QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

HOME  |   SEARCH  |   ARCHIVE  |   SUBSCRIBE  |   CONTACT  |   HELP

The Journal of Neuroscience, June 27, 2007, 27(26):7054-7069; doi:10.1523/JNEUROSCI.1744-07.2007

This Article Abstract Full Text (PDF) Supplemental Data Submit an eLetter Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in ISI Web of Science Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Rosenkranz, J. A. Articles by Johnston, D. Search for Related Content PubMed PubMed Citation Articles by Rosenkranz, J. A. Articles by Johnston, D. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB DOPAMINE

 Previous Article  |  Next Article 

Cellular/Molecular
State-Dependent Modulation of Amygdala Inputs by Dopamine-Induced Enhancement of Sodium Currents in Layer V Entorhinal Cortex

J. Amiel Rosenkranz and Daniel Johnston

Center for Learning and Memory, University of Texas at Austin, Austin, Texas 78712

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Interaction between the entorhinal cortex (EC) and basolateral amygdala (BLA) may be a fundamental component in the consolidation of many forms of affective memory, such as inhibitory avoidance. Dopamine (DA) in the EC is necessary for, and may facilitate, this form of learning. This effect of DA on affective behaviors may be accomplished in part through modulation of amygdala inputs. Although it is known that DA can modulate neuronal activity in the EC, it is not known whether DA modulates inputs from the BLA. In this study, we used in vitro patch-clamp recordings and Ca²⁺ imaging of layer V neurons in the rat lateral EC to determine whether DA modulates the integration of inputs from the BLA and the mechanism for this modulation. We found that DA exerted actions that depended on the neuronal state. Near resting membrane potentials, DA suppressed integration of inputs, whereas at depolarized potentials, DA enhanced integration. DA enhanced the integration by a D₂-mediated enhancement of Na⁺ currents, via phospholipase C. These experiments demonstrate that DA can exert actions in the EC that depend on the membrane voltage. This effect of DA may affect a wide range of inputs. Functionally, by enhancement of amygdala inputs that arrive in concert with other inputs, or during depolarized states, DA can facilitate the impact of affect on memory in a subset of conditions.

Key words: dopamine; dendrite; sodium channel; synaptic integration; entorhinal cortex; rat

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
The temporal lobe is composed of several structures, the interaction of which is necessary for many forms of learning and memory (Zola-Morgan et al., 1986; Squire et al., 2004), including the hippocampus, amygdala, and entorhinal cortex (EC). Abnormalities of the EC are seen in several disorders, such as Alzheimer's disease (Hyman et al., 1984; Hejl et al., 2004) and schizophrenia (Falkai et al., 1988; Arnold et al., 1991; Krimer et al., 1997; Prasad et al., 2004), and may provoke disruptions of memory and context-dependent behaviors (Otto et al., 1991; Baxter and Murray, 2001; Buckmaster et al., 2004). The rat EC receives a large input from the basolateral complex of the amygdala (BLA) (Krettet and Price, 1977; Pikkarainen et al., 1999), a region necessary for many forms of emotional learning. The interaction between the amygdala and the EC may be especially significant in the formation of memories that contain an affective component (Maren and Fanselow, 1997; Coutureau et al., 1999; Ferbinteanu et al., 1999; Ferry et al., 1999). For example, pharmacological suppression of EC activity interferes with consolidation of inhibitory avoidance (Ferreira et al., 1992; Quillfeldt et al., 1996), and neurons in the amygdala that project to the EC are activated during fear conditioning (Majak and Pitkanen, 2003).

Pathology of the dopamine (DA) system in the EC has been observed in postmortum examination of the brains of patients with Alzheimer's disease (Ryoo and Joyce, 1994) and schizophrenia (Goldsmith et al., 1997; Lahti et al., 1998; Akil et al., 2000). Pharmacological manipulations of the entorhinal DA system, and its second messengers, have a potent impact on several forms of emotional learning, including inhibitory avoidance and latent inhibition (Izquierdo et al., 1998; Barros et al., 2001; Roesler et al., 2002). However, it is not known whether DA modulates amygdala inputs to the EC. Given the behavioral effects of DA in the EC on emotional behaviors, it is hypothesized that DA will enhance the impact of amygdala inputs on EC neurons.

Layer V pyramidal neurons of the lateral EC have long apical dendrites that are potential targets for BLA inputs. DA may influence integration of synaptic inputs by modulation of dendritic properties. Layer V neurons contain mRNA for DA receptors (Weiner et al., 1991) and DA receptor proteins, and tyrosine hydroxylase (TH)-positive fibers are present in most layers of the EC (Diop et al., 1988; Huang et al., 1992; Goldsmith and Joyce, 1994; Tarazi et al., 1998, 1999). A previous study has examined the effects of DA on the excitability of layer V pyramidal neurons of the EC (Rosenkranz and Johnston, 2006). In addition, studies have demonstrated that DA can modulate single postsynaptic events in the EC (Pralong and Jones, 1993; Caruana et al., 2006). This study uses in vitro whole-cell recordings and calcium imaging to test how DA modifies amygdala inputs to the EC, the underlying ionic mechanisms, and whether a portion of this modulation occurs by effects on layer V apical dendrites.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Slice preparation. Male Sprague Dawley rats (age, 4–6 weeks) were anesthetized with an intraperitoneal injection of a combination of ketamine (42 mg/ml), xylazine (8 mg/ml), and, in some experiments, acepromazine (1.4 mg/ml). No differences were observed in the results from animals in which acepromazine was included in the anesthetic mixture. Once anesthetized, rats were perfused intracardially with an ice-cold solution containing (in mM) 110 choline chloride, 2.5 KCl, 1.25 NaH₂PO₄, 25 NaHCO₃, 0.5 CaCl₂, 7 MgCl₂, 7 dextrose, 1.3 ascorbic acid, and 3 pyruvic acid (all drugs were from Sigma, St. Louis, MO, unless specified otherwise). The solution was saturated with a 95% O₂/5% CO₂ gas mixture. After perfusion, the rat was quickly decapitated, and the brain was removed and sliced in 350 µm sections in the same solution (Vibratome Series 1000; Vibratome, St. Louis, MO). Brain slices were then incubated at 37°C for 10–15 min in a solution containing (in mM) 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 25 NaHCO₃, 2 CaCl₂, 2 MgCl₂, 10 dextrose, 1.3 ascorbic acid, and 3 pyruvic acid. After incubation, brain slices stabilized at room temperature for at least 50 min before transfer to the recording chamber.

