Oxycodone Self-Administration in Female Rats is Enhanced by ∆9-tetrahydrocannabinol, but not by Cannabidiol, in a Progressive Ratio Procedure

Epidemiological evidence suggests that the legalization of cannabis may reduce opioid-related harms. Preclinical evidence of neuropharmacological interactions of endogenous cannabinoid and opioid systems prompts further investigation of cannabinoids as potential therapeutics for the non-medical use of opioids. In these studies female rats, previously trained to self-administer oxycodone (0.15 mg/kg/infusion) intravenously in 6 h sessions, were allowed to self-administer oxycodone after exposure to cannabidiol (CBD) and Δ9-tetrahydrocannabinol (THC) by vapor inhalation and THC by injection (5.0–20 mg/kg, i.p.). Self-administration was characterized under Progressive Ratio (PR) and Fixed Ratio (FR) 1 schedules of reinforcement in 3 h sessions. THC decreased IVSA of oxycodone in a FR procedure but increased reward seeking in a PR procedure. CBD decreased the IVSA of oxycodone in the FR but not the PR procedure. The results are consistent with an anti-reward effect of CBD but suggest THC acts to increase the reinforcing efficacy of oxycodone in this procedure.


Introduc on
Reduc on of opioid use concomitant with cannabis use (Kim et al., 2016;Nguyen et al., 2023;Pardo, 2017;Piper et al., 2017;Shi, 2017) may arise because cannabis acts Address Correspondence to: Dr. Jacques D. Nguyen, One Bear Place, Waco, TX, 76798-7334 Email: jacques_nguyen@baylor.eduOr Dr. Michael A. Taffe, Department of Psychiatry, 9500 Gilman Drive; University of California, San Diego, La Jolla, CA 92093; USA; Email: mtaffe@ucsd.edutherapeu cally to reduce the mo va on to seek opioids, because it is subs tu ng for the opioid or because it is enhancing the impact of a unit dose of the opioid.Cannabis smoking slightly increased the abuse-related subjec ve effects of oxycodone in a human laboratory study (Cooper et al., 2018), but self-administra on was not directly assessed.It is difficult to experimentally determine if cannabis reduces opioid use in humans, but animal models can assist in dis nguishing pro-reward from an -abuse effects.
A recent inves ga on in a nonhuman primate model concluded there was no impact of cannabinoids on opioid rewarding efficacy (Maguire and France, 2020) however that study used a full agonist cannabinoid and the selfadministra on of a very short ac ng opioid, remifentanil, with the explicit intent that animals could not reach a sa ety point for remifentanil under the experimental condi ons.Use of a drug-food choice procedure (Carey et al., 2023) may suggest that the cannabinoid would have no effect on rela ve drug vs food mo va on, but this poten al interac on is not well established by the model.Δ 9 -tetrahydrocannabinol (THC) reinstates beer-and sucrose-seeking behavior in rats, for example (McGregor et al., 2005).THC has also been shown to increase the reinforcing efficacy of cocaine in a monkey model, but THC is not itself readily self-administered intravenously by rhesus monkeys (John et al., 2017;Mansbach et al., 1994) or by rats (Takahashi and Singer, 1979); for review see Jus nova et alia (Jus nova et al., 2005).
In our recent work we concluded that parallel shi s in the dose-response func on for THC treatment across different unit doses of oxycodone, combined with an increase in PR breakpoint, is most compa ble with the interpreta on that THC enhances the rewarding value of a unit dose of self-administered oxycodone (Nguyen et al., 2019).
In animal models it is possible that a generally seda ng effect of cannabinoids may disrupt behavioral responding non-specifically, effects that are misinterpreted as being specific to the mo va on to seek drug effects in a selfadministra on procedure.Cau on should be used when interpre ng the impact of full-agonist compounds that reduce response rate where THC does not do so.This is also a reason why demonstra on of increased behavioral responding in a PR procedure associated with THC is cri cal addi onal support for the interpreta on of enhanced reinforcer efficacy.
Although Δ 9 -tetrahydrocannabinol is undoubtedly the cons tuent of cannabis that drives most use, for both recrea onal and therapeu c purposes, addi onal cons tuents may also produce significant effects.Cannabidiol (CBD) reduces alcohol or cocaine seeking behavior in rats when applied via a dermal gel (Gonzalez-Cuevas et al., 2018), as well as cocaine (Galaj et al., 2020) or methamphetamine (Hay et al., 2018) seeking in rats and alcohol seeking in rats (Tringali et al., 2023) (Mirmohammadi et al., 2022).There does not appear to be any evidence whether CBD can alter the rewarding value of opioids.A recent review of studies involving human substance users found most studies are of low quality and only hinted at a general beneficial effect of CBD on opioid craving and anxiety (Lo et al., 2023), e.g., (Suzuki et al., 2022); also see (Fernandes et al., 2023;Paulus et al., 2022).The effect may be route and/or species specific since effects of CBD in rats have only variably been demonstrated with parenteral injec on (Boggs et al., 2018).
There is no effect of oral CBD on sustained a en on in rats (Moore et al., 2023).
Furthermore, CBD vapor inhala on decreases rat body temperature in a par ally serotonin 1A receptor-dependent manner (Javadi-Paydar et al., 2019), but no effects on body temperature are observed a er CBD injec on (Taffe et al., 2015).The impact of CBD may also be species dependent since oral cannabidiol did not affect the self-administra on of alcohol in baboons (Moore et al., 2023).
This study was therefore conducted to determine if treatment with THC or CBD by vapor inhala on alters the intravenous selfadministra on of oxycodone in a rat model.

