Sex specific endophenotypes of in-utero cannabinoid exposure

Cannabinoids can cross the placenta, thus may interfere with fetal endocannabinoid signaling during neurodevelopment, causing long-lasting deficits. Despite increasing cannabis consumption during pregnancy, the protracted consequences of prenatal cannabinoid exposure (PCE) remain incompletely understood. Here we report sex-specific differences in behavioral and neuronal deficits in the adult progeny of rat dams exposed to low doses of cannabinoids during gestation. In males, PCE reduced social interaction, ablated endocannabinoid long-term depression (LTD) and heightened excitability of prefrontal cortex pyramidal neurons, while females were spared. Group 1 mGluR and endocannabinoid signaling regulate emotional behavior and synaptic plasticity. Notably, sex-differences following PCE included levels of mGluR1/5 and TRPV1R mRNA. Finally, positive allosteric modulation of mGlu5 and enhancement of anandamide levels restored LTD and social interaction in PCE adult males. Together, these results highlight marked sexual differences in the effects of PCE and introduce strategies for reversing detrimental effects of PCE.

Regarding their cognitive abilities, adult SHAM and WIN-exposed animals of both sexes displayed identical discrimination index in both the temporal order memory task (Figure 2 C) and the novel object recognition test (Figure 2 D). Thus, while social interaction was specifically impaired in PCE males, locomotion, anxiety and cognition were spared in both sexes. These data reveal discrete and sex-specific behavioral consequences of PCE at adulthood.

Prenatal cannabinoid exposure alters synaptic plasticity specifically in the PFC of adult male.
In theory, cannabinoid exposure during in-utero neurodevelopment may perturb synaptic functions in most brain areas. The finding of selective impairments of social interaction in the absence of emotional and cognitive deficits does not favor the hypothesis of widespread synaptic deficits. The prefrontal cortex and nucleus accumbens play prominent roles in social behaviors (van Kerkhof et al. 2013) and prefronto-accumbens glutamatergic circuits modulate reward-related behaviors (Mateo et al., 2017;Floresco, 2015). Furthermore, the endogenous cannabinoid (eCB)-system of the accumbens core is instrumental to social interaction behavior (Manduca et al., 2016;Trezza et al., 2012). In-vivo cannabinoid exposure desensitizes CB1R and ablates eCB-mediated synaptic plasticity in the accumbens (Mato et al., 2004;Mato et al., 2005). Thus, we first compared eCB-mediated LTD (eCB-LTD) in our experimental groups. The data indicated that eCB-LTD in the accumbens is not affected by PCE (Figure 3). The extensive repertoire of synaptic plasticity displayed by medial prefrontal synapses is a consistent target of environmental and genetic insults (Iafrati et al., 2014;Kasanetz et al., 2013;Labouesse et al., 2017;Lafourcade et al., 2011a;Manduca et al., 2017;Thomazeau et al., 2014). We first established that tetanic stimulation induced a robust eCB-LTD of excitatory synapses onto layer 5 mPFC pyramidal neurons in mPFC slices prepared from adult SHAM male and female rats (Figure 4 A-B-C-D). In marked contrast to the accumbens, eCB-LTD in the mPFC was absent in WIN-exposed males (Figure 4 A-B). Strikingly, LTD was normal in PCE female littermates . In a second series of experiments a similar sex-specific ablation of eCB-LTD in the progeny of dams exposed to THC, the main psychoactive component of cannabis, was observed (5mg/kg, s.c. GD5-20; Figure 5). To test if the effects of PCE extend to other forms of plasticity expressed at excitatory mPFC synapses we assessed two critical actors of bidirectional plasticity and reliable prefrontal endophenotypes of neuropsychiatric diseases: mGlu2/3-dependent LTD and NMDARdependent LTP (Iafrati et al., 2014(Iafrati et al., , 2016Kasanetz et al., 2013;Thomazeau et al., 2014). In contrast to the lack of eCB-LTD, mGlu2/3-mediated LTD was indistinguishable in SHAM-and WIN-exposed male and female ( Figure 6 A-B). Similarly, theta-burst stimulation (TBS) induces similar levels of NMDAR-dependent LTP at mPFC pyramidal synapses independent of sex and prenatal treatment (Figure 6 C-D-E-F). We also confirmed that prenatal THC does not affect prefrontal LTP at adulthood (Figure 7). Together the data reveal eCB-LTD in the mPFC as sex-specific target of in-utero cannabinoid exposure.