Whole-cell recordings. Brain slices were transferred to the recording chamber perfused with a solution containing (in mM) 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 25 NaHCO₃, 2 CaCl₂, 2 MgCl₂, 10 dextrose, and 0.075 sodium metabisulfite (an antioxidant that prevents degradation of DA). Layer V of the lateral EC was visually identified using infrared differential interference contrast microscopy. Pyramidal neurons were selected based on the protrusion of a long apical dendrite and a nonstellate appearance. Patch electrodes were constructed from borosilicate glass with an outer diameter of 2.0 mm and an inner diameter of 1.16 mm (Sutter Instruments, Novato, CA). Electrodes were pulled using a Flaming/Brown micropipette puller (model P-97; Sutter Instruments). Electrodes were filled with (in mM) 120 potassium methylsulfate, 20 KCl, 0.2 EGTA, 10 HEPES, and 2 MgCl₂, with a pH of 7.3. On the day of recordings, 4 mM Na₂-ATP, 0.3 mM Tris-GTP, 7 mM phosphocreatine, and 0.1% neurobiotin were added to the internal recording solution (yielding an osmolarity of 290–295 mOsm). Whole-cell recordings were performed at 32–34°C. The electrode junction potential was corrected in the bath, but the diffusion potential was not compensated after seal formation and membrane rupture. Based on relative ionic mobilities and charge, this potential was calculated to be 7–8 mV. Therefore, the actual membrane potentials could be up to 8 mV more hyperpolarized than the measured voltage. The liquid junction potential between the intracellular and extracellular solutions was measured to be 3–4 mV using the method described by Neher (1992), suggesting that this difference between measured and actual membrane potentials may be even less. In some experiments, Sylguard was applied to the electrode tips, or the tips were wrapped with Parafilm, to minimize electrode capacitance. Before formation of a seal, the electrode tip potential was negated, and the open-tip resistance was measured (typically 4–6 M). After successful transition to whole-cell configuration, the neuron was given at least 5 min to stabilize before data were collected. Voltage signals were amplified and filtered in most experiments at 3 kHz (SEC-05L; NPI, Tamm, Germany). When action potential (AP) parameters were examined, voltage signals were filtered at no less than 5 kHz. Series resistance was monitored and compensated throughout the experiment using built-in bridge circuitry. Membrane potential was tracked throughout the experiment in the presence and absence of a holding current. Only neurons that had a resting membrane potential of at least –60 mV and APs that overshoot 0 mV were examined. To hold the membrane at various potentials, direct current was applied through the recording electrode. After completion of the experiment, brain slices were transferred to a solution containing 4% paraformaldehyde in 8% sucrose, where they remained for at least 24 h. Slices were stained for neurobiotin using the ABC Vectastin kit (Vector Laboratories, Burlingame, CA).

Calcium imaging. In some of the whole-cell experiments, 100 µM bis-Fura-2 was included in the recording pipette, and EGTA and neurobiotin were excluded. After whole-cell configuration was achieved, dye was permitted to diffuse and equilibrate throughout the cell for at least 15 min before experiments began. Bis-Fura-2 was excited with 380 nm light via a xenon lamp and a 380/13 excitation filter. Fluorescent emissions were filtered through a high-pass 495 nm filter and collected with a modified Photometrics CH250A cooled CCD camera (Roper Scientific, Duluth, GA) at a minimum of 40 Hz. Changes of intracellular calcium were quantified as F/F, where F was the background fluorescence of the cell before stimulation and F was the change of fluorescence during stimulation. F values were corrected by subtraction of autofluorescence, and F was corrected for bleaching during each exposure. A decrease in F/F indicates an increase in intracellular calcium. Calcium signals were inverted for quantification and display in figures. Electrical and optical signals were simultaneously acquired and synchronized (IgorPro; Wavemetrics, Lake Oswego, OR).

Cell-attached recordings. To directly measure Na⁺ currents, cell-attached recordings were obtained from the soma of layer V EC pyramidal neurons in the slice. Patch electrode shafts were coated with Sylguard. In these conditions, the extracellular solution was the same as above (in mM: 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 25 NaHCO₃, 2 CaCl₂, 2 MgCl₂, 10–15 dextrose, and 0.075 sodium metabisulfite). The solution inside the patch electrode contained (in mM) 110 NaCl, 10 HEPES, 2 CaCl₂, 2 MgCl₂, 1–3 3,4-diaminopyridine (3,4-DAP), 5 4-aminopyridine (4-AP), 30 tetraethylammonium (TEA)-Cl, and 10 dextrose. In some experiments, 1 µM TTX was also included in the pipette. Cell-attached recordings were performed at room temperature (23°C). Signals were collected and filtered at 5 kHz with an Axopatch 200B (Molecular Devices, Sunnyvale, CA). Electrode capacitance was compensated. Leak was arithmetically subtracted during analysis using scaled traces from smaller voltage steps that did not evoke a Na⁺ current. To examine activation of this current, the membrane patch was held 20 mV hyperpolarized to rest, and depolarizing steps were applied at 10 mV intervals. A liquid junction potential of –2.1 mV was calculated, based on the relative ionic mobilities of the internal and external solutions. Mean values used in activation curves were corrected by 2.1 mV. The actual membrane potential was determined at the end of the experiment by rupturing the patch and measuring V_rest in whole-cell configuration. DA, when used, was applied before the patch recording was obtained. These neurons that were exposed to DA were compared with unexposed control neurons. Because DA was shown to cause a small depolarization of the membrane potential of these neurons (Rosenkranz and Johnston, 2006), DA was not applied during the course of an individual recording to avoid the potentially difficult interpretation of apparent shifts in the activation curve that could be the result of an unquantified shift of membrane potential during the experiment.

Nucleated patch recordings. To examine the effects of DA on Na⁺ currents within the same neuron, without potential space-clamp errors or problems attributable to a change of membrane potential after application of DA, nucleated patch recordings were obtained from layer V pyramidal neurons at 32°C. After a whole-cell configuration was obtained, the pipette was slowly withdrawn while maintaining negative pressure at the tip of the pipette. The same intracellular solution was used for nucleated patch recordings and whole-cell current clamp recordings (above). To block K⁺ currents, 20 mM TEA-Cl, 3 mM 3,4-DAP, and 5 mM 4-AP were included in the bath, in addition to 0.2 mM CdCl₂ to block Ca⁺ currents. The amount of dextrose in the bath solution was decreased (2 mM) to maintain an osmolarity near 300 mOsm. A liquid junction potential of 8.1 mV was calculated based on relative ionic mobilities. All voltage values were corrected by 8.1 mV. Currents were activated with voltage steps from a holding potential of approximately –100 mV (–90 mV before liquid junction correction). Inactivation was examined by stepping the membrane potential to 0 mV, starting from –100 mV and progressing to more depolarized holding potentials in 10 mV steps.

Synaptic stimulation. Synaptic inputs to the recorded neuron were activated by discrete electrical stimulation using theta glass pipettes with a tip size between 2 and 4 µm [filled with (in mM) 125 NaCl, 2.5 KCl, 1.25 NaH₂PO₄, 10 HEPES, 2 CaCl₂, 2 MgCl₂, and 25 dextrose] or self-assembled bipolar tungsten electrodes that tapered to a fine tip (made of parts from A-M Systems, Carlsborg, WA). The electrodes were positioned 10–20 µm to the side of the apical dendrite, at a distance of 0–450 µm from the soma. Stimulations were performed as single pulses (0.2 ms duration; 0.005–0.05 mA) or as trains (five stimuli, 20 Hz, unless indicated otherwise). A horizontal slicing procedure has been used that preserves connections between the lateral nucleus of the amygdala (LAT) and EC, as demonstrated by von Bohlen und Halbach and Albrecht (2002). Amygdala inputs were activated by positioning the stimulating electrode in the LAT. Stimulation parameters were the same as above. The distance between the amygdala stimulation electrode and the entorhinal recording electrode ranged from 1.5 to 3.0 mm. Synaptic responses that did not have a monotonic rise, or had a variable latency, or had clear multiple events indicative of polysynaptic inputs were excluded from analysis.

