Influence of reconsolidation in maintenance of cocaine-associated contextual memories formed during adolescence or adulthood

Adolescents are at increased risk to develop substance use disorders and suffer from relapse throughout life. Targeted weakening of drug-associated memories has been shown to reduce relapse-like behavior in adult rats, however this process has been understudied in adolescents. We aimed to examine whether adolescent-formed, cocaine-associated memories could be manipulated via reconsolidation mechanisms. To accomplish this objective, we used an abbreviated operant cocaine self-administration paradigm (ABRV Coc-SA). Adult and adolescent rats received jugular catheterization surgery followed by ABRV Coc-SA in a distinct context for 2 h, 2×/day over 5 days. Extinction training (EXT) occurred in a second context for 2 h, 2×/day over 4 days. To retrieve cocaine-context memories, rats were exposed to the cocaine-paired context for 15 min, followed by subcutaneous injection of vehicle or the protein synthesis inhibitor cycloheximide (2.5 mg/kg). Two additional EXT sessions were conducted before a 2 h reinstatement test in the cocaine-paired context to assess cocaine-seeking behavior. We find that both adult and adolescent cocaine-exposed rats show similar levels of cocaine-seeking behavior regardless of post-reactivation treatment. Our results suggest that systemic treatment with the protein synthesis inhibitor cycloheximide does not impair reconsolidation of cocaine-context memories and subsequent relapse during adulthood or adolescence.

Surgery.Intravenous (IV) catheters were constructed in house, with a 10 cm length for adults and 9.7 cm length for adolescents.A silicone ball was placed 3.25 or 3.0 cm from the inserted end of the catheter for adults or adolescents [27][28][29] .On the surgery day, rats were fully anesthetized via intraperitoneal (IP) injection of ketamine + xylazine (80-100 mg/kg + 5-10 mg/kg, respectively; Covetrus, Cats:11695-6840, 33198).Catheters were implanted into the right jugular vein, exiting subcutaneously at the back below the shoulder blade 27,30,31 .Rats were administered oral meloxidyl 2 days pre-and post surgery (Covetrus, Cat:50401), topical lidocaine and gentamycin applied to suture areas, and catheters flushed with 0.1 mL of cefazolin (0.1 mg/mL, Covetrus, Cat:54846) dissolved in heparinized saline (Hep-Sal; 70 U/mL; Covetrus, Cat:49130), followed by 0.1 mL of 10 U/mL Hep-Sal.Catheters were flushed before Coc-SA with 70U Hep-Sal, and after with 10U Hep-Sal.Cefazolin was administered after the last Coc-SA session of the day.When catheter patency was uncertain, 0.1 mL propofol was administered IV and rats monitored for brief loss of movement (Covetrus, Cat:54899).

General Coc-SA training.
As shown in Fig. 1, Coc-SA occurred in an operant conditioning chamber with distinct odor, tactile, auditory and visual stimuli, as previously described 27 .Operant chambers (29.5 × 24 × 28 cm; Med Associates Inc., St. Albans, NY) were configured with the following contextual stimuli: Context A contained a continuous red house light (0.4 fc brightness), intermittent pure tone (80 dB, 1 kHz; 2 s on, 2 s off), pinescented air freshener (Car Freshener Corp., Watertown, NY), and wire mesh flooring (26 cm × 27 cm).Context B contained an intermittent white stimulus light over the inactive lever (1.2 fc brightness; 2 s on, 2 s off), continuous pure tone (75 dB, 2.5 kHz), vanilla-scented air freshener (Car Freshener Corp., Watertown, NY), and a slanted black acrylic panel bisecting the bar flooring (19 cm × 27 cm)..Immediately after MR, rats were given a subcutaneous (SC) injection of vehicle (VEH) or cycloheximide (CHX, 2.5 mg/ kg) and returned to their home cage.We used a 2.5 mg/kg dose based on previous studies which showed that 2.2 mg/kg given immediately after retrieval of a cocaine-cue memories diminished subsequent relapse 21 .To assess whether CHX affected responding in the EXT context, two additional 2 h EXT sessions were administered (post-MR EXT).After the final post-MR EXT, rats were returned to the original cocaine-paired context for a 2 h reinstatement test (Test).During the Test, both active and inactive lever presses resulted in no programmed consequences; therefore, active lever responses served as an index of drug-seeking behavior.On average, MR and Tests occurred on P49 and P51 for adolescent rats and P80 and P82 for adult rats.

