Concurrent maternal stress and THC exposure during pregnancy alters adolescent behavioral outcomes and corticolimbic molecular programs
aDepartment of Neurobiology, University of Maryland School of Medicine, Baltimore, Maryland
bDepartment of Psychiatry, University of Maryland School of Medicine, Baltimore, Maryland
cDepartment of Biological Sciences, University of Southern California, Los Angeles, California
dDepartment of Anatomy and Cell biology, University of Illinois, College of Medicine, Chicago, Illinois
eDepartment of Obstetrics, Gynecology and Reproductive Sciences, University of Maryland School of Medicine, Baltimore, Maryland
*Corresponding author: Mary Kay Lobo (mklobo@som.umaryland.edu), 20 Penn Street, HSF 2, 21201, MDAbstract
Cannabis use during pregnancy is increasing, often to alleviate stress and anxiety, yet the long-term effect of prenatal cannabis exposure alone or in combination with psychosocial stress on offspring neurodevelopment or maternal behaviors remains unclear. Here, we developed a translational rodent model combining prenatal Δ⁹-tetrahydrocannabinol (THC) exposure with chronic psychosocial stress using the maternal witness defeat stress (MWDS) paradigm. Pregnant C57BL/6 mice were exposed to MWDS from gestational day (GD) 3-12 and received daily subcutaneous THC (2 mg/kg) or vehicle until birth. All exposure groups showed impaired maternal behavior, with negative postnatal outcomes and caregiving, with additive effects observed in the combined exposure group. In adolescence, male and female offspring exhibited exposure-specific behavioral alterations. Prenatal stress and combined exposures led to increased anxiety-like behavior and reduced motivated behavior in both sexes, while THC alone primarily impacted female self-care and social behavior. Transcriptomic profiling of the prefrontal cortex (PFC) and nucleus accumbens (NAc) of adolescent offspring revealed sex- and region-specific gene expression changes across all exposure groups. Prenatal THC-, stress-, and combined exposures each altered distinct molecular pathways related to mitochondrial function, synaptic organization, and glial signaling. Comparative analysis with a perinatal fentanyl model revealed shared transcriptional substrates involved in synaptic signaling and circadian regulation. These findings indicate that THC and stress independently and additively impair maternal behaviors with lasting neurodevelopment signatures in offspring.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
INTRODUCTION
The increasing legalization of cannabis across the United States has led to a surge in its use among vulnerable populations, including pregnant people, many of whom cite its anxiolytic and stress-relieving effects as their reasons for continued use during pregnancy (1,2). While cannabis may confer short-term anxiolytic effects, its long-term consequences on maternal health, maternal-infant bonding, and offspring neurodevelopment are not well understood. The placental barrier serves as a critical checkpoint, protecting the fetus from maternal glucocorticoids that may ensue from acute stressors (3,4). Chronic maternal stress can compromise placental protection, increasing fetal exposure and altering inflammatory responses in the placenta and fetal brain (5–7). Additionally, prenatal stress (PNS) exposures can disrupt the fetus’s endocannabinoid (eCB) system, crucial in shaping synaptic plasticity, stress regulation, and emotional behaviors (8–10). As both PNS and prenatal cannabis exposure (PCE) independently alter brain development, their combination may cause cumulative or synergistic neurodevelopmental disruptions in fetal brain development, particularly in brain regions involved in emotional regulation, motivation, and reward processing.
The prefrontal cortex (PFC) and nucleus accumbens (NAc) are central to cognition, emotional regulation, and reward-driven behaviors, and are both modulated by the eCB system (11,12). Studies have demonstrated that early-life stress and cannabinoid exposure result in measurable neuroanatomical changes, including reduced cortical thickness, synaptic loss, and dendritic atrophy in the PFC, alongside decreased dopamine receptor expression and sensitivity in the NAc (13–16). While such alterations may not manifest as overt behavioral symptoms immediately after birth, they may predispose offspring to psychiatric disorders later in life (17–19). Therefore, there is a dire need to identify molecular substrates that could provide insights into mechanisms of endocannabinoid actions to inform novel therapeutics.
