Cannabis sativa L. is an indigenous plant originating from Central Asia. It has been used for medicinal uses for centuries and is a significant source of phytocannabinoids. Today, its compounds are being used as antiemetic drugs for cancer patients. In the United States, patients with advanced stages of cancer are given cannabis medicine as part of palliative care to manage symptoms such as chronic pain and nausea and to stimulate appetite. A cross-sectional study by the National Cancer Institute in 2017 reported that 24% of adult cancer patients in Washington are active users of cannabis to relieve cancer treatment-related symptoms, whereas 74% expressed interest in gaining more information about cannabis usage from cancer providers [10]. THC, CBD, and THCV are some of the most studied and abundant bioactive compounds found in Cannabis sativa L. These compounds have been demonstrated by in vitro studies to attenuate cell proliferation, invasion, and angiogenesis of human cancer cell lines such as human glioma [11], breast, and prostate cancers [12].

Moreover, they have 23 shown antitumor activity in numerous rodent models, which further establishes their potential as anticancer agents.

In silico approaches such as CADD and virtual screening are utilized to undertake ADME/Tox studies since it allows for efficient screening and synthesizing of a large number of drug-like compounds in a shorter period. Vimblastine, vincristine, taxol, podophyllotoxin, captothecin and cytarabine are some examples of successful anticancer agents developed through CADD techniques [13]. In the present study, the ADMET properties, toxicity profiles, and pharmacological activities of THC, CBD, and THCV were predicted using several free web servers to provide an initial evaluation of prospective drug candidates in the hit-to-lead stage in drug discovery and later, lead-optimization.

There is scant research and data on the physicochemical and pharmacokinetic properties of cannabinoids in humans; therefore, there is not enough understanding of factors that influence their drug action and effectiveness. SwissADME is a free, simple, robust, and accurate web tool to determine the ADME parameters of small compounds. The physicochemical properties, lipophilicity, water solubility, pharmacokinetics, drug likeness, and medicinal chemistry properties of THC, CBD, and THCV were predicted to assess their potential as anticancer agents and integrate these findings in modern drug development. SwissADME provides a one-panel-per-molecule output that comprises all the physical and chemical information of the molecules entered. Although there are variations in the computed findings, the projected physicochemical characteristics of THC, CBD, and THCV follow a regular pattern, with little difference in each parameter.

The freely available pkCSM-pharmacokinetics web tool relies on graph-based signatures and experimental data to predict and optimize small-molecule ADME/Tox properties [14]. The investigated compounds (THC, CBD, and THCV) must be non-carcinogenic and non-hepatotoxic. Accordingly, the pkCSM web server along with SwissADME were utilized to estimate the pharmacokinetics and toxicity of the aforementioned compounds. The pkCSM server makes it possible to predict ADME/Tox properties quickly and accurately.

SwissADME was used to predict the physicochemical properties of THC, CBD, and THCV. The transportation of an unmetabolized drug from the site of delivery to the body’s circulatory system is known as drug absorption. Some patients opt for their chemotherapeutics to be delivered orally, instead of intravenously due to low cost and lack of infusion-related inconveniences [15]. However, in terms of pharmacokinetic considerations, this route of administration does not allow for fast and efficient absorption since they are affected by the first-pass metabolism that occurs in the intestine and liver [16]. In addition, cancer drugs generally have low and highly variable oral bioavailability, rendering the drug ineffective. The bioavailability radar of the screened cannabinoids is depicted in Figure 3, which provides a graphical depiction of the drug-likeness of the submitted molecules. The physicochemical properties that characterize the oral druggability of the selected cannabinoids include lipophilicity, size, polarity, solubility, and saturation. The pink region in the diagram represents the ideal range for these parameters. The aforementioned properties were analyzed based on the guidelines of Lipinski’s rule-of-five (L-Ro5) [17] for drug-likeness. L-Ro5 is a set of criteria for determining whether a compound has the qualities to be regarded as a good oral drug.