Acquisi on:
Female rats were implanted with intravenous catheters and trained to selfadminister oxycodone (0.15 mg/kg/infusion) on a Fixed Ra o (FR) 1 schedule of reinforcement using methods previously described (Nguyen et al., 2019;Nguyen et al., 2021;Nguyen et al., 2018;Nguyen et al., 2017).In this study, separate groups were ini ally injected with the vehicle (N=7) or with THC (N=8) 30 minutes before each 6 h self-administra on session; treatments were the same for housing pairs.The first 10 sessions were conducted on sequen al days and the THC dose was 5 mg/kg, i.p. for the first 10 sessions, 10 mg/kg, i.p. for Sessions 11-14 and 20 mg/kg, i.p. for Sessions 15-19.IVSA was suspended for a week between Sessions 19 and 20 to permit assessment of the acute effects of THC (5, 10, 20 mg/kg, i.p.) on rectal temperature and nocicep on assays.For Sessions 22-28, the original Vehicle group was injected with 5 mg/kg, THC, i.p., prior to the IVSA sessions and the original THC group was injected with the vehicle.
The response requirement was incremented to FR5 for Sessions 30-32.

Progressive Ra o:
On sessions 33-37 animals were tested on a Progressive Ra o (PR) procedure with 0.15 mg/kg/infusion oxycodone available.In the PR paradigm, the required response ra o was increased a er each reinforcer delivery, within a session (Hodos, 1961;Segal and Mandell, 1974) as determined by the following equa on (rounded to the nearest integer): Response Ra o=5e^(injec on number*j)-5 (Richardson and Roberts, 1996).In this study the j value was set to 0.2.Sessions were a maximum of 3 h in dura on.Animals were injected with either Veh or THC (5 mg/kg, i.p.) before the 34th and 35th sessions in a counter-balanced order.The 36th and 37th sessions were conducted under PR, without any pre-session treatment.The per-infusion dose of oxycodone was changed to 0.06 mg/kg/inf for Sessions 38-41 and the rats were injected with either Veh or THC (5 mg/kg, i.p.) before the 39th and 40th sessions in a counterbalanced order.For analysis, the day following the injec on days was used as the no-injec on comparison.
The study next evaluated injec on of 0, 5, 10 and 20 mg/kg of  In our approach a single priming infusion is delivered if no responses have been made within 30 minutes of session ini a on under FR (this is not included for PR).These are rare a er the ini al few sessions of acquisi on and are thus reported qualita vely where present, but not formally analyzed.In all analyses, a criterion of P<0.05 was used to infer that a significant difference existed.Any significant main effects were followed with post-hoc analysis using Tukey  ) in a counter-balanced order, followed by a noinjec on day.The effect on self-administra on of the oxycodone 0.15 mg/kg/infusion dose was evaluated first and the 0.06 mg/kg/infusion dose evaluated the following week.A significant difference between pretreatment condi ons, across oxycodone dose, is indicated with *.THC vapor inhala on produced a small reduc on in the infusions obtained (Figure 3B) which did not reach sta s cal significance (P=0.054).Considering that 5 mg/kg, i.p., may be a threshold dose (Nguyen et al., 2019), and perhaps subthreshold given prior repeated exposure of these rats to THC during acquisi on or maintenance phases, we determined the effects of pretreatment with THC (5, 10, 20 mg/kg, i.p.) in a counter-balanced order.THC significantly [F (3, 33) = 2.93; P<0.05] reduced the infusions obtained and the Dunne post-hoc test further confirmed that THC significantly reduced infusions obtained rela ve to vehicle injec on at the 10 mg/kg, i.p., dose, but not at the 20 mg/kg, i.p. dose (P=0.065),(Figure 3C).There was a priming infusion delivered to one rat a er 5 mg/kg and another rat a er 10 mg/kg injec on in this experiment.

