Contribution of Dopamine D1/5 Receptor Modulation of Post-Spike/Burst Afterhyperpolarization to Enhance Neuronal Excitability of Layer V Pyramidal Neurons in Prepubertal Rat Prefrontal Cortex

Dopamine (DA) receptors in the prefrontal cortex (PFC) modulate both synaptic and intrinsic plasticity that may contribute to cognitive processing. However, the ionic basis underlying DA actions to enhance neuronal plasticity in PFC remains ill-defined. Using whole-cell patch-clamp recordings in layer V-VI pyramidal cells in prepubertal rat PFC, we showed that DA, via activation of D1/5, but not D2/3/4, receptors suppress a Ca2+-dependent, apamin-sensitive K+ channel that mediates post-spike/burst afterhyperpolarization (AHP) to enhance neuronal excitability of PFC neurons. This inhibition is not dependent on HCN channels. The D1/5 receptor activation also enhanced an afterdepolarizing potential (ADP) that follows the AHP. Additional single-spike analyses revealed that DA or D1/5 receptor activation suppressed the apamin-sensitive post-spike mAHP, further contributing to the increase in evoked spike firing to enhance the neuronal excitability. Taken together, the D1/5 receptor modulates intrinsic mechanisms that amplify a long depolarizing input to sustain spike firing outputs in pyramidal PFC neurons.


Introduction
The prefrontal cortex (PFC) is functionally engaged in organizing and processing cognitive information that are critical to the planning and execution of forthcoming behaviors [1]. Neural activities in the PFC that represent working memory processing can be modulated by mesocortical DA input, predominantly via activation of D1/5 receptors [2,3].
At the cellular levels, DA is known to modulate not only synaptic, but also intrinsic plasticity in the PFC to generate repetitive and persistent firing [4][5][6]. Intrinsic plasticity is integral to synaptic plasticity, and occurs through modulation of postsynaptic soma-dendritic ion channels that regulate neuronal excitability, thus allowing coupling of potentiated excitatory postsynaptic potentials (EPSPs) (from synaptic inputs) to reach spike firing (output) [7]. The resultant prolonged changes in spike firing are now termed long-term potentiation (or depression) of neuronal excitability (LTP/D-IE) [7].
During sustained repetitive firing of a PFC neuron, the firing of single spike or bursts of spikes is temporarily terminated by a postspike/burst afterhyperpolarization (AHP) that provides varying inter-spike or inter-burst inactive intervals. AHP is both intracellular and extracellular Ca 2+ -dependent [8][9][10] and consists of three subtypes: 1) fast AHP, (mediated by BK channel) [11], 2) apamin-sensitive medium AHP (mediated by a Ca 2+ -dependent K + ion conducting SK channel), and 3) slow AHP (mediated by an as yet unknown channel [12]. Recent pharmacological and knockout mice data suggest that both SK and Kv7/KCNQ/M channels can also mediate cortical mAHP [13,14].
Modulation of the AHP will alter neuronal refractoriness, interspike intervals, and consequently, spike timing, firing patterns and frequencies, which ultimately lead to changes in neuronal excitability [7,15,16]. An increase in post-burst slow AHP in hippocampal CA1 neurons is found in aged, cognitively-impaired, but not cognitively-unimpaired rodents. In contrast, in prepubertal animals, a predominant presence of the smaller AHP facilitates neuronal excitability increase following different forms of associative learning [17,18], presumably the smaller AHP is being modulated by learning-related release of neuromodulators, including dopamine.
Membrane hyperpolarization during AHP can, in turn, activate an h-current (I h ) that is mediated by the hyperpolarizationactivated and cyclic nucleotide-gated cationic HCN1 and/or HCN2 channels. Activated I h slowly depolarizes the cell back towards its equilibrium potential (V R = 230 mV). I h is functionally important for regulating neuronal excitability, resting potential and rhythmic pacemaker activity [19][20][21]. Thus, AHP and I h interact to shape the amplitude and duration of post-spike/burst AHP, as shown in hippocampal CA1, thalamic relay neurons and striatal cholinergic interneurons [22][23][24].
DA, through D1/5 receptor activation, enhances neuronal excitability of PFC neurons in pre-pubertal PFC neurons in vitro and mature PFC neurons in vivo (see Table 1 for a summary of the literature for this DA effects on PFC neuronal excitability). This suggests that the pre-pubertal PFC is already endowed with the functional D1/5 receptors [25] that can modulate intrinsic mechanisms that contributed to neuronal excitability enhancement. This DA increase in neuronal excitability may promote local interconnected network interaction and forms part of the mechanisms for a sustained increase in persistent firing necessary for cognitive processes [2,5].
The DA enhancement of neuronal excitability depends on an Ca 2+ -dependent intracellular signaling [4,26,27], which may in turn modulate the i[Ca 2+ ]-dependent post-spike/burst AHP. In the present study, we characterized the modulatory effects of DA on post-spike/burst AHP in layer V/VI pyramidal cells of prepubertal rat PFC, and the contribution of modulated AHP to the overall neuronal excitability changes. We found that the D1/5 receptor suppressed post-spike/burst mAHP and enhanced a post-AHP afterdepolarizing potential (ADP), all independent of I h activation, to contribute to the increase in neuronal excitability of layer V-VI PFC neurons in prepubertal rats. The findings also enable us to understand DA actions in the still developing PFC.

