Introduction

Major depressive disorder (MDD) is a prevalent illness that negatively impacts the adolescent population [1]. Juveniles who develop MDD often display comorbid conduct and anxiety disorders [2], and up to 25% develop substance abuse [3, 4]. Even more concerning, ~50% of young MDD-sufferers are unresponsive to traditional antidepressants, including the selective serotonin reuptake inhibitor fluoxetine—the only medication approved by the Food and Drug Administration for the treatment of depressive illnesses in the pediatric population. Consequently, alternative pharmaceutical treatments for MDD are actively being explored, particularly for youngsters, since the first incidence of depression is usually reported during adolescence, and suicidal ideation rises drastically during this time [5].

Clinical data indicates that ketamine (KET), a non-competitive N-methyl-d-aspartate receptor antagonist, shows superior promise for the management of MDD symptomatology [6]. Specifically, repeated sub-anesthetic doses of KET in MDD patients result in a rapid and long-lasting antidepressant response [7,8,9]. Importantly, animal studies support these clinical findings in models of despair [10,11,12], providing a platform for future work to uncover the neurobiological factors that mediate its rapid therapeutic properties. These approaches, however, have focused on examining acute KET exposure in mostly adult populations [10, 13]. While the mechanisms underlying KET’s antidepressant efficacy and duration have recently begun to be studied [6], its potential enduring neurobehavioral side effects are relatively unknown [14]. Aside from this, KET is a controlled substance that has abuse/dependence potential [15], so it is surprising that KET is currently being administered to juvenile populations [16], when the brain is still undergoing development [17], and adolescents display higher predisposition for drug abuse [18]. Indeed, exposing children and adolescents to psychotropic agents is of concern because animal research shows that psychotropic pharmacological insults across ontogeny induce neurobehavioral consequences in later adulthood [19,20,21,22,23,24], particularly alterations in subsequent drug preference [25,26,27,28,29,30].

To date, researchers have focused on examining KET’s anesthetic and analgesic effectiveness in pediatric populations [31, 32], with only few recent clinical studies examining its efficacy in treating adolescent mood-related illnesses [9, 16, 33, 34]. Although acute exposures to KET are being administered to treat MDD, it is imperative that the long-lasting consequences of chronic KET be assessed, given that both adult and adolescent patients on KET trials must return to the clinic for recurring treatment [33, 35]. Indeed, animal work has already shown that repeated KET exposure during adolescence induces long-term behavioral alterations to inescapable stressors [36]. Because vulnerability to stress influences drug-seeking behavior [37, 38], the goal of this investigation is to examine if adolescent KET treatment alters preference for reward-related stimuli in adulthood (i.e., sucrose and cocaine), and if so, whether or not this enduring effect is influenced by the age of KET exposure and/or social stress history.

Methods

Animals

See Supplement for details. Experiments were conducted with approval of the Institutional Animal Care and Use Committee at The University of Texas at El Paso.

Drugs

Ketamine (KET)- and cocaine-hydrochloride were dissolved in sterile 0.9% saline (SAL) and administered via intraperitoneal injections at a volume of 2 ml/kg. KET was purchased from Spectrum Chemicals (Gardena, CA), and cocaine from Sigma-Aldrich (St. Louis, MO).