The general characteristics of the compounds shown in Table 2 revealed that all compounds had MW ranging from 286.41 to 314.46 Da, which is within the acceptable MW (≤ 500 Da) set by L-Ro5 to be considered an orally available bioactive drug. Lipophilicity describes the permeability of compounds through biological systems. The lipophilicity property is measured by the partition coefficient P (log Po/w) of the compounds between n-octanol and water systems [18]. In SwissADME, lipophilicity is estimated as consensus Log P (Table 4), which is the average value of all Log P evaluated with various lipophilicity criteria. Among the compounds screened, only THCV met the lipophilicity cutoff (log P ≤ 5) set by L-Ro5 with a value of 4.58. In contrast, THC and CBD with 5.28 and 5.20 lipophilicity values, respectively, were out-of-range. High lipophilicity indicates a high insolubility rate and poor oral absorption. In addition, the water solubility of the compounds is ranked from highly water-soluble (THCV) to least water-soluble or insoluble (THC and CBD), respectively, in fats, oil, lipids, and nonpolar solvents [19].

The topological predicted polar surface area (TPSA) was 29.26 Å2 for THC and THCV and 40.46 Å2 for CBD. TPSA values should be between 20 to 130 Å2 to have a high probability of oral bioavailability [20]. In addition, Lipinski (1997) [17] suggested that less than 140-150 Å2 TPSA commonly exhibits an adequate intestinal permeability, which means that the obtained TPSA for the said compounds was not highly acidic. The penetration of these compounds into the cell membranes is not restricted. All compounds were found to satisfy L-Ro5 for the number of HBDs (≤5), HBAs (≤10), and rotatable bonds (≤10) (Table 3), but all compounds violated the cLogP [17] (Table 4). Nevertheless, the results obtained from L-Ro5 may indicate the potential for increased oral absorption [21]. For saturation, the ratio of sp3 hybridized carbons to total molecule count of fraction Csp3 should be at least 0.25 [14], in which all compounds met with values ranging from 0.52 to 0.62 (Table 3). Another parameter assessed to determine the oral bioavailability is molar refractivity (MR) as described by the Ghose filter [22] and molar solubility (log S). According to L-Ro5, MR should be between 40-130 for drug-likeness [17], while log S should not exceed 6 [23]. The cannabinoids had MR values ranging from 88.30 to 99.85 (Table 3). On the other hand, based on the calculations by the ESOL, Ali, and SILICOS-IT model, as shown in Table 5, the cannabinoids had log S values are within the accepted range; however, the aqueous solubility equation adapted from Ali et al. (2012) [24] shows that all compounds are poorly soluble.

The absorption properties of each compound based on computed lipophilicity and polarity of small molecules were depicted by the Brain or IntestinaL EstimateD permeation method or BOILED-Egg model shown in Figure 4. The BOILED-Egg model is an accurate predictive method to evaluate the passive gastrointestinal absorption (HIA) and blood-brain barrier permeation (BBB) of drug-like molecules [25]. Molecules found in the white region have a high probability for HIA, while the yellow region or yolk has a high probability for brain penetration [25]. Results showed that the positioning of the compounds was seen to be close and similar to one another. THC, CBD, and THCV were found on the yolk or yellow region of the BOILED-egg model, implying that they exhibit good BBB permeation and high GIT absorption. In addition, the blue dots represent P-gp substrates (PGP+) and the red dots for P-gp non-substrate (PGP-) [25]. All compounds were represented by a red dot, indicating that they are non-P-glycoprotein (Pgp) substrates; therefore, they will not be subjected to being actively effluxed out of the cells.

