Male and female adult Wistar rats (n = 64 per sex) were obtained from Charles River Laboratories aged 8 weeks old (Raleigh, NC, USA) and were single-housed in a 12-hr reverse light cycle humidity- and temperature-controlled (22°C) colony room. Rats were allowed to acclimate to the colony room for 4 days prior to the start of surgeries. Animals were handled daily following surgery. Testing occurred during the dark period (between 8:00 am and 8:00 pm). All animals were provided with additional enrichment (Nestlet, BedderNest, and Nylabone) in their home cages throughout the study. Food was available ad libitum except during the first 4 days of the experiment while lever training. Water was always available. All procedures were approved by the Institutional Animal Care and Use Committee of the Louisiana State University Health Sciences Center and were in accordance with the National Institute of Health guidelines. For final analyses, 39 total rats were excluded due to either loss of catheter patency, post-surgical complications, or outlier behavioral responses.

Rats were anesthetized with 4% isoflurane gas and maintained on 2.5% isoflurane gas throughout the surgery. They received an implanted chronic indwelling catheter into the right jugular vein using catheters made in house. The catheters were inserted under the skin directly posterior to the shoulder blades and the Silastic tubing was threaded over the right shoulder and inserted into the right jugular vein. Catheter cannulas were covered with a small piece of clean plastic tubing and a metal screw top to prevent debris from entering the catheter. Rats were allowed to recover from surgery for 1 h on a heating pad before being returned to a clean home cage. 3 mL of sterile saline was administered subcutaneously to aid in recovery time. Flunixin meglumine (Covetrus; Portland, Maine, USA) was administered at the start of the surgery subcutaneously (2.5 mg/kg). After the sterilized (Maxicide supplied by Henry Schein; Mandeville, LA, USA) catheter was set in place, the catheter was flushed with 0.2 mL of cefazolin/heparin (1,000 USP units per mL) solution to prevent infections and catheter clogging. Heparin was supplied by Sagent Pharmaceuticals; Schaumburg, IL, USA, and cefazolin was supplied by WG Critical Care LLC.; Paramus, New Jersey, USA. Post-operative care included weighing and visually inspecting animals daily after surgery and a catheter flush with 0.2 mL cefazolin/heparin solution once daily until the end of the experiment. Rats were allowed to recover from surgery for at least 5 days before starting experiments.

Cocaine (-) HCl was generously provided by the National Institute for Drug Abuse (NIDA) Drug Supply Program and Research Triangle Institute (RTI). Cocaine was dissolved in sterile saline and delivered at a dose and volume of 0.5 mg/0.1 mL/infusion. Rats were weighed daily before the start of each session and unit cocaine dose for each rat was calculated based on body weight.

To establish a baseline of cocaine intake per animal and facilitate lever-response training, rats were food restricted (20 g/day for males and 15 g/day for females) 1 day prior to the start of acquisition sessions and during the first 3 days of acquisition. Rats were placed into sound- and light-attenuated operant-conditioning chambers (30 cm × 20 cm × 24 cm; Med Associates Inc., St. Albans, VT) for 2 h/day for 10 consecutive days beginning at the same time each day. Rats began sessions 1–3 h into the dark cycle every day for 7 days/week. Each chamber contained two retractable levers, a stimulus light above each lever, a food trough between the levers, a fan, and a speaker connected to a tone generator (ANL-926, Med Associates). A single lever press on the active lever (right lever) resulted in the delivery of a single cocaine infusion (FR1; 0.5 mg/kg/0.1 mL) along with a 20 s presentation of 1) light above the active lever, 2) tone cue, and 3) 20 s timeout period during which time no subsequent cocaine could occur. Presses of the left lever had no consequence but were recorded. To meet criteria for progression to the next phase of the study, rats were required to press the active lever at least 10 times during a 2-hr session for the final three consecutive days of acquisition (days 8–10)—most animals achieved this during the 10-day baseline period.

