Characterization of cannabinoid plasma concentration, maternal health, and cytokine levels in a rat model of prenatal Cannabis smoke exposure

Cannabis sativa has gained popularity as a “natural substance”, leading many to falsely assume that it is not harmful. This assumption has been documented amongst pregnant mothers, many of whom consider Cannabis use during pregnancy as benign. The purpose of this study was to validate a Cannabis smoke exposure model in pregnant rats by determining the plasma levels of cannabinoids and associated metabolites in the dams after exposure to either Cannabis smoke or injected cannabinoids. Maternal and fetal cytokine and chemokine profiles were also assessed after exposure. Pregnant Sprague–Dawley rats were treated daily from gestational day 6–20 with either room air, i.p. vehicle, inhaled high-Δ9-tetrahydrocannabinol (THC) (18% THC, 0.1% cannabidiol [CBD]) smoke, inhaled high-CBD (﻿0.7% THC, 13% CBD) smoke, 3 mg/kg i.p. THC, or 10 mg/kg i.p. CBD. Our data reveal that THC and CBD, but not their metabolites, accumulate in maternal plasma after repeated exposures. Injection of THC or CBD was associated with fewer offspring and increased uterine reabsorption events. For cytokines and chemokines, injection of THC or CBD up-regulated several pro-inflammatory cytokines compared to control or high-THC smoke or high-CBD smoke in placental and fetal brain tissue, whereas smoke exposure was generally associated with reduced cytokine and chemokine concentrations in placental and fetal brain tissue compared to controls. These results support existing, but limited, knowledge on how different routes of administration contribute to inconsistent manifestations of cannabinoid-mediated effects on pregnancy. Smoked Cannabis is still the most common means of human consumption, and more preclinical investigation is needed to determine the effects of smoke inhalation on developmental and behavioural trajectories.

General trends of Cannabis legalization/decriminalization have changed social narratives surrounding use and risk perception in Canada and around the world.As social perception has changed, rates of use have increased for all ages by 7.3% since legalization, where smoking still accounts for 79% of Cannabis consumption in Canada 1 .Importantly, Cannabis has gained popularity as a "natural substance, " with the false assumption that it is therefore safe.Emerging human data indicate that children exposed to Cannabis in utero are at a higher risk of being born pre-term, underweight, and developing persistent behavioural psychopathologies and cognitive deficits across their lifetime [2][3][4][5][6][7][8][9][10][11] .Although long-term outcomes are still not understood, these data indicate the need for more preclinical research to establish the risks and investigate the molecular and physiological consequences of in utero Cannabis smoke exposure.
Over 120 unique phytocannabinoids are produced by the Cannabis sativa plant, with the acid forms of Δ 9 tetrahydrocannabinol (THC) and cannabidiol (CBD) present at the highest relative concentrations 12 .These
From these data, we conclude that circulating levels of THC or CBD, but not their respective metabolites, increased by approximately threefold between GD6 to GD20 in all treatments, regardless of administration method.

Pharmacokinetic evaluation and comparison of maternal plasma phytocannabinoid levels during gestation: route of exposure influences observed concentrations
Mean GD6 and GD20 plasma levels of THC, CBD, and their metabolites were compared among treatments.As these analyses revealed similar differences among the treatments at each timepoint, we depicted the data from GD20 in Fig. 3 and the data from GD6 in Supp.Fig S1 .For GD20 samples, THC levels were above the LLOQ for i.p THC, high-THC smoke, and high-CBD smoke.The Kruskal-Wallis (KW) test revealed a significant difference among treatments (H(3) = 13.70,p < 0.0001), where the post-hoc showed that levels of THC were significantly higher following i.p.THC and high-THC smoke than high-CBD smoke (Fig. 3A).It is noteworthy that mean levels of THC in the i.p.THC-exposed dams are approximately tenfold higher than those of the high-THC smoke group; however, due to unequal variances between treatments, this difference did not emerge as significant in the analysis.Mean plasma levels of 11-OH-THC were above the LLOQ for the i.p.THC and high-THC smoke groups only (Fig. 3B).Levels of 11-OH-THC in the i.p.THC treatment were significantly higher than levels in the high-THC smoke group (KS: p = 0.047).Mean plasma levels of 11-COOH-THC were greater than the LLOQ in the i.p.THC and high-THC smoke treatments only (Fig. 3C).Plasma levels of 11-COOH-THC were higher in the i.p.THC group compared to high-THC smoke (KS: p = 0.047) (Fig. 3C).
Plasma levels of CBD were greater than the LLOQ for the i.p.CBD and high-CBD smoke groups only and were significantly higher with i.p.CBD compared to the high-CBD smoke-exposed group (KS: p = 0.0003) (Fig. 3D).7-OH-CBD was only detectable above the LLOQ in rats of the i.p.CBD group (Fig. 3E).