Drug application. Chemicals were bath applied in the recording solution, unless noted otherwise, and included DA (10 µM), 6-chloro-2,3,4,5-tetrahydro-1-phenyl-1H-3-benzazepine hydrobromide (SKF81297; 10 µM), quinpirole (10 µM), 7-chloro-8-hydroxy-3-methyl-5-phenyl-2,3,4,5-tetrahydro-1H-3-benzazepine (SCH23390; 5 µM), sulpiride (5 µM), TTX (0.5–2 µM), 4-AP (3 mM), BaCl₂ (250 µM), NiCl₂ (50 µM), nimodipine (10 µM), and 4-ethylphenylamino-1,2-dimethyl-6-methylaminopyrimidinium chloride (ZD7288; 20 µM applied in bath solution for 10 min before patching). To examine signaling cascades, kinase inhibitors 2-[1-(3-dimethylaminopropyl)indol-3-yl]-3-(indol-3-yl) maleimide (GF109203X; 1 µM) and chelerythrine chloride (2 µM) were bath applied at least 15 min before patching a neuron. Alternatively, chelerythrine chloride (2 µM), 1-[6-[[(17ß)-3-methoxyestra-1,3,5(10)-trien-17-yl]amino]hexyl]-1H-pyrrole-2,5-dione [U73122; 3 µM in DMSO; phospholipase C (PLC) inhibitor], forskolin (5 µM or 25 µM in DMSO; adenylyl cyclase activator), cyclosporin A [1 µM; protein phosphatase 2B (PP2B; calcineurin) inhibitor], or okadaic acid (100 nM in DMSO; PP2A/PP1 inhibitor) was applied alone during the recording, 15 min before quinpirole. The kinase inhibitor (9R,10S,12S)-2,3,9,10,11,12-hexahydro-10-hydroxy-9-methyl-1-oxo-9,12-epoxy-1H-diindolo[1,2,3-fg:3',2',1'kl]pyrrolo[3,4-i][1,6]benzodiazocine-10-carboxylic acid, hexyl ester (KT5720; 1 µM) and the PP2B inhibitor calcineurin autoinhibitory peptide (100 µM) were included in the recording pipette. When BAPTA was included in the pipette, the following internal solution was used (in mM): 20 BAPTA-tetrapotassium, 40 KMeSO4, and 10 sucrose (to adjust solution to 290–295 mOsm). The remainder of the solution was the same as above. Where indicated, synaptic transmission was blocked by the inclusion of CNQX (10 µM), DL-AP-5 (50 µM), bicuculline (10 µM; in DMSO), and (2S)-3-[[(1S)-1-(3,4-dichlorophenyl)ethyl]amino-2-hydroxypropyl](phenylmethyl)phosphinic acid (CGP55845) (2 µM; in DMSO) in the bath solution. All other experiments were performed with DL-AP-5 (50 µM), bicuculline (10 µM; in DMSO), and CGP55845 (2 µM; in DMSO) in the extracellular recording solution, but only 1 µM CNQX to allow examination of the evoked AMPA/kainate EPSP without contamination by epileptiform activity that could be evoked when GABA receptors were blocked. Higher concentrations of DL-APV did not exert additional actions, indicating that the NMDA component of the EPSP was blocked by 50 µM DL-APV. The total concentration of DMSO in the bath remained below 0.15%.

In a subset of experiments, drugs were applied by puff application through a small-tipped glass pipette (1–2 µm, similar to recording electrodes). Drugs were added to a puffing solution of (in nM) 125 NaCl, 10 HEPES, 2.5 KCl, 1.25 NaH₂PO₄, 2 MgCl₂, 2 CaCl₂, 10 dextrose, and 0.075 sodium metabisulfite, pH 7.3. The drugs that were puff applied included 1–2 µM TTX and 50 µM DA. Puffing pipettes were placed within 20 µm of the apical dendritic trunk or the soma and were pulled back when not in use. The spatial specificity of dendritic puff application was verified by puffing TTX (1–2 µM) to the dendritic region. In contrast to somatic application, dendritic puff application had no effect on somatic AP amplitude or threshold using these methods, indicative of a spatially limited action of puff application.

Data collection and analysis. Signals were collected and digitized at 20–50 kHz via an ITC-18 interface board (Instrutech Corporation, Port Washington, NY), transmitting to an Apple Computer (Cupertino, CA) PowerPC G5 running Igor Pro software (WaveMetrics).

The AP rise time was quantified as the peak of the first derivative (dV/dT) of the AP waveform. The AP amplitude was defined as the difference between the AP peak voltage and the initial membrane potential (–70 mV in these experiments). AP firing adaptation was quantified as the ratio of the first interspike interval to the last interspike interval within a 700 ms current injection. A current amplitude that evoked four to five APs was used for analysis. PSP amplitude was measured as the peak voltage deflection in response to synaptic stimulation. PSP summation was quantified as the ratio of the amplitude of the last PSP to the first PSP. The PSP rise time was quantified as the time between 10 and 90% of the PSP amplitude. The decay time constant of PSPs was fit with a least-squares best-fit double-exponential function. Paired-pulse facilitation was quantified as the amplitude of the second PSP divided by the amplitude of the first PSP. In some experiments, a current shaped similar to a synaptic current was injected into the neuron [PSP, current shape described by A(t/)e^(–), where A is the amplitude of the injected current, t is time, and 1/ is the time to peak]. The PSPs evoked by this current injection were quantified and analyzed in the same manner as synaptic PSPs. For determination of V_1/2 of activation and inactivation of Na⁺ currents in nucleated patches, a sigmoidal curve was fit to the data from each neuron. A V_1/2 value was determined from this fit for each neuron before and after DA treatment.

All planned comparisons were made with paired t tests, with an level of 0.05 considered significant. Additional post hoc comparisons were made with Student's t tests. Bonferroni corrections of p values were used for multiple comparisons. In all plots, markers and error bars indicate the mean value ± SEM.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Pyramidal neurons in layer V of the lateral EC were selected based on soma size and shape and the presence of a prominent apical dendrite directed toward the pial surface. The lateral EC was targeted because it receives the heaviest projections from the LAT (Pikkarainen et al., 1999). A total of 469 neurons were included in the analysis. The neurons recorded within layer V of the lateral EC were located between the lamina dissecans, which was apparent because of the density of fibers and layer VI, which had neurons that appeared smaller and closer to the external capsule. When recorded in current clamp, these neurons displayed electrophysiological characteristics consistent with previously described entorhinal cortical pyramidal neurons (Hamam et al., 2002). In current clamp, the mean resting membrane potential was –71.3 ± 0.8 mV, and input resistance was 87.5 ± 3.3 M. These neurons displayed a range of regular or adapting AP firing patterns (adaptation ratio ranging from 0.77 to 0.21), a baseline AP amplitude of 106.1 ± 0.7 mV, and a half-width of 1.14 ± 0.02 ms. Furthermore, when filled with neurobiotin (Fig. 1A) or the calcium-sensitive dye bis-Fura 2 (below), the neurons displayed pyramidal neuron morphology.

View larger version (40K): [in this window] [in a new window]   Figure 1. Characteristics of EPSPs. A, Example of a neurobiotin-filled layer V pyramidal neuron recorded in lateral EC. B, Stimulation of the LAT or local stimulation close to the apical dendrite of layer V EC neurons evokes an EPSP in the presence of blockers of NMDA, GABAA, and GABAB receptors. The remaining component of this EPSP is entirely blocked by 10 µM CNQX, an AMPA/kainate receptor blocker. For maximal clarity, in all panels, examples of LAT or local stimulation are shown, unless indicated otherwise. C, Traces show an EPSP evoked by stimulating electrodes 40 and 250 µm from the soma and an EPSP evoked by LAT stimulation (stim; average of 5 traces). The rise time of EPSPs is a rough index of the distance of a synapse from the recording electrode at the soma. The rise time of the locally evoked EPSPs is plotted as a function of distance of the stimulation electrode from the soma. By plotting the rise times of EPSPs evoked by LAT stimulation along the regression constructed from locally evoked EPSPs, one can derive a rough estimate of the distance of the LAT inputs from the soma. The histogram plots the number of LAT-evoked events as a function of their extrapolated input distance. The majority of LAT inputs display rise times that correspond to EPSPs evoked by local stimulation at 150–200 µm from the soma. D, The decay time of EPSPs was strongly voltage dependent, with a slower decay time constant observed at depolarized membrane potentials (gray). Three overlayed traces at each voltage are displayed. E, A consequence of this slower decay time was increased summation of EPSPs at depolarized membrane potentials (traces displayed are averages of 5 sweeps). Membrane potentials of –70 and –55 mV were chosen for statistical comparison. *p < 0.05, significant difference in a paired t test comparison between the two membrane potentials.

