Chronic inflammatory pain promotes place preference for fentanyl in male rats but does not change fentanyl self-administration in male and female rats

Abstract

The current opioid epidemic is a national health crisis marked by skyrocketing reports of opioid misuse and overdose deaths. Despite the risks involved, prescription opioid analgesics are the most powerful and effective medications for treating pain. There is a clear need to investigate the risk of opioid misuse liability in male and female adults experiencing chronic pain. In the present study, we tested the hypothesis that chronic inflammatory pain would increase fentanyl intake, motivation to acquire fentanyl, and drug seeking in the absence of fentanyl in rats. Fentanyl intake, motivation for fentanyl, and drug seeking were tested under limited and extended access conditions using intravenous fentanyl self-administration. Fos activity in ventral tegmental area (VTA) dopamine neurons following intravenous fentanyl challenge (35 μg/kg) was examined using immunohistoehemistry. Finally, we tested whether low-dose fentanyl supports development of conditioned place preference under an inflammatory pain state in rats. Contrary to our hypothesis, fentanyl self-administration and VTA Fos activity were unaffected by inflammatory pain status. During acquisition, males exhibited increased fentanyl intake compared to females. Animals given extended access to fentanyl escalated fentanyl intake over time, while animals given limited access did not. Males given extended access to fentanyl demonstrated a greater increase in fentanyl intake over time compared to females. During the dose-response test, females given limited access to fentanyl demonstrated increased motivation to acquire fentanyl compared to males. Both sexes displayed significant increases in responding for fentanyl as unit fentanyl doses were lowered. Following fentanyl challenge, females exhibited higher numbers of Fos-positive non-dopaminergic VTA neurons compared to males. Using conditioned place preference, we found that chronic inflammatory pain promotes fentanyl preference in males, but not females. These findings suggest that established fentanyl self-administration is resistant to change by inflammatory pain manipulation in both sexes, but chronic inflammatory pain increases the rewarding properties of low-dose fentanyl in males.

1. Introduction

The current opioid epidemic is a national health crisis marked by skyrocketing reports of opioid misuse, opioid use disorder diagnoses, and overdose deaths (2019 National Survey on Drug Use and Health, 2020; Hedegaard et al., 2020). The opioid epidemic has been defined by three waves of increasing mortality with the first wave resulting in a rapid rise in prescription opioid use and opioid-related deaths (Ciccarone, 2019; Kolodny et al., 2015). The second wave of the opioid epidemic was associated with an increase in illicit heroin use and rising heroin overdose deaths as a portion of individuals switched from prescription pill use to illicit heroin use as an easier to attain and low-cost alternative (Ciccarone, 2019; Cicero et al., 2015; Mars et al., 2014). Currently in its third wave, the opioid epidemic involves a massive rise in overdose deaths from synthetic opioids including fentanyl and its potent analogs, which were introduced into the illicit heroin and counterfeit opioid pill supplies (Ciccarone, 2019). Recent epidemiological information indicates that the third wave of the opioid epidemic has continued to worsen during the COVID-19 pandemic (Hedegaard et al., 2021), and it is clear that addiction risks are just as present for pain sufferers as they are for others (Volkow et al., 2018).

Despite the risks involved, prescription opioid analgesics including synthetic opioids are the most powerful and effective medications for treating acute pain as well as surgical pain and are widely used for end-of-life palliative care and cancer pain (Corder et al., 2018; Pahng and Edwards, 2021). High incidence of chronic pain is a parallel health crisis that impacts an estimated 20% of the population with 8–10% of individuals experiencing high impact chronic pain, which limits their vocational and recreational activities (Dahlhamer et al., 2018; Goldberg and McGee, 2011; Yong et al., 2021). More recent reports have recommended non-opioid medications as the preferred treatments for chronic pain due to the limited efficacy and high abuse potential of opioids for the treatment of chronic pain (Chou et al., 2015; Dowell et al., 2016; Krebs et al., 2018). Part of the limited efficacy of opioids for long-term use can be attributed to the development of analgesic tolerance (Martyn et al., 2019). Extensive opioid use can lead to development of analgesic tolerance as well as the development of hyperalgesia, which is a pronociceptive system sensitization that manifests as heightened pain sensitivity (Mercadante et al., 2019; Pahng and Edwards, 2021; Park et al., 2015). Furthermore, there is preclinical evidence that underlying pain conditions may dose-dependently enhance the rewarding effects of opioids and potentially increase risk of opioid misuse (Cahill et al., 2013; Colpaert et al., 2001; Hipolito et al., 2015; Hou et al., 2015; Lim et al., 2014; Zhang et al., 2014). In contrast, there is also preclinical evidence demonstrating that pain conditions may dose-dependently suppress the rewarding effects of opioids or only promote reinforcement at higher opioid doses (Ewan and Martin, 2011, 2013; Martin et al., 2007; Narita et al., 2005; Nazarian et al., 2021; Ozaki et al., 2002; Suzuki et al., 1996). Additional preclinical studies have found a lack of effect of pain manipulations on opioid preference, use, and seeking (Reiner et al., 2021; Shippenberg et al., 1988).

During preclinical investigations of pain and opioid interactions, there is a complex interplay of experimental factors to consider including type of opioid, opioid potency, type of pain condition modeled, opioid access conditions, and sex of the subject. Due to the destructive contributions of synthetic opioids in the current opioid epidemic, we chose to investigate the highly potent synthetic opioid, fentanyl, for this study. Chronic inflammatory pain caused by persistent inflammation and chronic neuropathic pain caused by nerve injury are two broad classifications of chronic pain. While both types of chronic pain conditions are devastating, inflammatory pain conditions including arthritis make up the majority of chronic pain complaints and are the leading cause of disability in adults (Hootman et al., 2016). In the present study, we used complete Freud’s adjuvant (CFA) to induce chronic inflammatory pain in male and female rats. Relevant to the current study, investigations of pain and opioid interactions have been almost exclusively completed in males and under short access conditions. Inclusion of females in these preclinical investigations is particularly important, because women report higher incidences of acute and chronic pain and demonstrate higher use of prescription opioids compared to men (Bartley and Fillingim, 2013; Cicero et al., 2009; Goetz et al., 2021; Hales et al., 2020; Schieber et al., 2020; Zelaya et al., 2020). Additionally, there is limited research available on sex differences with fentanyl use and misuse (Goetz et al., 2021).

In the present study, we examined the effects of chronic inflammatory pain and fentanyl access (limited vs. extended) on fentanyl intake, motivation to acquire fentanyl, and drug seeking in the absence of fentanyl in rats using intravenous opioid self-administration. We also examined the effects of chronic inflammatory pain and fentanyl access on Fos activity in ventral tegmental area (VTA) neurons expressing tyrosine hydroxylase (TH) following acute fentanyl challenge using immunohistochemistry. Finally, we investigated the effects of chronic inflammatory pain on low dose fentanyl preference using conditioned place preference. All experiments were completed in both male and female rats. We specifically tested the hypothesis that chronic inflammatory pain would increase fentanyl intake, motivation to acquire fentanyl, and drug seeking for fentanyl as well as promote conditioned place preference for fentanyl. Additionally, we expected blunted Fos activation in VTA dopamine neurons of animals experiencing chronic inflammatory pain. We also tested the hypothesis that animals given extended access to fentanyl would demonstrate escalated fentanyl intake, increased motivation to acquire fentanyl, and increased drug seeking for fentanyl compared to animals given limited access to fentanyl.

2. Methods and materials

2.1. Animals

One hundred and nine adult male and female Long-Evans rats were purchased from Charles River (Hollister, CA, USA) and weighed approximately 275 and 225 g, respectively when received (approximate age 9–12 weeks). The animal numbers for each experiment are as follows: intravenous (IV) self-administration of fentanyl and immunohistochemistry (N = 77, n = 48 males, n = 29 females), low dose fentanyl conditioned place preference (N = 32, n = 16 males, n = 16 females). The complete timelines for these experiments are shown in Fig. 1. Male and female rats were pair-housed separately, had ab libitum access to food (5053 Irradiated PicoLab Diet 20, LabDiet, St. Louis, MO, USA) and water, and were kept on a reverse 12-h light/dark cycle (lights off at 8:00 a.m. and lights on at 8:00pm). The animals were habituated in the colony room for one week prior to the start of experimental procedures and were handled daily during this time. All animal care, use, and procedures in this study were approved by the Institutional Animal Care and Use Committee of Southeast Louisiana Veterans Health Care System (SLVHCS) and were in accordance with the National Institute of Health and the Department of Veterans Affairs guidelines.

Fig. 1. Experiment timelines.

(A) Male and female rats were allowed to self-administer fentanyl (2.5 μg/kg/infusion) over the course of 10 operant sessions (5 sessions per week). Animals were split into equivalent conditions of either ShA-saline, ShA-CFA, LgA-saline, or LgA-CFA based on the total active lever responding. CFA animals received intra-plantar injections of CFA to induce persistent inflammation and mechanical hypersensitivity, while saline controls received intra-plantar injections of saline. Over 14 sessions (5 sessions per week), ShA animals (CFA & saline) were allowed to self-administer fentanyl (2.5 μg/kg/infusion) 1 h per day, while LgA animals (CFA & saline) were allowed to self-administer fentanyl (2.5 μg/kg/infusion) 12 h per day. Animals underwent 6-h progressive ratio (PR) testing at 2.5 μg/kg/infusion (ShA on day 1, LgA on day 2). For DR testing, animals received daily doses of fentanyl that were adjusted each day in descending order (5, 2.5, 1.25 μg/kg/infusion) for 1 h per day. Animals were given two additional days of testing where the fentanyl syringes were replaced with sterile saline syringes. Rats received an IV fentanyl challenge (35 μg/kg) and were perfused 90 min later. (B) Male and female rats were given one week for housing room acclimation, handling, and baseline testing of mechanical hypersensitivity (von Frey). Animals were given either CFA or saline intra-plantar injections and were tested for mechanical hypersensitivity 1 week later. Next, animals underwent low-dose (4 μg/kg) fentanyl CPP, which consists of a day 1 pre-test, 8 conditioning sessions (4 fentanyl pairings, 4 saline pairings) over 4 days, and a day 6 post-test.