One day after the final day of the acquisition/baseline period, predator odor place conditioning began. No cocaine was given during this part of the experiment. All rats completed conditioned place aversion (CPA) procedure as previously described (38, 45) using a 3-chamber apparatus. Briefly, on day 1 of this procedure, each rat is given 5 minutes to freely explore the three connected chambers in the apparatus (3-chamber pre-test), each with unique visual cues on the walls and tactile cues on the floor. To maintain an unbiased design, the chamber each rat preferred most or least relative to the other two was excluded for the remainder of the experiment. On day 2, each rat explored the two chambers it spent the most similar amounts of time in during day 1 for 5 minutes (2-chamber pre-test). Rats were then separated into predator odor-exposed or unstressed control groups. Again, to maintain an unbiased design, the chamber in which odor exposure would occur was assigned in a counterbalanced fashion based on chamber preference. On day 3 of the procedure, rats were confined to the “no odor” chamber in the absence of any odor for 15 min. On day 4 of the procedure, rats were confined to the “odor” chamber for 15 min, during which time rats in the odor group were exposed to bobcat urine odor and rats in the control group were exposed to no odor (same as day 3). On day 5 of the procedure, rats were once again allowed to freely move about the two chambers (2-chamber post-test). Rats were classified as “Avoiders,” or “Non-avoiders” based on the differences in time spent in the “odor” chamber between the 2-chamber pre-test and post-test. Rats with greater than a 10 s decrease in time spent in the “odor” chamber from pre-test to post-test were categorized as “Avoiders,” and all other odor-exposed animals were categorized as “Non-avoiders.” Unstressed controls were never exposed to bobcat urine.

Rats were assigned to ShA or LgA self-administration groups based on their avoidance score in a counterbalanced fashion, such that unstressed controls, Non-avoiders, and Avoiders were equally represented in the two cocaine access groups. One day following the 2-chamber post-test of the place conditioning procedure, rats were once again allowed to self-administer cocaine (0.5 mg/0.1 mL/infusion) on an FR1 schedule of reinforcement for either 1 h/day (ShA) or 6 h/day (LgA) for 10 consecutive days in the same operant chambers used for acquisition and baseline. Specific operant boxes were used for males only, and others were used for females only during the entirety of the experiment. At the end of the study, 1 day after the 10th day of self-administration, rats were allowed to self-administer cocaine for 1 h and then anesthetized with 4% isoflurane gas, decapitated, and the brains were flash frozen in 2-methylbutane for future analysis.

Catheter patency was checked once per week by flushing catheters with 0.2 mL Brevitol Sodium (distributed by Henry Schein; Mandeville, LA, USA) and observing body response. Rats that failed to go limp after 3 s were excluded from the study.

All data was analyzed using Prism Graph Pad version 9.0 (GraphPad Software, Inc., La Jolla, CA, USA). Data were analyzed using ANOVAs with stress groups, sex, and self-administration access groups as between-subjects factors and session days as a within-subjects factor when appropriate. Significant effects were followed up with Tukey’s or Bonferonni post hoc tests where appropriate (Bonferonni post hoc tests were used for all repeated measures ANOVAs). Outliers were detected using the interquartile range (IQR) rule, and data analyzed after outliers were detected were analyzed using Mann-Whitney test. Correlations were performed using Pearson correlations. Statistical significance was set at p < 0.05.

Adult male and female Wistar rats were exposed to bobcat urine and indexed for avoidance behavior of the odor-paired chamber (see Figure 1), and 62% of males and 46% of females were categorized as Avoiders. A Pearson correlation revealed no association between avoidance and cocaine deliveries in ShA stressed males (r = 0.30; p = 0.37), LgA stressed males (r = 0.25; p = 0.41), ShA stressed females (r = −0.05; p = 0.89), or LgA stressed females (r = −0.07; p = 0.83). We analyzed the effects of avoidance phenotype and time on cocaine deliveries using separate two-way repeated measures (RM) ANOVAs for LgA males, LgA females, ShA males, and ShA females. There was no effect of avoidance phenotype on post-stress cocaine deliveries in any of these groups (p > 0.05 in all cases). Because we did not find differences in cocaine intake between Avoiders and Non-avoiders, we collapsed Avoiders and Non-avoiders into a single group of “Stressed” rats for each sex, and we compared stressed rats to unstressed controls in all subsequent analyses.

