www.nature.com/npp ARTICLE OPEN Long-term hippocampal interneuronopathy drives sex-dimorphic spatial memory impairment induced by prenatal THC exposure Adán de Salas-Quiroga1,2, Daniel García-Rincón1,2, Daniel Gómez-Domínguez3, Manuel Valero3, Samuel Simón-Sánchez1,2, Juan Paraíso-Luna1,2, José Aguareles1,2, Mitona Pujadas4, Carolina Muguruza5, Luis F. Callado 5, Beat Lutz6, Manuel Guzmán1,2, Liset Menéndez de la Prida 3 and Ismael Galve-Roperh 1,2 Prenatal exposure to Δ9-tetrahydrocannabinol (THC), the most prominent active constituent of cannabis, alters neurodevelopmental plasticity with a long-term functional impact on adult offspring. Specifically, THC affects the development of pyramidal neurons and GABAergic interneurons via cannabinoid CB1 receptors (CB1R). However, the particular contribution of these two neuronal lineages to the behavioral alterations and functional deficits induced by THC is still unclear. Here, by using conditional CB1R knockout mice, we investigated the neurodevelopmental consequences of prenatal THC exposure in adulthood, as well as their potential sex differences. Adult mice that had been exposed to THC during embryonic development showed altered hippocampal oscillations, brain hyperexcitability, and spatial memory impairment. Remarkably, we found a clear sexual dimorphism in these effects, with males 1234567890();,: being selectively affected. At the neuronal level, we found a striking interneuronopathy of CCK-containing interneurons in the hippocampus, which was restricted to male progeny. This THC-induced CCK-interneuron reduction was not evident in mice lacking CB1R selectively in GABAergic interneurons, thus pointing to a cell-autonomous THC action. In vivo electrophysiological recordings of hippocampal LFPs revealed alterations in hippocampal oscillations confined to the stratum pyramidale of CA1 in male offspring. In addition, sharp-wave ripples, a major high-frequency oscillation crucial for learning and memory consolidation, were also altered, pointing to aberrant circuitries caused by persistent reduction of CCK+ basket cells. Taken together, these findings provide a mechanistic explanation for the long-term interneuronopathy responsible for the sex-dimorphic cognitive impairment induced by prenatal THC. Neuropsychopharmacology (2020) 45:877–886; https://doi.org/10.1038/s41386-020-0621-3 INTRODUCTION contribute to the cognitive deficits observed in adulthood [7, 8]. Exposure of the immature brain to cannabinoids exerts deleter- The consequences of cannabinoid exposure during the highly ious functional consequences in adulthood by interfering with vulnerable adolescence period have been the subject of intense neural cell development, as well as with neuronal differentiation research [9]. In adolescence, THC induces a subcortical hyperdo- and connectivity [1]. Within the more than 100 active compounds paminergic state and adaptations in the prefrontal cortex, and in the cannabis plant, Δ9-tetrahydrocannabinol (THC) is the most these plasticity changes are associated to decreased social relevant molecule considering both its high abundance and its interaction, higher anxiety, and altered sensorimotor gating [10]. high potency to engage cannabinoid CB1 receptors (CB1R). The In contrast, in the adult brain, THC blunts dopaminergic function consequences of prenatal THC exposure are intrinsically different and is associated with negative emotionality and addiction from administration in adulthood, because, during the develop- severity [11]. Similarly, maternal cannabis consumption results in mental window, the pattern of CB1R receptor expression diverges a selective impairment of D2R-dependent (but not D1R-depen- from that in the mature nervous system, which allows to control dent) dopaminergic neurotransmission in the developing meso- various neural progenitor cell functions [2, 3]. Embryonic THC limbic system [6, 12, 13]. exposure impairs N-methyl-D-aspartate receptor-dependent long- THC interferes with the delicate neurodevelopmental role of the term depression [4] and modulates dopamine release and endocannabinoid system [2, 3]. CB1R, upon engagement by receptor availability [5, 6]. Perinatal THC exposure also induces endocannabinoids (2-arachidonoylglycerol and anandamide), reg- changes in glutamatergic and noradrenergic signaling that may ulates crucial steps of cortical development, including neural 1 Department of Biochemistry and Molecular Biology, Instituto Ramón y Cajal de Investigación Sanitaria (IRYCIS), Instituto Universitario de Investigación Neuroquímica (IUIN), Complutense University, 28040 Madrid, Spain; 2Centro de Investigación Biomédica en Red sobre Enfermedades Neurodegenerativas (CIBERNED), 28049 Madrid, Spain; 3Instituto Cajal, CSIC, Avda Dr Arce 37, 28002 Madrid, Spain; 4Integrative Pharmacology and Systems Neuroscience Research Group, Neurosciences Research Program, Hospital del Mar Medical Research Institute, Barcelona, Spain; 5Department of Pharmacology, University of the Basque Country UPV/EHU and Centro de Investigación Biomédica en Red de Salud Mental (CIBERSAM), Leioa, Spain and 6Institute of Physiological Chemistry, University Medical Center of the Johannes Gutenberg University Mainz, Mainz, Germany Correspondence: Adán de Salas-Quiroga (a.salas@ucm.es) or Ismael Galve-Roperh (igr@quim.ucm.es) These authors contributed equally: Adán de Salas-Quiroga, Daniel García-Rincón Received: 7 August 2019 Revised: 13 January 2020 Accepted: 13 January 2020 Published online: 26 January 2020 © The Author(s) 2020 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 878 progenitor proliferation, neuronal differentiation, and neuronal around the setup, and mounting/dismounting the head during migration. Hence, pharmacological or genetic manipulation of brief periods of time. After 2–3 days, animals were water-deprived CB1R function in the embryonic brain results in long-lasting and they started daily sessions of 10 min running for reward alterations. Specifically, prenatal cannabinoid signaling regulates (water), till they were fully habituated. We increased the time in interneuron migration and morphogenesis [14], and administra- the setup over steps of 5 min, till animals got