Effect of abstinence on amphetamine-induced conditioned place preference

To determine the effect of abstinence length on the magnitude of amphetamine-induced CPP, we compared the CPP scores between baseline preference and cue-exposure tests performed 1 or 14 days after the last acquisition session (Fig. 1b). A three-way mixed ANOVA yielded statistically significant main effects of Conditioning (F(1, 30) = 57.18, p < 0.001) and Group (F(1, 30) = 48.29, p < 0.001), but not for Abstinence (F(1, 30) = 2.05, p = 0.16). Moreover, multiple significant interaction effects were detected between Conditioning by Group (F(1, 30) = 59.10, p < 0.001), Conditioning by Abstinence (F(1, 30) = 6.65, p < 0.05), and Group by Abstinence (F(1, 30) = 10.23, p < 0.01). Two-way mixed ANOVA for each treatment group resulted in significant main effects of Conditioning for amphetamine-treated subjects (F(1, 16) = 93.23, p < 0.001), Abstinence (F(1, 16) = 13.22, p < 0.01), and a significant interaction between these factors (F(1, 16) = 4.96, p < 0.05). Bonferroni multiple comparison analysis showed a significant increase in CPP scores after amphetamine conditioning for both ABS1 (p < 0.001) and ABS14 groups (p < 0.001) when compared to their initial preference. Additionally, posttest CPP scores for ABS14 were significantly higher (p < 0.01) than ABS1. No differences between initial sessions were detected. For saline-treated subjects, no significant main effects were found for Conditioning (F(1, 14) = 0.01, p = 0.91), Abstinence (F(1, 14) = 1.27, p = 0.27), or interaction between these factors (F(1, 14) = 2.02, p = 0.17). Since CPP relies on intact motor activity, we compared the distance traveled during tests (Fig. 1c). A three-way mixed ANOVA found no significant main effects for Conditioning (F(1, 30) = 3.54, p = 0.07), Group (F(1, 30) = 1.89, p = 0.17), Abstinence (F(1, 30) = 0.06, p = 0.80), or any interaction between these factors.

These results demonstrate that contrary to saline, amphetamine treatment produced a robust increase in the time spent in the paired chamber, effectively reversing initial preference. Moreover, following amphetamine treatment, CPP expression was notably higher after 14 days of abstinence than after 1 day (Fig. 1b), with no substantial differences in motor activity between tests or abstinence periods (Fig. 1c).

Fig. 1

Abstinence-exacerbated CPP expression at 14 days following chronic amphetamine treatment. a Schematic of CPP procedure. b Baseline preference and CPP expression following early or late abstinence. *p < 0.001 vs. baseline preference, #p < 0.01 vs. ABS1. c Motor activity during place preference tests

To evaluate whether CPP expression may have been affected by differences during conditioning between abstinence groups, we compared the distance traveled during acquisition sessions corresponding to the first (AMPH1) and last (AMPH5) exposure to amphetamine (Fig. S2a). A three-way mixed ANOVA yielded statistically significant main effects of the Pairing session (F(1, 30) = 44.09, p < 0.001), Group (F(1, 30) = 88.46, p < 0.001), and a significant interaction between these factors (F(1, 30) = 33.49, p < 0.001). No significant main effects were found for Abstinence (F(1, 30) = 0.23, p = 0.63) or any other interaction. Two-way mixed ANOVA for each treatment group resulted in a significant main effect of the Pairing session for amphetamine-treated mice (F(1, 16) = 49.28, p < 0.001) with no significant main effect of Abstinence (F(1, 16) = 0.02, p = 0.88) or interaction between these factors (F(1, 16) = 0.93, p = 0.34). For saline-treated mice, no significant main effects were found for the Pairing session (F(1, 14) = 1.42, p = 0.25), Abstinence (F(1, 14) = 0.49, p = 0.49), or interaction between these factors (F(1, 14) = 0.00, p = 0.98).

These results demonstrate that amphetamine treatment significantly increased motor activity during acquisition, compared to saline. Additionally, motor sensitization developed due to chronic amphetamine exposure, with no differences between abstinence periods (Fig. S2a).

To further evaluate whether the abstinence-exacerbated CPP expression could be countered by recent non-contingent drug exposure, we administered conditioned subjects either saline or amphetamine in their home-cage, 24 h before the cue-exposure test performed 14 days after the last acquisition session (Fig. S3a). Two-way mixed ANOVA yielded a significant main effect of Conditioning (F(1, 16) = 198.67, p < 0.001), with no significant main effect of Group (F(1, 16) = 0.60, p = 0.44) or interaction between these factors (F(1, 16) = 2.64, p = 0.12). Regarding motor activity during preference tests (Fig. S3b), two-way mixed ANOVA found no significant main effects of Conditioning (F(1, 16) = 0.74, p = 0.40), Group (F(1, 16) = 0.23, p = 0.63), or interaction between these factors (F(1, 16) = 0.02, p = 0.86). Two-way mixed ANOVA of motor activity during acquisition (Fig. S2b) showed a significant main effect of the Pairing session (F(1, 16) = 52.56, p < 0.001), with no significant main effect of Group (F(1, 16) = 0.06, p = 0.79) or interaction between these factors (F(1, 16) = 0.15, p = 0.69).

