Comparative metabolism of THCA and THCV using UHPLC-Q-Exactive Orbitrap-MS

Abstract Delta(9)-tetrahydrocannabinolic acid (THCA) and delta(9)-tetrahydrocannabivarin (THCV) are phytocannabinoids with a similar structure derived from Cannabis sativa and possess a variety of biological activities. However, the relationship between the metabolic characterisation and bioactivity of THCA and THCV remains elusive. To explore the relationship between the metabolism of THCA and THCV and their underlying mechanism of activity, human/mouse liver microsomes and mouse primary hepatocytes were used to compare the metabolic maps between THCA and THCV through comparative metabolomics. A total of 29 metabolites were identified containing 7 previously undescribed THCA metabolites and 10 previously undescribed THCV metabolites. Of these metabolites, THCA was transformed into an active metabolite of delta(9)-tetrahydrocannabinol (THC) in these three systems, while THCV was transformed into THC and CBD. Bioactivity assays indicated that all of these phytocannabinoids exhibited anti-inflammatory activity, but the effects of THCA and THCV were slightly different in macrophages RAW264.7. Prediction of ADMET lab demonstrated that THCV and its metabolites were endowed with the advantage of blood–brain barrier (BBB) penetration compared to THCA. In conclusion, this study highlighted that metabolism plays a critical role in the biological activity of phytocannabinoids.


Introduction
Phytocannabinoids of multiple pharmacological activities were extensively used in the scientific field and pharmaceutical industry, which have been legalised in the USA for the past few years (Smart and Pacula 2019). More than 150 phytocannabinoids were characterised from Cannabis sativa L, including two naturally abundant phytocannabinoids of delta(9)-tetrahydrocannabinol (THC) and cannabidiol (CBD) (Nigro et al. 2021). Apart from THC and CBD, other bioactive compounds could interact with G-protein-coupled receptor (GPCR) and/or peroxisome proliferator-activated receptor (PPAR) proteins expressed in many different neuronal types, such as delta(9)-tetrahydrocannabivarin (THCV) and delta(9)tetrahydrocannabinolic acid (THCA) (Figure 1) (Morales et al. 2017;Palomares et al. 2020). THCA is the acidic precursor of THC and it is also the main form existing in fresh hemp tissue, which is easily decarboxylated to THC by under smoking, heating, or lighting. It exerts a variety of biological activities, including immunomodulatory, anti-inflammatory, neuroprotective, and antineoplastic effects (Moreno-Sanz 2016; Skell et al. 2021). The compound of THCV exists in Cannabis sativa, with a small content, while a broad medicinal value. It could interact with the endogenous cannabinoid system by binding with CB1 and CB2 receptors, and it displays an inhibition (lower than 3 mg/kg) or activation effect (10 mg/kg) with different concentrations (Pertwee 2008). Of the three compounds, the abundance of THCA in cannabis was highest, followed by THC and finally THCV Li et al. 2022).
Metabolomics is an effective tool that reflects the normal changes and pathophysiological effects of diseases in the biological system (Zhao et al. 2010). Pharmaco-metabonomics is a branch of metabolomics that helps to shed light on the mechanisms underlying individual diversity in drug toxicity and side effects, as well as on the pathways that help to determine the pharmacokinetic and pharmacodynamic characteristics of the drug-associated response (Corona et al. 2012). By accurately screening the exogenous metabolites and excluding the false positives and endogenous metabolites interference in a complex biological sample to profile biological fluids for small molecules, it could provide substantial insight into drug metabolism pathways concerning pharmacology Zhao et al. 2019). In this study, we employed ultra-high-performance liquid chromatography (UHPLC)-mass spectrometer (MS)-based metabolomics to compare a series of cannabinoid compounds (THCA, THCV, and THC). A total of 29 drug metabolites have been identified by metabolomics, including 12 THCA metabolites and 17 THCV metabolites. The metabolite V17 was produced from the formation of THCV and cysteine (Cys). In silico ADMET prediction revealed that THCA and THCV metabolites might show potential human hepatotoxicity (H-HT). Moreover, biological assay proved THCA and THCV, as well as their main metabolites (THC and CBD) capable of different anti-inflammatory activities against LPSinduced macrophages RAW 264.7. This study provides new insights into the relationship between the metabolism and biological activity of two phytocannabinoids (THCA and THCV).