Cannabinoid prenatal exposure causes sex-specific modifications in pyramidal neuron properties.
Activity-induced modulation of intrinsic excitability is a powerful mean to control cortical circuits (Bacci et al., 2005;Debanne and Poo, 2010) postulated to contribute to disorders of the CNS. To test the hypothesis that PCE altered prefrontal excitability, intrinsic firing properties of pyramidal neurons in our animal groups were compared. Sexual differences to the effects of prenatal exposure to cannabinoids were visible at this cellular level. Specifically, mPFC neurons in WIN-exposed rats showed impaired membrane reaction profiles in response to a series of somatic current steps (Figure 8 A and E). Independent of the sex and treatment, the resting potential was similar (Figure 8 B and F). While the rheobase (Figure 8 C and G) was decreased in males prenatally exposed to WIN, females were spared. Moreover, this hyper-excitability in exposed-males was accompanied by an increased number of action potentials in response to somatic currents steps (Figure 8 D). Similar results have been obtained in the rat progeny of dams treated with THC during gestation ( Figure 9). Thus, PCE leads to an increased excitability of deep layers mPFC pyramidal cells in male progeny only.

Sex-differences in mRNA expression levels
Motivated by the eCB-LTD deficiency in layer 5 mPFC of the male PCE progeny, we surveyed mRNA levels of key components of the endocannabinoid system in adult PCE rats using real time quantitative PCR ( Figure 10 and Table 3). Key differences were decreased levels of mRNA for TRPV1, mGlu5, and DAGLα in the female PCE offspring and decreased levels of mGlu5 mRNA in the male offspring ( Figure 10). A mild increase was seen in mGlu1 in the PCE males (Table 3). No differences were seen in other endocannabinoid system mRNAs after PCE, including CB1, CB2, DAGLβ, NAPE-PLD, FAAH, ABHD6, or CRIP1a (Table 3).

Positive allosteric modulation of mGlu5 corrects synaptic and behavioral deficits associated with prenatal cannabinoid male exposure.
The present data reveal a down-regulation of mGlu5 RNA expression levels in the mPFC of PCE males. Previous work from our group underlies the efficiency of mGlu5 positive allosteric modulation in correcting synaptic and behavioral deficits (Martin et al., 2017;Manduca et al., 2017). Thus, we tested whether amplification of mGlu5, an effector of eCB-signaling complex in the PFC, could normalize LTD in PCE male. To this aim, we used CDPPB (10 μM), a well-described positive allosteric modulator (PAM) of mGlu5 (Doria et al., 2015;Manduca et al., 2017;Martin et al., 2016a;Nickols and Conn, 2014;Won et al., 2012). We found that CDPPB fully restored the ability of excitatory mPFC synapses to express LTD in the prenatally WINexposed male group (Figure 11 A -B). To verify that the curative properties of CDPPB were indeed due to eCB production, we tested the effects of CB1R and TRPV1 antagonists on the CDPPB rescue. Interestingly, preincubation with either SR141716A or capsazepine prevented the CDPPB rescue of synaptic plasticity (Figure 11 A-B). Thus, the data showed that both receptors are engaged in the presence of the mGlu5 PAM. Strikingly, we found that systemic CDPPB normalized social cognition in the PCE group but remained without effect in the SHAM group, indicating selectivity of the drug's effects to the disease-state (Figure 11 C). In agreement with the synaptic plasticity experiments, pretreatment with either SR141716A (Figure 11 D) or capsazepine prevented the ameliorative actions of CDPPB (Figure 11 D). Thus, potentiating mGlu5 signaling normalized synaptic and behavioral deficits induced by prenatal exposure to cannabinoids via CB1R and TRPV1 activation.

TRPV1, not CB1R, mediate LTD in PCE females.
The observation that PCE did not alter CB1R RNA levels in either sex but lowered TRPV1R in the female PCE progeny ( Figure 10, Table 3) drew our attention to the role of these two receptors at PFC synapses. Comparing full dose-response curves for the CB1 agonist CP55,940 in male SHAM vs PCE revealed no differences ( Figure 12). Thus, the lack of LTD in PCE males cannot readily be attributed to altered sensitivity of presynaptic CB1R. We and others have shown that central excitatory synapses can engage TRPV1R and/or CB1R to trigger LTD (Chávez et al., 2010;Grueter et al., 2010;Martin et al., 2016b;Puente et al., 2011). In PCE females, LTD was fully expressed in the presence of a specific CB1R antagonist (SR141716A 5 μM, Figure 13 A-B-C) but completely prevented by TRPV1R antagonists (capsazepine 10µM, Figure 13 A-B-C). Thus, TRPV1R rather than CB1R mediates LTD In female PCE. This contribution of TRPV1R to LTD in female PCE could be part of the adaptive response to PCE or a previously unnoticed sexual difference in the mechanism of prefrontal synaptic plasticity. We next investigated the substrate of LTD in naive male and female rats ( Figure 14). As expected from over a decade of studies in male rodents (Bender et al., 2006;Chevaleyre et al., 2006;Gerdeman et al., 2002;Iremonger et al., 2011;Lafourcade et al., 2007;Marsicano et al., 2002;Nevian and Sakmann, 2006;Puente et al., 2011;Robbe et al., 2002a)

