Towards a machine-learning assisted diagnosis of psychiatric disorders and their operationalization in preclinical research: Evidence from studies on addiction-like behaviour in individual rats.

Over the last few decades, there has been a progressive transition from a categorical to a dimensional approach to psychiatric disorders. Especially in the case of substance use disorders, interest in the individual vulnerability to transition from controlled to compulsive drug taking warrants the development of novel dimension-based objective stratification tools. Here we drew on a multidimensional preclinical model of addiction, namely the 3-criteria model, previously developed to identify the neurobehavioural basis of the individual's vulnerability to switch from controlled to compulsive drug taking, to test a machine-learning assisted classifier objectively to identify individual subjects as vulnerable/resistant to addiction. Datasets from our previous studies on addiction-like behaviour for cocaine or alcohol were fed into a variety of machine-learning algorithms to develop a classifier that identifies resilient and vulnerable rats with high precision and reproducibility irrespective of the cohort to which they belong. A classifier based on K-median or K-mean-clustering (for cocaine or alcohol, respectively) followed by artificial neural networks emerged as a highly reliable and accurate tool to predict if a single rat is vulnerable/resilient to addiction. Thus, each rat previously characterized as displaying 0-criterion (i.e., resilient) or 3-criteria (i.e., vulnerable) in individual cohorts was correctly labelled by this classifier. The present machine-learning-based classifier objectively labels single individuals as resilient or vulnerable to developing addiction-like behaviour in a multisymptomatic preclinical model of addiction-like behaviour in rats. This novel dimension-based classifier increases the heuristic value of these preclinical models while providing proof of principle to deploy similar tools for the future of diagnosis of psychiatric disorders.

The temporal origin of dentate granule neurons dictates their role in spatial memory.

The dentate gyrus is one of the only brain regions that continues its development after birth in rodents. Adolescence is a very sensitive period during which cognitive competences are programmed. We investigated the role of dentate granule neurons (DGNs) born during adolescence in spatial memory and compared them with those generated earlier in life (in embryos or neonates) or during adulthood by combining functional imaging, retroviral and optogenetic tools to tag and silence DGNs. By imaging DGNs expressing Zif268, a proxy for neuronal activity, we found that neurons generated in adolescent rats (and not embryos or neonates) are transiently involved in spatial memory processing. In contrast, adult-generated DGNs are recruited at a later time point when animals are older. A causal relationship between the temporal origin of DGNs and spatial memory was confirmed by silencing DGNs in behaving animals. Our results demonstrate that the emergence of spatial memory depends on neurons born during adolescence, a function later assumed by neurons generated during adulthood.

Depleting adult dentate gyrus neurogenesis increases cocaine-seeking behavior.

The hippocampus is the main locus for adult dentate gyrus (DG) neurogenesis. A number of studies have shown that aberrant DG neurogenesis correlates with many neuropsychiatric disorders, including drug addiction. Although clear causal relationships have been established between DG neurogenesis and memory dysfunction or mood-related disorders, evidence of the causal role of DG neurogenesis in drug-seeking behaviors has not been established. Here we assessed the role of new DG neurons in cocaine self-administration using an inducible transgenic approach that selectively depletes adult DG neurogenesis. Our results show that transgenic mice with decreased adult DG neurogenesis exhibit increased motivation to self-administer cocaine and a higher seeking response to cocaine-related cues. These results identify adult hippocampal neurogenesis as a key factor in vulnerability to cocaine addiction.

Not all smokers appear to seek nicotine for the same reasons: implications for preclinical research in nicotine dependence.

Tobacco use leads to 6 million deaths every year due to severe long-lasting diseases. The main component of tobacco, nicotine, is recognized as one of the most addictive drugs, making smoking cessation difficult, even when 70 percent of smokers wish to do so. Clinical and preclinical studies have demonstrated consistently that nicotine seeking is a complex behavior involving various psychopharmacological mechanisms. Evidence supports that the population of smokers is heterogeneous, particularly as regards the breadth of motives that determine the urge to smoke. Here, we review converging psychological, genetic and neurobiological data from clinical and preclinical studies supporting that the mechanisms controlling nicotine seeking may vary from individual to individual. It appears timely that basic neuroscience integrates this heterogeneity to refine our understanding of the neurobiology of nicotine seeking, as tremendous progress has been made in modeling the various psychopharmacological mechanisms driving nicotine seeking in rodents. For a better understanding of the mechanisms that drive nicotine seeking, we emphasize the need for individual-based research strategies in which nicotine seeking, and eventually treatment efficacy, are determined while taking into account individual variations in the mechanisms of nicotine seeking.