In the present study, we investigated whether additive exposure to Δ9 tetrahydrocannabinol (THC), the main psychoactive component of cannabis, modulates maternal behavioral responses to psychosocial stress and whether these combined exposures disrupt offspring development. To model maternal stress, we employed the chronic witness defeat stress (CWDS) paradigm. CWDS is a translationally relevant paradigm used to induce psychological stress (20,21) in rodents without direct physical harm. This model mimics vicarious trauma relevant to aspects of chronic stress in pregnant people (22,23). Using this approach, pregnant dams were subjected to a 10-day CWDS from gestation onset, which we term maternal witness defeat stress (MWDS), while receiving moderate daily doses of THC via subcutaneous injections (24). This experimental design allowed us to model the co-occurrence of prenatal stress and cannabis exposure in a controlled manner and assess their cumulative effects on maternal behavior pre-birth, maternal care after birth, and offspring behaviors and neurobiological outcomes. Immediately following cessation of MWDS, dams were assessed for anxiety-like behavior to evaluate the emotional impact of combined prenatal stress and THC exposure. To uncover the molecular substrates underlying these phenotypes, we performed transcriptomic profiling of the PFC and NAc. Finally, we compared these data to published transcriptomic analysis of perinatal fentanyl exposed (PFE) mice, shown to exhibit sex-specific affective disturbances (25). Our study provides mechanistic insight into how prenatal environments program long-term vulnerability to emotional and motivational dysfunction.
METHODS AND MATERIALS
Experimental Subjects
All experiments were performed in accordance with the Institutional Animal Care and Use Committee Guidelines at the University of Maryland School of Medicine (UMSOM) and in accordance with NIH guidelines for the use of laboratory animals. Mice were given food and water ad libitum and housed in the UMSOM vivarium on a 12-h light/dark cycle. Experimental mice were 8-9 weeks old male chronic social defeat stress (CSDS) and pregnant females (MWDS) C57BL/6 mice. CD-1 retired breeder males (>4 months) were used as the aggressors for CSDS/MWDS. Mice were purchased from Charles River Laboratories and acclimated in the vivarium for at least one week. Dams were randomly assigned to four groups across three cohorts (6–8 dams each), with data combined for analysis (Fig. 1B-I). The groups included: (I) dams receiving vehicle injections without stress (vehicle-exposed; n = 20), (II) dams receiving THC injections without stress (THC-exposed; n = 21), (III) dams receiving vehicle injections following MWDS (stress-exposed; n = 22), and (IV) dams receiving THC injections following MWDS (THC/stress-exposed n = 22). Adolescent behaviors included mice from no more than 3 pups/litters/sex/condition.
Behavioral Testing
Maternal witness defeat stress
The chronic witness defeat stress (CWDS) paradigm (20) was adapted for use in pregnant dams and referred to herein as MWDS. Pregnant dams were housed on one side of a perforated divided cage to witness a 10-minute resident-intruder interaction. The CSDS mouse was then housed opposite the resident, and the MWDS dam was removed and housed opposite to a novel CD-1 aggressor. After 24 h of sensory interaction, the CSDS mouse was defeated by a new CD-1 resident while a different MWDS dam witnessed the agonistic interaction. This process was repeated daily for 10 days with new CD-1–CSDS–MWDS pairings to prevent habituation. Unstressed vehicle and THC dams were pair-housed across perforated dividers in cages containing woodchip bedding for 10days.
Open field test
Mouse was placed in an acrylic open-field arena (43 × 43 × 43 cm) virtually divided into a peripheral zone and a central zone (15 × 15 cm) for 15 minutes. Total time spent in the central zone (26), recorded using ANY-maze 7.4 video tracking software, was used as an index of anxiety-like behavior, with reduced center time indicating increased anxiety. Total distance traveled within the arena was measured as an indicator of locomotor activity.
Elevated plus maze test
This consisted of a plus-shaped acrylic apparatus elevated 94 cm above the floor, with two open arms (30 × 5 cm) and two closed arms (30 × 5 × 16 cm) connected by a central platform (5 × 5 cm). Mouse was placed in the center of the apparatus facing an open arm, and behavior was recorded for 5 minutes (26). Total arm entries and time spent in and exploring the open arms were tracked using ANY-maze 7.4 video tracking software. Decreased time spent in the open arms was interpreted as heightened anxiety-like behavior.
Splash test
The dorsal coat of mice was sprayed three times with a 10% (v/v) sucrose solution and immediately placed in an empty 4-liter PYREX glass cylinder (15 cm diameter). Behavior was recorded for 5 minutes using a CCTV camera (Model -WV-CP294), and total grooming duration was manually scored by a blinded experimenter using Kinoscope open-source software (27).
