglion (Coggan et al., 2005; Stanchev and Sargent, 2011). Our data cannot distinguish synaptic from perisynaptic localization of α7 receptors since both depend on short diffusion pathways from single release sites.
Mechanisms underlying the slow component
In contrast to the fast component, we found that the slow component was highly sensitive to perturbations of AChE (Figs. 2, 3), suggesting that the slow component is mediated by transmitter diffusing over a relatively long distance. Under these conditions, it is expected that many release sites would contribute to the slow response, resulting in low trial-to-trial variability. Indeed, we found that the slow component was much less variable than the fast component. This disparity was not explained by a difference in release probabilities between varicosities mediating the two components, as the fast and slow EPSCs recovered from paired-pulse depression at similar rates (Fig. 5).
Recent anatomical studies indicate that a prominent fraction of cholinergic varicosities do not form synaptic junctions with postsynaptic neurons (Yamasaki et al., 2010; for review, see Lendvai and Vizi, 2008). Together with these anatomical findings, our data suggest that the slow non-α7 response is mediated by nonsynaptic varicosities whereas the fast α7 response is mediated by relatively sparse synaptic contacts with cortical interneurons (Fig. 7).
Comparison with other nicotinic synapses
There are relatively few studies investigating the synaptic mechanisms underlying endogenous nicotinic signaling in the CNS. Here we demonstrate a dual-component nicotinic EPSC comprised of a fast α7 and slow non-α7 receptor-mediated response and provide physiological evidence that these two components reflect synaptic and nonsynaptic transmission, respectively. Interestingly, the kinetics of these cortical responses closely resemble the kinetics of the fast (Alkondon et al., 1998; Frazier et al., 1998) and slow (Bell et al., 2011; English et al., 2012) nicotinic responses reported in the hippocampus and striatum. While these studies did not examine the synaptic mechanisms generating the nicotinic response, their similarity to our data suggests that nicotinic signaling may operate by conserved mechanisms throughout the forebrain.
In the interpeduncular nucleus (IPN), Ren et al. (2011) observed a very slow nicotinic response (5 s decay τ) elicited only by tetanic stimulation of cholinergic fibers, suggesting nonsynaptic ACh transmission. Notably, our slow response is tenfold faster than that reported in the IPN and is elicited by single stimulations of cholinergic axons, indicating a significant difference in structure between these two cholinergic synapses. In the spinal cord, heteromeric nicotinic receptors generate a biphasic response whose slow component is selectively modulated by perturbation of AChE, leading the authors to conclude that these receptors are likely activated by spillover of ACh (d'Incamps et al., 2012). Similarly, the slow component of the cortical nicotinic EPSC, which we propose is mediated by nonsynaptic transmission, is selectively modulated by addition or blockade of AChE.
In the spinal cord, d'Incamps et al. (2012) found that a significant portion of the slow nicotinic EPSC remained when the cell was depolarized to 0 mV, suggesting that much of the response is transmitted through gap junctions. Indeed, the authors demonstrated that the response at 0 mV was abolished by gap junction blockers. In the cortex, L1 interneurons are known to be interconnected by gap junctions; thus, some fraction of the nicotinic EPSC must be mediated by gap junctions. However, we found that the slow nictonic EPSC was largely diminished at 0 mV, suggesting that, in the cortex, gap junctional coupling is unlikely to mediate a large fraction of the slow nicotinic EPSC.
Functional significance
Recent studies suggest that subcortical neuromodulatory systems may exert temporally precise control over postsynaptic targets. For example, serotonergic fibers from the median raphe nucleus have been shown to evoke rapid EPSCs in hippocampal interneurons (Varga et al., 2009). Similarly, in the interpeduncular nucleus, cholinergic axons from the habenula co-release glutamate and ACh to produce both rapid EPSCs and a slow nicotinic inward current (Ren et al., 2011). Thus, by providing evidence for conventional cholinergic synapses mediating fast α7 receptor-mediated EPSCs in the cortex, our data further supports an emerging view that ascending subcortical systems operate not only by slow modulation but also by fast excitation. Such rapid modulatory input could serve to synchronize specific postsynaptic cell types. Additionally, spatially defined synaptic contacts onto the calcium-permeable α7 nicotinic receptor could provide a local plasticity signal for nearby inputs. Indeed, recent studies suggest that activation of α7 nicotinic receptors may have profound consequences for plasticity of glutamatergic and GABAergic inputs in the hippocampus (Wanaverbecq et al., 2007; Gu and Yakel, 2011).
The time course of the ACh transient after release from BF axons is widely debated (Sarter et al., 2009). Here, we show that the decay of the slow EPSC is prolonged by inhibition of AChE. Thus, the waveform of the slow EPSC (duration, ∼500 ms) provides an upper limit for the time course of ACh following photostimulation and release. This time course is relatively fast, indicating that the cholinergic system may exert temporally precise modulation of behavioral responses, as previously proposed (Parikh et al., 2007).
Is there any functional advantage to signaling via volume transmission? Cholinergic neurons represent a small fraction of the total number of cells in the basal forebrain (Henny and Jones, 2008), yet they prominently innervate the entire cortical mantle and exert a profound influence on cortical processing (Mesulam et al., 1983; Rye et al., 1984; Jones and Wonnacott, 2004; Disney et al., 2007; Goard and Dan, 2009). Cholinergic modulation of cortical activity likely involves activation of acetylcholine receptors located at both postsynaptic and presynaptic sites (Gil et al., 1997; Ji et al., 2001; Wanaverbecq et al., 2007; Kruglikov and Rudy, 2008; Lucas-Meunier et al., 2009). Operating by volume transmission in addition to conventional synaptic transmission enables a limited number of cholinergic axons to efficiently interact with a larger number of receptors than would be possible by synaptic transmission alone.
Currently, AChE inhibitors are the only treatment shown to improve cognitive function in patients with Alzheimer's disease (Birks and Harvey, 2003; Courtney et al., 2004; Howard et al., 2012). We found that inhibiting AChE preferentially impacts the non-α7 nicotinic response. Thus, our data suggest that selectively activating non-α7 nicotinic receptors may provide an alternative treatment without the unwanted side effects of global AChE downregulation. Moreover, given that α7 nicotinic receptors are unaffected by AChE inhibition, new treatments targeting both α7 and non-α7 receptors may confer additional therapeutic benefits.
Footnotes
This work was supported by NIH Grant EY012114, the Weston Havens Foundation, the Stanford Graduate Fellowship, the National Science Foundation Graduate Research Fellowship Program, the Stanford Medical Scientist Training Program, and the Ruth L. Kirschstein Graduate Fellowship. We thank Tony Ricci for contributing equipment used in the rapid application experiments.
The authors declare no competing financial interests.
- Correspondence should be addressed to Shaul Hestrin, 300 Pasteur Drive, Edwards R314, Department of Comparative Medicine, Stanford University School of Medicine. Stanford, CA 94305. shaul.hestrin{at}stanford.edu