Ethics Statement
The present study was strictly in compliance with Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All the experimental protocols were approved by the Committee on the Ethics of Animal Experiments of the Fudan University (Permit Number: 2007-0002). All surgery was performed under deeply anesthesia with sodium pentobarbital, and all efforts were made to minimize suffering.
APV (50 mM), DNQX (20 mM) and Picrotoxin (100 mM) were bath-applied throughout the experiment to block excitatory (glutamate) and inhibitory (GABAa) inputs to the recorded cells, so that the recorded cells were synaptically isolated from other cells.
Whole-cell recordings in current-clamp mode were used to test the neuronal excitability and post-burst/spike AHP of the mPFC neurons. The membrane potential was held at 255 mV manually by DC injection throughout the entire experiment.
The neuronal excitability test of the mPFC pyramidal cells was obtained by performing a spike activation protocol. The spike activation protocol contained a 200-ms and 30 pA hyperpolarizing pre-pulse (to monitor changes in the input resistance of recorded cells) and a 500-msecs depolarizing pulse that was adjusted to evoke 3-4 spikes in baseline recordings. The hyperpolarizing and depolarizing pulses were separated by 1 second. The post-burst AHP test consisted of a 13 pulse-train (50 Hz, 2 msec pulse-width, 2.8 nA). The excitability test and the post-spike/burst AHP recording were performed sequentially every 30 secs.

Data Acquisition and Analyses
The biophysical data were digitized and recorded on-line using Clampex (version 9.2.0.09; Axon Instruments, Molecular Devices). The data analyses were performed using Clampfit (version 9.2.0.09; Axon Instruments, Molecular Devices). Cursors-defined mean amplitudes, area and decay time of AHP from 5 consecutive responses before drug application and before wash were taken for paired comparison. Significant main effects were evaluated by paired t-test or unpaired t-test or one-way ANOVA using Origin 8.1 (OriginLab, Northampton, MA) or SPSS Statistic 21 (SPSS, Chicago, IL). P,0.05 was deemed as significant difference. All data are reported as the means 6 SEM.
The peak amplitude and integrated area of the post-burst AHP are measured when the neurons were held at 255 mV. The decay time constant of the post-burst AHP is defined as the time span from 10% to 90% of the amplitude, when the AHP is decaying back to the holding potential of 255 mV. For the ADP measurements, because the variation of the ADP time courses in the recorded cells was large, we fixed a time window (the third second after the beginning of each sweep of recording) to measure the amplitude and area of ADP at that time window after postburst AHP.
For the post-single spike AHP analyses, there are three potential components that are analyzed in the present work. We name the three components as fAHP, fADP and mAHP. The amplitudes of the fAHP, fADP and mAHP are measured at their peak when the cell was held at 255 mV.
All measurements were made .10 min after membrane rupture to allow for adequate internal solution equilibration within the neuron. All traces of potentials shown in figures are average of 5 consecutive responses.
Stock solutions of PTX (50 mM) was prepared in DMSO and stored at 4uC. Stock solutions of DNQX (20 mM) was also prepared in DMSO and stored at 220uC. Stock solutions of APV (25 mM) and quinpirole (10 mM) were prepared in ultrapure deionized water (Millipore Q-Gard 1, Billerica, MA) and stored at 220uC. SKF81297 was prepared as 2.5 mM stock in ascorbic acid (also 2.5 mM in stock) in the lab with lights turned off, and the stock solution was stored at 220uC for no longer than one month. When DA and SKF81297 were bath administered, the microscope light was also dimmed down to prevent photo degradation of the compounds.