Pup retrieval test
Testing was conducted in a quiet behavior room during the morning light cycle. Prior to testing, dams in their home cages were habituated to the testing environment for 5 minutes. Before the start of the test, the total number of live pups per dam, including any litter mortalities recorded from the day of birth (P0), was documented. Following habituation, dams were temporarily removed from the home cage, during which pups were placed individually at designated positions approximately 15 cm away from the nest site. The dam was then returned to the cage, and behavior was recorded for 10 minutes (28). Latency to retrieve first pup, and all pups to the nest was manually scored by a blinded experimenter. Pups not retrieved within the 10-minute observation period were assigned a maximum latency of 600 seconds.
RNA isolation
At P35, behavior-naive adolescent mice were weighed prior to tissue collection. A minimum of four mice per group and sex from mixed litters were used (Vehicle: Females n = 5, Males n = 6; prenatal THC: Females n = 5, Males n = 8; prenatal Stress: Females n = 5, Males n = 6; prenatal THC/Stress: Females n = 7, Males n = 4). Tissue punches of PFC and NAc were collected between 10:00 AM and 12:00 PM as previously described (33) and stored at −80 °C until ready for RNA extraction. Total RNA was isolated via Trizol (Invitrogen) homogenization and chloroform phase separation, followed by purification using the RNeasy Mini Kit (#74104, Qiagen) with a DNase step (#79254, Qiagen). RNA concentration and purity were assessed using a Nanodrop spectrophotometer (ThermoScientific).
Nanostring
RNA quality was assessed using the RNA 6000 Nano assay on an Agilent 2100 Bioanalyzer (Agilent Technologies, Santa Clara, CA) (36), and RNA integrity numbers (RIN) were generated using the 2100 Expert Software. Only samples with RIN > 8 were used for downstream transcriptomic analysis. Gene expression was quantified using the Nanostring nCounter Analysis System (Bruker, Seattle, WA) (30) with a custom-designed codeset targeting 337 genes. For each sample, 100 ng of total RNA was hybridized to reporter and capture probes according to the manufacturer’s protocol. Samples were processed in batches of 12 using the Nanostring Prep Station, and digital counts were acquired on the nCounter Digital Analyzer. Raw expression data were normalized and analyzed using nSolver Analysis Software (Nanostring Technologies).
Prenatal THC exposure
THC was prepared by dissolving 100 mg/ml THC in ethanol, followed by emulsification in 2% Tween-80, sonication, and dilution in physiological saline (24). All dams received daily s.c. injections of THC (2 mg/kg, 2 ml/kg) or vehicle from GD3 to birth.
Perinatal fentanyl exposure (PFE)
Detailed experimental description of the PFE paradigm has been discussed elsewhere (25, 31). In brief, pregnant dams were administered 10ug/ml fentanyl citrate in drinking water from GD3 through birth till weaning at postnatal day 21.Following weaning, juvenile pups were grouped housed by sex until ready for transcriptome profiling at adolescent age (postnatal day 35).
Statistics
All statistical analyses were performed using GraphPad Prism (version 10.4.1). Behavioral data from both dams and adolescent offspring were analyzed using two-way analysis of variance (ANOVA) with MWDS/no MWDS and vehicle/THC exposure as between-subject factors. For adolescent behavioral assessments, males and females were analyzed separately to examine sex-specific effects. Where significant interactions or main effects were detected, Dunnett’s post hoc tests were used to compare each experimental group to the vehicle-treated control group. For all analyses, statistical significance was set at p < 0.05. Data are reported as mean ± SEM, with exact F- and p-values presented in the Results. Sample sizes are indicated in figure legends, and assumptions of normality and homogeneity of variance were confirmed prior to analysis. For gene expression analysis, normalized RLF files were analyzed using nSolver and GraphPad Prism (v10). Gene annotations done using Genecard resource, and ontology analysis was performed using Metascape (32) and ChEA3 (33).