2.2. Drugs

Flunixin (analgesic) and cefazolin (antibiotic) were purchased from MWI Animal Health (Boise, ID, USA). A stock solution of fentanyl (50 μg/ml) was purchased from the SLVHCS VA pharmacy and was diluted with sterile saline for IV self-administration (1.25, 2.5, 5 μg/kg/infusion). These doses were previously shown to establish fentanyl self-administration in male rats (Wade et al., 2015). A stock solution of ketamine (100 mg/ml) was purchased from the SLVHCS VA pharmacy and was diluted with sterile saline to assess catheter patency (10 mg/kg, administered intravenously at 1 ml/kg). For conditioned place preference, the fentanyl stock (50 μg/ml) was diluted with sterile saline to 4 μg/kg and administered subcutaneously (s.c) at 2 ml/kg.

2.3. Intravenous catheterization

Catheters (P1 Technologies, Roanoke, VA, USA) with silastic tubing (Fisher Scientific, USA) were made and sterilized for use in intrajugular surgeries to allow for IV self-administration of fentanyl in rats. Rats were anesthetized with 5% isoflurane for 5 min and anesthesia was maintained between 1 and 3% isoflurane for the remainder of the surgical procedure. Rats received a 2.5 mg/kg dose of flunixin s.c for analgesia. The shaved surgical sites were prepped with an alcohol prep pad followed by betadine. Incisions were made on the back and over right jugular vein for positioning of the catheter base and insertion of the catheter tubing. After the catheter was surgically implanted, it was secured with surgical silk and proper placement was checked by pulling back with a sterile saline-filled syringe to observe blood. The incisions on the neck and back were closed with suture silk and wound clips, respectively. The animals were removed from anesthesia and the catheters were flushed with 0.2 mL of heparinized water (33 unit/ml) with cefazolin (1 g/10 ml), which was given 5 days per week for the remainder of the experiment. Rats received additional flunixin injections (2.5 mg/kg s.c.) 12 h and 24 h post-surgery. For the catheter patency test, the catheters were flushed with 0.2 mL of heparinized water with cefazolin, which was followed by an infusion of ketamine (10 mg/kg, administered intravenously at 1 ml/kg). If the rat lost muscle tone within 5 s, the catheter was considered patent. If any rat failed the catheter patency test, they were removed from the experiment.

2.4. Induction of chronic inflammatory pain

Induction of chronic inflammatory pain in rats was conducted as previously described (Edwards et al., 2022). CFA was purchased from Sigma-Aldrich (St. Louis, MO, USA) and was diluted to 50% in sterile saline each morning of use. Rats were anesthetized with isoflurane and received a subcutaneous injection in the left hindpaw of 150 μL of 50% CFA or sterile saline. Once animals had righted themselves, they were returned to their home cage for recovery. Fentanyl self-administration was re-introduced 72 h after animals received CFA and saline injections.

2.5. Testing mechanical hypersensitivity

One week following CFA or saline injections, mechanical hypersensitivity was measured with von Frey testing as previously described (Edwards et al., 2022; Pahng et al., 2017). Rats acclimated to the testing room for 30 min prior to testing, then were placed in chambers elevated on wire mesh flooring and acclimated for 5 min. Once all four paws were on the ground, an electronic von Frey anesthesiometer (Ugo Basile; Germonio, Italy) was used to apply pressure to the middle region of the hindpaw. When the animal withdrew the hindpaw, the peak force (grams) was recorded as the paw withdrawal threshold. This was done for both the left (injected) and right (non-injected) hindpaws.

2.6. Fentanyl self-administration

2.6.1. Operant chambers

Standard rat operant chambers (Med Associates, St Albans, VT, USA) were used for IV fentanyl self-administration. Each operant chamber had plexiglass front and back walls, metal side walls, a metal grid floor, and a removable tray below the grid floor filled with bedding. The chambers contained two retractable levers with a cue light located above each lever. Fentanyl was delivered through Tygon tubing (Cole-Parmer, Vernon Hills, IL, USA) that connected to the rat’s catheter and a 30 ml BD luer lock syringe (Mckesson, USA) attached to an infusion pump (Med Associates) outside of the operant chamber. The tubing was protected by a modified leash (Med Associates) and held in place by a drug delivery arm (Med Associates) and swivel (Med Associates), which allowed the animals to move freely in the chambers. For all operant sessions, we provided animals with hard plastic chew toys (Bio-Serv Nylon Bones, Fisher Scientific) to minimize destructive chewing or self-injury.

2.6.2. Self-administration procedures

The complete timeline for this experiment is shown in Fig. 1A. All animals underwent fentanyl self-administration training with 1-h daily sessions 5 days per week for a total of 10 sessions with two days off per week (weekends off). At the start of each operant session both levers (active and inactive) extended into the chamber. These sessions occurred on a fixed-ratio 1 (FR1) schedule, where the active lever press delivered one infusion (0.1 mL for 3 s) of fentanyl at 2.5 μg/kg. The infusion was followed by a 5 s timeout in which the active lever press did not result in an infusion of fentanyl. The cue light above the active lever turned on when the active lever was pressed and remained on during the infusion and timeout period (8 s total). The maximum allowable number of reinforcers was 300 infusions. Pressing the inactive lever did not result in an infusion. During the training sessions, if a rat did not press the active lever within the first 5 min of the session, they received a priming dose. To administer the priming dose, the rats’ forepaws were placed on the active lever by the experimenter to mimic the rat pressing the lever on their own. If the rat did not press the active lever again after 20 min (25 min into the session), they received a second priming dose. By session 4, 14 animals out of 77 received 1 or 2 priming doses. By session 7, 6 animals out of 77 received 1 or 2 priming doses. By session 10, 4 animals out of 77 received 1 priming dose, while no animals received 2 priming doses. All initial priming dose infusions were recorded and subtracted out from the presented data (e.g., if a rat made 16 infusions but needed 2 priming doses, then the reported number of infusions is 14). Catheters were flushed at the end of each operant session to maintain catheter patency and to monitor for potential clogs. After the completion of 10 training sessions, both male and female groups were each split into four subgroups with approximately equal means and standard error for the average active lever presses of the last three sessions. Groups were also split to keep pre-existing cage mates housed under the same pain conditions (CFA or saline). The male and female subgroups were then randomly assigned to each condition (ShA saline, ShA CFA, LgA saline, LgA CFA). Specific group numbers are as follows: N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

2.6.3. Escalation of fentanyl intake

To measure escalation of fentanyl intake, rats continued fentanyl self-administration on a FR1 schedule at 2.5 μg/kg/infusion. CFA and saline rats of both sexes completed either 1-h (ShA) or 12-h (LgA) sessions 5 days per week for a total of 14 sessions (animals did not self-administer on weekends). Allowing ShA and LgA animals to have short periods of forced abstinence was done to mitigate the negative physical effects of LgA responding (e.g., weight loss) and to lower chances of mortality. Similar short periods of forced abstinence were given to LgA animals self-administering opioids in previous studies (Kimbrough et al., 2020; Wade et al., 2015), which does not seem to reliably impact subsequent drug consumption. These sessions followed the same procedure as the training sessions. During these self-administration sessions, rats had access to food and water in the operant chamber. Food pellets were placed on a small metal platform on the grid floor of the operant chambers. A small hole was drilled into the metal plates of each operant chamber to allow the spout of a water bottle (Ancare, Bellmore NY, USA) to be accessible to animals during testing.

2.6.4. Progressive ratio testing

Following the 14 sessions of ShA or LgA fentanyl self-administration, all rats completed a progressive ratio (PR) test to measure motivation to acquire fentanyl at a dose of 2.5 μg/kg/infusion. The PR session was 6 h and occurred over 2 days, with the ShA group running on the first day and the LgA group running on the second day. For the PR schedule, we used a PR schedule similar to the PR schedule used by Wade and Colleagues (Wade et al., 2015). The number of active lever presses required to receive one infusion increased as follows: 1, 1, 2, 2, 3, 3, 4, 4, 5, 5, 7, 7, 9, 9, 11, 11, 13, 13, 15, 15, 17, 17, 19, 19, 21, 21, …,73, 73. Each successive infusion required increased lever pressing. The point at which an animal stopped pressing and receiving a reward was considered that animal’s breakpoint.

2.6.5. Dose-response testing and drug seeking for fentanyl

After the PR test, rats completed a dose-response (DR) test. For this test, rats completed 1-h daily sessions on a FR1 schedule for 3 days. Rats self-administered one dose of fentanyl that was changed each day in a descending order (5, 2.5, and 1.25 μg/kg/infusion). Infusions times were the same between doses. The following 2 days, rats were tested for drug seeking behavior with 1-h daily sessions on a FR1 schedule. For this test, the fentanyl syringes were replaced with sterile saline so that rats received 0 μg/kg/infusion of fentanyl.

2.7. TH and Fos immunohistochemistry

After the completion of operant testing (3–4 days after drug seeking, but 5–6 days after last fentanyl exposure), animals received an acute fentanyl challenge of 35 μg/kg administered intravenously. The dose of fentanyl was chosen to be within the analgesic dose range for fentanyl (Khomula et al., 2019; Laboureyras et al., 2022). The analgesic and hyperalgesic profiles of fentanyl at a wide range of doses in male rats can be found in a recent publication (Laboureyras et al., 2022). This dose of fentanyl was also chosen to approximate the average amount of total fentanyl animals self-administered during a ShA session (equivalent to 14 infusions at once).

A single bolus delivery of fentanyl totaling 35 μg/kg and fentanyl infusions spread out over 1 h totaling 35 μg/kg are not directly comparable. However, to measure time-locked Fos activity in response to fentanyl challenge, we needed to give fentanyl in a single bolus infusion instead of spreading infusions out over an hour. Ninety minutes later, rats were anesthetized with isoflurane and transcardially perfused with cold 1X phosphate buffered saline (PBS) followed by cold 4% paraformaldehyde (PFA) in PBS. This entire process was completed over 2 days with half of the animals receiving infusions 3 days after drug seeking (5 days after last fentanyl exposure) and the other half receiving infusions 4 days after drug seeking (6 days after last fentanyl exposure). The brains were extracted and stored in 4% PFA for 24 h at 4 °C then switched to 20% sucrose in PBS for 48 h. Brains were snap-frozen in 2-methylbutane at −20 °C and stored at −80 °C. Brains were moved to −20 °C the day prior to brain sectioning. 40 μm-thick coronal sections containing the VTA between bregma −4.92 and −5.88 mm (Paxinos and Watson, 1998) were collected and stored at 4 °C in PBS with 0.1% sodium azide until immunostaining.