habituated to tion of the CB1R/CB2R-mixed agonist WIN-55,212-2 selectively behave freely in the cylinder. They typically alternated periods of reduces hippocampal cholecystokinin (CCK) basket cell (BC) running and immobility during a maximum of 60 min each day. neurotransmission, without altering pyramidal neurons [15]. Upon Water port was removed after the first 10 min session to keep prenatal exposure to WIN-55,212-2, mice exhibit reduced deprivation level. Once mice were habituated to a 30–60 min depolarization-induced suppression of inhibition and feedforward session, they were anesthetized to open the craniotomy. After- inhibition, as well as altered social interaction. However, embryo- wards, craniotomy was covered with low toxicity silicone nic THC administration can also interfere with pyramidal neuron elastomer (Kwik-SilTM, World Precision Instruments) and record- development, such as deep-layer (Vb) corticospinal motor neurons ings started the day after. Every animal was recorded in two [16]. These changes are associated to an unbalanced excitation/ independent sessions of 1 h. inhibition state, thus resulting in a higher susceptibility to seizures. For recordings, we used 16-channel silicon probes consisting in Studies based on conditional CB1R genetic rescue from a CB1R-null a linear (100 µm resolution, 413 µm2 electrode area, Neuronexus). background have demonstrated a cell-autonomous action of THC Extracellular signals were pre-amplified (4x gain) and recorded in dorsal telencephalic principal neurons associated to skilled with a 16-channel AC amplifier (Multichannel Systems), further motor impairment, while a proconvulsive THC-induced status amplified by 100, filtered by analog means at 1 Hz to 5 kHz, and relies, at least in part, on alterations of both pyramidal neurons sampled at 20 kHz/channel with 12 bits precision. The animal and GABAergic interneurons [16]. Similarly, prenatal administra- speed was stored to evaluate periods of running and immobility. tion of WIN-55,212-2 exerts alterations in pyramidal neuron Analysis of electrophysiological signals was implemented in excitability of the prefrontal cortex in adult male rats [17]. In the MATLAB 9.3 (MathWorks). Local-field potential (LFP) signals from present study, evaluation of the neurodevelopmental impact of sites at stratum lacunosum moleculare (slm) were used for prenatal cannabinoid exposure revealed a striking sex-dependent identifying θ periods during running (bandpass 4–12 Hz). impact on hippocampal plasticity, with males being selectively Forward–backward zero-phase finite impulse response filters of affected. Hence, in the context of the progressive worldwide order 512 were used to preserve temporal relationships between legalization of cannabis use for both medical and recreational channels and signals. The mean power spectra during θ was fitted purposes [18], this study strongly supports that the functional to the 1/f decay for frequencies >60 Hz and a reference level was consequences of prenatal THC exposure should be assessed established at 0 dB. Spectral values fitted to 1/f were similar considering the existence of sex-dependent differences in both between groups and were discarded for the analysis. Conse- intensity and nature of THC-induced alterations. quently, only the θ (4–12 Hz) and low γ bands (30–60 Hz) were included in the analysis. For detection of sharp-wave ripple (SWR), LFP signals from MATERIALS AND METHODS stratum radiatum (sr) were low-pass filtered (100 Hz), whereas Detailed methods for immunofluorescence and confocal micro- signals from stratum pyramidale (sp) were bandpass filtered scopy, in situ hybridization (ISH), immunohistochemistry (IHC), (100–600 Hz). Filtered signals were smoothed by a Gaussian stereological analysis, pentylenetetrazol (PTZ)-induced seizures kernel and candidate events were detected by thresholding (>4 assay, novel object recognition (NOR) and object location (OL) SDs). The power spectra was evaluated in a window of ±0.2 ms tasks, cannabinoid-induced analgesia and hypothermia, THC around each detected event. Time–frequency analysis was measurements, and WIN-55,212-2-stimulated [35S]GTPγS-binding performed by applying the multi-taper spectral estimation in assays are provided as Supplementary Materials and Methods. sliding windows with 97.7% overlap and frequency resolution of 10 Hz in the 90–600 Hz frequency range. The normalized power in Animals the 90–600 Hz band was treated as a statistical distribution, as Experimental designs and procedures were approved by the previously described [20]. Slow (90–120 Hz) and fast ripples (>120 Complutense University Animal Research Committee in accor- Hz) were separated using individual spectra for visualization dance with the European Commission regulations. The generation purposes. and genotyping of Nex-CB1−/−, Dlx5/6-CB1−/−, and control CB1f/f littermates has been reported elsewhere [19]. Neuronal population Data and statistical analyses changes and CB1R status were analyzed at different time points Results shown represent the means ± SEM and the number of (P20 and P90) after electrophysiological characterization and experiments is indicated in every case. Statistical analysis was behavioral determinations performed at P60. performed with GraphPad Prism 6.07 (GraphPad Software, La Jolla, CA, USA). All variables were first tested for normality In vivo electrophysiological recordings (Kolmogorov–Smirnov) and homoscedasticity (Levene’s). When Mice were implanted with chamber/fixation bars under isoflurane variables satisfied these conditions, two-way analysis of variance anesthesia (1.5–2% mixed in oxygen 400–800 ml/min). The and Fisher’s least significance difference post hoc test were used recording chamber was aligned to target the right dorsal to assess differences between groups. P-values < 0.05 were hippocampus at −1.8 mm posterior from the bregma and 1.25 regarded as statistically significant. mm lateral from midline. At least two jeweller’s screws were inserted into the skull for providing additional anchoring and reference/ground connections (over the cerebellum). An intracer- RESULTS ebellar silver wire was used as the main reference/ground Embryonic THC exposure