RESULTS
Prenatal stress, THC, and combined exposures induce sex-specific deficits in adolescent anxiety-like and motivated behaviors
To examine long-term behavioral consequences of prenatal exposures, adolescent offspring (P35–P45) were assessed for anxiety-like, and effortful motivated behaviors. In the SI test (Fig. 2A,A’), both male and female offspring exposed to prenatal stress or combined THC/stress displayed reduced social interaction, while prenatal THC exposure alone had no significant effect. In the OFT (Fig. 2B,B’), all male exposure groups showed reduced center zone exploration, indicative of increased anxiety-like behaviors. In females, this anxiogenic effect was limited to the prenatal stress and THC/stress groups, while THC alone had no effect. Importantly, total locomotion remained comparable across groups and sexes (Fig. 2C,C’). On the EPM (Fig. 2D,D’), both sexes exhibited reduced open arm exploration following combined prenatal THC/stress exposure. Males were particularly sensitive to prenatal stress exposures alone, while females showed heightened anxiety only following THC or combined exposures. The splash test (Fig. 2E,E’) revealed reduced grooming behavior, a proxy for self-care motivation in both sexes following prenatal THC or THC/stress exposures. Prenatal stress exposure alone also impaired grooming in females but not males. In the Y-maze barrier task, a measure of effort-based reward motivation, prenatal stress and THC exposures reduced HR arm selection under increasing effort demands. In the 10 cm barrier (Fig. 2F,F’), both sexes exhibited reduced HR-arm preference following combined THC/stress exposure, while prenatal stress only males and THC only females also showed decreased effortful motivation for sucrose reward. In the 15 cm barrier (Fig. 2G, G’), HR-arm selection reduced across groups, especially in males.
Prenatal THC and stress exposure drive brain region and sex specific transcriptional reprogramming in adolescent offspring
To uncover molecular correlates of behavioral changes, we conducted gene expression profiling using the Nanostring single molecular detection (nCounter) assay on PFC and NAc punches from adolescent male and female mice. The assay targeted 337 curated genes (Suppl. data 1) involved in synaptic function, neurodevelopment, mitochondrial metabolism, stress response, and immune signaling, previously found to be implicated perinatal fentanyl exposure (31). Principal component analysis (PCA) was first applied to reduce dimensionality and assess the contribution of sample characteristics to transcriptome variability and indicated that brain region was the primary driver of transcriptomic variance (Fig. 3A), with no distinct clustering by treatment or sex indicating minimal influence on sample variance. Treatment-driven effects were thus segregated by brain region. Broadly, prenatal stress induced upregulation of transcripts in the PFC and downregulation in the NAc (Fig. 3B). Volcano plots confirmed that differentially expressed transcripts (DETs) were primarily upregulated in the PFC with fewer downregulated DETs across all treatment groups (Fig. 3C, D). Venn diagrams show that seventeen and twelve DETs were shared across all three conditions in the PFC and NAc respectively (Fig. 3E, E’). Although sex did not emerge as a major contributor to global transcriptomic variance, we found sex-specific DETs not detected in combined-sex datasets (Fig. 3F, F’), underscoring distinct transcriptional signatures in males and females. Gene Ontology (GO) analysis showed that in the PFC (Fig. 3G), upregulated genes in the prenatal stress group were enriched for precursor metabolite and energy generation and transmembrane transporter complexes, indicating alterations in cellular metabolism and transport. In prenatal THC exposed mice, upregulated genes were enriched for synapse organization, excitatory synapse components, and semaphorin receptor activity, suggesting enhanced synaptic signaling and axon guidance processes. Combined THC/stress exposure led to upregulation of genes enriched for ribonucleoside triphosphate metabolism, respiratory chain complexes, and transmembrane transporter activity, reflecting increased bioenergetic demand. In contrast, downregulated genes in this group were enriched for cell–cell adhesion, indicating potential disruptions in structural connectivity. In the NAc (Fig. 3H), prenatal stress exposure resulted in upregulation of genes associated with metabolite and energy production and transcription factor binding, while downregulated genes were enriched for gliogenesis, cell adhesion, and integrin binding, suggesting suppression of glial and structural pathways. Prenatal THC exposure upregulated genes involved in metabolite generation, cation channel complexes, and ion channel activity, with downregulated genes enriched for gliogenesis, respiratory chain components, and integrin binding. In the Prenatal THC/stress group, upregulated genes were enriched for startle response, dendritic spine development, and glutamate receptor binding, while downregulated genes were associated with cellular respiration, plasma membrane signaling receptor complexes, and integrin binding, indicating altered synaptic excitability and cellular energetics.