Results
Whole-cell recordings were performed in rat layer V-VI mPFC pyramidal cells identified by their morphological and electrophysiological features. As shown in Fig. 1A and 1B, the recorded pyramidal cell had a prominent apical dendrite and showed typical spike frequency adaptation upon injection of 800 ms depolarizing pulse. All cells included in the database of the present study had resting potentials more negative than 260 mV and evoked spikes with amplitude greater than +70 mV.
For recordings of AHP, the basal membrane potential was held at 255 mV manually by DC injection. Post-spike AHP was elicited by a single short pulse (2 ms, +2.8 nA) (Fig. 1C). In our pilot studies, to optimize the recordings of the spike numberdependent post-burst AHP we found that a train of 13 pulses evoked at 50 Hz (2 ms per pulse, +2.8 nA), elicited post-burst AHPs with consistent amplitude and integrated area (Fig. 1C). To prevent synaptic feedback influences on the spike evoked AHP, all recordings were made in the presence of glutamate and GABA receptor antagonists.

DA Enhances Neuronal Excitability and Suppresses Postburst AHP
The effects of DA on neuronal excitability and post-burst AHP were tested alternately in the same individual cells ( Fig. 2A). After establishing a stable baseline recording for at least 5 minutes, DA (30 mM) was bath-applied for 15 minutes. As shown in Fig. 2B, there was an overall significant enhancement of the depolarization pulse-evoked spike number upon DA application (Excitability: from 6.1460.15 to 7.3760.37 Hz, p,0.01, paired t test, n = 14) and the evoked spike responses continued to increase with time, with the peak increase much after the termination of DA application (peak response in excitability: 8.0060.59 Hz, p,0.01, paired t test, n = 14) (Fig. 2B1) [4].
It is notable that with the analysis window placed at the last 5 min of the DA application, 8 of the 14 tested cells (57%) were considered to have an enhanced excitability during the DA treatment, and the remaining 6 cells showed no obvious response during the DA application. However, 3 of these 6 cells showed enhanced excitability after the termination of DA application. We found that the time course of the post-burst AHP change was quite different from that of the excitability change following DA application. In 12/14 tested neurons, the reduced post-burst AHP recovered to baseline level within a few minutes after the termination of DA application, whereas the post-DA enhanced excitability persisted well after termination of DA application and showed only a partial recovery during the washout.
To determine which neurons show this correlated responses of DA-induced reduction of AHP amplitude and increase in neuronal excitability, we plotted the % change of neuronal excitability (during the last 5 minutes of of DA application) against the % change of post-burst AHP for all the 14 tested neurons. As shown in Fig. 2F, the cells (n = 8) showing increased excitability to DA application also showed considerable suppression in post-burst AHP amplitude (31.0463.39%). The other cells (n = 6) showing no increase in excitability during the last 5 mins of DA application showed smaller suppression of post-burst AHP amplitude (12.4364.87% vs. 31.0463.39%, p,0.01, one-way ANOVA). The decay time change in the two groups of cells are also significant different in responding to DA application (37.2665.57% vs. 15.6165.64%, p,0.05, one way ANOVA). These results suggest that there might be two groups of PFC neurons exhibiting different AHP responses during DA application to contribute to the neuronal excitability enhancement. Furthermore, it is also likely that DA enhanced neuronal excitability is biphasic, ie., during DA exposure, DA enhanced neuronal excitability through suppression of AHP, and during post-DA application, the excitability increase is maintained and reached peak responses slowly via other mechanisms not related directly to AHP changes.

