Impact of Serine Racemase Deletion on Nicotine Discrimination

Abstract

Introduction

The high comorbidity between schizophrenia and cigarette smoking points to a possible shared heritable factor predisposing individuals with schizophrenia to nicotine addiction. The N-methyl-d-aspartate (NMDA) receptor has been highly implicated in both schizophrenia and nicotine addiction.

Methods

In the present study, we used mice with a null mutation on the serine racemase gene (srr), an established risk gene for schizophrenia, which encodes the enzyme to produce the NMDA receptor co-agonist d-serine, to model the pathology of schizophrenia and to determine whether NMDA receptor hypofunction reduced the ability of srr^−/− mice to identify nicotine’s subjective effects. Established nicotine discrimination procedures were used to train srr^−/− and wild-type (WT) mice to discriminate 0.4 mg/kg nicotine under a 10-response fixed-ratio (FR10) schedule of food reinforcement.

Results

Results show that WT mice reliably acquired 0.4 mg/kg nicotine discrimination in about 54 training sessions, whereas srr^−/− mice failed to acquire robust 0.4 mg/kg nicotine discrimination even after extended (>70) training sessions. These results show that NDMA receptor hypofunction in srr^−/− mice decreased sensitivity to the interoceptive effects of nicotine.

Conclusions

Projected to humans, NMDA receptor hypofunction caused by mutations to the serine racemase gene in schizophrenia may reduce sensitivity to nicotine’s subjective effects leading to increased nicotine consumption to produce the same effects as those unaffected by schizophrenia.

Implications

There is high comorbidity between schizophrenia and nicotine dependence as well as possible shared genetic risk factors between the two. The serine racemase knockout mouse (srr^−/−) with NMDA receptor hypofunction has been developed as a model for schizophrenia. We found that srr^−/− mice were unable to acquire 0.4 mg/kg nicotine discrimination, while WT mice readily discriminated nicotine. These results show that decreased NMDA receptor function present in srr^−/− mice and patients with schizophrenia may result in reduced sensitivity to nicotine’s interoceptive effects, leading to increased nicotine consumption to produce the same subjective effects as those unaffected by schizophrenia.

Introduction

Nearly 90% of individuals with schizophrenia (SZ) are heavy cigarette smokers, compared to about 20% of the general population.^1–3 Moreover, smokers with SZ consume more cigarettes and extract significantly greater levels of nicotine—the primary psychoactive chemical in tobacco—from each cigarette.² Consequently, SZs are more likely to develop nicotine dependence, experience pronounced nicotine withdrawal symptoms during abstinence, and suffer a higher incidence of tobacco-related diseases and death.³

Genetic factors play a key role in SZ (~80% heritability), and substance abuse (SA)—especially tobacco addiction—may similarly exhibit significant heritability.^3,4 Genome-wide association studies in SZ have revealed over 100 risk genes with 30 encoding proteins clustered at the glutamate synapse.⁵ Thus, shared heritable factors may render individuals with SZ more vulnerable to nicotine addiction. Hypofunction of glutamatergic N-methyl-d-aspartate receptors (NMDARs) is a core feature of SZ.⁶ The NMDAR has also been implicated in neuroplastic changes underlying addictive behaviors, and inhibition of NMDAR activation has been associated with increased nicotine self-administration.^2,7 The high comorbidity of nicotine dependence in SZ suggests a convergent pathway of NMDAR hypofunction in both conditions with the risk genes for SZ increasing the likelihood for tobacco addiction. To further explore comorbidity between SA and SZ, we have previously utilized mice with a null mutation of serine racemase (SR), an established risk gene for SZ, which prevents the synthesis of the forebrain NMDAR co-agonist d-serine, thus impairing NMDAR function.⁸ These srr^−/− mice exhibit numerous neurobiological homologies to SZ and display reduced rewarding and neurochemical effects to challenge with psychomotor stimulants.^6,9,10 These data suggest that abused substances may be less rewarding in srr^−/− mice, possibly requiring higher doses to achieve a hedonic response and thus increasing dependence liability. Here, we tested whether this mutation could reduce the ability of srr^−/− mice to learn nicotine discrimination compared to wild-types (WT). We used drug discrimination methods to study the discriminative stimulus effects of nicotine because such effects in animals have been related to subjective (interoceptive) effects that contribute to tobacco consumption^11–13 and therefore may provide insights into how this SZ risk genotype influences nicotine’s subjective effects.

