Activation of Pyramidal Neurons in Mouse Medial Prefrontal Cortex Enhances Food-Seeking Behavior While Reducing Impulsivity in the Absence of an Effect on Food Intake

Abstract

The medial prefrontal cortex (mPFC) is involved in a wide range of executive cognitive functions, including reward evaluation, decision-making, memory extinction, mood, and task switching. Manipulation of the mPFC has been shown to alter food intake and food reward valuation, but whether exclusive stimulation of mPFC pyramidal neurons (PN), which form the principle output of the mPFC, is sufficient to mediate food rewarded instrumental behavior is unknown. We sought to determine the behavioral consequences of manipulating mPFC output by exciting PN in mouse mPFC during performance of a panel of behavioral assays, focusing on food reward. We found that increasing mPFC pyramidal cell output using designer receptors exclusively activated by designer drugs (DREADD) enhanced performance in instrumental food reward assays that assess food seeking behavior, while sparing effects on affect and food intake. Specifically, activation of mPFC PN enhanced operant responding for food reward, reinstatement of palatable food seeking, and suppression of impulsive responding for food reward. Conversely, activation of mPFC PN had no effect on unconditioned food intake, social interaction, or behavior in an open field. Furthermore, we found that behavioral outcome is influenced by the degree of mPFC activation, with a low drive sufficient to enhance operant responding and a higher drive required to alter impulsivity. Additionally, we provide data demonstrating that DREADD stimulation involves a nitric oxide (NO) synthase dependent pathway, similar to endogenous muscarinic M3 receptor stimulation, a finding that provides novel mechanistic insight into an increasingly widespread method of remote neuronal control.

Keywords: DREADD; food; impulsivity; operant behavior; prefrontal cortex.

Figure 1

Expression of hM3Dq designer receptors exclusively activated by designer drugs (DREADD). Anti-mCitrine staining shows expression of the adeno-associated virus (AAV)-CaMKIIa-HA-hM3D(Gq)-IRES-mCitrine vector reporter in the mPFC (A). Positively stained neurons had a characteristic pyramidal morphology (B). (C) Single cell RT-PCR of 10 mCherry-positive cells confirmed that DREADD-expressing cells were predominantly glutamatergic (9/10 cells). 8/10 cells also expressed additional tested markers. Lanes 1–10 are picked cells, while lanes 11 and 12 are no-template control reactions. GAPDH, Glyceraldehyde 3-Phosphate Dehydrogenase; VGLUT1, Vesicular Glutamate Transporter 1; VGLU2, Vesicular Glutamate Transporter 2;GAD65a, GAD65b, Glutamic Acid Decarboxylase 65; GAD67, Glutamic Acid Decarboxylase 67; PARV, Parvalbumin; SOM, Somatostatin; VIP, Vasoactive Intestinal Peptide.

Figure 2

Activity of hM3Dq DREADD. Injection of a depolarizing current step (300 pA) evoked action potential firing in labeled mPFC neuron (n = 10; 6 mice: A). Application of CNO (5 mins) caused membrane depolarization and triggered action potential bursting (n = 19; 6 mice: B). (C) Quantification of the dose dependent increases in membrane depolarizations and firing rates (D). CNO induced depolarizations and firing bursts were abolished by L-NAME (500 μM; E). (F) Quantification of reduction in action potential frequency by L-NAME. Values represent means ± SEM. *p < 0.001 ANOVA followed by Tukey’s post hoc test.

Figure 3

Excitation of mPFC pyramidal neurons (PN) enhances operant responding. Saline (white bars) had no effect on operant responding in DREADD-expressing mice (N = 16) relative to wildtype mice (N = 8) in either the food restricted (A) or fed (B) state, whether measured as breakpoint, total nosepokes delivered, or rewards obtained. In contrast, CNO (striped bars) did enhance operant responding in DREADD-expressing mice relative to wildtype mice in the food-restricted state, but not in the fed state. Error bars are ±SEM. *p < 0.05, 2-way rmANOVA followed by Sidak’s multiple comparisons test.

