Microglial cannabinoid receptor type 1 mediates social memory deficits in mice produced by adolescent THC exposure and 16p11.2 duplication

Abstract

Adolescent cannabis use increases the risk for cognitive impairments and psychiatric disorders. Cannabinoid receptor type 1 (Cnr1) is expressed not only in neurons and astrocytes, but also in microglia, which shape synaptic connections during adolescence. However, the role of microglia in mediating the adverse cognitive effects of delta-9-tetrahydrocannabinol (THC), the principal psychoactive constituent of cannabis, is not fully understood. Here, we report that in mice, adolescent THC exposure produces microglial apoptosis in the medial prefrontal cortex (mPFC), which was exacerbated in a model of 16p11.2 duplication, a representative copy number variation (CNV) risk factor for psychiatric disorders. These effects are mediated by microglial Cnr1, leading to reduction in the excitability of mPFC pyramidal-tract neurons and deficits in social memory in adulthood. Our findings suggest the microglial Cnr1 may contribute to adverse effect of cannabis exposure in genetically vulnerable individuals.

Fig. 1. Cnr1 expression in the microglia of the mouse brain.

a Experimental flow of Fluorescence-Activated Cell Sorting (FACS)-based microglia (CD45⁺CD11b⁺TMEM119⁺), astrocyte (ACSA-2⁺), and neuron (NeuN⁺) isolation using specific cell markers. b Relative mRNA expression levels (arbitrary units: a.u.) of Cnr1 in microglia, astrocytes, and neurons isolated from wild type mice were measured by quantitative real time PCR (qPCR) using the TaqMan assay protocol. n = 6 mice per group. ***p < 0.001, **p < 0.01 (p values are Microglia versus Astrocytes: p = 0.0010, Microglia versus Neurons: p < 0.0001, Astrocytes versus Neurons: p < 0.0001), determined by one-way ANOVA with post hoc Tukey test. c Relative mRNA expression levels (arbitrary units: a.u.) of Cnr1 in microglia, astrocytes, and neurons isolated from Cx3cr1^CreER/+;Cnr1^+/+ and Cx3cr1^CreER/+;Cnr1^flox/flox mice were measured by qPCR. n = 6 mice per group. ***p < 0.001 (p values are p < 0.0001), determined by unpaired two-tailed Student’s t test. d Microglia-enriched CD11b⁺ cells, ACSA-2⁺ astrocytes, and remaining cells including neurons were collected from the cerebral cortex of Cx3cr1^CreER/+;Cnr1^flox/flox mice and littermate controls (Cx3cr1^CreER/+;Cnr1^+/+) by magnetic activated cell sorting (MACS). For each cell type, expression of Cnr1, marker proteins (Iba1, GFAP, NeuN) and a loading control (GAPDH) in the total protein (arbitrary units: a.u.) were analyzed with SDS-PAGE followed by Western blotting with 10 μg of protein sample loaded in each well. n = 3 mice per group. *p < 0.05 (p values are Cnr1^+/+ Cd11b⁺ cells versus Cnr1^+/+ Astrocytes: p = 0.0495, Cnr1^+/+ Cd11b⁺ cells versus Cnr1^+/+ Remaining cells: p = 0.0204, Cnr1^+/+ Cd11b⁺ cells versus Cnr1^flox/flox Cd11b⁺ cells: p = 0.0479), determined by two-way ANOVA with post hoc Tukey test. b–d Each symbol represents one animal. Data are presented as the mean ± s.e.m.

Fig. 2. Microglial changes in the mPFC of 16p11dup male mice produced by adolescent THC treatment.

