Drug initiation, including the first experimentation with a rewarding substance, sets the stage for subsequent behavioral patterns. Even a single encounter with a substance can induce enduring changes in the central nervous system that persist beyond the immediate presence of the drug (1, 2, 3, 4). This holds true during cannabis initiation, where the psychoactive properties of the plant trigger a cascade of biochemical, neuronal, and synaptic consequences. Novel psychoactive substances, a rising public health concern, emulate naturally existing molecules or conventional illicit psychoactive substances (5,6). Synthetic cannabinoids (SCs; also known as Spice) are intentionally designed to mimic the active components in natural cannabis (5,7). Recreational use of SCs has increased considerably in recent years, and the prevalence of SC use tends to be higher among young males (5,7, 8, 9).

Considering the acknowledged risks associated with SC use during adolescence (8,10) and the substantial prevalence of experimentation during this critical developmental period, it is crucial to address the lack of understanding regarding the effects of SCs initiation on central synapses. This knowledge void is especially critical in relation to the prefrontal cortex (PFC). The PFC is a brain region essential for decision making, impulse control, and social interaction, which undergoes significant development throughout adolescence and early adulthood in both humans and rodents, making it one of the last structures in the central nervous system to mature (11,12). This developmental period is characterized by profound changes in PFC structure and function, including the refinement of PFC connectivity and the maturation of cognitive, emotional, and social abilities in both species (13,14). Concurrently, adolescence is marked by increased exposure to new and diverse environmental factors, which extends the vulnerability of the PFC to various environmental insults, notably exogenous cannabinoids (10,15,16). Exposure to cannabinoids during adolescence causes lasting alterations in PFC-dependent behaviors, as well as abnormalities in prefrontal architecture and synaptic functions (17, 18, 19, 20, 21). The endocannabinoid (eCB) system, which includes G protein–coupled cannabinoid CB₁ and CB₂ receptors (CB1R and CB2R) and endogenous cannabinoid ligands, among other components, plays a significant role in these processes (22). The PFC exhibits a high density of CB1Rs (23,24), rendering it susceptible to eCB synaptic dysfunctions from SCs that result in various cognitive alterations (25). To understand the relationship between SC exposure during the critical period of adolescence, altered behavior, and abnormalities in the prefrontal circuitry, identifying the specific molecular mechanisms involved is necessary.

Reelin, a secreted glycoprotein of the extracellular matrix, plays a pivotal role in neuronal migration and layer formation during embryonic development (26). As a synaptic organizer, reelin is essential for the proper functioning of both postnatal and adult brain physiology. Reelin regulates the developmental trajectory of pyramidal PFC neurons and associated behaviors (27, 28, 29). During adolescence, reelin haploinsufficiency has adverse effects on synaptic transmission, excitatory-inhibitory balance, plasticity, and PFC-dependent behaviors (27,29). Furthermore, reelin is a critical molecular mediator of PFC dysfunction due to nutritional stress (15) and influences behavioral abnormalities associated with high Δ⁹-tetrahydrocannabinol (THC) use in adolescence (30). Considering the vulnerability of reelin to early-life environmental insults (15,31, 32, 33, 34), we hypothesized that it could serve as a link between initiation of SC use during adolescence and prefrontal dysfunctions. Because SCs consumption is more prevalent among young men (7,8), we focused our study on investigating the effects of SC exposure on reelin expression and PFC functioning in male adolescent mice.

Results

Effect of a Single In Vivo SC Exposure on Reelin’s Cellular Density in Specific PFC Layers

Cannabis and SCs acutely impair cognitive functions such as attention, memory, and decision making in humans and animals (10); however, research on persistent effects (i.e., one or more days after use) of the first exposure are limited (2,21,39,40). We hypothesized that cannabinoids may interact with the reelin system in the PFC. We chose to utilize the SC WIN 55,212-2 (WIN) instead of THC, the principle psychoactive component of cannabis, due to its increased potency for CB1R, as well as the public health concerns associated with SCs. Adolescent male mice were administered a single intraperitoneal injection of WIN or its vehicle, and the reelin expression pattern in the PFC was studied the following day by immunofluorescence. SC exposure led to an 18.2% ± 1.6% reduction in the density of reelin-positive cells throughout the entire PFC compared with naïve and vehicle-treated mice (Figure 1A). To investigate whether the observed effect was consistent across the PFC layers, we examined the cumulative distributions of reelin-positive cell densities in naïve mice, vehicle-treated mice, and WIN-treated mice. In layer 1, these distributions were superimposable, indicating no WIN effect in the most superficial layer of the PFC (Figure 1B). Conversely, in layers 2/3 and 5/6, the WIN-treated group distribution shifted to the left compared with naïve and vehicle-treated mice, suggesting decreased reelin-positive cell density in these layers due to WIN exposure (Figure 1B). Co-administration of the CB1R antagonist SR141716A with WIN prevented the total density reduction of reelin-positive cells (Figure 1A) and normalized the density distributions in layers 2/3 and 5/6, as evidenced by the overlap of SR+WIN, vehicle, and naïve cumulative probability curves (Figure 1B). Therefore, a single administration of an SC to a drug-naïve adolescent male mouse significantly disrupted reelin PFC expression in a layer-specific pattern through the activation of CB1Rs.