THC, but not Cannabidiol, Inhala on
Increases Oxycodone Self-Administra on in a PR Procedure It was a completely novel observa on that CBD inhala on reduced oxycodone IVSA, so we then went on to further contrast the effects of inhaled PG, CBD and THC in a counter balanced order using a PR procedure.This was conducted first with the 0.15 mg/kg/infusion oxycodone dose and then with the 0.30 mg/kg dose (Wade et al., 2015), across sequen al weeks.
THC inhala on increased drug seeking in the PR procedure (Figure 4).In contrast, CBD inhala on did not significantly alter oxycodone seeking rela ve to the PG inhala on condi on.The  In addi on, rats reached higher breakpoints (Figure 4A) and emi ed more total correct responses (Figure 4B) a er THC vapor inhala on compared with either PG or CBD vapor inhala on.

Discussion
This study replicates and generalizes our con ngency.Our previous study was conducted in male rats and while several condi ons of oxycodone IVSA were evaluated, it was all conducted with a FR1 schedule of reinforcement (Nguyen et al., 2019).In that prior study, however, THC did increase breakpoints in a PR procedure when the animals were switched to the IVSA of heroin (0.006, 0.06 mg/kg/infusion doses) and the present study found that the effect of FR vs PR reinforcement schedule on the impact of THC extended to oxycodone IVSA in female rats.The increased behavioral output (e.g., total drug-associated responses) in the PR procedure associated with THC (Figure 1, 4) is also inconsistent with an inference that reduc ons in drug seeking in the FR1 procedure are due to general seda ng proper es of the THC.
Although inhaled THC did not significantly reduce oxycodone IVSA (Figure 3B) as it did in our prior study, this is almost certainly a ma er of THC dose, possibly related to tolerance created by the repeated THC pretreatments during the prior study in these rats.
This interpreta on is supported by the injec on experiments (Figure 3C), which established that THC delivered intraperitoneally reduces oxycodone IVSA under a FR1 con ngency in the female rats, just as it did with males in our prior study (Nguyen et al., 2019).Thus, the lack of effect of the inhaled THC was likely due to dose, since the THC injec on study showed that 10 mg/kg, but not 5 mg/kg, was able to reduce oxycodone IVSA.
It was somewhat unexpected that rats reached similar breakpoints for 0.06 and 0.15 mg/kg/infusion in the PR under control condi ons (PG here, also true in the notreatment acclima on sessions) given an apparent difference across this dose range that was observed in prior work with male rats trained in sessions of 1 h or 12 h dura on (Nguyen et al., 2021;Nguyen et al., 2018;Wade et al., 2015).In par cular, Wade and colleagues showed that the largest difference was between the 0.06 and 0.15 mg/kg/infusion doses in the animals trained in 12 h sessions and between the 0.15 and 0.30 mg/kg/infusion doses in the animals trained in 1 h sessions.In this study rats did, however, exhibit differen al intake for 0.06, 0.15 and 0.3 mg/kg/infusions under the FR1 response con ngency (Figure 5A), demonstra ng the expected behavioral compensa on.In our prior study, male rats selfadministered more infusions at the 0.06 mg/kg dose compared with the 0.15 mg/kg dose under an FR1 response con ngency, and this did not interact with the effects of THC (Nguyen et al., 2019).However, there was no effect of heroin unit dose (0.006 vs 0.06 mg/kg/infusion) in a PR procedure, nor any interac on of the effect of THC with heroin unit dose, in that study.
Therefore, the prior results were quite similar to the impact of unit dose of oxycodone across FR1 and PR procedures in the current study.access IVSA in male and female rats (Mavrikaki et al., 2017;Pravetoni et al., 2014).Similarly, behavioral escala on, and disrup ons of brain reward systems have been observed with extended access IVSA (Blackwood et al., 2019a;Blackwood et al., 2019b;Matzeu and Mar n-Fardon, 2020;Nguyen et al., 2021;Nguyen et al., 2017;Wade et al., 2015).