Metabolism of THCA and THCV in MLMs, HLMs
In vitro metabolism of THCA and THCV was carried out in 96well plates and each incubation was conducted in triplicates. Co-incubations system (180 lL) contained 25 lM THCA or 25 lM THCV, individually with pooled HLMs or MLMs at 0.5 mg protein/mL PBS (pH ¼ 7.4). Incubation of microsomes with THCA, THCV, and THC was operated to start metabolic reactions in 37 C at 40 min. The HLMs and MLMs incubation methods were previously described methods (Rao et al. 2022). Five microlitres aliquot of the supernatant was collected for LC-MS characterisation using a Thermo Scientific TM Vanquish TM flex UHPLC Q-Exactive MS with the orbitrap resolution of 70 000.

Metabolic rates %
where A is the relative abundance of parent compound not containing NADPH and B is the relative abundance of parent compound containing NADPH.

Cell cultures and THCA/THCV treatment
The animal study was performed following the Institute of Laboratory Animal Resources guidelines and approved by the Institutional Animal Care and Use Committee of West China Hospital, Sichuan University (no. 20211527A). Mouse primary hepatocytes were separated from male C57BL/6 mice by a two-step in situ collagenase perfusion method (Hu et al. 2018). The MTT assay was used to measure cell viability after treatment with THCA and THCV in primary hepatocytes (Supplementary Figure 1). After being cultured for 8 h at 37 C with 5% CO 2 , primary hepatocytes were treated with 25 lM THCA or THCV for 24 h at 37 C. Cell mixtures were collected, vortexed, and centrifuged at 1500 rpm for 5 min at 4 C. Control groups were established in the absence of drug treatments.

Metabolite extraction
The suspension of mouse liver primary hepatocytes cells, containing both the cell pellet and supernatants, was extracted with isovolumetric ethyl acetate three times. The combined organic extracts were vortexed and centrifuged at 13 000Âg for 20 min at 4 C. Next, the extracts were dried under nitrogen flow. The residue was re-dissolved in 100 lL acetonitrile and centrifuged at 13 000Âg for 15 min at 4 C. The incubation supernatant of mouse primary hepatocytes was injected 5 lL into ultra-high-performance liquid chromatography Q-Exactive TM Hybrid Quadrupole-Orbitrap TM Mass spectrometer (UHPLC Q-Exactive MS) for analysis. The incubation was conducted in triplicates.

UHPLC Q-Exactive MS analysis
The microsomal analysis was conducted by the UHPLC Q-Exactive (Thermo Fisher Scientific, San Jose, CA) system and analyte separations were achieved on an HSS T3-C18 column (2.1 Â 100 mm, 1.8 lm) at 25 C. The analytes were eluted with a gradient elution of 0.1% formic acid in water (A) and acetonitrile (B). The gradient procedure was performed as follows: 1% B at 0-1 min, 1-70% B at 1-3 min, 70-100%B at 3-15 min, and 100-1%B at 15-18 min. Other conditions were set as follows: flow rate: 0.3 mL/min, capillary temperature: 350 C, gas flow: 15 L/min, sheath gas flow rate: 35 L/min, capillary voltage: 3.7 kV, and injection volume: 5 lL. The mass spectrometer was operated in the positive mode. The MS/MS fragment information was detected with the following MS/MS conditions: the collision energies were set at 15, 25, and 35 eV, respectively. The orbitrap resolution was set at 70 000. Xcalibur software (Version 4.3, Thermo Fisher Scientific, San Jose, CA) was used to control the LC-HRMS system and for data acquisition.

Multivariate data analysis for screening metabolites and metabolite identification
Chromatographic analysis of THCA and THCV metabolites and data transformation were performed using Thermo Xcalibur Qual Browser software (Thermo Fisher Scientific, San Jose, CA) and MS-DIAL software. The parameters were set as follows: retention time range ¼ 0-15 min, minimum peak height ¼ 10 000, and the parent ions of [M þ H] þ were chosen for screening metabolites in the positive mode. The processed data matrix was exported as an excel file containing the peak number, retention time, m/z, and integrated peak area. Discriminatory features were characterised using principal component analysis (PCA) with SIMCA 14.1 software.
In silico ADMET prediction of THCA, THCV, and their metabolites Metabolic behaviour prediction was performed using ADMET lab 2.0 (https://admetmesh.scbdd.com/) to systematically evaluate the ADMET properties of the drug and its metabolites (Xiong et al. 2021). Drug absorption transporter P-gp and blood-brain barrier (BBB) penetration were applied to predict drug absorption and the ability of transport to the brain. The toxicity model comprises submodules for cardiac toxicity, hepatotoxicity, AMES mutagenicity, and skin sensitisation. Cardiac toxicity parameters were specifically evaluated by relevant hERG blockers. After these models were accumulated, the criteria of affinity behaviour between THCA/THCV and all tested indexes were assessed and scored.
Cell culture and NO production assay RAW 264.7 cells were obtained from the West China Hospital. MTT assay was employed to measure the cytotoxicity of the four compounds in RAW 264.7 cells. Cells were grown in DMEM containing 10% heat-inactivated FBS with 100 units/mL penicillin-streptomycin in a humidified atmosphere of 5% CO 2 at 37 C. Cells were plated into 96-well plates at a density of 4 Â 10 4 /well and cultured overnight. Cells were pre-treated for 2 h with THC, THCA, CBD, and THCV at concentrations of 5 lM, respectively. One hundred micromolars of DEX was used as the positive control. Then, LPS (1 lg/mL) was added to stimulate inflammation for 24 h (Park et al. 2020). The collected supernatant was used to determine the concentrations of NO according to kit protocol.