evokes long-term interneuron connection. The implant was secured with dental cement. Animals alterations in a sex-dimorphic manner were recovered from anesthesia and were returned to To assess the impact of prenatal THC administration throughout home cages. interneuron embryonic development, we defined a temporal After surgery, mice were habituated to the head-fixed setup window of drug delivery at 3 mg/kg (intraperitoneally) from E10.5 consisting on a cylinder (20 cm radius) coupled to a stereotactic to E17.5. At P20, analysis of CB1R immunoreactivity in the dorsal frame. Habituation sessions included handling, running freely hippocampus of wild-type mice prenatally exposed to THC Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 879 Fig. 1 Prenatal THC exposure induces a sex-dimorphic long-lasting hippocampal CB1R adaptation and CCK+ basket cell interneurono- pathy. a, b Representative images and quantification of CB1R immunoreactivity measured in the CA1 hippocampal region of P20 mice prenatally exposed to THC (3 mg/kg) or its vehicle. c, d CB1R immunoreactivity quantified in the stratum pyramidale of CA1 in the same mice. e, f Representative images of CCK in situ hybridization and GABA immunohistochemistry in the hippocampus, and quantification of interneurons identified as CCK/GABA double-positive cells. n = at least 7 (a) or 6 (d, f) animals per group. *p < 0.05 vs. corresponding vehicle; **p < 0.01 vs. corresponding vehicle. Scale bars: a 50 µm; c 10 µm; e left 150 µm, center 25 µm, right 5 µm. showed a male-specific decrease of CB1R levels with respect to THC during embryonic development exhibit an overt interneur- vehicle-treated mice (Fig. 1a, b and Supplementary Fig. S1a, b). onopathy with a remarkable sexual dimorphism. To rule out the Interestingly, no significant differences were found for sex or possibility that sexual differences emerge from pharmacokinetic treatment in other areas as the prefrontal or somatosensory cortex factors, such as THC bioavailability, we measured THC levels in (Supplementary Fig. S1c, d). Of note, the vast majority of E17.5-brain tissue (i.e., 12 h after the last THC injection) and found immunoreactivity that is detected under these experimental similar values in both sexes (Supplementary Fig. S1f). In addition, conditions corresponds to CB1R+ terminals of GABAergic inter- as protracted embryonic THC exposure leads to a transient neurons [16, 21]. Hence, we decided to focus on CA1 CCK+ BCs, downregulation of CB1R [16], we asked whether a sex-dependent the population of hippocampal interneurons that expresses the CB1R downregulation might be responsible for the observed highest levels of CB1R in the adult mouse forebrain [22]. effects. [35S]GTPƳS-binding analysis in E17.5-brain tissue revealed Perisomatic CB1R immunoreactivity at CA1 sp was reduced in an evident downregulation of CB1R after embryonic THC exposure, the THC-exposed male progeny compared with its vehicle-treated but comparable curves were obtained for both sexes (Supple- counterpart (Fig. 1c, d). No differences were found for females mentary Fig. S1g). treated with either THC or its vehicle. Next, we evaluated whether these alterations are due to a Prenatal THC exposure alters main hippocampal oscillations in previously described morphogenetic role of CB1R in developing adult males interneurons [23] or to a decrease in the total number of CCK+ CCK+ BCs are essential for proper hippocampal physiology and interneurons. To address this question, we performed ISH with a function [24–26]. Hence, we decided to further characterize the riboprobe against CCK, combined with anti-GABA IHC to unequi- aforementioned THC-induced deficits via intrahippocampal vocally identify CCK-containing hippocampal interneurons, given recordings. Specifically, we performed in vivo electrophysiological the broad expression of CCK by CA1 pyramidal neurons (Fig. 1e recordings of the LFP throughout CA1 in the hippocampus of and Supplementary Fig. S1e). Stereological analysis showed a head-fixed P60 adult mice (Fig. 2a). During running, θ (4–12 Hz) selective reduction of CA1 CCK+ interneuron density in THC- and γ oscillations (30–60 Hz) typical of exploratory behavior were treated males as compared either with vehicle-treated males or recorded across layers (Fig. 2b). Spectral analyses of the LFP signal with females irrespective of the treatment (Fig. 1f). Taken together, revealed changes in the θ band that were confined to the sp of these findings support that those mice that had been exposed to CA1 (Fig. 2c, d and Supplementary Fig. S2a, b). THC treatment Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 880 Fig. 2 Prenatal THC exposure induces sex-dimorphic alterations of hippocampal oscillations and spatial cognitive deficits. a Scheme of experimental setup. P60-P90 mice were recorded head-fixed while running freely on a wheel. b Laminar recordings of LFP signals from a representative male mouse from each group (vehicle and THC). The running speed is shown at the bottom. so, stratum oriens; sp, stratum pyramidale; sr, stratum radiatum; slm, stratum lacunosum moleculare. c Mean power spectra of running episodes show representative peaks at θ and γ recorded at the pyramidal cell layer (sp). n = 15 vehicle-treated males, 15 THC-treated males, 9 vehicle-treated females, 4 THC- treated females. d Group differences of θ power were specific for male THC mice. *p < 0.05. Same dataset as above. e No differences of γ power. f SWR events were recorded during immobility. Note higher event rate in THC-treated males (arrows). g Distribution of spectral peaks per SWR events in the different groups. Note higher number of events in males THC and a larger contribution of slow ripple events. h Group differences of SWR rate, **p < 0.01. i Consistent larger SWR rate across detection thresholds (in number of SD). j Examples of slow (80–110 Hz) and fast (110–200 Hz) SRW events. k The proportion of slow SWR was significantly higher in the male THC group. Same dataset as above. *p < 0.05. l Mean group values of discrimination index in the NOR task. n = 6 vehicle-treated males, 7 THC-treated males, 3 vehicle-treated females, 5 THC-treated females. m Mean group values of discrimination index in the OL task. Same dataset as NOR task. n Latency to PTZ-induced seizures was significantly lower in the male THC group. *p < 0.05 n = 16 vehicle-treated males, 13 THC-treated males, 9 vehicle-treated females, 14 THC-treated females. Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 881 tended to reduce the γ power at sp consistently in both males and interneurons in CB1f/f males prenatally exposed to THC compared females (Fig. 2e). These data align with the idea of a with vehicle-treated littermates (Fig. 3d), whereas CB1f/f females complementary role of CCK BCs in hippocampal θ and γ exhibited comparable densities irrespective of treatment (Fig. 3e). oscillations [25], and the impact of THC administration in network Likewise, similar data were obtained in Glu-CB1-KO mice, whereas oscillations and psychiatric traits [27]. GABA-CB1-KO mice appeared refractory to THC impact in both Next, we analyzed SWRs (Fig. 2f), a major hippocampal high- males and females (Fig. 3d, e). Remarkably, conditional deletion frequency oscillation recorded during immobility and sleep that is of CB1R in the GABAergic lineage per se led to a decrease in the crucial for learning and memory consolidation [28]. SWR emerge density of CCK+ hippocampal interneurons, only reaching from the oscillatory firing of pyramidal cell ensembles, controlled statistical significance in the male population (Fig. 3d). These by the interplay between excitation and BC-mediated inhibition findings demonstrate the involvement of CB1R located on [29]. In diseases affecting the hippocampal formation such as hippocampal GABAergic interneurons as the main target for the temporal lobe epilepsy and Alzheimer’s disease, SWR are sex-dimorphic impact of embryonic THC exposure. pathological and associated to severe cognitive deficits and epileptiform activities [30, 31]. Spectral analysis of SWR in mice Cell population-specific sex dimorphism of CB1R signaling is treated with THC confirmed some differences with a clear sex critical for adult hippocampal function selectivity (Fig. 2g). First, the SWR rate was significantly higher To unequivocally identify the neuronal population responsible for in males exposed to THC as compared with females or the prenatal THC-induced cognitive impairment, we performed the vehicle-treated males, independently of the detection threshold NOR and OL tests in conditional CB1R-deficient mice. Prenatal (Fig. 2g–i). Second, the typical bimodality of the spectral SWR THC administration did not induce significant differences in peaks [32] was remarked in THC-treated male mice (Fig. 2g), with the NOR task by sex or treatment in CB1f/f and Glu-CB1-KO mice many more slower (<120 Hz) than fast events (>120 Hz) (Fig. 2j). (Fig. 4a, b and Supplementary Fig. S3d–i). In contrast, we found a This reflects a larger proportion of high-γ SWR in THC-exposed robust spatial memory impairment in THC-treated CB1f/f and Glu- males as compared with the other groups (Fig. 2k and Supple- CB1-KO male mice in the OL task (Fig. 4c and Supplementary mentary Fig. S2f). No significant changes were found in other Fig. S3j–o). THC exerted no effect in GABA-CB1-KO males, which measures of spectral distribution such as entropy and fast ripple exhibited worse spatial memory as compared with vehicle-treated index (Supplementary Fig. S2c–e). Taken together, these observa- CB1f/f and Glu-CB1-KO males (Fig. 4c). Surprisingly, a different tions provide functional evidence for aberrant hippocampal mechanism seems to operate in females, which exhibited an microcircuit function caused by a persistent reduction of CCK+ interaction between genotype and treatment (Fig. 4d), hence interneurons and oscillatory defects, which predict cognitive suggesting that additional CB1R neuron populations may con- alterations in males prenatally exposed to THC. To test this idea tribute to spatial cognition. Interestingly, vehicle-treated GABA-CB1- further, we conducted NOR and OL memory tests. The former KO male mice presented impairment of conceptual memory, which evaluates non-spatial, conceptual learning of object identity, was not evident upon embryonic THC administration (Fig. 4a, b). In which depends on multiple brain regions, while the latter provides this regard, NOR performance is known to be affected in GABA-CB1- a measure of spatial learning, which strongly relies on hippocam- KO mice [34]. Together, these data support a requirement of CB1R pal function [33]. We did not observe any interference of prenatal signaling in developing GABAergic interneurons for the adequate THC in the NOR paradigm, but THC-treated males performed maturation of CCK-dependent hippocampal function in male mice, poorly in the OL test as compared with vehicle-treated counter- and point to remarkable sex-dimorphic actions of CB1R signaling parts (Fig. 2l–m and Supplementary Fig. S3a–c), consistent with along development. To further assess the specific involvement of the aforementioned electrophysiological data. In addition, we CB1R located on GABAergic interneurons in the developmental analyzed susceptibility to PTZ-induced seizures and found consequences of THC exposure, we analyzed the impact of THC on selective effects in the male offspring (Fig. 2n). Altogether, these two behavioral traits that depend mainly on CB1R located on data support deleterious sex-dimorphic effects of prenatal THC principal neurons. Thus, no genotype or sex interaction was exposure caused by a CCK interneuronopathy. observed in THC-induced analgesia, a trait that relies on principal neurons located outside the neocortex [19] (Fig. 5a, b). Moreover, CB1R located on GABAergic neurons is responsible for the sex- cannabinoid-induced hypothermia, which is mediated mainly by dimorphic interneuronopathy induced by embryonic THC CB1R on dorsal telencephalic glutamatergic neurons [19], did not exposure reveal sex or treatment interaction in Glu-CB1-KO mice, but showed To understand in further detail the developmental mechanisms treatment interaction and no sex dimorphism in GABA-CB1-KO leading to CCK interneuronopathy in adulthood, we sought for a animals (Fig. 5c, d). direct link between THC action and specific cellular targets. CB1R is expressed