Materials and Methods

Materials

Transgenic Mice

Details on the transgenic mice have been described previously.¹³ Briefly, srr^−/− mice (n = 7) expressed the SR gene with a deleted exon 1, which encodes the catalytic domain of the enzyme.⁸ The SR mutant mice were backcrossed onto a C57Bl/6J background for >10 generations and bred in-house by crossing heterozygous srr^+/− male and female mice. WT (n = 5) littermates were used for the srr^+/+ strain. Only male mice were used due to the sexual dimorphism seen in the srr^−/− phenotype. Animals were group-housed with littermates in a temperature- and humidity-controlled vivarium on a 12:12-hour light/dark cycle (lights on at 7:00 am). All experiments were conducted 5 days/week during the light cycle (10:00 am to 5:00 pm). Mice were maintained at ~85% free-feeding weight and were fed ~3 g of standard chow after daily behavioral sessions. Water was available ad libitum except during behavioral sessions. Mice were maintained in accordance with the Guide for the Care and Use of Laboratory Animals (National Academy of Sciences, 2011) and experiments were approved by the McLean Hospital Institutional Animal Care and Use Committee.

Drugs

(−)-Nicotine hydrogen tartrate [(2)-1-methyl-2-(3-pyridyl)pyrrolidine(1)-bitartrate salt] was obtained from Sigma-Aldrich (St. Louis, MO) and dissolved in sterile 0.9% saline solution. The pH of nicotine was adjusted to pH 7.0 with sodium hydroxide (0.1 N). Doses of nicotine are expressed in terms of free base. All injections were administered subcutaneously (s.c.) at a volume of 10 mL/kg.

Methods

Apparatus

Standard mice operant conditioning chambers were used (Med Associates, St. Albans, VT). Each chamber consists of two nose-poke apertures, a food tray, stimulus lights above each of the nose-poke apertures, and a house light for ambient illumination. A beam break within each nose-poke aperture produced an audible click in the chamber and was recorded as a response. Computers with Med Associates operating software and equipment were used to control experimental variables and to record data (MED-PC; Med Associates).

Nicotine Discrimination

The discrimination training procedure used to establish 0.4 mg/kg nicotine discrimination has been described previously.^14,15 Prior studies have shown that the substitution profiles for many drugs depend on training dose,¹¹ and that higher doses generally are more pharmacologically restrictive than lower doses. Thus, we selected a lower training dose of 0.4 mg/kg nicotine in mice to better detect differences in sensitivity between the srr^−/− and WT mice. Briefly, prior to the first injection, all mice in both groups underwent similar magazine (autoshaping) training such that they consistently responded on either nose-poke hole under a gradually increasing fixed-ratio (FR) schedule (i.e., from FR1 to FR10) of food reinforcement over 20–30 training sessions. Completion of FR10 nose-poke responses resulted in the delivery of one 45-mg food pellet (BioServe, Frenchtown, NJ). Once reliable responding for food was achieved for both nose-poke apertures, mice were trained to discriminate s.c. injections of 0.4 mg/kg nicotine from saline. Following nicotine injections, responses to one aperture were rewarded and following saline injections, responses only on the other aperture were reinforced. Responses on the incorrect aperture reset the FR response requirement. Through differential reinforcement, over daily training sessions the mice learn to discriminate between the presence and absence of the training drug, that is, discriminate between the interoceptive (subjective) effects of 0.4 mg/kg nicotine from saline. A double-alternation training schedule (ie, drug–drug–saline–saline) was used. Injections were given 10 minutes prior to the behavioral session; each session started with the illumination of the house and stimulus lights. Delivery of each food pellet initiated a 10-second time-out period and sessions ended once subjects obtained 20 reinforcements or after 15 minutes. Substitution test sessions in WT mice were only initiated once mice achieved ≥80% correct responses for the overall and the first food reinforcement over four consecutive training sessions. After that, test sessions were conducted when subjects met the above criteria for two successive sessions (eg, nicotine–saline or saline–nicotine). Test sessions were identical to training sessions, except 10 consecutive responses on either nose-poke aperture resulted in the delivery of a food reinforcer.

Data Analysis

Drug effects are described as group means (±SEM) of percentage of nicotine-associated responding overall and for first FR and overall response rates. The percentage of nicotine-associated responding was calculated by dividing the number of responses on the drug aperture by the total number of responses on both apertures. The response rate was calculated by dividing the total number of responses by the session duration excluding time-out periods. One-way ANOVA with Tukey’s test for multiple pairwise comparisons was used when appropriate to evaluate statistical significance defined at the 95% level of confidence and p < .05.