Enhancing anandamide levels corrects synaptic and behavioral deficits associated with prenatal cannabinoid male exposure.
We thought of an additional strategy to reverse PCE induced deficits based on our finding that TRPV1R mediates LTD in PCE females in spite of a down-regulation of TRPV1R RNA expression levels. Anandamide is an endogenous ligand of TRPV1 (Di Marzo et al., 2002) and we hypothesized that inhibition of anandamide's main degrading enzyme, fatty acid amid hydrolase (FAAH), would increase anandamide levels to the threshold for LTD induction in the PCE males (Rubino et al. 2015). Indeed, incubation of mPFC slices prepared from PCE male rats with the FAAH inhibitor URB597 effectively restored eCB-LTD ( Figure 15 A-B). Interestingly, we found that pre-incubation with the specific TRPV1 antagonist capsazepine did not prevent the curative effect of URB597. Rather, the CB1R antagonist SR141716 blocked the URB597 rescue ( Figure 15 A-B). Previous studies have shown that endocannabinoids, including anandamide, mediate social behavior in rodents (Trezza et al., 2012). We directly tested the hypothesis that elevating circulating levels of a main eCB, anandamide, could normalize social deficits. We found that in vivo treatment with the FAAH inhibitor (URB597, 0.1 mg/kg) normalized social deficits in adult male PCE rats (Figure 15 C). In-line with our mPFC slice results, SR141716A but not capsazepine prevented the curative actions of URB597 (Figure 15 D). Thus, blocking anandamide degradation normalized social deficits via CB1R.