in various neuronal lineages throughout development [3]. To unequivocally assess the involvement of a specific subset of DISCUSSION CB1R in the abovementioned defects, we made use of conditional The demonstration of a functional role of endocannabinoid knockout mice lacking CB1R exclusively in dorsal telencephalic signaling in modulating crucial neurodevelopmental processes glutamatergic pyramidal cells (CB1floxed/floxed;Nex-Cre/+ mice; herein [2, 3, 35] has contributed to the understanding of the referred to as Glu-CB1-KO) or in forebrain GABAergic neurons consequences of prenatal cannabinoid exposure in developing (CB1floxed/floxed;Dlx5/6-Cre/+ mice; herein referred to as GABA-CB1-KO) neuronal circuits [12]. Here we demonstrate that prenatal THC [21]. Immunofluorescence analysis of perisomatic CB1R+ BC exposure exerts a sex-dependent interference with CCK BC synapses at CA1 sp revealed a selective long-term decrease in development, thereby leading to long-term interneuronopathy, CB1f/f males exposed to THC, which was preserved in Glu-CB1-KO altered hippocampal function, and impaired spatial cognition. animals, pointing to an involvement of CB1R located on Several studies have investigated the consequences of prenatal developing GABAergic interneurons in mediating THC actions exposure to cannabinoids on GABAergic interneuron develop- (Fig. 3a–c). In contrast, the remnant hippocampal CB1R immunor- ment. Our results are in partial agreement with previous findings eactivity of GABA-CB1-KO mice exhibited no differences by showing that prenatal WIN-55,212-2 administration interferes with sex or treatment (Fig. 3a–c). In addition, we carried out anti- the development of CCK BCs, which in turn impacts feedforward CCK ISH combined with anti-GABA IHC to label CCK+ and feedback inhibition [15]. This effect was associated with hippocampal interneurons in every genotype, sex, and treatment. reduced social interaction, but not with anxiety. Likewise, WIN- Stereological analysis confirmed a persistent reduction of CCK+ 55,212-2 administration during early and mid, but not at late, Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 882 Fig. 3 Sex-specific interneuronopathy elicited by prenatal THC exposure requires CB1R located on GABAergic neurons. a Representative images of CB1R immunofluorescence in the sp of the CA1 hippocampal region of Glu-CB1-KO, GABA-CB1-KO and CB1f/f P90 mice prenatally exposed to THC or its vehicle. b, c Quantification of CB1R immunoreactivity in the area shown above both in males and females. d, e Quantification of CCK+ interneurons in the same animal groups segregated in males and females. n = at least 5 animals per group. **p < 0.05 vs. corresponding vehicle; ***p < 0.001 vs. corresponding vehicle; ##p < 0.01 vs. CB1f/f vehicle group; ###p < 0.001 vs. CB1f/f vehicle group. Scale bar: 10 µm. adolescence interferes with maturation of GABAergic function, GABAergic circuits can contribute to cognitive impairment [38]. leading to layer V prefrontal cortex disinhibition [36]. Hence, both CB1R signaling is critical in cognition and memory, and hence perinatal and adolescent exposure to cannabinoids result in cannabinoid exposure interferes with working and episodic GABAergic hypofunction. The contribution of this GABAergic memory [38]. Although in the adult brain acute and chronic THC hypofunction to the neuropsychiatric traits induced by develop- administration impairs NOR in males and females [34], our results mental cannabinoid exposure has been mainly associated to indicate that embryonic THC exposure does not affect conceptual increased risk of developing psychosis or schizophrenia [37]. learning, but instead blunts spatial memory, and this occurs in a Alternatively, cannabinoid-induced remodeling and plasticity of striking sex-dependent manner. Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 883 Fig. 4 Hippocampal-dependent spatial memory, but no object recognition memory, is altered by prenatal THC exposure specifically in males that preserve CB1R in GABAergic neurons. a, b Novel object recognition test was performed in Glu-CB1-KO, GABA-CB1-KO, and CB1f/f P60 mice, and a discrimination index was calculated for both males and females. c, d Object location test was performed in the mentioned groups, a similar index was then calculated for males and females. n = 3–14 animals per group. *p < 0.05 vs. corresponding vehicle; **p < 0.01 vs. corresponding vehicle; ##p < 0.01 vs. CB1f/f vehicle group. Our results also show a reduction of CCK interneurons and a emerging concept of different microcircuit organization along the selective impairment of spatial memory, consistent with the major deep-superficial hippocampal sublayers [43, 44]. The immediate role of hippocampal CCK BCs in spatial information coding and early gene transcription factor NPAS4 mediates experience-driven control of hippocampal oscillations [25, 39]. A reduction in both recruitment of CCK-evoked cannabinoid inhibition [45]. Hence, CCK BC density and inhibitory tone decreases the power of θ enriched environment may constitute a valuable strategy to oscillations during exploratory behavior and disrupts place cell- counteract the detrimental consequences of prenatal cannabinoid dependent spatial coding, hence impairing spatial learning [39]. exposure as an alternative to ongoing studies aimed to prevent We found changes in the θ and the low ripple band exclusively in deleterious cannabinoid maladaptive plasticity by pharmacological THC-treated male mice. Strikingly, θ-nested γ oscillations were not manipulation [13, 46]. affected, thus confirming frequency- and circuit-specific effects. Our findings highlight the importance of addressing sex CB1R differentially contributes to perisomatic inhibition of super- differences when investigating the neurodevelopmental changes ficial vs. deep CA1 pyramidal cells [26]. Hence, given the ability of induced by cannabis exposure and the functional consequences in CCK interneurons that express CB1R to influence SWR [26], and the the offspring [47]. Males and females possess different expression role of SWR in memory [40], our findings support the notion that levels of endocannabinoid system elements, respond