All chemicals were obtained from Sigma (St. Louis, MO) unless otherwise indicated. Coelenterazine 400a was purchased from Nanolight™ Technology. 2-AG, Δ⁹-THC and CBD are from Cayman Chemical (Ann Arbor, MI). The sp-hCB1 encoding plasmid (signal peptide human CB1) was a gift from Michel Bouvier, (University of Montreal).

Research grade cigarettes (3R4F) with a filter were acquired from the Kentucky Tobacco Research Council (Lexington, KY). Research grade cigarettes (3R4F) contain 0.73 mg of nicotine, 9.4 mg of tar, and 12.0 mg of CO as described by the manufacturer. CSE was produced as previously described by us (Baglole et al., 2008a; Zago et al., 2013; Guerrina et al., 2021a; Guerrina et al., 2021b). Briefly, CSE was prepared by bubbling smoke from a cigarette through 10 ml of serum-free cell culture medium with the exception that some extracts were prepared with 30% methanol (MeOH). The CSE was then sterile-filtered with a 0.45-μm filter (25-mm Acrodisc; Pall Corp., Ann Arbor, MI). Standardization was done for each CSE preparation by spectrophotometer using an OD₃₂₀ nm of 0.65 to represent 100% CSE as described (Baglole et al., 2006; Zago et al., 2013).

Cannabis was purchased from the Société québécoise du cannabis SQDC online store (Quebec, Canada). Whole flower cannabis that was selected for purchase contained varying cannabinoid profiles based on THC/CBD content. Those purchased were as follows: 1) Indica-THC dominant; contains 16–22% THC and 0–0.1% CBD (#688083002311). 2): Sativa-CBD dominant; contains 0.1–2% THC and 13–19% CBD (#694144000219) and 3) Hybrid-Balanced: contains 5–11% THC and 5–11% CBD (#688083002588). Cannabis joints (cigarettes) were hand-rolled by grinding the dried cannabis flower with a plastic grinder and packing the product into classic 1 1/4 size rolling paper (RAW^®). Each cannabis cigarette contained 0.5 ± 0.05 g of cannabis. A slim unrefined cellulose filter (RAW^®) was added to the end of the joint. Then, CaSE was produced as previously described for CSE (Baglole et al., 2008a; Zago et al., 2013; Guerrina et al., 2021b) where the smoke from the lit cannabis cigarette was bubbled through 10ml of serum-free cell culture Dulbecco’s modified Eagle’s medium (DMEM) with or without 30% methanol (MeOH) or 30% ethanol (EtOH). CaSE was then filtered using a 0.45-μm filter (25-mm Acrodisc; Pall Corp., Ann Arbor, MI). Because the tar components in tobacco and cannabis are similar (Tashkin, 2013), and chemical species of tobacco tar absorb light at 320 nm (Taylor et al., 2020), we standardized each CaSE preparation as previously described for CSE (Baglole et al., 2008a; Zago et al., 2013; Guerrina et al., 2021a; Guerrina et al., 2021b) to ensure consistency in CaSE preparations between experiments. Similar to CSE preparation described above, an optical density of 0.65 was considered to represent 100% CaSE. Then, the CaSE solution was diluted with serum-free MEM for further analysis. The pH of 2% CaSE and 5% CaSE was 7.3 ± 0.06 and 7.7 ± 0.08, respectively.

Δ⁹-THC concentration in CaSE was analyzed by a direct competitive THC Forensic ELISA kit (NEOGEN^®) according to manufacturer’s instructions. The concentration of interleukin-8 (IL-8) in the cell culture supernatant was determined by ELISA (Human IL-8 ELISA Duo Set, R&D Systems, United States) according to the manufacturer’s instructions. The absorbance was read at 450 and 570 nm within 15 minutes by infinite TECAN (M200 pro, TECAN, CA).