Results

The acquisition of 0.4 mg/kg nicotine discrimination in WT and srr^−/− mice are shown in Figures 1 and 2, respectively. All data shown represent effects after the first injection of either drug or vehicle following initial magazine (autoshaping) training in all subjects. Due to this initial nose-poke training, both groups were performing above the chance level of accuracy (~50%) drug-associated responding for the first 20 training sessions (compare Figures 1A vs. 2A). Thereafter, WT mice were able to discriminate 0.4 mg/kg nicotine from vehicle as evident by the clear separation between nicotine- (>70%) and vehicle- (<30%) associated responding (Figure 1A). This ability to discriminate 0.4 mg/kg nicotine gradually increased across all WT subjects with the overall and first food-reinforcement nicotine-associated responding reaching criteria (≥80%) for nicotine discrimination after 54 (±3.19 SEM) training sessions (Figure 1A and B, respectively). In contrast, while some separation was observed between 0.4 mg/kg nicotine and saline training sessions in srr^−/− mice (ie, 51.6 ± 4.64% vs. 12.3 ± 2.44%, respectively, from sessions 20–40 and 69.4 ± 3.07% vs. 15.4 ± 1.77%, respectively, from sessions 41–75), they failed to consistently discriminate nicotine (Figure 2A and B). While rates across WT and srr^−/− subjects increased steadily over discrimination training sessions, no substantial differences in average response rates were observed between the groups (Figures 1C and 2C, respectively).

Figure 1.

Acquisition of 0.4 mg/kg nicotine as a discriminative stimulus from vehicle in WT mice after the first injection of either drug or saline (n = 5). Ordinates: percentage of overall nicotine-associated responding (Figure 1A), first food-reinforcement nicotine-associated responding (Figure 1B), response rate as responses/second (Figure 1C), and percentage of overall nicotine-associated responding and response rate as responses/second in nicotine substitution testing (Figure 1D). Abscissae: number of nicotine discrimination training sessions (Figure 1A–C), and controls and tested nicotine doses (Figure 1D). Control points (V = vehicle; N = training dose of nicotine) are the mean nicotine-associated responding for the last 5 training sessions of vehicle or nicotine, respectively, before substitution testing began (Figure 1D). Horizontal dashed lines at 80% and 20% nicotine-associated responding (Figure 1A, B, and D) indicate the criteria for evidence of recognizing nicotine as a discriminative stimulus. Each point represents the mean and the standard error of the mean.

Figure 2.

Acquisition of 0.4 mg/kg nicotine as a discriminative stimulus from vehicle in srr^–/– mice after the first injection of either drug or saline (n = 7). Ordinates: percentage of overall nicotine-associated responding (Figure 2A), first food-reinforcement nicotine-associated responding (Figure 2B), and response rate as responses/second (Figure 2C). Abscissae: number of nicotine discrimination training sessions. Horizontal dashed lines at 80% and 20% nicotine-associated responding (Figure 2A and B) indicate the criteria for evidence of recognizing nicotine as a discriminative stimulus. Each point represents the mean and the standard error of the mean.

Nicotine substitution tests were only conducted in WTs as the srr^−/− mice were unable to fully acquire nicotine discrimination (eg, vehicle = 12.6% ± 3.77 SEM; 0.4 mg/kg nicotine = 67.5% ± 11.5 SEM), precluding determination of the nicotine dose–response function. In the WT mice, the vehicle and 0.4 mg/kg nicotine training sessions engendered, respectively, 8.29% (±2.20 SEM) and 88.1% (±3.63 SEM) nicotine-associated responding. Nicotine (0.032–0.4 mg/kg) produced a full and dose-dependent substitution for the training does of 0.4 m/kg nicotine, with 0.32 and 0.4 mg/kg engendering 90.5% (±2.81 SEM) and 98.9% (±0.69 SEM) nicotine-associated responding, respectively (Figure 1D). Nicotine did not significantly alter rates of food-maintained behavior (F_(5,36) = 1.998, p > .05; Figure 1D).

Discussion

The NMDAR has been implicated in both SZ and nicotine addiction.^6,7 Here, we determined if a null mutation of SR, an established risk gene for SZ needed to produce NMDAR co-agonist d-serine, affected nicotine discrimination learning. Consistent with previous findings, WT mice readily discriminated 0.4 mg/kg nicotine,¹⁶ whereas the srr^−/− mice were unable to reliably acquire 0.4 mg/kg nicotine discrimination. These results suggest that the null mutation in the SR gene, and consequent lack of d-serine, blunted nicotine’s subjective effects. The first food-reinforcement data further emphasizes these findings—unlike their WT counterparts, srr^−/− mice do not clearly distinguish between saline and 0.4 mg/kg nicotine, showing that the mice are not switching from the inactive to the active aperture. These data align with our predictions that the srr^−/− mice may require higher doses of nicotine to achieve a subjective response and are consistent with our earlier observations in srr^−/− mice showing a blunted behavioral and neurochemical response to cocaine.¹³ A similar mechanism may underlie the comorbidity between SZ and nicotine addiction wherein those with SZ have decreased sensitivity to nicotine’s subjective response, thus increasing the risk for greater nicotine consumption and dependence.