Figure 4

Exciting mPFC PN enhances reinstatement of food seeking. Note: Saline or CNO was administered only during the reinstatement test; the groups are separated during Acquisition and Extinction to demonstrate nose poking in each group. Mice in both the Saline (N = 8) and CNO (N = 9) groups successfully reinstated nose poking behavior during the Reinstatement test. Mice given CNO poked significantly more than mice given saline during the test. Error bars are ± SEM. *p < 0.05, ***p < 0.001, 2-way rmANOVA followed by Sidak’s multiple comparisons test.

Figure 5

Exciting mPFC PN reduces impulsivity. CNO did not significantly alter premature responding in DREADD-expressing mice as a group (N = 12) in a 3-choice serial reaction time task when administered at 0.5 mg/kg (striped bars) or 2.5 mg/kg (black bars), relative to saline (white bars) (A). However, in highly impulsive mice (N = 8) CNO did decrease premature responding (B), an effect attributed to higher doses of CNO (2.5 mg/kg). Error bars are ± SEM. *p < 0.05 ANOVA followed by Tukey’s post hoc test.

Figure 6

Exciting mPFC PN does not alter binge-like consumption. CNO did not alter consumption of a high fat diet in sated mice (N = 10) whether administered at 0.5 mg/kg (A) or 2.5 mg/kg (B) when food intake was measured after 15, 30, and 60 min. Error bars are ± SEM, rmANOVA.

Figure 7

Exciting mPFC PN does not alter food intake after fasting. Regular chow intake in fasted mice (N = 10) was similar after saline (white bars) and CNO 2.5 mg/kg (black bars). Error bars are ± SEM, rmANOVA.

Figure 8

Exciting mPFC PN does not alter behavior in the open field. CNO, whether administered at 0.5 mg/kg (striped bars) or 2.5 mg/kg (black bars) had no effect on time spent in the center or time spent in the border of an open field arena, relative to saline (white bars). CNO also had no effect on total distance moved (CNO Panel A,C, N = 7; Saline Panel A,C, N = 7; CNO Panel B,D N = 12, Saline Panel B,D N = 12). Error bars are ± SEM. Unpaired t-test.

Figure 9

Exciting mPFC PN does not alter social interaction. In a social interaction task CNO, whether administered at 0.5 mg/kg (A, striped bar, N = 8) or 2.5 mg/kg (B, black bar, N = 10) had no effect on the ratio of time spent interacting with a novel mouse to time spent interacting with an empty animal restrainer, relative to saline (white bars, (A) N = 8, (B) N = 12). Error bars are ± SEM. Unpaired t-test.

Figure 10

Targeted neurons exhibit partial overlap with the DR1 expressing population. (A) Co-injection of AAV-CamKII-hM3D(Gq) and Cre-dependent SUN2-myc into DR1-Cre mice demonstrated that the neurons targeted in this study only partially overlap with the DR1-expressing population. (A₁) SUN2 expression, demonstrating DR1 promoter activity; (A₂) mCherry expression, showing AAV-hM3D(Gq) expression; (A₃) Merged image, with example cells of interest identified. (B) A parallel experiment using AAV-CamKII-eYFP instead of AAV-CamKII-hM3D(Gq) showed similar results. (B₁) eYFP epifluoresence; (B₂) SUN2 expression; (B₃) Merged image. In (A,B), Filled arrowheads: DR1/SUN2 expression only; Open arrowheads: dual expression of DR1/SUN2 and eYFP/mCherry/DREADD expression; Arrows: eYFP/mCherry/DREADD expression only. In (C), arrows indicate examples of the expression of both SUN2-myc (C₁), tdTomato (C₂), and Merge (C₃) in a control experiment, in VGAT-Cre-tdTomato mice (Vong et al., 2011). No Sun2-myc single positive neurons were observed. Scale bars are 100 μm.