D1/5-receptor Stimulation Enhances Neuronal Excitability and Suppresses Post-burst AHP
To identify which DA receptor is likely to regulate this AHP mechanism at this time window, we first examined the effects of the D1/5 receptor agonist SKF81297 on neuronal excitability and post-burst AHP (Fig. 3A). Consistent with previous studies [29], stimulation of D1/5 receptors with SKF81297 (10 mM) significantly enhanced the neuronal excitability (Excitability: from 6.0060.00 to 6.7660.36 Hz, p,0.05, paired t test, n = 19) (Fig. 3B). Towards the end of the 15 min agonist application, 8 out of the 19 cells (42%) showed an increased spike numbers. In the remaining 11 cells, 5 cells showed enhanced excitability only after the termination of D1/5 agonist. The other 6 cells (32%) showed no obvious response to SKF81297 throughout the recordings. Thus, in a total of 68% (13 of 19 cells) of the tested cells, the enhanced excitability persisted and peaked after the termination of SKF81297 application (Excitability: from 6.0060.00 to 7.6260. 38 (Fig. 3C, 3D and 3E). Notably, following application of the agonist SKF81297, the suppressed AHP amplitude persisted longer than that after DA application and showed little recovery during the wash period (compare Fig. 2C 1 with Fig. 3C 1 ).
The D1/5 agonist-induced enhancement of excitability and suppression on post-burst AHP was blocked in the presence of 10 mM SCH23390 before the termination of SKF81927 application (Excitability: from 6.0060.00 to 6.0860.08 Hz, p = 0.37; Post-burst AHP amplitude: from 7.4460.70 to 7.2260.82 mV, p = 0.11; Area: from 1.4660.2 to 1.2560.2 mV*s, p = 0.68, paired t test, n = 5) (Fig. 4B, 4C and 4D). Furthermore, the DA effects on excitability and post-burst AHP were also blocked by the co-administered SCH23390 before the termination of DA application (Excitability: 5.7360.32 to 5.6060.4 Hz, p = 0.17; Post-burst AHP Amplitude: from 10.0061.43 to 10.0161.55, p = 0.89; Area: 6.7362.15 to 6.6361.98 mV*s p = 0.73, paired t test, n = 6) (Fig. 4B, 4C and 4D). Notably, in the absence of SCH23390, a large number of cells showed enhanced excitability after the termination of SKF81297 (13/19) or DA (11/14), while, in the presence of SCH23390, no cell showed obvious change in excitability during the same post-drug time period after SKF81297 (n = 5) or DA (n = 6) application, suggesting that the D1/5 agonist or DA influences on neuronal excitability and post-burst AHP were mediated through the D1/5 receptors.
In the correlation plots in Fig. 3F1 and 3F2, we found that the cells (n = 8) showing increased excitability to SKF81297 application also showed considerable suppression in post-burst AHP amplitude (25.4762.79%), while the other cells (n = 11) showed a smaller suppression of post-burst AHP amplitude (6.7063.57% vs. 25.4762.79%, p,0.01, one-way ANOVA). The decay time changes in the two parts of cells are also significantly different in responding to SKF81297 application (3.2562.94% vs. 21.9564.13%, p,0.01, one-way ANOVA).
Seong and Carter [30] recently reported that D1 receptors are present in a portion of layer V pyramidal cells in the PFC. We found that cells that respond to DA and D1/5-receptor agonist in excitability typically have a higher input resistance (198

D2/3/4-receptor Stimulation Produces No Effect on Neuronal Excitability and Post-burst AHP
In the continuous presence of blockers for glutamate and GABA-mediated synaptic transmission, we then investigated the effects of the D2/3/4 receptor agonist quinpirole on neuronal excitability and post-burst AHP in the same tested neurons. As shown in Figure 5, bath-application of quinpirole (10 mM) induced no significant effect on both the neuronal excitability (Excitability: from 6.3460.36 to 5.8060.14 Hz, p = 0.16, paired t test, n = 6) (  (Fig. 5B, 5C and 5D), suggesting that D2/3/4 receptors were not involved in the DA regulation on neuronal excitability changes and post-burst AHP. In a few neurons, there appeared to have a ''rebound'' excitability increase effect after the termination of quinpiorle (5 minutes) but this was not statistically significant (Excitability: 6.8061.04 Hz, p = 0.33, paired t test, n = 6).