a Schematic diagram of the adolescent THC treatment protocol. 16p11dup mice and wild type littermate controls (WT) were treated with THC or vehicle (Veh) during adolescence, followed by microglial phenotyping at P51 upon completion of THC treatment. b (Top left) Immunohistochemistry of Iba1 (green) and P2ry12 (red) in the medial prefrontal cortex (mPFC) at P51. Scale bar, 50 μm. (Top right) The percentage of Iba1⁺P2ry12⁻ cells among all Iba1⁺ cells in the mPFC, ventral hippocampus (vHPC), dorsal hippocampus (dHPC), nucleus accumbens (NAc), and amygdala (Amy). (Bottom left) The number of Iba1⁺P2ry12⁺ cells and (bottom right) Iba1⁺P2ry12⁻ cells in these brain regions, presented as % of control (n = 6 slices in 3 mice per condition). c Microglial morphology analysis of individual Iba1⁺ cells in these brain regions. (Top) Representative tracing images (red) along with images of cellular processes (green) and cell bodies (yellow) of Iba1⁺ cells. Scale bar, 10 μm. (Bottom left) Quantification of cellular process area of Iba1⁺ cells (n = 50 cells in 5 mice per condition). (Bottom right) Quantification of the ratio of cell body size to total cell size of Iba1⁺ cells (n = 50 cells in 5 mice per condition). ***p < 0.001, *p < 0.05 (p values are (b) (Top right) all: p < 0.0001, (Bottom left) all: p < 0.0001, (Bottom right) all: p < 0.0001, (c) (Left) all: p < 0.0001, (Right) WT-Veh versus WT-THC: p < 0.0001, WT-Veh versus 16p11dup-Veh: p = 0.0229, WT-THC versus 16p11dup-THC: p < 0.0001, 16p11dup-Veh versus 16p11dup-THC: p < 0.0001), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one slice (b) and one animal (c). Data are presented as the mean ± s.e.m.

Fig. 3. THC-induced Cnr1-mediated microglial apoptosis is exacerbated by 16p11dup.

a Apoptosis assay of primary microglia cultures produced from WT and 16p11dup male mice. Quantification of signal intensity (arbitrary units: a.u.) of apoptosis marker apopxin (n = 12 fields in 3 mice per condition). b Necrosis assay of primary microglia cultures produced from WT and 16p11dup male mice. Quantification of signal intensity (arbitrary units: a.u.) of necrosis marker 7-AAD (n = 12 fields in 3 mice per condition). c Quantification of cellular process area of phalloidin-stained microglia cultures produced from WT and 16p11dup male mice (n = 12 fields in 3 mice per condition). d Quantification of cellular process number of phalloidin-stained microglia cultures produced from WT and 16p11dup male mice (n = 12 fields in 3 mice per condition). e Representative images of microglia cell cultures produced from WT and 16p11dup male mice in apoptosis and necrosis assays. Apopxin (green) and 7-AAD (red) are shown. Scale bar, 50 μm. f Immunohistochemistry with antibody against phalloidin (green) of primary microglia cultures produced from WT and 16p11dup male mice. Scale bar, 25 μm. g Apoptosis assay of primary microglia cultures produced from genetic deletion of Cnr1 (Cnr1 KO) and genetic deletion of Cnr2 (Cnr2 KO) male mice. Quantification of signal intensity (arbitrary units: a.u.) of apopxin (n = 12 fields in 3 mice per condition). h Necrosis assay of primary microglia cultures produced from Cnr1 KO and Cnr2 KO male mice. Quantification of signal intensity (arbitrary units: a.u.) of 7-AAD (n = 12 fields in 3 mice per condition). i Quantification of cellular process area of microglia cultures produced from Cnr1 KO and Cnr2 KO male mice. (n = 12 fields in 3 mice per condition). j Quantification of cellular process number of microglia cultures of Cnr1 KO and Cnr2 KO male mice (n = 12 fields in 3 mice per condition). k Representative images of microglia cell cultures produced from Cnr1 KO and Cnr2 KO male mice in apoptosis and necrosis assays. Apopxin (green) and 7-AAD (red) are shown. Scale bar, 100 μm. l Immunohistochemistry with antibody against phalloidin (green) of primary microglia cultures of Cnr1 KO and Cnr2 KO male mice. Scale bar, 25 μm. m Representative images of immunohistochemistry of Iba1 (green) and Casp3-p17 (red) (top left) as well as TUNEL signals (red) and DAPI (blue) (bottom left) in the mPFC at P51. Scale bar, 50 μm. Quantification of signal intensity (arbitrary units: a.u.) of Casp3-p17 and TUNEL (right) (n = 6 mice per group). ***p < 0.001, **p < 0.01, *p < 0.05 (p values are (a) all: p < 0.0001, (c) all: p < 0.0001, (d) WT-vehicle (Veh) versus WT-THC: p < 0.0001, WT-THC versus 16p11dup-THC: p = 0.0410, 16p11dup-Veh versus 16p11dup-THC: p < 0.0001, (m) (Left) WT-Veh versus WT-THC: p = 0.0230, WT-Veh versus 16p11dup-Veh: p = 0.0471, WT-THC versus 16p11dup-THC: p = 0.0249, 16p11dup-Veh versus 16p11dup-THC: p = 0.0119, (Right) WT-Veh versus WT-THC: p < 0.0001, WT-Veh versus 16p11dup-Veh: p = 0.0376, WT-THC versus 16p11dup-THC: p < 0.0001, 16p11dup-Veh versus 16p11dup-THC: p < 0.0001), determined by two-way ANOVA with post hoc Tukey test. ***p < 0.001 (p values are (g) p < 0.0001, (i) p < 0.0001, (j) p < 0.0001), determined by unpaired two-tailed Student’s t test. Each symbol represents one field (a–d, g–j) and one animal (m). Data are presented as the mean ± s.e.m.