 
Voltage dependence of EPSP shape and summation
Synaptic inputs were activated by electrical stimulation in the LAT or 10–20 µm from the apical dendrite. Although study of the modulation of inputs from the amygdala to the EC was the main goal, locally evoked EPSPs were examined in addition to LAT-evoked EPSPs to facilitate Ca-imaging experiments of evoked EPSPs (below). Furthermore, it allows more direct comparisons with previous studies. In the presence of blockers for GABA and NMDA receptors (see Materials and Methods), an EPSP was evoked, the amplitude of which depended on the intensity of the electrical stimulation. An intensity of electrical stimulation was chosen that evoked a somatic EPSP of 2–4 mV (mean, 2.8 ± 0.2 mV; n = 39). With the addition of 10 µM CNQX, the remaining component of the EPSP was entirely blocked (n = 8) (Fig. 1B). When the LAT was stimulated, the measured rise time of the evoked EPSP was 10.2 ± 2.9 ms (n = 17). For initial experiments that used local stimulation, the distance of the stimulation electrode from the soma varied from 20 to 400 µm along the apical dendrite. The rise time of these evoked EPSPs was measured and plotted as a function of the distance of the stimulation electrode from the soma. The measured rise time correlated with the distance of the stimulation electrode from the soma (Fig. 1C) (r² = 0.76; slope of change, 1 ms/36 µm). In theory, from this slope and the mean measured rise times of LAT-evoked EPSPs (10.2 ms), one can extrapolate the approximate distance from the soma of synapses of LAT inputs. The average mean rise time of LAT inputs is similar to synaptic inputs activated by a stimulation electrode 175 µm from the soma. Therefore, for the remainder of the experiments, the stimulation electrode was placed along the dendrite at 150–200 µm from the soma, unless noted otherwise.

To further characterize the LAT-evoked and locally evoked EPSPs, the decay time was measured. It was found that the EPSP decay time strongly depended on the membrane voltage (Fig. 1D). For comparison, a voltage near V_rest (–70 mV) was chosen and compared with a depolarized subthreshold membrane voltage (–55 mV) that is close to the depolarized "up" state in EC neurons observed during in vivo intracellular recordings (J. A. Rosenkranz and A. A. Grace, unpublished observation). The mean decay time of LAT-evoked EPSPs near V_rest (–70 mV) was 59.5 ± 3.6 ms, whereas the decay time at the depolarized membrane potential (–55 mV) was 77.3 ± 3.1 ms (n = 17; t = 3.75; p < 0.001, paired t test). A similar difference was observed with EPSPs evoked by local stimulation (near V_rest: –70 mV, 56.4 ± 4.0 ms; depolarized potential: –55 mV, 74.5 ± 3.8 ms). One consequence of this voltage-dependent difference in EPSP decay time was that when a train of five EPSPs was evoked by stimulation at 50 ms intervals, there was greater EPSP summation at the depolarized membrane potential (Fig. 1E) (LAT evoked: –70 mV, 1.73 ± 0.26, –55 mV, 2.17 ± 0.25 mV, n = 17, t = 2.77, p < 0.05, paired t test; locally evoked: –70 mV, 1.61 ± 0.19, –55 mV, 2.13 ± 0.13, n = 22, t = 2.85, p < 0.05, paired t test). For the remainder of this study, summation of PSPs in this manner was used as an assay for temporal integration.

DAergic modulation of EPSPs
To examine modulation of EPSP integration, 10 µM DA was bath applied for 2–3 min. This application of DA did not have a significant effect on the peak amplitude of evoked EPSPs near V_rest (Fig. 2A) [–70 mV; LAT-evoked EPSP: amplitude baseline, 2.7 ± 0.2 mV; post-DA, 2.6 ± 0.3 mV; n = 17; t = 1.87; not significant (n.s.); paired t test; local-evoked EPSP: amplitude baseline, 2.9 ± 0.3 mV; post-DA, 2.3 ± 0.5 mV; n = 22; t = 1.81; n.s.; paired t test] or at a depolarized membrane potential (–55 mV; LAT-evoked EPSP: amplitude baseline, 2.4 ± 0.4 mV; post-DA, 2.4 ± 0.4 mV; n = 17; t = 1.64; n.s.; paired t test; local-evoked EPSP: amplitude baseline, 2.6 ± 0.3 mV; post-DA, 2.7 ± 0.4 mV; n = 22; t = 1.72; n.s.; paired t test). To obtain preliminary indication of a presynaptic or postsynaptic locus of action, paired-pulse facilitation of EPSCs was examined at a range of interstimulus intervals (5–140 ms). There was no significant change of the paired-pulse ratio at any interval (Fig. 2B).

View larger version (29K): [in this window] [in a new window]   Figure 2. DA has minimal effects on EPSP amplitude. A, Application of DA (10 µM for 2–3 min) did not have a significant effect on the amplitude of a single LAT- or local-evoked EPSP, regardless of the membrane potential. Five traces in baseline and five after DA are overlayed and displayed for each panel. Black traces are baseline, and gray traces are after DA. B, Paired-pulse facilitation of EPSCs did not change after application of DA, examined over a range of frequencies. Displayed are overlays of averaged traces of paired stimulations at varying intervals. C, Application of DA resulted in a slower decay time of EPSPs (see A). This was observed specifically at depolarized membrane potentials. *p < 0.05, significant difference compared with baseline conditions (paired t test).

 
When briefly examined in a previous study, and replicated here, we found that application of DA can suppress the summation of a train of EPSPs. However, when examined in further detail, there was a membrane state dependency for this effect of DA, and DA only suppressed the summation of EPSPs near V_rest (Fig. 3A). Thus, despite no significant effect on the amplitude of a single EPSP, at depolarized membrane potentials DA enhanced the summation of a train of EPSPs (Fig. 3A). As demonstrated previously (Rosenkranz and Johnston, 2006), the effect of DA on EPSP summation near V_rest was blocked by ZD7288 (10 µM, intracellular; n = 6), without blocking the effect of DA at depolarized membrane potentials (Fig. 3B). Below we will examine the mechanism of this DAergic facilitation of EPSP integration at depolarized membrane potentials, which contributes to the state dependency of the effects of DA. To examine this, trains of 5 EPSPs were used below instead of 10 EPSPs for two reasons: (1) a shorter train of EPSPs can minimize the potential long-term effects of repeated synaptic stimulation, and (2) using a shorter train of EPSPs allows further isolation of the DAergic effect to be examined here, because the effect of DA on EPSP integration at V_rest is weaker when a shorter train of EPSPs is examined (Fig. 3). This is consistent with DAergic modulation of h-channels near V_rest, which is expected to have a temporally constrained impact on EPSP summation because of its slower kinetics.