Individual variations in motives for nicotine self-administration in male rats: evidence in support for a precision psychopharmacology.

The significant heterogeneity in smoking behavior among smokers, coupled with the inconsistent efficacy of approved smoking cessation therapies, supports the presence of individual variations in the mechanisms underlying smoking. This emphasizes the need to shift from standardized to personalized smoking cessation therapies. However, informed precision medicine demands precision fundamental research. Tobacco smoking is influenced and sustained by diverse psychopharmacological interactions between nicotine and environmental stimuli. In the classical experimental rodent model for studying tobacco dependence, namely intravenous self-administration of nicotine, seeking behavior is reinforced by the combined delivery of nicotine and a discrete cue (nicotine+cue). Whether self-administration behavior is driven by the same psychopharmacological mechanisms across individual rats remains unknown and unexplored. To address this, we employed behavioral pharmacology and unbiased cluster analysis to investigate individual differences in the mechanisms supporting classical intravenous nicotine self-administration (0.04 mg/kg/infusion) in male outbred Sprague-Dawley rats. Our analysis identified two clusters: one subset of rats sought nicotine primarily for its reinforcing effects, while the second subset sought nicotine to enhance the reinforcing effects of the discrete cue. Varenicline (1 mg/kg i.p.) reduced seeking behavior in the former group, whereas it tended to increase in the latter group. Crucially, despite this fundamental qualitative difference revealed by behavioral manipulation, the two clusters exhibited quantitatively identical nicotine+cue self-administration behavior. The traditional application of rodent models to study the reinforcing and addictive effects of nicotine may mask individual variability in the underlying motivational mechanisms. Accounting for this variability could significantly enhance the predictive validity of translational research.

Integrating operant behavior and fiber photometry with the open-source python library Pyfiber.

Despite the popularity of fiber photometry (FP), its integration with operant behavior paradigms is progressing slowly. This can be attributed to the complex protocols in operant behavior - resulting in a combination of diverse non-predictable behavioral responses and scheduled events, thereby complicating data analysis. To overcome this, we developed Pyfiber, an open-source python library which facilitates the merge of FP with operant behavior by relating changes in fluorescent signals within a neuronal population to behavioral responses and events. Pyfiber helps to 1. Extract events and responses that occur in operant behavior, 2. Extract and process the FP signals, 3. Select events of interest and align them to the corresponding FP signals, 4. Apply appropriate signal normalization and analysis according to the type of events, 5. Run analysis on multiple individuals and sessions, 6. Collect results in an easily readable format. Pyfiber is suitable for use with many different fluorescent sensors and operant behavior protocols. It was developed using Doric lenses FP systems and Imetronic behavioral systems, but it possesses the capability to process data from alternative systems. This work sets a solid foundation for analyzing the relationship between different dimensions of complex behavioral paradigms with fluorescent signals from brain regions of interest.

A decrease in gamma-synuclein expression within the nucleus accumbens increases cocaine intravenous self-administration in the rat.

Except as a marker of cancer progression, gamma-synuclein (GSyn) had received little attention. Recent data showed however that GSyn modulates cocaine-induced locomotor effects, suggesting that it could also play a role in cocaine reinforcing effects. In the rat, siRNAs targeting GSyn expression were injected in the nucleus accumbens and cocaine reinforcing effects were evaluated by means of intravenous self-administration. A dose-response curve was followed by procedures of progressive ratio, extinction, cocaine- and cue-induced reinstatements. Decrease of GSyn expression increased self-administration over a large range of doses. This effect was associated with an increase in cocaine-induced reinstatement. The present data reveal that GSyn exert a specific negative control on cocaine-induced reinforcing and incentive effects.

Nicotine inhibits the VTA-to-amygdala dopamine pathway to promote anxiety.

Nicotine stimulates dopamine (DA) neurons of the ventral tegmental area (VTA) to establish and maintain reinforcement. Nicotine also induces anxiety through an as yet unknown circuitry. We found that nicotine injection drives opposite functional responses of two distinct populations of VTA DA neurons with anatomically segregated projections: it activates neurons that project to the nucleus accumbens (NAc), whereas it inhibits neurons that project to the amygdala nuclei (Amg). We further show that nicotine mediates anxiety-like behavior by acting on ß2-subunit-containing nicotinic acetylcholine receptors of the VTA. Finally, using optogenetics, we bidirectionally manipulate the VTA-NAc and VTA-Amg pathways to dissociate their contributions to anxiety-like behavior. We show that inhibition of VTA-Amg DA neurons mediates anxiety-like behavior, while their activation prevents the anxiogenic effects of nicotine. These distinct subpopulations of VTA DA neurons with opposite responses to nicotine may differentially drive the anxiogenic and the reinforcing effects of nicotine.