DA and D1/5-receptor Regulation of Excitability and Post-burst AHP does not Involve HCN
Membrane hyperpolarization induced by post-burst AHP typically lasted .100 ms, which is sufficient to activate I h via hyperpolarization-activated cyclic nucleotide-gated (HCN) channels. Thus, the activation of the Ca 2+ -activated K + current (mediating AHP) overlaps with the activation of the opposing I h current that depolarizes the membrane towards resting potential. The interaction between hyperpolarizing AHP and the depolarizing I h are likely to be responsible for the self-sustained recurrent burst firing in pyramidal cells [31].
i) Effect of HCN-channel blocker ZD7288 alone. As shown in Figure 6, when the tested neurons were held at 255 mV, application of ZD7288 (20 mM) alone significantly enhanced the neuronal excitability and increased the input resistance of the examined cells (Excitability: from 6.0860.09 to 7.4460.59 Hz, p,0.05; Input resistance: from 209.97610.03 to 258.46616.22 MOhm, p,0.05, paired t test, n = 6) (Fig. 6A 1 , 6B and 6C), suggesting that HCN channels were opened for I h current flux at the holding potential (255 mV), although, at such a more positive holding potential at 255 mV, a portion of the I h current is likely to be deactivated [32]. When HCN channels were fully blocked with ZD7288, the input resistance increased, resulting in a greater membrane voltage deflection upon a given depolarizing pulse and thereby, facilitating neuronal excitability.
Treatment with ZD7288 (20 mM) alone significantly prolonged the decay time constant of post-burst AHP (From 0.4660.04 to 0.8860.15 s, p,0.05, paired t test, n = 6; Fig. 6E), with no effect on the amplitude of post-burst AHP (From 8.5060.552 to 8.1160.21 mV, p = 0.16, paired t test, n = 6; Fig. 6D), indicating    Fig. 7C and 7D). Further, the DA enhancement of excitability demonstrated no difference in the absence or presence of ZD7288 (Excitability: 119.8865.01% without ZD7288 vs.121.7665.75% with ZD7288, p.0.05, unpaired t test, n = 14 and n = 17 cells, respectively). Notably, in the presence of ZD7288, the reduction of post-burst AHP amplitude and decay time constant did not recover back to the baseline values after the termination of DA application (Fig. 7D), although these two parameters did recover to baseline in the absence of ZD7288 (see Fig. 2E).

DA and D1/5-receptor Regulation of Post-burst Afterdepolarizing Potentials
Spike burst activity not only results in a post-burst AHP, but also afterdepolarizing potential (ADP). Attenuation of AHP has been shown to unveil an afterdepolarization in dopamine neurons and striatal cholinergic interneurons [24,33]. The interaction of AHP and ADP may overlap in time and are likely to be responsible for the self-sustained recurrent burst firing in pyramidal cells [31].
In this study, DA not only reduced the decay time constant of post-burst AHP, but also augmented the ADP component following the post-burst AHP (Fig. 8A). This may be a direct or an indirect effect of DA, since the emerged enhancement of the ADP could be a result of a DA suppression of the AHP, or DA could directly augment the intrinsic mechanisms that mediate the ADP, but are normally and largely masked by the AHP. In the lateral amygdala, DA has also been shown to induce an ADP after a train of action potential [34].
As mentioned above, in the absence of the HCN blocker ZD7288 (20 mM), we found that application of DA (30 uM) not only resulted in a significant reduction in the decay time constant of the post-burst AHP, but also unveiled an emerging longduration (.2 s) ADP (ADP Amplitude: from 1.4960.19 to 2.6260.43 mV, p,0.01, paired t test, n = 14) (Fig. 8B 1 and  8B 3 ). In the presence of ZD7288 (20 mM) the DA augmentation of ADP persisted (Amplitude: from 2.0860.40 to 3.9460.33 mV, p,0.01, paired t test, n = 17, 14 out of 17 cells showed an increase in ADP) (Fig. 8B 2 and 8B 3 ). No difference was found between DA effect on ADP in the absence and presence of ZD7288 (p.0.05, one way ANOVA, Fig. 8B 3 ), suggesting that the DA-induced ADP was also independent of I h . Thus, regardless of whether the HCN channels were blocked or not, DA still significantly suppressed the post-burst AHP and enhanced the ADP.
Similar results were obtained when the D1/5 receptor agonist SKF81297 (10 mM) was applied in the absence or presence of ZD7288 (Amplitude of ADP in the absence of ZD7288: from 1.7060.16 to 2.3460.18 mV p,0.01, paired t test, n = 19; and Amplitude of ADP in the presence of ZD7288: from 1.3860.23 to 2.0560.40 mV, p,0.01, paired t test, n = 10) (Fig. 8C). No difference was found between SKF81297 effect on ADP in the absence and presence of ZD7288 (p.0.05, one way ANOVA). Notably, during the drug application, the onset of the responses in enhancing the ADP amplitude was faster with DA than the D1/5 agonist.