Three sections per animal were selected for immunohistochemistry to represent the anterior, mid, and posterior VTA. Sections were washed in PBS (3 × 10 min), incubated in 3% hydrogen peroxide in PBS for 5 min, and washed in PBS again (3 × 10 min). Next, sections were blocked for 1 h at room temperature (RT) in PBS containing 10% normal donkey serum (NDS) and 0.1% Triton X-100. Sections were then incubated overnight at RT in rabbit anti-cFos primary antibody (1:1000, abcam, ab190289) diluted in PBS with 0.01% Triton X-100 and 2% NDS. The next day, sections were washed in PBS (3 × 10 min) before incubation in donkey anti-rabbit AlexaFluor 488 secondary antibody (1:1000, ThermoFisher, A21206) diluted in PBS with 0.01% Triton X-100 and 1%NDS for 1 h at RT. The sections were washed in PBS (3 × 10 min) and blocked for 1 h at RT in PBS containing 10% NDS and 0.1% Triton X-100. Sections were then incubated overnight at 4 °C in sheep anti-TH primary antibody (1:1000, Millipore, AB1542) diluted in PBS with 0.01% Triton X-100 and 2% NDS. The following day, sections were washed in PBS (3 × 10 min) and then incubated in donkey anti-sheep AlexaFluor 594 secondary antibody (1:1000, ThermoFisher, #A-11016) diluted in PBS with 0.01% Triton X-100 and 1% NDS for 1 h at RT. Finally, sections were washed in PBS (3 × 10 min), mounted on slides, and coverslipped with Fluoro-Gel II with DAPI (Electron Microscopy Sciences, Hatfield, PA, USA).

Images were captured with a Keyence BZ-X810 all-in-one fluorescence microscope (Keyence, Itasca, IL, USA). Three VTA-containing sections at the same approximate bregma coordinates (anterior, mid, and posterior) were analyzed per animal. 2x images were collected and the boundaries of VTA were identified using TH expression and the Paxinos and Watson atlas as guides (Paxinos and Watson, 1998). Then three representative 20x images (544 μm μ 724 μm per image) of medial and lateral (right & left) VTA were taken to quantify Fos and TH expression for each anterior (bregma −4.92 mm), mid (bregma −5.40 mm), and posterior (bregma −5.88 mm) VTA-containing section. The number of Fos+ cells and Fos+/TH+ co-stained cells were counted manually and expressed as number of neurons. The number of Fos+ cells in TH− neurons was determined by subtracting Fos+/TH+ co-stained cells from total Fos+ cells. Specific group numbers are as follows: N = 76, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 6, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7. One animal’s IHC data was not included, because only the posterior VTA section could be quantified (anterior and mid VTA sections were damaged during the IHC process).

2.8. Conditioned place preference

The complete timeline for this experiment is shown in Fig. 1B. In a separate experiment, male and female rats were trained on a conditioned place preference (CPP) task with low dose fentanyl (4 μg/kg). Four custom-made Plexiglas CPP apparatuses (61 cm × 31 cm x 30.5 cm) were used for the CPP experiment. The CPP apparatuses were placed in individual cabinets containing an overhead Basler GenICam camera (Noldus, Leesburg, VA, USA) and were illuminated by 90 lux. In each CPP apparatus, the two chambers (26 cm × 31 cm x 30.5 cm) were separated by a small middle chamber (9 cm × 14 cm x 30.5 cm), which contained two doors to allow passage between chambers during the pretest and post-test. The tops of the CPP apparatuses were covered with a Plexiglas lid (61 cm × 31 cm x 0.5 cm). Each CPP chamber had removable textured flooring and walls of particle board fixed to the outside of the see-through Plexiglas. One chamber had black and white horizontal striped walls with beige clay tile flooring (Context A), while the other chamber had walls with multiple large black circles on a white background and white plastic flooring with raised rectangles (Context B). We counterbalanced which context was drug-paired for each CPP apparatus, so that half of the animals were drug-paired with Context A and the other half of animals were drug-paired with Context B.

Baseline mechanical hypersensitivity was measured with von Frey testing. Both male and female groups were each split into two subgroups with approximately equal means and standard error for baseline von Frey thresholds and then were randomly assigned to either CFA or saline conditions with cage mates housed under the same pain conditions (CFA or saline). We have previously used this procedure for group splitting and condition assignment (Edwards et al., 2022). One week following CFA or saline injections, mechanical hypersensitivity was measured again. 48 h later, animals were given a CPP pre-test. On the pre-test day (day 1), the doors in the middle chamber were opened and animals were allowed to freely explore both contexts in a drug-free state for 20 min (total test duration = 1200 s). We tracked time spent in each chamber using EthoVision XT 16.0 (Noldus). On conditioning days (days 2–5), the doors in the middle chamber were closed and animals were placed in one context immediately following a s.c. injection of fentanyl (4 μg/kg) or saline. We used twice-a-day conditioning similar to protocols used for CPP conditioning in rats with morphine (Zadina et al., 2016) and fentanyl (Vitale et al., 2003). In the morning, animals were given saline injections and immediately placed in the control-paired context for 20 min. In the afternoon, animals were given fentanyl injections and immediately placed in the drug-paired context for 20 min. In previous studies, typical conditioning times for fentanyl CPP ranged between 20 and 40 min in rats (Gaulden et al., 2021; Sustkova-Fiserova et al., 2020; Vitale et al., 2003) and mice (Bryant et al., 2021; Du et al., 2021). The shorter 20 min conditioning time for fentanyl CPP was chosen based on previous studies that tested fentanyl CPP using low dose fentanyl (4 μg/kg) (Gaulden et al., 2021; Vitale et al., 2003). Animals received 4 pairings of each treatment (4 h between treatments) and context for a total of 8 conditioning sessions over 4 days. For the post-test day (day 6), the doors in the middle chamber were opened and animals were allowed to freely explore both contexts in a drug-free state for 20 min.

Data are presented as preference duration (sec) and change in preference duration (sec). Change in preference duration (sec) for the drug-paired chamber was calculated by (time in the drug-paired chamber post-test) – (time in the drug-paired chamber pre-test). Change in preference duration (sec) for the saline-paired chamber was calculated by (time in the saline-paired chamber post-test) – (time in the saline-paired chamber pre-test). Change in preference duration (sec) for the center chamber was calculated by (time in the center chamber post-test) – (time in the center chamber pre-test). These measures account for inherent (pre-test) bias during CPP. Specific group numbers are as follows: N = 32, male saline n = 8, male CFA n = 8, female saline n = 8, and female CFA n = 8.

2.9. Statistical analysis

All data were analyzed using Prism 9 (GraphPad Software, Inc; La Jolla, CA, USA). Fentanyl self-administration acquisition data were analyzed using two-way ANOVAs with sex (male vs. female) as the between-subjects factor and session (operant sessions 1–10) as the within-subjects factor. Male and female animals were then split into 4 groups (8 groups total) where animals were assigned to a fentanyl access condition (ShA vs. LgA) and a CFA treatment condition (CFA vs. saline): ShA-saline, ShA-CFA, LgA-saline, and LgA-CFA. Baseline differences in fentanyl self-administration prior to group assignments were analyzed using a three-way between-subjects ANOVA with sex (male vs. female), fentanyl access assignment (pre-ShA vs. pre-LgA), and CFA treatment assignment (pre-CFA vs. pre-saline) as factors. For the self-administration experiment, changes in mechanical hypersensitivity following CFA treatment were analyzed using two-way between-subjects ANOVAs with CFA treatment (CFA vs. saline) and sex (male vs. female) as factors. After splitting animals into different groups, fentanyl self-administration data were analyzed using two-way and three-way ANOVAs with fentanyl access (ShA vs. LgA) and CFA treatment (CFA vs. saline) as the between-subjects factors and session (operant sessions 1–14) as the within-subjects factor. Changes in fentanyl intake were analyzed using a two-way between-subjects ANOVA with CFA treatment (CFA vs. saline) and sex (male vs. female) as factors.

Pearson’s r correlations and linear regressions were used to analyze the relationship between von Frey thresholds and fentanyl intake during operant session 2 (session prior to von Frey testing). PR data were analyzed using three-way between-subjects ANOVAs with sex (male vs. female), fentanyl access (ShA vs. LgA), and CFA treatment (CFA vs. saline) as factors. DR data were analyzed using two- and three-way ANOVAs with fentanyl access (ShA vs. LgA), and CFA treatment (CFA vs. saline) as the between-subjects factors and fentanyl dose (1.25, 2.5, 5 μg/kg/infusion) as the within-subjects factor. Drug seeking after fentanyl dose-response data (infusions) were analyzed using three-way ANOVAs with fentanyl access (ShA vs. LgA), and CFA treatment (CFA vs. saline) as the between-subjects factors and fentanyl dose (0–5 μg/kg/infusion) as the within-subjects factor. Drug seeking data (active lever presses) were analyzed using three-way ANOVAs with fentanyl access (ShA vs. LgA), and CFA treatment (CFA vs. saline) as the between-subjects factors and fentanyl seeking tests (sessions 1–2) as the within-subjects factor.

Fos+, Fos+/TH+, FOS+/TH− cell counts in the VTA were analyzed using three-way between-subjects ANOVAs with sex (male vs. female), fentanyl access (ShA vs. LgA), and CFA treatment (CFA vs. saline) as factors. For the CPP experiment, changes in mechanical hypersensitivity following CFA treatment were analyzed using two-way ANOVAs with CFA treatment (CFA vs. saline) as the between-subjects factor and time (baseline vs. 1 week post) as the within-subjects factor. CPP preference duration was analyzed using two-way ANOVAs with CFA treatment (CFA vs. saline) as the between-subjects factor and conditioning (pretest vs. post-test) as the within-subjects factor. Change in preference duration was analyzed using two-way between-subjects ANOVAs with sex (male vs. female) and CFA treatment (CFA vs. saline) as factors. Post hoc analysis of CPP data were analyzed using Sidak’s and Tukey’s multiple comparisons tests. Significance levels for statistical tests were set at p < 0.05.

