Dopamine attenuates prefrontal cortical suppression of sensory inputs to the basolateral amygdala of rats

Abstract

The basolateral complex of the amygdala (BLA) plays a significant role in affective behavior that is likely regulated by afferents from the medial prefrontal cortex (mPFC). Studies suggest that dopamine (DA) is a necessary component for production of appropriate affective responses. In this study, prefrontal cortical and sensory cortical [temporal area 3 (Te3)] inputs to the BLA and their modulation by DA receptor activation was examined using in vivo single-unit extracellular recordings. We found that Te3 inputs are more capable of driving BLA projection neuron firing, whereas mPFC inputs potently elicited firing from BLA interneurons. Moreover, mPFC stimulation before Te3 stimulation attenuated the probability of Te3-evoked spikes in BLA projection neurons, possibly via activation of inhibitory interneurons. DA receptor activation by apomorphine attenuated mPFC inputs, while augmenting Te3 inputs. Additionally, DA receptor activation suppressed mPFC-induced inhibition of Te3-evoked spikes. Thus, the mPFC may attenuate sensory-driven amygdala-mediated affective responses via recruitment of BLA inhibitory interneurons that suppress sensory cortical inputs. In situations of enhanced DA levels in the BLA, such as during stress and after amphetamine administration, mPFC regulation of BLA will be dampened, leading to a disinhibition of sensory-driven affective responses.

Fig. 1.

Example of Pontamine sky blue-labeled recording site. The recording site could be effectively determined by examination of the Pontamine sky blue iontophoresed from the tip of the electrode at the conclusion of the recording. The nuclei were determined after cresyl violet staining of the tissue sections.

Fig. 2.

Characteristics of BLA neuronal activity. A, Evoked spikes were characterized as originating from projection neurons or interneurons using the criteria of firing rate and spike duration. Aligned with they-axis is a distribution histogram of firing rates, and aligned with the x-axis is a distribution histogram of spike durations (from a randomly selected sample of neurons,n = 59). The circles represent each individual neuron plotting its spike duration as a function of its firing rate. The presumed interneurons (black circles) consistently show faster firing rate and shorter spike duration than do the presumed projection neurons (gray circles).B, Antidromic responses of BLA neurons that project to the mPFC (AD; n = 34) display longer latencies than mPFC-evoked responses in BLA interneurons (IN; n = 40). Therefore, the significantly shorter latency of mPFC-evoked responses on BLA interneurons compared with projection neurons (PN;n = 42) cannot be attributable to antidromic activation of a BLA neuron that projects to the mPFC. Antidromic activation of BLA projection neurons is confirmed by the ability of the spikes to follow high-frequency stimulation (300 Hz, 0.6 mA, 0.4 msec duration, three stimuli at arrows) (C) and constant response latency (1), and collision (3) with a spontaneous spike (2) (D).

Fig. 3.

Placement of stimulating and recording electrodes.A, Stimulating electrode placements in the Te3 (3; −4.3 to −6.7 mm bregma) that evoked short-latency responses in the BLA (2; black andwhite circles, −2.8 to −4.2 mm bregma), and mPFC stimulation sites (1; +4.2–2.2 mm bregma) that suppressed Te3-evoked responses (2; white circles). B, Stimulating electrode placements in the infralimbic (gray circles) and prelimbic (black circles) cortex subdivisions of the mPFC (1; +4.2–2.2 mm bregma) that evoked short-latency responses in the BLA (2; white circles indicate infralimbic cortex-evoked responses, and black circles indicate prelimbic cortex-evoked responses, −2.3 to −3.8 mm bregma). This figure included only those neurons used for analysis in this study.

Fig. 4.

Prefrontal cortical stimulation evokes bursts of spikes in BLA interneurons and suppresses the activity of many projection neurons. A, mPFC stimulation (0.4 mA, 0.3 msec duration) evokes a short-latency burst of spikes in a BLA interneuron. B, Peristimulus time histogram (PSTH) of mPFC-evoked short-latency responses in a single BLA interneuron (10 sweeps, 0.6 Hz, 0.4 mA, 0.3 msec duration). C, PSTH of mPFC-induced suppression (*) of spontaneous spike discharge of a BLA projection neuron (60 sweeps, 0.5 mA, 0.6 Hz, 0.3 msec duration). Stimulation occurs at time = 0 in each PSTH.

Fig. 5.

mPFC and Te3 monosynaptic inputs exert different effects on BLA interneurons and projection neurons.A, mPFC stimulation at increasing stimulus intensities is more effective at driving BLA interneurons than is stimulation of Te3 inputs. B, Te3 stimulation at increasing intensities is more effective at driving projection neuron firing than is stimulation of mPFC inputs. To compare stimulation–response curves, threshold stimulation intensity (T) is the lowest stimulation that evokes a spike at least once in >40 consecutive attempts (ranges, 2–12% spike probability). Stimulation intensities 0.05 mA lower than T do not evoke any spikes in at least 50 stimulations. From threshold stimulation intensity, the intensity is increased in steps of 0.1 mA. Each 0.1 mA step is labeled consecutively, beginning at 2. The Te3 and mPFC inputs significantly differ from each other at every stimulus intensity plotted, other than threshold stimulation intensities (t test;p < 0.05).

Fig. 6.