DISCUSSION
We discovered that fetal exposure to cannabinoids causes sex-specific alterations of behavioral and synaptic functions in the adult progeny. Specifically, we found that prenatal cannabinoid exposure (PCE) reduces social interactions in males but not females, while other behaviors were spared. In parallel, PCE specifically altered neuronal excitability and synaptic plasticity in the prefrontal cortex of male rats. We also identify a pharmacological strategy to normalize synaptic functions and socialization in adult prenatally-exposed males. PFC malfunctions are a common denominator in several neuropsychiatric diseases (Goto et al., 2010;Iafrati et al., 2014;Kasanetz et al., 2013;Labouesse et al., 2017;Lafourcade et al., 2011a;Manduca et al., 2017;Martin et al.). Our data indicate that in the PCE group, plasticity deficits were circumscribed to eCB-LTD as two other types of synaptic plasticity, which act as biomarkers for prefrontal deficits in other model of neuropsychiatric diseases, were spared. Namely neither NMDAR-mediated LTP (Iafrati et al., 2014;Manduca et al., 2017;Martin and Manzoni, 2014;Thomazeau et al., 2017) nor mGluR2/3-LTD (Kasanetz et al., 2013) were affected in PCE males. Furthermore, while excitatory synapses of the nucleus accumbens display reduced eCB-mediated plasticity in models of depression (Bosch-Bouju et al., 2016), autism (Jung et al., 2012a, adult cannabis exposure (Mato et al., 2004(Mato et al., , 2005 or nutritional deficits (Manduca et al., 2017), accumbal eCB-LTD was normal in both PCE males and females. Our behavioral exploration reveals that social interactions were reduced in males after PCE while female social behavior and other behavioral domains were spared in both sexes. Our results greatly extend that of Vargish and collaborators, who observed reduced social interaction in PCE mice (Vargish et al., 2017). In the latter study, female progeny was unfortunately not tested. The current results add to previous work showing covariation between defects in PFC synaptic plasticity and social interaction behaviors (Manduca et al., 2017). Together, these data suggest a certain level of regional PFC specificity to the effects of PCE. The fact that the PFC develops late in to postnatal life (i.e. late adolescence/early adulthood) (Arain et al., 2013;Fuster, 1991;Kolb et al., 2012) may explain this sensitivity. In the hippocampus, PCE impaired eCB-dependent short-term depression at GABAergic synapses (Vargish et al., 2017) and NMDAR-dependent LTD (Tortoriello et al., 2014) Mereu et al, PNAS 2003. In contrast with the latter hippocampal study, we found no evidence that PCE modified release properties of PFC synapses (supplementary results), reinforcing the idea that region-specific mechanisms engage to alter PFC synapses in response to PCE. Our qRT-PCR data showed similar levels of CB1R mRNA in the PFC of male and female progeny exposed to cannabinoids during fetal life, in line with previous observations in the hippocampus and neocortex (Tortoriello et al. 2014). While our work clearly shows previously undisclosed sexual divergence in the consequences of fetal cannabinoids, they also reveal sex differences in the main elements involved in eCB signaling: CB1R was similar in both sexes while lower mRNA levels of the 2arachidonoylglycerol (2-AG) synthesizing (DAGL⍺) enzyme were found in naive females compared to naive males. Previous work highlighted sexual differences at the molecular level. CB1R affinity is higher in males while receptor density is similar in both sexes in the limbic forebrain; however, in the mesencephalon both parameters are higher in males (Rodríguez de Fonseca et al., 1994). In the PFC, CB1R density is lower in males than in females (Castelli et al., 2014). At the structural level, the medial PFC of rats is also sexually different: female prelimbic cortex pyramidal neurons have smaller/less complex apical dendritic arbors than in males (Garrett and Wellman, 2009;Kolb and Stewart, 1991). Thus, it is likely that both nanoscale differences in the relative position of the molecular elements of the eCB system and dissimilarities of mRNA/protein expression underlie these sex differences. At the behavioral level, sexual differences in fear-conditioning have also been identified (Gruene et al., 2015;Gupta et al., 2001;Pryce et al., 1999). Sexual differences are unlikely exclusive to the eCB-LTD/PFC/fear conditioning. For instance, active avoidance (Dalla et Shors 2009), spatial memory (Dalla and Shors, 2009;Qi et al., 2016) and hippocampal LTP (Gupta et al., 2001;Maren, 1995;Qi et al., 2016) display sex differences. The data provide new impetus for the urgent need to investigate the functional and behavioral substrates of neuropsychiatric diseases in both sexes. Reduced bioavailability of endocannabinoids during LTD induction likely underlies the lack of synaptic plasticity found in PCE males. In support of this idea, basal levels of anandamide are reduced in limbic areas of prenatally WIN-exposed rats (Castelli et al., 2007). We hypothesized that enhancing anandamide levels may lower the LTD-induction threshold and indeed found that FAAH inhibition with URB597 restored eCB-LTD in PCE males. Furthermore, based on the well-known, prominent role of mGlu5 in synaptic eCB-signaling (Araque et al., 2017), we found that positive allosteric modulation of mGlu5 with CDPPB (Manduca et al., 2017;Martin et al., 2016b) restored LTD in PCE males. Strikingly, this finding translated well from the ex-vivo slice preparations to the behavioral level, and in-vivo treatments with either the FAAH inhibitor or the mGlu5 PAM normalized social interactions in PCE exposed adult males. In conclusion, these results provide compelling evidences for sexual divergence in the longterm functional and behavioral consequences of PCE and introduce strategies for reversing its detrimental effects.

Animals:
The experiments were performed following the ARRIVE (Animals in Research: Reporting In Vivo Experiments) guidelines (Kilkenny et al. 2010). Wistar female rats (Charles River, France) weighing 250-280 g were housed at constant room temperature (20 ± 1°C) and humidity (60%), and exposed to a light cycle of 12 h/day (08:00 a.m. to 08:00 p.m.), with food and water available ad libitum. For mating, pairs of females were placed with single male rat in the late afternoon. Vaginal smears were taken the following morning at 09:00 a.m. The day on which sperm was present was designated as the day 0 of gestation (GD 0). Sub-cutaneous (s.c.) injections were performed daily with the synthetic cannabinoid WIN55,212-2 (WIN, 0.5 mg/kg), the phytocannabinoid Delta9-Tetrahydrocannabinol, THC (5 mg/kg) or their respective vehicles from GD 5 to GD 20. WIN was suspended in 5% DMSO, 5% cremophor and saline, at a volume of 1 ml/kg. Control dams (SHAM group) received a similar volume injection of vehicle solution. THC was suspended in 5% Ethanol, 5% cremophor and saline, and injected s.c. at a volume of 1 ml/kg. Control dams (SHAM group) received a similar volume injection of vehicle solution. Newborn litters found up to 05:00 p.m. were considered to be born on that day (postnatal day (PND) 0). On PND 21, the pups were weaned and housed in groups of three. The experiments were carried out on the male and female offspring at adulthood (PNDs 90-130). One pup per litter from different litters per treatment group was used in the behavioral experiments. Sample size (n and scatter dot plot) is indicated in the figure legends. Animals were treated in compliance with the European Communities Council Directive (86/609/EEC) and the United States National Institutes of Health Guide for the care and use of laboratory animals. Body weights of the dams were taken daily throughout pregnancy and the length of pregnancy was determined. Litter size, weight gain of pups and postnatal vitality were also measured, see Table 1.