Experimental design

To assess for enduring alterations in responsivity to reward-related stimuli, as a function of juvenile KET pretreatment, we conducted three experiments (Fig. 1). In Experiment 1 (Fig. 1a) adolescent male and female C57BL/6 mice received KET (0 [vehicle; VEH] or 20 mg/kg) for 15 days (PD35-49). After KET or VEH exposure, male and female mice were left undisturbed in their home-cage for a 21-day washout period. At PD70 (i.e., adulthood), KET or VEH pre-exposed male and female mice were tested on the sucrose- or cocaine-preference tests. The age at the start and duration of KET exposure (PD35-49) was selected because it approximates adolescence in humans [39]. Similarly, the KET dose (20 mg/kg/day) was chosen because it yields significant effects on depression-related behavior in adolescents [16, 36]. Because KET induces long-term alterations on despair-related behavior (i.e., forced-swim immobility) in both adolescent and adult male rodents [36], we conducted a separate set of similar experiments using PD70 male and female mice (as a positive control group for the age of KET exposure). Specifically, Experiment 2 (Fig. 1b) examined if 15-days of KET exposure in adult male and female mice (PD70-84) would alter sucrose and cocaine reactivity 21-days post treatment (PD105+). Lastly, because KET has distinct effects between healthy and depressed patients [40], in Experiment 3 (Fig. 1c) we evaluated how KET treatment along with social defeat stress (SDS) or vicarious defeat stress (VDS), during adolescence (PD35-44), would influence preference for cocaine or sucrose in adulthood (PD70+). To do this, we first validated the SDS and VDS paradigms as models of juvenile stress-induced behavior (i.e., reducing sociability; Fig. 4a) and antidepressant-efficacy (i.e., KET preventing stress-induced decreases in social behavior; Fig. 4c). We specifically selected SDS and VDS given that social stress is linked to the development of mood-related illnesses [41, 42]. Only male mice were used in Experiment 3 since juvenile KET pre-exposure increased reward-preference in adult male, but not female, mice (i.e., Experiments 1–2). Separate groups of animals were used across all experiments in order to avoid potential testing carryover effects (see Table S1).

Fig. 1: Experimental design.
figure 1

a To evaluate if juvenile ketamine (KET) exposure mediates enduring alterations to reward-related stimuli in adulthood, adolescent (postnatal day [PD] 35) male and female C57BL/6 mice received KET (20 mg/kg) or saline (vehicle, VEH) for 15 consecutive days (PD35-49). Twenty-one days after the last KET exposure (PD70), mice were evaluated on the cocaine or sucrose preference tests (PD70-75). b To assess if KET pretreatment influenced preference for rewards as a function of age, adult (PD70) male and female mice were exposed to KET or VEH for 15 days (PD70-84), and after a 21-day washout period (PD105), they were evaluated on the cocaine or sucrose preference tests (PD105-110). c Lastly, we examined how juvenile exposure to KET, in combination with either physical or psychological stress history would influence reward reactivity in adulthood. Specifically, adolescent male mice were exposed to either social defeat stress (SDS) or the vicarious defeat stress (VDS) paradigm for 10 consecutive days (PD35-44). Non-stressed control (CON) and stressed mice were administered with KET or VEH immediately after each SDS (KET+SDS) or VDS (KET+VDS) episode. Twenty-four hour post stress/KET treatment (PD45), one group of mice was evaluated on social interaction (SI) behavior. All other groups were evaluated on the cocaine or sucrose preference test once they reached adulthood (PD70-75).

Full size image

Conditioned place preference

CPP was carried out as we have previously described [43] to assess the rewarding effects of cocaine [44, 45]. For specifics, see Supplement.

Sucrose preference

This single-day test consisted of a 2-bottle procedure [46] in which mice had the choice between consuming water or a 1% sucrose solution (details in Supplement).

Stress and social behavior

The juvenile SDS approach [47] with a witnessing component (i.e., VDS; 12, 48) was adopted to induce physical or psychological stress, as a function of KET exposure. Twenty-four hour after stress and/or KET exposure, one group of mice was tested on the social interaction test (details in Supplement).

Statistical analysis

Data were analyzed using ANOVA techniques, with KET-alone or KET-and-stress history, and cocaine post-treatment (between measures) as sources of variance. Tukey post hoc tests were used to examine all pairwise comparisons. Planned comparisons were also conducted to examine the hypothesis that KET pretreatment altered cocaine/sucrose-induced reward. Two-tailed Student’s t tests were used for analyses implicating two-group comparisons. Statistical significance was defined as p ≤ 0.05. Data are presented as mean + SEM.