RNA isolation and quantitative reverse transcription PCR (QPCR)
RAW 264.7 cells were plated into 12-well plates at a density of 1 Â 10 6 /well at 37 C in a humidified incubator containing 5% CO 2 to cultured overnight. Cells were pre-treated for 2 h with THC, THCA, CBD, and THCV at concentrations of 5 lM, respectively. One hundred micromolars of DEX were used as the positive control. Then, LPS (1 lg/mL) was added to stimulate inflammation for 24 h. After being cleaned twice with PBS, the RAW 264.7 cell pellets were centrifuged with 1000 rpm, 5 min at 4 C. Total RNA was obtained from the RAW 264.7 cell pellets using TRIzol reagent on ice, reverse transcribed to cDNA, and subsequently subjected to QPCR. QPCR was employed to determine the expressions of iNOS and Ilb mRNAs. The program for amplification was one cycle of 95 C for 30 s, followed by 40 amplification cycles of 95 C for 10 s, 55 C for 30 s, and 72 C for 40 s. The PCR primers used in this study were listed as follows: (1) iNOS: GTTCTCAGCCCAACAATACAAGA (forward), CTGGA CGGGTCGATGTCAC (reverse); (2) Il1b: CCCTGCAGCTGGAGA GTGTGGA (forward), TGTGCTCTGCTTGTGAGGTGCTG (reverse); (3) Gapdh: TTGATGGCAACAATCTCCAC (forward), CGTCCCGTA GACAAAATGGT (reverse). Relative gene expression levels were normalised to Gapdh expression levels.

Statistical analysis
All data were presented as means±SD. The differences between the means of the individual groups were assessed by one-way analysis of variance (ANOVA) with Duncan's multiple-range test. Between-group comparison was performed using Student's t-test. Differences were considered to be significant when p < 0.05.

Metabolomic profiling of HLMs, MLMs treated with THCA or THCV
The profiling of metabolomic analysis on the positive ions produced by Q-Exactive MS assay of THCA, THCV, and THC groups in HLMs system is shown in Figure 2. The PCA revealed three clusters (Figure 2(A)) corresponding to the control, THCA and THCV groups in the score plots, which indicated metabolic differences for these two structurally similar cannabinoids compared to the control group. The loading scatter plots (Figure 2(B)) generated from OPLS-DA showed significant differences among the metabolites of these three groups in HLMs. The top-ranking ions were identified as THCA, THCV, and their metabolites, which were marked in the loading scatter plots. Most of the THCA and THCV metabolites were found in both HLMs and MLMs systems (Tables 1 and 2). The relative abundance of discriminatory metabolites was visualised with tread plots, exemplified by THC (m/z 315.2317 þ ) and CBD (m/z 315.2319 þ ) that was typical in the THCA-treated group and THCV-treated groups, respectively (Figure 2(C,D)).
Comparison of THCA and THCV metabolism in HLMs, MLMs, and primary hepatocytes A total of 29 metabolites of THCA and THCV were identified by the analysis of data from positive mode. Twelve of the metabolites in phase I were derived from THCA, which were all produced in MLMs (Table 1). The metabolic rate of THCA was 11% in HLMs, while that of THCV was 31% (Figure 3(A)). Similarly, the metabolic rate of THCA was 31% in MLMs, while that of THCV was 96%. These results confirmed the metabolic rate of THCA is higher than THCV in liver microsomes while MLMs are much higher than HLMs.
Among these THCA metabolites, six metabolites of THCA were consistently found across all three systems, 10 metabolites were found in HLMs and MLMs, and three metabolites were found in MLMs and mouse primary hepatocytes. In THCV metabolism, a total of 17 THCV metabolites were identified in HLMs, MLMs, and primary hepatocytes, including 13 metabolites found across all three systems (Table 2). V1-V17 were consistently found in MLMs system, whereas V17 was only observed in MLMs and mouse primary hepatocytes.