Behavior:
All animals were experimentally naïve and used only once. Social interaction: Social behavior was assessed as previously described (Manduca et al., 2015). The test was performed in a sound attenuated chamber under dim light conditions. The testing arena consisted of a Plexiglas cage measuring 45 × 45 cm, with 2 cm of wood shavings covering the floor. Rats were individually habituated to the test cage for 5 min on each of the 2 days prior to testing. Before testing, animals were socially isolated for 24 h to enhance their social motivation and thus facilitate the expression of social behaviors during testing. Drug treatments were counterbalanced so that cage mates were allocated to different treatment groups. The animals of each pair were similarly treated, did not differ more than 10 g in body weight and were not cage mates. Behavior was assessed per pair of animals and analyzed by a trained observer who was unaware of treatment condition using the Ethovision XT 13.0 video tracking software (Noldus, Wageningen, The Netherlands). The test consisted of placing two similarly treated animals into the test cage for 10 min. The total frequency of active social interactions was obtained as the sum of the frequency of the following behavioral elements scored per 10 min (Manduca et al., 2015;Segatto et al., 2014): • Play-related behaviors: pouncing (on rat solicits another rat to play by attempting to nose or rub the nape of its neck), pinning (the rat that is pounced upon can respond in different ways -'pinning' is the result when the animal lies with its dorsal surface on the floor with the other rat standing over it) and boxing (rats push or paw at each other with their forepaws, usually around the head, neck, shoulders of their opponent). • Social behaviors unrelated to play: social exploration (sniffing any part of the body of the test partner), social grooming (one rat licks and chews the fur of the conspecific, while placing its forepaws on the back or the neck of the other rat), following/chasing (walking or running in the direction of the partner which stays where it is or moves away), crawling under/over (one animal crawls underneath or over the partner‫׳‬s body, crossing it transversely from one side to the other), kicking (the rat kicks backwards at the conspecific with one or both hind paws). • Aggressive behavior: attack. Open field: The test was performed as we previously published (Manduca et al., 2017;Jung et al., 2012). The apparatus consisted of a Plexiglas arena 45 × 45 cm, illuminated by fluorescent bulbs at a height of 2 m above the floor of the open field apparatus (light intensity of 30 Lux). The floor was cleaned between each trial to avoid olfactory clues. Each animal was transferred to the open-field facing a corner and was allowed to freely explore the experimental area for 10 min. A video tracking system (Ethovision XT, Noldus Information Technology) recorded the exact track of each rat as well as total distance traveled (Lafourcade et al., 2011;Larrieu et al., 2012). Elevated plus-maze: The elevated plus-maze apparatus comprised two open (50 × 10 × 40 cm) and two closed arms (50 × 10 × 40 cm) that extended from a common central platform (10 × 10 cm). The test was performed as previously described (Manduca et al. 2015). Rats were individually placed on the central platform of the maze for 5 min. Each 5-min session was recorded with a camera positioned above the apparatus for subsequent behavioral analysis carried out an observer, unaware of animal treatment, using the Observer 3.0 software (Noldus, The Netherlands). The following parameters were analyzed: Temporal order: Animals were habituated to the experimental arena (40 × 40 cm) without objects for 10 min daily for 2 d before testing. This task consisted of two sample phases and one test trial (Barker et al., 2007). In each sample phase, rats were allowed to explore two copies of an identical object for a total of 4 min. Different objects were used for sample Phases 1 and 2, with a delay between the sample phases of 1 h. After 3 h from sample Phase 2, rats performed the test trial (4 min duration) where a third copy of the objects from sample Phase 1 and a third copy of the objects from sample Phase 2 were used. The positions of the objects in the test and the objects used in sample Phase 1 and sample Phase 2 were counterbalanced between the animals. An intact temporal order memory requested the subjects to spend more time exploring the object from Sample 1 (i.e., the object presented less recently) compared with the object from Sample 2 (i.e., the "new" object). The discrimination ratio was calculated as the difference in time spent by each animal exploring the object from sample Phase 1 compared with the object from sample Phase 2 divided by the total time spent exploring both objects in the test phase.
Novel object recognition: After 2 days of 10 min habituation to the experimental arena, rats were exposed to the procedure of an acquisition or sample phase, followed by a preference test after a delay of 30 min (Campolongo et al., 2014). In the sample phase, duplicate copies (A1 and A2) of an object were placed near the two corners at either end of one side of the arena (8 cm from each adjacent wall). Rats were placed into the arena facing the center of the opposite wall and allowed a total of 4 min in the arena. At test (4 min duration), animals were placed in the arena, presented with two objects in the same positions: one object (A3) was a third copy of the set of the objects used in the sample phase, and the other object was a novel object (B). The positions of the objects in the test and the objects used as novel or familiar were counterbalanced between the rats. The discrimination ratio was calculated as the difference in time spent by each animal exploring the novel compared with the familiar object divided by the total time spent exploring both objects. Exploration was scored when the animal was observed sniffing or touching the object with the nose and/or forepaws. Sitting on objects was not considered to indicate exploratory behavior.