Results

Adolescent KET exposure increases cocaine and sucrose preference in adult male mice

Figure 2a, b displays the effects of male adolescent KET exposure (PD35-49) on cocaine (0, 5, or 10 mg/kg) CPP in adulthood (PD70+; N = 60). Specifically, a 2-way ANOVA (with KET-pretreatment and cocaine post-treatment as sources of variance) indicated that preference scores for the cocaine-paired side varied as a function of adolescent KET exposure (pretreatment main effect: F1,54 = 13.54, p < 0.05) as well as cocaine exposure in adulthood (post-treatment main effect: F2,54 = 162.69, p < 0.05). Adolescent VEH (n = 10) or KET (n = 10) pretreatment did not influence preference for any of the compartments of the apparatus, when mice were conditioned to SAL (p > 0.05). On the other hand, VEH-pretreated mice conditioned to 5 (n = 10) or 10 mg/kg (n = 10) cocaine displayed reliable conditioning for the cocaine-paired compartment, when compared to control (VEH-pretreated/SAL-conditioned) mice (αp < 0.05). KET-pretreated mice also displayed reliable conditioning to the compartment paired with 5 (n = 10) or 10 (n = 10) mg/kg cocaine, when compared to KET-pretreated/SAL-conditioned mice (p < 0.05, respectively). Interestingly, KET-pretreated mice conditioned to 5 or 10 mg/kg cocaine, displayed higher preference scores for the drug-paired compartment (*p < 0.05, respectively) when compared to VEH-pretreated mice receiving the same dose of cocaine in adulthood (i.e., within dose comparison). Importantly, there were no differences in distance traveled (cm), as a function of adolescent KET or VEH pretreatment, during the preconditioning phase (p > 0.05; Fig. 2b); indicating no differences in general locomotor activity between the groups.

Fig. 2: Enduring effects of adolescent ketamine (KET) exposure for cocaine and sucrose preference in adult male mice.
figure 2

a Twenty-one days after adolescent KET exposure (postnatal day [PD] 70+), KET-pretreated mice displayed an enhanced preference to 5 and 10 mg/kg cocaine, when compared to saline (VEH)-pretreated control mice receiving the same doses of cocaine (n = 10 per experimental group; p < 0.05). Asterisk represent within cocaine-dose group comparison (p < 0.05). Alpha represent significantly different when compared to age-matched controls conditioned to saline (p < 0.05). b No differences in distance traveled (cm) were evident during the preconditioning phase, between the experimental groups. c Adolescent KET history increased preference for a 1% sucrose solution, three weeks after KET exposure (n = 10 per group; Φp < 0.05). d No differences in total liquid intake were observed between the experimental groups (p > 0.05).

Full size image

Figure 2c shows the enduring effects of male adolescent KET exposure (PD35-49) on sucrose preference in adulthood (PD70+). A t test indicated that adult male mice with juvenile KET history (n = 10) displayed a significantly higher preference for a 1% sucrose solution when compared to VEH-pretreated (n = 10) controls (Φt18 = 2.20, p < 0.05). No differences in total liquid intake (water+sucrose) were noted (p > 0.05; Fig. 2d).

Adult KET exposure in male mice does not influence cocaine and sucrose preference later in life

Figure 3a, b displays the enduring effects of adult KET pretreatment (PD70-84) on cocaine CPP, in male mice (PD105+). Here, a 2-way ANOVA (with KET-pretreatment and cocaine post-treatment as sources of variance) showed that preference scores were influenced by cocaine exposure (post-treatment main effect: F2,54 = 203.17, p < 0.05). Specifically, KET (n = 10) or VEH (n = 10) pretreatment did not influence preference for any of the compartments conditioned to SAL (p > 0.05). In contrast, VEH- or KET-pretreated mice conditioned to 5 (n = 10) or 10 mg/kg (n = 10) cocaine, displayed reliable preference scores, when compared to VEH-pretreated/SAL-conditioned mice (ςp < 0.05). Similarly, KET-pretreated male mice displayed a preference for environments paired to 5 (n = 10) or 10 mg/kg cocaine (n = 10), when compared to KET-pretreated/SAL-conditioned mice (p < 0.05, respectively). Planned comparisons indicated that no differences in preference scores were evident between SAL- and KET-pretreated mice receiving the same doses of cocaine (p > 0.05, respectively). No differences in distance traveled (cm) during the preconditioning phase, as a function of KET pretreatment, were observed between the groups (p > 0.05; Fig. 3b).

