Dopamine modulates persistent synaptic activity and enhances the signal-to-noise ratio in the prefrontal cortex

Abstract

Background: The importance of dopamine (DA) for prefrontal cortical (PFC) cognitive functions is widely recognized, but its mechanisms of action remain controversial. DA is thought to increase signal gain in active networks according to an inverted U dose-response curve, and these effects may depend on both tonic and phasic release of DA from midbrain ventral tegmental area (VTA) neurons.

Methodology/principal findings: We used patch-clamp recordings in organotypic co-cultures of the PFC, hippocampus and VTA to study DA modulation of spontaneous network activity in the form of Up-states and signals in the form of synchronous EPSP trains. These cultures possessed a tonic DA level and stimulation of the VTA evoked DA transients within the PFC. The addition of high (> or = 1 microM) concentrations of exogenous DA to the cultures reduced Up-states and diminished excitatory synaptic inputs (EPSPs) evoked during the Down-state. Increasing endogenous DA via bath application of cocaine also reduced Up-states. Lower concentrations of exogenous DA (0.1 microM) had no effect on the up-state itself, but they selectively increased the efficiency of a train of EPSPs to evoke spikes during the Up-state. When the background DA was eliminated by depleting DA with reserpine and alpha-methyl-p-tyrosine, or by preparing corticolimbic co-cultures without the VTA slice, Up-states could be enhanced by low concentrations (0.1-1 microM) of DA that had no effect in the VTA containing cultures. Finally, in spite of the concentration-dependent effects on Up-states, exogenous DA at all but the lowest concentrations increased intracellular current-pulse evoked firing in all cultures underlining the complexity of DA's effects in an active network.

Conclusions/significance: Taken together, these data show concentration-dependent effects of DA on global PFC network activity and they demonstrate a mechanism through which optimal levels of DA can modulate signal gain to support cognitive functioning.

Figure 1. Properties of the organotypic triple co-culture system and the DA innervation of the PFC as demonstrated by tyrosine-hydroxylase (TH) containing fibers.

A) Schematic representation of the triple co-culture consisting of the PFC, VTA, and hippocampus. Electrical stimulation of the afferents from the VTA (indicated as green lines) or ventral hippocampus (red lines) induces Up-states in the PFC. B-D) Photomicrographs illustrating putative DAergic (TH-positive) neurons in the VTA and the distribution of TH-fibers in the PFC. Co-cultures were made from mice expressing green fluorescent protein under the control of the TH gene promoter. C) Properties of putative DAergic (green TH-positive) neurons in the VTA. Cell-attached recordings (top left inset) show that DA neurons are tonically active. Bottom right inset: Membrane properties and firing response in whole-cell mode in response to a series of hyperpolarizing and depolarizing current steps (−150 to+120 pA). The recorded cell was filled with Alexa 594 after break-in. D) shows the laminar distribution of fibers in the PFC. E) Morphological properties of a pyramidal cell (top) and interneuron in the PFC of organotypic co-cultures. Cells were loaded with Alexa 594 during recording and visualized using series of confocal images. Images are montages of convoluted z-stacked images at 40× magnification in C-F. All images were contrast-enhanced for clarity. F) Electrochemical detection of phasic DA release in the PFC following stimulation of the VTA. Stimulation trains (3–100 pulses) were initiated at time 0, and evoked an increase in extracellular DA. Scale bar is 200 nM. The insert shows background-subtracted cyclical voltammograms taken at the peak of the response for each of the stimulations. Abbreviations: VTA, ventral tegmental area; PFC, prefrontal cortex, Cg1, cingulate cortex; WM, white matter.

Figure 2. Cortical Up-states in organotypic co-cultures are a network phenomenon.

The membrane potential of cortical neurons in PFC-Hipp-VTA co-cultures alternates between a hyperpolarized Down-state close to the resting membrane potential and a depolarized Up-state during which action potential firing occurs. Up-states could be evoked synaptically by short burst stimulation of the VTA, the hippocampus, or the contralateral PFC, respectively (see text for details). Inclusion of the Na⁺ channel blocker QX-314 in the recording pipette did not alter the occurrence or duration of Up-states. In contrast, glutamatergic transmission at both non-NMDA and NMDA receptors is required to initiate and sustain Up-states, respectively. In the presence of the NMDA receptor antagonist CPP (10 µM; n = 7), stimulation of the VTA or the hippocampus evoked large EPSPs, but these failed to evoke recurrent activity and Up-states. Bath application of CNQX (20 µM; n = 18) blocked all evoked responses following stimulation of either the VTA or the hippocampus, as well as a large portion of locally evoked PSPs.