Statistical analysis.
Separate analyses of variance (ANOVAs) or independent t-tests were conducted to examine for pre-existing differences between VEH and CHX groups for: lever responses during Coc-SA (Mean ± SEM, final 3 sessions), last EXT session, 15 min MR session, final post-MR EXT session, and number of sessions to reach acquisition criteria for Coc-SA and EXT.
To assess behavioral differences by age, data from adult and adolescent VEH and CHX groups pre-treatment were collapsed.Interaction effects for within subject factors.Coc-SA, EXT, MR and the between subject factor of Age were assessed (Fig. 2).For all behavioral phases (Coc-SA, EXT, MR, final post-MR EXT and Test), no Age × Context × Treatment interaction effects were found and therefore subsequent ANOVAs were analyzed separately for adult or adolescent groups.
To assess behavioral differences for adults (Fig. 3) or adolescents (Fig. 4), interaction effects for within-subjects factors: Coc-SA, EXT, MR, Context (final post-MR EXT session vs Test), 20 min Test Bins, 5 min MR or 5 min Test Bins and between subjects factor of Treatment (VEH vs CHX) were conducted.Significant effects, when

Results
Adolescent and adult behavior during Coc-SA, EXT, MR, and tests.To examine potential behavioral differences between adult and adolescent rats, lever response data for all phases before VEH or CHX administration (Coc-SA, EXT, MR) were combined by age.Overall, there were no pre-existing differences between adult and adolescent rats in the number of sessions to acquire Coc-SA, lever responses, or cocaine infusions during the last three sessions of Coc-SA (Fig. 2A).The 2 × 10 ANOVAs during Coc-SA did not reveal significant Coc-SA session × Age or Coc-SA infusion × Age interaction effects.During EXT, there were no pre-existing differences between adult and adolescent rats in the number of sessions needed to complete EXT training, or lever responses on the last EXT session (Fig. 2B).The 2 × 8 ANOVA of lever responses revealed a significant EXT session main effect (active: F 7, 28 = 44.395,p < 0.001; inactive: F 7, 28 = 28.558,p < 0.001), but no EXT session × Age interaction or Age main effects.Tukey's post-hoc comparisons revealed that both adult and adolescent groups decreased lever responses by the final EXT session (active and inactive: EXT S1 > S8, # p < 0.01).
Adult and adolescent rats had similar lever responses during the 15 min MR session before systemic VEH or CHX administration (Fig. 2C).The 2 × 3 ANOVA of lever responses during 5 min Bins of the 15 min MR session revealed a significant MR Bin main effect (active: F 2, 28 = 22.226, p < 0.001; inactive: F 2, 28 = 3.646, p = 0.032) with no MR Bin × Age interaction or Age main effects.Tukey's post-hoc comparisons revealed that both adult and adolescent groups displayed reduced responses on the active lever by the final 5 min of MR (active: Bin 1 > 3, # p < 0.05) (Suppl Fig. 2A).Tukey's post-hoc comparisons did not reveal significant differences in inactive lever responses during 5 min MR Bins (Suppl Fig. 2A).
To examine for potential behavioral differences between adult and adolescent rats during the reinstatement Test after VEH or CHX administration, we compared lever response data for (1) VEH groups only or (2) VEH and CHX groups combined by age and observed similar results.Therefore we presented data for all groups.Adult and adolescent groups displayed increased cocaine-seeking behavior when exposed to the cocaine-paired context (Fig. 2D).The 2 × 2 ANOVAs of active lever responses during the final post-MR EXT vs Test session revealed a significant Context main (active: F 1, 14 = 56.826,p < 0.001), but no Context × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that lever responses in the VEH and CHX groups were higher during the Test compared to the final post-MR EXT session (active: # p < 0.01) therefore, both VEH and CHX-treated groups exhibited cocaine-seeking behavior.The 2 × 2 ANOVAs of inactive lever responses during the final post-MR EXT vs Test session did not reveal significant Context × Treatment interaction, Context or Treatment main effects.
The 2 × 6 ANOVA of lever responses during 20 min Test Bins of the 2 h Test revealed significant Test Bin main (active: F 5, 28 = 42.136,p < 0.001; inactive: F 5, 28 = 13.150,p < 0.001), with no Test Bin × Age interaction or Age main effects (Fig. 2E).Tukey's post-hoc comparisons revealed that both adult and adolescent groups decreased active lever responses by the final 20 min of the Test (active: Bin 1 > 6, # p < 0.01).Tukey's post-hoc comparisons did not reveal significant differences in inactive lever responses during 20 min Test Bins.
For all behavioral phases, no Context × Age × Treatment interaction effects were found, therefore adult and adolescent groups were further independently analyzed.