Human embryonic kidney (HEK) 293 cells were cultured in DMEM supplemented with 10% fetal bovine serum (FBS) and gentamicin (20 μg/ml). Cells were grown at 37°C in 5% CO₂ and 90% humidity. HEK293 cells were seeded at a density of 1 × 10⁶ cells per 100-mm dish and transfected the next day with 3 µg of sp-hCB1 with 120 ng of PKN-RBD-RlucII and 480 ng of rGFP-CAAX using PEI methods as described previously (Boussif et al., 1995; Namkung et al., 2018). Briefly, a total of 6 µg of DNA (adjusted with pcDNA3.1 zeo (+)) in 0.5 ml of PBS was mixed with 12 µl of PEI (25 kDa linear, 1 mg/ml) in 0.5 ml PBS and then incubated for 20 min at RT prior to applying to the cells. After 24 h, cells were detached and seeded onto poly-ornithine-coated 96-well white plates at a density of 25,000 cells per well for the BRET assays, which were performed 48 h after transfection.

Primary human lung fibroblasts (HLFs) were isolated from cancer-free lung tissue by explant procedure as described (Baglole et al., 2005). This study was approved by the Research Ethics Board of St. Joseph’s Healthcare Hamilton and informed written consent was obtained from each patient. Experiments were conducted with fibroblasts from three different individuals of the non-smoker group (Normal; M/F = 1/2; age 68 ± 9 years) and within passage six to nine. HLFs were cultured in 10% MEM and treated with THC dominant CaSE for 6 and 24 h.

BRET assay for detecting Rho activation was performed as previously described (Namkung et al., 2018). Briefly, cells in 96 well plates were washed once with 150 µl/well of Tyrode’s buffer (140 mM NaCl, 2.7 mM KCl, 1 mM CaCl₂, 12 mM NaHCO₃, 5.6 mM D-glucose, 0.5 mM MgCl₂, 0.37 mM NaH₂PO₄, 25 mM HEPES, pH 7.4) and left in 80 µl/well of Tyrode’s buffer. 2-AG, THC, and CBD were serially diluted in 15% MeOH in Tyrode’s buffer. The final concentration of MeOH in the assay is 3.75%. For BRET assay, the cells were loaded with 10 µl of coelenterazine 400a (final concentrations of ∼3.5 µM) and then the cells were stimulated with 30 µl of ligands or two-fold diluted CaSE in Tyrode’s buffer for 4 min prior to BRET measurement. Thus, final concentrations of CaSE were 12.5% (8-fold dilution of original CaSE). BRET signals were measured using a Synergy2 (BioTek) microplate reader. The filter was set at 410/80 nm and 515/30 nm for detecting the RlucII Renilla luciferase (donor) and rGFP (acceptor) light emissions, respectively. The BRET ratio was determined by calculating the ratio of the light emitted by rGFP over the light emitted by the RlucII.

CaSE culture media samples were diluted 1:20 v/v by adding 10 µl to 190 µl of MeOH containing an internal standard CBD-d9 (10 pmol); 10 µl was subsequently analyzed by LC-MS/MS. In some cases, a 1:2 dilution was prepared by mixing 100 µl of CaSE culture medium with 100 µl of methanol containing internal standard CBD-d9 (10 pmol). CBD was chromatographed on a Waters UPLC reversed phase column (100 × 2.1 mm i.d.) using a blend of water and acetonitrile containing 0.1% acetic acid with a flow rate of 0.2 ml/min. The eluate was directed into a Thermo Quantum Access Max triple-quadrupole mass spectrometer and the CBD and CBD-d9 detected by single-reaction monitoring. The peak area for CBD was normalized by the peak area for the internal standard (CBD-d9) and the ratio compared to an external calibration curve for CBD prepared in MeOH. The limit of quantitation for CBD was 10 nM.

HLFs were grown to approximately 70–80% confluence and cultured with serum-free MEM for 18 h before the treatment. Total cellular protein was extracted using RIPA lysis buffer (Thermo Scientific, Rockford) containing Protease Inhibitor Cocktail (PIC, Roche, United States). Ten to 20 μg of protein lysate were subjected to 10% SDS-PAGE gels and transferred onto Immuno-blot PVDF membranes (Bio-Rad Laboratories, Hercules, CA). Then, the membrane was blocked for 1 hour at room temperature in blocking solution (5% w/v of non-fat dry milk in 1x PBS/0.1% Tween-20). The primary antibodies, COX-2 (1:1,000; Cell Signaling Technology, CA) and β-Tubulin (1:50000; Sigma, CA) were added to the membranes and incubated overnight at 4°C or 1 h at room temperature. After several washes, membrane was incubated with secondary antibodies goat anti-rabbit IgG HRP-linked (1:10000, Cell Signaling Technology, CA) or HRP-conjugated horse anti-mouse IgG (1:10000, Cell Signaling Technology, CA). Detection of protein was done by enhanced chemiluminescence (ECL) and visualized using a ChemiDoc™ MP Imaging System (Bio-Rad, CA). Densitometric analysis was performed using Image Lab™ Software Version 5 (Bio-Rad, CA). Protein expression was normalized to β-tubulin and the data presented as the fold-change relative to the untreated condition.