Drug-likeness and medicinal chemistry properties of THC, CBD, and THCV are shown in Tables 6 and 7. SwissADME gives access to five different rule-based filters: Lipinski (Pfizer) filter, Ghose (Amgen) filter, Veber (GSK) filter, Egan (Pharmacia) filter, and Muegge (Bayer) filter [14]. These filters come with a diverse range of properties inside of which the molecule is defined as drug-like. Most of the compounds followed the filtered rule invoked in SwissADME, and the violation exhibited by the molecules are minimal. All compounds did not have any PAINS alert and passed L-Ro5 for drug-likeness; however, THC and CBD had one violation due to their MLOGP values exceeding the lipophilicity threshold of >4.15. To broaden lead optimization prospects, the Brenk filter considered molecules that are smaller and less hydrophobic, as well as those that are not characterized by L-Ro5 [26]. Compounds with potentially mutagenic, reactive, and unfavorable groups are excluded [26]. All compounds had one alert in Brenk filter due to the presence of isolated alkene in their compounds. Moreover, all compounds failed the leadlikeness criteria due to a violation in their out-of-range XLOGP3 values (should be in the range –0.7 to +6.0). Finally, the synthetic accessibility (SA) score estimates the easiness of the compounds to be synthesized. SA scores range from 1 (very easy to synthesize) to 10 (very difficult to synthesize) (Daina et al., 2017) [14]. The cannabinoids had SA scores ranging from 4.05 to 4.27, suggesting that they could be relatively easy to synthesize.

The absorption, distribution, metabolism, and excretion (ADME) profiles of THC, CBD, and THCV according to pkCSM web server are presented in Table 8. Absorption is the movement of a drug into the systemic circulation from an extravascular administration site. Compounds with values less than (more negative than) −6 in the pkCSM Log mol/L scale are considered poorly soluble [27]. Water solubility (log mol/L) values for the compounds THC, CBD, and THCV were predicted to range from −4.901 to −6.275, with CBD and THCV having high solubility and THC having low solubility, indicating poor absorption and elimination by the urinary tract, as shown in Table 8. The data shows high negative values, which indicate the presence of moderately to poorly soluble compounds. The pkCSM solubility predictions are consistent with results obtained from SwissADME.

The Caco-2 cell line, which is made up of human epithelial colorectal adenocarcinoma cells, is widely used as an in vitro model of the human intestinal mucosa for predicting drug absorption by measuring the log of the evident permeability coefficient (log Papp; log cm/s). A compound with a log Papp value >0.90 cm/s is considered to have high Caco-2 permeability [28]. Table 8 shows that all phytocannabinoids have a high Caco-2 permeability. All compounds have nearly the same Caco-2 permeability value, which is predicted to be 1.5 cm/s, except for cannabidiol (1.7 cm/s).

Furthermore, all compounds have a good profile of absorption rate in the intestine. Interestingly, the human intestine absorption (HIA) values are high, indicating that phytocannabinoids derivatives have a greater than 90% chance of being absorbed by the human intestine, with tetrahydrocannabivarin having the highest probability (93.772) and cannabidiol having the lowest (90.657). According to Cerqueira et al. (2015) [29], the recommended value of the skin permeability (log Kp) for a drug molecule, an important consideration for improving drug efficacy and is particularly pertinent in developing transdermal drug delivery is set at more than -2.5 cm/h. All compounds have log Kp values ranging from -2.4 to -2.7 cm/h. THC and CBD compounds are predicted to have good skin permeability with 2.538 cm/h and 2.795 cm/h, respectively. THCV, on the other hand, has a lower skin permeability of 2.441 cm/h. The bioavailability score of 0.55 confirms that all investigated compounds have good absorption since they may have more than 10% of bioavailability in the rat [30].

The volume of distribution at steady state (VDss) and the blood−brain barrier (BBB) are two important parameters to consider when evaluating a drug’s ability to be distributed in the body. The higher the VD is, the larger the amount of a drug is distributed to tissue rather than plasma. Pires et al. (2015) [28] reported that a compound has good distribution if its VDss value is greater than 0.45. All the compounds have VDss values that are twice as high as the recommended value. In terms of the BBB, which defines a drug’s ability to cross into the brain while improving efficacy (fewer side effects), a compound is capable of passing through the blood-brain barrier when log BB is greater than 0.3. The pkCSM web server reveals that THC and THCV can easily cross the BBB except for CBD. Since the log BB value of CBD is less than 0.3, hence, it can only penetrate the blood-brain barrier moderately. BBB permeability results for THC and THCV are higher than 0.3, suggesting that this compound has a high probability of being absorbed by the gastrointestinal tract and permeating into the brain. Testing with pKCSM gives results that are slightly different from SwissADME, especially on the distribution parameters.