Figure 1.
Experimental timeline for gestational Cannabis exposure.Rats were bred and treated once daily between gestational day (GD) 6 and 20 for 15 consecutive days of treatment.Maternal and fetal tissues were harvested 30 min following the first (GD6) and last (GD20) treatment.THC, CBD, and their metabolites were quantified in maternal plasma using high performance liquid chromatography tandem mass spectroscopy (HPLC-MS/MS).Protein samples from placenta and whole fetal brain were prepared for cytokine and chemokine quantification.Created with Biorender.com.
From these data, we conclude that exposure of pregnant dams to high-THC smoke was associated with quantifiable levels of THC, 11-OH-THC, and 11-COOH-THC that were approximately 10% of the respective levels of THC, 11-OH-THC, and 11-COOH-THC in pregnant dams treated with 3 mg/kg/day THC i.p. (Fig. 3A-C).Similarly, treatment with high-CBD smoke was associated with quantifiable levels of CBD on GD20.Levels of CBD and 7-OH-CBD in high-CBD smoke-exposed dams were 1% and 10% of the respective plasma levels observed in pregnant dams treated with 10 mg/kg/day CBD i.p. (Fig. 3D,E).Therefore, the smoke exposure paradigm used in these studies yields quantifiable levels of plasma cannabinoids in pregnant dams at levels lower than those observed with conventionally used rodent injection models when collected 30 min after the final treatment.

THC injection was unique among treatment groups increasing or maintaining cytokine and chemokine levels relative to controls whereas other treatments were associated with reduced cytokine and chemokine levels
To survey the inflammatory state of placental and fetal tissue after phytocannabinoid treatment, 27 cytokines and chemokines were quantified in placental and fetal brain tissue collected on GD20, 30 min after final treatment.
IL-4, IL-6 and IL-10 are pro-and anti-inflammatory cytokines depending on how they interact with various cell types in pregnancy and other conditions 29 .In this study, IL-6 levels were elevated in the placenta of dams receiving THC injections compared to control (p < 0.01) or compared to dams receiving high-THC smoke (p < 0.001) (Fig. 5J).IL-4 (p < 0.05) and IL-10 (p < 0.05) levels were elevated in the placenta of dams receiving THC injections compared to dams receiving high-THC smoke (Fig. 5K,L).
From these data, three general trends emerge: (i) the majority of changes observed were for pro-inflammatory and not anti-inflammatory cytokines; (ii) when comparing to controls, injection of either THC or CBD was associated with the upregulation of four species (IL-1β, RANTES, TNFα, and IL-6), three of which are proinflammatory, whereas smoke exposure was generally associated with decreases in pro-inflammatory species; and (iii) upregulation of a subset of cytokines (eotaxin, IFNγ, IL-1β, IL-4, IL-6, IL-10, IL-17A, and TNFα) was observed in placenta tissue of rats receiving THC injection relative to high-THC smoke and three species (LIX, MIP-1α, and RANTES) were similarly up-regulated in rats receiving CBD injection relative to high-CBD smoke.
Together, these data highlight the widespread changes in inflammatory state that occurred in both injection and inhalation rodent models of prenatal cannabinoid exposure and illustrate that those injection and smoke exposure produced disparate outcomes in placental tissue.
EGF is considered an anti-inflammatory chemokine 98 .EGF levels were elevated in the fetal brain tissues of dams that received THC injections compared to dams that received CBD injections (p < 0.05, Fig. 6O).
From the fetal brain samples two patterns arise: (i) exposure to Cannabis smoke (high-THC or high-CBD) or injected CBD was associated with widespread downregulation of several cytokines and chemokines in fetal brains; and (ii) several analytes were notably elevated in fetal brain tissue from THC-injected groups relative to high-THC Cannabis smoke, but no post-hoc elevation was observed when THC-injected sample means were compared to control means.These data further demonstrate the widespread changes in inflammatory state that occur in both injection and inhalation rodent models of prenatal cannabinoid exposure and present several instances where changes observed in the placenta are similarly observed in the fetal brain.

