Published online 2022 Dec 5. doi: 10.1016/j.addicn.2022.100050
Associated Data
Abstract
Chronic pain patients report analgesic effects when using cannabidiol (CBD), a phytocannabinoid found in whole-plant cannabis extract (WPE). Several studies suggest that cannabis-derived products may serve as an analgesic adjunct or alternative to opioids, and importantly, CBD may also attenuate the abuse potential of opioids. Vaping is a popular route of administration among people who use cannabis, however both the therapeutic and hazardous effects of vaping are poorly characterized. Despite the fact that chronic pain is more prevalent in women, the ability of inhaled high-CBD WPE to relieve pain and reduce opioid reward has not been studied in females. Here, we present a comprehensive analysis of high-CBD WPE vapor inhalation in female rats. We found that WPE was modestly efficacious in reversing neuropathy-induced cold allodynia in rats with spared nerve injury (SNI). Chronic exposure to WPE did not affect lung cytoarchitecture or estrous cycle, and it did not induce cognitive impairment, social withdrawal or anxiolytic effects. WPE inhalation prevented morphine-induced conditioned place preference and reinstatement. Similarly, WPE exposure reduced fentanyl self-administration in rats with and without neuropathic pain. We also found that WPE vapor lacks of reinforcing effects compared to the standard excipient used in most vapor administration research. Combined, these results suggest that although high-CBD vapor has modest analgesic effects, it has a robust safety profile, no abuse potential, and it significantly reduces opioid reward in females. Clinical studies examining high-CBD WPE as an adjunct treatment during opioid use disorder are highly warranted.
1. Introduction
Medical and recreational cannabis use has rapidly increased in the United States since the approval of the 2018 Farm Bill and other legislative changes [1, 2]. The management of chronic pain is the predominant medical condition related to cannabis use [3, 4], and inhalation is the most common route of administration used by humans [5]. Although smokable flower is the most common form of cannabis sold in the United States (41% market share), vaporizable cannabis extracts constitute an additional 22% of sales in regulated markets [6]. Cannabis consumers inhale vaporized cannabis extracts through electronic cigarette (eCIG) devices for recreational and medical purposes [7]. The popularity of vaping amongst medical patients has risen steadily due to the discreet characteristics of eCIGs, the immediate onset of drug effects, as well as the perception of safety compared to smoking [7–9]. Nevertheless, some studies have highlighted the potential health hazards associated with the use of eCIG devices and their excipients (e-liquids), especially after chronic use [10, 11]. Although inhalation is the most common route of cannabis consumption in humans, cannabis vapor effects in preclinical research remain critically understudied. The majority of cannabinoid research evidence relies on non-inhalation routes of administration which also have a less favorable pharmacokinetic profile. Previous studies lack face validity, may have limited translational value and, the comparative therapeutic efficacy of inhaled vs. other delivery methods is not well-defined.
Several extraction techniques are used to concentrate the active pharmaceutical ingredients of cannabis plant [12] while removing undesirable components, resulting in a phytochemically-rich whole-plant cannabis extract (WPE) [13]. Through vaporization, people consume WPE which contains a mixture of terpenoids, flavonoids and phytocannabinoids. These constituents have different pharmacologic profiles and interactions (antagonistically, additively, or synergistically), inducing consequently a wide range of effects [14–16]. However, these poorly characterized interactions are often dependent upon the experimental species and methodology, and results are difficult to replicate or generalize [17–19]. Furthermore, the majority of preclinical cannabinoid studies have analyzed the effects of synthetic and/or isolated compounds which have different pharmacologic profiles and intrinsic activity than the phytocannabinoids found in commercial products [20]. The therapeutic and side effects of WPE have not been thoroughly examined; these studies are highly warranted given the millions of people consuming WPE products worldwide [5].
The most well-characterized constituent in WPE is Δ9-tetrahydrocannabinol (THC), which is the main psychoactive compound in the cannabis plant and is also directly linked to the rewarding and analgesic effects [15, 21, 22]. Cannabidiol (CBD) is the most abundant cannabinoid in hemp-derived WPE varieties [23, 24]. Unlike THC, CBD does not produce rewarding effects or psychomotor impairment [25], and it has numerous potential therapeutic applications, including anti-inflammatory and analgesic effects [3, 4]. Although numerous CBD products are marketed for pain relief, this potential therapeutic effect may vary by pain condition, and there is likely a substantial expectancy effect [23, 26–28]. Moreover, chronic pain produces adaptations in the endocannabinoid system in brain regions that are critical for pain processing [29]. Thus, it is critical to conduct systematic research about high-CBD WPE’s antinociceptive effects and health impacts in the context of persistent pain and prolonged drug administration. Opioids are the most widely-used pharmacological tool for chronic pain management, however they produce multiple unwanted and potentially lethal side effects (e.g. constipation, impaired sleep, tolerance, abuse liability, respiratory depression, and overdose risk), which are limitations for their therapeutic use [30]. Clinical studies have demonstrated the opioid-sparing effects of cannabis [31–34]. For example, the combination of opioids and THC or other CB1R agonists produces potentiated antinociception in different species and assays, while reducing the development of tolerance [35–38]. However, the role of CBD in these effects is poorly defined and the potential of inhaled high-CBD WPE to attenuate tolerance and enhance morphine antinociception has not been characterized.
Moreover, the promising results of previous clinical and preclinical studies have largely relied upon male subjects. For instance, in one seminal study only 14% of human research participants were female [34] and, the ability of CBD to reduce morphine reward and the reinstatement of heroin-seeking behavior has only been studied in male rats [39, 40]. This incomplete examination of cannabinoid effects in females is of immense concern, given that women 1) have higher prevalence of chronic pain conditions [41, 42]; 2) consume more CBD products compared to men [42, 43], and 3) are more vulnerable to opioid abuse and withdrawal [44]. Therefore, there is a critical need to analyze the antinociceptive and opioid-modulating effects of inhaled high-CBD WPE in females.
Novel therapeutic approaches to analgesia, opioid sparing, and harm reduction must be efficacious in females, if they are to be deployed in the general population. Using a highly translational approach, this study assessed the ability of inhaled high-CBD WPE to enhance the antinociceptive effects of opioids, reduce opioid tolerance, and attenuate opioid reward in female rats. Our in-depth analysis included a robust behavioral and physiological characterization of acute and chronic effects of WPE vapor inhalation in pain-naïve animals as well as those with neuropathic pain.
2. Materials and Methods
2.1. Animals
Young adult female Long Evans rats (N = 196, PND 40 at time of arrival) were purchased from Envigo (Indianapolis, IN) and were paired-housed in a vivarium on a reverse 12-hour light/dark cycle (lights on at 19:30) with ad-libitum access to water and standard chow. Rats were acclimatized to the animal facilities for one week before starting any procedure. Some animals were implanted with a subcutaneous RFID transponder from Unified Information Devices (Lake Villa, IL) to monitor body temperature. Animals chronically exposed to vapor were monitored every day using a rat vaginal impedance meter (Muromachi MK-12, Tokyo, Japan), that measures the electrical impedance of the vaginal mucosa. A reading of 3kOhm (range of 0–19kOhm) is an indication of proestrus phase, which corresponds to the human follicular phase and aligns with the surge of estrogen and progesterone [45]. All experiments were performed during the dark phase and in accordance with the guidelines of the National Institute of Health Guide for the Care and Use of Laboratory Animals and were approved by the Institutional Animal Care and Use Committee at Legacy Research Institute.
2.2. Drugs
WPE (hemp-derived glandular trichome oil, Om extracts, Applegate, OR) with a composition of 64.2% CBD and 7.1% THC (verified by HPLC; Chem History Cannabis Testing Laboratory, Portland, OR) was used for vapor inhalation experiments. A mixture of 1:1 propylene glycol (PG; Sigma-Aldrich, St. Louis, MO) and vegetable glycerin (VG; Biopharm, Hatfield, AR) was used as a control (PGVG) in all vapor delivery experiments.
Morphine sulfate (Sigma-Aldrich, St. Louis, MO) was dissolved in 0.9% sterile saline. For repeated morphine administration experiments, morphine (10 mg/kg s.q.) was administered twice daily (0900 and 1600). To assess the development of morphine tolerance, a within-session cumulative quarter-log dose-response analysis was conducted (1.8, 3.2, 5.6, 10 and 18 mg/kg) as previously described [46]. Fentanyl HCl (National Institute on Drug Abuse, Drug Supply Program, Rockville, MD) was dissolved in 0.9% sterile saline for intravenous (i.v.) self-administration experiments.