Metabolism was also predicted using the pkCSM web server in accordance with the inhibition of the main cytochromes (CYP) of the P450 superfamily, namely CYP1A2, CYP2C19, CYP2C9, CYP2D6, and CYP3A4. A common mechanism for metabolism-based drug–drug interactions is CYP enzyme inhibition, which involves competition with another drug for the same enzyme binding site. All clinically used drugs, including several anticancer agents, are impaired by enzyme inhibition, resulting in higher plasma levels of drugs that influence the therapeutic outcome. The effect is reduced if the drug is a prodrug. Therefore, inhibition of CYPs can result in drug toxicity or ineffectiveness [31]. According to Thorn et al. (2012) [32], CYP1A2 is involved in the metabolism of xenobiotics in the body. In contrast, CYP2C19 is responsible for the metabolism of several drugs as well as the detoxification of carcinogens and bioactivation of some environmental procarcinogens [33]. CYP2C9, on the other hand, is the primary enzyme that metabolizes drugs with a narrow therapeutic index [34]. CYP2D6 is highly polymorphic, and its metabolism is variable; people with reduced or no activity of this enzyme would be at risk of reduced efficacy of drugs or present adverse effects [35]. Lastly, CYP3A4 is responsible for the metabolism of drugs, carcinogens, steroids, and eicosanoids [36]. Based on the results obtained, THC, CBD, and THCV were all predicted to be CYP2C19, CYP2C9, and CYP2D6 inhibitors. Additionally, the results of metabolism prediction for THC and THCV are promising, as all compounds were found to have a low likelihood of inhibiting the CYP1A2 enzyme. The findings also showed that both THC and THCV compounds are non-inhibitors of CYP3A4 enzyme, except for CBD, which was found to be an inhibitor of the said enzyme.

The excretion profiles of the three compounds were predicted using total clearance and OCT2 (organic cation transporter2) substrate renal. Total clearance (log ml/min/kg) is a combination of hepatic and renal clearance. OCT2, on the other hand, is a renal uptake transporter that plays a vital role in drug uptake, disposition, and clearance of drugs in the kidneys. Hence, the total clearance is proportional to renal OCT2. OCT2 evaluation of a candidate compound’s transfer provides information on both its clearance and potential contraindications [37]. The pKCSM prediction results (Table 8) show that the total clearance of cannabidiol is the highest, followed by tetrahydrocannabinol, while tetrahydrocannabivarin has the lowest. This means that tetrahydrocannabivarin has the highest bioavailability. The findings also revealed that none of the compounds are OCT2 substrates. From the results above, it can be inferred that these phytocannabinoids are excreted through the kidney via a mechanism other than OCT2.

Toxicity refers to how much a substance can harm an organism or its organs, such as cells and tissues (Duran-Iturbide et al., 2020) [38]. Vital parameters such as the Ames test, hepatotoxicity, and oral rat acute toxicity LD50 values were used to predict the toxicity levels of THC, CBD, and THCV. The toxicity results show that none of the compounds are mutagenic and are unlikely to be hepatotoxic. On the other hand, all compounds obtained LD50 values that are in the range of 482-500 mg/kg/day, which fall into toxicity class 4 of the Globally Harmonized System (300 < Category 4 ≤ 2000 mg/kg/day), indicating that they are slightly toxic if swallowed and thus can be considered safe [9].

The pre-assessment of the toxic properties of a compound is essential for drug discovery, especially on the increasing number of compounds and their mixtures. There is also increased exposure of humans to chemicals, particularly to the abused usage of cannabinoids. THC, CBD, and THCV are the major psychoactive components of marijuana plants. Hence, Protox-II was utilized to predict the toxicity of these cannabinoid compounds. Protox-II results exhibited that it has active effects on immunotoxicity; three compounds are positive to have an activity on immune cells. Several pieces of literature supported the immunotoxicity effects where the amount of various immune system cells was reduced and the immune response of mice to influenza virus infection after the administration of the cannabinoids. In addition, pregnant mice were also subjected to a cannabinoid compound and noticed the immunosuppressive effect on the fetus. These immunotoxicity effects are due to the ligation of G protein-coupled receptors, CB 1 and 2. Varying expression of the receptors modifies the response to immune stimulation, although CB2 has a more significant expression than CB1.