To explore upstream regulatory mechanisms, we identified all significantly expressed transcription factors (TFs) across exposures and brain regions in a sex-wise manner using GeneCard annotations (Figs. 4, 5). In the PFC, the number of significant TFs in males vs. females were prenatal stress (1 vs. 8), prenatal THC (3 vs. 7), and prenatal THC/stress (4 vs. 6) (Fig. 4A). We compared our transcription factor list to ChEA3-predicted regulators (33), a tool developed based on RNA-seq and ChIP-seq datasets, to validate enrichment and support the functional relevance of TFs identified in our dataset. This revealed that in the PFC, upregulated DETs in prenatal stress-exposed males were enriched for regulation of secretion, with Olig1 as the predicted TF (Fig. 4B). In females, enriched processes included anatomical structure maturation and regulation of neurogenesis, predicted to be regulated by Myrf and Sox10 (Fig. 4C). Sox10 was identified to be significantly upregulated in females in our dataset (Fig. 4A). In THC-exposed mice, DETs were enriched for cation transport in males with top TF Hivep2 (Fig. 4D). Upregulated DETs in females were enriched in GABAergic neuron differentiation with top TF Foxg1,(Fig. 4E). Downregulated DETs in THC-exposed males were associated with cell-cell adhesion, with Sox8 as the top TF (Fig. 4F). In THC/stress-exposed mice, upregulated genes were enriched in aerobic respiration with top predicted TF Alx1 (Fig. 4G) while female upregulated DETs were enriched for epithelial cell proliferation with Mterf2 as the predicted TF (Fig. 4H). Downregulated genes in THC/stress-exposed females were enriched for semaphorin-plexin signaling, with Egr1 identified as the top regulator (Fig. 4I). Notably, Egr3, a related TF, was identified to be significantly downregulated in our dataset (Figs. 4A, I). In the NAc, the number of significant TFs in males vs. females were prenatal stress (5 vs. 4), prenatal THC (3 vs. 8), and prenatal THC/stress (3 vs. 2) (Fig. 5A). In stress-exposed females, upregulated DETs were enriched for memory formation with top predicted TF Znf484 (Fig. 5B). In males, prenatal stress exposure downregulated genes enriched for myelin sheath development and predicted to be regulated by Znf654 (Fig. 5C). In prenatal THC-exposed females, upregulated genes were associated with development and predicted to be regulated by Lcor (Fig. 5D). Downregulated genes in THC-exposed males were enriched in NAD metabolic process with Znf654 as the top predicted TF (Fig. 5E) and in females, synaptic signaling process was top enrichment for downregulated DETs with Nfiz as predicted TF (Fig. 5F). In THC/stress-exposed mice, upregulated genes in males were enriched for cell-cell adhesion with Ahdc1 as top predicted TF (Fig. 5G), and females show enrichment of DETs involved in response to growth factor with predicted TF Myrf (Fig. 5H). Downregulated genes in THC/stress-exposed females showed enrichment of cell-cell adhesion, with Bcl6b as the predicted TF (Fig. 5I).
Cross-model transcriptomic convergence reveals common molecular signatures of early-life adversity
Developmental exposures to addictive substances or early life stress independently disrupts reward processing and increases vulnerability to negative affect. We build on our previous findings that low-dose of PFE induces persistent affective deficits and mesolimbic circuit disruption (25). We investigated whether shared behavioral outcomes across prenatal THC, stress and fentanyl exposures reflect convergence at the transcriptomic level. Using our curated list of 337 genes, we compared gene expression patterns in the PFC and NAc across exposures and sex (Fig. 6). In the PFC, THC- and stress-exposed females showed general transcript upregulation relative to PFE females and all male groups (Fig. 6A). In males, 5 DETs Cspg5, Nr1d1, Ptger3, S1pr1 and Syn3 were upregulated across all exposure groups (Fig. 6B, C). Four DETs overlapped between PFE and stress-exposed males including Cd200, Mtfr1l, Mzt1, and downregulated Cldn5, while Shank1 was shared between PFE and THC-exposed males. Among females, four DETs, Arhgef1, Atp5g2, Mtfr1l and Syn3 were shared across all exposures (Fig. 6D, E). Six DETs were shared between PFE and stress-exposed females, Bin1, C1qc, Foxo1, Nr1d1, RhoA, Tmed7 and Plxna3 was the only gene shared between PFE and THC in females. Notably, Syn3 was upregulated across all groups and sexes in the PFC. Functional enrichment of shared DETs showed postsynaptic density and secretion regulation in males, and cell differentiation along with extrinsic component of membrane in females (Fig. 6F). In the NAc, a predominant downregulation of gene profile across all groups was observed (Fig. 6G). Four DETs Cldn5, Hsp90, Lhx2 and Nr1d1 were shared across all male exposures with Nr1d1 consistently upregulated across exposures (Fig. 6H, I). Nine DETs were shared between PFE and prenatal stress-exposed males, with Mzt1, P2ry12, Tprkb and Zfp748 showing opposing expression profile. No shared DETs were found between PFE and prenatal THC exposures. In females, 11 DETs were shared across all three groups, including six downregulated genes and four genes, Atp5k, Dnm1l, Gpbp1, Ndufv2 upregulated in THC and stress groups. Gria1 was uniquely upregulated by prenatal stress exposure but downregulated in prenatal THC and PFE. Ten DETs were shared between PFE and prenatal stress exposure with 5 DETs consistently downregulated in both exposures and 5 DETs upregulated only in prenatal stress exposure. Twelve DETs overlapped between PFE and THC with 11 DETs downregulated in both conditions, and 1 DET, Atp5j2 upregulated only in prenatal THC exposure (Fig. 6J, K). Enrichment analysis of shared DETs reveal negative regulation of cell differentiation and regulation of protein localization process males, and ATP synthesis, gliogenesis and regulation of microglial migration in females (Fig. 6L).
