Prenatal exposure to maternal immune activation (MIA) and chronic adolescent cannabis use have both been identified as environmental risk factors for neuropsychiatric disorders. However, most individuals exposed to a single risk factor do not typically develop major mental illness, which suggests that multiple exposures may be required for illness onset. Here, we examine whether combined exposure to prenatal MIA and adolescent delta-9-tetrahydrocannabinol (THC), the main psychoactive component of cannabis, lead to enduring neuroanatomical and behavioural changes in adult offspring, potentially reflecting changes in humans indicative of mental illness.

Mice were prenatally exposed to a viral mimetic, poly I:C (5mg/kg), or vehicle at gestational day (GD)9, and then postnatally exposed to chronic THC (5mg/kg) or vehicle by intraperitoneal injections during adolescent development (postnatal day [PND]28-45). Longitudinal in vivo whole-brain magnetic resonance imaging (MRI) was performed pre-treatment, PND25, post-treatment, PND50, and in adulthood, PND85, followed by a series of behavioural tests aimed at assessing anxiety-like and locomotor, social, and sensorimotor gating behaviour. Post-mortem assessment of cannabinoid (CB)1 and 2 receptor expressing cells was performed in developmentally altered regions identified by MRI (anterior cingulate and somatosensory cortices, striatum, and hippocampus). We hypothesized that there would be differential, but synergistic effects of each exposure.

Briefly, we found subtle deviations in neurodevelopmental trajectory and subthreshold anxiety-like behaviours were observed in mice exposed to both risk factors. Sex-dependent effects were observed in patterns of shared brain-behaviour covariation, suggesting that exposure to MIA and THC may affect males and females in different ways. Density of CB1 and CB2 receptor positive cells was significantly decreased in all regions assessed for all mice exposed to either one or both risk factors, relative to controls.

These findings suggest that there may be a cumulative effect of risk factor exposure on gross neuroanatomical and behavioural development, and that the endocannabinoid system may be sensitive to both prenatal MIA, adolescent THC, or the combination. For full details, see our publication: .

In this dataset, you will find a total of 243 preprocessed structural MRIs (in MINC format) acquired at postnatal day ~25, ~50, and ~85 in mice exposed to poly I:C or vehicle control (0.9% sterile saline) at GD9, and then postnatally treated with vehicle or THC from PND 28-45. These are T1-weighted structural images at 100 micron isotropic resolution acquired on a 7 Tesla Bruker Biospec 70/30; matrix size of 180 x 160 x 90; 14.5 minutes, 20 degrees and TE/TR of 4.5/20 ms (2 averages, ~14 minutes). Anesthesia was induced with 3% isoflurane in oxygen and a (0.075 mg/kg bolus) dexmedetomidine injection. Anesthesia was maintained during the scan between 1.5-0.5% isoflurane, and a constant infusion of dexmedetomidine (0.05mg/kg/h continuous mg/kg during scan). T1-weighted scans were preprocessed by stripping native coordinates, flipping left-right to maintain fidelity, denoising, correcting inhomogeneities in the bias field using the N4 algorithm, and registering in LSQ6 alignment (i.e. 6 degrees of freedom are allowed for imagine alignment: translations and rotations along x, y, and z dimensions). The demographics information for each animal is included in the demographics.csv file.

Behavioural tests were performed following the postnatal day 85 scans in all animals with a 2 day rest period. These include: open field test, three chambered social approach, and prepulse inhibition. The data for all of these tests is presented in its own individual .csv spreadsheet.

Included in this data set are the structural MRIs in MINC format, the behavioural .csv data, and a readme.txt file providing further detail on the data structure and content, and on how to interpret the data column titles. DICOMS are also available for the structural MRI data, as are the raw (not-preprocessed) MINC files, available upon request to the authors.

Finally, the authors would like to acknowledge the funding bodies that supported the completion of this work including the Canadian Institute for Health Research, the Fonds de Recherche du Québec en Santé, and the Healthy Brains for Healthy Lives at McGill University.

Prenatal maternal immune activation (MIA) is a risk factor for neurodevelopmental disorders. How the gestational timing of MIA-exposure differentially impacts downstream development remains unclear. The data presented here includes longitudinal structural magnetic resonance imaging (MRI) data from weaning to adulthood, and behavioural testing in adolescence and adulthood on C57BL/6 mice exposed to MIA induced by the viral mimetic, polyinosinic:polycytidylic acid (poly I:C) either early (gestational day [GD]9) or late (GD17) in gestation.

The data published here was collected and analyzed for the following publication, where more details can be found (Guma et al., 2021 https://doi.org/10.1016/j.biopsych.2021.03.017). Briefly, we found that early MIA-exposure was associated with accelerated brain volume increases in adolescence/early-adulthood that normalized in later adulthood, in regions including the striatum, hippocampus, and cingulate cortex. Similarly, alterations in anxiety-like, stereotypic, and sensorimotor gating behaviours observed in adolescence normalized in adulthood. In contrast, MIA-exposure in late gestation had less impact on anatomical and behavioural profiles.

In addition to the univariate analyses described above, we also undertook a multivariate analysis (partial least squares) to relate imaging and behavioural variables for the time of greatest alteration, i.e. adolescence/early adulthood. We further explored the molecular underpinnings of region-specific alterations in early MIA-exposed mice in adolescence using RNA sequencing (data for differentially expressed genes in the anterior cingulate cortex, dorsal hippocampus, and ventral hippocampus are available via the original publication https://doi.org/10.1016/j.biopsych.2021.03.017 for a separate cohort of adolescent mice prenatally exposed to MIA or vehicle at GD9).

In this dataset, you will find a total of 376 preprocessed structural MRIs (in MINC format) acquired at postnatal day ~21, ~38, ~60, and ~90 in mice exposed to poly I:C or vehicle control (0.9% sterile saline) at GD9 or 17. These are T1-weighted, manganese enhanced (50mg/kg 24 hours pre-scan), structural images at 100 micron isotropic resolution acquired on a 7 Tesla Bruker Biospec 70/30; matrix size of 180 x 160 x 90; 14.5 minutes, 2 averages, using 5% isoflurane for induction, 1.5% for maintenance of anesthesia during the scan. T1-weighted scans were preprocessed by stripping native coordinates, flipping left-right to maintain fidelity, denoising, correcting inhomogeneities in the bias field using the N4 algorithm, and registering in LSQ6 alignment (i.e. 6 degrees of freedom are allowed for imagine alignment: translations and rotations along x, y, and z dimensions). The demographics information for each animal is included in the demographics.csv file.

Behavioural tests were performed following the postnatal day 38 and 90 scans in all animals with a 2 day rest period. These include: open field test, marble burying test, three chambered social approach, and prepulse inhibition. The attentional set shifting task was also performed following the final behavioural test in the postnatal day 90 wave of behaviours. The data for all of these tests is presented in its own individual .csv spreadsheet and includes data for both the timepoints evaluated.

Included in this data set are the structural MRIs in MINC format, the behavioural .csv data, and a readme.txt file providing further detail on the data structure and content, and on how to interpret the data column titles. DICOMS are also available for the structural MRI data, as are the raw (not-preprocessed) MINC files, available upon request to the authors.

Finally, the authors would like to acknowledge the funding bodies that supported the completion of this work including the Canadian Institute for Health Research, the Fonds de Recherche du Québec en Santé, and the Healthy Brains for Healthy Lives at McGill University.