Characterization of cognitive and mesofrontal deficits in Arc mutant mice.

(a) Diagram showing the navigation choices for mice in a Y-maze. At the center of the maze, a mouse has a choice to enter either a new arm (alternation) or a previously visited old arm. (b) Alternation percentage in the Y-maze task showing significant reduction in the Arc-/- animals compared to wildtype animals (**p=0.001, t-test, t(15)=3.975, WT N = 8, Arc-/- N = 9 mice, both groups passed Shapiro-Wilk normality test at alpha=0.05). (c) Total arm entries are comparable between Arc-/- and WT. (d) Top, schematic for AAV injection in TH-Cre animals to label dopaminergic neurons. Bottom, confocal image showing tdTomato (red) and SypGFP (green) labeling in the VTA. Scale bar, 100 µm. (e) Left, confocal image showing labeled dopaminergic axons in the frontal cortex. The dotted line indicates the region-of-interest for M2. Scale bar, 100 µm. Right, zoomed-in region showing labeled axons (tdTomato, red) and boutons (tdTomato+SypGFP, yellow). Scale bar, 20 µm. (f, g) The normalized axon length (f) is not significantly different. The normalized bouton density (g) is significantly reduced in Arc-/- animals compared to WT (*p=0.034, t-test, t-test, t(12)=2.393, N=7 mice for each group). The axon length is normalized by the number of labeled cells in VTA, the bouton density is normalized by the axon length, and both are expressed as a percentage of the group average in WT mice. All the error bars indicate SEM.

Task-coordinated frontal neuronal ensemble activity is disrupted in Arc mutant mice.

(a) Diagram showing the setup for miniaturized microscope imaging of frontal cortical activity in mice performing the Y-maze task. The example image represents the projection (by standard deviation of ΔF/F) of a calcium activity movie (∼500 s), showing labeled M2 frontal cortical neurons. Scale bar, 100 µm. (b) Top, an example plot showing the positions of a wildtype mouse relative to the Y-maze center during navigation. Bottom, raster plot showing the calcium activity of M2 neurons simultaneously recorded during navigation. (c) Top, the average activity of individual frontal cortical neurons at binned positions relative to the center of Y-maze in WT (938 neurons from 8 animals) and Arc-/- (1338 neurons from 7 animals) animals. Bottom, the proportion of neurons showing maximal activation at each maze position. This proportion peaked right before the maze center in WT mice but reduced significantly in Arc-/- mice. Shaded areas indicate 95% confidence intervals. (d) Examples of neurons showing differential (alternation-selective, traces from three neurons on the left) or similar (non-selective, traces from three neurons on the right) activity between new and old arm visits. (e) The proportion of alternation-selective neurons peaked right before the maze center in WT mice but reduced significantly in Arc-/- mice. Shaded areas indicate 95% confidence intervals.

Adolescent dopamine neuron stimulation leads to long-term reversal of mesofrontal circuit deficits.