A 2-way RM ANOVA showed that LgA males escalated cocaine drug deliveries regardless of stress history (F [2.41, 55.64] = 6.29; p < 0.01), but there was no effect of stress (F [1.23] = 0.02; p = 0.88) (Figure 2A). Neither stress (F [1.20] = 2.11; p = 0.16) nor day (F [1.93, 38.53] = 1.92; p = 0.16) had effects on ShA males’ cocaine deliveries (Figure 2A). Figure 2B highlights that LgA males exhibit greater increases in cocaine deliveries over time relative to ShA males (F [1.43] = 27.94; p < 0.0001) in the absence of stress effects (F [1.43] = 1.90; p = 0.18).

A separate 2-way RM ANOVA showed that LgA females escalated cocaine deliveries regardless of stress history (F [2.15, 45.22] = 26.18; p < 0.0001; Figure 2C), and again there was no effect of stress (F [1.21] = 0.21; p = 0.65). A separate 2-way RM ANOVA in ShA females revealed a main effect of day (F [2.48, 42.13] = 3.29; p = 0.04) but not of stress (F [1.17] = 0.05; p = 0.83) (Figure 2C). Figure 2D highlights LgA females exhibit greater increases in cocaine intake over time relative to ShA females (F [1.38] = 54.74; p < 0.0001) in the absence of stress effects (F [1.38] = 2.738; p = 0.11).

In separate analyses, we compared self-administration across sexes in animals with the same access conditions (LgA or ShA) collapsed across stress history. In a 2-way RM ANOVA analysis of cocaine delivery data in LgA animals, we observed that all rats escalate cocaine deliveries over 10 days (F [2.50, 115.0] = 22.93; p < 0.0001; Figure 3A). There was no main effect of sex (F [1.46] = 3.35; p = 0.07; Figure 3A). ShA males and females did not differ in intake (F [1.39] = 0.23; p = 0.64), but there was a main effect of day (F [2.46, 95.82] = 3.81; p = 0.02) on ShA cocaine delivery showing all rats escalated cocaine deliveries over time. A 2-way ANOVA on cocaine deliveries in LgA rats showed LgA males and females take more cocaine during the second half of post-CPA self-administration (days 6–10) than during days 1–5 (D1-5) of post-CPA self-administration (main effect of access days: F [1.46] = 21.52; p < 0.0001; Figure 3B). In LgA rats, females take more cocaine than males during self-administration days 6–10 (D6-10) (days x sex interaction: F [1.46] = 2.750; p < 0.01; Figure 3B).

A follow-up 2-way ANOVA highlights that LgA rats display greater escalation of cocaine deliveries than ShA rats (F [1.85] = 76.46; p < 0.0001; Figure 3C). LgA males and females escalated drug deliveries similar amounts across all 10 days post-CPA self-administration (F [1.85] = 1.06; p = 0.31). When comparing total cocaine deliveries, we found all LgA rats received more cocaine deliveries than ShA rats (F [1.85] = 876.9; p < 0.0001; Figure 3D). There was not a significant difference between sexes in total deliveries in 10 days, but there was a trend that showed females self-administered more than males (F [1.85] = 3.53; p = 0.06; Figure 3D). Figure 3E illustrates LgA males and females increase total active lever presses (with and without drug delivery) over time (F [3.20, 107.5] = 4.21; p < 0.01) with no sex differences (F [1.34] = 2.96; p = 0.09). There was no change in active lever presses across days (F [1.81, 51.56] = 1.07; p = 0.34) nor sex (F [1.30] = 0.04; p = 0.84) in ShA rats (Figure 3E). A 3-way ANOVA on total active lever presses during the time-out period, when no cocaine was available, revealed that LgA rats press the active lever more during 20 s timeouts than ShA rats (F [1.60] = 18.70; p < 0.001; Figure 3F); there was no effect of stress (F [1.60] = 0.11; p = 0.74) nor sex (F [1.60] = 0.013; p = 0.91) on responding during time-out periods.