Physiology
Slice preparation: Adult male and female rats were anesthetized with isoflurane and killed as previously described (Martin and Manzoni, 2014;Manduca et al., 2017). The brain was sliced (300 μm) in the coronal plane with a vibratome (Integraslice, Campden Instruments) in a sucrose-based solution at 4°C (in mm as follows: 87 NaCl, 75 sucrose, 25 glucose, 2.5 KCl, 4 MgCl2, 0.5 CaCl2, 23 NaHCO3 and 1.25 NaH2PO4). Immediately after cutting, slices containing the medial prefrontal cortex (PFC) or accumbens were stored for 1 h at 32°C in a low calcium ACSF that contained (in mm) as follows: 130 NaCl, 11 glucose, 2.5 KCl, 2.4 MgCl2, 1.2 CaCl2, 23 NaHCO3, 1.2 NaH2PO4, and were equilibrated with 95% O2/5% CO2 and then at room temperature until the time of recording. During the recording, slices were placed in the recording chamber and superfused at 2 ml/min with either low Ca 2+ ACSF for mPFC or normal ACSF for the accumbens. All experiments were done at 32°C or room temperature for mPFC and accumbens respectively. The superfusion medium contained picrotoxin (100 µM) to block gamma-aminobutyric acid types A (GABA-A) receptors. All drugs were added at the final concentration to the superfusion medium. Electrophysiology: Whole cell patch-clamp of visualized layer 5 pyramidal neurons mPFC and field potential recordings were made in coronal slices containing the mPFC or the accumbens as previously described (Kasanetz and Manzoni, 2009;Kasanetz et al., 2010Kasanetz et al., , 2013. Neurons were visualized using an upright microscope with infrared illumination. The intracellular solution was based on K + gluconate (in mM: 145 K + gluconate, 3 NaCl, 1 MgCl2, 1 EGTA, 0.3 CaCl2, 2 Na2 + ATP, and 0.3 Na + GTP, 0.2 cAMP, buffered with 10 HEPES). To quantify the AMPA/NMDA ratio we used a CH3O3SCs-based solution (in mM: 128 CH3O3SCs, 20 NaCl, 1 MgCl2, 1 EGTA, 0.3 CaCl2, 2 Na2 + ATP, and 0.3 Na + GTP, 0.2 cAMP, buffered with 10 HEPES, pH 7.2, osmolarity 290-300 mOsm). The pH was adjusted to 7.2 and osmolarity to 290-300 mOsm. Electrode resistance was 4-6 MOhms. Whole cell patch-clamp recordings were performed with an Axopatch-200B amplifier as previously described (Robbe et al., 2002;Kasanetz and Manzoni, 2009;Lafourcade et al., 2011;Jung et al., 2012;Thomazeau et al., 2014Thomazeau et al., , 2016Martin et al., 2016a). Data were low pass filtered at 2 kHz, digitized (10 kHz, DigiData 1440A, Axon Instrument), collected using Clampex 10.2 and analyzed using Clampfit 10.2 (all from Molecular Device, Sunnyvale, USA). A −2 mV hyperpolarizing pulse was applied before each evoked EPSC in order to evaluate the access resistance and those experiments in which this parameter changed >25% were rejected. Access resistance compensation was not used, and acceptable access resistance was <30 MOhms. The potential reference of the amplifier was adjusted to zero prior to breaking into the cell. Cells were held at -76 mV. Current-voltage (I-V) curves were made by a series of hyperpolarizing to depolarizing current steps immediately after breaking into the cell. Membrane resistance was estimated from the I-V curve around resting membrane potential (Kasanetz et al., 2010;Martin et al., 2015b). Field potential recordings were made in coronal slices containing the mPFC or the accumbens as previously described (Kasanetz and Manzoni, 2009;Kasanetz et al., 2010Kasanetz et al., , 2013. During the recording, slices were placed in the recording chamber and superfused at 2 ml/min with either low Ca 2+ ACSF for mPFC or normal ACSF for the accumbens. All experiments were done at 32°C or room temperature for mPFC and accumbens, respectively. The superfusion medium contained picrotoxin (100 μM) to block GABA Type A (GABA-A) receptors. All drugs were added at the final concentration to the superfusion medium. The glutamatergic nature of the field EPSP (fEPSP) was systematically confirmed at the end of the experiments using the ionotropic glutamate receptor antagonist CNQX (20 μM), which specifically blocked the synaptic component without altering the non-synaptic component (data not shown). Both fEPSP area and amplitude were analyzed. Stimulation was performed with a glass electrode filled with ACSF and the stimulus intensity was adjusted ∼60% of maximal intensity after performing an input-output curve (baseline EPSC amplitudes ranged between 50 and 150 pA). Stimulation frequency was set at 0.1 Hz. Data acquisition and analysis: The magnitude of plasticity was calculated 35-40 min after induction as percentage of baseline responses. Statistical analysis of data was performed with Prism (GraphPad Software) using tests indicated in the main text after outlier subtraction. All values are given as mean ± SEM, and statistical significance was set at p < 0.05.