Figure 3. Dopamine-modulation of cortical Up-states is concentration-dependent.

DA was bath-applied to VTA-PFC-Hipp co-cultures and Up-states were evoked by VTA stimulation (see insert). A) Representative traces showing the effects of high (10 µM) DA on VTA-evoked Up-states. B) At concentrations of 1 µM exogenous DA or higher the duration and number of spikes during the initial 500 ms of the Up-state were significantly reduced. These effects on Up-states were abolished when DA receptors were blocked by combined pre-application of the D1 receptor antagonists SCH 23390 and sulpiride (5 µM each) to the bath for 10 minutes before application of DA (10 µM). C) In marked contrast to the reductions in Up-state duration and action potential firing due to network activity, the number of spikes evoked by somatic current injection was consistently increased across a wide range of exogenous DA concentrations, starting at 100 nM. C1) In the presence of DA, the same cell as shown in A) displays a significant increase in evoked spikes in response to a square pulse current injection. C2) Summary graph of the effects of various bath-applied DA concentrations on spike firing evoked by somatic current injection. Similar to the effects on Up-states shown in B), increases in current-evoked spike firing depended on DA receptor activation, and accordingly pre-application of SCH-23390 and sulpiride blocked the effects of 10 µM DA. Statistical comparisons used paired t-tests after repeated measures ANOVA. Levels of significance for multiple comparison were * P<0.01, and ** P<0.005. The number of cells in each group used for comparisons in B) and C) are indicated in B1. The same cells were used for measurements in B) and C).

Figure 4. Cocaine enhances endogenous DA activity to reduce VTA-evoked Up-states.

A) Representative traces illustrating the effects of 5 µM cocaine on cortical Up-states evoked by VTA-stimulation. B) Altering DA transmission with cocaine (5 or 10 µM, N = 10) resulted in transient reductions in Up-state duration and spike number during the Up-state (Level of significance * P<0.05, and ** P<0.01, compared to baseline, paired Student's t-tests).

Figure 5. Acute blockade of DA receptors does not affect properties of cortical Up-states in VTA-PFC-Hipp co-cultures.

A) Representative traces of cortical Up-states synaptically evoked by brief burst stimulation of the VTA (2–6 pulses at 20 Hz), before (top), during (middle) and 20 minutes after bath application of the DA receptor antagonists SCH23390 and sulpiride (both 5 µM). The insert shows a diagram of the recording configuration with the stimulation electrode in the VTA and the recording electrode in the PFC. B) Bath application of either the DA D1 receptor antagonist SCH23390 (5 µM) alone (n = 9), or in combination with the D2 receptor antagonist sulpiride (5 µM; n = 5) had no significant effect on Up-state duration, or the number of spikes during the first 500 ms of VTA-evoked Up-states.

Figure 6. In co-cultures that show no or reduced DAergic tone bath application of DA can increase Up-state duration.

A-C) PFC-PFC-Hipp co-cultures were prepared to study the acute effects of DA in the absence of DAergic innervation from the midbrain. Up-states were evoked by electrical stimulation of either the ventral Hipp (n = 23) or the contralateral PFC (n = 12) and the data were pooled (see text for details). The insert in A) shows a diagram of the 2 possible recording configurations. A) Representative traces of Up-states evoked by stimulation of the contralateral PFC before, during and after bath application of 1 µM DA. B) Summary graph of the effects of various doses of bath-applied DA on Up-states in PFC-PFC-Hipp co-cultures. At low to moderate doses (100 nM – 1 µM) DA augmented Up-state duration (B1) and the number of spikes in evoked Up-states (B2). Further increasing exogenous DA concentrations (10 µM) significantly shortened Up-states and the number of spikes in the Up-state, similar to the effects observed in VTA-PFC-Hipp co-cultures. C) Summary graph of the effects of various bath-applied DA concentrations on spike firing evoked by somatic current injection. With the exception of the lowest dose (10 nM) DA consistently increased the number of spikes evoked by somatic current pulses. Statistical comparisons used paired t-tests after repeated measures ANOVA. Levels of significance for multiple comparison were * P<0.0125, and ** P<0.00625. The numbers of cells in each group used for comparisons are indicated in B1 and C), respectively. D) DA levels in VTA-PFC-Hipp cultures were reduced by adding reserpine (10 µM) and AMPT (100 µM) to the culture media for 5 hours prior to recordings. Up-states were evoked by VTA stimulation and 100 nM DA were bath applied. E) In co-cultures in which DA release was reduced over several hours application of 100 nM DA significantly increased the duration of Up-states as well as the number of spikes in the Up-state. The number of spikes evoked by somatic current injection was also increased (paired t-tests; n = 12).