Therefore, in silico results from Protox-II were confirmed by several in vitro experiments performed. The Protox-II results also demonstrated that CBD has immunotoxicity effects, supported by the study by Zwicker et al. (2015) [39]. In their study, the immunotoxic potential of CBD was evaluated using in vitro approach. Based on the proteomic, functional, and metabolomic data acquired, it was discovered that CBD has immunotoxic effects on T cells. According to Lee et al. (2008) [40], immune cell apoptosis is caused by the formation of reactive oxygen species that is induced by the application of CBD.

On the other hand, these cannabinoid compounds have shown inactive results on hepatotoxicity, carcinogenicity, mutagenicity, and cytotoxicity. In a study by Stohs & Ray (2020) [41], where epilepsy patients received a higher oral dose of the drug with cannabinoid compounds for 48 weeks, four patients withdrew from the study stating the reasons of diarrhea, fatigue, somnolence, and convulsions. Nevertheless, patients did not mention any elevated serum aminotransferases or hepatoxicity. Therefore, based on in vitro studies performed, there is a low incidence of hepatotoxicity caused by cannabinoids in low and therapeutic doses. Furthermore, cannabinoid compounds are also found to be inactive with carcinogenicity, as supported by the study of Melamede (2005) [42], discovering that cannabinoids initiate tumor regression and inhibit angiogenic factors in mice. In addition, in vitro studies in treating glioma have resulted to depreciating levels of VEGF with the administration of cannabinoids. This is due to the special immune regulatory activity of cannabinoids that down-regulates immunologically produced free radical by promoting Th2 immune cytokine profile.

Among the compounds, CBD has the highest probability of being inactive in mutagenicity, with a value of 0.85 compared to 0.77 for THC and 0.74 for THCV. The lack of mutagenic effects of CBD is also evident in the study of Jones et al. (2012) [43], wherein CBD was administered in rats. There has been no observation of significant toxicity, genotoxicity, or mutagenicity in vivo. As for THC, it has no mutagenic properties in vitro, according to Zimmerman et al. (1978) [44]. After exposure to THC to the cultured fibroblasts derived from healthy individuals and Xeroderma pigmentosum patients, the amount of chromosomal breakage and chromatid exchanges did not appear to have increased. Similarly, THCV was found to be non-mutagenic in the Ames test prediction [45]. Nonetheless, in terms of cytotoxicity, the results show that the three compounds are all inactive. Cannabinoids have selective cytotoxic properties against tumor cells while also preventing the apoptosis of healthy, normal cells [46]. Several studies in different cancer cell lines have demonstrated that these cannabinoids activate autophagy and ultimately causes cell death [47, 48]. Therefore, based on the predicted results generated from Protox-II that are supported by in vitro and in vivo studies, these cannabinoid compounds are not toxic and safe for cancer treatment uses.

Both cannabinoid and non-cannabinoid molecules contribute significantly to the cannabis extract’s pharmacological profile. Research among phytoplants contributed to developing the defined chemotypes of high-yield cannabis cultivars. Cannabis sativa L. is herbal medicine from which cannabinoid compounds like THC, CBD, and THCV can be derived. This herbaceous plant has at least 554 compounds, including about 150 terpenophenolic active molecules and phytocannabinoids that are only exclusive to this plant. Aside from this, it also contains various phytomolecular families such as oxylipins, flavonoids, and terpenoids. This has been a subject of interest mainly due to its pharmaceutical applications since it has metabolites that exhibit potent bioactivities on human health. The cannabis extracts contain large amounts of pharmacological properties, and two major cannabinoid molecules are responsible for this, namely cannabidiol (CBD) and tetrahydrocannabinol (THC) [49]. CBD is the primary molecule found in hemp, while THC is the main component of drug-type cannabis. THC is responsible for psychotropic activity, while CBD lacks this and is reported to significantly reduce adverse effects like anxiety and psychosis [50].