3. Results

3.1. Fentanyl self-administration

3.1.1. Male rats given extended access to fentanyl demonstrate a greater increase in fentanyl intake over time compared to females

We trained male and female rats to self-administer fentanyl (2.5 μg/kg/infusion) over the course of 10 operant sessions (5 days per week) at a fixed-ratio of 1 (FR-1). For fentanyl intake, we found significant effects of session (operant sessions 1–10) [F(2.291,171.5) = 20.10, p < 0.0001] and sex (male vs. female) [F(1,75) = 8.468, p = 0.0048], but no interaction [F(9,674) = 1.521, p = 0.1365] Fig. 2A. For active lever presses, we found significant effects of session [F(2.669,199.6) = 11.59, p < 0.0001] and sex [F(1,75) = 4.603, p = 0.0351], but no interaction [F(9,673) = 1.242, p = 0.2662] Fig. 2B. Both fentanyl intake and active lever pressing increased over time in male and female rats, but there was increased intake and active lever pressing in males compared to females Fig. 2A & B. For inactive lever presses, there was an effect of sex [F(1,75) = 23.61, p < 0.0001], but no effect of session [F(4.289,321.2) = 1.443, p = 0.2162] and no interaction [F(9,674) = 1.231, p = 0.2725]. Both male and female rats demonstrated a preference for the active lever over the inactive lever [F(1,150) = 280.6, p < 0.0001], while males demonstrated increased inactive lever pressing compared to females Fig. 2B.

Fig. 2. Male rats exhibit increased operant responding compared to females during acquisition of fentanyl self-administration.

(A) During acquisition of fentanyl self-administration, fentanyl intake increased over time in male and female rats (####p < 0.0001). Male rats demonstrated increased fentanyl intake compared to female rats (**p < 0.01). (B) Active lever pressing increased over time in male and female rats (####p < 0.0001), but inactive lever pressing did not change over time (p > 0.05). Male rats demonstrated increased active lever pressing (*p < 0.05) and inactive lever pressing (****p < 0.0001) compared to female rats. (C) There were no baseline differences in operant responding between group assignments (ShA-saline, ShA-CFA, LgA-saline, or LgA-CFA) (p > 0.05). (D) In the injected left paw, CFA animals demonstrated a decrease in paw withdrawal thresholds compared to saline controls (****p < 0.0001), indicating mechanical hypersensitivity. There was no effect of sex (p > 0.05). (E) In the non-injected right paw, there was no effect of CFA treatment or sex (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

We averaged the last 3 sessions of operant responding for each animal and split male and female animals into equivalent conditions of either ShA-saline, ShA-CFA, LgA-saline, or LgA-CFA Fig. 2C. Our analysis revealed that there were no baseline differences in operant responding between group assignments: no effect of sex [F(1,69) = 0.9342, p = 0.3371], no effect of fentanyl access assignment (pre-ShA vs. pre-LgA) [F(1,69) = 0.0671, p = 0.7964], no effect of CFA treatment assignment (pre-CFA vs. pre-saline) [F(1,69) = 0.0048, p = 0.9448], and no interactions (p > 0.05) Fig. 2C. To model chronic inflammatory pain in rats, we used left intra-plantar CFA injections to induce persistent inflammation and mechanical hypersensitivity (measure of pain-like behavior) as previously described (Edwards et al., 2022). One week later, we measured paw withdrawal thresholds to examine the effects of CFA treatment on mechanical hypersensitivity in the injected and non-injected paws Fig. 2D & E. In the injected left paw, there was a significant effect of CFA treatment [F(1,73) = 55.81, p < 0.0001], but no effect of sex [F(1,73) = 0.3491, p = 0.5564] and no interaction [F(1,73) = 0.5524, p = 0.4597] Fig. 2D. In the non-injected right paw, there were no effects of CFA treatment [F(1,73) = 0.0155, p = 0.9013], sex [F(1,73) = 2.851, p = 0.0956], and no interaction [F(1,73) = 1.736, p = 0.1918] Fig. 2E. This demonstrates that CFA produces significant mechanical hypersensitivity in the injected paw but did not produce mechanical hypersensitivity in the non-injected paw.

Fentanyl self-administration was re-introduced 72 h after animals received CFA and saline injections. We measured escalation of fentanyl intake after induction of inflammatory pain in ShA and LgA animals of both sexes during FR-1 responding. In male rats, there were significant effects of session (operant sessions 1–14) [F(3.873,170.4) = 21.95, p < 0.0001], fentanyl access (ShA vs. LgA) [F(1,44) = 77.45, p < 0.0001], and a significant interaction [session x (ShA vs. LgA)] [F(13,572) = 21.92, p < 0.0001], but no effect of CFA treatment [F(1,44) = 0.0001, p = 0.9917] Fig. 3A. In female rats, there were significant effects of session [F(4.321,107.7) = 8.260, p < 0.0001], fentanyl access [F(1,25) = 38.38, p < 0.0001], and a significant interaction [session x (ShA vs. LgA)] [F(13,324) = 6.969, p < 0.0001], but no effect of CFA treatment [F(1,25) = 0.0213, p = 0.8851] Fig. 3B. In both sexes, there was significant escalation of fentanyl intake in LgA animals (males: [F(3.842,103.7) = 34.26, p < 0.0001], females: [F(4.123,53.28) = 8.227, p < 0.0001]) over time, but no escalation of fentanyl intake in ShA animals (males: [F(3.086,52.46) = 1.265, p = 0.2961], females: [F(3.584,43.01) = 1.319, p = 0.2796]). Escalated fentanyl intake was unaffected by inflammatory pain status.

Fig. 3. Male rats given extended access to fentanyl demonstrate a greater increase in fentanyl intake over time compared to females.

(A) In male rats, there was an effect of fentanyl access (ShA vs. LgA) (****p < 0.0001), but no effect of CFA treatment (p > 0.05). There was significant escalation of fentanyl intake in LgA males (****p < 0.0001), but not in ShA males (p > 0.05). (B) In female rats, there was an effect of fentanyl access (****p < 0.0001), but no effect of CFA treatment (p > 0.05). There was significant escalation of fentanyl intake in LgA females (****p < 0.0001), but not in ShA females (p > 0.05). (C) LgA males demonstrated a greater increase in fentanyl intake between operant sessions 1 and 14 compared to LgA females (*p < 0.05). There was no effect of CFA treatment on the change in fentanyl intake between operant sessions 1 and 14 (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

To examine sex differences in escalated fentanyl intake of LgA animals, we compared the change in fentanyl intake between the first and final operant sessions. This analysis revealed a significant effect of sex [F(1,40) = 5.609, p = 0.0228] where LgA males demonstrated a greater increase in fentanyl intake compared to LgA females Fig. 3C. There was no effect of CFA treatment [F(1,40) = 0.1854, p = 0.6691] and no interaction [F(1,40) = 0.6861, p = 0.4124] Fig. 3C. To examine the relationship between prior fentanyl intake and subsequent mechanical hypersensitivity in ShA and LgA animals, we correlated fentanyl intake in FR-1 session 2 with paw withdrawal thresholds (1 week-post CFA) during opioid abstinence. FR-1 session 2 was the last session animals received prior to von Frey testing. ShA animals, there was no correlation between fentanyl intake and paw withdrawal thresholds in the right paw (r = −0.0342, p = 0.8501) or left paw (r = 0.0093, p = 0.9589) Supplementary Figs. 1A and 1B. In LgA animals, there was a significant negative correlation between fentanyl intake and paw withdrawal thresholds in the right paw (r = −0.3283, p = 0.0296), demonstrating that animals with the highest fentanyl intake had the lowest paw withdrawal thresholds (increased mechanical hypersensitivity) in the non-injected paw Supplementary Fig. 1C. In the injected paw of LgA animals, there was no correlation between fentanyl intake and paw withdrawal thresholds (r = −0.2442, p = 0.1102) Supplementary Fig. 1D.

We also examined active and inactive lever pressing in ShA and LgA animals of both sexes. For active lever presses in LgA male rats, we found a significant effect of session [F(2.389,64.51) = 20.67, p < 0.0001] and an interaction [F(13,351) = 2.069, p = 0.0154], but no effect of CFA treatment [F(1,27) = 0.2235, p = 0.6402] Fig. 4A. Active lever responding increased over time in LgA male rats. For inactive lever presses in LgA male rats, there were no effects of session [F(1.582,42.71) = 0.9515, p = 0.3752], CFA treatment [F(1,27) = 1.194, p = 0.2841], and no interaction [F(13,351) = 1.133, p = 0.3295] Fig. 4A. Male LgA animals demonstrated a preference for the active lever over the inactive lever [F(1,54) = 37.92, p < 0.0001]. For active lever presses in LgA female rats, we found a significant effect of session [F(2.980,38.51) = 4.572, p = 0.0079], but no effect of CFA treatment [F(1,13) = 0.3116, p = 0.5862] and no interaction [F(13,168) = 1.073, p = 0.3853] Fig. 4B. For inactive lever presses in LgA female rats, there was significant interaction [F(13,168) = 1.817, p = 0.0440], but no effect of session [F(4.578,59.16) = 1.082, p = 0.3774] or CFA treatment [F(1,13) = 0.2827, p = 0.6039] Fig. 4B. Similar to LgA males, LgA females increased active lever responding over time and demonstrated a preference for the active lever over the inactive lever [F(1,26) = 11.74, p = 0.0020]. We found that inflammatory pain status did not change active or inactive lever pressing in male and female LgA rats.

Fig. 4. Male and female rats given extended access to fentanyl escalate active lever responding for fentanyl over time.