DA and D1/5-receptor Regulation of Post-single Spike AHP
We have also analyzed the effects of DA and D1/5-receptor activation on post-single spike AHP. As shown in Figure 9A, an evoked single spike in mPFC pyramidal cells is usually followed by a post-spike AHP, which is often obscured by a small but fast postspike ADP (fADP). According to their temporal appearance, we subdivided the post-single spike events into three components: fAHP, fADP and mAHP.
We next analyzed the effects of DA and D1/5-receptor stimulation on alternatively tested excitability and post-single spike AHP (these cells have never been tested for post-burst AHP). As shown in Figure 9C, DA alone (30 mM) enhanced the neuronal excitability (from 6.8060.54 to 7.6760.60 Hz, 12.3764.45%, p,0.05, paired t test, n = 12); and apamin alone (0.1 mM) also produced an increase in excitability (from 7.3660.82 to 9.6061.17 Hz, 31.27611.00%, p,0.05, paired t test, n = 6). However, in the continuous presence of apamin (0.1 mM) that blocked the post-spike mAHP, DA no longer altered the neuronal excitability (from 7.9260.71 to 7.2060.55 Hz, 27.9566.79%, p.0.05, paired t test, n = 6). Apamin occluded the effects of DA on neuronal excitability.

Discussion
The present study first replicated previous findings to show that DA enhances the neuronal excitability of pyramidal cells of prepubertal rat PFC during and after its application, and such enhancement was mimicked by activation of D1/5, but not D2/3/ 4, receptors. The DA-enhancement of neuronal excitability was accompanied by a suppression of the apamin-sensitive post-burst AHP, and the post-spike mAHP to reduce spike frequency adaptation. The additional enhancement by DA via D1/5 receptor activation of a post-AHP afterdepolarization (ADP) also contributes to the overall enhancement of neuronal excitability in PFC neurons (Fig. 10).

Methodological Considerations
There are a few notable differences between our approaches in the present study, when compared with previously published studies on the influence of DA on neuronal excitability in PFC neurons. The overall magnitude of the excitability changes reported here is smaller than that reported in the previous studies [4,29]. This was attributed to the time points where the responses were taken. In the present study, the changes in neuronal excitability were measured during the last 5 min of the 15-min drug application period, when the typical maximal DA/D1 effect has not been reached, whereas in the previous study [4], the total mean changes in neuronal excitability were determined during the entire period of drug application and including the 30-min postdrug period, when the maximal effect of DA/D1 effects were achieved. In the present study, we chose the last 5 min of DA/D1 agonist application to quantify neuronal excitability change so that we could compare it with the peak changes in post-burst AHP in the same time window. Our data showed that D1/5 suppression of post-burst AHP contributed to the increase neuronal excitability in In addition, since the afterhyperpolarization is temperaturesensitive and can be evoked maximally at lower (eg, room) temperature [35,36], we adopted the approach of whole-cell recordings at room temperature of 25uC and current-clamped the neurons at 255 mV in order to maximize simultaneously the activation of AHP and neuronal excitability. In contrast, Chen et al. [4] carried out whole-cell recordings at a mean bath temperature of 33uC and current-clamped the neurons constantly at 265 mV [4]. Under both temperatures (25 or 33uC) or holding potentials, the profile of the increase in neuronal excitability by DA or its D1/5 agonist did not differ.