Fig. 3: Enduring effects of adult ketamine (KET) or saline (VEH) exposure (postnatal day [PD] 70-84) on cocaine and sucrose preference in male mice.
figure 3

a Twenty-one days after adult KET exposure (PD105+), VEH- and KET-pretreated mice conditioned to 5 or 10 mg/kg cocaine displayed reliable preference scores to the cocaine-paired side, when compared to VEH-pretreated/SAL-conditioned mice (ςp < 0.05; n = 10 per experimental condition). No differences in preference scores were evident between VEH- and KET-pretreated mice receiving the same doses of cocaine (p > 0.05, respectively). b No differences in distance traveled (cm) were evident during preconditioning phase, between the experimental groups. c Adult KET pre-exposure did not influence preference for a 1% sucrose solution three weeks after KET treatment (n = 5 per group; p < 0.05). d No differences in total liquid intake were observed between the KET or VEH pretreated groups (p > 0.05).

Full size image

Figure 3c shows that adult KET pretreatment (PD70-84), in male mice, did not alter preference for a 1% sucrose solution (p > 0.05; n = 5 per group), or total liquid intake (p > 0.05; Fig. 3d), 21-days later (PD105+).

Female KET exposure, during adolescence or adulthood, does not alter cocaine or sucrose preference in later life

The long-term effects of female juvenile (PD35-49) or adult (PD70-84) KET history, on cocaine and sucrose preference, are shown in Figs. S1 and S2, respectively. For specific data analyses see Supplemental and Table S2. Briefly, while cocaine mediated CPP across the groups (Figs. S1A and S2A; n = 10/group), there were no differences in preference magnitude between VEH- or KET-pretreated female mice receiving the same cocaine dose (independent of age of KET pre-exposure). No differences in locomotor activity, during the preconditioning phase, were noted between the groups (Figs. S1B and S2B). Similarly, female KET pre-exposure (during adolescence or adulthood) did not influence sucrose preference or total liquid intake, later in life (Fig. S1C, D, n = 9/group; and Fig. S2C, D, n = 10/group).

Establishing depression-related behavior and KET antidepressant-efficacy in juvenile male mice

Figure 4 shows the effects of social stress (VDS or SDS)-alone, or along with KET treatment, on social behavior in adolescent male mice. A one-way ANOVA (within vehicle pretreated mice) indicated that social interaction ratios varied as a function of stress exposure (F2,27 = 12.15, p < 0.05). Tukey post hoc comparisons indicated that, when compared to non-stressed controls (CON, n = 10), mice exposed to VDS (n = 10) or SDS (n = 10) displayed significantly lower social interaction ratios (Fig. 4a)–a depression-related phenotype [48, 49]. Conversely, in Fig. 4c, a one-way ANOVA (within KET exposed mice) indicated that social interaction ratios did not differ as a function of stress [p > 0.05; CON (n = 10), VDS (n = 10), or SDS (n = 10)], showing that KET prevents adolescent stress-induced decreases in sociability, in a similar fashion as adult rodents [12, 50]. Importantly, no differences in locomotor activity were noted as a function of stress type, independent of VEH (p > 0.05, Fig. 4b) or KET (p > 0.05, Fig. 4d) pretreatment.

Fig. 4: Ketamine mediates prophylactic effects on depression-related behavior in adolescent male mice (postnatal day 45).
figure 4

a When compared to CON’s [non-stressed], vicarious defeat stress (VDS/vehicle)- and social defeat stress (SDS/vehicle)-exposed mice displayed decreases in social interaction ratios, b without altering locomotor activity (p > 0.05). c Conversely, ketamine exposure (after each defeat episode) prevented the stress-induced reductions of social behavior across the groups (p > 0.05). d No differences in locomotor activity (distance traveled) were noted between the stress and ketamine-pretreated groups (n = 10 per group). *p < 0.05 when compared to CON.