2.3. Vapor chamber apparatus
Sealed vapor chambers (26 cm × 20 cm × 33.5 cm; La Jolla Alcohol Research, La Jolla, CA) were used for passive and operant vapor self-administration protocols. Briefly, each chamber delivered unidirectional flow of vapor through a port in the front of the chamber, the air intake port was connected to both an air flow meter (~1L/min) and a commercial e-cigarette cartridge (SMOK TFV8 X-Baby Tank with 0.4 Ω Q2 atomizer) filled with WPE or PGVG. A coil within the cartridge was heated and air was pulled through the tubing delivering the vapor, which was removed through an exhaust valve connected to a Whatman in-line HEPA filter. To minimize the formation of voltage-dependent eCIG toxins (e.g. methacrolein and benzene) [47], before any experimental manipulation, we determined the optimal wattage to be applied to the atomizer coil. At 19W, there was a visibly prominent vapor cloud, no combustion-like odor, > 2.5 mg CBD per 5 s vapor delivery puff, and < 6 mg total particulate matter per 5 s puff (verified by HPLC). Based on these measurements, we estimate 0.25 mg THC delivery per 5 s vapor puff. For passive exposure experiments, animals were acutely (single session) or chronically (two sessions a day for 20 days) exposed to 5 s puffs of vapor every 30 s during 30 min sessions. Air control animals were placed in the chamber for 30 min without receiving any vapor,
2.4. Vapor Self-administration
Animals were individually placed in the chambers fitted with two nose-poke ports and their corresponding cue lights (Med Associates, Fairfax, VT). After a nose-poke in the active port, a 5 s vapor puff was delivered and a yellow cue light located 9 cm above the active port was illuminated and remained on throughout the timeout period of 60 s, during which no vapor would be delivered. The active nose-poke port was right/left counterbalanced among animals. Responses in the inactive nose-poke were recorded but no consequences were associated. The animals were trained under a fixed ratio (FR)-1 reinforcement schedule for 10 consecutive days (30 min sessions), then transitioned to FR2 (Days 11–15) and FR4 (days 16–20) schedules. On the final day (Day 21), animals were tested under a progressive ratio (PR) schedule for 180 min and active responses were paired with vapor deliveries, as previously in [48].
2.5. Plasma Cannabinoid Quantification
To quantify plasma levels of CBD, THC, 11-nor-9-carboxy-Δ9-tetrahydrocannabinol (COOH-THC), 11-hydroxy-tetrahydrocannabinol (OH-THC), and 11-nor-9-carboxy-(−)-trans-Δ9-tetrahydrocannabinol-β-D-glucuronide (11-COOH-THC-Gluc), blood samples were collected 30–120 min after acute vapor exposure. Animals were deeply sedated with 2% isoflurane and intracardiac blood was collected (~200 μl). The samples were centrifuged at 4°C at 4000 × g for 15 min, and stored at −80°C until analysis of cannabinoids plasma concentrations, by liquid chromatography mass spectrometry as described previously [49].
2.6. Lung histology
After the last chronic vapor exposure on Day 20, rats were deeply anesthetized with 2.5% V/V isoflurane and intracardially perfused with 0.9% saline followed by buffered 4% paraformaldehyde (PFA). Lungs remained in 4% PFA for post-fixation during 24 h. The tissue was cryoprotected using 30% sucrose solution for 72 h. Using a cryostat (Leica CM3050 S), lungs were transversely sectioned at 20 μm and later stained with hematoxylin and eosin [50]. The stained slides were visualized using an optical microscope (Leica DM1000), digitally scanned and analyzed using ImageJ software (NIH, Bethesda, MD). As previously described [51], micrographs were used to quantify the mean linear intercept (MLI), which is a standardized measure of lung morphometric changes (free distance between gas exchange surfaces in the acinar airway complex).
2.6.1. Spared Nerve Injury Surgery
Rats were anesthetized with 2–2.5% V/V isoflurane, delivered in a mixture of 10% O2 and the surgery was performed as described before [52]. Briefly, the SNI model involved the ligation of the left tibial and the common peroneal nerves, leaving the sural nerve intact. In sham animals the sciatic nerve was visualized but all nerve branches remained intact. All animals received carprofen (2 mg/tab) and enroflaxacin (2 mg/tab) tablets the day before surgery and for 2 days post-surgery and were allowed to recover for 7 days prior to experimental manipulations.
2.7. Behavioral analysis
Animals were handled and habituated to experimental conditions under red light for at least 30 m for two consecutive days before any behavioral manipulations. In all cases, rats were randomly assigned to the experimental treatments and tested on different behavioral assessments within one hour after vapor exposure.
For nocifensive-related assays, rats were individually placed in clear acrylic chambers (19.5 × 10 × 12.5 cm), either mesh or glass flooring was used as required by each assay.
2.7.1. Pain-related behavior
Mechanical hyperalgesia (response threshold) was measured using an electronic von Frey apparatus from Bioseb (Pinellas Park, FL). On testing days, rats were placed in the chamber and were tested after a 15 m habituation. Each animal was tested three times, one minute apart per trial, and the average of the three trials was considered the mechanical threshold. Cold allodynia response was measured using acetone evaporation, where acetone was drawn in a repeater pipette programmed to dispense 50 μL of liquid. During 30 s, reactions such as paw flicking or licking were observed and scored as follows: 0 – no response, 1 – quick withdrawal, flick, or stamp of paw, 2 – prolonged withdrawal or repeated flicking of the paw, 3 – repeated flicking (> 2 s) with paw licking [53]. Each animal was tested three times, one minute apart per trial to obtain an average value. Thermal antinociception and hyperalgesia (paw withdrawal latency) to radiant heat was measured using the method originally described by Hargreaves [54]. Rats were acclimated to the apparatus (IITC Life Sciences, Woodland Hills, CA) during the 2-day habituation phase. The light was optimized to 25% intensity, resulting in a baseline behavioral withdrawal latency of 14.6 ± 1.2 s, with a cutoff of 30 s to prevent potential tissue damage. Three averaged paw-withdrawal latency values were obtained from each animal with at least 1 m apart in between trials. Responses to mechanical, thermal, and chemical stimuli were tested in naïve, sham-pain and SNI rats. Sham and SNI rats were tested before surgery (baseline), 7 days post-injury (DPI) and after stable neuropathic hypersensitivity was established at 14 DPI.
2.7.2. Catalepsy
Cannabinoid impact on voluntary motor behavior was measured using the static bar test. The assay started when rats’ forelimbs were gently positioned on an elevated bar set to a height of 13.5 cm from the apparatus floor, which required animals to stand in an upright position. The latency for the animal to remove both paws from the bar was recorded manually by an experimenter.
2.7.3. Locomotion
For locomotion, anxiety-like behavior, and conditioned place preference (CPP) tests, rats were placed in a three-compartment apparatus containing infrared sensors placed 4 cm above the floor (Med Associates, Fairfax, VT). The left and right compartments were of equal size (21 × 25.5 × 21 cm) but had different colored walls (black and white), connected by a middle gray-walled compartment (21 × 12 × 21 cm). The left compartment had a grid floor, the right compartment had a mesh floor, and the middle compartment had a smooth floor with two sliding doors on each side that allowed isolation of the compartments. Spontaneous locomotion was measured in the three-compartment apparatus. Each rat was placed in the chamber and the horizontal activity (total number of beam breaks) was assessed for 5 min.
2.7.4. Light Avoidance Test
To evaluate anxiety-like behavior the animals were placed in the gray-colored middle chamber at the beginning of the experiment. The black-walled chamber was blocked from any light using a black laminated cover, while the white-walled chamber was illuminated with a bright light (3000 K). Both the time spent and distance traveled (beam breaks) in each compartment were measured for a total of 5 min.