Mifepristone and spironolactone differently alter cocaine intravenous self-administration and cocaine-induced locomotion in C57BL/6J mice.

Corticosterone, the main glucorticoid hormone in rodents, facilitates behavioral responses to cocaine. Corticosterone is proposed to modulate cocaine intravenous self-administration (SA) and cocaine-induced locomotion through distinct receptors, the glucocorticoid receptor (GR) and the mineralocorticoid receptor (MR), respectively. However, this remains debatable. On one hand, modulation of both responses by the GR was tested in different experimental conditions, i.e. light versus dark nycthemeral phase and naïve versus cocaine-experienced animals. On the other hand, modulation of both responses by the MR was never tested directly but only inferred based on the ability of low plasma corticosterone levels (those for which corticosterone almost exclusively binds the MR) to compensate the effects of adrenalectomy. Our goal here was to test the involvement of the GR and the MR in cocaine-induced locomotor and reinforcing effects in the same experimental conditions. C57Bl/6J mice were trained for cocaine (1 mg/kg/infusion) intravenous SA over 40 SA sessions. The animals were then administered with mifepristone (30 mg/kg i.p.), a GR antagonist, or with spironolactone (20 mg/kg/i.p.), an MR antagonist, 2 hours before either cocaine intravenous SA or cocaine-induced locomotion. In a comparable nycthemeral period and in similarly cocaine-experienced animals, a blockade of the GR decreased cocaine-induced reinforcing effects but not cocaine-induced locomotion. A blockade of the MR decreased both cocaine-induced reinforcing (but to a much lesser extent than the GR blockade) and locomotor effects. Altogether, our results comforted the hypothesis that the GR modulates cocaine-related operant conditioning, while the MR would modulate cocaine-related unconditioned effects. The present data also reveal mifepristone as an interesting tool for manipulating the impact of corticosterone on cocaine-induced reinforcing effects in mice.

Varenicline Targets the Reinforcing-Enhancing Effect of Nicotine on Its Associated Salient Cue During Nicotine Self-administration in the Rat.

Nicotine is acknowledged as the key addictive compound of tobacco. Varenicline (Champix® or Chantix®), mainly acting as a partial agonist at the α4ß2 nicotinic receptor, is an approved smoking cessation pharmacotherapy, although with efficacy limited to a portion of smokers. Smokers differ in the motives that drive their drug seeking and Varenicline might be more efficient in some groups more than others. Studies in rodents revealed that nicotine-seeking is strongly supported by complex interactions between nicotine and environmental cues, and notably the ability of nicotine to enhance the reinforcing properties of salient environmental stimuli. It is not yet understood whether the decrease of nicotine-seeking by acute Varenicline in rats results from antagonism of the primary reinforcing effects of nicotine, of the reinforcement-enhancing effect of nicotine on cues, or of a combination of both. Thanks to a protocol that allows assessment of the reinforcement-enhancing effect of nicotine on cues during self-administration in rats, we showed that Varenicline targets both nicotine reinforcing effects and reinforcement-enhancing effect of nicotine on cues. Importantly, individual variations in the latter determined the amplitude of acute Varenicline-induced decrease in seeking. These results suggest that Varenicline might be more beneficial in smokers who are more sensitive to nicotine effects on surrounding stimuli.

Gene-environment interactions in vulnerability to cocaine intravenous self-administration: a brief social experience affects intake in DBA/2J but not in C57BL/6J mice.

RATIONALE: Individual differences in cocaine-taking behavior and liability to develop abuse are clearly observed, but underlying mechanisms are still poorly understood. A role for gene-environment interactions has been proposed but remains hypothetical. OBJECTIVES: We investigated whether gene-environment interactions influence intravenous cocaine self-administration (SA) in mice. We tested the effect of a past short group housing experience on cocaine SA in two inbred strains of mice, the C57BL/6J (C57) and DBA/2J (DBA). METHODS: Adult C57 and DBA mice were individually housed upon arrival in the laboratory. After 3 weeks, half of the animals of each strain were group housed for 19 days. One week after the end of group housing, cocaine SA or measurement of brain cocaine levels took place. RESULTS: Individually and ex-group-housed C57 mice did not differ for cocaine SA. On the contrary, the ex-group-housed DBA mice showed an upward shift in the dose-response curve as compared to individually housed DBA. Differences in brain cocaine levels could not account for the observed behavioral differences. CONCLUSIONS: These results demonstrate that vulnerability to cocaine reinforcing effects can be affected by gene-environment interactions. We propose a mouse model for the characterization of gene-environment interactions in the vulnerability to cocaine-taking behavior.