Full size image

Adolescent KET+VDS or KET+SDS history does not alter cocaine or sucrose preference in adult male mice

Figure 5a shows the long-term effects of concomitant juvenile KET (or VEH) and stress (KET+SDS or VEH+SDS; KET+VDS or VEH+VDS) pre-exposure (PD35-44) on cocaine CPP in male mice (PD70+). A three-way ANOVA (with KET and/or stress pre-exposure, and cocaine post-treatment, as sources of variance) revealed that preference scores varied as a function of cocaine (post-treatment main effect: F2,155 = 76.45, p < 0.05) and a marginal stress × KET × cocaine interaction (F4,155 = 2.12, p = 0.08). Neither stress history (CON, SDS, or VDS), with or without KET pre-exposure, influenced preference scores for the compartment conditioned to SAL (n = 8–10, per group; p > 0.05). In contrast, VEH+CON pre-exposed mice conditioned to 5 or 10 mg/kg cocaine displayed a significant preference for the cocaine paired-compartment, when compared to VEH+CON-pretreated/SAL-conditioned mice (αp < 0.05). Similarly, mice exposed to stress alone (VEH+VDS or VEH+SDS) or in combination with KET (KET+CON, KET+VDS, or KET+SDS) displayed significant preference for the compartment paired with 5 or 10 mg/kg cocaine, when compared to their respective SAL-conditioned groups (p < 0.05). Unexpectedly, VEH+VDS mice displayed enhanced conditioning scores to 5 and 10 mg/kg cocaine (*p < 0.05, respectively), when compared to VEH+CONs receiving the same doses of cocaine. Furthermore, KET-pretreated mice without stress history (KET+CON) also displayed enhanced CPP scores to 5 and 10 (*p ≤ 0.05) mg/kg cocaine, when compared to VEH+CON mice receiving the same cocaine doses—replicating the findings from Experiment 1 (Fig. 2a). Lastly, a three-way ANOVA (with stress, KET, and cocaine as sources of variance) indicated that distance traveled, during the preconditioning phase, varied as a function of KET history (βKET pretreatment main effect: F1,167 = 12.41, p < 0.05). Specifically, mice receiving KET (KET+CON, KET+VDS, KET+SDS) displayed significantly lower distance traveled (Fig. 5b), when compared to VEH-pretreated groups (VEH+CON, VEH+VDS, VEH+SDS).

Fig. 5: Enduring effects of male juvenile exposure to concomitant ketamine (KET)-and-stress [social defeat stress (SDS), vicarious defeats stress (VDS), or no-stress control (CON)] on cocaine and sucrose preference in adulthood.
figure 5

a Neither KET or stress combination treatment (n = 8–10 per group) influenced preference scores for the compartment conditioned to saline (i.e., 0 mg/kg cocaine; p > 0.05). In contrast, all experimental male mice displayed reliable conditioning to 5 and 10 mg/kg cocaine, when compared to respective groups conditioned to saline (cocaine main effect, p < 0.05). Furthermore, adult mice pre-exposed to saline-vehicle (VEH) and VDS during adolescence, showed an enhanced preference to 5 and 10 mg/kg cocaine, when compared to VEH+CON mice receiving the same doses (*p < 0.05, respectively). Similarly, adult mice pre-exposed to KET+CON showed a higher preference score to 5 and 10 (*p ≤ 0.05) mg/kg cocaine, when compared to VEH+CON mice receiving the same doses of cocaine. b KET-pretreatment, independent of stress history, resulted in lower distance traveled during the preconditioning phase (main effect of stress, βp < 0.05). c Adolescent stress history (VDS or SDS), independent of KET or VEH pretreatment, resulted in decreased preference for a 1% sucrose solution in adulthood (main effect of stress; χp < 0.05). d KET-pretreatment, independent of stress history, resulted in decreased liquid intake in adulthood (main effect of KET-pretreatment, βp < 0.05).