2.7.5. Conditioned place preference paradigm (CPP)
Rats were placed in the three-compartment apparatus, and movement and time spent in each compartment were recorded via beam breaks. To evaluate the baseline preference, animals were placed in the middle chamber and allowed to freely explore the entire apparatus for 15 min. Baseline preference (preconditioning CPP score) was evaluated for 3 days, the averaged value was obtained and used for later comparisons. On the fourth day, using an unbiased approach, the conditioning phase started by injecting either saline or morphine (5 mg/kg i.p.) and placing the rats in the randomly paired chamber with the door closed for 30 min. In all experiments, the drug-paired compartment was counterbalanced. The conditioning sessions were conducted at 08:00 and 16:00 for 5 consecutive days. To minimize any impact of opioid withdrawal on preference scores, in morphine-treated rats, the drug was administered only in the morning session for a total of 5 morphine and 5 saline injections. Control animals received a total of 10 saline injections. The post-conditioning test (Day 9) was performed 24 h after the last conditioning session. Animals were placed in the middle chamber without any injection and with the doors open to allow them freely move within the three chambers. The side preference was evaluated for 15 min. After the post-conditioning test, animals underwent CPP extinction in their home cage for 2 weeks. Once the side preference returned to baseline levels (Day 21), reinstatement was induced with 1 mg/kg i.p. of morphine and side preference was evaluated immediately after the injection. When evaluating the effect of vapor exposure in the expression and reinstatement of morphine-induced CPP, animals were acutely exposed to WPE or PGVG before the first post-conditioning session (Day 9) and before morphine-induced reinstatement injection (Day 21).
2.7.6. Novel Object Recognition Test (NOR)
Animals were habituated in a rectangular arena (55 cm × 35 cm × 31 cm) for 5 min without objects. On the next day, animals were acutely exposed to WPE or PGVG vapor and immediately after were placed in the arena to freely explore two identical objects for 5 min, after which rats were returned to their home cage. 30 min later, animals were placed back in the arena but one of the objects previously presented was replaced with a novel one. The rats were allowed to explore the novel and familiar objects for 5 min. All trials were video recorded and later analyzed by a blinded experimenter who quantified the exploratory behavior. Exploration was defined as the animals directing the snout toward the object at a distance ≤ 2 cm, sniffing, or touching it with the nose. The objects used in the test were 325-mL cylinders (15 cm in height) and a pyramid-shaped toy building block (13.5 cm in height). The use of each set of objects was counterbalanced, with each object equally used as novel or familiar. The location (right vs. left) of novel object was also counterbalanced.
2.7.7. Social Interaction
Immediately after vapor exposure two unfamiliar rodents matched in sex, weight, and vapor treatment were placed into an open arena (55 cm × 35 cm × 31 cm) for 10 min. Sessions were recorded and videos were later analyzed by a blinded experimenter who obtained the total (cumulative) time of interaction. Behaviors considered social interaction were: grooming, anogenital sniffing, running behind and crawling over/under the partner. All sessions were performed under red light and all animals were individually habituated to the arena the day before the experiment.
2.7.8. Fentanyl self-administration
Rats with neuropathic pain (SNI) and pain-sham animals were used to evaluate fentanyl reinforcing effects. Sham or SNI surgery was performed at least one week before the jugular catheterization. Rats were anesthetized with 2% isoflurane and implanted with chronic indwelling jugular vein catheters, as described previously [55]. One day before and immediately after the surgery, rats received carprofen (2 mg/tab) and enroflaxacin (2 mg/tab) tablets. To prevent infection and to assure optimal function, catheters were flushed daily with a mixture of cefazolin (10 mg/ml) and heparin (0.1 ml 100 U/mL) dissolved in sterile saline. Catheter patency was weekly tested with methohexital (2 mg/kg), only the rats showing the behavioral effects of the drug were included in the final analysis (n = 26). The animals were allowed to recover from surgery for one week after which they were trained in commercial operant conditioning boxes (Med Associates, Fairfax, VT) controlled by Med-PC IV software, and equipped with two retractable levers (active and inactive), two cue lights and one house light.
Rats underwent one or two training sessions in which each active lever press resulted in a fentanyl infusion (3 μg/kg/100 μL infusion over 6 s). During the training phase “fruit loops ” cereal pieces were used to shape the instrumental behavior acquisition. After training, animals self-administered fentanyl (2 μg/kg/100 μL infusion over 6 s) under FR1, FR3, and FR5 for 5 consecutive days each schedule. Every fentanyl infusion was paired with a cue light and was followed by the initiation of 20 s timeout period signaled by turning on the house light. All sessions lasted 2 h, after which animals were returned to their home cage.
After completion of FR5 sessions, following a Latin square design all animals were assigned to either a within-session dose-response or a PR session with previous exposure to WPE or PGVG. All animals were crossed over to each experimental condition: exposed to PGVG and WPE and tested in both behavioral assays, having only one vapor exposure and one behavioral task per day. For the within-session fentanyl dose response, animals were acutely exposed to PGVG or WPE and immediately after, placed in the operant boxes. One of the three tested fentanyl doses (1, 2 or 3 μg/kg/100 μL) was available for 30 min under an FR3 schedule, with a 10 min resting period between doses. The order of dose presentation was randomized in all experiments. For the PR session, animals were acutely exposed to PGVG or WPE and immediately after given access to fentanyl (2 μg/kg/100 μL) for 120 min. The number of responses needed to deliver a single infusion increased successively, as previously [56]. For reinstatement, only pain-sham animals were trained to self-administered fentanyl (2 μg/kg/100 μL), under FR1 and FR3, 5 days each. On the 11 th day, the extinction protocol started (1 h sessions/day for 7–10 days), in which lever presses had no consequences. Once un-reinforced active lever presses were reduced to 10% of FR3 responding for at least 5 days, the reinstatement was induced by fentanyl (3 μg/kg i.v.) administration. When evaluating the effects of vapor inhalation on fentanyl reinstatement, animals were acutely exposed to WPE or PGVG immediately before fentanyl availability.
2.8. Data Analysis
The same cohorts of animals were evaluated for (in order from first to last evaluation): body temperature, catalepsy, acetone allodynia, mechanical and thermal sensitivity, locomotion, and light avoidance. Values are expressed as°C, time in (s), paw withdrawal latency in (s) and number of beam breaks in 5 min, respectively. A two-way ANOVA analysis was used to compare vapor treatment effects.
Data from estrous cycle measures, NOR and CBD plasma levels after chronic exposure were collected from the same cohorts. NOR data is expressed as Discrimination Index (DI) which results from dividing the difference in exploration time for familiar object by the total amount of exploration of the novel and familiar objects [DI = (TN −TF) / (TN + TF)]. The values can vary between + 1 and −1, a positive score indicates more time exploring the novel object, a negative score indicates a preference to explore the familiar object, and a zero score indicates a null preference [57]. An independent cohort of animals was used for SI experiments and lung histology analysis. SI data is expressed as cumulative interaction time in (s) for a total of 10 min period. CBD plasma levels are expressed as ng/mL. In all cases, a two-way ANOVA analysis was used to compare vapor treatment effects. Lung histology data are expressed in m of MLI analysis that reflect the free distance between gas exchange surfaces in the acinar airway complex, which is measure of morphological changes [51]. The averaged value from three alternating sections per animal was obtained for each experimental condition. A two-way ANOVA analysis was used to determine differences between treatment groups.
A separate group of animals was used to assess the development of morphine tolerance and vapor treatment combination effects on mechanical sensitivity and locomotion. A two-way ANOVA analysis was used to compare morphine ED50s between groups.
An independent group of animals was used for CPP experiments. CPP score was calculated with the formula: CPP = the time spent in the drug-paired chamber/the total time spent in both chambers [58]. Data was analyzed using a two-factor repeated-measures ANOVA with treatment as the between-subjects factor and time as the within-subject factor, and a paired t-test was used to confirm CPP extinction.
Data pertaining to vapor self-administration were collected from an independent cohort and were expressed as number of active responses and number of vapor deliveries in 120 or 180 min session. Nose-poke discrimination index (DI) was calculated using the following formula DI = (active nose-pokes − inactive nose-pokes)/(active nose-pokes + inactive nose-pokes); where 0 reflects no discrimination between active and inactive nose-poke, 1 indicates complete preference for the active nose-poke, and −1 indicates complete preference for the inactive one [59]. A two-way ANOVA analysis was used to compare group differences. Vapor self-administration data were also analyzed by determining area under the curve (AUC) and, a t-test or one-way ANOVA was performed to further assess differences between groups in AUC values.