differently to disrupted hippocampal oscillations are responsible for particular THC, and hence are differently affected in various cannabinoid- forms of cognitive impairment induced by prenatal THC exposure. related parameters [48, 49]. In addition, whereas we cannot In adult mice, acute THC-induced alterations in synchronized completely exclude that changes of oestrous cycle may play a role neural oscillations in the γ (30–80 Hz) and θ (4–7 Hz) ranges have in the behavioral differences induced by prenatal THC exposure, been linked to psychosis-related alterations [27], and CB1R- the male-selective observed interneuronopathy makes this possi- dependent regulation of cortical and subcortical network syn- bility very unlikely. In animal models, prenatal cannabinoid chrony has been proposed to participate on THC-induced exposure induces sex-dimorphic changes in pyramidal neuron alterations of sensory perception [41]. In this regard, a thorough intrinsic properties and synaptic plasticity in the PFC, which are in characterization of hippocampal oscillatory activity in mice lacking turn associated to social interaction deficits [17]. In addition, sex- CB1R selectively in interneurons or principal neurons would shed dimorphic THC-induced CCK BC interneuronopathy contributes to additional light on the precise functional contribution and spatial cognitive impairment (present study) and THC-induced sensitivity to prenatal THC exposure of different CB1R-expressing hyperdopaminergic state is also male selective [13]. The existence neuronal populations to network activity and cognitive function, of a sex-dependent bias in cannabinoid-induced interference with an issue that deserves future research. Previous findings suggest social interaction mediated by a loss of GABAergic perisomatic that CB1R controls spatial memory by regulating hyperpolarization- inhibition remains unknown [15]. The nature of this bias would activated cyclic nucleotide-gated (HCN) cationic channels and thus require further investigation. Indeed, brain wide-mapping current Ih, an effect that is specific of CA1 superficial pyramidal studies have recently revealed cell-type-specific contributions to neurons [42]. Possibly, the frequency-specific effect found in ripple cortical and subcortical sexual dimorphism [50], suggesting that distribution of THC-treated males can be associated with the many effects should be identified at the system level. Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 884 Fig. 5 THC-induced actions that depend on principal neurons are not sex-dependent. a, b THC-induced analgesia was analyzed in vehicle- and THC-treated Glu-CB1-KO, GABA-CB1-KO, and CB1f/f P60 male and female mice. c, d THC-induced hypothermia was analyzed in the same animal groups. n = 3–14 animals per group. *p < 0.05 vs. corresponding vehicle; ***p < 0.001 vs. corresponding vehicle; #p < 0.05 vs. CB1f/f vehicle group; ###p < 0.001 vs. CB1f/f vehicle group; $p < 0.05 vs. control group within each genotype; $$p < 0.01 vs. control group within each genotype; $$$p < 0.001 vs. control group within each genotype. In our hands, we did not find that embryonic THC influences dependent analyses of PTZ-induced seizure susceptibility in mice psychotic-like features in the offspring (startle response and conditionally CB1R-rescued from a CB1R-null background and prepulse inhibition; data not shown). One possible explanation subjected to prenatal THC treatment [16] confirms the involvement is that, when modeling human cannabinoid exposure in of GABAergic interneurons, but not principal neurons, in the THC- laboratory animals, WIN-55,212-2 is likely to induce a stronger induced hyperexcitability of the male progeny (Supplementary impact than THC and other phytocannabinoid molecules. Fig. S4). Recently, some consequences of prenatal THC exposure in Phytocannabinoids possess important differences in solubility, the male offspring have been shown to be rescued by co- potency, and hence pharmacokinetic and pharmacodynamic administration of pregnenolone, a CB1R modulator [13]. In summary, behavior, compared with synthetic cannabinoid drugs used for the study of functional consequences (hyperexcitability, cognition, research purposes (e.g., WIN-55,212-2, HU-210). Moreover, and other behavioral traits), in vivo electrophysiological recordings, different cannabinoid ligands contribute differentially to biased and neuron-lineage tracking in conditional CB1R-deficient mice, as CB1R signaling and can target additional receptors and binding shown herein, demonstrates that neurodevelopmental exposure to proteins. Noteworthy, a severe cannabinoid-induced impair- THC exerts a sex-dependent interneuronopathy that selectively ment of cortical oscillations was observed upon administration affects spatial cognitive function in male offspring. of the WIN-55,212-2 compound [36, 51]. Hence, it would be desirable that, when attempting to extrapolate to humans the consequences of cannabinoid exposure from experimental FUNDING AND DISCLOSURE models based on small laboratory animals, exquisite care is The authors declare no conflict of interest. This work was supported taken in the pharmacological regulation strategy and the by grants PI18-00941 to IG-R cofinanced by the European experimental design that is used. Evidence for the impact of Development Regional Fund “A way to achieve Europe”; RTI2018- sex-dimorphic prenatal cannabinoid exposure on human brain 095311-B-100 to MG, BFU2015-66887-R to LM-P, and 2017-SGR-138 development is scarce and its interpretation is extremely to MP from the Generalitat de Catalunya. DG-R was supported by complex due to a wide array of confounding factors. None- Fundación Tatiana Pérez de Guzmán; DG-D was supported by a PhD theless, early human cannabinoid exposure contributes differ- fellowship from the Spanish Ministry of Economy and Competitive- ently to drug addiction vulnerability [52] and the development ness (BES-2013-064171). JP-L and JA were supported by FPI and FPU of aggressiveness in male and female offspring [53]. program fellowships, respectively (Ministerio de Educación, Cultura y Here we demonstrate that conditional CB1R ablation in the Deporte) and S. S-S. was supported by