Fig. 4. Up-regulation of p53 signaling pathway identified by mPFC microglia-specific transcriptome profiling.

a Experimental flow of Fluorescence-Activated Cell Sorting (FACS)-based mPFC microglia isolation by using 3 markers (CD45/CD11b/P2ry12) at P51, followed by RNA sequencing (RNA-seq). b Significantly enriched pathways of genes associated (two-tailed uncorrected p < 0.05) with the 16p11dup with THC condition compared to control, in the Reactome Pathway analysis. c Functional gene interaction network from the upstream regulators/Ingenuity Pathway Analysis (IPA) for the genes associated with the 16p11dup with THC treatment condition showing predicted inhibition (blue), activation (orange), and unknown directionality (gray). d Relative mRNA expression level of Trp53 in microglia isolated from mPFC at P51. n = 6 mice per condition. e Apoptosis (left) and necrosis (right) assays using primary microglia cultures treated by THC or vehicle (Veh) with and without pifithrin-α. (Left) Quantification of signal intensity (arbitrary units: a.u.) of apopxin. n = 6 fields in 3 mice per condition. (Right) Quantification of signal intensity (arbitrary units: a.u.) of 7-AAD. n = 6 fields in 3 mice per condition. d, e ***p < 0.001, **p < 0.01, *p < 0.05 (p values are (d) WT- Veh versus WT-THC: p = 0.0270, p values are WT-Veh versus 16p11dup-Veh: p = 0.0465, WT-THC versus 16p11dup-THC: p = 0.0037, 16p11dup-Veh versus 16p11dup-THC: p = 0.0021, (e) all: p < 0.0001), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one animal (d) and one field (e). Data are presented as the mean ± s.e.m.

Fig. 5. Social novelty recognition and memory deficits synergistically produced by adolescent THC treatment and 16p11dup.