A separate cohort of animals was used for fentanyl i.v. self-administration acquisition, within-session dose response and PR experiments. A two-way ANOVA analysis was used to compare group differences with SNI vs. sham condition as the between-subjects factor and time as the within-subjects factor. For fentanyl dose response and PR sessions analysis, a three-way ANOVA was performed with vapor treatment and pain condition as the between-subjects factors and time as the within-subjects factor. The fentanyl extinction and reinstatement data were collected from a separate cohort of animals, a two-way ANOVA analysis was used to compare group differences, and t-test was used to confirm extinction.
When ANOVA analyses were performed, significant effects were further probed using Bonferroni post hoc tests. In all cases data are expressed as mean ± SEM, values of p ≤ 0.05 were considered statistically significant, and values of 0.05 < p < 0.1 were considered trending.
3. Results
3.1. Plasma levels of cannabinoids after acute WPE vapor exposure
Animals exposed to WPE in a single 30 min session (a total of 15 five-second vapor deliveries) had measurable plasma levels of the following cannabinoids and metabolites (in ng/mL): 9.8 ± 2.65 CBD, 1.9 ± 0.49 THC, 1.4 ± 0.24 11-OH-THC and 1.8 ± 0.26 11-COOH-THC. No cannabinoid analytes were detected in the plasma from animals acutely exposed to PGVG.
3.2. Cannabinoid tetrad effects of WPE vapor
Because cannabinoids are well known for inducing the classic tetrad behaviors [60], it was important to characterize the degree to which WPE produced these effects in pain-naïve animals. First, we evaluated the antinociceptive effects of a single acute (T1) and chronic (T40, twice-daily for 20 days) vapor exposure in response to different nocifensive stimuli (Figs. 1A–1C). Using the Hargreaves apparatus, we found no significant differences in the paw withdrawal latency induced by thermal stimulation between PGVG and WPE-treated animals either after T1 or T40 vapor exposure (treatment: F(1, 14) = 2.2, p = 0.16; time: F(2, 28) = 1.6, p = 0.21, Fig 1A).
WPE has mild antinociceptive effects and partially reduces neuropathic hypersensitivity in female rats. In pain-naïve animals, PGVG (gray bars) and WPE (green bars) produced similar nocifensive responses to thermal (A) and mechanical (C) stimulation after acute (T1) or chronic exposure (T40, on the 20th day). However, compared to PGVG, acute WPE vapor reduced cold allodynia (B). (G-I) Animals with spared nerve injury (SNI, red bars), showed the expected hyperalgesia two weeks (T14) after the SNI procedure compared to PGVG vapor (N = 8–10 per group, * p < 0.05, * * p < 0.01, * * * p < 0.001, Bonferroni post hoc). In SNI rats, acute WPE vapor inhalation (shaded bars) partially reversed acetone/cold allodynia (H), but did not affect the hypersensitivity produced by thermal (G) or mechanical stimulation (I). Schematic representation of Hargreaves (D), acetone-induced cold allodynia (E) and von Frey (F) assays.
WPE, but not PGVG, reduced acetone-induced allodynia at T1 (treatment: F(1, 14) = 6.5, p < 0.05. Fig. 1B). However, this anti-allodynic effect did not persist with chronic WPE exposure (p = 0.72). Similar to what we saw in thermal test, the Von Frey assessment revealed no differences in mechanical sensitivity between WPE and PGVG-exposed rats (treatment: F(1, 42) = 0.035 p = 0.85) at either T1 or T40 (time: F(2, 42) = 1.2, p = 0.31, Fig. 1C).
Because these behavioral paradigms are more sensitive to the reversal of hyperalgesia than antinociception, we next sought to evaluate whether acute WPE inhalation could reverse pain-induced sensory hypersensitivity in animals with neuropathic pain (SNI). As expected, fourteen days after nerve injury (DPI14), when SNI-induced changes in stimuli sensitivity were fully stabilized, we observed that SNI rats were more sensitive to nociceptive stimulation; that is, we observed shorter paw withdrawal latencies in response to thermal (F(1, 18) = 11.6, p = 0.006, Fig. 1G) and mechanical stimuli (F(1, 18) = 40, p < 0.001, Fig. 1I). SNI rats also had higher scores on the acetone-induced allodynia test compared to sham-pain animals (F(1, 18) = 48.4, p < 0.001) both at T1 and T14 (F(2, 42) = 23.4, p < 0.001; Fig. 1H).
In animals with neuropathic pain, a single acute exposure to WPE partially reversed the allodynia produced by acetone (p < 0.001, Fig. 1H). However, WPE vapor did not reverse the hypersensitivity produced by thermal (p > 0.9. Fig. 1G) or mechanical stimulation (p > 0.9; Fig. 1I).
Naïve rats exposed to PGVG or WPE exhibited similar locomotor behavior after acute and chronic vapor exposure (vapor: F(1, 14) = 0.0008, p = 0.98, time: F(2, 26) = 0.12, p = 0.86. Fig. 2A). The SNI procedure did not affect motor activity, as rats with neuropathic pain had a similar number of beam breaks compared to pain-sham animals (F(1, 18) = 0.08, p = 0.77; Fig. 2B). Acute exposure to WPE vapor did not produce any changes in locomotor activity of SNI or sham animals (p = 0.89; Fig. 2B).
WPE does not affect locomotion or induce hypothermia. (A) In pain-naïve rats, locomotion (number of beam breaks) was similar for PGVG-exposed (gray bars) and WPE-exposed rats (green bars) after both acute (T1) and chronic (T40, 20 days later) vapor exposure. (B) Rats with neuropathic injury (SNI, red bars) had similar locomotor behavior compared to sham-pain animals (gray bars). WPE vapor inhalation (shaded bars) did not produce any changes in locomotor activity in either SNI or sham animals. (C) A transitory drop in body temperature was seen on the second day of WPE exposure but ANOVA revealed no difference between vapor treatments. N = 8–10 per group. Data are expressed as mean ± SEM.
Because WPE contains 7% THC and may produce hypothermia, we registered body temperature over the course of chronic WPE or PGVG administration (Fig. 2C). Repeated-measures ANOVA revealed no significant differences between PGVG and WPE-exposed animals (F(1, 14) = 0.025, p = 0.88). Although there was no overall difference, there was a transient drop in body temperature on the second day of WPE exposure (p = 0.76), which we attribute to the low percentage of THC present in high-CBD WPE. By the 4 th day of vapor exposure, both groups had similar body temperatures. This data is consistent with previous reports of tolerance to THC’s hypothermic effects (Fig. 2C) [61].
To complete the cannabinoid tetrad assessment, catalepsy was assessed using the static bar test after acute and chronic WPE or PGVG vapor exposure. Regardless the vapor treatment, we did not observe catalepsy in any animals (data not shown).
3.3. Psychoactivity of WPE vapor
Δ9-THC is well-known for its ability to produce reward, cognitive impairment and also induces a biphasic anxiogenic effect [62]. In contrast, the administration of isolated CBD has anxiolytic effects in male rats. However, these effects have not previously been examined using the inhalation route of administration. Considering that CBD’s effects could be modulated by other plant-derived molecules present in WPE [63–66], particularly the low percentage of THC, it was critical to thoroughly evaluate the behavioral effects of WPE vapor.
When evaluating cognitive function, we found no differences between PGVG and WPE in rats’ performance in the NOR test (F(1, 14) = 1.3, p = 0.28), either after acute or chronic vapor exposure (F(1, 14) = 0.13, p = 0.73; Fig. 3A). All animals spent more time exploring the novel object than the familiar one, reflected by the positive values in the discrimination index, suggesting that vapor treatment did not affect working memory (Fig. 3A). Similarly, our analysis revealed no significant differences between WPE and PGVG in spontaneous social behaviors (F(1, 14) = 0.94, p = 0.35); however, both groups exhibited significantly reduced social interactions by the end of the chronic treatment time frame (F(1, 14) = 5, p = 0.04; Fig. 3B).
WPE does not produce cognitive impairment or anxiolytic behavior. (A) Cognitive function, expressed as the discrimination index on the novel object recognition test, was similar in PGVG-(gray bars) and WPE-exposed rats (green bars) after both acute (T1) or chronic treatment (T40, on the 20th day). (B) Cognitive function, expressed as the cumulative interaction time on the social interaction test was similar between PGVG- and WPE-treated groups at T1 and T40. (C) Anxiety like behavior, as measured by chamber exploration time in the light avoidance test: all animals spent more time in the dark compartment versus the illuminated chamber, without differences between vapor treatments. N = 8–10 per group. Data are expressed as mean ± SEM.