We have previously reported that NMDAR hypofunction and consequent decrease in sensitivity to cocaine in srr^−/− mice may be mediated by reduced dopamine and glutamate efflux.¹³ This may also hold true for nicotine. Dopamine is a secondary mediator for nicotine discrimination¹¹ and thus a blunted nicotine-induced increase of dopamine in srr^−/− mice would also lead to decreased sensitivity to nicotine’s discriminative stimulus effects. The relationships between dopamine, glutamate, NMDARs, and the rewarding effects of nicotine have been previously highlighted. For example, antagonist-induced decreases in NMDAR activity in the ventral tegmental area can abolish nicotine’s rewarding effects and decrease nicotine self-administration, suggesting that the glutamate neurotransmission is critical to nicotine’s rewarding properties.¹⁷ The hypofunction of NMDARs evident in srr^−/− mice, and individuals with SZ, could decrease reward sensitivity to nicotine and nicotine’s subjective effects leading to increased nicotine consumption in individuals affected with SZ. This is consistent with our working hypothesis that NMDAR hypofunction creates a state in which nicotine becomes less potent and effective, and therefore, considerably higher doses of nicotine are required to achieve a subjective response in individuals with SZ than in unaffected individuals. Thus, contrary to our expectation that the high prevalence of tobacco addiction in individuals with SZ was due to greater responses to abused substances, our results suggest that the high prevalence of tobacco use in these individuals may be related to their need for considerably higher doses of the drug to obtain a hedonic response.¹³

Although the NMDAR has been implicated in learning and memory, reducing d-serine levels are unlikely to produce global deficits in learning and memory that could account for the inability of srr^−/− mice to fully discriminate nicotine. For example, electrophysiologic studies indicate that while d-serine is the major co-agonist in forebrain NMDARs, glycine is also a co-agonist that supports hippocampal NMDAR function and modulates memory.^18,19 Moreover, srr^−/− mice exhibit a partial but not complete reduction in NMDAR function in electrophysiologic studies at the hippocampal CA1 pyramidal neuron synapse. Regarding memory, in the Morris Water maze, srr^−/− exhibited a subtle impairment that affected performance on the probe task but not acquisition or retention.⁸srr^−/− animals were unimpaired in the detection of novel objects and in spatial displacement and showed intact relational memory in a test of transitive inference. In addition, srr^−/− mice exhibited normal sociability and preference for social novelty.²⁰ Thus, it is unlikely that genetic manipulation caused global deficits in their learning and memory to inhibit the ability of srr^−/− mice to acquire nicotine discrimination. Although, at present, it is unclear whether the srr^−/− mice can acquire nicotine discrimination at higher training doses (eg, 0.8 mg/kg),¹¹ such data would further support for our observations that decreased d-serine and NMDAR function reduces sensitivity to nicotine’s interoceptive effects.

Conclusions

NMDAR hypofunction reduces sensitivity to nicotine’s interoceptive effects which likely necessitates higher doses in individuals with SZ, thus increasing the likelihood of developing nicotine dependence. Thus, enhancing NMDAR function may have therapeutic effects in SZ and reduce tobacco consumption. Consistent with this hypothesis, clozapine, which uniquely lowers cigarette consumption and negative symptoms in SZ, enhances brain NMDAR function.¹

Funding

This research was supported by NIH grant RO1MH51290-18 to JTC and DA031231 K01 NIH/NIDA grant to RID.

Declaration of Interests

JTC reports a patent on the clinical use of d-serine that is owned by Partners Health Care and consulting with Novartis and Forum Pharm over the last 3 years. No other conflicts are reported.

Author Contributions

Isabel Yu (Formal analysis [supporting], Writing—original draft [lead], Writing—review & editing [supporting]), Joseph Coyle (Conceptualization [equal], Investigation [equal], Writing—original draft [supporting], Writing—review & editing [equal]), and Rajeev Desai (Conceptualization [lead], Formal analysis [equal], Investigation [lead], Methodology [lead], Resources [lead], Supervision [lead], Writing—original draft [equal], Writing—review & editing [equal])

Data Availability

All data in the manuscript are available from the corresponding author upon request. Correspondence and requests for materials should be addressed to RID.