Adult behavior during Coc-SA, EXT, MR and tests: effect of CHX on reconsolidation of cocaine context memories.
Adult rats in the VEH (n = 8) and CHX (n = 8) groups displayed similar responses on both levers during Coc-SA, with no pre-existing differences observed in the number of sessions to acquire Coc-SA, lever responses or cocaine infusions during the last three sessions of Coc-SA (Fig. 3A).The 2 × 10 ANOVAs during Coc-SA did not reveal significant Coc-SA session × Treatment or Coc-SA infusion × Treatment interaction effects.
Adult rats in the VEH and CHX groups had similar lever responses during EXT training (Fig. 3B).The 2 × 8 ANOVAs of lever responses revealed a significant EXT session main effect (active: F 7, 14 = 26.439,p < 0.001; inactive: F 7, 14 = 13.301,p < 0.001), with no EXT session × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that both groups decreased lever responses by the final EXT session (active and inactive: EXT S1 > S8, # p < 0.01).
Adult rats in the VEH and CHX groups had similar lever responses during the MR session (Fig. 3C).The 2 × 3 ANOVA of lever responses during 5 min bins of the 15 min MR session revealed a significant MR Bin main effect (active: F 2, 14 = 14.646, p < 0.001) with no MR Bin × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that both VEH and CHX groups displayed reduced active lever responses by the final 5 min of MR (active: Bin 1 > 3, # p < 0.01).Tukey's post-hoc comparisons did not reveal significant differences in inactive lever responses during 5 min MR Bins (Suppl Fig. 3A).
Adult rats in VEH and CHX groups displayed increased cocaine-seeking behavior when exposed to the cocaine-paired context (Fig. 3D).The 2 × 2 ANOVAs of active lever responses during the final post-MR EXT vs Test session revealed a significant Context main (active: F 1, 14 = 56.826,p < 0.001), but no Context × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that lever responses in the VEH and CHX groups were higher during the Test compared to the final post-MR EXT session (active: # p < 0.01) therefore, both VEH and CHX-treated groups exhibited cocaine-seeking behavior.The 2 × 2 ANOVAs of inactive lever responses during the final post-MR EXT vs Test session did not reveal significant Context × Treatment interaction, Context or Treatment main effects.