Figure 7. Dopamine modulation of synaptic short-term plasticity in the Down-state.

A) Representative traces of EPSPs under baseline conditions (black trace) and following bath application of 100 nM DA. Trains of EPSPs (15 pulses at 20 Hz) were evoked by local stimulation of afferents in the PFC in PFC-Hipp-VTA co-cultures. Under control conditions trains of EPSPs typically showed a mixture of synaptic depression and summation. The glutamatergic nature of the synaptic response was confirmed at the end of the experiment through bath application of the AMPA antagonist CNQX (20 µM; right trace) Traces represent averages of 20 sweeps. The insert illustrates the measurements (amplitude and area) obtained for each EPSP in the train. The EPSPs shown in this example are indicated by the shaded area in the train on the left. B, C) Dopamine (red symbols) at 100 nM increased both the amplitude B), and area under the EPSP C) over baseline values (black symbols). The DA-induced changes in EPSP amplitude and area became significant after short repetitive stimulation, starting with the 4^th pulse. Statistical comparisons used paired t-tests after repeated measures ANOVA (* P<0.0033, and ** P<0.00165). D, E) The low concentration of 10 nM DA did not alter EPSP amplitude D) or area under the curve E). At high levels of exogenous DA (10 µM) the amplitude of the EPSPs, F), was reduced across all pulses in the train (repeated measures ANOVA; P<0.05). G) The area under the curve showed a similar trend but this change did not reach significance in our sample (n = 8).

Figure 8. Dopamine enhances EPSP-spike coupling at moderate concentrations.

A) Example trace illustrating the recording set-up used to test DA modulation of EPSPs during active network states. Up-states were evoked by stimulation of the hippocampus in PFC-Hipp-VTA co-cultures. After a minimum of 500 ms (but typically between 1000–1500 ms) into the Up-state, trains of EPSPs (15 pulses at 20 Hz, indicated by the red box) were evoked by local stimulation of afferents close to the neuron recorded in the PFC using the same neurons and stimulation parameters as shown in Figure 7 for the Down-state (N = 10). B) Representative traces showing the effectiveness of EPSPs to induce action potential firing during the Up-state under the baseline (black trace) and 100 nM DA condition (red trace). The green trace shows the averaged synaptic response during the Down-state before DA application. C) Summary graph showing the overall increase in spike number during the period of synaptic stimulation C2) and the change in EPSP-spike coupling in the 100 nM DA condition (red symbols) over baseline (black symbols). The plot shows for each pulse in the train the probability that a spike occurred within 10 ms of the onset of the stimulation. In the 100 nM DA condition the probability that an EPSP evoked an action potential was generally increased across all pulses. The relative magnitude of this effect was greater at later pulses in the train, with pairwise comparisons showing significant increases in spike probability over baseline starting at the 4^th pulse. D) In contrast to the effects during synaptic stimulation, bath application of 100 nM DA had no significant overall effect on Up-state duration (top) or the number of spikes before local synaptic stimulation (during the first 500 ms of the Up-state). Post-hoc comparisons used paired t-tests after repeated measures ANOVA (Bonferroni-adjusted level of significance * P<0.0033, and ** P<0.00165). E) The low dose of 10 nM DA had no effect of EPSP spike-coupling during the Up-state, or the overall properties of the Up-state (the inserts show measures for total Up-state duration, top bar graph; or number of spikes during the period before local synaptic stimulation, bottom; n = 7). F) In contrast, the high concentration of DA (10 µM) significantly reduced both Up-state duration (top) and the number of spikes within the Up-state. Thus, under these conditions both the synaptic signal (c.f. Fig. 7) as well as the background network activity were reduced. For comparisons at all concentrations the same cells were used as in Figure 7. In the 10 µM DA condition one cell dropped out because the Up-states evoked by hippocampal stimulation were too brief to allow stimulation of EPSP trains during the Up-state.