Fondo Social Europeo-YEI interneuron lineage induces a similar sex-dependent interneurono- (CT101/18-CT102/18PEJD-2018-PRE/BMD-7933). CM is recipient of a pathy and spatial memory impairment as that evoked by prenatal Marie Curie program fellowship (747487). THC treatment, thus pointing to CB1R receptor loss of function as the main mechanism of THC action on the analyzed traits. Overall, repeated cannabinoid exposure in the immature brain results in ADDITIONAL INFORMATION functional antagonism of CB1R signaling in principal neurons [16] Supplementary Information accompanies this paper at (https://doi.org/10.1038/ and CCK BCs (present study). As a consequence of THC treatment, s41386-020-0621-3). CB1R desensitizes and the resulting cannabinoid signaling impair- ment would favor the development of seizures and spatial cognitive Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims deficits in a sex-dimorphic manner. In agreement, post-hoc sex- in published maps and institutional affiliations. Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 885 REFERENCES 24. Klausberger T, Somogyi P. Neuronal diversity and temporal dynamics: the unity of 1. Grant KS, Petroff R, Isoherranen N, Stella N, Burbacher TM. Cannabis use during hippocampal circuit operations. Science 2008;321:53–7. pregnancy: pharmacokinetics and effects on child development. Pharm Ther 25. Klausberger T, Marton LF, O’Neill J, Huck JHJ, Dalezios Y, Fuentealba P, et al. 2018;182:133–51. Complementary roles of cholecystokinin- and parvalbumin-expressing 2. Maccarrone M, Guzmán M, Mackie K, Doherty P, Harkany T. Programming of GABAergic neurons in hippocampal network oscillations. J Neurosci 2005;25: neural cells by (endo)cannabinoids: from physiological rules to emerging thera- 9782–93. pies. Nat Rev Neurosci. 2014;15:786–801. 26. Valero M, Cid E, Averkin RG, Aguilar J, Sanchez-Aguilera A, Viney TJ, et al. 3. Galve-Roperh I, Chiurchiù V, Díaz-Alonso J, Bari M, Guzmán M, Maccarrone M. Determinants of different deep and superficial CA1 pyramidal cell dynamics Cannabinoid receptor signaling in progenitor/stem cell proliferation and differ- during sharp-wave ripples. Nat Neurosci 2015;18:1281–90. entiation. Prog Lipid Res. 2013;52:633–50. 27. Skosnik PD, Cortes-Briones JA, Hajós M. It’s all in the rhythm: the role 4. Tortoriello G, Morris CV, Alpar A, Fuzik J, Shirran SL, Calvigioni D, et al. Miswiring of cannabinoids in neural oscillations and psychosis. Biol Psychiatry 2016;79: the brain: Δ9-tetrahydrocannabinol disrupts cortical development by inducing an 568–77. SCG10/stathmin-2 degradation pathway. EMBO J 2014;33:668–85. 28. Buzsáki G. Hippocampal sharp wave-ripple: a cognitive biomarker for episodic 5. Bloomfield MAP, Ashok AH, Volkow ND, Howes OD. The effects of δ9- memory and planning. Hippocampus 2015;25:1073–188. tetrahydrocannabinol on the dopamine system. Nature 2016;539:369–77. 29. Stark E, Roux L, Eichler R, Senzai Y, Royer S, Buzsáki G. Pyramidal cell- 6. DiNieri JA, Wang X, Szutorisz H, Spano SM, Kaur J, Casaccia P, et al. Maternal interneuron interactions underlie hippocampal ripple oscillations. Neuron cannabis use alters ventral striatal dopamine D2 gene regulation in the offspring. 2014;83:467–80. Biol Psychiatry 2011;70:763–9. 30. Gillespie AK, Jones EA, Lin Y-H, Karlsson MP, Kay K, Yoon SY, et al. Apolipoprotein 7. Campolongo P, Trezza V, Cassano T, Gaetani S, Morgese MG, Ubaldi M, et al. E4 causes age-dependent disruption of slow gamma oscillations during hippo- Perinatal exposure to delta-9-tetrahydrocannabinol causes enduring cognitive campal sharp-wave ripples. Neuron 2016;90:740–51. deficits associated with alteration of cortical gene expression and neuro- 31. Valero M, Averkin RG, Fernandez-Lamo I, Aguilar J, Lopez-Pigozzi D, Brotons-Mas transmission in rats. Addict Biol 2007;12:485–95. JR, et al. Mechanisms for selective single-cell reactivation during offline sharp- 8. Mereu G, Fa M, Ferraro L, Cagiano R, Antonelli T, Tattoli M, et al. Prenatal exposure wave ripples and their distortion by fast ripples. Neuron 2017;94:1234–47. to a cannabinoid agonist produces memory deficits linked to dysfunction in 32. Sullivan D, Csicsvari J, Mizuseki K, Montgomery S, Diba K, Buzsaki G. hippocampal long-term potentiation and glutamate release. Proc Natl Acad Sci Relationships between hippocampal sharp waves, ripples, and fast gamma USA. 2003;100:4915–20. oscillation: influence of dentate and entorhinal cortical activity. J Neurosci 9. Rubino T, Parolaro D. The impact of exposure to cannabinoids in adolescence: 2011;31:8605–16. Insights from animal models. Biol Psychiatry 2016;79:578–85. 33. Langston RF, Wood ER. Associative recognition and the hippocampus: differential 10. Renard J, Rosen LG, Loureiro M, De Oliveira C, Schmid S, Rushlow WJ, et al. effects of hippocampal lesions on object-place, object-context and object-place- Adolescent cannabinoid exposure induces a persistent sub-cortical hyper-dopa- context memory. Hippocampus 2009;20:1139–53. minergic state and associated molecular adaptations in the prefrontal cortex. 34. Puighermanal E, Marsicano G, Busquets-Garcia A, Lutz B, Maldonado R, Ozaita A. Cereb Cortex 2016;27:1297–310. Cannabinoid modulation of hippocampal long-term memory is mediated by 11. Volkow ND, Wang G-J, Telang F, Fowler JS, Alexoff D, Logan J, et al. mTOR signaling. Nat Neurosci 2009;12:1152–8. Decreased dopamine brain reactivity in marijuana abusers is associated with 35. Galve-Roperh I, Aguado T, Rueda D, Velasco G, Guzmán M. Endocannabinoids: a negative emotionality and addiction severity. Proc Natl Acad Sci. 2014;111: new family of lipid mediators involved in the regulation of neural cell develop- E3149–56. ment. Curr Pharm Des. 2006;12:2319–25. 12. Morris CV, Dinieri Ja, Szutorisz H, Hurd YL. Molecular mechanisms of maternal 36. Cass DK, Flores-Barrera E, Thomases DR, Vital WF, Caballero A, Tseng KY. cannabis and cigarette use on human neurodevelopment. Eur J Neurosci. CB1 cannabinoid receptor stimulation during adolescence impairs the matura- 2011;34:1574–83. tion of GABA function in the adult rat prefrontal cortex. Mol Psychiatry 13. Frau R, Miczán V, Traccis F, Aroni S, Pongor CI, Saba P, et al. Prenatal THC 2014;19:536–43. exposure produces a hyperdopaminergic phenotype rescued by pregnenolone. 