On the other hand, THCV is a minor cannabinoid that is known to stimulate alertness and euphoria among its users. Although cannabinoids are mostly known for their psychoactive components, non-psychoactive compounds such as CBD give new light to future research. They have been found moderately effective in clinical trials of multiple brain injury, neuropathic pain, multiple sclerosis, and arthritis [50].

LD50 is the drug’s concentration at which half of the cells undergo death. This implies that the higher amount of LD50 means lesser drug action making it necessary for there to be a higher drug concentration to cause death in cells or cause harmful effects. On the other hand, lower LD50 values mean that the drug action is more potent at lower concentrations.

The pharmacokinetics of CBD is complex, and the bioavailability of oral CBD is low across species due to extensive Phase I metabolism. Comprehensive studies in animals, including rodents, indicate that much of the administered CBD is excreted intact or as its glucuronide [51]. In acute rat toxicity, CBD falls mainly in Class 4 (Rat IP LD50, Rat IV LD50, and Rat Oral LD50). With this, intake by oral and intraperitoneal CBD is classified as non-toxic. However, in Rat SC LD50, it is in Class 3, which is categorized as slightly toxic upon intravenous intake.

For the rat toxicity prediction of CBD and THC, The LD50 value at rat oral is 799,200 mg/kg, making its concentration comparatively low than THCV. Moreover, THCV has the lowest LD50 concentration in rat oral which is 663,300 mg/kg. In LD50 concentrations, CBD and THC yield the highest value at rat IV-0,912 mg/kg) while THCV is within the rat IV (-0,841 mg/kg). The classification of the routes for the cannabinoids shows that THC has moderate toxicity in Rat IV and low toxicity in Rat IP, Oral, and SC. CBD has the same toxicity classification as THC. On the other hand, THCV is moderately toxic when administered subcutaneously compared to when it is administered intravenously, orally, or intraperitoneally as these routes have a classification of low toxicity. Based on these data, it can be inferred that cannabinoids, when administered intravenously, cause more toxicity compared to other routes at a given concentration as it is considered as least harmful when administered orally [52].

ADVER-Pred is a free online service for in silico prediction of adverse pharmacological effects of drug candidates, including arrhythmia, cardiac failure, hepatotoxicity, myocardial infarction, and nephrotoxicity on the cardiovascular and hepatobiliary systems. The estimated Pa and Pi values predict their probable activity or inactivity to show toxic effects. Compounds with Pa > Pi or high Pa and Pa-Pi values are considered active and have a greater probability of activity in in vivo studies or clinical trials [53]. Results shown in Table 11 revealed that all compounds had higher Pi values and lower Pa values, indicating that THC, CBD, and THCV are unlikely to lead to such adverse effects; thus, these compounds are designated to the subset of “inactive” in the PASS (Prediction of Activity Spectra for Substances) training set consisting of drug-like biologically active compounds. “Inactive” compounds in the PASS training set refer to those that lack definitive information about their biological activity found in an authoritative data source [54].

Pi values influence how these “inactive” correlate to their minimal adversity property. CBD had the highest probable activity (Pa= 0.128) among the compounds, which can lead to arrhythmia. In comparison, THCV had the highest probability of being inactive (Pi = 0.792), also corresponding to arrhythmia. THCV had the lowest activity probability of arrhythmia (Pa = 0.048) whereas CBD had the lowest inactivity probability of cardiac failure (Pi = 0.468).