a Schematic diagram of the adolescent THC treatment protocol. b (Left) Schematic diagram of the three-chamber social interaction test for sociability. (Middle) Sociability phenotypes in WT and 16p11dup mice with adolescent THC or vehicle (Veh) treatment as indicated by discrimination index ([Stranger 1 − Inanimate mice sniffing time]/[Stranger 1 + Inanimate mice sniffing time] × 100 (%)). (Right) Representative heat maps depict movements of the WT-Veh mice vs. 16p11dup-THC mice. c (Left) Schematic diagram of the three-chamber social interaction test for social novelty preference. (Middle) Preference of social novelty in WT and 16p11dup mice with adolescent THC or Veh treatment as indicated by discrimination index ([Stranger 2 − Stranger 1 sniffing time]/[Stranger 2 + Stranger 1 sniffing time] × 100 (%)). (Right) Representative heat maps depict movements of the WT-Veh vs. 16p11dup-THC mice. d (Left) Schematic diagram of the 5-trial social memory test. (Right) Interaction time of WT and 16p11dup mice receiving adolescent THC or Veh treatment with an ovariectomized female mouse. b–d WT-Veh (n = 17 mice), WT-THC (n = 15 mice), 16p11dup-Veh (n = 11 mice), and 16p11dup-THC (n = 11 mice). e (Left) Preference of novel object in novel object recognition test (NORT) in WT and 16p11dup mice with adolescent THC or Veh treatment. (Right) Preference of novel place in novel place recognition test (NPRT) in WT and 16p11dup mice with adolescent THC or Veh treatment. e WT-Veh (n = 8 mice), WT-THC (n = 8 mice), 16p11dup-Veh (n = 11 mice), 16p11dup-THC (n = 7 mice). ***p < 0.001, **p < 0.01, *p < 0.05 (c p values are WT-Veh versus 16p11dup-Veh: p < 0.0001, WT-THC versus 16p11dup-THC: p < 0.0001, 16p11dup-Veh versus 16p11dup-THC: p = 0.0074, d THC × 16p11dup interaction for Trial 3 [F_1,50 = 11.98, p = 0.0011], THC × 16p11dup interaction for Trial 4 [F_1,50 = 9.226, p = 0.0038], THC × 16p11dup interaction for Trial 5 [F_1,50 = 4.095, p = 0.0484]), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one animal. Data are presented as the mean ± s.e.m.

Fig. 6. Synergistic reduction in intrinsic excitability in PT neurons produced by adolescent THC and 16p11dup.

a Schematic diagram of the adolescent THC treatment protocol. b (Left) Schematic structure of multiple layers of the mPFC. Pyramidal tract (PT, red) and intra-telencephalic (IT, green) neurons in layer 5 of the mPFC are depicted. (Right) Representative voltage traces elicited by applying −100 pA current steps. These traces show typical hyperpolarizing responses of PT and IT neurons. c (Left) Representative voltage traces recorded from PT neurons in response to current step injections. (Right) The intrinsic excitability assessed by measurement of input resistance (left), rheobase (middle), and spike frequency (right). WT-Veh (n = 11 cells in 6 mice), WT-THC (n = 11 cells in 7 mice), 16p11dup-Veh (n = 11 cells in 6 mice), and 16p11dup-THC (n = 10 cells in 5 mice). d (Left) Representative voltage traces recorded from IT neurons in response to current step injections. (Right) The intrinsic excitability assessed by measurement of input resistance (left), rheobase (middle), and spike frequency (right). WT-Veh (n = 21 cells in 8 mice), WT-THC (n = 13 cells in 7 mice), 16p11dup-Veh (n = 10 cells in 6 mice), and 16p11dup-THC (n = 17 cells in 7 mice). c ***p < 0.001, **p < 0.01, *p < 0.05 (p values are (Left) WT-Veh versus 16p11dup-Veh: p = 0.0395, WT-THC versus 16p11dup-THC: p = 0.0394, 16p11dup-Veh versus 16p11dup-THC: p = 0.0406, (Right) WT-Veh versus 16p11dup-Veh: p = 0.0463, WT-THC versus 16p11dup-THC: p = 0.0004, 16p11dup-Veh versus 16p11dup-THC: p = 0.0368), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one cell. Data are presented as the mean ± s.e.m.

Fig. 7. Cnr1 deletion in the microglia normalizes microglial abnormalities produced by adolescent THC treatment and 16p11dup.