DISCUSSION
The rising use of cannabis during pregnancy, often motivated by its perceived anxiolytic properties, raises serious concerns about long-term neurodevelopmental consequences in the offspring (34,35). Chronic psychosocial stress during gestation is a well-established disruptor of fetal brain development, acting either through direct alterations to neurodevelopmental trajectories (36,37) or by epigenetically priming neural circuits for heightened vulnerability to future stressors (38,39). Although findings on the effects of prenatal cannabis exposure are mixed (40,41), the critical role of the endocannabinoid (eCB) system in early brain development (42) suggests that cannabis use during gestation, particularly under conditions of chronic stress, may synergistically impair neurodevelopmental outcomes. Here, we established a novel rodent model combining prenatal THC exposure with chronic psychological stress to examine their additive effects on maternal behavior, adolescent behavioral outcomes, and transcriptomic signatures in the PFC and NAc. We employed the MWDS paradigm to model vicarious trauma, a form of psychosocial stress with translational relevance to human experiences of chronic threat or violence (20,43,44). Daily subcutaneous administration of 2 mg/kg THC from gestational day 3 to birth modeled moderate, clinically relevant exposure (45,46) while avoiding cannabinoid tetrad effects (47). Co-exposure exacerbated maternal caregiving deficits, as reflected by reduced litter size, increased pup mortality, and delayed pup retrieval. These findings challenge assumptions of THC’s anxiolytic benefit during pregnancy and suggest that combined THC and stress exposures disrupt maternal regulation.
Adolescent offspring exposed to combined prenatal THC and chronic stress exhibited robust impairments in anxiety-related, and motivated behaviors. Both sexes showed heightened anxiety-like phenotypes, consistent with studies indicating that prenatal THC sensitizes the HPA axis and increases emotional reactivity (48,49). Similarly, gestational stress alone disrupts glucocorticoid signaling and impairs brain development of emotional related circuitry (50,51), and our findings suggest that concurrent THC exposure exacerbates these effects. Social interaction deficits in both sexes align with evidence that prenatal cannabinoid exposure impairs sociability, likely via reduced endocannabinoid signaling and oxytocin function (52–54). These behavioral abnormalities are characteristic of early-life adversity and have been linked to persistent synaptic dysregulation in corticolimbic circuits (55–57). These findings support a “double-hit” model in which prenatal stress and cannabinoid exposure interact to potentiate long-term neurobehavioral dysfunction. Female mice exhibited more pronounced deficits in disrupted grooming behavior, consistent with human data showing a higher prevalence of depression among women (58,59). In contrast, male mice displayed more severe anxiety-like behaviors, a pattern that diverges from epidemiological trends. This discrepancy may reflect sex differences in symptom presentation or healthcare utilization, with males potentially underreporting or avoiding treatment for anxiety, thereby skewing population-level data (60).