RNAscope Profiling Reveals a Significant Prevalence of CB1R and Reelin Coexpression in PFC Neurons

The layer-specific decrease in reelin levels induced by the SC prompted us to investigate the expression profile of reelin in the PFC, particularly its spatial and cellular association with the CB1R. SmFISH analysis was used to explore the specific RNA coexpression patterns of reelin (Reln) and various neuronal markers in the PFC of naïve male mice, covering layers 2 to 6 (Figure 2 and Figure S1). We found that Reln-positive neurons are molecularly heterogeneous, expressing GABAergic (Slc32a1) or glutamatergic (Slc17a7) markers (Figure 2). In terms of distribution, Reln-positive cells are primarily positive for Slc32a1, and their presence among GABAergic or glutamatergic populations is similar in both superficial layers (layers 2/3) and deep layers (layers 5/6) of the PFC (Figure 2A). Reln is coexpressed with the neuropeptide cholecystokinin (Cck), and the proportion of Reln neurons positive for Cck was similar across layers (Figure 2B). Considering the observed reduction in reelin expression upon CB1R activation (Figure 1), the significant proportion of Reln-positive neurons expressing Cck in the PFC (Figure 2B), and previous studies showing a strong coexpression of CCK with CB1R in various forebrain regions (41,42), we investigated the presence of CB1Rs (Cnr1) in Reln-positive neurons. The CB1R was expressed across all layers 2 to 6, and Reln demonstrated a high level of coexpression with the Cnr1 in both superficial and deep layers (Figure 2C). Additionally, reelin was found to be expressed with neuropeptide Y (NPY) and somatostatin (Sst) (Figure S1).

These findings demonstrate a diverse array of Reln-positive neurons in the layers of the adolescent mouse PFC, exhibiting unique coexpression patterns with the Cnr1 and/or Cck. Furthermore, our results revealed the presence of Reln in glutamatergic neurons within the PFC. Given the observed coexpression of the CB1R and reelin in prefrontal neurons, next we investigated how adolescent cannabinoid exposure interferes with reelin expression and function in the PFC.

Cannabinoid-Induced Reduction in PFC Reelin Expression Is Selectively Mediated Through the Activation of CB1Rs on GABAergic Neurons

We found that the reduction in PFC reelin-positive cell density following WIN exposure correlated with alterations in reelin cellular protein levels. Initially, Western blot analysis of whole PFC lysates using the anti-reelin G10 antibody revealed that WIN treatment decreased reelin levels compared with mice treated with vehicle alone. Importantly, this effect was effectively blocked by the CB1R antagonist SR141716A (Figure 3A). Furthermore, densitometric analysis of total reelin levels in mice exposed to the SC revealed that the reduction of reelin levels induced by WIN was not present in global CB1R knockout mice (CB1R-KO) in comparison to their wild-type littermates (Figure 3B). These findings, which complement the immunostaining observations (Figure 1), demonstrate that in the PFC, the SC acted on CB1Rs to decrease reelin protein levels.

Next, we aimed to obtain genetic confirmation substantiated by our smFISH analysis suggesting that most neurons coexpressing reelin and CB1Rs in the PFC are of GABAergic nature (Figure 2). To specify the cell types of CB1R-expressing neurons responsible for the reduction in reelin expression, we utilized conditional knockouts targeting either CB1Rs in forebrain GABAergic (i.e., Dlx5/6-CB1R-KO mice, named GABA-CB1R-KO) or cortical glutamatergic (i.e., Nex-CB1R-KO mice, named Glu-CB1R-KO) neurons (43, 44, 45). The data showed that the effects of WIN on the day after were preserved in Glu-CB1R-KO mice, whereas these effects were abolished in GABA-CB1R-KO mice (Figure 3C). These findings strongly suggest that the reduction of prefrontal reelin expression by cannabinoids relies on the activation of the CB1R present on GABAergic neurons.