When evaluating anxiety-like behavior, as expected, all animals spent more time in the dark compartment versus the illuminated chamber (F(1, 28) = 54.7, p < 0.0001; Fig. 3C). We found no differences between vapor treatments (F(1, 28) = 1.28, p = 0.28. Fig. 3C) after acute or chronic exposure (F(1, 28) = 0.97, p = 0.3, Fig. 3C) in the light avoidance test.
In vapor self-administration experiments, rats responding for PGVG had a significantly higher number of nose-pokes compared to WPE-responding rats (F(3, 28) = 17, p < 0.0001). As shown in Fig. 4A, the PGVG group had 63 ± 11 instrumental responses while WPE-exposed animals had 15 ± 3 nose-pokes during the FR1. An overall one-way ANOVA of AUC values from the number of responses on the active nose-poke further confirmed significant differences between PGVG and WPE (F(3, 28) = 5.5, p = 0.004). Accordingly, PGVG animals received more vapor deliveries than WPE rats (F(1, 14) = 6.1, p = 0.027; Fig. 4B). Similarly, during the PR session, PGVG animals showed a trend toward a higher number of nose-pokes than the WPE vapor group (F(1, 14) = 3.2, p = 0.09; Fig. 4C). The analysis of total number of responses between WPE (89 ± 18) and PGVG (195 ± 55) groups further confirm this statistical trend (t14 = 1.8, p = 0.09. Fig. 4D). Moreover, the PGVG group quickly discriminated between the active nose-poke vs. the inactive one, compared to WPE rats (F(1, 14) = 1.7, p = 0.21), which could be interpreted as a greater motivation to self-administer PGVG vapor (Fig. 4E).
WPE vapor is less rewarding than PGVG vapor in an operant vapor self-administration. PGVG-exposed animals (gray symbols) had higher number of nose-pokes (A) and vapor deliveries (B) compared to WPE-treated rats (green symbols) during self-administration training. (C, D) Analysis of instrumental responses during the PR session revealed higher motivation to self-administer PGVG than WPE. (E) PGVG-treated rats were faster to discriminate between active vs. inactive nose-poke, compared to WPE-exposed animals. N = 8 per group, * p < 0.05, * * p < 0.01, * * * p < 0.001 vs. WPE group, Bonferroni post hoc. Data are expressed as mean ± SEM.
3.4. Effects of chronic vapor inhalation on lung health and estrous cycle
Chronic cannabinoid exposure has been shown to have negative impacts on female endocrine physiology [49]. Furthermore, the extent to which chronic cannabis vapor exposure contributes to lung injury has yet to be defined in awake/behaving animals [67]. Thus, it was important to assess whether chronic WPE exposure alters physiological variables including estrous cycle and lung cytoarchitecture. As shown in Fig. 5, histology analysis revealed that compared to the air-exposed controls (MLI = 5.12) exposure to both PGVG and WPE induces maladaptive changes in lung morphology, as measured by MLI (F(2, 30) = 25.6, p < 0.001). That is, both vapor conditions increased the mean distance between gas exchange surfaces in the acinar airway complex There were no differences in lung morphometrics of rats chronically exposed to PGVG vs WPE (mean MLI = 10.22 and 10.12, respectively). Measuring daily vaginal impedance over the course of chronic vapor treatment, we found no differences in the amount of time animals spent in the proestrus phase between rats exposed to PGVG and WPE (t14 = 0.28, p = 0.8).
WPE and PGVG induce similar lung adaptations. (A) Quantification of lung histology using mean linear intercept. Compared to air-exposed controls (open bar, n = 3, mean MLI = 5.12) PGVG (gray bar) and WPE (green bar) induce maladaptive changes in lung morphology (F(2, 30) = 25.6, p < 0.001). There were no differences in lung morphometrics of rats chronically exposed to PGVG vs WPE (n = 8, mean MLI = 10.22 and n = 7, mean MLI = 10.12, respectively). (B – D). Representative lung slices (20 μm, stained with hematoxylin and eosin) from rats exposed to air (B), PGVG (C) and WPE (D).
3.5. Nociceptive interactions between morphine and WPE
Previous studies from our research group have demonstrated that co-administration of cannabinoids potentiates the antinociceptive effects of morphine and mitigates the development of morphine tolerance [36, 68]. However, the ability of inhaled WPE vapor to modulate morphine-induced antinociception and tolerance has not previously been assessed. Morphine (10 mg/kg s.c.) or saline was administered to rats twice daily for two days. After every morphine or saline injection, rats were exposed to PGVG or WPE, for a total of four 30 m vapor exposures. The behavioral assays were conducted immediately before the first injection (baseline) and within an hour after first (T1) and fourth (T4) vapor exposure.
At T1, regardless the vapor treatment, all morphine-injected rats showed a two-fold increase in paw withdrawal latency on the Von Frey test compared to saline-injected groups (F(3, 30) = 13.08, p < 0.001). Morphine antinociception was similar in animals exposed to PGVG and WPE vapor (p = 0.9467). As expected, at T4, we observed antinociceptive tolerance in animals pre-treated with morphine (F(2, 28) = 34, p < 0.01 vs. T1, Fig. 6A). WPE vapor had no impact on the development of morphine tolerance, compared to PGVG-exposed rats (p = 0.8395; Fig. 6A).
Analysis of WPE vapor inhalation on morphine-induced analgesia, sedation and tolerance. (A) WPE vapor inhalation (green bars) did not affect acute morphine analgesia after first (T1) morphine injection (shaded bars), nor did it attenuate the normal development of morphine tolerance, as assessed on the Von Frey test after the fourth injection + vapor exposure (T4). (B) Morphine administration (shaded bars) produced the expected reduction in locomotor behavior compared to saline-treated animals: this effect was similar between PGVG (gray bars) and WPE vapor treatments at T1 and T4. (C) Morphine tolerance was assessed on the next day (Day 3), using a cumulative dose-response analysis. Saline-pretreated animals (gray symbols) showed the typical analgesic dose response to morphine administration, while rats pretreated with morphine on the previous two days exhibited a marked development of antinociceptive tolerance. WPE vapor treatments had no effect on the development of morphine tolerance. N = 8–10 per group, * p < 0.05, * * p < 0.01, * * * p < 0.001, Bonferroni post hoc. Data are expressed as mean ± SEM.
In the locomotion assessment, morphine-treated rats showed the expected reduction in locomotor behavior compared to saline-treated animals (F(3, 30) = 16, p < 0.001, Fig. 6B). Compared to PGVG, WPE vapor had no impact on morphine-induced sedation after the first injection (T1, p = 0.9). Regardless of vapor treatment, tolerance did not develop to the sedative effects of morphine, as measured at T4 (p = 0.1 vs. T1), and morphine-induced sedation was similar in WPE- and PGVG-exposed animals (p = 0.1; Fig. 6B).
On Day 3 (no vapor exposure), morphine tolerance was assessed on the Von Frey apparatus using a cumulative dose-response analysis (1.8, 3.2, 5.6, 10 and 18 mg/kg, as previously reported) [68]. Animals did not exhibit any signs of withdrawal before starting the cumulative dose-response. In saline-pretreated animals, exposure to WPE vapor did not impact the morphine ED50, compared to PGVG-exposed animals (F(1, 102) = 1.65, p = 0.20). As expected, rats which had been pretreated with morphine on the previous two days exhibited a marked development of antinociceptive tolerance, demonstrated by a right-ward shift in the dose-response curve compared to saline-pretreated animals (Fig. 6C). The morphine ED50’s were identical for WPE and PGVG-exposed animals that had been pretreated with morphine (18 mg/kg, Table 1). Because the largest dose of morphine we used in the cumulative dosing experiment was 18 mg/kg, the reported morphine ED50s likely underestimate the actual degree of tolerance in morphine-pretreated animals (a ceiling effect was observed).
Table 1
Morphine ED50 values obtained from the Von Frey test.
| Pretreatment | Morphine ED50 on Day 3 (mg/kg) |
|---|---|
| Saline+PGVG | 7.12 |
| Saline+WPE | 9.60 |
| Morphine+PGVG | 18.0 |
| Morphine+WPE | 18.04 |
3.6. WPE vapor modulation of opioid reward
Both clinical and preclinical studies have demonstrated that CBD reduces opioid-seeking behavior, cue-induced anxiety and craving, and opioid withdrawal symptoms [34, 39, 40, 69]. However, most of these studies were in males, and none have examined the inhalation route of administration. Using two different approaches, we next tested the ability of WPE vapor to reduce opioid reward.