Preexposure during or following adolescence differently affects nicotine-rewarding properties in adult rats.

RATIONALE: Many people come in contact with psychoactive drugs, yet not all of them become addicts. Epidemiology shows that a late approach with cigarette smoking is associated with a lower probability to develop nicotine dependence. Exposure to nicotine during periadolescence, but not similar exposure in the postadolescent period, increases nicotine self-administration in rats, but underlying mechanisms remain poorly understood. OBJECTIVE: We investigated whether exposure to nicotine during or after adolescence would alter rewarding properties of the same drug at adulthood, as assessed by place conditioning. MATERIALS AND METHODS: Periadolescent (PND 34-43) or postadolescent (PND 60-69) rats were injected with saline or nicotine (0.4 mg kg(-1)) for 10 days. The rats received three pairings with saline and three pairings with nicotine (0, 0.3, or 0.6 mg kg(-1)) 5 weeks after pretreatment. The rats were then tested for place conditioning in a drug-free state. RESULTS: Upon first exposure to the apparatus, animals pretreated with nicotine during adolescence showed elevated novelty-induced activation. The 0.3 (but not the 0.6) mg kg(-1) dose failed to produce both ongoing locomotor sensitization and place conditioning in animals pretreated with nicotine following adolescence. This suggests a rightward shift in the dose-response curve, namely, a reduced efficacy of nicotine. Conversely, the same dose was effective in saline-pretreated controls and noteworthy in rats pretreated during adolescence. CONCLUSION: Exposure following the adolescent period might diminish the risk to develop nicotine dependence. As for human implications, findings are consistent with a reduced vulnerability to nicotine addiction in people who start smoking late in their life.

Differential Control of Cocaine Self-Administration by GABAergic and Glutamatergic CB1 Cannabinoid Receptors.

The type 1 cannabinoid receptor (CB1) modulates numerous neurobehavioral processes and is therefore explored as a target for the treatment of several mental and neurological diseases. However, previous studies have investigated CB1 by targeting it globally, regardless of its two main neuronal localizations on glutamatergic and GABAergic neurons. In the context of cocaine addiction this lack of selectivity is critical since glutamatergic and GABAergic neuronal transmission is involved in different aspects of the disease. To determine whether CB1 exerts different control on cocaine seeking according to its two main neuronal localizations, we used mutant mice with deleted CB1 in cortical glutamatergic neurons (Glu-CB1) or in forebrain GABAergic neurons (GABA-CB1). In Glu-CB1, gene deletion concerns the dorsal telencephalon, including neocortex, paleocortex, archicortex, hippocampal formation and the cortical portions of the amygdala. In GABA-CB1, it concerns several cortical and non-cortical areas including the dorsal striatum, nucleus accumbens, thalamic, and hypothalamic nuclei. We tested complementary components of cocaine self-administration, separating the influence of primary and conditioned effects. Mechanisms underlying each phenotype were explored using in vivo microdialysis and ex vivo electrophysiology. We show that CB1 expression in forebrain GABAergic neurons controls mouse sensitivity to cocaine, while CB1 expression in cortical glutamatergic neurons controls associative learning processes. In accordance, in the nucleus accumbens, GABA-CB1 receptors control cocaine-induced dopamine release and Glu-CB1 receptors control AMPAR/NMDAR ratio; a marker of synaptic plasticity. Our findings demonstrate a critical distinction of the altered balance of Glu-CB1 and GABA-CB1 activity that could participate in the vulnerability to cocaine abuse and addiction. Moreover, these novel insights advance our understanding of CB1 neuropathophysiology.

Evidence for enhanced neurobehavioral vulnerability to nicotine during periadolescence in rats.

Epidemiological studies indicate that there is an increased likelihood for the development of nicotine addiction when cigarette smoking starts early during adolescence. These observations suggest that adolescence could be a "critical" ontogenetic period, during which drugs of abuse have distinct effects responsible for the development of dependence later in life. We compared the long-term behavioral and molecular effects of repeated nicotine treatment during either periadolescence or postadolescence in rats. It was found that exposure to nicotine during periadolescence, but not a similar exposure in the postadolescent period, increased the intravenous self-administration of nicotine and the expression of distinct subunits of the ligand-gated acetylcholine receptor in adult animals. Both these changes indicated an increased sensitivity to the addictive properties of nicotine. In conclusion, adolescence seems to be a critical developmental period, characterized by enhanced neurobehavioral vulnerability to nicotine.