Metabolic identification of THCA and THCV
A total of 29 metabolites of THCA and THCV were identified in HLMs, MLMs, and mouse primary hepatocytes. A chromatogram of the representative THCA and THCV metabolites is presented in Figure 4. Among these metabolites, seven of THCA metabolites (A2-4, A7, A9, and A11-A12) and 10 of THCV metabolites (V5-V10, V13, and V15-V17) were not previously described. The structures of THCA and THCV metabolites were identified by detailed analysis of THCA and THCV parent compounds based on accurate mass measurements and their MS/MS spectrum. The same metabolic pathways of THCA and THCV were monohydroxylation, dihydroxylation, ketone, and aldehyde reactions, which yielded metabolites A2-A12, V3-V7, and V10-V15. Cys-THCV adduct (V17) and demethylation metabolites of THCV (V8-V9) were the unique products from the metabolic pathway of THCV. Additionally, both THCA and THCV were able to form THC by metabolic pathway, but THC in THCV metabolism without b-nicotinamide adenine dinucleotide 2 0 -phosphate reduced (NAPDH) was further metabolised to CBD ( Table 2). The MS/MS spectra of THCA and THCV and their representative metabolites from their fragmentation pathways are presented in Figure 5

Identification of hydroxylation in THCA and THCV metabolism
The chemical formulas of metabolites A2-A6 were C 22 H 30 O 5 according to the observed [M þ H] þ at m/z 375.2148 þ to 375.2163 þ . A2-A6 were the isomers and were eluted successively from 5.91 to 8.76 min. The diagnostic product ion at m/z 357 was detected by the loss of 18 Da, suggesting that a molecule of H 2 O was lost from a protonated molecular ion. The possible positions of the oxidation reaction for five metabolites (A2-A6) may occur at 4 0 , 2 0 , 5 0 , 11, and an  unidentified position, respectively. Similarly, metabolites V3 and V4 were observed in the extracted chromatogram from m/z 303.1956 þ and 303.1950 þ , which were 18 Da (1 H 2 O) higher than that of THCV, indicating that these metabolites were the hydroxylated products of THCV. The formation of A7-A8 and V3-V4 indicated that they were produced in the metabolism of THCA and THCV, respectively. Those metabolites were 32 Da (two oxygen atoms) higher than the parent structure (Supplemental Figures 3 and 4), which implied that the positions of dihydroxylation might happen to alkyl side chain or C-11.

The unique metabolites of THCV metabolism
In the incubation of THCV in HLMs, MLMs and mouse primary hepatocytes, two demethylated metabolites (V8 and The relative abundance of THCV metabolites in HLM, MLM, and mouse primary hepatocyte. All samples were analysed by UHPLC Q-Exactive MS. The sum of peak areas of identified metabolites was integrated as 100%. The data were represented as the means ± SD (n ¼ 3). Statistical analysis between the groups was conducted using Student's independent t-test. ÃÃ p< 0.01, and ÃÃÃ p< 0.001.

V9
) were observed and characterised in these three systems (  Figure 4). The Cys conjugated position could occur at the benzene ring.