These results indicate that both groups displayed higher CPP scores after amphetamine conditioning than their corresponding baseline preference. Moreover, no differences in CPP expression were observed following a non-contingent amphetamine dose administered 24 h before the late abstinence test (Fig. S3a). Motor sensitization developed following chronic amphetamine exposure similarly between groups (Fig. S2b), and no differences in motor activity were observed across tests or between groups (Fig. S3b).

aIC photoinhibition on abstinence-exacerbated CPP expression

To assess whether aIC activity affects CPP expression following different abstinence lengths, subjects were bilaterally infused with a virus encoding eNpHR3.0 or eYFP alone into the aIC. We conducted the previously described amphetamine-induced CPP procedure and delivered 532 nm photoinhibition to aIC during a cue-exposure test 1 or 14 days after the last acquisition session (Fig. 2a, b).

Fig. 2

aIC photoinhibition reduces the abstinence-exacerbated CPP expression. a Schematic of vector delivery and aIC photoinhibition. b Representative images of opsin expression in aIC. c Baseline preference and CPP expression following early or late abstinence. #p < 0.001 vs. eYFP ABS1, ^p < 0.001 vs. eYFP ABS14. d Motor activity during place preference tests

By comparing CPP scores between baseline preference and cue-exposure test (Fig. 2c), three-way mixed ANOVA yielded statistically significant main effects of Conditioning (F(1, 34) = 217.80, p < 0.001) and Abstinence (F(1, 34) = 6.10, p < 0.05), but not for Group (F(1, 34) = 1.09, p = 0.30). Moreover, multiple significant interaction effects were detected between Conditioning by Group by Abstinence (F(1, 34) = 4.95, p < 0.05) and Group by Abstinence (F(1, 34) = 6.56, p < 0.05). By comparing posttest CPP scores, two-way factorial ANOVA found no significant main effects for Group (F(1, 34) = 3.74, p = 0.06) or Abstinence (F(1, 34) = 3.92, p = 0.05) but revealed a significant interaction between these factors (F(1, 34) = 10.43, p < 0.01). Bonferroni multiple comparison analysis showed significantly higher CPP scores on ABS14 compared to ABS1 for eYFP control groups (p < 0.001). However, no such differences were observed between the eNpHR3.0 groups. Additionally, while both groups exhibited similar CPP scores on ABS1, aIC photoinhibition significantly reduced place preference on ABS14 (p < 0.001) compared to eYFP controls.

Regarding motor activity during tests (Fig. 2d), three-way mixed ANOVA found no significant main effects for Conditioning (F(1, 34) = 3.07, p = 0.08), Group (F(1, 34) = 0.00, p = 0.94), Abstinence (F(1, 34) = 2.79, p = 0.10), or any interaction between these factors. Three-way mixed ANOVA on motor activity during acquisition (Fig. S2c) showed a statistically significant main effect of the Pairing session (F(1, 34) = 36.72, p < 0.001), but no significant main effects of Group (F(1, 34) = 0.51, p = 0.47), Abstinence (F(1, 34) = 0.86, p = 0.35), or any interaction between these factors.

These results indicate that all groups displayed higher CPP scores after conditioning than their initial preference. Moreover, while abstinence-exacerbated CPP expression was observed for control conditions, aIC photoinhibition significantly reduced CPP expression after 14 days but did not impair expression after one day of abstinence (Fig. 2c). As previously described, chronic exposure to amphetamine produced similar motor sensitization between groups (Fig. S2c), and no differences in motor activity were observed between tests or groups (Fig. 2d).

aIC-AMY photoinhibition on abstinence-exacerbated CPP expression

To further assess whether aIC-AMY glutamatergic projection modulates CPP expression following different abstinence lengths, subjects were bilaterally infused with eYFP or eNpHR3.0 vectors into the aIC and followed the amphetamine-induced CPP procedure. Photoinhibition was delivered to AMY during a cue-exposure test held after the same periods (Fig. 3a). A quantitative analysis was performed to measure aIC-AMY projection magnitude between CeA and BLA. Although eYFP positive axons were detected on both nuclei (Fig. 3b, c), the mean intensity of aIC axonic projections was significantly higher on BLA compared to CeA (t(8) = 8.04, p < 0.001).