Despite their high in silico predicted probability of inactivity values, there have been reports showing that there is still a possibility for THC, CBD, and THCV to be active upon intake and exert negative effects on human health. Findings from Holland (2020) [55] reported that THC and CBD cause no minimal side effects such as increased heart rate, coordination problems, fatigue, and diarrhea. However, studies have shown that high dosages of cannabinoids, particularly THCV, have a profound effect on specific brain nerve cells, which could lead to long-term complications [56]. Hence, it can be deduced that excessive intake of THC and CBD could potentially lead to adversity affecting the brain. Despite these reported side effects, cannabinoids are still medically used to treat related disorders due to their vast therapeutic properties.

CLC-pred web service was accessed to predict the cytotoxicity of drug-like compounds against tumor and normal human cell lines [57]. Pa>Pi threshold was used to determine active compounds. Table 12 reveals that CBD has the highest probability for their top cell lines with a Pa value of 0.542 and Pi value of 0.004 against breast adenocarcinoma (MDA-MB-453). This is comparatively high compared to the probability of THC against Glioma (U-251) with only 0.341 as its Pa value (Pi value = 0.048) and the probability of THCV against non-small cell lung carcinoma (HOP-18) with a Pa value of 0.369 (Pi value = 0.040). Following this, CBD has a probability of only 0.372 (Pi value = 0.021) and 0.385 (Pi value = 0.056) against thyroid gland undifferentiated (anaplastic) carcinoma (8505C) and adult B acute lymphoblastic leukemia (NALM-6), respectively. THC also has a probability of 0.321 (Pi value = 0.034) and 0.319 (Pi value = 0.064) against gastric carcinoma (MKN-7) and non-small cell lung carcinoma (HOP-18), making these its top 2 and 3 cell lines. On the other hand, THCV was found to have a probability of 0.317 (Pi value = 0.058) against Glioma (U-251) and a probability of 0.318 (Pi value = 0.059) against Melanoma (M19-MEL). Comparing the top three results of these cannabinoids, CBD having the highest probability against adenocarcinoma infers that it shows promising results and potential as an alternative component for cancer treatment. In addition to this, in terms of addressing Glioma, THC is slightly more effective against this disease than THCV. There is a slight difference between the probability of the two, with THC having a higher probability against the said disease.

The Pi values signify each cannabinoid’s probability of inactivity. Looking at the data gathered, CBD, which has the highest Pa value compared to THC and THCV, also has the lowest Pi value among the top three cancer cell lines. This further proves that, to an extent, CBD has active components that can be effective against breast adenocarcinoma. It is also the only cannabinoid that has exemplified a Pa value of greater than 0.5, which means that CBD is the only cannabinoid with a considerably high probability of action against the disease compared to the rest of the data in the table.

Although some inferences can be made from this study’s cell line cytotoxicity prediction, further investigation and research are recommended to determine their bioactive components, applicability, and efficiency against diseases like cancer, which could pave the way for future clinical studies and the development of treatment plans.

Pass Target Prediction predicts the biological potential of THC, CBD, and THCV by analyzing their biological activity, pharmacological effects, mechanism of action, toxic and adverse effects, interactions with metabolic enzymes and transporters, and influence on gene expression (Filimonov et al., 2014) [58]. The structure of THC, CBD, and THCV is represented in PASS predicting proteins with possible direct and mediated target interactions. The result in PASS prediction comes in the form of confidence p-values for each class [59].

The predicted results of THC produced a list of proteins that have direct interactions with the compound in descending order Cannabinoid CB2 receptor, Cannabinoid CB1 receptor, Cytochrome P450 2C9, DNA Topoisomerase II-a, and Estrogen receptor-b. Meanwhile, the list of proteins that have mediated interactions with THC is Glycine receptor subunit a-1, Cannabinoid CB2 receptor, Cannabinoid CB1 receptor, Serotonin 1e (5HT1e) receptor, and Nuclear receptor subfamily o group B member 1. Whereas the predicted results of CBD produced a list of proteins that have direct interactions with the compound in descending order are as follows: Phosphatidylinoitol-4-phosphate-5-kinasetype 1g, Cannabinoid CB1 receptor, Serine/threonine-protein kinase 32A, Rhodopsin kinase, and Cannabinoid CB2 receptor. Meanwhile, the list of proteins that have mediated interactions with CBD is Cannabinoid CB2 receptor, Glycine receptor subunit a-1, Tyrosine-protein kinase FYN, Cannabinoid CB1 receptor, and N-arachidonyl glycine receptor. Lastly, the predicted results of THCV produced a list of proteins that has direct interactions with the compound also in descending order are Phosphatidylinoitol-4-phosphate-5-kinasetype 1g, Rhodopsin kinase, Serine/threonine-protein kinase 32A, Cannabinoid CB1 receptor, and tyrosine-protein kinase Srms. Correspondingly, the list of proteins that have mediated interactions with THCV is Glycine receptor subunit a-1, Cannabinoid CB2 receptor, Tyrosine-protein kinase FYN, Cannabinoid CB1 receptor, and N-arachidonyl glycine receptor.