(A) LgA male rats demonstrated a preference for the active lever over the inactive lever (p < 0.0001). Active lever responding increased over time in LgA male rats (####p < 0.0001), while inactive lever responding did not change over time (p > 0.05). There was no effect of CFA treatment on active or inactive lever responding in LgA male rats (p > 0.05). (B) LgA female rats demonstrated a preference for the active lever over the inactive lever (p < 0.01). In LgA female rats, active lever responding increased over time (##p < 0.01), while inactive lever responding did not change over time (p > 0.05). There was no effect of CFA treatment on active or inactive lever responding in LgA female rats (p > 0.05). (C) ShA male rats demonstrated a preference for the active lever over the inactive lever (p < 0.0001). In ShA males, there was no change in active or inactive lever pressing over time (p > 0.05). There was no effect of CFA treatment on active or inactive lever responding in ShA male rats (p > 0.05). (D) ShA female rats demonstrated a preference for the active lever over the inactive lever (p < 0.0001). In ShA females, there was no change in active or inactive lever pressing over time (p > 0.05). There was no effect of CFA treatment on active or inactive lever responding in ShA female rats (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

For active lever presses in ShA male rats, we found no effects of session [F(3.811,64.79) = 1.080, p = 0.3724], CFA treatment [F(1,17) = 1.352, p = 0.2610], and no interaction [F(13,221) = 0.6254, p = 0.8316] Fig. 4C. For inactive lever presses in ShA male rats, we found no effects of session [F(5.128,87.17) = 1.718, p = 0.1373], CFA treatment [F(1,17) = 0.2354, p = 0.6337], and no interaction [F(13,221) = 0.9841, p = 0.4678] Fig. 4C. In contrast to LgA males, there was no increase in active lever pressing over time in ShA males. However, ShA males demonstrated a preference for the active lever over the inactive lever [F(1,34) = 66.23, p < 0.0001]. For active lever presses in ShA female rats, we found no effects of session F(3.613,43.35) = 1.488, p = 0.2260], CFA treatment [F(1,12) = 1.332, p = 0.2709], and no interaction [F(13,156) = 0.7806, p = 0.6796] Fig. 4D. For inactive lever presses in ShA female rats, we found no effects of session [F(4.020,48.24) = 1.211, p = 0.3184], CFA treatment [F(1,12) = 0.0697, p = 0.7962], and no interaction [F(13,156) = 1.039, p = 0.4169] Fig. 4D. Similar to ShA males, ShA females did not increase active lever responding over time, but demonstrated a preference for the active lever over the inactive lever [F(1,24) = 117.9, p < 0.0001]. We found that inflammatory pain status did not change active or inactive lever pressing in male and female ShA rats. In summary, we found no effects of inflammatory pain on fentanyl intake in non-escalating ShA rats or escalating LgA rats of either sex. However, we found that males self-administer more fentanyl than females during acquisition and that LgA males demonstrate a greater increase in fentanyl intake over time compared to LgA females.

3.1.2. Female rats given limited access to fentanyl show increased responding for fentanyl across doses compared to males

To examine differences in the willingness to work for fentanyl (i.e. motivation to acquire fentanyl), all groups underwent 6 h PR testing, which involves a multiplicative increase in the number of operant responses needed to receive a single infusion of fentanyl (2.5 μg/kg/infusion). As the PR steps increase, the point at which an animal stops responding is considered that animal’s breakpoint. For breakpoint, there was a significant effect of fentanyl access [F(1,69) = 7.966, p = 0.0062], but no effects of sex [F(1,69) = 0.1841, p = 0.6692], CFA treatment [F(1,69) = 0.1992, p = 0.6567], and no interactions (p > 0.05) Fig. 5A. For active lever presses, there was a significant effect of fentanyl access [F(1,69) = 6.954, p = 0.0103], but no effects of sex [F(1,69) = 0.3077, p = 0.5809], CFA treatment [F(1,69) = 1.042, p = 0.3109] and no interactions (p > 0.05) Fig. 5B. In both male and female rats, LgA animals had higher breakpoints and more active lever presses than ShA animals, demonstrating increased motivation to acquire fentanyl in LgA animals. There was a preference for the active lever over the inactive lever in both ShA [F(1,58) = 53.70, p < 0.0001] and LgA [F(1,80) = 34.12, p < 0.0001] animals Fig. 5B & C. There was no effect of inflammatory pain status on motivation to acquire fentanyl in either males or females. For inactive lever presses, there was a significant effect of sex [F(1,69) = 4.826, p = 0.0314], but no effects of fentanyl access [F(1,69) = 0.4400, p = 0.5093], CFA treatment [F(1,69) = 0.0149, p = 0.9031], and no interactions (p > 0.05) Fig. 5C. Male rats pressed the inactive lever more than female rats during PR testing.

Fig. 5. Male and female rats given extended access to fentanyl exhibit increased motivation to acquire fentanyl (PR testing) compared to animals given limited access to fentanyl.

(A) LgA animals had higher breakpoints than ShA animals (**p < 0.01). There was no effect of CFA treatment or sex on breakpoint (p > 0.05). (B) LgA animals had more active lever presses than ShA animals (*p < 0.05). There was no effect of CFA treatment or sex on active lever presses (p > 0.05). (C) Male rats had more inactive lever presses than female rats (p < 0.05), but there was no effect of CFA treatment or fentanyl access (ShA vs. LgA) on inactive lever presses (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

We next examined motivation to acquire fentanyl in all groups by completing a DR experiment. For DR testing, animals received daily doses of fentanyl that were adjusted each day in descending order (5, 2.5, 1.25 μg/kg/infusion) for 1 h per day. We measured changes in number of fentanyl infusions and intake during DR testing. The data are graphically represented in ascending dose order (1.25, 2.5, 5 μg/kg/infusion). For number of fentanyl infusions in male rats, there were significant effects of fentanyl dose [F(1.370,60.26) = 122.4, p < 0.0001], fentanyl access [F(1,44) = 19.54, p < 0.0001], and a dose by fentanyl access interaction [F(2,88) = 15.08, p < 0.0001], but no effect CFA treatment [F(1,44) = 0.0345, p = 0.8536] Fig. 6A. For fentanyl intake in male rats, there were significant effects of fentanyl dose [F(1.526,67.14) = 40.34, p < 0.0001] and fentanyl access [F(1,44) = 17.76, p = 0.0001], but no effect of CFA treatment [F(1,44) = 0.1028, p = 0.7500] and no interactions (p > 0.05) Fig. 6B. There is a significant DR effect for male animals where there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses, but an increase in total fentanyl intake at higher fentanyl doses. Male LgA animals demonstrated an upward shift in operant responding during DR testing for both number of fentanyl infusions and total fentanyl intake compared to male ShA animals. Inflammatory pain status did not change the number of fentanyl infusions or total fentanyl intake in male rats during DR testing.

Fig. 6. Female rats given limited access to fentanyl exhibit increased motivation to acquire fentanyl (DR testing) compared to males.

(A) For male rats, there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses in ShA (****p < 0.0001) and LgA (****p < 0.0001) animals. LgA male animals demonstrated an upward shift in operant responding for number of fentanyl infusions compared to ShA male animals (****p < 0.0001). There was no effect of CFA treatment on number of fentanyl infusions in male animals (p > 0.05). (B) There was an increase in total fentanyl intake at higher fentanyl doses in ShA (***p < 0.001) and LgA (****p < 0.0001) male animals. LgA male animals demonstrated an upward shift in operant responding for fentanyl intake compared to ShA male animals (***p < 0.001). There was no effect of CFA treatment on fentanyl intake in male animals (p > 0.05). (C) For female rats, there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses in ShA (***p < 0.001) and LgA < 0.0001) animals. There was no effect of CFA treatment or fentanyl access (ShA vs. LgA) on number of fentanyl infusions in female animals (p > 0.05). (D) There was an overall increase in total fentanyl intake at higher fentanyl doses in female animals (p < 0.05). Looking at ShA and LgA separately, there was an increase in total fentanyl intake at higher fentanyl doses in LgA (**p < 0.01), but not ShA (p > 0.05) female animals. There was no effect of CFA treatment or fentanyl access on fentanyl intake in female animals (p > 0.05). (E) For LgA rats, there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses in females (****p < 0.0001) and males (****p < 0.0001). There was no difference between LgA females and LgA males for number of fentanyl infusions (p > 0.05). There was no effect of CFA treatment on number of fentanyl infusions (p > 0.05). (F) For ShA rats, there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses in females (***p < 0.001) and males (****p < 0.0001). There was increased fentanyl responding in ShA females compared to ShA males (*p < 0.05). There was no effect of CFA treatment on number of fentanyl infusions (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

For number of fentanyl infusions in female rats, there was a significant effect of fentanyl dose [F(1.313,31.51) = 85.02, p < 0.0001] and a dose by fentanyl access interaction [F(2,48) = 4.445, p = 0.0169], but no effects of fentanyl access [F(1,24) = 2.370, p = 0.1367] or CFA treatment [F(1,24) = 1.631, p = 0.2138] Fig. 6C. For fentanyl intake in female rats, there was a significant effect of fentanyl dose [F(1.077,25.86) = 6.910, p = 0.0127], but no effects of fentanyl access [F(1,24)= 0.7645, p = 0.3906], CFA treatment [F(1,24) = 1.778, p = 0.1949], and no interactions (p > 0.05) Fig. 6D. Similar to male rats, there is a significant DR effect for female animals where there was a decrease in the number of fentanyl infusions taken at higher fentanyl doses, but an increase in total fentanyl intake at higher fentanyl doses. In contrast to male animals, there was no difference between LgA and ShA animals for fentanyl responding during DR testing in female rats. This lack of a difference between LgA and ShA animals was most likely due to an upward shift of female ShA animals. When comparing sexes, there was no difference between LgA males and females for number of fentanyl infusions [F(1,40) = 0.9747, p = 0.3295] Fig. 6E, but there was increased fentanyl responding in ShA females compared to ShA males [F(1,28) = 5.403, p = 0.0276] Fig. 6F. Similar to the males, inflammatory pain status did not change the number of fentanyl infusions or total fentanyl intake in female rats during DR testing.

To examine drug seeking for fentanyl after completion of DR testing, all groups were given two additional days of testing where the fentanyl syringes were replaced with sterile saline syringes. These data are graphically represented in the order of descending fentanyl dose (5, 2.5, 1.25 μg/kg/infusion) and saline presentation over the course of 5 days of testing with the seeking days (0, 0 μg/kg/infusion) occurring during the last 2 days of testing. For number of fentanyl infusions in male rats across the 5 days, there were significant effects of dose [F(1.454,63.99) = 117.6, p < 0.0001] and fentanyl access [F(1,44) = 31.51, p < 0.0001], but no effects of CFA treatment [F(1,44) = 1.362, p = 0.2495] and no interactions (p > 0.05) Fig. 7A. When examining active lever pressing during the two fentanyl seeking sessions in males, there were significant effects of test day [F(1,44) = 24.64, p < 0.0001] and fentanyl access [F(1,44) = 11.13, p = 0.0017], but no effect of CFA treatment [F(1,44) = 3.962, p = 0.0528] and no interactions (p > 0.05) Fig. 7B. For number of fentanyl infusions in female rats across the 5 days, there was a significant effect of dose [F(1.246,29.90) = 51.39, p < 0.0001], but no effects of fentanyl access [F(1,24) = 1.693, p = 0.2056], CFA treatment [F(1,24) = 0.0001, p = 0.9768], and no interactions (p > 0.05) Fig. 7C. When examining active lever pressing during the two fentanyl seeking sessions in females, there was a significant effect of test day [F(1,24) = 27.41, p < 0.0001], but no effect of fentanyl access [F(1,24) = 2.138, p = 0.1567], CFA treatment [F(1,24) = 0.01124, p = 0.9165], and no interactions (p > 0.05) Fig. 7D. Both males and females, demonstrated increased operant responding during the drug seeking when the fentanyl syringes were replaced with saline as well as decreased operant responding to saline during the second day of the fentanyl seeking test. However, there was a significant increase in operant responding in LgA males compared to ShA males, but no difference between LgA females and ShA females. In summary, we found no effects of inflammatory pain on motivation to acquire fentanyl or drug seeking for fentanyl in either sex. However, we found that ShA females show increased responding for fentanyl across doses compared to ShA males.