View larger version (45K): [in this window] [in a new window]   Figure 3. DA modulates integration of EPSPs. A, Overlayed traces of a train of 10 EPSPs demonstrate that DA (gray traces) enhances summation at a depolarized membrane potential but decreases summation near the resting membrane potential. The right panel plots the time course of the effects of DA in two neurons (black circles are summation of 10 EPSPs; gray circles are summation of 5 EPSPs). The time course of the effects of DA in this example indicate that it began shortly after application of DA (at solid bar) and returned close to baseline. B, As demonstrated previously, the effects of DA on summation near Vrest are blocked by ZD7288 but not the effects of DA at a depolarized membrane potential. C, DAergic modulation of the summation of five EPSPs also displayed a voltage dependence and allows more isolated examination of processes involved in DAergic enhancement of integration at depolarized membrane potentials because effects on summation of EPSPs near Vrest are minimal. *p < 0.05, significant difference in a paired t test comparison between baseline and post-DA conditions.

 
When using a train of five EPSPs, DA significantly increased the summation of LAT-evoked and locally evoked EPSPs at a depolarized V_m (Fig. 3A,C) (–55 mV; LAT-evoked EPSP: summation baseline, 2.17 ± 0.25; post-DA, 2.59 ± 0.19; n = 17; t = 2.67; p < 0.05, paired t test; local-evoked EPSP: summation baseline, 2.1 ± 0.13; post-DA, 2.47 ± 0.18; n = 22; t = 2.52; p < 0.05, paired t test). This effect was not observed near V_rest when five EPSPs are evoked (–70 mV; LAT-evoked EPSP: summation baseline, 1.73 ± 0.26; post-DA, 1.54 ± 0.28; n = 17; t = –1.94; n.s.; paired t test; local-evoked EPSP: summation baseline, 1.61 ± 0.19; post-DA, 1.55 ± 0.19; n = 22; t = –1.83; n.s.; paired t test). The effect of DA on EPSP summation was at least partly reversible after extended washout in all cases examined (Fig. 3A) (after 60 min washout, EPSP summation was measured at –55 mV; LAT-evoked EPSP: summation, 2.21 ± 0.30; n = 11; t = 1.51; n.s.; paired t test with baseline; local-evoked EPSP: summation, 2.19 ± 0.31; n = 11; t = 1.44; n.s.; paired t test with baseline).

To further test whether postsynaptic factors are sufficient to account for the effects of DA, synaptic inputs were blocked, and EPSC-shaped currents were injected to the soma. The voltage responses evoked (PSPs) by these currents mimicked synaptic inputs. The EPSC-shaped current amplitude and decay time were modified until they approached the kinetics of EPSPs evoked by electrical stimulation 150–200 µm from the soma (mean amplitude, 2.9 ± 0.7 mV; n = 19) including the voltage dependence of the PSP decay time (mean decay time constant: –70 mV, 57.5 ± 3.6 ms; –55 mV, 72.7 ± 3.1 ms; n = 19; t = 3.13; p < 0.001, paired t test). Similar to synaptic EPSPs, depolarization enhanced summation of a train of five EPSPs (supplemental Fig S1A, available at www.jneurosci.org as supplemental material) (summation –70 mV, 1.63 ± 0.26; summation –55 mV, 2.0 ± 0.27; n = 19; t = 2.98; p < 0.01, paired t test).

To test whether two separate mechanisms that use different temporal kinetics and voltage parameters contribute to the effects of DA at depolarized and resting membrane potentials, PSPs were injected at different frequencies and different numbers, and the membrane potential was shifted between two different membrane potentials, –70 and –55 mV. The effects of DA on PSPs summation near V_rest were minimal throughout a range of frequencies when five PSPs were examined (supplemental Fig S1A, available at www.jneurosci.org as supplemental material). When longer trains of PSPs were used, DA caused a suppression of PSP integration near V_rest. However, at depolarized membrane potentials, DA greatly enhanced summation of five PSPs at a variety of frequencies. Longer trains of PSPs (n = 10) facilitated the actions of DA when the membrane potential was depolarized to –55 mV. The voltage dependence of the direction of the effects of DA indicates that DAergic modulation of voltage-dependent ion channels may contribute to this voltage state dependency. The dependence on the number of PSPs (functionally, the duration of the input) near V_rest, but not at depolarized membrane potentials, implies that DA modulates a faster conductance at depolarized membrane potentials and a slower conductance near V_rest. What emerges is a divergent DAergic modulation of PSP integration that depends on the membrane potential.

To gain initial insight into the mechanism of the effect of DA on EPSP summation, the EPSP decay time was examined. DA significantly prolonged the EPSP decay time, specifically at depolarized membrane potentials (Fig. 2A,C) (–55 mV; LAT-evoked EPSP decay time: baseline, 77.3 ± 3.1 ms; post-DA, 88.6 ± 3.9 ms; n = 17; t = 2.57; p < 0.05, paired t test; local-evoked EPSP decay time: baseline, 74.5 ± 3.8 ms; post-DA, 86.2 ± 4.0 ms; n = 22; t = 2.71; p < 0.05, paired t test). There was no significant effect near V_rest (–70 mV; LAT-evoked EPSP decay time: baseline, 59.5 ± 3.6 ms; post-DA, 53.8 ± 3.0 ms; n = 17; t = –1.99; n.s.; paired t test; local-evoked EPSP decay time: baseline, 56.4 ± 4.0 ms; post-DA, 52.7 ± 3.9 ms; n = 22; t = –2.08; n.s.; paired t test). This prolongation of the EPSP can increase the time window for EPSP integration and may account for the facilitation of EPSP summation.

To determine what DA receptor subtypes were involved in the actions of DA on summation of LAT-evoked or locally evoked EPSPs, specific agonists and antagonists were applied. The DA D₂ agonist quinpirole (10 µM) mimicked the effects of DA on EPSP summation at depolarized membrane potentials (Table 1). The DA D₁ agonist SKF81297 (10 µM) exerted the opposite effects (Table 1). Furthermore, the actions of DA on EPSP decay and summation were blocked by sulpiride (5 µM) but not SCH23390 (10 µM) (Table 1).

View this table: [in this window] [in a new window]   Table 1. Effects of DA receptor subtype on EPSP decay time and summation

 
DAergic modulation of postsynaptic channels
To verify the postsynaptic locus of this action of DA and to probe the mechanism involved, two approaches were used: (1) a blocker of ion channels [N-(2,6-dimethylphenylcarbamoylmethyl)-triethylammonium bromide (QX-314)] was included in the intracellular solution, or (2) all synaptic inputs were blocked, and the mechanism of DAergic modulation of PSPs were examined. Intracellular QX-314 (3 mM) suppressed the enhancement of EPSP summation observed at depolarized membrane potentials (supplemental Fig S2A, available at www.jneurosci.org as supplemental material) (EPSP summation: –70 mV, 2.17 ± 0.16; –55 mV, 1.77 ± 0.21; n = 7; t = –2.60; p < 0.05, paired t test; LAT-evoked EPSPs were examined). In addition, intracellular QX-314 blocked the DAergic enhancement of EPSP summation at depolarized membrane potentials (supplemental Fig S2B, available at www.jneurosci.org as supplemental material) (EPSP summation at –55 mV: baseline, 1.77 ± 0.21; post-DA, 1.68 ± 0.20; n = 7; t = -1.52; n.s.; paired t test; EPSP summation at –70 mV: baseline, 2.17 ± 1.6; post-DA, 2.04 ± 0.28; n = 7; t = 1.77; n.s.; paired t test). Blockade of the effects of DA with a postsynaptic manipulation is consistent with a postsynaptic action of DA.