First, we evaluated whether acute WPE inhalation was able to reduce the expression of morphine-induced CPP in pain-naïve rats. As shown in Fig. 7, acute exposure to WPE was able to prevent the expression of morphine-induced CPP compared to saline controls (F(3, 28) = 1.8, p = 0.1; p > 0.9). In contrast, PGVG pre-treated animals showed the typical increase in side preference produced by morphine (p = 0.04 vs. saline, Fig. 7). Morphine CPP extinction was evident 14 days after the conditioning phase, when the CPP score from morphine-treated animals returned to baseline values (F(3, 84) = 3.8, p > 0.9). In rats pre-exposed to PGVG, the administration of 1 mg/kg of morphine completely produced CPP reinstatement (p = 0.05 vs. saline); however, in rats pre-exposed to WPE, 1 mg/kg morphine was not able to reinstate the drug-seeking behavior (p > 0.9 vs. saline; Fig. 7).
WPE inhalation reduces morphine-induced CPP. Acute exposure to WPE (green bars) was able to prevent the expression of morphine-induced CPP, while PGVG pre-treated animals (gray bars) showed the typical increase in side preference produced by morphine (shaded bars), as measured 24 h after the last condition session. Following 14 days of extinction, PGVG-exposed animals returned to baseline preferences. On the 15th day (far right), morphine-induced reinstatement of preference was assessed: WPE, but not PGVG vapor, prevented the reinstatement induced by 1 mg/kg of morphine. N = 8 per group * p < 0.05, Bonferroni post hoc. Data are expressed as mean ± SEM.
According to clinical data, pain is an important factor associated with the development of addiction to prescribed opioids [70, 71]. Our previous studies have also demonstrated that persistent pain promotes increased opioid self-administration in male rats [72]. Because cannabinoids could be used as an adjunct treatment to reduce opioid abuse liability [73], it was critical for us to evaluate the ability of WPE to reduce opioid self-administration behavior in the context of chronic neuropathic pain.
First, using fentanyl (2 μg/kg i.v.), we assessed the acquisition of self-administration behavior in SNI and pain-sham animals. The pain-sham group progressively increased the number of active lever presses; however, SNI animals did not show any significant differences in the number of active responses over the course of various reinforcement schedules (F(3, 38) = 3.6, p = 0.02; Fig. 8A). Unlike our previous studies with heroin in male rats with inflammatory pain [72], pain-sham female rats consistently received more fentanyl infusions than animals with chronic neuropathic pain (F(1, 19) = 5, p = 0.04; Fig. 8B).
WPE vapor reduces fentanyl self-administration. Sham-pain animals (gray symbols) readily acquire fentanyl-seeking behavior, while SNI rats (red symbols) maintained a lower number of lever presses (A) and drug infusions (B) across different schedules of reinforcement. (C, D) In a within-session dose-response analysis, the dose of fentanyl (1, 2, and 3 μg/kg, in randomized order), did not significantly impact the number of responses or infusions, regardless of pain condition. However, WPE vapor pre-treatment (green shaded bars) reduced fentanyl intake in both sham and SNI groups. (E, F) Similarly, on a progressive ratio schedule of reinforcement, sham animals showed higher motivation to self-administer fentanyl compared to rats with neuropathic pain: WPE treatment did not impact fentanyl motivation in either sham-pain or SNI animals. N = 8–10 per group. Data are expressed as mean ± SEM.
Next, we examined the ability of WPE to reduce drug-seeking behavior in rats with neuropathic pain. Using a within-session dose-response experimental design (1, 2, and 3 μg/kg, in randomized order) we analyzed differences in fentanyl intake between pain-sham and SNI rats that were pre-exposed to WPE or control vapor (Fig. 8C–8D). Similar to our fixed-ratio acquisition experiments, sham animals had greater fentanyl intake compared to SNI rats during the 90-minute dose-response session, which was reflected by higher number of lever presses (F(1, 54) = 11.78, p < 0.001; Fig. 8C) and infusions (F(1, 54) = 6.90, p = 0.01; Fig. 8D), regardless of the dose. For both the pain-sham and SNI animals, there was no dose-dependent relationship between the dose of fentanyl and the number of lever presses for drug infusions (F(2, 54) = 0.84, p = 0.44; Fig. 8D). However, pretreatment with WPE vapor reduced fentanyl intake in both sham and SNI groups, as measured by the number of lever presses (F(1, 54) = 15.83, p < 0.001 vs. PGVG pre-treatment; Fig. 8C).
Next, we assessed the impact of WPE on rats’ motivation to seek fentanyl using a PR schedule. Similar to our fixed-ratio results, sham animals showed higher motivation to self-administer fentanyl compared to rats with neuropathic pain (F(1, 52) = 12, p < 0.001; Fig. 8E–8F). We found no differences between PGVG and WPE treatments in the performance of the task neither in pain-naïve or SNI groups (p = 0.6, Fig. 8E–8F).
Because clinical studies have shown that oral CBD reduces opioid craving [34], we next tested the hypothesis that inhaled high-CBD WPE could reduce reinstatement to fentanyl-seeking behavior (a model of drug-cue induced relapse). We tested this hypothesis only in sham-pain animals, because rats with neuropathic pain did not express sufficient drug-seeking behavior (Fig. 8B). After self-administration acquisition under FR1 and FR3, animals underwent extinction (Fig. 9A), and drug-induced reinstatement was later evaluated (Fig. 9B). Using a within-subjects crossover experimental design, we observed that although 3 μg/kg i.v. fentanyl partially restored drug-seeking behavior (t8 = 2, p = 0.08; Fig. 9B), a single 30 m exposure to WPE or PGVG vapor had no impact on the number of active lever presses (F(2, 12) = 0.1, p = 0.9; Fig. 9B).
WPE does not affect fentanyl-induced reinstatement of drug-seeking behavior. Sham-pain animals were trained to self-administer fentanyl (2 μg/kg) under FR1 and FR3 reinforcement schedules (A), after which extinction was induced (A, B). Although 3 μg/kg i.v. fentanyl only partially restored drug-seeking behavior (t8 = 2, p = 0.08, 5th day of extinction compared to air/no vapor + 3 μg/kg i.v. fentanyl), neither PGVG (gray bars) or WPE vapor pre-treatment affected fentanyl-seeking behavior. N = 5 per group. * p < 0.05, * * * p < 0.001, Bonferroni post hoc. Data are expressed as mean ± SEM.
4. Discussion
To our knowledge, this is the first comprehensive study of the behavioral and physiological effects of inhaled vapor from high-CBD whole-plant cannabis (hemp) extract. This work also fills in critical gaps about the effects of cannabinoids in females with neuropathic pain, which have been critically understudied. We found that 30 minutes of vapor exposure to WPE was sufficient to produce physiologically-relevant levels of cannabinoids and their metabolites in the rats’ plasma. In the tetrad assessment, we observed no locomotor or cataleptic effects, and a mild-but-transient hypothermic effect of WPE. Although WPE did not produce antinociception in pain-naïve animals, it was modestly efficacious in reversing neuropathy-induced cold allodynia. Furthermore, we observed that inhaled WPE did not induce cognitive impairment, social withdrawal or anxiolytic effects. Interestingly, rats self-administered control (PGVG) vapor much more readily than WPE, suggesting that WPE is less rewarding than the standard control excipient used in the vast majority of vapor inhalation studies [59, 74, 75].
In female rats chronically exposed to WPE, we found no detrimental effects on lung histology or estrous cycle. WPE vapor did not enhance the acute antinociceptive effects of morphine, nor did it modulate the normal development of morphine tolerance. Despite the lack of effect on morphine antinociception and tolerance, WPE successfully reduced the rewarding effects of morphine, as measured on the CPP paradigm. Similarly, in i.v. opioid self-administration experiments, we found that WPE vapor significantly reduced opioid seeking behavior in both SNI and pain-sham rats. Unlike our previous studies in male rats with inflammatory pain [72], the presence of neuropathic pain in females reduced opioid-seeking behavior. Finally, WPE had no impact on either the motivation to engage in drug-seeking behavior, or the reinstatement of drug-seeking behavior after extinction.