Full size image

Figure 5c, d shows the long-lasting effects of adolescent KET and stress history (PD35-44) on sucrose preference in adulthood (PD70+). A two-way ANOVA (with stress and KET as sources of variance) indicated that juvenile stress exposure altered sucrose preference in adulthood (Χstress main effect: F2,54 = 18.33, p < 0.05). Specifically, mice with stress history (VEH+VDS, VEH+SDS, KET+VDS, KET+SDS) displayed significantly lower preferences for sucrose, when compared to non-stressed groups (VEH+CON, KET+CON), in adulthood. Lastly, a two-way ANOVA, with stress and KET history as sources of variance, indicated that total liquid intake ([water+sucrose]/weight) varied as a function of KET pretreatment (βKET-pretreatment main effect: F1,54 = 10.92, p < 0.05). Specifically, independent of stress pre-exposure (VEH+CON, VEH+VDS, VEH+SDS), KET pretreated animals (KET+CON, KET+VDS, KET+SDS) displayed lower total liquid intake (M = 0.15 mL/g) when compared to mice pre-exposed to VEH (M = 0.20 mL/g). Lastly, Fig. S3 demonstrates the long-lasting effects of adolescent KET and stress history (PD35-44) on body weight in adulthood (PD70). A two-way ANOVA indicated that body weight did not differ between the experimental groups, as a function of KET-pretreatment, stress-type, or their interaction (p > 0.05, respectively).

Discussion

KET is currently being dispensed to juvenile populations for the management of MDD, given its rapid antidepressant therapeutic effects [9, 16, 33, 34]. Such off-label pharmacological approach is concerning, since KET is a controlled substance with abuse potential [51] and could result in unexpected long-lasting side effects [36, 52]. Therefore, the purpose of this investigation was to assess if adolescent KET exposure alters preference for reward-related stimuli in adulthood; a preclinical approach that could inform about KET-induced predisposition to drug use-vulnerability in later life [29, 44, 53]. We report that juvenile KET exposure increases preference for the rewarding properties of cocaine and sucrose in male, but not female, adult mice. Moreover, this enduring male-specific effect is dependent on the age of KET pre-exposure (i.e., adolescence) as well as the absence of physical (SDS) or psychological (VDS) stress history.

CPP is a behavioral paradigm that is commonly implemented to evaluate the rewarding properties of drugs of abuse [29, 54]. Adopting Pavlovian learning principles, CPP is used to measure increases in the incentive motivation value of rewarding stimuli, as they become associated with a specific environment and thus it is considered a model of reward-seeking behavior [44]. Accordingly, implementing this well-established approach, we found that independent of juvenile KET pretreatment, adult male and female mice (PD75+) displayed a preference for the cocaine-paired compartment (Fig. 2a and S1A, respectively). Interestingly, within the male groups (Fig. 2a), adult mice (PD75+) pre-exposed to KET during adolescence (PD35-49) displayed enhanced CPP magnitude, when compared to corresponding controls receiving the same cocaine doses. This suggests that juvenile KET history increases reactivity to the rewarding effects of cocaine in adulthood [44]. In line with this, KET-pretreated adult male mice displayed higher preference for sucrose (Fig. 2c), when evaluated on a single-day two-bottle test. Conversely, no alterations to either reward-related stimuli were observed in adult female mice pretreated with KET during adolescence, when compared to respective controls (Fig. S1). Importantly, KET pretreatment did not alter general locomotor activity (Fig. 2b, S1B, and 52) or total liquid intake in neither male nor female mice (Fig. 2d and S1D, respectively). These findings unmask an enduring KET-induced increase in preference for cocaine and sucrose in males specifically.

Previous work indicates that repeated exposure to KET, during adolescence or adulthood, alters sensitivity to inescapable stressors two-months after drug pretreatment [36], highlighting a long-lasting KET-induced effect. Therefore, we exposed adult (PD70) male and female mice to KET for the same amount of time (15-days) and evaluated their preference for cocaine or sucrose at a similar post-treatment interval (21-days). As expected, PD105+ male and female mice acquired cocaine CPP (Fig. 3 and Fig. S2, respectively). However, in this case, neither PD105+ male nor female mice (pre-exposed to KET from PD70-84) displayed enhanced CPP scores when compared to respective VEH-controls administered with the same doses of cocaine. Similarly, no differences in sucrose preference were noted as a function of adult KET pre-exposure in either male or female mice (Fig. 3c and Fig. S2C). Together, these sex- (compare Fig. 2a and S1A) and age-dependent (compare Figs. 2a and 3a) effects confirm that adolescence is a developmental window of vulnerability (to KET exposure) resulting in long-lasting alterations to reward reactivity in males, but not females. Furthermore, these data support previous work demonstrating that juvenile psychotropic drug exposure leads to enduring neurobiological changes that are persistent in adulthood [26, 27], yet, here we expand these findings to KET-induced alterations for enhanced cocaine and sucrose reward in a sex- and age-specific manner.