Discussion
In the present study, pregnant female rats were treated daily from GD6 until GD20 with either a control (i.p. vehicle injection or room air), 3 mg/kg/day i.p.THC, 300 mg/day high-THC Cannabis smoke (SW), 10 mg/kg/ day i.p.CBD, or 300 mg/day high-CBD Cannabis smoke (TI).Results indicate that the duration of exposure (Fig. 2) and route of administration (Fig. 3) significantly affect the levels of cannabinoids and their metabolites in maternal plasma 30 min following administration.Analyses of maternal and fetal bodyweight, resorptions, litter size, fetal brain to BW and BW to placenta and sex ratios, over the course of pregnancy suggest relatively subtle effects of these treatment regimes, except for some evidence of reduced fetal viability following repeated injections of either THC or CBD (Fig. 4).Following repeated treatment of the dams, levels of cytokines and chemokines in the placenta (Fig. 5) and fetal brain (Fig. 6) were also altered in a cannabinoid-and route-dependent manner.Taken together, our results suggest that the route of administration of cannabinoids in developmental studies is an important factor in determining the specific effects obtained.Thus, care should be taken to use methods of administration to closely resemble those of humans to improve the construct validity of preclinical models.

Comparing plasma levels of THC and CBD following repeated injections or smoke administrations to pregnant rats
Analyses of plasma levels of cannabinoids 30 min after exposure revealed that plasma levels of THC and CBD, but not their metabolites, were significantly higher when sampled on GD20, after 15 administrations, compared to GD6, after one administration (Fig. 2).Selection of the 30 min timepoint enabled comparisons to existing plasma concentrations measured in adult male rats from our lab 93,94 .To the best of our knowledge, this is the first-time plasma cannabinoid levels have been directly compared between acute and repeated administrations during pregnancy in rats.While the potential for accumulation during pregnancy is interesting, these results may be explained by several factors including-but not limited to-homeostatic changes to the metabolism of cannabinoids during repeated exposures or binding and depositing of these cannabinoids on plasma proteins and other tissue during repeated dosing.It is noteworthy that levels of THC and CBD following acute exposure on GD6 were lower than typically observed in male, but not female, rats.Barnard and colleagues 94 reported plasma levels of THC of 125 ng/ml (vs.approximately 77 ng/ml for GD6 here) and 15 ng/ml (vs.approximately 8 ng/ml for GD6 here) 30 min following acute treatment of male rats with the same THC injection and high-THC smoke protocols used in the present study.Plasma levels of CBD following high-CBD smoke exposure were approximately 15 ng/ml in the same study 94 , which is also considerably higher than the roughly 5 ng/ml reported following the first administration of high-CBD smoke to the pregnant rats in this study.However, Baglot and colleagues 99 reported sex differences in the plasma concentrations of injected THC (2.5 mg/kg, i.p.) with females showing levels similar to those reported herein, and males showing higher levels similar to the Barnard study 94 .In a comparison of THC (dronabinol) Volcano ® inhalation treatment to i.p. injection in adult male rats, Manwell and colleagues 100 reported comparable plasma levels of THC at 20 min and at 40 min time points between 1 mg/pad THC and 2.5 mg/kg i.p.THC to males injected in Barnard 94 .Using a different protocol www.nature.com/scientificreports/ of vaporized THC exposure, Baglot and colleagues found higher plasma levels of THC in male and female rats 30 min after inhalation than reported in the present study and showed no difference in circulating concentrations between their inhalation and injected THC groups at this time point 99 .Notably, Sandini and colleagues detected behavioural changes in adult offspring after repeated prenatal exposures to high-THC smoke like those used in the present study 81 .In another inhalation model, Jenkins and colleagues 101 observed suppressed oscillatory power and coherence with acute THC flower vapour exposure in male rats; however, neither Jenkins et al. 101 nor Sandini et al. 81 provide pharmacokinetic support for their models.