Fig. 7. Both male and female rats display significant increases in operant responding in the absence of fentanyl (drug seeking).

(A) LgA (****p < 0.0001) and ShA (****p < 0.0001) male rats demonstrated increased operant responding during the drug seeking test when the fentanyl syringes were replaced with saline. There was a significant increase in operant responding in LgA males compared to ShA males (****p < 0.0001). There was no effect of CFA treatment on drug seeking behavior in male animals (p > 0.05). (B) LgA and ShA male rats decreased operant responding to saline during the second day of fentanyl seeking test (p < 0.0001). There was a significant increase in operant responding in LgA males compared to ShA males (**p < 0.01). There was no effect of CFA treatment on drug seeking behavior in male animals (p > 0.05). (C) LgA (****p < 0.0001) and ShA (***p < 0.001) female rats demonstrated increased operant responding during the drug seeking test when the fentanyl syringes were replaced with saline. There was no effect of CFA treatment or fentanyl access (ShA vs. LgA) on drug seeking in female animals (p > 0.05). (D) LgA and ShA female rats decreased operant responding to saline during the second day of fentanyl seeking test (p < 0.0001). There was no effect of CFA treatment or fentanyl access (ShA vs. LgA) on drug seeking in female animals (p > 0.05). N = 77, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 7, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7.

3.2. Following fentanyl challenge, female rats exhibit higher numbers of Fos-positive non-dopaminergic VTA neurons compared to males

After the completion of fentanyl self-administration (3–4 days), rats were given an acute fentanyl challenge of 35 μg/kg administered intravenously, and then 90 min later brains were perfused, extracted, and stored for brain sectioning and immunostaining. Immunohistochemistry was performed on VTA-containing sections to stain for Fos, a marker of neuronal activity, and TH, a marker of dopaminergic neurons Fig. 8A–H. See Supplementary Figs. 2A–H for representative images of Fos and TH co-staining from all eight groups. For Fos immunoreactivity in VTA neurons, there was an effect of sex [F(1,68) = 11.36, p = 0.0012], but no effects of fentanyl access [F(1,68) = 0.0785, p = 0.7801], CFA treatment [F(1,68) = 0.0034, p = 0.9540], and no interactions (p > 0.05) Fig. 8I. This demonstrates that females had more Fos + neurons in the VTA than males following fentanyl challenge Fig. 8I. We next examined Fos immunoreactivity in VTA dopamine neurons by measuring TH and Fos co-staining in the VTA. For Fos immunoreactivity in TH+ VTA neurons, there were no effects of sex [F(1,68) = 1.493, p = 0.2260], fentanyl access [F(1,68) = 0.1352, p = 0.7143], CFA treatment [F(1,68) = 0.4788, p = 0.4913], and no interactions (p > 0.05) Fig. 8J. Due to the cellular heterogeneity of the VTA (Morales and Margolis, 2017), we followed up this analysis by examining Fos immunoreactivity in TH− (non-dopaminergic) VTA neurons where we once again found an effect of sex [F(1,68) = 12.04, p = 0.0009], but no effects of fentanyl access [F(1,68) = 0.0265, p = 0.8712], CFA treatment [F(1,68) = 0.1572, p = 0.6930], and no interactions (p > 0.05) Fig. 8K. This demonstrates that the increased VTA Fos immunoreactivity in females following fentanyl challenge is occurring in a population of non-dopaminergic neurons. In summary, we found no effects of inflammatory pain on Fos immunoreactivity in dopaminergic VTA neurons following fentanyl challenge. However, we found increased Fos immunoreactivity in non-dopaminergic VTA neurons of females compared to males following fentanyl challenge.

Fig. 8. Following fentanyl challenge, female rats exhibit higher numbers of Fos-positive non-dopaminergic VTA neurons compared to males.

(A) 2x image of TH (red) staining in the VTA of a male LgA saline rat. (B) 2x image of TH (red) staining in the VTA of a female LgA saline rat. (A) & (B) White boxes represent the medial and lateral VTA sampling areas (544 μm × 724 μm per box). (C) 20x image of Fos (green) and DAPI (blue) staining in the VTA of a male LgA saline rat. (D) 20x image of TH (red) and DAPI (blue) staining in the VTA of a male LgA saline rat. (E) 20x image of Fos (green), TH (red), and DAPI (blue) staining in the VTA of a male LgA saline rat. White arrows denote Fos and TH co-staining. (F) 20x image of Fos (green) and DAPI (blue) staining in the VTA of a female LgA saline rat. (G) 20x image of TH (red) and DAPI (blue) staining in the VTA of a female LgA saline rat. (H) 20x image of Fos (green), TH (red), and DAPI (blue) staining in the VTA of a female LgA saline rat. White arrows denote Fos and TH co-staining. (I) Following acute fentanyl challenge, female rats had more Fos immunoreactivity in VTA neurons than males (**p < 0.01). There were no effects of fentanyl access (p > 0.05) or CFA treatment (p > 0.05) on Fos immunoreactivity in VTA neurons. (J) Following acute fentanyl challenge, there were no effects of sex (p > 0.05), fentanyl access (p > 0.05), or CFA treatment (p > 0.05) on Fos immunoreactivity in dopaminergic (TH+) VTA neurons. (K) Following acute fentanyl challenge, female rats had more Fos immunoreactivity in non-dopaminergic (TH−) VTA neurons than males (***p < 0.001). There were no effects of fentanyl access (p > 0.05) or CFA treatment (p > 0.05) on Fos immunoreactivity in non-dopaminergic (TH−) VTA neurons. N = 76, ShA male saline n = 10, ShA male CFA n = 9, LgA male saline n = 14, LgA male CFA n = 15, ShA female saline n = 6, ShA female CFA n = 7, LgA female saline n = 8, and LgA female CFA n = 7. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)

3.3. Chronic inflammatory pain promotes fentanyl place preference in male rats, but not females

In a separate experiment, male and female rats were given either CFA or saline injections and then trained on a CPP task with low dose fentanyl (4 μg/kg) one week later. Both male and female animals given CFA paw injections demonstrated significant mechanical hypersensitivity at the start of CPP conditioning in the injected paw (CFA treatment in males: [F(1,14) = 13.38, p = 0.0026]; CFA treatment in females: [F(1.14) = 16.03, p = 0.0013]) Supplementary Figs. 3A and 3B, but not in the non-injected paw (CFA treatment in males: [F(1,14) = 0.1787, p = 0.6789]; CFA treatment in females: [F(1,14) = 0.0127, p = 0.9118]) Supplementary Figs. 3C and 3D. For female rats, we found no effect of conditioning (pre-test vs. post-test) [F(1,14) = 3.245, p = 0.0932], CFA treatment [F(1,14) = 0.3554, p = 0.5606], and no interaction (conditioning x CFA treatment) [F(1,14) = 0.6511, p = 0.4332] for the duration spent in the fentanyl-paired chamber Fig. 9A. For male rats, we found a significant interaction (conditioning x CFA treatment) [F(1,14) = 16.48, p = 0.0012], but no main effects of conditioning [F(1,14) = 0.1118, p = 0.7431] or CFA treatment [F(1,14) = 2.179, p = 0.1621] for the duration spent in the fentanyl-paired chamber Fig. 9B. Post hoc analysis of the significant interaction (conditioning x CFA treatment) revealed decreased time spent in the fentanyl-paired chamber for saline males during the post-test (p = 0.0389), but increased time spent in the fentanyl-paired chamber for CFA males during the post-test (p = 0.0154) Fig. 9B. In the fentanyl-paired chamber when comparing the change in preference duration between the pre-test and post-test, we found a significant effect of CFA treatment [F(1,28) = 11.03, p = 0.0025], a significant interaction (sex x CFA treatment) [F(1,28) = 4.517, p = 0.0425], but no effect of sex [F(1,28) = 2.434, p = 0.1300] Fig. 9C. Post hoc analysis of the significant interaction (sex x CFA treatment) revealed there was increased preference for the fentanyl-paired chamber in CFA males compared to saline males (p = 0.0012), but no difference in preference for the fentanyl-paired chamber between CFA females and saline females (p = 0.6455) Fig. 9C. This demonstrates significant preference for low dose fentanyl in CFA males only.

Fig. 9. Chronic inflammatory pain promotes fentanyl place preference in male rats, but not females.

(A) In female rats, there was no effect of CFA treatment on the duration of time spent in the fentanyl-paired chamber (p > 0.05). (B) There was a decrease in the duration of time spent in the fentanyl-paired chamber with saline males (*p < 0.05), but an increase in the duration of time spent in the fentanyl-paired chamber with CFA males (*p < 0.05). (C) In the fentanyl-paired chamber, there was a significant effect of CFA treatment on the change in preference duration (**p < 0.01). Additionally, there was an increase in preference duration [(time in the drug-paired chamber post-test) – (time in the drug-paired chamber pre-test)] for the fentanyl-paired chamber in CFA males compared to saline males (**p < 0.01). N = 32, male saline n = 8, male CFA n = 8, LgA, female saline n = 8, and female CFA n = 8.

We also analyzed differences in time spent in the saline-pared and center chambers as well as the change in preference duration in male and female rats. For female rats, we found no effect of conditioning (pre-test vs. post-test) [F(1,14) = 0.4356, p = 0.5200], CFA treatment [F(1,14) = 0.0582, p = 0.8129], and no interaction (conditioning x CFA treatment) [F(1,14) = 0.2351, p = 0.6353] for the duration spent in the saline-paired chamber Supplementary Fig. 4A. For male rats, we found no effect of conditioning (pre-test vs. post-test) [F(1,14) = 0.2903, p = 0.5985], CFA treatment [F(1,14) = 0.0464, p = 0.8326], and no interaction (conditioning x CFA treatment) [F(1,14) = 1.741, p = 0.2081] for the duration spent in the saline-paired chamber Supplementary Fig. 4B. In the saline-paired chamber when comparing the change in preference duration between the pre-test and post-test, we found no effect of CFA treatment [F(1,28) = 0.7796, p = 0.3848], no effect of sex [F(1,28) = 0.0151, p = 0.9029], and no interaction (sex x CFA treatment) [F(1,28) = 1.908, p = 0.1781] Supplementary Fig. 4C.