In contrast with THC, the impact of CBD was to reduce oxycodone seeking under FR1 and to leave it unaffected under the PR1 con ngency.
CBD has been previously shown to have andrug-seeking effects in rats (Gonzalez-Cuevas et al., 2018) when delivered by the transdermal route and to diminish withdrawal associated with chronic nico ne exposure (Smith et al., 2021) when injected repeatedly with the nico ne.CBD i.p. also reduced cocaine IVSA in rats under FR1 and PR schedules of reinforcement with the effects significant in lower unit doses of cocaine but not with higher unit doses of cocaine (Galaj et al., 2020).These effects were mediated by CB2, TRPV1 and 5HT1A receptors but not by CB1, GPR55 or MOR receptors.Effects of inhaled CBD in the present study were therefore consistent with results of Gonzalez-Cuevas and colleagues (2018), since oxycodone infusions were reduced in the FR1 study.CBD also failed to affect cocaine IVSA under a PR schedule in rats in prior study (Mahmud et al., 2017).This la er finding in the present study func ons as a nega ve control to further emphasize the impact of THC is to increase, not decrease, the rewarding value of a unit dose of oxycodone.
Interes ngly, the inhala on of vapor from cannabis extracts with high CBD content (64.2% CBD and 7.1% THC) decreased fentanyl IVSA under a FR, but not a PR, schedule in rats with and without induced neuropathic pain (Rivera-Garcia et al., 2023).Interac ve effects of CBD administered with THC remain unclear in rodent models (Boggs et al., 2018), however this result implies that the impact of the CBD overcame the impact of the THC in that study, rather than enhancing the impact of THC, e.g., via metabolic inhibi on (Varvel et al., 2006).
In conclusion these studies provided addi onal evidence that THC reduces oxycodone seeking under easy-access condi ons because it enhances the rewarding value of a unit dose of oxycodone.When access requires more work in a PR procedure, THC increases oxycodone seeking responding.In contrast, the effect of CBD was to decrease oxycodone self-administra on in easy access condi ons but to have no effect in harder access condi ons.This cau ons that cannabis is not necessarily reducing addic on liability in associa on with any observed reduc on in oxycodone intake.

2. 5
Data Analysis: Infusions obtained, the percentage of responses directed at the drug-associated lever, total correct responses and breakpoints (for Progressive Ra o) were analyzed by ANOVA, or by mixed-effect models where there were missing values.Within-subjects factors of Dose or Pre-treatment were included where relevant.

(
mul -level factors), Sidak (two-level factors) or Dunne (to assess change rela ve to one of the Treatment Condi ons.) correc on.All analysis used Prism for Windows (v.9.5.1;GraphPad So ware, Inc, San Diego CA)., i.p.) increases breakpoints reached by female rats in the progressive ra o procedure (Figure 1), confirmed by a significant effect of pre-session Treatment Condi on [F (2, 22) = 3.86; P<0.05].One subject did not complete the 0.15 mg/kg with THC condi on, thus the analysis was mixed effects.The impact of THC was sta s cally indis nguishable across oxycodone unit doses (0.06, 0.15 mg/kg/infusion) as there was no significant effect of oxycodone Dose nor an interac on of Dose with Treatment Condi on.The post-hoc Dunne test of the marginal mean confirmed significantly lower breakpoints were reached in the No-Injec on condi on compared with the THC condi on, across oxycodone Dose.ons (3 h sessions to match the maximum dura on of the PR sessions), to determine if these were func onally different doses in IVSA.This demonstrated that the 0.15