Adolescent behavior during Coc-SA, EXT, MR and tests: effect of CHX on reconsolidation of cocaine context memories.
Adolescent rats in the VEH (n = 6) and CHX (n = 8) groups displayed similar responses on both levers with no pre-existing differences observed in the number of sessions to acquire Coc-SA, lever responses, or cocaine infusions during the last three sessions of Coc-SA (Fig. 4A).The 2 × 10 ANOVAs during Coc-SA did not reveal significant Coc-SA session × Treatment or Coc-SA infusion × Treatment interaction effects.Adolescent rats in the VEH and CHX groups had similar lever responses during EXT training (Fig. 4B).The 2 × 8 ANOVA of lever responses revealed a significant EXT session main effect (active: F 7, 12 = 19.713,p < 0.001; inactive: F 7, 12 = 14.304, p < 0.001), with no EXT session × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that responding decreased in both VEH and CHX groups by the final EXT session (active and inactive: EXT S1 > S8, # p < 0.01).
Adolescent rats in the VEH and CHX groups had similar lever responses during the MR session (Fig. 4C).The 2 × 3 ANOVA of lever responses during 5 min bins of the 15 min MR session revealed significant MR Bin main effect (active: F 2, 12 = 8.311, p = 0.002) with no MR Bin × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that only VEH groups displayed reduced active lever responses by the final 5 min of MR (VEH: Bin 1 > 3, # p < 0.01; CHX: Bin 1 ≠ 3, 2 ≠ 3).Tukey's post-hoc comparisons did not reveal significant differences in inactive lever responses during 5 min MR Bins (Suppl Fig. 4A).
Adolescent rats in the VEH and CHX groups displayed increased cocaine-seeking behavior when exposed to the cocaine-paired context (Fig. 4D).The 2 × 2 ANOVA of active lever responses during the final post-MR EXT vs Test sessions revealed a significant Context main (active: F 1, 12 = 31.633,p < 0.001; inactive: F 1, 12 = 7.010, p = 0.021) with no Context × Treatment interaction or Treatment main effects.Tukey's post-hoc comparisons revealed that lever responses in the VEH and CHX groups were higher during the Test compared to the final post-MR EXT session (active: # p < 0.01; inactive: # p < 0.05) therefore, both VEH and CHX-treated adolescent groups exhibited cocaine-seeking behavior.