Using the Aurum™ Total RNA Kit (Bio-Rad, CA), total RNA was isolated according to the manufacturer’s instructions. Quantification of RNA was conducted on a Nanodrop 1,000 spectrophotometer. Reverse transcription of RNA was carried out using iScript™ Reverse Transcription Supermix (Bio-Rad, CA). Then, using this cDNA template, mRNA levels of PTGS2, CXCL8 and S9 were analyzed by quantitative PCR (qPCR) by using 1 µl of cDNA (10 ng/μl) and 0.5 µM primers with SsoFast™ EvaGreen^® (Bio-Rad, CA). Sequences of gene-specific primers are listed in Table 1. PCR amplification was performed using a CFX96 Real-Time PCR Detection System (Bio-Rad, CA). Thermal cycling was initiated at 95°C for 3 min and followed by 39 cycles denaturation at 95°C for 10 s and annealing at 59°C for 5 s. Gene expression was analyzed using the ΔΔC_T method, and results are presented as fold-change normalized to housekeeping gene (S9).

Using GraphPad Prism 6 (v. 6.02; La Jolla, CA), statistical analysis was performed using a one-way analysis of variance (ANOVA) followed by Dunn’s multiple comparisons test to assess differences between treatments. Groups of two were analyzed by paired t-test. A two-way analysis of variance (ANOVA), followed by Tukey’s multiple comparisons test was used to evaluate differences between groups and conditions of more than two. Results are presented as mean ± standard error of the mean (SEM) or as mean ± standard deviation (SD) of the fold-changes compared to control cells. Experimental readings were done in triplicate and averaged; statistical analysis was therefore done using averaged values from three to five independent experiments unless otherwise indicated. In all cases, a p value <0.05 is considered statistically significant. For Tables 4, 5, the standard THC concentration response curve was obtained from a nonlinear regression curve fitting in GraphPad Prism software. The mean, upper limits, and lower limits of the unknowns were interpolated from the fitted standard curve with a confidential interval of 95%.

Like tobacco, cannabis smoke contains hundreds of combustion products. However, cannabis also contains cannabinoids that exert biological and pharmacological effects. Standardized preparations of aqueous cigarette smoke extract (CSE) are well-described in the literature and are used to understand the consequences of tobacco exposure (Carnevali et al., 2003; Baglole et al., 2006; Baglole et al., 2008b; Hecht et al., 2014); no such standardized extract for cannabis smoke exists. Moreover, CSE prepared in cell culture media or PBS contains water soluble gas and particle phases of cigarette smoke (Kim et al., 2018). While many of these same compounds would be captured from cannabis smoke, cannabis also contains cannabinoids which are hydrophobic (Huestis, 2007) and unlikely to be present in an aqueous extract suitable for in vitro testing. Therefore, we sought to develop a cannabis smoke extract (CaSE) that contains biologically active cannabinoids. First, we utilized a semi-quantitative THC Forensic ELISA kit for which we developed a standard curve using Δ⁹-THC to allow for subsequent quantification. The standard curve was first prepared to calculate the relative concentration of Δ⁹-THC relative to the absorbance. We diluted Δ⁹-THC (in the ELISA buffer) from a starting concentration of 1 mg/ml to an upper limit of 4 μg/ml. The concentration of this Δ⁹-THC standard curve therefore ranged from 0 μg/ml (buffer only)- 4 μg/ml (0–12.7 µM) (Figure 1) and was used for analysis with all CaSE preparations.