(a) Schematic for labeling dopamine neurons with DREADD-Gq-mCherry and imaging frontal cortical activity with GCaMP6. The example image at the top shows a two-photon image of labeled cortical neurons. Scale bar, 20 μm. (b) Example traces of spontaneous cortical activity averaged from the whole image frame in control mCherry only (blue) and DREADD-Gq (red) animals before and 1hr after CNO injection. (c) Gq animals show significantly higher change in cortical activity after CNO injection compared to control mCherry only animals, suggesting CNO-induced activation of the mesofrontal circuit (*p=0.018, t-test, t(10)=2.814, N=6 mice for each group). Cortical activity is summarized by the standard deviation (SD) of the spontaneous activity traces. Activity change is calculated as (SD2-SD1)/SD1, where SD1 is before and SD2 after treatment. (d) Diagram showing experimental procedures to evaluate the effect of adolescent dopamine neuron stimulation on the structure of frontal dopaminergic projections. (e, f) Normalized axon length (e) is not significantly different. Normalized bouton density (f) is significantly increased in Gq animals compared to Ctrl (*p=0.013, t-test, t(18)=2.763, N = 10 for each group). The axon length is normalized by the number of labeled cells in VTA, the bouton density is normalized by the axon length, and both are expressed as a percentage of the group average in Ctrl mice. (g) Diagram showing procedures to determine the effect of adolescent dopamine neuron stimulation on mesofrontal circuit activity in Arc-/-;TH-Cre mice labeled with DREADD-Gq or mCherry-Ctrl. (h) Schematic showing the experimental setup to measure the mesofrontal circuit activity by VTA electrical stimulation and GCaMP6 imaging in the frontal cortex. (i) Time courses of cortical calcium signals in response to VTA stimulation in Arc-/-;mCherry-Ctrl and Arc-/-; DREADD-Gq mice. 1, 5, or 10 pulses of electrical stimuli (50 Hz) was delivered at 20 seconds after the start of imaging. The cortical calcium activity at each time point is represented by the change in image fluorescence relative to the baseline image fluorescence (ΔF/F). (5 pulse F(1,10)=6.0, *p=0.034, 10 pulse F(1,10)=9.5, *p=0.012, Two-way ANOVA, N=6 mice per group).

Adolescent dopamine neuron stimulation leads to restoration of coordinated frontal neuronal activity and cognition in adulthood.

(a) Diagram showing procedures for adolescent stimulation of dopamine neurons and functional imaging of frontal cortical neuron activity in adult Arc-/-;TH-Cre mice. (b) The average activity of individual frontal cortical neurons at binned positions relative to the center of Y-maze in Arc -/-; mCherry-Ctrl (1288 neurons from 7 mice) and Arc-/-; DREADD-Gq (1008 neurons from 7 mice) animals. Bottom, the proportion of neurons showing maximal activation at each maze position. This proportion reached a higher peak right before the maze center in Arc-/-; DREADD-Gq mice compared to Arc-/-; mCherry-Ctrl mice. Shaded areas indicate 95% confidence intervals. (c) The proportion of alternation-selective neurons reached a significantly higher peak right before the maze center in Arc-/-; DREADD-Gq compared to Arc-/-; mCherry-Ctrl mice. Shaded areas indicate 95% confidence intervals. (d) Y-maze alternation percentage for the animals used in the miniaturized microscope imaging experiments shows significant increase in Arc-/-;DREADD-Gq animals compared to Arc-/-; mCherry-Ctrl (*p=0.010, t-test, t-test, t(12)=3.043, N = 7 mice for each group, both groups passed Shapiro-Wilk normality test at alpha=0.05). (e) Total arm entries are not significantly different (p=0.399, t-test, t(12)=0.875). All the error bars indicate SEM.

Efficacy requirements for adolescent dopamine neuron stimulation.

(a) Top, diagram showing the base procedure for the stimulation of midbrain dopamine neurons and Y-maze testing in Arc-/-;TH-Cre mice labeled with DREADD-Gq or mCherry control viruses. Animals were injected with CNO (1 mg/kg) once per day for 3 days in adolescence (5 weeks old) and then tested in the Y-maze at adulthood (∼1 month later). Arc-/-; DREADD-Gq animals show significantly higher alternation compared to Arc-/-; mCherry-Ctrl at adulthood (*p=0.023, t-test, t(12)=2.598, N=7 mice for each group, both groups passed Shapiro-Wilk normality test at alpha=0.05). (b) The behavioral effect at the test interval of 1 day after CNO injection. Arc-/-; DREADD-Gq animals show significantly higher alternation compared to Arc-/-; mCherry-Ctrl (*p=0.031, t-test, t(20)=2.322, N=11 mice for each group, both groups passed Shapiro-Wilk normality test at alpha=0.05). (c) The behavioral outcome following a long duration stimulation starting in adolescence (2 times per day, 5 days per week, for 3 weeks). These animals did not show any improvement but a declining trend in performance (p=0.12, t-test, t(9)=1.727, Arc-/-; mCherry-Ctrl N=5, Arc-/-; DREADD-Gq, N=6 mice, both groups passed Shapiro-Wilk normality test at alpha=0.05). (d) The behavioral outcome following three days of CNO stimulation administered in adult mice (2-3 months). These animals did not show significant difference (p=0.97, t-test, t(11)=0.037, Arc-/-;mCherry-Ctrl, N=6; Arc-/-;DREADD-Gq, N=7 mice, both groups passed Shapiro-Wilk normality test at alpha=0.05).