For female rats, we found no effect of conditioning (pre-test vs. posttest) [F(1,14) = 3.257, p = 0.0926], CFA treatment [F(1,14) = 0.5686, p = 0.4633], and no interaction (conditioning x CFA treatment) [F(1,14) = 0.8733, p = 0.3659] for the duration spent in the center chamber Supplementary Fig. 4D. For male rats, we found no effect of conditioning (pre-test vs. post-test) [F(1,14) = 0.1491, p = 0.7052], CFA treatment [F(1,14) = 3.594, p = 0.0788], and no interaction (conditioning x CFA treatment) [F(1,14) = 1.881, p = 0.1918 for the duration spent in the center chamber Supplementary Fig. 4E. In the center chamber when comparing the change in preference duration between the pre-test and post-test, we found no effect of CFA treatment [F(1,28) = 2.636, p = 0.1157], no effect of sex [F(1,28) = 1.072, p = 0.3093], and no interaction (sex x CFA treatment) [F(1,28) = 0.0754, p = 0.7856] Supplementary Fig. 4F. Together these findings demonstrate that we did not observe a significant change in preference for either the saline-paired chamber or center chamber in male and female rats. In summary, our findings demonstrate that CFA males were the only group to show fentanyl preference with a low dose of fentanyl.

4. Discussion

The present study was designed to test the effects of chronic inflammatory pain and fentanyl access (limited vs. extended) on fentanyl intake, motivation to acquire fentanyl, drug seeking for fentanyl, and Fos activity in VTA dopamine neurons of male and female rats. We hypothesized that chronic inflammatory pain would increase fentanyl intake, motivation to acquire fentanyl, and drug seeking using operant self-administration in male and female rats. We expected that chronic inflammatory pain would blunt pain-induced Fos activation in VTA dopamine neurons. We also hypothesized that chronic inflammatory pain would promote conditioned place preference for fentanyl. We found that chronic inflammatory pain promotes fentanyl place preference in males but does not change fentanyl self-administration or Fos activation in VTA dopamine neurons in either males or females. Additionally, we found that male and female rats given extended access to fentanyl escalate use over time, while animals given limited access to fentanyl do not. We also found that males given extended access to fentanyl demonstrate a greater increase in fentanyl intake over time than females given extended access to fentanyl. During DR testing, females given limited access to fentanyl demonstrate increased motivation to acquire fentanyl compared to males given limited access to fentanyl. After acute fentanyl challenge, we found that fentanyl-exposed females displayed more Fos activity in non-dopaminergic VTA neurons than fentanyl-exposed males.

Male and female rats acquired fentanyl self-administration over the course of 10 operant sessions. During this acquisition period, males demonstrated increased fentanyl intake and active lever pressing for fentanyl compared to females. Our data is consistent with findings showing increased acquisition of opioid self-administration in male rats compared to female rats with fentanyl (Bakhti-Suroosh et al., 2021; Bardo et al., 2022) and oxycodone (Mavrikaki et al., 2017). In contrast, our data differs from findings showing either increased acquisition of opioid self-administration in female rats (Malone et al., 2021; Thorpe et al., 2020) or no sex differences during acquisition of opioid self-administration (George et al., 2021; Mavrikaki et al., 2021). We observed that increases in operant responding for fentanyl in males may be partially due to general increases in operant responding because male rats demonstrated increased inactive lever pressing compared to female rats during acquisition of fentanyl self-administration. However, males and females demonstrated similar fentanyl intake and active lever pressing for fentanyl during the last three operant sessions, suggesting similar operant responding for fentanyl by the end of the acquisition period.

Male and female rats given either limited or extended access to fentanyl demonstrated significant mechanical hypersensitivity following induction of chronic inflammatory pain. Contrary to our hypothesis, we found that fentanyl intake by ShA and LgA animals was unaffected by inflammatory pain status. Our findings do not support that there is a greater risk of increasing opioid intake in males and females experiencing chronic inflammatory pain. Reiner and colleagues examined the effects of different pain manipulations including intraplantar capsaicin, visceral lactic acid, and intraplantar CFA on opioid self-administration, reinstatement of opioid seeking, and opioid vs. food choice (Reiner et al., 2021). In this study, the authors report a lack of effect of different pain-related manipulations on opioid intake, seeking, and opioid vs. food choice. Consistent with our findings, the authors conclude based on their findings and previous findings that once higher unit dose opioid self-administration is established, operant responding is resistant to change by pain manipulations (Reiner et al., 2021). In a very recent study, Higginbotham and colleagues report increased fentanyl intake in CFA males, but not CFA females compared to non-pain controls during week 3 of limited access to fentanyl (2 h per day) (Higginbotham et al., 2022). There are several methodological differences between this study and our own, including fentanyl dose, change of dose during the FR-1 schedule, operant session length, access conditions, and food restriction, but one critical difference of note is the timing of the pain manipulation. Here, we established stable operant responding to fentanyl prior to inducing chronic inflammatory pain. In contrast, in the Higginbotham et al. (2022) study, animals received CFA injections prior to acquisition of opioid self-administration. In a separate study, animals were given extended access to morphine (12 h per day) either early or late after a neuropathic spinal cord injury (SCI) (Woller et al., 2014). Intake of morphine varied depending on the phase of the neuropathic injury. During the acute phase (24 h), rats with SCI administered less morphine than sham animals, whereas during the early chronic (14 days) and late chronic (35 days) phases, rats with SCI administered more morphine than sham animals (Woller et al., 2014). Repeated use of opioids can lead to the development of analgesic tolerance, physical dependence, and hyperalgesia (Pahng and Edwards, 2021). Accordingly, the repeated exposure of animals to fentanyl prior to the introduction of pain may have unknown consequences on fentanyl-related behaviors. In sum, the timing of the introduction of pain and prior opioid exposure are likely to be critical factors in whether or not pain alters opioid self-administration.

In rats given limited access to fentanyl, we did not observe an escalation of fentanyl intake over time. This is consistent with data showing that male rats given limited access to fentanyl at the same dose do not escalate fentanyl intake over time (Wade et al., 2015). Our data shows that this is also the case for female rats given limited access to fentanyl. As expected, both ShA male and female rats exhibited preference for the active lever over the inactive lever. In contrast to our acquisition data, we did not observe sex differences in fentanyl intake, active lever responding, or inactive lever responding in ShA animals during these 14 operant sessions. Our finding of no sex differences with ShA animals is consistent with similar reports of animals given limited access to fentanyl (Malone et al., 2021) and oxycodone (Mavrikaki et al., 2021), but contrasts with evidence showing increased heroin intake in ShA female mice compared to ShA male mice (Towers et al., 2019). The absence or presence of sex differences in opioid self-administration seems to be at least partly dependent on the dose of opioid used. For example, Towers and colleagues found increases in heroin intake in females compared to males with 30 μg/kg/infusion, but observed no sex differences at a higher dose of 60 μg/kg/infusion (Towers et al., 2019). In a separate study, Towers and colleagues found increased operant responding in females with low doses (0.25 μg/kg/infusion and 0.75 μg/kg/infusion), but not with higher doses (1.5 μg/kg/infusion and 3.0 μg/kg/infusion) in animals that did not escalate fentanyl use over time (Towers et al., 2022). Comparably, in our non-escalating ShA rats, we observed no sex differences in number of fentanyl infusions with a dose of 2.5 μg/kg/infusion.

In our study, male and female rats given extended access to fentanyl significantly escalated fentanyl intake over time, demonstrating key symptomatology associated with opioid misuse and opioid use disorder (Edwards and Koob, 2013). These findings are consistent with data showing escalation of fentanyl intake over time in male rats at the same dose (Wade et al., 2015). Additionally, our data shows that female rats given extended access to fentanyl escalate intake over time. As expected, LgA animals demonstrated increased fentanyl intake and active lever responding compared to ShA animals. When examining the relationship between fentanyl intake and mechanical hypersensitivity, we observed that individual fentanyl intake levels were negatively correlated with paw withdrawal thresholds during opioid abstinence in LgA animals, but not ShA animals. This shows that LgA animals with the highest fentanyl intake in the previous operant session were the most sensitive to tactile stimulation during the subsequent test of mechanical hypersensitivity. Similarly, paw withdrawal thresholds during heroin withdrawal were negatively correlated with individual levels of heroin intake following extended access self-administration (Edwards et al., 2012). In our study, this relationship was only observed in the non-injected hindpaw of CFA and saline animals. This is most likely due to reduced individual variability (i.e., ceiling effect) of mechanical hypersensitivity in the injected hindpaw of CFA animals.

In contrast to our findings with ShA animals, we observed sex differences in fentanyl intake with LgA animals. We found that LgA males demonstrated a greater increase in fentanyl intake between the first and last operant sessions compared to LgA females. Our data are consistent with findings from Bardo and colleagues demonstrating increased fentanyl intake in male rats compared to female rats under extended access conditions (6 h per day) and at the same dose of fentanyl that we used in our study (Bardo et al., 2022). However, both our findings and those of Bardo et al. (2022) contrast with reports showing no reliable sex differences with fentanyl (Malone et al., 2021) or oxycodone (Fredriksson et al., 2020) under extended access conditions (6 h per day). Additionally, increased opioid self-administration has been reported in females compared to males with heroin (Towers et al., 2019) and oxycodone (Kimbrough et al., 2020) under extended access conditions (6 h per day and 12 h per day, respectively). While varying methodological differences could explain these contrasting findings, it is not possible at this time to conclude which experimental factors contributed to these mixed results. One underexamined factor and a limitation of our current study, is the examination of the estrous cycle in normal cycling female rats during opioid self-administration. It has been reported that heroin intake is decreased during the pro-estrus phase of the estrous cycle when estrogen levels increase to their highest levels (Lacy et al., 2016). Using estrogen- and progesterone-specific antagonists, Smith and colleagues determined that this effect was driven by estrogen (Smith et al., 2021). Furthermore, chronic administration of estrogen decreased heroin and remifentanil intake in intact female rats (Sharp et al., 2021), and this decrease in heroin intake has been shown in Lewis, Sprague-Dawley, and Long-Evans rats (Schmidt et al., 2021).