To further examine postsynaptic mechanisms, PSPs were examined in more detail. An additional advantage of PSPs is that DAergic modulation of postsynaptic glutamate receptors can be excluded. In accord with synaptic EPSPs, DA prolonged the PSP decay time constant only at depolarized membrane potentials (Fig. 4A) (–55 mV decay time: baseline, 72.7 ± 3.1 ms; post-DA, 84.3 ± 3.1 ms; n = 19; t = 2.21; p < 0.05, paired t test) but not near V_rest (–70 mV decay time: baseline, 57.5 ± 3.3 ms; post-DA, 52.6 ± 3.1 ms; n = 19; t = –1.81; n.s.; paired t test). Furthermore, as noted above, DA enhanced the summation of five PSPs at depolarized membrane potentials (Fig. 4B) (PSP summation at –55 mV: baseline, 2.0 ± 0.27; post-DA, 2.8 ± 0.23; n = 19; t = 2.53; p < 0.05, paired t test) but not near V_rest (PSP summation at –70 mV: baseline, 1.63 ± 0.26; post-DA, 1.50 ± 0.30; n = 19; t = 1.64; n.s.; paired t test). This indicates that the DAergic modulation of PSP time constant and summation can be accounted for with postsynaptic mechanisms. These effects of DA on PSP were mimicked by the DA D₂ agonist quinpirole (supplemental Fig S3A, available at www.jneurosci.org as supplemental material) and not mimicked by the DA D₁ agonist SKF81297 (Table 2).

View this table: [in this window] [in a new window]   Table 2. Effects of DA receptor subtype on PSP decay time and summation

 

View larger version (22K): [in this window] [in a new window]   Figure 4. DAergic modulation of postsynaptic PSPs is blocked by TTX. All synaptic inputs were blocked, and an EPSC-shaped current was injected. The resulting PSP displayed similarities to EPSPs. A, There was a slower PSP decay time at depolarized membrane potentials. DA (gray) caused an additional prolongation of the PSP decay. B, Also similar to EPSPs, there was greater PSP summation at depolarized membrane potentials, and DA further enhanced summation at depolarized membrane potentials. C, D, Bath application of TTX (1 µM; gray) blocked the prolongation of the EPSP decay that is observed at depolarized membrane potentials in baseline conditions (C) and blocked the enhanced summation of PSP observed at depolarized membrane potentials in baseline conditions (D). E, Preapplication of TTX blocked the effect of DA on EPSP decay time and summation at depolarized membrane potentials. All PSP traces are averages of five sweeps. *p < 0.05, significant difference in a paired t test comparison between baseline and drug conditions.

 
To gain insight into the second-messenger cascade altered by DA D₂ receptor activation, the effects of quinpirole (10 µM) on PSP summation were examined after preincubation with protein kinase inhibitors. The specific PKA blocker KT5720 (1 µM, intracellular) diminished the effects of quinpirole on PSP summation (Table 3 and supplemental Fig S3B, available at www.jneurosci.org as supplemental material) but did not completely block these effects. However, the protein kinase C (PKC) inhibitor chelerythrine (2 µM) caused two effects: the summation at depolarized membrane potentials was much greater when PKC was blocked, and the effects of quinpirole were reversed (Table 3 and supplemental Fig S3C, available at www.jneurosci.org as supplemental material). This implies that DA D₂ receptors may exert their actions on PSP summation by reduction in PKC activity, or by actions that oppose PKC.

View this table: [in this window] [in a new window]   Table 3. Effect of the DA D2 receptor agonist quinpirole of PSP summation when protein kinases are inhibited

 
To verify that PKC may be exerting action by modulation of Na⁺ channels, TTX (0.5 µM) was applied before chelerythrine (2 µM). TTX blocked the effects of chelerythrine. However, as seen above, TTX diminished PSP summation at depolarized membrane potentials. To compensate for this, the size of the PSP was increased, and its decay prolonged to enhance the maximal potential reached by the summating PSPs. Even when the PSPs were increased in this manner, TTX still blocked the effect of chelerythrine (Table 3).

To further examine the potential role of a PKC-dependent path in the effects of D₂ receptor activation, other inhibitors were also applied. The PLC inhibitor U73122 [GenBank] (3 µM) and internal BAPTA (20 mM) also blocked the effect of quinpirole (Table 3 and supplemental Fig S3C, available at www.jneurosci.org as supplemental material). A combination of nickel and nimodipine (50 µM and 10 µM) did not block the effects of DA (below), indicating that internal calcium stores are likely involved in the actions of D₂ receptor activation. The PP2B inhibitors cyclosporin A (1 µM) (supplemental Fig S3C, available at www.jneurosci.org as supplemental material) and calcineurin autoinhibitory peptide (100 µM, intracellular) also both blocked the effects of quinpirole (Table 3), whereas the PP2A/PP1 inhibitor okadaic acid (100 nM) diminished but did not block the effect of quinpirole (Table 3). This implies that D₂ receptor activation likely modulates PP2B via PLC and internal Ca²⁺ stores and thereby induces an action on PSP summation opposite to PKC activation, similar to the effects of PKC inhibition. To further exclude a role for PKA in this effect of DA, modulators of adenylyl cyclase were examined. SQ22536 (100 µM, intracellular) did not block the effect of quinpirole (Table 3 and supplemental Fig S3B, available at www.jneurosci.org as supplemental material), whereas forskolin (5 µM) decreased PSP summation and higher doses of forskolin (25 µM) did not block the effect of quinpirole on PSP summation (Table 3 and supplemental Fig S3B, available at www.jneurosci.org as supplemental material).

However, it was unclear why quinpirole reduced PSP summation when PKC was blocked with chelerythrine (above). When PKC activity was blocked with GF109203X at a concentration that also reduces PKA activity (10 µM), the effects of quinpirole on PSP summation are completely blocked (Table 3). Blockade of PKC may unmask PKA-dependent effects, or PKC can bias the coupling of DA receptors between G-proteins that may be preferentially linked to PLC (such as Gq) or adenylyl cyclase (i.e., Gs and Gi/o).

To verify that DA uses a similar signaling cascade in enhancement of PSP summation, its effects (10 µM DA) were examined in the presence of blockers of PLC or PP2B (Table 4). Both U73122 [GenBank] (3 µM) and cyclosporin A (1 µM) blocked the DAergic enhancement of PSP summation.