The WPE used in this study contained tenfold more CBD than THC. Thirty minutes of high-CBD WPE inhalation resulted in physiologically-relevant levels of CBD in rats’ plasma, which is consistent with a previous study describing a concentration-dependent increase in CBD plasma levels after isolated CBD vapor inhalation in female rats [76]. Hemp-derived products often contain low percentages of THC (typically ≤ 0.3%) [77]. Thus, we found measurable levels of THC and its metabolites in the plasma of rats acutely exposed to WPE vapor. Because CBD inhibits several members of the cytochrome P450 family [78, 79], 77], the inhibition of THC metabolism may enhance THC plasma levels after CBD administration [58, 78]. In fact, exposure to high-CBD WPE vapor may produce an increase in THC concentrations in brain tissue [59]. Thus, some of our observations may have been mediated by the interaction between CBD and THC. For example, we observed a transient drop in body temperature on the second day of vapor exposure. Similar results were described before on female rats after synthetic THC vapor exposure [61]. Indeed, the time course we observed for body temperature to return to baseline is consistent with the development of hypothermic tolerance to THC, as described by others [61, 80]. Thus, the transient drop in body temperature could be directly related to the low percentage of THC in WPE, the interactions between THC and other compounds in WPE [18], or both. Overall, we found no significant differences in body temperature between PGVG- and WPE-exposed animals after chronic vapor treatment, which is consistent with the lack of hypothermic effects of systemic CBD injection in rats [81, 82] and mice [83, 84].
Because hemp-derived products (high-CBD and low-THC) are widely available and poorly regulated, it was critical to characterize the safety profile of inhaled WPE, particularly in light of the recent increase in electronic cigarette and vaping-related lung injuries [85]. Although lung morphology was similar in animals chronically treated with either PGVG or WPE, both vapor conditions produced lung abnormalities compared to the air-exposed controls. This data is consistent with previous clinical reports showing that chronic exposure to vaporized products may pose pulmonary risks [86]. The frequent use of THC-containing vape devices has also been linked with lung injuries [87]. Further research is required to confirm the mechanisms by which WPE may pose pulmonary risks; that is, whether low-dose THC, aerosolized toxins, or particulate matter are implicated [47].
Continuing with the analysis of WPE safety profile, we found that high-CBD WPE did not produce catalepsy, nor did it affect locomotion. Similar findings were found after i.p. CBD administration, and in fact, CBD appears to attenuate catalepsy induced by dopaminergic compounds through a partial activation of 5-HT1A receptors [88]. Some previous studies have described a reduction in locomotor activity after oral or vaporized CBD both in male and female rats [80, 89]. The discrepancy between our findings and previous results could be partially explained by CBD’s interaction with other molecules (including THC) in the phytochemically-rich WPE vs. isolated compounds used in other studies [90, 91].
Compared with the control treatment, WPE vapor exposure did not affect memory function, social interaction, or anxiety-like behavior. Several reports have suggested that CBD reduces social withdrawal, has anxiolytic properties, and improves memory impairment in different preclinical models of disease [92–96]. However, there are conflicting findings in healthy control animals that receive CBD which demonstrate that, similar to our results in pain-naïve animals, CBD has no impact on social withdrawal or anxiety-like behavior [93–96]. Although THC is well known to produce cognitive impairment, the low percentage of THC in WPE was insufficient to impact rats’ performance in this study. Future studies in our laboratory will evaluate whether WPE reduces social withdrawal or anxiety-like behavior in animals with chronic neuropathic pain, and additional studies analyzing-vaporized CBD effects in models of other pathological states are warranted.
The endocannabinoid system (ECS) is critically involved in female reproduction, and there are complex interactions between the ECS and the hypothalamic-pituitary-ovarian axis (reviewed in [97]). Chronic exposure to cannabinoids has previously been shown to disrupt the menstrual cycle and other reproductive functions [49, 98]. Because chronic use of high-CBD WPE in female pain patients could potentially interfere with reproduction, it was critical to examine the impact of WPE vapor on the estrous cycle. Although we found that 20 days of twice-daily WPE vapor exposure had no impact on the amount of time female rats spent in proestrus, follow up studies are needed to confirm the reproductive safety of high-CBD WPE, particularly over longer periods of time and higher exposure frequency.
Although humans readily self-administer various ratios of CBD:THC, using an operant vapor self-administration approach we found that WPE is not rewarding in female rats, which is consistent with previous reports describing the inability of CBD to produce reinforcing effects in males across various paradigms [59, 99, 100]. Historically in preclinical research, the self-administration of THC has been particularly difficult to establish compared to other drugs of abuse (e.g. opioids and psychostimulants), which could be explained by THC’s mild rewarding effects at low doses, and aversive effects at high doses [101]. Although one previous study demonstrated that the vaporized high-THC WPE produced stable self-administration in male rats [59], it is likely that in the current work, the low percentage of vaporized THC was insufficient to elicit self-administration behavior. This effect could also potentially be mediated by the interaction between low-dose THC and CBD, however studies conflict as to whether CBD can antagonize or potentiate the effects of THC [65, 66, 91, 102]. Antagonism and potentiation between THC and CBD are very likely to be administration route, sex, and concentration-dependent. Furthermore, given the difficulty in reverse-translation of cannabinoid self-administration in rodents, there may be more suitable models for future studies which fully characterize CBD:THC reward modulation.
Mixtures of PG and VG are the most commonly used control excipients in vapor administration studies, but their effects on reward are poorly understood [59, 74, 75]. Our data showed that, in female rats, PGVG produced significantly higher values of drug-seeking and drug-taking behavior compared to WPE. In our experiments, the PGVG-seeking behavior was evident only during FR-1 training, while the number of instrumental responses during FR-2 and FR-4 schedules was similar between PGVG and WPE groups, which may reflect mild reinforcing effects of PGVG. Although humans are less likely to self-administer PGVG compared to flavored e-liquids [103], It is plausible that PG is reinforcing in rats because it tastes sweet [104]. In comparison, the taste and smell of WPE could be aversive to rats, however this is not supported by our observations. First, although responding for WPE on the first two days of FR-1 was lower than the inactive PGVG port, this transient difference failed to reach significance. Second, responding on the active WPE port was significantly higher than responding on the inactive WPE port across the experiment (FR-1 through FR-4, Days 1 – 20). Finally, there was a mild but steady increase in WPE responses during the PR session. Taken together, these results strongly suggest that WPE is not aversive, and that it is less rewarding than PGVG. Thus, results from previous operant and addiction-adjacent studies using PGVG as a vehicle should be interpreted with caution, given that rodents reliably self-administer PGVG [105], which has previously been regarded as innocuous. Although commercially-available cannabis vape devices contain numerous other additives which could potentially serve as neutral control diluents in preclinical studies (such as phytol and coconut-derived medium-chain triglyceride oil), these vehicles are likely to be problematic given their negative impacts on pulmonary function [67, 106–109]. The current results suggest that for some cannabinoid vapor experiments (such as high-THC administration), where reward and reinforcement are important experimental endpoints, hemp-derived WPE may be a more suitable and translationally-relevant control excipient than PGVG.
PGVG is an extremely common diluent in nicotine and cannabis eCIG devices [110]. Although some studies suggest that PGVG does not induce toxicity in the liver or lungs after chronic vapor inhalation in rats [111], others have shown robust inflammatory and immune responses, apoptotic neurodegeneration, and microglial activation after PGVG inhalation [112–115]. Differences in the experimental delivery of PGVG may partially explain these conflicting reports, given that some studies rely upon nebulized (compressed air or aerosol-like conversion methods) [110], rather than the more translationally-relevant heated and vaporized PGVG. Although most nicotine eCIG devices are limited to a peak operating temperature of 215°C [115], commercially-available cannabis eCIG devices often reach temperatures in excess of 450°C [116]. When exposed to high temperatures (> 200°C) PGVG produces carbonyl, acetaldehyde, acrolein, and formaldehyde [117, 118]. Thus, cannabis eCIG users are at a much higher risk for toxicant exposure from additives and excipients than nicotine eCIG users, because they are exposed to higher device operating temperatures. However, the cytotoxic effects of cannabis eCIG devices remains severely understudied.