Clinical studies report that KET mediates differential effects between normal human volunteers and MDD patients [40]. Thus, we evaluated if adolescent KET-pretreatment along with social stress history would influence reward preference in adulthood. We selected SDS and VDS given that these physical and psychological forms of stress reduce social behavior—a depression-related phenotype frequently reported in adult rodents [47, 49, 55] that is rescued by chronic administration of traditional antidepressants [56] and acute KET exposure [12, 50]. Adopting these stress paradigms, we forced adolescent male C57BL/6 mice to intrude the territorialized cage of a resident aggressor CD1 mouse, resulting in an agonistic physical encounter (SDS), while an observer adolescent C57BL/6 mouse vicariously experienced the stress about from the safety of an adjacent compartment (VDS), once per day for 10 consecutive days [12, 48, 57]. After each stress episode, SDS, VDS, and non-stressed (CON) mice were administered KET or VEH (Fig. 1c). To validate this approach as a model for the study of adolescent stress-induced depression, one subgroup of mice was evaluated on social behavior the following day (PD45). We found that adolescent mice exposed to SDS [47] and VDS [55] displayed lower social interaction ratios, when compared to CONs (Fig. 4a). Importantly, we show for the first time that exposure to KET prevents stress-induced decreases of social behavior in juvenile mice (Fig. 4c), establishing pharmacological validity for the study of depressive- and antidepressant-like efficacy in the juvenile SDS/VDS model [57]. Consequently, after additional groups of ketamine-and-stress pre-exposed mice reached adulthood (PD70), they were evaluated for cocaine and sucrose reward; matching the age of behavioral testing of stress-naive animals pretreated with KET or VEH in our initial experiment (Fig. 1a). Here, PD75 mice with adolescent history of psychological stress (VEH+VDS), or KET+CON, displayed higher preference scores to cocaine, when compared to non-stressed controls (VEH+CON) exposed to the same doses of the stimulant (p < 0.05). This indicates that juvenile exposure to VDS-, or KET-alone, mediates long-term increases for cocaine seeking behavior, in a similar fashion as early-life exposure to stress or pharmacological insults [26, 27, 58, 59]. Notably, while previous work suggests that SDS-alone mediates an enduring increase for cocaine CPP [60], here, VEH+SDS pre-exposed mice did not differ from the VEH+CON group receiving the same cocaine dose. This discrepancy is likely due to differences in mouse strain (OF1 vs C57BL/6), stress protocol, and dose of cocaine used (1 vs 5–10 mg/kg) between the studies. Interestingly, the cocaine-induced CPP scores in animals exposed to concomitant KET and stress (KET+VDS or KET+SDS) did not differ in magnitude from saline/non-stressed control mice (VEH+CON). Hence, revealing that adolescent exposure to combined stress and KET—modeling antidepressant treatment for MDD—does not result in long-lasting changes to reward-related stimuli.