The glucocorticoid receptor as a potential target to reduce cocaine abuse.

Several findings suggest that glucocorticoid hormones are involved in determining the propensity of an individual to develop cocaine abuse. These hormones activate two related transcription factors, the glucocorticoid receptor (GR) and the mineralocorticoid receptor. In this study, we show that the selective inactivation of the GR gene in the brains of mice profoundly flattened the dose-response function for cocaine intravenous self-administration and suppressed sensitization, two experimental procedures considered relevant models of addiction. Furthermore, administration of a GR antagonist dose-dependently reduced the motivation to self-administer cocaine. Importantly, the absence of GR did not modify the basal behavioral and molecular effects of cocaine but selectively modified the excessive response to the drug spontaneously present in certain vulnerable individuals or induced by repeated drug exposure in others. In conclusion, we provide the first genetic evidence that the GR gene can modulate cocaine abuse. This suggests that targeting GR function in the brain could provide new therapeutic strategies to treat cocaine addiction for which there is no available treatment.

Psychobiology of cocaine addiction: Contribution of a multi-symptomatic animal model of loss of control.

Transition to addiction is the shift from controlled to uncontrolled drug use that occurs after prolonged drug intake in a limited number of drug users. A major challenge of addiction research in recent years has been to develop models for studying this pathological transition. Toward this goal, a DSM-IV/5-based multi-symptomatic model of cocaine addiction has been developed in the rat. It is based on an operational translation of the main features of the disease. 1. Addiction is not just taking drug; it is a non-adaptive drug use: The procedure models addiction in relation to its clinical definition. 2. All drug users do not face the same individual risk of developing addiction: The model includes an individual-based approach. 3. Addiction develops after protracted periods of controlled drug use: This procedure allows for the study of the long-term shift from controlled drug use to addiction. We describe this model in detail and show how it can contribute to our understanding of the pathophysiology of cocaine addiction. This article is part of a Special Issue entitled 'NIDA 40th Anniversary Issue'.

Frequency of cocaine self-administration influences drug seeking in the rat: optogenetic evidence for a role of the prelimbic cortex.

High-frequency intake and high drug-induced seeking are associated with cocaine addiction in both human and animals. However, their relationships and neurobiological underpinnings remain hypothetical. The medial prefrontal cortex (mPFC), basolateral amygdala (BLA), and nucleus accumbens (NAc) have been shown to have a role in cocaine seeking. However, their involvement in regulating high-frequency intake and high cocaine-induced seeking is unclear. We manipulated frequency of cocaine self-administration and investigated whether it influenced cocaine seeking. The contribution of the aforementioned structures was evaluated using changes in expression of the immediate early gene c-Fos and targeted optogenetic manipulations. Rats that self-administered at High frequency (short inter-infusion intervals allowed by short time-out) showed higher cocaine-induced seeking than low frequency rats (long inter-infusions intervals imposed by long time-out), as measured with cocaine-induced reinstatement. c-Fos was enhanced in High frequency rats in the prelimbic (PL) and infralimbic (IL) areas of the mPFC, the BLA, and the NAc core and shell. Correlational analysis of c-Fos revealed that the PL was a critical node strongly correlated with both the IL and NAc core in High frequency rats. Targeted optogenetic inactivation of the PL decreased cocaine-induced reinstatement, but increased cocaine self-administration, in High frequency rats. In contrast, optogenetic activation of the PL had no effect on Low frequency rats. Thus, high-frequency intake promotes a PL-dependent control of cocaine seeking, with the PL exerting a facilitatory or inhibitory effect, depending on operant contingencies. Individual differences in cocaine-induced PL activation might be a source of vulnerability for poorly controlled cocaine-induced seeking and/or cocaine intake.

Pregnenolone can protect the brain from cannabis intoxication.

Pregnenolone is considered the inactive precursor of all steroid hormones, and its potential functional effects have been largely uninvestigated. The administration of the main active principle of Cannabis sativa (marijuana), Δ(9)-tetrahydrocannabinol (THC), substantially increases the synthesis of pregnenolone in the brain via activation of the type-1 cannabinoid (CB1) receptor. Pregnenolone then, acting as a signaling-specific inhibitor of the CB1 receptor, reduces several effects of THC. This negative feedback mediated by pregnenolone reveals a previously unknown paracrine/autocrine loop protecting the brain from CB1 receptor overactivation that could open an unforeseen approach for the treatment of cannabis intoxication and addiction.