Specific stimulation of adolescent frontal dopaminergic axons leads to reversal of both cognitive and psychomotor deficits.

(a) Schematic for labeling dopamine neurons with SSFO and two-photon imaging of frontal cortical activity with GCaMP6. (b) Example traces of spontaneous cortical activity averaged from the whole image frame in control EGFP only (blue) and SSFO (orange) labeled animals before and 30 min after frontal cortical blue light stimulation. (c) SSFO animals show significantly higher change in cortical activity (summarized by the standard deviation of spontaneous activity traces) compared to control EGFP animals after the light activation, suggesting light activation of SSFO expressing dopamine neurons (*p=0.001, t-test, t(8)=5.135, N=5 mice for each group). Cortical activity is summarized by the standard deviation (SD) of the spontaneous activity traces. Activity change is calculated as (SD2-SD1)/SD1, where SD1 is before and SD2 after treatment. (d) Diagram showing procedures for local light activation of frontal dopaminergic projections and Y-maze testing in Arc-/-;TH-Cre mice labeled with SSFO or Ctrl-GFP viruses. Light activation was delivered in adolescence (5 weeks old) once per day for 3 days. Animals were first tested in the Y-maze 1 day after the last light activation and then tested again in the Y-maze at adulthood, followed by an amphetamine induced locomotion test. (e-h) SSFO expressing animals show significantly higher alternation compared to control animals 1 day after light activation e, (*p=0.028, t-test, t(15)=2.440, N=9 EGFP, N=8 SSFO mice, both groups passed Shapiro-Wilk normality test at alph =0.05), with no difference in total entries f. These animals also show higher alternation at adulthood g, (*p=0.015, t-test, t(15)=2.748, EGFP N=9, SSFO N=8, both groups passed Shapiro-Wilk normality test at alpha=0.05) with no difference in total entries h. (i) In Arc-/-;TH-Cre mice that received adolescent frontal light stimulation, amphetamine induced locomotion is significantly reduced at adulthood in SSFO animals compared to GFP-control animals (F(1,14)=5.3, *p=0.037, Two-way ANOVA, N=8 mice per group). All the error bars indicate SEM.

Adolescent dopamine neuron stimulation reverses cognitive deficits in DISC1 mutant mice.

(a) Schematic showing the experimental setup to measure the mesofrontal circuit activity by VTA electrical stimulation and GCaMP6 imaging in the frontal cortex. (b) Time courses of cortical calcium signals in response to VTA stimulation in WT and DISC1+/- mice. 1, 5, or 10 pulses of electrical stimuli (50 Hz) was delivered at 20 seconds after the start of imaging. The cortical calcium activity at each time point is represented by the change in image fluorescence relative to the baseline image fluorescence (ΔF/F). (1 pulse F(1,10)=5.7, *p=0.038; 10 pulse F(1,10)=16.5, **p=0.002, Two-way ANOVA, N=6 per group). (c) DISC1+/- mice shows significant reduction in Y-maze alternation compared to WT (*p=0.044, t-test, t(18)=2.171, N=10 mice for each group, both groups passed Shapiro-Wilk normality test at alpha=0.05). (d) DISC1+/-;TH-Cre mice with DREADD-Gq labeled dopamine neurons show significantly higher alternation in the Y-maze at adulthood compared to mCherry-Ctrl labeled mice after 3-day adolescent CNO injections (*p=0.012, t-test, t(9)=3.139, DISC1+/-; mCherry N = 5, DISC1+/-;DREADD-Gq, N=6 mice, both groups passed Shapiro-Wilk normality test at alpha=0.05). All the error bars indicate SEM.