DA or D1/5-receptor Activation Enhances Neuronal Excitability
DA has been reported to exert conflicting influences on neuronal excitability in neurons from different, and even in the same brain region(s). A few previous studies have reported that DA negatively regulates neuronal excitability in hippocampus, entorhinal cortex and striatum [37][38][39] while enhances neuronal excitability in the hippocampus [40]. In the rat PFC, DA enhances the overall neuronal excitability in a prolonged manner (beyond the time when the compound application was terminated) [4,29,30,41,42], and suppresses intracellularly-evoked simulated EPSPs [43](also see Table 1 for a summary). showing that DA enlarges the amplitude of fAHP and fADP, and suppresses mAHP (*p,0.05 for DA vs. baseline, paired t-test); apamin (0.1 mM) itself suppresses mAHP (**p,0.01 for apamin vs. baseline, paired t-test); in the presence of apamin, DA enlarges the amplitude of fAHP and fADP (*p,0.05 for DA+apamin vs. baseline, paired t-test), but does not suppress mAHP anymore (p.0.05 for DA+apamin vs. baseline, paired t-test) (For the grey column with back slash, its baseline refers to the data obtained in the presence of apamin before DA application). (C) Group histograms showing that DA enhances the neuronal excitability; apamin itself enhances the neuronal excitability (*p,0.05 for DA vs. baseline, paired t-test); in the presence of apamin, DA does not change neuronal excitability anymore (p.0.05 for DA+apamin vs. baseline, paired t-test) (For the grey column with back slash, its baseline refers to the data obtained in the presence of apamin before DA application). (D) Group histograms showing SKF81297 enlarges the amplitude of fAHP and fADP, suppresses mAHP (*p,0.05 for SKF81297 vs. baseline, paired t-test); in the presence of apamin, SKF81297 enlarges the amplitude of fAHP and fADP (*p,0.05 for SKF81297+apamin vs. baseline, paired t-test), but does not suppress mAHP anymore (p.0.05 for SKF81297+apamin vs. baseline, paired t-test) (For the grey column with slash, its baseline refers to the data obtained in the presence of apamin before SKF81297 application). (E) Group histograms showing SKF81297 enhances the neuronal excitability; in the presence of apamin, SKF81297 does not change neuronal excitability anymore (p.0.05 for SKF81297+apamin vs. baseline, paired t-test) (For the grey column with slash, its baseline refers to the data obtained in the presence of apamin before SKF81297 application). doi:10.1371/journal.pone.0071880.g009 Not all the recorded cells in our study exhibited excitability enhancement upon DA or D1/5 agonist application. This suggests that there might be two separate DA responsive and nonresponsive pyramidal cells. We did not find their correlation with the amplitude of post-burst AHP or postnatal age (data not shown). A recent study in BAC (bacteria artificial chromosome) transgenic mice suggest that in layer V pyramidal D1+ neurons that fire in spike bursts upon intracellular depolarization pulse activation, has higher input resistance and with smaller Ih than D1-PFC neurons, and shows an increased neuronal excitability in response to a D1 agonist stimulation [30]. In the present analyses, the cluster of neurons that showed increased excitability by D1 agonist SKF81297 also showed the electrophysiological features of typical D1+ neurons, i.e. higher input resistance (p = 0.023) and smaller Ih (p = 0.052) compared with the non-responding cells.
The excitability of some of the cells were partially enhanced during the application of DA or its agonist, but the peak increase excitability occurred after the termination of DA or its D1/5 agonist, when the initially suppressed AHP (during DA or D1/5 agonist application) has already returned back to control levels. This suggested that the suppressed AHP may be contributing to the early phase of increase excitability, but the post-DA excitability increase and maintenance may be mediated via other mechanisms.

DA or D1/5-receptor Activation Suppresses Post-burst Afterhyperpolarization
Both DA and D1/5 agonist were shown to suppress post-burst AHP in the present study. The D1/5 receptor antagonist SCH23390 blocked SKF81297 suppression of post-burst AHP, suggesting SKF81297 specifically acts on D1/5 receptors to inhibit post-burst AHP. Although, D2/3/4 receptor activation led to no change in the neuronal excitability of prepubertal (postnatal days #35 days) PFC pyramidal (present study) and interneurons [44], considering that D2/3/4 receptor modulation of PFC function can be age-dependent [45], our present conclusions for an absence of D2/3/4 effects in neuronal excitability may be restricted to prepubertal PFC neurons. Nevertheless, the D1/5 receptor mechanisms in enhancing neuronal excitability appeared to be less dependent on the age of rat PFC pyramidal neurons (see Table 1). A recent study also showed that norepinephrine induced long duration synaptic depression in PFC is similar in young and adult rats [46]. D1/5 receptor activation-induced increase of neuronal excitability can be through multiple cellular pathways in both PFC pyramidal cells and interneurons, including modulation of K + , Ca 2+ and Na + channels [41,[47][48][49][50]. There are additional interactions between D1 and NMDA receptors [51,52], PKAdependant modulation of K + channels [29,47] and intracellular Ca 2+ dependent signaling [4,26] that mediate the D1-dependent increase in neuronal excitability. In addition to all these mechanisms, we found that D1/5 receptor activation suppressed the Ca 2+ activated K + current that mediates post-spike/burst AHP, to result in an increase in neuronal excitability.

Functional Roles of Intracellular Ca 2+ in Mediating D1/5 Receptor Mediated Changes in AHP and Neuronal Excitability
The enhanced excitability by DA or D1/5 receptor agonist is dependent on intracellular Ca 2+ elevation in the PFC [4]. Yet, both Ca 2+ influx and ryanodine receptor-mediated Ca 2+ release from intracellular Ca 2+ stores are known to activate AHP [53], which will result in a decreased excitability. Hence, the time course of this D1/5 receptor mediated intracellular Ca 2+ changes, the physical compartmentalization of the intracellular pool(s) of Ca 2+ [54] that led to SK channels activation, as well as the dynamic sensitivity of the SK channel to the Ca 2+ , may all contribute differentially to the complex mechanisms that mediate intracellular Ca 2+ dependent neuronal excitability changes in PFC neurons through AHP modulation. Furthermore, we cannot rule out the possibility that D1/5 receptor activation may also suppress SK channels directly, but independent on the early intracellular Ca 2+ level rise following the D1/5 receptor activation. An understanding of the time course and dynamic relationships of DA-dependent Ca 2+ elevation that leads to SK channel activation, DA inhibition of SK channel (and ADP activation, see below), would provide further insight into a fuller understanding of how DA increases iCa 2+ and suppress a Ca 2+ dependent SK channel to enhance neuronal excitability.