Among the direct protein targets involved in THC, the Cannabinoid CB2 receptor had the highest confidence of 0.4686. According to Vučković et al. (2018) [60], THC acts as a partial agonist toward the Cannabinoid CB2 receptor, which, regardless of the quantity of cannabinoid used, cannot fully activate the said receptor—hence affecting their binding interactions. Although, as presented in hyperalgesia models, THC binds to the Cannabinoid CB2 receptor, which indicates the relation of analgesic qualities with the attenuation of nerve growth factor and consequent inhibition of adenylyl cyclase [61]. On the other hand, CBD has been demonstrated to be a promising therapeutic and pharmacological medication target, with both direct and indirect interactions. Phosphatidylinoitol-4-phosphate-5-kinasetype 1g had the highest confidence interaction of 0.7681 out of the five proteins that can interact with CBD directly. D’Souza and Epand (2013) [62] found that phosphatidyl is substantially enriched with acyl chains, supporting this finding. Furthermore, the Cannabinoid CB2 receptor had the highest binding affinity with CBD, with 0.7529. The functional relevance of the Cannabinoid CB2 receptor in neuropathic pain and other neuroinflammatory illnesses underscores the therapeutic promise of CB2 receptor-targeting medicines [63]. Lastly, the compound THCV directly interacted with Phosphatidylinoitol-4-phosphate-5-kinasetype 1g with a confidence p-value of 0.8084. It is known that THCV is a homolog of THC where it only differs by a propyl side chain. It appears that it is a potential treatment for neurological and psychiatric conditions. Furthermore, it has a mediated interaction with Glycine receptor subunit α-1 at 0.6769. The interaction of THCV and glycine receptors may exhibit similar or more potent neuroprotective properties like THC and CBD. Moreover, THCV is a homolog of THC where it only differs at a propyl side chain. Hence, THCV acts as a CB1 receptor agonist that showed promise as an antiepileptic agent and protected neurons [64].

In conclusion, the physicochemical properties obtained from in silico ADME indicate that CBD had significant cell line toxicity and direct and mediated protein target interaction results, but had a poor oral absorption due to its insolubility, inactivity in mutagenicity, an inhibitor of only the CYP3A4 enzyme; THC exhibited excellent HIA and BBB, low possibility of blocking the CYPA2 enzyme, show a violation in drug-likeness analysis, insolubility, resulting in poor oral absorption, and higher toxicity than THCV. Among the three cannabinoids, THCV was a potential candidate for chemotherapeutic drugs obtaining the most favorable results in drug-likeness analysis, physicochemical properties, acute rat toxicity, and ADMET results.

To further improve this study and evaluate the effects of THC, CBD, and THCV, Biopharmaceutics Drug Disposition Classification System (BDDCS) must be performed to distinguish the transporters that participate in the drug absorption and elimination pathway. Furthermore, molecular docking analysis is also recommended to determine the binding affinity of the ligand and the structure of the protein-ligand interaction, which is vital for lead optimization. Due to the broad list of cannabinoid compounds with anticancer potential, there is a struggle to choose the cannabinoid with optimal anticancer development. Thus, a systematic investigation and more comprehensive ADMET profiles and different anticancer mechanisms must be conducted to generate broader antitumor spectra on cannabinoids to address this problem.