DISCUSSION
Here we investigated, for the first time, the contribution of thePLCB1 gene to cocaine addictive properties using Plcb1+/-mice. We found that the heterozygous Plcb1 genotype resulted in a phenotype of resistance to cue-induced reinstatement of cocaine-seeking behavior. Furthermore, we found relevant transcriptomic differences inPlcb1+/- mice compared to WT after cue-induced reinstatement of cocaine seeking in mPFC, a brain area essential in relapse (Jasinska et al., 2015; Moorman et al., 2015). This study supports a role forPLCB1 in cocaine addiction, confirming previous findings in humans (Cabana-Domínguez et al., 2017), and suggests it may be relevant in relapse to cocaine addiction.
The phenotype of Plcb1+/- mice was characterized at different levels to evaluate consummatory and general behavior. In general, mutant mice presented normal levels of body weight, food and water intake, locomotor activity, and motor coordination, demonstrating that this animal model is valid to study the effects of the heterozygous deletion of the Plcb1 gene in other behavioral responses. Plcb1+/-showed increased anxiety in the elevated plus maze, reduced short-term memory in the novel object recognition paradigm, without any sign of depressive-like behavior in the anhedonia sucrose preference test. Thus, the single targeting of the Plcb1 gene has an impact on selective emotional and cognitive responses.
The short-term memory impairment observed in the mutants did not affect the acquisition of operant associative conditioning nor the instrumental learning that drives the goal-directed action, as shown by similar levels of operant cocaine self-administration and extinction learning than WT. In agreement, both genotypes trained with cocaine accomplish cue-induced reinstatement criterion with high enhancement of responding during this test compared to extinction. Thus, different brain circuits are involved in each kind of learning, with perirhinal-hippocampal structures (Winters et al., 2004) participating in short-term memory in the novel object recognition paradigm and mPFC-dorsal striatum pathway in cue-associated seeking. Also differences in each paradigm such as the acute retrieval or repetitive exposure involved respectively in each task may have a crucial influence in cognition. Hence, cocaine seeking can be multidimensional, involving different types of associative learning that together lead to an extensive repertoire of conditioned and instrumental responding.
The anxiogenic profile of Plcb1+/- is in accordance with the results previously reported in the Plcb4 -/- mice, which were associated with alterations in the cholinergic activity of the medial septum (Shin et al., 2009). However, the selective knock-down ofPlcb1 in the mPFC did not replicate this phenotype in a previous study (Kim et al., 2015), suggesting that other areas may be involved. The anxiogenic profile of Plcb1+/- mice had no effect on cocaine self-administration since the acquisition and extinction of this operant behavior was not modified in the mutants. Besides, mutants showed protection against cue-induced reinstatement of cocaine seeking, instead of the expected cocaine seeking promoted by an anxiogenic phenotype (Mantsch et al., 2016). Therefore, the association of an anxiogenic profile with a phenotype resilient to cocaine seeking in thePlcb1+/- mice suggests modifications in specific brain areas involved in cocaine relapse, such as the mPFC. Recently, a phenotype resilient to develop food addiction has been associated with increased strength of pyramidal glutamatergic synaptic transmission in the mPFC related to decreased compulsivity in the face of negative consequences (Domingo-Rodriguez et al., 2020) in mice with an anxiogenic profile (Lutz et al., 2015). Concerning cue-induced cocaine seeking, the mPFC has a crucial role and the network of glutamatergic projections from the prelimbic mPFC to the dorsal striatum participates in this phenotype (Lüscher et al., 2020). In our model, Plcb1 haploinsufficiency is linked to reduced cue-induced cocaine seeking possibly associated with modifications in this top-down corticolimbic brain network. Furthermore, these glutamatergic projections from mPFC to dorsal striatum receive mesocortical dopaminergic inputs from the VTA that could be crucially involved in the protective effects of Plcb1, since this gene plays an essential role in the dopaminergic signal transmission in the mPFC and many of the genes encoding for proteins in this pathway are differentially expressed in Plcb1+/- mice after cue-induced reinstatement (Figure 4).
Transcriptomic analyses performed after the cue-induced reinstatement of cocaine-seeking behavior revealed some of the molecular underpinnings that underlie the cocaine relapse-related phenotype observed in the mutant mice. Interestingly, differences in gene expression betweenPlcb1+/- and WT were predominantly found in the mPFC, whereas almost no differences were observed in the HPC. This evidence highlights mPFC as a key region to explain these differences in cocaine-seeking reinstatement observed in Plcb1+/- mice. Consistently, in DEGs in mPFC, we found enrichment on pathways that are essential for the development of addiction, including the dopaminergic neurotransmission (Figures 3-4) (Wise and Robble, 2020). PLCB1 is part of the dopamine-DARPP-32 signaling pathway (Figure 4), which plays a key role in cocaine reward (Borgkvist and Fisone, 2007; Nishi and Shuto, 2017). In a previous study, we also found this pathway enriched in DEGs in the frontal cortex and ventral striatum of mice that showed frustrated expected reward, produced by the cue in the absence of expected reward after a high level of effort in a progressive ratio schedule of reinforcement with palatable food (Martín-García et al., 2015b). Notably, Plcb1 was upregulated in the frontal cortex of those frustrated mice with increased responses to obtain the reward (palatable food). These data are in line with our findings in which decreased expression of Plcb1 (Plcb1 +/- mice) results in decreased cue-induced responses to obtain cocaine after extinction. All these results support the participation of Plcb1 in cue-induced cocaine seeking revealed in the present study and the involvement of mPFC .