To the best of our knowledge, there are no previous reports of CBD accumulation during pregnancy and the direct comparison of plasma levels of CBD following injected CBD or inhalation of high-CBD smoke in this study is novel.We found maternal plasma CBD levels at 30 min following 10 mg/kg/day i.p. were 239 ng/mL on GD6, increasing to 1,473 ng/mL after 15 daily injections on GD20 (Fig. 3).In the high-CBD smoke group, levels of CBD were dramatically lower than in injected CBD dams, where on GD6 plasma levels were 7 ng/ml which increased to 17 ng/mL on GD20 at 30 min after administration.Our injected CBD GD6 plasma values more closely align with acute vapor administration in female rats at 100 ng/mL and 10 mg/kg i.p 102 .whereas our GD20 injected CBD plasma values resemble those from Ochiai and colleagues (2021), where a single dose of 10 mg/ kg i.v. of CBD resulted in a C max in maternal plasma of ~ 1200 ng/mL and 900 ng/mL in whole fetal tissue 82 .Of note, Ochiai and colleagues (2021) reported the half-life of injected CBD (10 mg/kg; i.v.) as 5 h in pregnant mice 82 , which casts doubt on significant accumulation from repeated once daily injections.Thus, comparing plasma cannabinoid levels over a time course by sex and reproductive status will be an important consideration for future research.In addition, assessing plasma levels of cannabinoids without treatment on GD20 would confirm whether accumulation of the cannabinoids was occurring over repeated exposures.
Our experimental design also allowed for a direct comparison of plasma levels of cannabinoids following either smoke exposure or injections.Levels of THC and CBD, as well as some of their metabolites, were significantly higher 30 min following repeated injections than repeated smoke exposure on GD20.In general, we chose to compare injected doses of THC and CBD commonly used in rodent models 41,42,45,49,51,[103][104][105][106] , with the smoke inhalation protocol developed in our laboratory 81,[93][94][95] .The inhalation protocol resembles other 'hotbox' methods 100,101,107 , and has been shown to result in acute behavioural and oscillatory effects [93][94][95] , as well as long-term changes in the offspring following exposure during pregnancy 81 .Despite previous acute and chronic behavioural disturbances associated with inhalation methods, this is the first pharmacokinetic evaluation comparing acute (GD6) and chronic (GD20) exposure in a prenatal model of whole cannabis flower smoke exposure.As reported previously for injected THC and smoke exposure 94 , levels of THC were lower following inhalation than injections.Determining the peak and total exposure is difficult with one sampling point and the passive exposure methods we used with smoke administration; however, it is likely that our smoke exposure protocol results in lower levels of exposure when compared to some other inhalation methods 40 .In addition, different pharmacokinetics following these administration methods may dramatically affect the peak and total exposure of the pregnant woman and fetus to cannabinoids.Our selection of timepoint and quantification of metabolites indicate that the doses of major phytocannabinoids received from smoked whole flower differs from injected phytocannabinoids and may contribute to different long-term outcomes in the offspring.

Relatively subtle effects of repeated cannabinoid administration on maternal and fetal parameters
Injected cannabinoids and inhaled Cannabis did not significantly impact maternal weight gain, fetal weight, fetal brain to BW or BW to placenta ratios, or sex ratios in this study.Previously, we observed no effect of repeated high-THC smoke exposures on maternal weight gain, litter size, or litter weight following high-THC smoke exposure 81 , although in that study an increase in the ratio of male:female offspring was significant.A limitation of this study is the relatively small sample sizes for the treatment groups that were carried to GD20, which may have limited the emergence of more nuanced differences in markers of maternal and fetal health.However, injected CBD did increase fetal resorptions and injected THC decreased litter size.Natale and colleagues (2020) reported increased rates of uterine resorption in mice treated with 3 mg/kg/day i.p.THC, alongside indicators of intra-uterine growth restriction (IUGR) and placenta abnormalities 42 .Although CBD has not been previously investigated in these models in vivo, in vitro assays of trophoblasts show that CBD has a dose-dependent negative impact on cell cycle progression, viability, and migration; all crucial mechanisms in fetal trophoblast development 108 .Analysis of clinical maternal and childhood health outcomes has yielded different results due to confounding variables such as polysubstance use and inaccurate reporting of Cannabis use frequency.However, clinical retrospectives and pre-clinical investigation have been associated with adverse neonatal outcomes such as decreased birthweight and APGAR scores, preterm births, miscarriages, growth restrictions, and placenta abnormalities (reviewed in 36 ), and new epidemiological data show geospatial associations in the US between congenital abnormalities and Cannabis availability, specifically CBD products 109 .It is therefore possible that injected THC and CBD produce similar restrictions to healthy litter growth resulting in the loss of pups.Because neither smoke treatment used in the current study produced any differences in measures of maternal and fetal health, we suggest that injected and inhaled Cannabis models differ in their effects on maternal and fetal health.