37. Volk DW, Lewis DA. The role of endocannabinoid signaling in cortical inhibitory Nat Neurosci 2019;22:1975–85. 2019 neuron dysfunction in schizophrenia. Biol Psychiatry 2016;79:595–603. 14. Berghuis P, Dobszay MB, Wang X, Spano S, Ledda F, Sousa KM, et al. 38. Curran HV, Freeman TP, Mokrysz C, Lewis DA, Morgan C, Parsons JA, et al. Keep Endocannabinoids regulate interneuron migration and morphogenesis off the grass? Cannabis, cognition and addiction. Nat Rev Neurosci. by transactivating the TrkB receptor. Proc Natl Acad Sci USA. 2005;102: 2016;17:293–306. 19115–20. 39. del Pino I, Brotons-Mas JR, Marques-Smith A, Marighetto A, Frick A, Marín O, et al. 15. Vargish GA, Pelkey KA, Yuan X, Chittajallu R, Collins D, Fang C, et al. Persistent Abnormal wiring of CCK+ basket cells disrupts spatial information coding. Nat inhibitory circuit defects and disrupted social behaviour following in utero exo- Neurosci 2017;20:784–92. genous cannabinoid exposure. Mol Psychiatry. 2016;22:56–67. 40. Fernández-Ruiz A, Oliva A, de Oliveira EF, Rocha-Almeida F, Tingley D, Buzsáki G. 16. de Salas-Quiroga A, Díaz-Alonso J, García-Rincón D, Remmers F, Vega D, Gómez- Long-duration hippocampal sharp wave ripples improve memory. Sci (80-) Cañas M, et al. Prenatal exposure to cannabinoids evokes long-lasting functional 2019;364:1082–6. alterations by targeting CB 1 receptors on developing cortical neurons. Proc Natl 41. Sales-Carbonell C, Rueda-Orozco PE, Soria-Gómez E, Buzsáki G, Marsicano G, Acad Sci USA. 2015;112:13693–8. Robbe D. Striatal GABAergic and cortical glutamatergic neurons mediate con- 17. Bara A, Manduca A, Bernabeu A, Borsoi M, Serviado M, Lassalle O, et al. Sex- trasting effects of cannabinoids on cortical network synchrony. Proc Natl Acad Sci dependent effects of in utero cannabinoid exposure on cortical function. Elife. USA. 2013;110:719–24. 2018;7:e36234. 42. Maroso M, Szabo GG, Kim HK, Alexander A, Bui AD, Lee SH, et al. Cannabinoid 18. Mark K, Terplan M. Cannabis and pregnancy: maternal child health control of learning and memory through HCN channels. Neuron implications during a period of drug policy liberalization. Prev Med (Balt). 2017; 2016;89:1059–73. 104:46–9. 43. Soltesz I, Losonczy A. CA1 pyramidal cell diversity enabling parallel information 19. Monory K, Blaudzun H, Massa F, Kaiser N, Lemberger T, Schütz G, et al. Genetic processing in the hippocampus. Nat Neurosci 2018;21:484–93. dissection of behavioural and autonomic effects of Delta(9)-tetra- 44. Valero M, de la Prida LM. The hippocampus in depth: a sublayer-specific per- hydrocannabinol in mice. PLoS Biol 2007;5:e269. spective of entorhinal–hippocampal function. Curr Opin Neurobiol. 20. Foffani G, Uzcategui YG, Gal B, Menendez, de la Prida L. Reduced spike-timing 2018;52:107–14. reliability correlates with the emergence of fast ripples in the rat epileptic hip- 45. Hartzell AL, Martyniuk KM, Brigidi GS, Heinz DA, Djaja NA, Payne A, et al. NPAS4 pocampus. Neuron 2007;55:930–41. recruits CCK basket cell synapses and enhances cannabinoid-sensitive inhibition 21. Monory K, Massa F, Egertová M, Eder M, Blaudzun H, Westenbroek R, et al. The in the mouse hippocampus. Elife 2018;7:1–24. endocannabinoid system controls key epileptogenic circuits in the hippocampus. 46. Vallée M, Vitiello S, Bellocchio L, Hébert-Chatelain E, Monlezun S, Martin-Garcia E, Neuron 2006;51:455–66. et al. Pregnenolone can potect the brain from cannabis intoxication. Science (80). 22. Marsicano G, Lutz B. Expression of the cannabinoid receptor CB1 in distinct 2014;343:94–8. neuronal subpopulations in the adult mouse forebrain. Eur J Neurosci. 1999;11: 47. Sanchis-Segura C, Becker JB. Why we should consider sex (and study sex dif- 4213–25. ferences) in addiction research. Addict Biol 2016;21:995–1006. 23. Berghuis P, Rajnicek AM, Morozov YM, Ross RA, Mulder J, Urban GM, et al. 48. Cuttler C, Mischley LK, Sexton M. Sex differences in cannabis use and effects: a Hardwiring the brain: endocannabinoids shape neuronal connectivity. Science cross-sectional survey of cannabis users. Cannabis Cannabinoid Res. 2016;1: (80-). 2007;316:1212–6. 166–75. Neuropsychopharmacology (2020) 45:877 – 886 Long-term hippocampal interneuronopathy drives sex-dimorphic spatial. . . A de Salas-Quiroga et al. 886 49. Rubino T, Parolaro D. Sexually dimorphic effects of cannabinoid compounds on Open Access This article is licensed under a Creative Commons emotion and cognition. Front Behav Neurosci. 2011;5:1–5. Attribution 4.0 International License, which permits use, sharing, 50. Kim Y, Yang GR, Pradhan K, Venkataraju KU, Bota M, García del Molino LC, et al. adaptation, distribution and reproduction in any medium or format, as long as you give Brain-wide maps reveal stereotyped cell-type-based cortical architecture and appropriate credit to the original author(s) and the source, provide a link to the Creative subcortical sexual dimorphism. Cell 2017;171:456–69.e22. Commons license, and indicate if changes were made. The images or other third party 51. Raver SM, Keller A. Permanent suppression of cortical oscillations in mice after material in this article are included in the article’s Creative Commons license, unless adolescent exposure to cannabinoids: receptor mechanisms. Neuropharmacol- indicated otherwise in a credit line to the material. If material is not included in the ogy 2014;86:161–73. article’s Creative Commons license and your intended use is not permitted by statutory 52. Wang X, Dow-Edwards D, Anderson V, Minkoff H, Hurd YL. In utero marijuana regulation or exceeds the permitted use, you will need to obtain permission directly exposure associated with abnormal amygdala dopamine D 2 gene expression in from the copyright holder. To view a copy of this license, visit http://creativecommons. the human fetus. Biol Psychiatry 2004;56:909–15. org/licenses/by/4.0/. 53. El Marroun H, Hudziak JJ, Tiemeier H, Creemers H, EAP Steegers, VWV Jaddoe, et al. Intrauterine cannabis exposure leads to more aggressive behavior and attention problems in 18-month-old girls. Drug Alcohol Depend. 2011;118:470–4. © The Author(s) 2020 Neuropsychopharmacology (2020) 45:877 – 886