ADMET lab
To evaluate the properties of THCA and THCV metabolites with exact structure in drug discovery and hazard risk assessment, ADMET lab 2.0 in silico was employed to systematically assess the risk of drug toxicity and safety. The results of in silico toxicological profile of THCA, THCV, and their metabolites are presented in Figure 6.
On the ADMET predictor, the parameter of hERG, H-HT, DILI, AMES toxicity, skin sensitisation, blood-brain barrier penetration (BBB penetration), and P-glycoprotein substrate (P-gp-substrate) were evaluated by Multi-task Graph Attention framework (MGA) to precisely predict drug properties and toxicity (Xiong et al. 2021). Its predictive results, like hERG, H-HT, etc., were displayed by the parameter values and possessed different evaluation ranges. In these evaluation systems, red to green represents the activity of compounds from high to low. For example, the green and black (CBD, V9 and V12) showed lower H-HT than other metabolites in H-HT evaluation. Additionally, 10 compounds with red (THCA, A2-A5, A9-A12, and V3) represent higher DILI. It represents that THCA and its metabolites predicted that most metabolites showed a higher potential risk of hepatotoxicity than THCV metabolites. However, the specific hepatotoxicity still needs to be further verified. It was anticipated that THCV metabolites were the P-gp inhibitors except for V6, V7, and V17, which could impact drug bioavailability. Notably, THCV was forecast to be a higher BBB penetration than THCA. BBB penetration is a drug distribution parameter for the central nervous system, and effective brain exposure to medications could influence the therapeutic effect (Reichel 2009;Vastag and Keseru 2009). The BBB penetration of THCV and its metabolites (V3, V6, V7, V13, and V16) were higher than THCA and its metabolites (A2-A5 and A9-A12), suggesting that the metabolism of THCV might improve brain exposure and activate the central nervous system.
Anti-inflammatory activities of THC, THCA, THCV, and CBD in RAW 264.7 macrophages The viability of RAW 264.7 cells measured by MTT assay was used to evaluate the cytotoxicity of THC, THCA, THCV, and CBD. The result showed no significant difference in cell viability between the control group and the groups treated with various concentrations of THC, CBD, THCA, and THCV, indicating that they did not affect the normal cell growth at the concentration of 3.125-25 lM (Supplemental Figure 2).

Inhibition of THC, THCA, THCV, and CBD on LPS-induced NO production in RAW 264.7 cells
The anti-inflammatory activities of THC, CBD, THCA, and THCV on LPS-induced NO production in RAW 264.7 cells were investigated, NO production assay was used in this study. Compared with the control group, the concentrations of NO were significantly increased in the LPS group, and it was suppressed by the THC group at the concentration of 5 lM (Figure 7(A)). The results showed that only THC inhibited NO production induced by LPS in RAW 264.7 cells among these four compounds.

Inhibition of THC, THCA, THCV, and CBD on LPS-induced over expression of pro-inflammatory mediators
To compare the anti-inflammatory effects of THC, CBD, THCA, and THCV in RAW 264.7 macrophages, the mRNA expression of NO synthase (iNOS) and interleukin-1 (Il1b) was determined by QPCR. These results revealed that THC and CBD effectively inhibited iNOS mRNA overexpression induced by LPS (Figure 7(B)). Moreover, the results suggested that THC, CBD, and THCV remarkably suppressed macrophages inflammatory cytokine Il1b mRNA level at a concentration of 5 lM (Figure 7(C)). Therefore, THC and CBD showed more potent anti-inflammatory activities than THCV and THCA.