We also investigated the combined effects of CFA and fentanyl access on motivation to acquire fentanyl (PR, DR) and drug seeking for fentanyl. Like our FR-1 data, PR testing in males and females was unaffected by inflammatory pain status in either limited access or extended access conditions. Higginbotham et al. (2022) report higher breakpoints in ShA males experiencing pain compared to ShA females experiencing pain. In addition to the methodological differences mentioned previously, the PR parameters that vary between our two studies include fentanyl dose, session duration, and the steepness of the PR breakpoint step schedule. We modeled our PR breakpoint step schedule based on that used in Wade et al. (2015), which is less steep than the exponential PR breakpoint step schedule used in Higginbotham et al. (2022). In the PR test, formally LgA animals had higher breakpoints and more active lever presses than formally ShA animals at a dose of 2.5 μg/kg/infusion, demonstrating increased motivation to acquire fentanyl in LgA animals following escalation of fentanyl intake. Our findings are consistent with data showing that LgA male rats have higher breakpoints than ShA male rats at the same dose (Wade et al., 2015), and our findings also replicate this effect in female rats. During the PR test, there was a preference for the active lever over the inactive lever in ShA and LgA animals of both sexes, but males demonstrated increased inactive lever pressing compared to females. We did not observe any sex differences in breakpoint of either ShA or LgA animals. In the absence of pain, a similar lack of sex differences in breakpoint during PR testing was observed in ShA animals with oxycodone self-administration (Mavrikaki et al., 2021) and fentanyl self-administration (Higginbotham et al., 2022).

For DR testing, there was a significant DR effect for male and female rats demonstrating a decrease in the number of fentanyl infusions taken at higher fentanyl doses, but an increase in total fentanyl intake at higher fentanyl doses. Similar to the FR-1 and PR data, DR testing was unaffected by inflammatory pain status in animals either given limited or extended access to fentanyl. Previous work found that fentanyl self-administration at differing doses was not maintained in ShA animals with a neuropathic injury (Martin et al., 2007). In contrast, we found that ShA animals regardless of pain status maintained responding during DR testing. Our findings in ShA males also contrasts with a previous report showing that CFA altered heroin self-administration in a dose-dependent manner where high unit doses (0.2 mg/kg/infusion) were more reinforcing, but low unit doses (0.05 mg/kg/infusion) were less reinforcing in male rats (Hipolito et al., 2015) and the Higginbotham et al. (2022) study in which it was reported that ShA male rats experiencing pain have greater fentanyl intake at 5 μg/kg. As mentioned above, these discrepant results are likely due to other methodological differences including whether the pain manipulation occurs before or after establishing opioid self-administration. It possible, but less likely that the lack of pain effects on fentanyl self-administration were due to the pain state diminishing over the course of the multi-week experiment. In the present study, we only tested paw-withdrawal thresholds 1 week post CFA, however, our group has reliably demonstrated pain effects that last several weeks using the same CFA model with Long-Evans rats (Edwards et al., 2022) and Wistar rats (McGinn et al., 2020; Sharfman et al., 2022). Formally LgA males demonstrated an upward shift in the DR function compared to formally ShA males, indicating an increase in the reinforcing qualities of fentanyl consistent with a phenotype that is predictive of drug dependence and resistance to extinction (Edwards et al., 2007; Piazza et al., 2000; Ahmed and Koob, 1998). In contrast, there was no change in the reinforcing effects of fentanyl between formally ShA and LgA females, demonstrating similar motivation between females given limited and extended access to fentanyl. Our interpretation is that this was due to an upward shift in the DR function in ShA females compared to ShA males, indicating that ShA females showed greater motivation to acquire fentanyl than ShA males. A similar upward shift in the DR function in females compared to males under limited and extended access conditions was previously shown with oxycodone (Mavrikaki et al., 2017) and heroin (George et al., 2021).

During the drug seeking test, we found that both males and females previously given either limited or extended access to fentanyl demonstrated increased operant responding when the fentanyl syringes were replaced with saline syringes. This demonstrates that all groups exhibited significant drug seeking for fentanyl during this test. We did not observe any CFA effects on drug seeking for fentanyl. Changes in drug seeking behavior for morphine have been reported in rat models of inflammatory pain (Hou et al., 2015) and neuropathic pain (Gutierrez et al., 2021). In Hou et al., 2015, the authors found CFA had no effect on morphine intake (0.2 & 1 mg/kg/infusion) or drug seeking after abstinence from morphine at the higher dose (1 mg/kg/infusion). However, at the lower morphine dose (0.2 mg/kg/infusion), only CFA animals maintained drug seeking after abstinence (Hou et al., 2015). In another study, only rats with spared nerve injury (SNI) exhibited drug seeking for morphine at the beginning of extinction from morphine self-administration at a low dose (0.1 mg/kg/infusion). No changes in morphine intake with SNI were observed (Gutierrez et al., 2021). In both studies, non-pain controls did not demonstrate drug seeking behavior, which contrasts with our data and highlights the increased potency and abuse liability of fentanyl relative to morphine. This also suggests that potential ceiling effects are more likely to emerge with more potent opioids. Similar to the DR testing, there was a significant difference in drug seeking behavior between LgA males and ShA males, but no difference in drug seeking behavior between LgA females and ShA females.

We next induced time-locked Fos activation through acute fentanyl challenge and examined if chronic inflammatory pain would blunt Fos activity in VTA dopamine neurons of fentanyl exposed male and female rats. This expected outcome was based on evidence that chronic pain states (Ren et al., 2016; Watanabe et al., 2018) and withdrawal from chronic morphine exposure (Diana et al., 1999) reduce the intrinsic of excitability of VTA dopamine neurons in male mice and rats. We found equivalent Fos immunoreactivity in TH+ VTA neurons of all groups, demonstrating that Fos activation in VTA dopamine neurons of male and female rats was not blunted by chronic inflammatory pain. However, we found an overall increase in VTA Fos immunoreactivity in female rats compared to male rats, which was attributed to females having more Fos immunoreactivity in non-dopaminergic (TH−) VTA neurons than males. The VTA is a brain region with high cellular heterogeneity (Morales and Margolis, 2017). The majority of VTA neurons are dopaminergic (55–65%), while the remaining neurons are either GABAergic (30–35%) or glutamatergic (2–5%) (Avegno et al., 2021; Dobi et al., 2010; Margolis et al., 2006; Nair-Roberts et al., 2008; Tan et al., 2012; Yamaguchi et al., 2007). Our findings suggest that increased VTA Fos immunoreactivity in females following fentanyl challenge is occurring in a population of GABAergic or glutamatergic VTA neurons. Since all animals received fentanyl challenge, we do not know if there are baseline sex differences in non-dopaminergic VTA Fos immunoreactivity between males and females with a history of fentanyl exposure. This was a limitation of the current experiment, which was designed to test the effects of an analgesic level of fentanyl on VTA Fos activity in male and female animals with differing histories of pain and fentanyl exposure. Inclusion of vehicle-injected controls would have added 8 more treatment groups to the design, which was not practical.

In a final experiment, we tested male and female rats with and without CFA in a fentanyl place conditioning procedure. Previous studies with morphine CPP have shown that various pain manipulations either increase morphine CPP (Cahill et al., 2013; Lim et al., 2014; Zhang et al., 2014) or suppress morphine CPP (Narita et al., 2005; Ozaki et al., 2002, 2003; Petraschka et al., 2007; Suzuki et al., 1996) in male animals, which may be attributable to various methodological differences (Cahill et al., 2013). Compared to morphine, the fully synthetic opioid, fentanyl, has a relative potency 100–150 times greater than that of morphine (World Health Organization, 2018). Here, we tested a low dose of fentanyl (4 μg/kg) in previously opioid naïve animals, and we found that male and female rats not experiencing pain did not show significant fentanyl preference at 4 μg/kg. Our data are consistent with prior findings showing significant fentanyl preference in male rats of the same strain at 16 μg/kg, but not at 4 μg/kg (Vitale et al., 2003), and are partially consistent with prior work showing that male and female rats demonstrate fentanyl preference at 16 μg/kg, while only male rats demonstrate fentanyl preference at 4 μg/kg (Gaulden et al., 2021). Differences across studies in male rats could be due to strain differences (Long-Evans vs. Sprague Dawley rat) or differing experimental parameters including number of CPP chambers (3 chambers vs. 2 chambers), number of conditioning days (4 days vs. 8 days), and conditioning frequency (twice-a-day vs. once-a-day). Both our experimental parameters and findings in non-pain males at 4ug/kg were most aligned with the experiment parameters (i.e., Long-Evans, 3 chambers, 4 conditioning days, twice-a-day conditioning) and findings (i.e., no CPP in males at 4ug/kg) in Vitale et al. (2003).

In contrast to non-pain males and females, we report that CFA males demonstrate CPP for a context paired with a low dose of fentanyl. CFA females did not show significant fentanyl place preference. Together our findings suggest that chronic inflammatory pain promotes low-dose fentanyl preference in male rats, but not female rats. It is possible, but less likely that animals which did not demonstrate place preference did so because fentanyl impaired their ability to learn the context. Without a state-dependent control we cannot rule out this possibility. However, it is unlikely that such a low dose of fentanyl would impair an animal’s ability to learn the context. The use of only one dose of fentanyl is a limitation of this experiment. It remains to be seen whether inflammatory pain increases fentanyl place preference at moderate or higher fentanyl doses in males and females. In contrast to our self-administration experiment, animals were opioid naïve when inflammatory pain was introduced in the CPP experiment. As mentioned above, prior opioid history and the timing of the introduction of pain may be critical factors in whether pain alters fentanyl-related behaviors.

In conclusion, the present study highlights the effects of chronic inflammatory pain, fentanyl access conditions, and sex on Fos activity in VTA dopamine neurons and fentanyl self-administration measures including fentanyl intake, motivation to acquire fentanyl, and drug seeking for fentanyl. Our study also highlights the effects of chronic inflammatory pain and sex on preference for low dose fentanyl. These findings suggest that established fentanyl self-administration is resistant to change by inflammatory pain manipulation in males and females, but chronic inflammatory pain can increase the rewarding properties of low-dose fentanyl in males.

Data availability

Data will be made available on request.