Discussion
Overall, systemic protein synthesis inhibition after retrieval of cocaine-associated contextual memories did not reduce relapse-like behavior in adult or adolescent cocaine-exposed groups.The current results did not reveal age-dependent differences during operant phases of behavioral training (Fig. 2).Adult and adolescent rats had similar lever responses during Coc-SA and EXT training phases and these data are in line with previous reports in which self-administration of cocaine occurred in the presence of diffuse environmental stimuli or paired with explicit cues [27][28][29]35,36 .
There is evidence that adolescent rats have increased resistance to extinguish fear and cocaine-related behaviors 24,29,37 .A previous study determined that adolescent rats were less responsive to a rodent version of cue-exposure therapy (CET).In this study, investigators combined a Coc-SA paradigm with passive presentation of a cocaine-associated cue to mimic human CET and observed that adult and adolescent rats displayed similar levels of EXT training behavior, as observed in our study, but were resistant to passive cue extinction compared to adults 29 .We have also noted a potential resistance to extinguish lever responses in adolescent rats within the original cocaine-paired context, but not within the EXT context 28 .Similar to our previous reports, adolescent rats had higher total inactive lever presses during reinstatement tests; however both age groups displayed increased inactive presses during the first 20 min bin of the reinstatement test.It is possible that across the entire test session, adolescent cocaine-exposed rats had increased motivation or response generalization to both levers in the drug-paired context 38,39 .
We modified our ABRV Coc-SA paradigm to incorporate standard procedures to test whether protein synthesis was required for the reconsolidation of adult-and adolescent-formed cocaine-context memories.Importantly, lever responses were similar between adult and adolescent cocaine-exposed rats during a reactivation session used to retrieve cocaine-context memories.We chose 15 min of exposure to the previous cocaine-paired context since this timeframe is sufficient to elicit retrieval in adult rats without engaging extensive extinction processes 11,12,40 .We examined 5 min intervals of the 15 min MR session but did not observe age-dependent differences in the curves of lever responses (Suppl Fig. 2A) or in the latency for the first lever press during MR; therefore, the MR session likely served to retrieve cocaine-associated memories to a similar extent for both age groups.Overall, operant lever response behavior was similar between adult and adolescent rats during Coc-SA and EXT training and MR, which suggests that an age-dependent difference in the strength of drug-context associations did not contribute to the lack of effect of CHX on reconsolidation and subsequent relapse.
The current results revealed that adult and adolescent cocaine-exposed rats had similar magnitudes of contextinduced cocaine-seeking behavior regardless of post-retrieval treatment with VEH or CHX (Figs. 3, 4).Surprisingly, CHX treatment failed to reduce cocaine-seeking behavior in adult rats which contrasts with reports of attenuated cocaine reward and cue-induced reinstatement following systemic CHX administration 21,41 .A subcutaneous dose of 1.0 mg/kg CHX has been shown to reduce protein synthesis in the brain by up to 57% 42 .A range of CHX doses, from 1.0 to 15 mg/kg, has been used for cocaine CPP and operant self-administration studies.A systemic dose of 2.2 mg/kg CHX reduced cue-induced cocaine-seeking behavior; whereas a 1.0 mg/kg dose did not reduce seeking behavior 21 .Higher doses of systemic CHX (15 mg/kg) resulted in reduced cocaine CPP 41,43 .We therefore chose the lowest dose of CHX that was within the range expected to observe behavioral effects, to avoid potential negative effects related to toxicity with protein synthesis inhibitors in general 44,45 .Weight loss has been noted after SC administration of CHX at a dose of 1.5 mg/kg; therefore, we examined weight at similar intervals-2 days before and after VEH or CHX administration 45 .Adult and adolescent groups both gained weight at similar magnitudes, with no reductions in weight observed in CHX-treated groups (Suppl Fig. 1).The most parsimonious explanation would suggest unsuccessful CHX delivery, however differences in experimental procedures and food delivery could also account for these results.Rats in the current study received restricted, rather than ad libitum feed.Both age groups gained weight on restricted feed, but in combination with a history of cocaine intake, our experimental procedures may have obscured the ability to detect subtle differences in weight loss due to CHX administration.
The current study used comparable lengths of memory reactivation and doses of CHX, but a major difference from previous reports was the type of stimuli used to elicit retrieval.Previous studies that reported CHX-induced impairment in reconsolidation and drug-seeking behavior used a cocaine-associated explicit cue to elicit retrieval, while the current study exposed rats to a previous cocaine-paired context.It is known that different brain regions and circuits support cue vs contextual cocaine-seeking behavior 7,32,46 .Furthermore, signaling cascades that are important for memory reconsolidation are shown to be differentially involved in cue vs contextual retrieval and subsequent relapse behavior.For example, post-MR inhibition of protein kinase A activity in the BLA impairs both cue-and context-elicited reconsolidation and subsequent cocaine-seeking; whereas post-MR inhibition of CAMKII in the BLA only impairs cue-elicited drug-seeking behavior 12,18,40 .Another difference from previous studies is that reinstatement Tests in the current study were conducted 2 days after MR and CHX treatment, whereas previous reports conducted Tests 3 days after MR + CHX treatment.The lack of a behavioral effect on relapse from adult systemic CHX manipulation likely is specific to retrieval of contextual vs cue memories, as has also been observed in the fear conditioning literature 47 .
While beyond the scope of the current work, future studies could focus on use of the unconditioned stimuli (US) of cocaine to retrieve and destabilize cocaine-associated memories.Previous reports that used the US to retrieve and destabilize both drug CPP and operant drug-associated memories observed reduced reward and cue-induced cocaine-seeking in adult rats [48][49][50][51] .Use of the US to retrieve cocaine-linked memories could be a particularly relevant approach in adolescent rats, given that this age group displays higher stress and cocaineprimed reinstatement compared to their adult counterparts, suggesting potential age differences in reactivity to external vs interoceptive cues 33,52 .In line with modifying the stimuli used to retrieve cocaine-context memories, future studies could present a CS-US pairing that differs in duration from the original CS-US used during training to provide for a larger prediction error during the memory retrieval session 53 .
Lastly, it is also possible that systemic administration of CHX may have exerted effects in several brain regions involved in reconsolidation and motivated behavior, thereby obscuring potential reductions in cocaine-seeking behavior.A large body of literature has demonstrated brain-specific reconsolidation mechanisms contribute to relapse-like behavior 11,12,20,34,54 .Delivery of protein synthesis inhibitors to specific brain regions to investigate adult vs adolescent differences in drug memory reconsolidation may be a more promising avenue for future experiments.