View this table: [in this window] [in a new window]   Table 4. The enhanced summation of PSPs after DA is blocked by inhibitors of PLC and PP2B

 
We observed above that intracellular application of the ion channel blocker QX-314 blocked the effects of DA on EPSP summation. To examine what ion channels may underlie this postsynaptic effect of DA, blockers of ion channels were applied. QX-314 blocks several ion channels, including Na⁺ channels. The effect of a more specific blocker of Na⁺ channels, TTX, on PSPs was examined. Application of TTX (1 µM) abolished the voltage-dependent enhancement of PSP summation (Fig. 4D, Table 5). Thus, whereas depolarization usually enhanced PSP summation, in the presence of TTX, summation was reduced at depolarized membrane potentials (PSP summation at –55 mV: baseline, 2.35 ± 0.15; TTX, 0.76 ± 0.07; n = 9; t = –4.07; p < 0.01, paired t test). TTX had minimal effects near V_rest (PSP summation at –70 mV: baseline, 1.37 ± 0.14; TTX, 1.33 ± 0.20; n = 9; t = 1.77; n.s.; paired t test). Furthermore, TTX blocked the effect of DA on PSP summation (Fig. 4E) (–55 mV: TTX, 0.76 ± 0.07; TTX plus DA, 0.80 ± 0.20; n = 9; t = 1.85; n.s.; paired t test). Similar to baseline conditions, there was a minimal effect of DA on PSP summation near V_rest in the absence (above) or presence of TTX (PSP summation at –70 mV: TTX, 1.33 ± 0.20; TTX plus DA, 1.10 ± 0.20; n = 9; t = 2.16; n.s.; paired t test). Application of 4-AP (3 mM), while greatly enhancing PSP summation, did not block the DAergic enhancement of PSP summation at depolarized membrane potentials (Table 5). Neither did application of blockers of several other voltage-dependent ion channels (Table 5), such as BaCl₂ (250 µM), ZD7288 (20 µM), and NiCl₂ (10 µM) with nimodipine (50 µM). These data are consistent with a postsynaptic DAergic enhancement of signal integration via modulation of Na⁺ channel activity.

View this table: [in this window] [in a new window]   Table 5. The effects of DA are diminished by Na+ channel blockers but not by blockers of other voltage-gated ion channels

 
If Na⁺ channels, in fact, underlie the actions of DA, one may expect modification of other indices of Na⁺ channel activation, such as AP parameters. Consistent with enhancement of Na⁺ channel activity, DA increased the amplitude and rise time of antidromic APs (supplemental Fig S4, available at www.jneurosci.org as supplemental material) (AP amplitude: baseline, 101.3 ± 2.7 mV; post-DA, 107.1 ± 2.5 mV; n = 14; t = 2.26; p < 0.05, paired t test; AP rise time: baseline, 167.2 ± 13.8 mV/ms; post-DA, 175.0 ± 13.2 mV/ms; n = 14; t = 2.19; p < 0.05, paired t test).

Modulation of Na⁺ channels can be more directly examined by measurement of Na⁺ currents. Nucleated patch recordings were used to measure the activation, inactivation, and peak amplitude of the Na⁺ current (see Materials and Methods). TTX (1 µM) entirely blocked this current in a separate group of neurons (n = 5) and in a portion of the neurons examined below after application of DA (n = 4). Application of DA induced a small leftward shift in the activation of Na⁺ currents (Fig. 5B) (baseline V_1/2, –38.7 ± 0.6 mV; post-DA V_1/2, –42.3 ± 0.7 mV; n = 11; t = 2.74; p < 0.05, paired t test) without a significant shift in the inactivation (baseline V_1/2, –70.3 ± 0.7 mV; post-DA V_1/2, –69.3 ± 0.7 mV; n = 6; t = 1.69; n.s.; paired t test). There was a small increase in the peak Na⁺ current (Fig. 5A) (baseline, 326.5 ± 29.7 pA; post-DA, 352.9 ± 30.15 pA; n = 11; t = 2.34; p < 0.05, paired t test). The DA D₂ agonist quinpirole (10 µM) also increased the peak Na⁺ current (Fig. 5C) (baseline peak Na⁺ current, 337.8 ± 32.5 pA; post-quinpirole, 372.1 ± 34.4 pA; n = 5; t = 2.72; p < 0.05, paired t test) and induced a leftward shift in the activation curve (baseline V_1/2, –39.5 ± 0.7 mV; post-quinpirole, –43.7 ± 0.7 mV; n = 5; t = 3.28; p < 0.08, paired t test). The DA D₁ agonist SKF81297 (10 µM) did not mimic the effect of DA but decreased the peak Na⁺ current (Fig. 5C) (baseline, 311.1 ± 32.8 pA; post-SKF81297, 282.5 ± 33.5 pA; n = 5; t = –2.95; p < 0.05) and had minimal effects on activation (baseline V_1/2, –38.8 ± 0.8 mV; post-SKF81297 V_1/2, –39.1 ± 0.8 mV; n = 5; t = 1.73; n.s.; paired t test).

View larger version (32K): [in this window] [in a new window]   Figure 5. DA modulates Na+ current measured from nucleated patches. A, Application of DA caused a small increase in the peak INa. B, DA (gray) caused a leftward shift in the voltage of activation of INa, without a significant change in the inactivation. C, The effects of DA were mimicked by a D2 agonist quinpirole (10 µM). A D1 agonist, SKF81297 (10 µM), exerted the opposite effect, decreasing Na+ currents. D, To test whether INa was modulated within the limited voltage ranges examined in current-clamp recordings (above), INa was evoked with a 10 mV step to –55 mV, grossly mimicking the depolarization evoked by summation of EPSPs at depolarized membrane potentials. The peak amplitude of this INa was greatly enhanced by DA. Overlayed are averaged traces of the current evoked by a 10 mV step to –55 mV (black) and to –80 mV (gray) for comparison. E, DA also prolonged the decay time of a Na+ current evoked by a voltage step from –100 mV to approximately –60 mV. All INa traces depicted are averages of 5–10 sweeps. *p < 0.05, significant difference in a paired t test comparison between baseline and DA conditions.

 
To examine whether DA modulates Na⁺ currents activated within the voltage range applicable to modulation of EPSP integration at depolarized membrane potentials, a Na⁺ current was activated with a 10 mV step to approximately –55 mV to grossly mimic a train of EPSPs evoked from depolarized membrane potentials. This current was greatly enhanced by DA (Fig. 5D) (baseline, 6.9 ± 1.2 pA; post-DA, 17.7 ± 2.5 pA; n = 7; t = 3.10; p < 0.05, paired t test).

To further examine modulation of Na⁺ channels, cell-attached recordings were performed. One advantage of this approach is that less cellular disruption occurs compared with nucleated patch recordings. Macropatch recordings of multiple Na⁺ channels were performed from control neurons and in other neurons within 5–10 min after application of DA (Fig. 6A). Na⁺ channels were activated with voltage steps in 10 mV increments from a holding potential 20 mV hyperpolarized to V_rest. The slope conductance was 18.0 ± 0.9 pS, measured with voltage steps from 20 mV hyperpolarized to between 80 and 110 mV depolarized from rest (approximately between membrane potentials of +10 and +40 mV from a V_initial of –90 mV). No current was observed when TTX (1 µM) was included in the recording pipette (n = 5). There was no significant difference in the peak Na⁺ current between control neurons and DA-exposed neurons (Figs. 6B, C) (control neurons: 5.2 ± 0.8 pA, n = 12; DA: 6.3 ± 1.2 pA, n = 13; t = 0.81; n.s.; unpaired t test). Similar to nucleated patch recordings, neurons exposed to DA displayed a more hyperpolarized activation curve (Fig. 6D) (control: V_1/2, –30.6 ± 0.9 mV; n = 12; DA: V_1/2, –36.0 ± 0.6 mV; n = 13; t = 5.07; p < 0.001, unpaired t test). In control neurons, Na⁺ channel activation was typically brief and lasted for a short duration after the onset of the voltage step between –50 to –60 mV. In neurons exposed to DA, the activity of the channels lasted longer (Fig. 6A). In some rare instances (1–2 of every 100 voltage steps), the channel openings appeared persistent throughout the duration of the voltage step. To quantify this, the current responses were averaged, and the decay of the current was fit with a single exponential. DA prolonged the decay time constant of the averaged current (Fig. 6B,C) (control, 2.46 ± 0.77 ms; range, 1.37–7.29 ms; n = 12; DA, 5.36 ± 0.67; range, 1.41–16.54 ms; n = 13; t = 2.85; p < 0.01, unpaired t test). Based on this finding, we also measured the decay time constants of Na⁺ currents activated by voltage steps from approximately –100 mV to approximately –60 mV in nucleated patches. Two time constants could be seen: a faster time constant, 1, of 3.69 ± 1.44 ms and a slower time constant, 2, of 18.01 ± 4.93 ms. Application of DA significantly prolonged the slower time constant (Fig. 5E) (1, 7.19 ± 2.01 ms; n = 11, t = 1.88; n.s.; paired t test with baseline 1; 2, 46.42 ± 11.29 ms; n = 11; t = 2.64; p < 0.05, paired t test with baseline 2).