Next, we evaluated Δ⁹-THC level by ELISA and CBD level LC-MS/MS in various aqueous CaSE preparations. Given that cannabinoids are hydrophobic, we compared Δ⁹-THC levels in CaSE prepared in standard cell culture media with and without MeOH. As additional controls, we also evaluated Δ⁹-THC concentrations in CSE prepared from research grade cigarettes. As expected, cell culture media alone with 30% MeOH as well as CSE (with 30% MeOH) contained no Δ⁹-THC or CBD (Table 2). We also measured Δ⁹-THC and CBD concentrations in CaSE prepared from the different strains of cannabis with reported varying amounts of Δ⁹-THC and CBD. Δ⁹-THC levels in CaSE generated from the Δ⁹-THC dominant and THC/CBD balanced strains with 10% MeOH were 0.62 ± 0.2 and 0.36 ± 0.02, respectively, and were therefore similar to the level in CaSE without MeOH. However, in CaSE generated from the Δ⁹-THC dominant strain with 30% MeOH, there were significantly higher levels of Δ⁹-THC compared to the CaSE without MeOH (THC dominant CaSE; Table 2). CBD levels were below the limit of detection by LC-MS/MS. Here, the estimated Δ⁹-THC concentration was 6.7 ± 0.29 µM in CaSE prepared in cell culture media with 30% MeOH. Preparation of CaSE from the balanced cannabis strain with 5–10% THC and 5–11% CBD also yielded significant Δ⁹-THC levels only when CaSE was prepared in media containing 30% MeOH. Finally, CaSE prepared from the CBD dominant cannabis strain in media with 30% MeOH has less Δ⁹-THC compared to CaSE prepared from the other two cannabis strains (Table 2). In CaSE generated from the balanced cannabis strain with 5–10% THC and 5–11% CBD, there were higher levels of CBD compared to the CaSE without MeOH (THC/CBD balanced CaSE; Table 2). CaSE prepared from the CBD dominant cannabis strain in media with 30% MeOH has higher CBD compared to CaSE prepared from the THC dominant strains. However, CBD levels are similar between CBD dominant and THC/CBD balanced strains (Table 2). We also generated CaSE from the Δ⁹-THC dominant strain in media with 30% EtOH. We found that the Δ⁹-THC level was slightly less in CaSE containing EtOH (5.7 ± 0.35 µM) comparing to CaSE with MeOH (6.7 ± 0.29 µM). These data show that preparation of CaSE in cell culture media with MeOH yields significantly higher concentrations of Δ⁹-THC and CBD compared to CaSE prepared without MeOH. Thus, the remainder of experiments were conducted with CaSE prepared in media with 30% MeOH and is refereed to hereafter as CaSE.

The tar components in tobacco and cannabis are similar (Tashkin, 2013), and chemical species of tobacco tar absorb light at 320 nm (Taylor et al., 2020). Thus, to ensure consistency in CaSE preparations between experiments, we standardized each CaSE preparation as previously described for CSE (Baglole et al., 2008a; Zago et al., 2013; Guerrina et al., 2021a; Guerrina et al., 2021b). Nine extracts from THC dominant cannabis were prepared and two measurements were taken for fresh extracts and after thawing of the same extracts that had been frozen at −80°C for 16 weeks. The optical density (OD) at 320 was 0.7 ± 0.05 and 0.64 ± 0.05 for fresh and frozen extracts, respectively (Table 3). Given that an OD of 0.65 is used to represent 100% CaSE, the percentage of CaSE averaged to be 110% ± 8 and 99% ± 7.7 for fresh and frozen extracts, respectively. We also evaluated Δ⁹-THC content by ELISA. The estimated Δ⁹-THC concentration of fresh and frozen extracts was similar and was approximately 12 µM. These data suggest that storage of CaSE extracts up to 16 weeks at −80°C does not affect Δ⁹-THC concentration and that an OD₃₂₀ can be used to standardize aqueous CaSE to minimize batch-to-batch variability.

Δ⁹-THC has high affinity to CB1 and CB2 receptors (Pertwee, 2010), which are G protein coupled receptors (GPCRs). CB1 couples to not only G_i/o but also to the G_12/13 subfamily and activates the down-stream protein Rho (Inoue et al., 2019; Krishna Kumar et al., 2019; Avet et al., 2020). To determine whether there is sufficient Δ⁹-THC in the CaSE preparations to activate CB1, we used a BRET-based Rho biosensor (Namkung et al., 2018). We transiently transfected HEK293 cells with signal-peptide-human CB1 (CB1) along with PKN-RBD-RLucII and rGFP-CAAX (Rho sensor) and stimulated the cells with Δ⁹-THC, CBD and 2-arachidonoylglycerol (2-AG), an endogenous CB ligand (Figure 2). The BRET signal increased in response to Δ⁹-THC and 2-AG but not to CBD (Figure 2A). Further, we observed that AM251, a CB1-specific antagonist, abolished the THC- and 2-AG- promoted BRET signals (Figure 2B). To verify the specificity of AM251 on CB1-mediated Rho activation, we examined the effect of AM251 on angiotensin II type 1 receptor (AT1R)-mediated Rho activation, which also couple to this pathway (Namkung et al., 2018). AM251 showed no effect on the basal BRET whereas AngII induced a BRET signal in HEK293 cells expressing AT1R along with Rho sensor (Figure 2C). These data show that Δ⁹-THC- and 2-AG- promoted CB1 activation and signaling to the G_12/13-Rho pathway.