To investigate molecular mechanisms underlying the observed behavioral deficits, we performed targeted transcriptomic profiling of 337 genes previously implicated in PFE (31), focusing on the PFC and NAc, regions central to executive function, emotion regulation, and reward processing (61,62). In males, combined THC/stress exposure upregulated Mtfr1l and Atp5g2 in the PFC, suggesting disrupted mitochondrial fission and ATP synthesis. This aligns with reports that cannabinoids impair mitochondrial dynamics and neuronal energy balance (63,64). Enrichment of transcription factors Olig1 and Alx1, involved in oligodendrocyte development and neurogenesis (65,66), may reflect heightened allostatic load impairing myelination during adolescence. In females, PFC upregulation of Fgf9, Pax6, and Jam2 suggests altered neurogenesis, cell fate, and BBB integrity (67–69). Downregulation of Egr3, a key regulator of synaptic plasticity and stress adaptation (70), alongside enrichment of semaphorin-plexin signaling, suggests impaired circuit refinement contributing to affective and motivational deficits (71). In male NAc, broad downregulation of genes related to cell adhesion and metabolism, including Cldn5, Gria1, Eno3, and Mdh1, was observed. Reduced Cldn5 and Gria1 implicate BBB dysfunction and impaired glutamatergic signaling, respectively (72,73), while Eno3 and Mdh1 downregulation signals compromised energy metabolism (74). Upregulated Arhgef1 and enriched Ahdc1 implicate disrupted cytoskeletal dynamics and neurodevelopmental vulnerability (75,76). In females, upregulation of Ascl1 and Fgf13, genes regulating neurogenesis and sodium channel function (77,78) suggests altered excitability and differentiation of medium spiny neurons. Like males, Eno3 and Mdh1 downregulation in females highlights convergent mitochondrial dysfunction. Collectively, these findings underscore the importance of sex-specific investigations and support targeting mitochondrial and synaptic pathways for intervention. Public health messaging around prenatal cannabis use, particularly under stress, should consider these additive developmental risks.
Cross-model comparisons revealed convergent upregulation of Syn3, a key regulator of dopaminergic vesicle cycling (79,80), across perinatal exposures and sexes in the PFC, consistent with disrupted reward sensitivity and cognition in early-life adversity (81). In males, shared PFC upregulation of Cspg5, Nr1d1, Ptger3, and S1pr1 suggests convergence on synaptic remodeling and circadian/metabolic pathways (82–84). Female PFC share upregulation of Arhgef1, Atp5g2, Mtfr1l, and Tmed7, reflecting structural and bioenergetic adaptations (85,86). Plxna3, a semaphorin receptor implicated in axonal guidance, was selectively upregulated in THC- and PFE-exposed females (87). In the NAc, male offspring shows shared downregulation of Cldn5, Hsp90, and Lhx2, with Nr1d1 uniquely upregulated across all exposures. Reduced Hsp90 and Lhx2 suggest impaired glucocorticoid signaling and neurogenesis, respectively (88,89), consistent with reduced plasticity and reward motivation. In females, a broader set of 11 genes were shared across models, but with divergent regulation. Notably, Atp5k, Dnm1l, Gpbp1, and Ndufv2 were downregulated in PFE but upregulated in THC and stress-exposed females, suggesting stimulus-specific metabolic reprogramming. Dnm1l and Gpbp1 regulate mitochondrial fission and redox balance, respectively (90,91), implicating mitochondrial pathways as a sex-specific vulnerability node. GO enrichment confirmed sex-biased patterns: males showed enrichment in postsynaptic density and secretion regulation, while females exhibited enrichment in ATP synthesis, gliogenesis, and microglial migration (92), suggesting distinct compensatory strategies to prenatal insults. These findings emphasize that while both sexes exhibit shared molecular adaptations across diverse prenatal exposures, the directionality, cellular processes, and potential compensatory capacity differ markedly.
Our findings underscore the need for further translational research to explore the complex interplay between prenatal cannabis exposure and maternal stress given that many pregnant people report using cannabis for its perceived anxiolytic effects and stress relief.
Supporting information
ACKNOWLEDGMENTS AND DISCLOSURES
This work was supported by the Matthew Osborne Foundation and the Kahlert Institute for Addiction Medicine [to JO]; Grant No. R01MH106500, R01DA054905, R33DA052101 [to MKL].
Conceptualization, JO, JC, KSM, MKL; Methodology, JO, MD, AK, MKL; Formal Analysis, JO, GK, DF; Investigation, JO, MD, MKL; Resources, JC, MKL; Writing-Original Draft, JO; Writing-Review & Editing, JO, JC, KSM, MKL; Visualization, JO; Supervision, JO, MKL; Funding Acquisition, JO, MKL.
The authors report no biomedical financial interests or potential conflicts of interest.