Repeated cannabinoid treatments differentially alter cytokine levels in placenta and fetal brain tissue
In GD20 placental tissues from rats repeatedly exposed to either THC or CBD, we observed significant differences in levels of 13 of 27 cytokines and chemokines compared to control (Fig. 5).Injected THC treatment upregulated 4 inflammatory cytokines and chemokines, whereas injected CBD only elevated TNFα (also observed with injected THC) and RANTES (not observed for injected THC).Smoke exposure, regardless of the cannabinoid www.nature.com/scientificreports/content, generally decreased cytokine and chemokine levels.Analyses of fetal brain tissue harvested from the same pregnancies as the placental tissue revealed significant alterations in levels of 14 of 27 cytokines and chemokines tested among the groups (Fig. 6).Although the effects of maternal Cannabis exposure on immune mechanisms in the mother and offspring are of current interest 110 , limited in vivo data exist regarding the maternal and fetal immune response to repeated cannabinoid treatments, and potential sex dependant differences.
Our results show intriguing changes 30 min following the final treatment on G20.Determining the time course and significance of these changes for placental and fetal health will require additional experimentation.IL-1β, IL-6, IP-10 (CXCL10), and TNFα are all pro-inflammatory cytokines and chemokines important for macrophage and natural killer (NK) cell activation, where the presence of NK cells correlate with fetal resorption, due to their damaging effects on trophoblast proliferation 26 .Elevations in IL-6, TNFα, and IP-10 (CXCL10), specifically amongst placenta and fetal tissue, have been associated with preterm births, intrauterine fetal death, and resorptions in addition to other downstream pathophysiological and behaviour changes [111][112][113] .

Conclusion
Our results show that injection of THC or CBD during gestation produces markedly different physiological effects when compared to smoke exposure with high-THC or high-CBD Cannabis.Smoke exposure did not cause any large-scale changes in markers of maternal or fetal health but did change the levels of immunomodulatory cytokines and chemokines in both placenta and fetal brain samples.Further in-depth behavioural analyses of offspring following exposure to Cannabis smoke compared to injected cannabinoids is required.An important limitation of our current model is that cannabinoids were administered to rats during the equivalent of the late first and second trimester of humans, and therefore our exposure model does not reflect potential use patterns and effects during critical neuro-developmental periods before conception, early in gestation, or the third trimester.Additionally, our model of smoke inhalation uses passive exposure, which standardizes the exposure that dams received.Active self-administration models (e.g., 78 ), may better represent voluntary usage patterns and reduce exposure stress.In addition, analysis of rodent breathing patterns during hot box exposure methods may also help to determine whether rodents make attempts to titrate exposure by limiting inhalation following repeated treatments.Overall, these data highlight the fundamental differences in outcomes that route, dose, and formulation of Cannabis can have.These results should be considered in our interpretations of rodent model data for gestation and other applications of smoke inhalation and cannabinoid injection.