The Roles of I h and ADP in DA Enhancement of Neuronal Excitability
Membrane hyperpolarization from the post-burst AHP can activate HCN channels that mediate I h . The depolarization mediated by I h may augment an ADP, especially following DA D1/5 receptor activation. As the voltage activation of these three currents overlaps, the resulting post-burst AHP is a product of an interaction among Ca 2+ -and apamin-sensitive AHP currents, I h and ADP.
HCN channel has a critical role in normalizing the distal dendritic EPSPs with respect to somatic EPSPs, and Ca 2+ spikes [55][56][57]. DA is known to modulate HCN channels to suppress neuronal excitability in layer I interneurons of the PFC and in the pyramidal neurons of rat entorhinal cortex [37,58]. In our present study, after blocking the HCN channels with ZD7288, we showed that DA or D1/5-receptor activation still clearly suppressed the post-burst AHP, and unmasked an ADP. It is known that HCN channels in the cortical pyramidal dendrites show a gradient distribution, with the densest distribution along apical dendrites in the superficial layers [59,60], where D1 receptors are co-located [61]. It is possible that DA modulation of HCN channels is brainregion and cortical-layer specific. Although our data suggests that the increased neuronal excitability is independent of I h in Layer V rat PFC pyramidal neurons, the measurement of neuronal excitability using somatic whole-cell recording by somatic current injection could not fully evaluate the exact roles of the DA on HCN channels in neuronal excitability changes in PFC pyramidal neurons because of the poor voltage control due to 'space-clamp' problem.
DA and D1/5 receptor activation also unveiled a post-AHP ADP. Activation from sub-threshold voltage of this slow ADP could normally be prevented by the post-burst AHP and this ADP could be unveiled when the AHP is suppressed. The induction of ADP is triggered by Ca 2+ influx in the neocortex [62,63], by a decrease of K + channel activity specifically in the olfactory cortex or lateral amygdala [34,64], or by an increase in a persistent Na + current in hippocampal CA1 and striatal interneurons [24,65]. The activated voltage and duration of the unveiled ADP in the present study resembled a dendritic 'hump' potential generated in the proximal dendrites [66]. D1/5 receptor stimulation, via PKA activation, has been shown to potentiate this depolarizing 'hump' potential, which is hypothesized to provide a mechanism to boost sub-threshold EPSPs to fire spike burst [50]. In a single-channel study in pyramidal PFC neurons, it was shown that a D1/5 agonist increased the open probability for a sub-threshold persistent Na + current, which can contribute to this ADP [41,67,68]. Functionally, ADP has been proposed to have a role in short-term memory storage [69] and the DA-induced ADP in rat lateral amygdala neurons plays a crucial role in fear conditioning [34].

Functional Roles of DA Modulation of Neuronal Excitability
In the Layer V pyramidal PFC neurons, following a spike train, several currents, including AHP, I h and ADP, are activated. Together, these currents play a concerted role in regulating the generation of the next spikes to fire and thereby, help to encode spike codes in the PFC. The DA enhancement of neuronal excitability via D1/5-receptor activation does not involve HCN channels (I h ). Instead, DA can suppress Ca 2+ influx via inhibition of dendritic L-or N-type Ca 2+ channel [50,70] that normally would activate SK channels to mediate the post-burst AHP [71,72]. In addition, DA, via D1/5 receptor also increase intracellular Ca 2+ [4,34,50] or a persistent Na + current [68,73], which can contribute to the afterdepolarization (ADP) following the AHP [50]. Since DA release in PFC is known to be associated with working memory and rule learning [74,75], the DA suppression of post-burst AHP, along with an augmentation of an ADP, to promote spike firing to result in a long-term potentiation of neuronal excitability, may represent a DAdependent rule-based learning, which subsequently is being translated to action outcomes in the PFC [75][76][77]. It appears that the prepubertal PFC is endowed with such functional mechanisms that can be fully engaged for the increasing higher cognitive demands during the process of maturation.