Conclusions and future directions
The current results suggest that both adult and adolescent-formed cocaine-context associations were not sensitive to, or weakened by, reconsolidation manipulations via systemic protein synthesis inhibition.A large body of literature demonstrates that brain-specific reconsolidation mechanisms contribute to cocaine-associated memories and relapse.Therefore, future directions should investigate the brain region specific mechanisms that strengthen cocaine-context memories in adult and adolescent rats.We recently demonstrated that adolescent cocaineexposed rats display a time-dependent increase in context-induced cocaine-seeking behavior after 15 days of abstinence from cocaine intake, whereas adult cocaine-exposed counterparts do not 28 .Future studies could aim to use similar procedures of memory reactivation in a cocaine-paired context after extinction training and then examine whether reduced context-induced incubation of craving is observed in adolescent cocaine-exposed rats.

Figure 3 .Figure 4 .
Figure 3.Effect of cycloheximide on reconsolidation of cocaine-context memories in adult cocaine-exposed rats.Mean ± SEM of active and inactive lever responses for adult rats during (A) cocaine self-administration (Coc-SA, 2 h sessions, 2×/day over 5 days), (B) extinction (EXT, 2 h sessions, 2×/day over 4 days), (C) memory reactivation (MR, 15 min session) in the previous cocaine-paired context to retrieve cocaine-associated memories.Adult rats received subcutaneous injection of vehicle or 2.5 mg/kg cycloheximide (VEH, CHX) immediately after the MR session.(D) Two additional sessions of EXT were given 1 day after MR (post-MR EXT), followed by a reinstatement test in the previous cocaine-paired context (Test, 2 h).(E) Active and inactive lever responses during 20 min Bins of the 2 h Test.Symbols indicate significant within-subject differences revealed by Tukey's test (B) # p < 0.01: EXT session 1 > 8, (D) final post-MR EXT session < Test, (E) 20 min Test Bin 1 > 6. Groups denoted by: Light blue = adult VEH (n = 8), Black = adult CHX (n = 8).

training, memory reactivation and reinstatement tests. As
Schematic and timeline of behavioral experiments to examine reconsolidation of cocaine-context memories in adult and adolescent cocaine-exposed rats.All experiments included abbreviated cocaine selfadministration training (Coc-SA, 2 h sessions, 2×/day over 5 days, minimum of 10 sessions) in a distinct context, followed by extinction training (EXT, 2 h sessions, 2×/day over 4 days, minimum of 8 sessions).On average, adolescent rats started Coc-SA at P38 and adults at P70.The average age at the start of EXT was P45 for adolescent rats and P76 for adult rats, due to differences in the number of sessions needed to acquire Coc-SA criteria.To retrieve cocaine-associated contextual memories, rats were returned to the previous cocaine-paired context for memory reactivation (MR, 15 min session), followed by a subcutaneous injection of vehicle or 2.5 mg/kg cycloheximide (VEH, CHX) immediately after the MR session.Two additional sessions of EXT were given 1 day after MR (post-MR EXT), followed by a reinstatement test in the previous cocaine-paired context (Test, 2 h).
shown in Fig. 1, extinction training (EXT) consisted of 2 h sessions, 2×/day over 4 days, for a minimum of 8 sessions (criteria = less than 25 responses on last 2 sessions).During EXT training, responses on the active and inactive lever had no programmed consequences.Rats received a 15 min memory reactivation session (MR) in the cocaine-paired context.The MR session functions to elicit brief memory reactivation, without extensive extinction responses during 5 min Bins of the 2 h Test did not reveal Bin × Treatment interaction, Bin or Treatment main effects at the active or inactive levers (Suppl Fig. ) Scientific Reports | (2023) 13:13936 | https://doi.org/10.1038/s41598-023-39949-ywww.nature.com/scientificreports/