Because chronic pain patients report using cannabis 2–3 times daily, every day [31], it was critical to evaluate the effects of therapeutically-relevant WPE exposure on pulmonary health. To our knowledge, this is the first study to examine the lung cytotoxicity of clinically-relevant WPE exposure in mammals. Our results revealed no differences in lung cytoarchitecture between PGVG and WPE after 20 days of twice-daily exposure; however, compared to the negative control group (exposed to air), both groups chronically exposed to vapor showed greater alveoli space, which may be indicative of lung injury. Although we experimentally optimized our vapor delivery apparatus to minimize the total particulate matter animals were exposed to, these safety measures are not uniform or guaranteed across commercially-available cannabis eCIG devices [116, 119]. Additional studies examining high device power, high temperature, and high frequency vapor delivery are extremely warranted, as are studies of other cannabis diluents such as medium chain triglycerides and plant-derived terpenes, which are known to produce toxicants upon heating and vaporization [120]. Nonetheless, the combined results of our hypothermic, locomotive, cognitive, operant, and histological analyses suggest that inhaled WPE exhibits a promising safety profile.
Because the most common use of cannabis-derived products is the management of chronic pain, we next examined the therapeutic capacity of WPE in the context of neuropathic pain. In pain-naïve animals, WPE vapor inhalation did not reduce nocifensive behaviors induced by thermal or mechanical stimulation, however it did reduce acetone-induced cold allodynia. These results are consistent with previous reports describing a lack of antinociceptive effects after oral, i.p. or vaporized CBD administration [76, 89, 121]. Thus, the mild-anti-allodynia we observed may have been mediated by low-dose THC. In our experiments, the anti-allodynic effect of WPE did not persist after chronic vapor exposure, which is similar to previous studies, suggesting the development of tolerance to THC [61]. The mechanisms through which CBD produces analgesia are not completely understood, but it has been proposed that the activation of TRPV1 and GPR55 receptors and serotoninergic modulation are involved [122, 123]. Previous studies of repeated oral and subcutaneous administration, but not vaporized CBD have shown to reduce allodynia in neuropathic pain models by enhancing 5-HT neurotransmission [111, 112]. It is also possible that the anti-allodynia effect we saw in pain-naïve rats was mediated by an interaction between CBD and THC. Indeed, the combination of CBD and THC has been shown to produce synergistic anti-allodynia and anti-hyperalgesia in mice with neuropathic pain [91, 124]. Further investigation is needed to determine whether this interaction between THC and CBD is dose, species, and/or pain model-dependent.
The results of this study suggest that the effects of WPE vapor in pain-naïve animals are mild, given its inability to reverse nocifensive responses in two out of three behavioral assessments. Similarly, in a study of healthy humans, CBD failed to impact pain threshold or pain tolerance, and in fact it increased the subjective painfulness of the experimental stimulus [113]. However, female chronic pain patients frequently report the analgesic efficacy of high-CBD WPE [23], which suggests that CBD may differentially modulate pain in typical people vs. those experiencing chronic pain. Thus, we replicated the antinociceptive battery in animals with neuropathic pain. As expected, the SNI procedure induced long-lasting hyperalgesia in response to mechanical, thermal, and cold stimulation relative to pain-sham rats. Although WPE vapor partially reversed acetone-induced cold allodynia, it had no effect on thermal or mechanical hyperalgesia. These results are similar to previous findings which demonstrate the anti-allodynic efficacy of oral cannabinoids during neuropathic pain [91], however they conflict with other reports where mechanical and thermal hyperalgesia were effectively reversed by systemically-injected CBD [125]. Our mild anti-allodynic results are not entirely surprising, given that the analgesic effects of vaporized cannabis in humans are likely mediated by THC, rather than CBD [126, 127]. Follow-up vapor administration studies in our laboratory will assess the concentration-dependent antinociceptive effects of THC in female rats.
Although opioids are the most effective drugs to relieve acute and severe pain, their efficacy for chronic pain is questionable [128]. The ineffective use of opioids in women, who experience a higher prevalence of chronic pain and are more vulnerable to opioid abuse than men [44], is particularly problematic given the alarming rise in opioid overdose deaths in recent years [129]. Thus, it is particularly important to develop novel therapeutic approaches which effectively reduce both pain and opioid risks in females. A growing body of evidence suggests that opioids and cannabinoids interact in several clinically-beneficial ways. Our previous preclinical studies demonstrate that co-administered cannabinoids potentiate antinociception and attenuate opioid tolerance, and our clinical results demonstrate the opioid-sparing effects of cannabis in chronic orthopedic pain patients [31, 32, 36]. These phenomena are partially mediated by CB1 receptor activation in the descending pain pathway [68]. Together, these data suggest that cannabis could serve as either an adjunct or alternative to opioids, however the efficacy of high-CBD WPE vapor in these effects had not been assessed prior to the current study. We found that unlike isolated cannabinoid receptor agonist administration [36], WPE inhalation did not enhance acute morphine antinociception, nor did it attenuate the development of morphine tolerance. It is well known that the predominant mechanism for THC’s analgesic effects is the activation of CB1 receptors [130]. Because our experiments relied on WPE with tenfold more CBD than THC, it is plausible that potentiated antinociception and attenuated tolerance may require a higher concentrations of THC. Future studies in our laboratory will directly test the hypothesis that high-THC vapor potentiates morphine antinociception by activating CB1 receptors in the descending pain pathway. Similar to our self-administration findings, the inability of low-dose THC to prevent morphine tolerance could potentially be mediated by the interaction between THC and CBD (that is, CBD may have antagonized the CB1 receptor, as in [66]). However, studies conflict as to whether CBD can antagonize or potentiate the effects of THC [65, 66, 91, 102]. Antagonism and potentiation between THC and CBD are very likely to be administration route, sex, and concentration-dependent.
Although we observed only mild anti-allodynic effects of WPE, and no potentiation of morphine antinociception, previous studies suggest that high-CBD WPE may still have a role in harm reduction by reducing opioid reward and craving [34, 39]. In the current study, WPE successfully reduced the rewarding effects of morphine. Our CPP data are aligned with previous reports in mice, where systemic injection of CBD was shown to attenuate morphine CPP [122]. Because WPE did not disrupt cognitive function as measured by the NOR test, it is likely that the reward-modulating effects of WPE were genuine (that is, rats were able to successfully establish morphine context associations). The morphine-reward modulating effects of WPE may be at least partially mediated by CBD’s activation of 5-HT1A receptors in the dorsal raphe [100]. In i.v. self-administration experiments, we found that inhaled WPE vapor reduced fentanyl intake in both SNI and sham-pain rats during the dose response session. Similar findings have been reported for male animals trained to self-administer heroin, and both AMPA GluR1 and CB1 receptors in the nucleus accumbens likely underlie the robust effects that CBD has on opioid seeking behavior [39]. Combined, these results suggest that CBD may have great therapeutic potential in opioid use disorder by reducing drug craving and relapse.
Our previous studies in male rats with inflammatory pain demonstrate that the presence of pain enhances motivation and intake for high, but not low doses of heroin [72]. However, in the current study, the presence of neuropathic pain reduced overall opioid-seeking behavior in females, compared to pain-sham animals. The discrepancies between these results could either be sex-dependent, pain model-dependent, or may be due to pharmacokinetic differences between heroin and fentanyl. Additional studies are required to fully characterize the potential of pain to modulate the abuse liability of opioids.
In summary, our results suggest that inhaled high-CBD WPE has modest anti-allodynic benefits, supporting the hypothesis that THC is the primary analgesic component of inhaled cannabis. However, the ability of WPE to reduce opioid reward and drug seeking behavior appears quite robust and of great clinical utility. Because of its promising safety profile and the absence of reinforcing effects compared to the standard excipient used in most vapor administration research, WPE may also be a more suitable and clinically-relevant control excipient for future vapor administration studies. Additional systematic research is required to fully evaluate the potential for CBD to serve as an adjunct treatment for opioid use disorder.
Acknowledgements
We would like to acknowledge Jiries Meehan-Atrash and Robert Strongin for HPLC analysis of CBD and total particulate matter in vapor. We would like to thank Ms. Lisa Bleyle for the cannabinoid analysis that was conducted in the Bioanalytical Shared Resource/Pharmacokinetics Core. The facility is part of the University Shared Resource Program at Oregon Health and Sciences University.
Funding disclosure
This work was supported by the National Institute on Drug Abuse [R00 DA041467].
Footnotes
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
References