Animals
Virgin female (n = 76) and male Sprague-Dawley rats (n = 24; Charles River, Senneville, QC) arrived at our animal housing facility and were allowed to habituate for at least 1 week in ventilated cages in a climate-controlled vivarium under a 12 h:12 h light-dark cycle (lights on at 07:00).Ad libitum food (Purina rat chow) and water was provided throughout all experiments.Animals were then handled for 1 week before males and females (8-10 weeks of age) were paired at 17:00 and left overnight to breed.Evidence of sperm following vaginal swab at 09:00 the following morning was considered GD0.Dams were left undisturbed until GD6, when drug treatments began (Fig. 1).Dams were randomized and assigned treatment with either high-THC Cannabis smoke (Skywalker Kush [SW]: 18% THC, 0.1% CBD, Aphria Inc., Lot# 5142072190); high-CBD Cannabis smoke (Treasure Island Kush [TI]: 13% CBD, 0.7% THC, Aphria Inc., Lot# 5142071556); or i.p. injections of 3 mg/kg/day THC (Cayman Chemical, Ann Arbor, MI Cat.No. 9003740), 10 mg/kg/day CBD (Cayman Chemical, Cat.No. 9003741), vehicle (i.p.) (1:1:18 ratio ethanol:kolliphor:phosphate buffered saline) or air exposure once daily between GD6-20 (Fig. 1).Once pregnant, dams were singly housed in standard cages with adequate cage enrichment.Dams and offspring were euthanized via overdose with 5% isoflurane, followed by cardiac puncture for plasma collection for dams and decapitation for offspring.All experiments were conducted according to standards set by Canadian Council on Animal Care and the University of Saskatchewan Research Ethics Board (AREB), and in accordance with the ARRIVE guidelines 114 .All experimental methods and protocols were approved by the University of Saskatchewan AREB (Protocol numbers 20210009, 20190067).

Smoke exposure
Pregnant rats (n = 5-8/group) were treated using a well-validated inhalation system (La Jolla Research Inc.) and method established in 81,[93][94][95] .Cannabis was shredded and ground using a coffee grinder (10 s).Each bowl (ceramic coil) was packed with 300 mg of the ground Cannabis flower.Airtight chambers (33 cm × 30.5 cm × 51 cm) equivalent to approximately 50 L each, could house a maximum of 2 rats, separated into plastic cages with metal grate roof used exclusively in the smoke chambers.Rats were habituated to the smoke chamber and pumps for 2 days for 20 min prior to smoke exposure.Before combustion, rats were placed in the chambers with pumps turned on for 5 min.Air was pumped through the chambers at 10 L/min and exhausted into a fume hood.Cannabis was combusted over a period of 1 min and the resulting smoke continuously pumped through the system.Pumps were turned off for 1 min following the combustion protocol before the smoke was vented gradually over 13 min.

Tissue collection
Maternal blood collection was performed via cardiac puncture.Rats were anesthetized with 5% isoflurane, and maternal blood, organs and fetal tissues were collected 30 min following the start of their last drug treatment.Fetal and maternal tissues were placed in 5 mL LoBind Eppendorf tubes (Cat.No. 0030122356), frozen in liquid nitrogen, and stored at -80 °C.Cardiac blood was collected by syringe and immediately transferred to 4 mL BD Vacutainer tube (Cat.No. CABD367884).All blood samples were centrifuged for 10 min at 4 °C at 2000×g under

Figure 2 .
Figure 2. Comparison of THC, CBD, and metabolite levels in maternal plasma according to treatment between first exposure on GD6 and last exposure on GD20.Plasma levels of phytocannabinoids were quantified on GD6 and GD20 30 min after cannabinoid exposure in the following treatment groups: (A) 3 mg/kg/day THC i.p., (B) 10 mg/kg/day CBD i.p., (C) High-THC smoke, and (D) High-CBD smoke.Note the different scaling of the y-axes of the panels.Data are mean ± S.E.M. n = 5-8 dams/per treatment.Values falling below LLOQ were deemed not detectable (n.d.).

Figure 4 .
Figure 4. Markers of maternal and litter health during gestation.(A) Rates of maternal weight gain, (B) uterine resorption discovered, (C) litter size (D), mean fetal body weight per litter, (E) mean fetal brain to body weight ratio per litter, and (F) mean fetal bodyweight to placenta ratio per litter, and (G) Ratio of male to female offspring per litter.Data are mean ± S.E.M, n = 4-8 dams/per treatment.Measurements in panels B-G were taken on GD20.*p < 0.05, as determined by one-way ANOVA followed by Holm-Šídák post-hoc analysis.