a Schematic diagram of the experimental design. b Immunohistochemical analysis of Iba1 (green) in the mPFC at P51. (Left) Representative images of the mPFC, representative tracing images (red), and images of cellular processes (green) and cell bodies (yellow) of Iba1⁺ cells. Scale bar, 50 μm (left) and 10 μm (middle and right). c The number of Iba1⁺P2ry12⁺ cells (left) and Iba1⁺P2ry12⁻ cells (middle) in the mPFC, presented as % of control. (Right) The percentage of Iba1⁺P2ry12⁻ cells among all Iba1⁺ cells in the mPFC. (n = 6 slices in 3 mice per condition). d Quantification of the ratio of cellular process area (left) and cell body size to total cell size (right) of Iba1⁺ cells. (n = 50 cells in 5 mice per condition). e Representative images of immunohistochemistry of Iba1 (green) and Casp3-p17 (red) (top left) as well as TUNEL signals (red) and DAPI (blue) (bottom left) in the mPFC at P51. Scale bar, 50 μm. f Quantification of signal intensity of Casp3-p17 in Iba1⁺ cells (left) and TUNEL (right). n = 6 slices in 3 mice per condition. c, d, f ***p < 0.001, **p < 0.01, *p < 0.05 (p values are (c) (Left) 16p11^wt;Cnr1^+/+ versus 16p11^wt;Cnr1^flox/flox: p = 0.0281, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p = 0.0001, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0221, 16p11^dup; Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (Middle) 16p11^wt;Cnr1^+/+ versus 16p11^wt;Cnr1^flox/flox: p = 0.0285, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p < 0.0001, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0337, 16p11^dup; Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (Right) all: p < 0.0001, (d) (Left) 16p11^wt;Cnr1^+/+ versus 16p11^wt;Cnr1^flox/flox: p < 0.0001, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p = 0.0017, 16p11^wt; Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (Right) 16p11^wt;Cnr1^+/+ versus 16p11^wt;Cnr1^flox/flox: p = 0.0225, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p = 0.0183, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0330, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (f) (Left) 16p11^wt;Cnr1^+/+ versus 16p11^wt;Cnr1^flox/flox: p = 0.0016, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p < 0.0001, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0032, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (Right) all: p < 0.0001), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one slice (a, f) and one animal (d). Data are presented as the mean ± s.e.m.

Fig. 8. Cnr1 deletion in the microglia normalizes deficits in PT neurons and social memory that are synergistically produced by adolescent THC treatment and 16p11dup.

a Schematic diagram of the experimental design. b (Left) Representative voltage traces recorded from PT neurons in response to current step injections. (Right) The intrinsic excitability of PT neurons, as quantified by input resistance (left), rheobase (middle), and spike frequency (right). 16p11^wt;Cnr1^+/+ (n = 9 cells in 2 mice), 16p11^wt;Cnr1^flox/flox (n = 5 cells in 3 mice), 16p11^dup;Cnr1^+/+ (n = 6 cells in 2 mice), and 16p11^dup;Cnr1^flox/flox (n = 8 cells in 3 mice). ***p < 0.001, **p < 0.01, *p < 0.05 (p values are (Left) 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p = 0.0042, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0001, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.0001, (Right) 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p = 0.0016, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0396, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p < 0.000 determined by two-way ANOVA with post hoc Tukey test. c (Left) Representative voltage traces recorded from IT neurons in response to current step injections. (Right) The intrinsic excitability of IT neurons, as quantified by input resistance (left), rheobase (middle), and spike frequency (right). 16p11^wt;Cnr1^+/+ (n = 4 cells in 2 mice), 16p11^wt;Cnr1^flox/flox (n = 4 cells in 2 mice), 16p11^dup;Cnr1^+/+ (n = 6 cells in 2 mice), and 16p11^dup;Cnr1^flox/flox (n = 5 cells in 2 mice). d Sociability phenotypes (left) and preference of social novelty (right) as indicated by the discrimination index in the three-chamber social interaction test. e, Time spent with an ovariectomized female mouse in the 5-trial social memory test. d, e 16p11^wt;Cnr1^+/+ (n = 6 mice), 16p11^wt;Cnr1^flox/flox (n = 5 mice), 16p11^dup;Cnr1^+/+ (n = 6 mice), 16p11^dup;Cnr1^flox/flox (n = 6 mice). d, e ***p < 0.001, **p < 0.01 (p values are (d) 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^+/+: p < 0.0001, 16p11^wt;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0004, 16p11^dup;Cnr1^+/+ versus 16p11^dup;Cnr1^flox/flox: p = 0.0026, (e) Cnr1^flox/flox x 16p11dup interaction for Trial 4 [F_1,19 = 11.30, p = 0.0033], Cnr1^flox/flox x 16p11dup interaction for Trial 5 [F_1,19 = 4.760, p = 0.0419]), determined by two-way ANOVA with post hoc Tukey test. Each symbol represents one cell (b, c) and one animal (d, e). Data are presented as the mean ± s.e.m.