Differences in latency of afferent-evoked responses between interneurons and projection neurons. A portion of Figure 2 is reproduced here for comparison. Distribution histograms of the latencies of mPFC-evoked responses (A) in BLA projection neurons (top panel) and interneurons (bottom panel) indicate that interneuronal responses often precede responses in projection neurons, indicative of the suppressive effect that the mPFC may exert over the BLA via feedforward inhibition. This can lead to preclusion of BLA projection neuron firing and, therefore, BLA output. However, the overlapping latencies of Te3-evoked responses (B) in BLA projection neurons (top panel) and interneurons (bottom panel) indicates that Te3-evoked inhibition may serve to inhibit competing paths or increase the acuity of Te3-evoked excitation. Average latencies are represented byvertical dashed lines.

Fig. 7.

Patterns of paired-pulse facilitation of Te3 and mPFC inputs to interneurons and projection neurons of the BLA. Response probability to the second stimulation pulse is divided by the first stimulation pulse. If there is no facilitation or depression, the result will be a value of 1 (dashed line). Values >1 indicate facilitation, whereas values <1 indicate depression. A >30% change from baseline is considered significant (*). A, Paired-pulse stimulation of mPFC (top panel) or Te3 (bottom panel) inputs led to either facilitation or depression of responses in BLA projection neurons. B, In BLA interneurons, paired-pulse stimulation of Te3 inputs (1) led to facilitation (F), early depression (ED), or late depression (LD; from top tobottom panel). In BLA interneurons, paired-pulse stimulation of mPFC (2) resulted in ED, LD, or F (from top to bottom panel).

Fig. 8.

mPFC suppresses Te3-evoked monosynaptic responses in projection neurons, and this suppression is attenuated by DA receptor activation. A, Te3 stimulation (0.7 mA, 0.3 msec duration) evokes a short-latency, presumably monosynaptic response in a BLA projection neuron. B1, Overlaid traces of 20 Te3-evoked responses demonstrate a relatively narrow distribution of latencies, consistent with a monosynaptic response. The stimulation intensity can be altered to evoke an ∼50% response probability.B2, mPFC stimulation (0.6 mA, 0.3 msec duration) 20 msec before Te3 stimulation decreases the probability of Te3-evoked responses, demonstrated by fewer evoked spikes in these traces.C1, After apomorphine administration (0.5 mg/kg, i.v.), the stimulation intensity of Te3 is altered until an ∼50% response probability is regained (to 0.7 mA). C2, After apomorphine administration, mPFC stimulation 20 msec before Te3 stimulation no longer produces the potent suppression of Te3-evoked responses. D, The response probabilities for the sample traces in B1–C2 are illustrated for this neuron.E, Overall, in the neurons tested (n= 23 of 27) Te3-evoked responses are significantly attenuated by mPFC stimulation, and this attenuation is removed by apomorphine administration (n = 6/7). F, The time course of the mPFC–Te3 interaction under control conditions (F1; * indicates a p < 0.05 significant difference relative to baseline Te3-evoked response probability) and after apomorphine administration (F2).

Fig. 9.

mPFC stimulation suppresses fiber bundle-evoked responses in BLA neurons. A, Stimulation of mPFC 20 msec before stimulation of the stria terminalis fiber bundle attenuates the probability of a stria terminalis-evoked monosynaptic spike (*p < 0.05). These data demonstrate that the mPFC is not likely attenuating Te3-evoked responses by an action at the Te3 cell body, but instead is probably having an effect within the BLA.B, The time course of mPFC-evoked suppression of stria terminalis-evoked responses (* indicates that stria terminalis-evoked response probability is significantly (p < 0.05) attenuated by previous mPFC stimulation, compared with baseline stria terminalis-evoked response probability).

Fig. 10.

DA receptor activation has opposite effects on mPFC and Te3 inputs to the BLA. In projection neurons, apomorphine causes a downward shift in spike probability after mPFC stimulation (A) while causing a decrease in threshold for responses after Te3 stimulation (B). For interneurons, apomorphine produces a potent attenuation of mPFC-evoked firing (C) while increasing the Te3-evoked spiking (D). Thus, apomorphine attenuates prefrontal cortical inputs to projection neurons and interneurons, whereas it augments Te3 inputs to projection neurons and interneurons. In all graphs (*) indicates that the comparison to control at the same stimulus intensity is significantly different (p < 0.05).

Fig. 11.

DA receptor activation alters paired-pulse facilitation of selected BLA afferents. A, DA receptor activation had negligible effects on paired-pulse facilitation or depression of Te3 inputs to BLA projection neurons. B,Paired-pulse facilitation, but not depression, of Te3 inputs to BLA interneurons was altered by DA receptor activation. C,Paired-pulse facilitation of mPFC inputs to BLA interneurons (and depression; data not shown) was altered by DA receptor activation. A change of >30% is considered significant (*) when comparing control and postapomorphine administration values of paired-pulse facilitation at a given ISI.

Fig. 12.

Schematic of the mPFC regulation of BLA output and its modulation by DA. In this figure, neurons of the mPFC and Te3 that project to the BLA are represented by triangles,and their level of activity is represented by firing rate histograms within the triangles. The dominant input is represented by abolder line connecting the cortical area with the BLA. BLA projection neuron (triangle) activity is represented by hypothetical voltage traces within the triangle. A,Enhanced mPFC activity (1) will decrease BLA output (hyperpolarization; 2) caused by activation of a BLA inhibitory interneuron (3). B,Enhanced sensory cortical activity (1) will lead to action potential firing (2) in BLA projection neurons. C, mPFC inputs (1) that occur concomitant with sensory cortical inputs (2) will dampen the spike firing that is induced by sensory cortical inputs under basal conditions (EPSP without action potential; 3). D, If the sensory stimulus has affective value, DA is released in the BLA. In the presence of DA (1), the BLA output will be enhanced (numerous action potentials; 4) by a combination of attenuated mPFC inputs (2) to interneurons, and augmented sensory cortical inputs (3).