Biochemistry
qRT-PCR: Frozen rat brains were coronally sectioned in a cutting block (Braintree Scientific, Inc., Braintree, MA, Cat.# BS-SS 605C) that had been pre-chilled to -20 o C. One-millimeter sections were kept frozen throughout dissection with brain regions stored at -80 o C until use. Total RNA was extracted using the RNeasy Plus Micro Kit (Qiagen, Hilden, Germany, Cat.# 74034 ). RNA was reverse-transcribed using the RevertAid Kit (Thermo Fisher Scientific, Waltham, MA Cat.# K1621) as per manufacturer's instructions. Taqman primers and probes were obtained through Applied Biosystems. Sequences used in qRT-PCR are as follows: TaqMan Gene Expression Master Mix from Applied Biosystems (Foster City, CA Cat.# 4369016) was used in generating expression data on a QuantStudio7 thermal cycler. Duplicates were run for each sample, and changes in gene expression were determined using the DDCt method. Data was analyzed using Excel (Microsoft, Redmond, WA) and Prism 7 (Graphpad, La Jolla, CA) software.

Statistics
All the behavioral parameters were expressed as mean ± SEM. Group comparisons used ttest and two-way repeated measures (RM) ANOVA (for different sex and pharmacological treatments), followed by Student-Newman-Keuls post hoc tests when appropriate.

Figure 1. Prenatal cannabinoid exposure induces sex-specific social deficits at adulthood.
A-B, Adult male progeny from dams exposed during gestation to WIN had less contact with their congeners than male SHAM-animals (A: F(sex x treat)1,28=15.54, p<0.05, two-way ANOVA). In contrast, the social behavior of female littermates was normal. During the social session, the number of sniffing events (B: F(sex x treat)1,28=16.30, p<0.05) and the frequency of play behavior (C: F(sex x treat)1,28=3.217, p>0.05, two-way ANOVA) were exclusively reduced in male rats exposed to WIN during gestation. Prenatal cannabinoid exposure did not induce aggressive behavior in adult rats (D: F(sex x treat)1,28=0.037, p>0.05, two-way ANOVA). Scatter dot plot represents a pair of animals. Error bars indicate SEM. *p<0.05. Student-Newman-Keuls test.