View larger version (24K): [in this window] [in a new window]   Figure 6. DA increases the activity of Na+ channels. A, Cell-attached recordings of multiple Na+ channels activated by a 40 mV step from 20 mV hyperpolarized to rest. Neurons exposed to DA (right) displayed a greater tendency for longer-lasting channel activity than nonexposed control neurons (left). B, The decay time of the averaged Na+ current was fit with a single exponential (gray). Neurons that were exposed to DA (bottom) displayed averaged current that decayed slower compared with control neurons. C, The decay time of the current was prolonged in neurons exposed to DA, and there was a trend toward increased average peak amplitude. The asterisk indicates a significant difference in a t test of control neurons compared with a separate group of neurons exposed to DA. D, Similar to nucleated patch recordings of INa, there was a shift in the activation of Na+ channels in neurons exposed to DA (gray).

 
Dendritic participation in modulation
Excitatory inputs predominantly synapse on spines in the dendrites of cortical pyramidal neurons. Dopaminergic modulation in the dendrites may be a means to potently modify synaptic integration. To gain preliminary insight into whether DA modulates dendritic signal propagation, calcium imaging of the apical dendrite was performed. A train of five APs (20–50 Hz) was evoked using either antidromic stimulation or five brief (2–5 ms) suprathreshold current injections, each evoking a single AP. The change in fluorescence normalized by baseline fluorescence (F/F) was measured over the length of the dendrite (see Materials and Methods) (Fig. 7A). This train of APs evoked a submaximal calcium response; more APs, or higher-frequency stimulation, evoked a greater response. As observed in other regions, there was a fairly steep decrement in the amplitude of the calcium signal as a function of distance from the soma (Fig. 7B). Application of DA enhanced the peak calcium signal at all dendritic regions examined (Fig. 7A,B) (n = 21; F = 64.1; p < 0.001, repeated-measures ANOVA), including apical branches. Furthermore, the percentage change in the calcium signal after DA increased as a function of distance from the soma (Fig. 7B). This provides preliminary evidence that DA modulates signal propagation in the dendrites. If the effects of DA were limited to the soma, one would expect that the absolute amplitude of the changes induced by DA would decrement as a function of the amplitude of the signal (or in this case, the distance of the signal from the soma). However, this was not observed. The relatively greater effect of DA as a function of distance implies that this effect is unlikely to be entirely accounted by somatic changes. Similar to DAergic enhancement of EPSP summation, the effects of DA were mimicked by the DA D₂ agonist quinpirole (Fig. 7C) (n = 6; F = 158.6; p < 0.001, repeated-measures ANOVA). The DA D₁ agonist SKF81297 exerted the opposite action (Fig. 7) (n = 6; F = 61.0; p < 0.01, repeated-measures ANOVA).

View larger version (34K): [in this window] [in a new window]   Figure 7. DA increases backpropagation of signals into dendrites. A, Calcium imaging of the backpropagation of APs was used as an initial index of dendritic signal propagation. DA enhanced this signal along the apical dendrite. Line trace color corresponds to the color of the box at a location on the dendrite. B, DA increased the amplitude of the signal along the entire extent of the apical dendrite examined. The effectiveness of DA on dendritic signals (percentage change) was greater at dendritic sites further from the soma (right). C, Similar to the effects of DA on EPSPs summation and INa, the effects of DA on the dendritic signal was mimicked by the DA D2 agonist quinpirole (10 µM), whereas the opposite effect was seen after application of the DA D1 agonist SKF81297 (10 µM).

 
Direct dendritic electrophysiological recordings of the backpropagating AP (bAP; 200–300 µm from the soma) (Fig. 8A) indicate that the enhancement of the calcium signal after application of DA can be explained by enhancement of the amplitude of the bAP (Fig. 8B) (baseline, 46.6 ± 4.1 mV; post-DA, 58.3 ± 4.4 mV; n = 11; t = 2.73; p < 0.05, paired t test). The percentage change of the dendritic bAP was greater than the percentage change of the somatic AP observed above (7.7 vs 25.1%).

View larger version (19K): [in this window] [in a new window]   Figure 8. DA increases the amplitude of antidromic bAPs. A, The amplitude of bAPs recorded in the dendrite 150–300 µm from the soma was increased by DA (right; in this example the bAP was recorded 200 µm from the soma). Top traces are recordings from the soma, and bottom traces are recordings from the apical dendrite. B, Plots of group data indicate that a significant increase in the amplitude of antidromic APs was observed in the soma and dendrites after DA. Similar to the calcium signal, the percentage change of the dendritic AP was greater than the change in the somatic AP. *p < 0.05, significant difference with a paired t test comparison between baseline and DA conditions. C, Inputs of varying distances from the soma were activated by varying the distance of the stimulation electrode from the soma, along the apical dendrite. Summation of five EPSPs after DA was normalized to baseline summation (see Results). DA had a greater impact on summation of EPSPs that were evoked more distally. indicates summation.

 
Although these data indicate that DA may facilitate dendritic signal propagation, modulation of subthreshold events that initiate in the dendrites are more relevant to the experiments above. To gain preliminary indication of whether DAergic modulation of dendrites contributes to its effects on EPSP summation, summation of EPSPs was examined as a function of the distance of the local stimulation electrode from the soma. When the distance of the stimulation electrode along the apical dendrite was varied, it was found that the effect of DA was greatest when the stimulation electrode was farther from the soma (Fig. 8C) (r² = 0.71; p < 0.05; for the relationship between the stimulation electrode distance and the normalized effect of DA on summation at –55 mV; this was quantified as the amplitude of summation after DA divided by the amplitude of summation before DA). This is preliminary evidence that the dendrites are involved in the modulatory effect of DA on EPSP summation.

To more directly examine a role of dendrites in EPSP modulation, the calcium signal evoked by local synaptic stimulation was examined. As above, train stimulation (five stimuli at 20 or 50 Hz) evoked EPSPs that could be measured electrically at the soma. In some instances, a calcium signal could be detected at a dendritic site in the vicinity of the stimulation electrode when the intensity of the stimulus trains was increased to evoke larger EPSPs (mean amplitude, 7.1 ± 1.2 mV; n = 9). This calcium signal could only be detected within 10–15 µm of the peak calcium signal at the apical dendritic trunk (Fig. 9A). Application of DA had no significant effect on the peak calcium signal evoked by subthreshold synaptic stimulation but increased the spread of the calcium signal further from the site of the peak response (Fig. 9B) [measured as the width of the Gaussian fit to the data (2) times the SD of the peak; baseline, 3.58 ± 0.25; post-DA, 5.36 ± 0.33; n = 9; t = 2.57; p < 0.05, paired t test]. This indicates that DA may facilitate propagation of signals that originate in the dendrite without contribution from the soma.

View larger version (28K): [in this window]