Fig. 3

aIC-AMY photoinhibition reduces the abstinence-exacerbated CPP expression. a Schematic of vector delivery and aIC-AMY photoinhibition. b Representative images of opsin expression in aIC-AMY projection. c Mean intensity of aIC projections to CeA and BLA. ***t-test p < 0.001 d Baseline preference and CPP expression following early or late abstinence. #p < 0.01 vs. eYFP ABS1, ^p < 0.05 vs. eYFP ABS14. e Motor activity during place preference tests

By comparing prior and after conditioning CPP scores (Fig. 3d), three-way mixed ANOVA resulted in statistically significant main effects of Conditioning (F(1, 34) = 246.67, p < 0.001), but not for Group (F(1, 34) = 1.87, p = 0.18) or Abstinence (F(1, 34) = 1.56, p = 0.21). Moreover, a significant interaction effect was detected between Conditioning by Abstinence (F(1, 34) = 7.30, p < 0.05). Two-way factorial ANOVA performed for posttest CPP scores showed a significant main effect of Abstinence (F(1, 34) = 5.12, p < 0.05) and a significant interaction between Group by Abstinence (F(1, 34) = 5.05, p < 0.05), while no main effect of Group was detected (F(1, 34) = 2.48, p = 0.12). Bonferroni multiple comparison analysis showed significantly higher CPP scores on ABS14 (p < 0.01) than ABS1 for eYFP control groups. Nevertheless, no such differences were observed between the eNpHR3.0 groups. Additionally, while both groups expressed similar scores on ABS1, aIC-AMY photoinhibition significantly reduced place preference on ABS14 (p < 0.05).

When comparing motor activity during tests (Fig. 3e), three-way mixed ANOVA found no significant main effects for Conditioning (F(1, 34) = 3.89, p = 0.05), Group (F(1, 34) = 1.03, p = 0.31), Abstinence (F(1, 34) = 0.25, p = 0.61), or any interaction between these factors. Three-way mixed ANOVA on motor activity during acquisition (Fig. S2d) showed a statistically significant main effect of the Pairing session (F(1, 34) = 33.49, p < 0.001), but no significant main effects of Group (F(1, 34) = 0.04, p = 0.83), Abstinence (F(1, 34) = 1.35, p = 0.25), or any interaction between these factors.

These results indicate that all groups exhibited higher CPP scores after conditioning than their initial preference. Moreover, while abstinence-exacerbated CPP expression was observed in control conditions, aIC-AMY photoinhibition significantly reduced CPP expression after 14 days of abstinence, but not after one day (Fig. 3d). Groups developed similar motor sensitization (Fig. S2d), and no differences in motor activity were observed between tests or groups (Fig. 3e).

aIC NMDAR antagonism on abstinence-exacerbated CPP expression

To assess whether aIC NMDAR signaling participates in CPP expression following different abstinence periods, subjects were bilaterally implanted with guide cannulae above aIC (Fig. 4a). Then, we conducted the amphetamine-induced CPP procedure and microinjected NMDAR antagonist AP5 or saline into aIC before a cue-exposure test performed at the same abstinence periods.

Fig. 4

aIC NMDAR antagonism reduces the abstinence-exacerbated CPP expression. a Schematic representation of cannulae end traces. Numbers adjoining each coronal section refer to distances from bregma (adapted from Paxinos and Watson, 2008). Circles, SAL; squares, AP5. b Baseline preference and CPP expression following early or late abstinence. #p < 0.01 vs. SAL ABS1, ^p < 0.01 vs. SAL ABS14. c Motor activity during place preference tests

By comparing pre- and post-conditioning CPP scores (Fig. 4b), three-way mixed ANOVA showed statistically significant main effects of Conditioning (F(1, 40) = 200.67, p < 0.001), but not for Group (F(1, 40) = 2.16, p = 0.14), or Abstinence (F(1, 40) = 0.93, p = 0.33). Moreover, a significant interaction effect was detected between Group by Abstinence (F(1, 40) = 4.23, p < 0.05). Two-way factorial ANOVA performed for posttest CPP scores showed no significant main effects for Group (F(1, 40) = 3.45, p = 0.07), or Abstinence (F(1, 40) = 2.91, p = 0.09) but revealed a significant interaction between these factors (F(1, 40) = 5.29, p < 0.05). Bonferroni multiple comparison analysis showed significantly higher CPP scores on ABS14 (p < 0.01) compared to ABS1 for saline groups. However, no such differences were observed between AP5 groups. Additionally, while both groups expressed similar scores on ABS1, NMDAR antagonism significantly reduced place preference for ABS14 (p < 0.01).

Three-way mixed ANOVA on motor activity during preference tests (Fig. 4c) found no significant main effects for Conditioning (F(1, 40) = 0.02, p = 0.86), Group (F(1, 40) = 0.02, p = 0.86), Abstinence (F(1, 40) = 1.16, p = 0.28), or any interaction between these factors. Regarding motor activity during acquisition (Fig. S2e), three-way mixed ANOVA resulted in a statistically significant main effect of the Pairing session (F(1, 40) = 73.08, p < 0.001), but no significant main effects of Group (F(1, 40) = 0.00, p = 0.95), Abstinence (F(1, 40) = 0.12, p = 0.72), or any interaction between these factors.

These results indicate that all groups displayed higher CPP scores after conditioning compared to their baseline preference. Moreover, while abstinence-exacerbated CPP expression was observed in control groups, aIC NMDAR antagonism significantly reduced CPP expression after 14 days but did not affect the expression after a single day of abstinence (Fig. 4b). During acquisition, motor sensitization developed due to chronic amphetamine treatment, with no differences between groups or abstinence periods (Fig. S2e), and no differences in motor activity were observed between tests or groups (Fig. 4c).