Figure 3: Prenatal cannabinoid exposure does not modify LTD in NAc of adult rats. A,
Average time-courses of mean fEPSPs showing that low-frequency stimulation (indicated by arrow) induces LTD at accumbens synapses in both SHAM-(white circles, n=5) and WIN-(green circles, n=5) exposed male rats. B, Individual experiments (grey) and group average (tan), before (baseline) and after (40 min) LTD induction showing that PCE does not affect eCB-LTD in male accumbens. In SHAM male rats: 2.518 ± 0.145 mV.ms before LTD versus 1.872 ± 0.124 mV.ms after LTD induction (n=5, p<0.05, paired t-test). In WIN male: 2.152 ± 0.224 mV.ms before LTD versus 1.696 ± 0.073 mV.ms after LTD induction (n=5, p<0.05, paired t-test). C, Average time-courses of mean EPSPs showing similar low-frequency LTD in SHAM (n=6; white circles) and WIN (n=4; orange circles) in utero-exposed female rats. D, Individual experiments (grey) and group average (white represents SHAM; orange represents WIN), before (baseline) and after (40 min) LTD induction showing that prenatal WIN exposure does not alter LTD in the female exposed group. In SHAM female: 2.350 ± 0.191 mV.ms before LTD versus 1.800 ± 0.182 mV.ms after LTD induction (n=6, p<0.05, paired ttest). In WIN female: 2.461 ± 0.439 mV.ms before LTD versus 1.833 ± 0.381 mV.ms after LTD induction (n=4, p<0.05, paired t-test). n=individual rats. Error bars indicate SEM. Figure 4: Sexual divergent ablation of LTD in the PFC of adult rats prenatally exposed to WIN. A, Average time-courses of mean field EPSPs showing that low-frequency stimulation (10 min at 10 Hz, arrow) induced LTD of evoked fEPSPs recorded in the mPFC in SHAM (n=7; white circles), but not in WIN (n=17; green circles) prenatally-exposed male rats. B, Individual experiments (grey) and group average (white represents SHAM; green represents WIN), before (baseline) and after (40 min) LTD induction showing the lack of LTD in the WIN exposed group. In SHAM rats: 0.185 ± 0.006 mV before LTD versus 0.150 ± 0.007 mV after LTD induction (n=7, p<0.05, paired t-test). In WIN animals: 0.197 ± 0.008 mV before LTD versus 0.189 ± 0.009 mV after LTD induction (n=17, p>0.05, paired t-test). C, Average timecourses of mean EPSPs showing similar low-frequency LTD in SHAM (n=5; white circles) and WIN (n=13; orange circles) in utero-exposed female rats. D, Individual experiments (grey) and group average (white represents SHAM; orange represents WIN), before (baseline) and after (40 min) LTD induction showing that prenatal WIN exposure does not alter LTD in the female exposed group. In SHAM female rats: 0.204 ± 0.009 mV before LTD versus 0.168 ± 0.010 mV after LTD induction (n=5, p<0.05, paired t-test). In WIN female rats: 0.255 ± 0.012 mV before LTD versus 0.214 ± 0.013 mV after LTD induction (n=13, p<0.05, paired t-test). *p<0.05. n=individual rats. Error bars indicate SEM. Figure 5: Sexual divergent ablation of LTD in the PFC of adult rats prenatally exposed to THC. A, Average time-courses of mean field EPSPs showing that low-frequency stimulation (10 min at 10 Hz, arrow) induced LTD of evoked fEPSPs recorded in the mPFC in SHAM (n=4; white circles), but not in THC (n=6; green circles) prenatally-exposed male rats. B, Individual experiments (grey) and group average (white represents SHAM; green represents THC), before (baseline) and after (40 min) LTD induction showing the lack of LTD in the THC exposed group. In SHAM rats: 0.239 ± 0.030 mV before LTD versus 0.202 ± 0.028 mV after LTD induction (n=4, p<0.05, paired t-test). In THC animals: 0.244 ± 0.023 mV before LTD versus 0.236 ± 0.019 mV after LTD induction (n=6, p>0.05, paired t-test). C. Average timecourses of mean EPSPs showing similar low-frequency LTD in SHAM (n=6; white circles) and THC (n=4; orange circles) in utero-exposed female rats. D, Individual experiments (grey) and group average (white represents SHAM; orange represents THC), before (baseline) and after (40 min) LTD induction showing that prenatal THC exposure does not alter LTD in the female exposed group. In SHAM female rats: 0.211 ± 0.011 mV before LTD versus 0.183 ± 0.010 mV after LTD induction (n=6, p<0.05, paired t-test). In THC female rats: 0.234 ± 0.030 mV before LTD versus 0.177 ± 0.030 mV after LTD induction (n=4, p<0.05, paired t-test). *p<0.05. n=individual rats. Error bars indicate SEM. induces LTP at layer V/VI mPFC synapses in both SHAM-(white circles, n=8) and WIN-(green circles, n=8) exposed male rats. D, Individual averaged fEPSP amplitude measurements before (baseline) and 40 min after LTP induction. Grey, individual experiments; tan, group averages. In SHAM rats: 0.138 ± 0.011 mV before LTP versus 0.185 ± 0.016 mV after LTP induction (n=8, p<0.05, paired t-test). In WIN male rats: 0.129 ± 0.012 mV before LTP versus 0.177 ± 0.015 mV after LTP induction (n=13, p<0.05, paired t-test). E, Identical TBS-LTP induction protocol but recording from female prenatally-exposed animals (SHAM, white circles, n=8; WIN, orange circles, n=8). F, Peak amplitude measurements before (baseline) and after (LTP) stimulation protocol from individual experiments (grey) and group averages in SHAM (white; 0.125 ± 0.014 mV before LTP versus 0.150 ± 0.020 mV after LTP; p<0.05, paired t-test) and WIN (orange; 0.133 ± 0.010 mV before LTP versus 0.167 ± 0.016 mV after LTP; p<0.05, paired t-test) animals. *p<0.05. n=individual rats. Error bars represent SEM.