Our transcriptomic analyses also found enrichment in genes related to learning and memory processes, such as LTP in the mPFC. The transition from drug use to drug addiction is a maladaptive process that directly affects learning and memory (Torregrossa et al., 2011; Goodman and Packard, 2016). LTP produces long-lasting activity-dependent synaptic modifications that underlie memory, and drugs of abuse alter this mechanism in several brain regions such as mPFC (Huang et al., 2007; Lu et al., 2010), mesocorticolimbic system (Thomas et al., 2008), VTA (Placzek et al., 2016) and hippocampus (Keralapurath et al., 2017; Preston et al., 2019), among others. Importantly, genes related to the glutamatergic system such as Gria2-4 (glutamate ionotropic receptor AMPA2-4 (alpha 2-4)) and Grik2 (ionotropic glutamate receptor kainate 2 (beta 2)) were found to be upregulated inPlcb1+/- mice. Meanwhile, Grin1, 2c, and 2d (glutamate ionotropic receptor NMDA1 (zeta 1, epsilon3 and epsilon4)) were downregulated, suggesting an increase in AMPAR/NMDAR ratio in cocaine-experienced and abstinent Plcb1 +/- mice as compared to WT, one of the key indicators of LTP induction and increased synaptic strength (Lüscher and Malenka, 2011). Furthermore, we found enrichment in other neuroplasticity (synapse organization, neuron projection development, and axon guidance) (Bahi and Dreyer, 2005; O’Brien, 2009; Cooper et al., 2017; Dong et al., 2017) and signaling pathways (MAPK, mTOR, and neurotrophin) (Li and Wolf, 2015; García-Pardo et al., 2016) related to addiction. The assessment of gene expression in the mPFC of the Plcb1+/- mice highlighted alterations in relevant pathways for the addictive process, which could contribute to the results observed in the cue-induced reinstatement.
Nowadays, there are few effective pharmacological treatments available for cocaine use disorders, and frequently, psychosocial interventions in combination with pharmacotherapy are needed (Kampman, 2019). Studies performed in mice have pinpointed multiple potential therapeutic approaches for the reinstatement of cocaine-seeking behavior, targeting the reward circuit (Buchta et al., 2020). In humans, treatment with bupropion, topiramate, or disulfiram has been widely used. Bupropion, a non-tricyclic antidepressant that inhibits dopamine and norepinephrine reuptake, is effective in reducing craving (Frishman, 2007). Topiramate, a GABA/glutamatergic medication, has also been used to treat cocaine use disorder as it reduces the activity of the mesocorticolimbic dopaminergic system (Kampman, 2019). A completely different strategy is the use of disulfiram which potentially inhibits the oxidoreductase dopamine β‐hydroxylase (DβH, encoded by the DBH gene), which converts dopamine to norepinephrine (Kampangkaew et al., 2019). Significantly, a genetic variation in the SLC6A3 gene (encoding DAT) has been associated with disulfiram treatment for cocaine addiction, with patients with higher DAT levels having better treatment outcomes than those with lower DAT levels (Kampangkaew et al., 2019). Thus, further studies and new therapeutic targets are needed to obtain effective treatment for cocaine addiction. The results obtained in the present study underscore the relevance of the Plcb1 gene in the cue-induced reinstatement of cocaine seeking after extinction. Together with previous findings in humans (Drgon et al., 2012; Cabana-Domínguez et al., 2017) and mice (Eipper-Mains et al., 2013), PLCB1 merits to be further evaluated as a promising novel therapeutic target for preventing relapse and treating cocaine addiction .
The experimental approach used in our study, the Plcb1+/- mouse model, allowed us to reproduce better the molecular context observed in humans in comparison to the use of a complete KO mouse (Hall et al., 2013), as these animals preserved, at least, half of the expression ofPlcb1 . However, the haploinsufficiency of Plcb1 during neurodevelopment in these animals could produce alterations that may contribute to the phenotype observed in the present study. Nevertheless, this may also be similar in humans with genetic risk variants that decrease PLCB1 expression. Therefore, this approach is appropriate to study a specific genetic alteration that confers susceptibility to drug addiction and to delineate the precise contribution of PLCB1.
To sum up, we studied, for the first time, the contribution of thePLCB1 gene to cocaine addictive properties using Plcb1+/-mice. Previous studies have revealed an up-regulation of Plcb1 in brain areas related to the reward circuit after cocaine exposure in animals (Eipper-Mains et al., 2013) and in human cocaine abusers (Cabana-Domínguez et al., 2017). These changes, together with our results, suggest that cocaine increases the expression of Plcb1,and this mechanism plays an essential role in cocaine addiction, as revealed now by the resistance to cue-induced reinstatement of cocaine-seeking behavior exhibited by Plcb1+/- mutant mice. These results highlight the importance of the Plcb1 gene in the development of cocaine addiction and relapse and pinpoint PLCB1 as a promising therapeutic target for cocaine addiction and perhaps other types of addiction.