When we evaluated reactivity to sucrose, as a function of juvenile stress and/or KET pre-exposure, we found that independent of KET history, stress (VDS and SDS)-exposed mice displayed reduced preference for this natural reward—replicating a well established long-term SDS- and VDS-induced anhedonia-like phenotype [48, 61, 62] that was not rescued by KET pretreatment during adolescence (Fig. 5c). This was anticipated, given that both animal models [50], as well as MDD patients [7,8,9], must receive repeated KET infusions to ameliorate depression-related phenotypes. However, unlike Experiment 1, no differences in sucrose preference were noted between KET+CON and VEH+CON groups. This inconsistent finding across experiments (see Figs. 2c and 5c) is likely due to differences in the number of KET/VEH injections received during adolescence (15 vs 10), as well as housing conditions during the drug washout period (grouped vs segregated). Unexpectedly, KET-history decreased liquid intake (Fig. 5d) and locomotor activity (Fig. 5b); yet, such factors were not likely to influence responses to reward-related stimuli, since no differences in body weight were noted between the groups (Fig. S3), and total liquid was normalized to body weight for the sucrose preference test (Fig. 5d). Of course, future work will be needed to evaluate the consequences of adolescent KET exposure on these physiological measures, particularly, since decreased liquid intake could be an indication of oligodipsia, a syndrome correlated with dissociative experiences [63, 64].

The neurobiological mechanisms by which early-life KET exposure increases reward-related preference in adulthood have not been evaluated. During adolescence, KET increases locomotor activity in a sex-specific manner [65], with males displaying a delayed response when compared to females. Interestingly, monoamine systems underlie, at least in part, this age and sex-specific KET-induced pattern of locomotor activation [66], which in turn, could provide insight to the male-specific consequences of KET on reward-related stimuli in adulthood (or enduring female resiliency to KET). For example, acute KET administration (i.e., 10–50 mg/kg) increases dopamine levels in the prefrontal cortex, nucleus accumbens, and striatum, while increasing the firing of dopamine neurons within the ventral tegmental area; [67] circuitry involved in reward-related behavior. Thus, repeated exposure to KET during adolescence may result in long-lasting and sex-specific changes within the dopamine system, altering reward-seeking behavior in adulthood [68, 69]. Supporting this hypothesis, human recreational KET users display increases in dopamine receptor 1 (DR1) within the medial prefrontal cortex [70], highlighting a potential neurobiological mechanism by which repeated early-life exposure to KET (without stress history) may increase preference for cocaine in later life. Indeed, the expression of cocaine CPP is modulated by DR1 antagonism within this brain region in male rodents [69], so it is possible that juvenile KET history may result in altered DR1 expression in a sex-specific manner, as it has been shown in other brain regions [71]. Conversely, the absence of long-lived increases in cocaine or sucrose preference, post juvenile exposure to concomitant stress-and-KET, could be attributed to the individual and oppositional effects between KET and stress on brain derived neurotropic factor (BDNF)-related signaling [10, 72], a molecular mechanism that modulates behavioral responses to cocaine and that is also implicated in MDD [73,74,75].

In rodent models of physical and/or psychological stress, KET acutely reverses or prevents the development of depression-related phenotypes (Fig. 4, and [12, 50, 76]. However, to our knowledge, the enduring effects of repeated juvenile KET exposure on cocaine- and sucrose-preference, with or without social stress history, have not been evaluated. Here we report, for the first time, that adolescent KET history in stress-naïve mice increases preference for cocaine and sucrose in adult male, but not female, C57BL/6 mice; suggesting a potential enduring male-specific increase in risk for drug use-vulnerability. However, juvenile KET exposure, along with physical or psychological stress, does not induce long-lasting alterations to drug or natural rewards. Collectively, this preclinical model provides first line evidence about the safety of chronic KET exposure during adolescence, post recreational use, or when administered as a treatment for stress-induced affect-related illnesses.

Funding and disclosure

No financial or non-financial interests in relation to the work described in this manuscript is declared by the authors. This project was supported by the National Institute of General Medical Sciences (NIGMS-1SC3GM130467, to SDI). Student co-authors were financially supported to attain basic research experience/training per their respective program objectives. Specifically, FJF-R was supported by a Summer Research Grant from The Graduate School at UTEP. IG-C and SAC were funded by the LSAMP (Louis Stokes Alliance for Minority Participation) Bridge to Doctorate support program. MAA was supported by the SMART-MiND (Summer Mentoring and Research Training: Methods in Neuroscience of Drug-abuse) program. JP-P was supported by the BUILD (Building Infrastructure Leading to Diversity) Scholars program at UTEP. The authors declare no conflict of interest.