Figure 1 :
Figure 1: Mean (N=11-12 female rats; ±SEM) andindividual breakpoints reached in the PR procedure following injec on with Vehicle or THC (5 mg/kg, i.p.) in a counter-balanced order, followed by a noinjec on day.The effect on self-administra on of the oxycodone 0.15 mg/kg/infusion dose was evaluated first and the 0.06 mg/kg/infusion dose evaluated the following week.A significant difference between pretreatment condi ons, across oxycodone dose, is indicated with *.
22) = 21.56;P<0.0001; Figure 2C], compared with the lower unit dose or vehicle.There was also a significant reduc on in responses on the alternate manipulandum during the me-out [Main Effect: F (2, 22) = 3.92; P<0.05; Figure 2D] compared with the vehicle session.3.3Cannabidiol and THC Decrease OxycodoneSelf-Administra on in a FR Procedure CBD was delivered by vapor inhala on because in our hands it produces an involuntary responses (lowering of body temperature) in rats(Javadi-Paydar et al., 2019;Javadi-Paydar et al., 2018) whereas the injec on of CBD at 20-60 mg/kg, i.p., has no effect(Taffe et al., 2015).The impact of 30 min inhala on of vapor from CBD (400 mg/mL in the Propylene Glycol vehicle) prior to the IVSA session was then contrasted with the impact of vapor inhala on of THC (100 mg/mL in the PG).CBD inhala on significantly [F (2, 18) = 31.42;P<0.0001] reduced the number of infusions obtained in comparison with PG vapor inhala on or a session in which there was no exposure prior to the IVSA session (Figure3A).Reponses on the drug-associated

Figure 2 :
Figure 2: Mean (N=12; ±SEM) A) Infusions, B) Percent Drug-associated responses, C) responses on the drug-associated manipulandum during the me out and D) responses on the alternate manipulandum during the me out for selfadministra on of oxycodone in 3 h sessions under a FR1 response con ngency.Tukey post-hoc test.A significant difference between pre-treatment condi ons is indicated with *.

Figure 3 :
Figure 3: A) Mean (N=10; ±SEM) oxycodone (0.15 mg/kg) infusions obtained a er exposure to vehicle (PG) or cannabidiol (CBD 400 mg/mL) vapor under a FR1 response con ngency.B) Mean (N=10; ±SEM) oxycodone (0.15 mg/kg) infusions obtained a er exposure to air, or PG or Δ 9 -tetrahydrocannabinol (THC; 100 mg/mL) vapor.C) Mean (N=12; ±SEM) oxycodone (0.06 mg/kg) infusions obtained a er THC injec on (0-20 mg/kg, i.p.)The corresponding responses on the drug-associated manipulandum is depicted in panels D, E and F. In panel F one individual responding 166 mes in the vehicle condi on and one individual responding 190 mes a er 10 mg/kg are omi ed to facilitate scale comparison with the other panels.A significant difference between pre-treatment condi ons is indicated with *.

Figure 4 :
Figure 4: Mean A) Breakpoint, B) Total responses on the drug-associated manipulandum, C) Infusions of oxycodone (0.15, 0.30 mg/kg) obtained and D) Responses on the drug-associated manipulandum during the me-out interval during self-administra on under a Progressive-Ra o response con ngency.Significant differences associated with pre-treatment condi on, across oxycodone dose, are indicated with *.
Self-Administra on in a FR Procedure FR1 oxycodone dose-response: We re-determined the impact of a range of oxycodone doses (0.06-0.3 mg/kg/infusion) under FR1 because the oxycodone dose did not alter self-administra on of oxycodone under a PR procedure.The one-way ANOVA confirmed a significant impact of oxycodone Dose [F (2, 14) = 60.19;P<0.0001] on infusions obtained and the Tukey post hoc further confirmed significant differences between all doses (Figure 5A).FR1 CBD dose-response: Because CBD did not alter oxycodone IVSA under a PR procedure, we re-determined the impact of CBD vapor under FR1 to rule out a poten al development of tolerance to the CBD.For this we tested two CBD vapor doses (100 and 400 mg/mL) and the 0.06 mg/kg/infusion dose of oxycodone.The one-way ANOVA confirmed a significant impact of Pre-Treatment condi on [F (2, 14) = 5.79; P<0.05] on infusions obtained and the Tukey post-hoc further confirmed significantly fewer infusions were obtained a er CBD 400 exposure compared to PG exposure (Figure 5B).
, while it reduces oxycodone selfadministra on under a Fixed Ra o 1 (FR1)

Table 1 :
Summary of experiments by cumula ve self-administra on session.