We next vetted three different extracts prepared from THC dominant or THC/CBD balanced cannabis prepared in media with or without 30% MeOH to verify that these CaSE preparations contained biologically active Δ⁹-THC; we utilized the same extracts as for the data presented in Table 2. First, the activation of CB1 in response to different concentrations of Δ⁹-THC (0.3–25 µM) was assessed. There was a concentration-dependent activation of CB1 by Δ⁹-THC (Figure 3). Furthermore, there was an increase in CB1 activation in cells treated with CaSE from THC dominant or THC/CBD balanced cannabis prepared in media with 30% MeOH (Figure 3). Extracts in media without 30% MeOH did not show BRET signals in our assay (data not shown). Based on CB1 activation by Δ⁹-THC (Figure 3), we extrapolated that CaSE prepared from THC-dominant cannabis activates CB1 in concentrations equivalent to 5–7 µM of Δ⁹-THC (Table 4). CaSE from THC/CBD balanced cannabis also activates the receptor, which is equivalent to 3–50 µM of Δ⁹-THC (Table 4).

We then tested whether the receptor itself was affected by the MeOH and evaluated the specificity of the system by adding CSE prepared in media with 30% MeOH; we also included CaSE from all three cannabis strains (see Table 2). We found that there was no Rho activation with media containing 30% MeOH or CSE (Figure 4). CaSE from THC dominant, THC/CBD balanced, and CBD-dominant cannabis all activated Rho signaling (Figure 4), at levels that corresponded approximately to between 4–22 µM of Δ⁹-THC present in the extracts (Table 5). Thus, CaSE, but not media containing MeOH or CSE, activates the CB1 receptor. Finally, we used the CB1 antagonist AM251 to confirm that CaSE is specific in its ability to activate CB1. AM251 inhibited THC-induced Rho activation. We also found that AM251 significantly inhibits CaSE-induced Rho activation for the CaSE prepared from the THC-dominant and THC/CBD balanced strains (Figure 5). Thus, CaSE induces Rho activation through CB1. Taken together, these data show that a standardized preparation of CaSE contains biologically active cannabinoids.

COX-2 and IL-8 are among the proinflammatory mediators that are induced by tobacco smoke (Martey et al., 2004; Li et al., 2007). IL-8 is also elevated in serum from cannabis smokers (Bayazit et al., 2017). To explore whether we could replicate these findings, we characterized the effect of CaSE exposure on the expression of COX-2 and IL-8 at the mRNA and protein levels in primary HLFs. For these experiments, HLFs were treated with either 2% or 5% CaSE that was prepared from THC dominant cannabis. Selection of these concentrations was based on our previous publications with CSE (Baglole et al., 2008a; Guerrina et al., 2021a). These concentrations of CaSE did not affect cell viability (data not shown). The concentration of Δ⁹-THC in 2% CaSE was 0.18 ± 0.003 µM and in 5% CaSE was 0.45 ± 0.006 µM (n = 3). The mRNA for PTGS2 did not increase with 6 h of CaSE (Figure 6A). However, there was a significant increase in PTGS2 mRNA upon exposure to 5% CaSE for 24 h- but not 2% CaSE. Accordingly, there was a significant increase in COX-2 protein with 5% CaSE (Figure 6B). There was also a significant increase in CXCL8 mRNA in response to 5% CaSE for 24 h (Figure 6C). At the protein level, IL-8 was also induced upon 5% CaSE treatment for 24 h (Figure 6D). These data indicate that a standardized CaSE preparation, containing biologically active cannabinoids, induces an inflammatory response in primary HLFs.