Single SC Exposure Modifies the Proteolytic Fragmentation of Reelin Rather Than Its Gene Expression

In the developing and adult brains, the full-length reelin protein (with relative molecular masses of 388 kDa and 450 kDa when glycosylated) (46) undergoes secretion and enzymatic cleavage at 2 primary sites, producing 5 fragments (47,48). In addition to the full-length reelin, the G10 antibody can detect 2 N-terminal region fragments (38,49,50). Analysis of the forms recognized by the G10 antibody revealed reduced amounts of full-length protein and the 320-kDa cleavage product in the WIN-exposed group (Figure 4A), suggesting diminished secretion and impacted proteolytic cleavage following CB1R activation. To further elucidate the mechanism that underlies reelin downregulation in SC-exposed mice, we performed quantitative polymerase chain reaction on the PFC from naïve, vehicle-, and WIN-exposed mice. Despite the delta delta CT variance, the study is adequately powered (1 − beta = 0.97) to detect a major effect in transcripts levels. No differences in Reln messenger RNA (mRNA) levels were observed across groups (Figure 4B), suggesting that adolescent cannabinoid initiation did not significantly influence reelin transcription. Taken together, the data support the idea that SC reduces reelin levels by disrupting its proteolytic cleavage rather than by altering its transcription.

Single In Vivo SC Exposure Durably Reduces PFC Long-Term Potentiation

Altered PFC long-term synaptic plasticity is commonly observed in various models of neurodevelopmental diseases, perinatal insults, and adolescent insults (15,16,21). In a previous study using reelin haploinsufficient mice, we demonstrated the abolishment of theta burst stimulation (TBS)–induced long-term potentiation (TBS-LTP) in the PFC (28). Building on this, we hypothesized that the reduction of reelin levels, triggered by SC and mediated through the CB1R, impacts TBS-LTP.

We observed a significant reduction in the magnitude of TBS-LTP at deep PFC excitatory synapses following a single injection of WIN, compared with naïve and vehicle-treated mice, the day after SC administration (Figure 5A–C). Notably, the LTP reduction was absent when the SC was administered together with a CB1R antagonist (Figure 5A–C), suggesting that CB1R activation plays a crucial role in the detrimental effects of WIN on prefrontal LTP. Notably, the decrease in TBS-LTP caused by a single SC exposure remained evident even on the fourth day following SC administration, indicating that the impact of in vivo SC on synaptic plasticity persisted long after the first encounter with the drug (Figure 5D). Variations in the excitability of layer 3/5 excitatory synapses cannot account for this effect because the input-output curves were found to be identical across all experimental groups (Figure S2).

In the PFC, TBS-LTP relies on NMDA receptor (NMDAR) activation (29). To explore whether the effects of WIN on deep layer excitatory synapses were associated with changes in NMDARs, we assessed the transcriptional levels of NMDAR subunits using quantitative polymerase chain reaction analysis. The results revealed no significant differences in mRNA levels of NMDAR subunits in the PFC of WIN-treated mice compared with the control groups (Figure S3), indicating that the reduction in TBS-LTP was not a result of altered transcription of NMDAR subunits.

Reelin Deficiency Underlies SC-Induced Impairment of Synaptic Plasticity

Our results thus far indicate that a single SC administration to drug-naïve adolescent male mice diminishes the density of reelin-positive cells in specific cortical layers through CB1R-mediated modification of reelin secretion and/or proteolytic cleavage. Furthermore, this SC exposure persistently impairs LTP.

To establish a functional relationship between CB1R-induced reelin reduction and LTP impairment, 2 specific predictions need to be satisfied: 1) LTP is impaired in another model of reelin reduction, and 2) intra-PFC supplementation of reelin will protect from the loss of LTP induced by SC exposure.

To address the first prediction, we investigated TBS-induced plasticity in a mutant mouse haploinsufficient for reelin (HRM) (28), aiming to explore the connection between reelin deficiency and decreased TBS-LTP after SC administration. We found that the time course and magnitude of TBS-LTP in adolescent HRM male mice closely resembled those observed in animals exposed to WIN (percentage of TBS-LTP: 12.2% ± 3.7%, n = 9 WIN-treated vs. 14.4% ± 4.1%, n = 9 HRM) (Figure 6A, B). Thus, a genetic reduction of endogenous reelin mimics the harmful effects of SC on PFC plasticity. Therefore, the alteration of TBS-induced plasticity following in vivo exposure to WIN is consistent with a decrease in endogenous reelin levels triggered by WIN exposure.

To validate that the decline in TBS-LTP after exposure to WIN was caused by reduced reelin expression, recombinant reelin (rRln) was bilaterally injected into the PFC, as we had done previously (15). Adolescent mice received WIN injections 2 days after rRln administration, and TBS-LTP was evaluated 17 ± 1 hours after WIN exposure. Administering rRln into the PFC completely prevented the WIN-induced decrease in TBS-LTP, restoring LTP magnitude to normal levels (Figure 6C, D). Importantly, this effect was not attributed to the influence of rRln on the excitability of layer 3/5 excitatory synapses because input-output curves were identical among all groups (Figure S4). These results further support the observation that reelin downregulation functionally contributes to the synaptic impairments induced by SC exposure.