Figure 3G shows inactive lever presses (lever presses resulting in no reinforcer delivery) for all LgA and ShA males and females during post-CPA sessions. We used a 2-way RM ANOVAs to analyze sex and time effects on inactive lever presses in LgA and ShA animals. After removing outliers (1 male) identified using the IQR test, a Mann-Whitney test revealed that there were no differences in inactive lever pressing during post-CPA self-administration between any group. There was no time x sex interactions between LgA males and LgA females (F [9, 288] = 0.82; p = 0.60), and no time x sex interactions between ShA males and females F [9, 225] = 0.87; p = 0.56). There was no difference in inactive lever presses between ShA and LgA animals (F [27, 513] = 0.65; p = 0.91). Outlier data appeared to be associated with traditional stimulant-induced stereotypy behavior.

Collapsing across stress groups, in the ShA male, ShA female, and LgA female groups, baseline cocaine intake (average cocaine intake [mg/kg] during acquisition days 8, 9, and 10) was positively correlated with post-CPA cocaine intake (ShA males [r = 0.66; p < 0.001; Figure 4A]; ShA females [r = 0.57; p = 0.01; Figure 4C]; LgA females [r = 0.50; p = 0.01; Figure 4D]). LgA males exhibited a trend toward the same correlation (r = 0.40; p = 0.058; Figure 4B).

Rats were weighed daily throughout the experiment. We analyzed percent change in body weight during the pre-CPA cocaine acquisition period (2 h/day cocaine access), and during the post-CPA LgA (6hrs/day cocaine access) and ShA (1 h/day cocaine access) period. Males and females increase body weight similarly during acquisition (unpaired t-test: p = 0.59; Figure 5A). During post-stress self-administration, all (control and stressed combined) LgA males gained less weight than all ShA males (F [1.85] = 7.04; p < 0.01); and there was a sex x access condition interaction effect on body weight gain (F [1.85) = 8.38; p < 0.01; Figure 5B).

Next, we collapsed data across LgA and ShA conditions to examine acute effects of predator odor exposure (i.e., during the 5-day conditioning procedure) on body weight gain in male and female rats. When comparing body weight (g) 48 h after the CPA test (i.e., day 1 of post-CPA self-administration prior to the start of the operant session) to body weight prior to day 1 of the conditioning procedure, we observed that stressed males and females gain less weight than controls (F [1.85] = 5.59; p = 0.02; Figure 5C), and females gained less weight overall than males (F [1.85] = 8.16; p < 0.01; Figure 5C) during the 5 days of the conditioning procedure, presumably due to acute stress effects of the predator odor exposure.

Finally, we used a 2-way ANOVA to analyze percent body weight gain in male rats during the 10 days of post-CPA self-administration. LgA males gain less weight than ShA males (F [1.43] = 20.85; p < 0.0001; Figure 5D), and stressed males gain more weight than control males during the 10 days of post-CPA self-administration (F [1.43] = 8.41; p < 0.01; Figure 5D). A separate 2-way ANOVA of percent body weight gain in female rats during the 10 days of post-CPA self-administration revealed there was no difference in post-CPA body weight gain between LgA and ShA females (F [1.38] = 0.002; p = 0.97), nor was there a stress effect (F [1.38] = 3.28; p = 0.08; Figure 5E) on body weight gain in females during the 10-day post-CPA self-administration period.