Discussion
THC, THCA, THCV, and CBD were the representative phytocannabinoids found in Cannabis sativa L with a similar structure characterised by an alkylresorcinol-monoterpene moiety (Gaston and Friedman 2017). THCV, a homologue compound of THC, differing by just a propyl side chain may act as a CB1 and CB2 receptor agonist, the same activation as THC, albeit the activation is lower than THC (Pertwee 2008). THCA is the acidic precursor of THC, which has poor stability and rapidly decarboxylates to a stable form THC. Additionally, our ADMET lab prediction identified THCA with extremely poor brain penetration, which was consistent with the previous report (Anderson et al. 2019). Numerous phytocannabinoids easy penetrate the BBB, and it has been previously shown that CBD was a stronger ability to penetrate the BBB by oral administration than THCV in vivo. Although the allergy to Cannabis sativa was first reported more than 40 years ago, there are few studies about its allergy. The prevalence of individuals who smoke the extract of cannabis reached to 14.6%, and the prevalence rate of skin prick test (SPT) by accepting cannabis leaf extract is 8.1%, which demonstrated that cannabis has skin sensitisation. However, it is speculated that component of cannabis has different skin sensitivity according to our ADMET lab prediction results. The map of the metabolic pathways of THCA ( Figure 8) and THCV ( Figure  9) was constructed based on a comprehensive profile of THCA-and THCV-related metabolites. Figure 6. ADMET lab prediction of THCA and THCV metabolites. hERG blockers: human ether-a-go-go related gene blockers; H-HT: human hepatotoxicity; DILIL: drug induced liver injury; AMES toxicity: Ames mutagenicity toxicity; BBB penetration: blood-brain barrier penetration; Pgp-substrate: P-glycoprotein substrate.
There was a significant difference in THCA metabolism between HLMs and MLMs systems by using the metabolomics approach. Twelve THCA metabolites (A1-A12) and 17 THCV metabolites (V1-V17) were identified in HLMs, MLMs, and mouse primary hepatocytes incubation systems. THCA and THCV shared the same metabolic reactions, including hydroxylation, dihydroxylation, oxidative to ketone and aldehyde, while the relative abundance of their oxidative metabolites was slightly different. Comparing the relative amounts of THCV oxidative metabolites (V10-V16) with that of THCA oxidative metabolites (A9-A12) indicated that the substituent of the C-2 carboxy group resulted in the different metabolic capacities of THCA and THCV. Aldehydes are one type of highly reactive molecule, and the endogenous and exogenous metabolites with aldehyde groups are implicated in a variety of human pathologies (Jinsmaa et al. 2009;Doorn et al. 2014). Their ability to react with biological macromolecules, such as proteins, to produce covalent adducts is acknowledged to be a common major mechanism of drug toxicity (Grilo et al. 2014). In addition, the number of THCA hydroxylated metabolites was more than that of THCV hydroxylated metabolites, which suggested that the number of side chain alkyl groups also affected the metabolic rate of hydroxylation and oxidation. Hydroxylation reaction is a correlation with the level of oxidative stress, such as 6-hydrolated dopamine, which was demonstrated that have an electrophilic effect and it could induce oxidative stress to reduce glutathione level by inhibiting antioxidant N-acetyl cysteine (Urano et al. 2018). Cysteine, known as an essential amino acid, is the precursor of GSH that is an important antioxidant and free radical scavenger that could drive harmful toxins to be eliminated from the body (Yin et al. 2016). In addition, it was previously reported that the sulphur-containing amino acid cysteine was particularly susceptible to reactive oxygen species (ROS) and reactive chlorine species (RCS), which could have destructive effects on protein structure and activity through oxidative damage even lead to cell death (Ezraty et al. 2017).
It is universally acknowledged that drug metabolic activation was mediated by hepatic cytochrome P450 and gut microbiota to produce chemical metabolites of pharmacology activity. On one hand, the biotransformation of a drug in metabolic activation was a way to induce hepatotoxicity, nephrotoxicity and pulmonary toxicities through conjugation with proteins or DNA in the formation of reactive or toxicity metabolites (Leung et al. 2012;Wang et al. 2021). For example, diosbulbin B is a common hepatotoxic compound in Chinese herbal medicine Dioscorea bulbifera L., and its metabolites were easily captured by cis-enedial protein through Cys, Schiff's, or Cys/Lys adducts (Wang et al. 2017). On the other hand, numerous of researchers isolated some activity metabolites (exogenous and endogenous) with potential utility in the pharmacology industry (Stresemann and Lyko 2008;Fan et al. 2020).
It was reported that many phytocannabinoids possessed a broad range of pharmacological roles (Devinsky et al. 2014). Inflammation is a complicated protective response that protects the body from harmful factors by triggering multiple factors. Cannabinoid compounds have been proven to be immune modulators, which could change the balance of pro-inflammatory and anti-inflammatory cytokines and inhibit cell-mediated immunity in different physiological systems Relative abundance of iNOS and Il1b mRNA level on LPS-induced. All data represented the mean ± SD of three independent experiments in triplicate. Ã p< 0.05, ÃÃ p< 0.01, and ÃÃÃ p< 0.001 vs. LPS group. (Greineisen and Turner 2010). The present investigation showed that THCV possessed anti-inflammatory activity by inhibiting the production of NO in macrophages (Romano et al. 2016). Besides, THCA was found to display neuroprotective activity by activating peroxisome proliferator-activated receptor-c (PPARc) and it could up-regulate proinflammatory markers induced by 3-nitro propionic acid (Nadal et al. 2017). However, the difference in anti-  inflammatory activities of THCA, THCV, THC, and CBD has been inconclusive (Arulselvan et al. 2016). In this study, we demonstrated that the anti-inflammatory effect of THC is more potent than CBD, THCA, and THCV in LPS-stimulated RAW 264.7 macrophages.

Conclusions
In summary, our study comprehensively analysed and compared metabolic intermediates and products of THCV and THCA in three systems in vitro. A total of 29 drug metabolites were identified by using UHPLC-MS. The metabolic rate of THCV in the liver microsome was higher than that of THCA. ADMET lab prediction highlighted that BBB penetration parameters of THCV and its metabolites were higher than THCA. In addition, THCV exhibited a more potent anti-inflammatory effect than THCA in LPS-induced RAW 264.7 cells. There results will support further studies on the pharmacological role and therapeutic potential of THCA and THCV.