In February 2026, a peer-reviewed journal Pharmaceuticals published a study that answers a compelling question — can natural compounds found in cannabis precisely target several molecular “switches” that trigger and sustain skin cancer? Researchers from Morocco and the UAE identified a number of such compounds — effective ones, and crucially, safe ones.
Three Enemies on One Field
To understand what this is all about, a brief detour into tumour biology is in order.
A cancer cell is not simply a “broken” cell. It is a cell that has learned to bypass all the control mechanisms built into the body. To do this, it exploits certain signalling proteins, continuously sending commands to “grow”, “divide”, “don’t die”. Three such proteins are particularly well studied in the context of skin cancer.
EGFR — the epidermal growth factor receptor. Under normal circumstances, it responds to external signals and triggers cell division when necessary — for example, during wound healing. In tumour cells, it operates in a permanent “on” mode, relentlessly pushing the cell towards uncontrolled proliferation.
BRAF V600E — a mutant version of the BRAF protein, found especially frequently in melanoma. The letter “V” and the numbers in the name refer to the specific point in the molecule where the “malfunction” occurred. Normal BRAF switches on and off as needed; the mutant version is stuck in the “on” position, triggering a chain of signals that leads to aggressive tumour growth.
TGF-β — transforming growth factor beta, the most ambiguous player of the three. In the early stages of disease it suppresses tumour growth, but as the tumour develops it effectively “switches sides”: it begins helping cancer cells spread throughout the body and evade the immune system.
Current drugs typically target one of these pathways. That is effective, but tumours have an unpleasant trait: remove one route and they find another. This is precisely why scientists are increasingly searching for so-called multi-targetmolecules — ones capable of blocking several signalling pathways simultaneously.
49 Candidates From a Single Plant
Cannabis sativa is an extraordinarily rich phytochemical reservoir. More than five hundred biologically active compounds have been identified in the plant (the best-known of them, cannabidiol, is covered in CBD explained). The study’s authors selected 49 of them — cannabinoids and terpenes — and used computational modelling to assess how well each one could “fit” into the active site of each of the three target proteins.
The method they used is called molecular docking. Think of a lock and key: the shape of a ligand molecule must match the shape of a pocket on the protein’s surface in order to form a stable bond. The computer calculates the energy of this interaction — the more negative the value, the more firmly the “key” sits in the “lock”, and the greater the likelihood that the molecule will block the protein’s activity.
It is worth noting that the researchers worked primarily with neutral cannabinoids. Yet their acidic precursors — CBDA, CBGA and others — are also showing growing therapeutic potential, which we explored in detail in a separate article.
Three reference drugs were used for comparison: erlotinib (an EGFR inhibitor), vemurafenib (a BRAF V600E inhibitor) and galunisertib (a TGF-β inhibitor). All three are approved medicines used in clinical practice.
Three Winners
Out of 49 candidates, a top three emerged with striking results. These are the compounds the authors labelled C5, C6 and C7 — known in everyday terms as tetrahydrocannabivarin (THCV), cannabinol (CBN) and delta-9-tetrahydrocannabinol (Δ⁹-THC, dronabinol).
All three demonstrated higher binding affinity to all three targets than the corresponding reference drugs.
But impressive numbers from docking are only the first step. To confirm that the binding remains stable not just momentarily but over time, the researchers ran molecular dynamics simulations — computational modelling of protein–ligand complex behaviour over 200 nanoseconds (an extremely long time by molecular standards).
Several parameters were analysed. RMSD shows how much the protein’s spatial structure changes — small values indicate stability. RMSF describes the mobility of individual regions — if the protein “settles down” after a ligand binds, that is a positive sign. The radius of gyration reflects the molecule’s overall compactness, while the surface areaindicates how the protein’s contact with surrounding water has changed.
The results were encouraging: all three complexes maintained stability throughout the entire simulation. Each compound remained seated in its pocket without dislodging or losing key contacts. CBN showed the greatest stabilising effect on EGFR, THCV on TGF-β, and Δ⁹-THC on both EGFR and BRAF V600E.
The final confirmation came from MM-GBSA binding free energy calculations — a more precise thermodynamic assessment of interaction strength than docking alone. The values confirmed that the complexes are thermodynamically favourable, meaning their formation is energetically “preferred” by nature.
What About Safety?
This is perhaps the most important question for any potential therapeutic agent.
The authors performed a detailed ADMET analysis — a computational assessment of how the molecules would behave in the body. The acronym stands for Absorption, Distribution, Metabolism, Excretion and Toxicity.
The picture that emerged was highly encouraging. All three compounds comply with the so-called Lipinski’s Rule of Five — a set of simple criteria that correlate well with a molecule’s suitability to become a drug (molecular weight, fat solubility, number of hydrogen bonds, etc.). The lipophilicity of all three — that is, their ability to dissolve in fats — fell within the range optimal for penetration through the lipid-rich layers of the skin.
Calculated skin permeability coefficients indicate moderate transdermal absorption: sufficient to deliver the active substance to the target layers, but without excessive systemic entry into the bloodstream. This is particularly important for topical (external) formulations.
None of the three compounds showed signs of hepatotoxicity, carcinogenicity, mutagenicity or cytotoxicity in computational predictions. The median lethal doses (LD₅₀) for THCV and Δ⁹-THC were 482 mg/kg, while for CBN the figure was 13,500 mg/kg — corresponding to extremely low acute toxicity.
None of the candidates inhibit the key enzymes of the cytochrome P450 system, which is responsible for metabolising most drugs — meaning a low risk of unwanted drug–drug interactions.
What This Means — and What It Doesn’t
Honesty is essential here.
This study is entirely computational. No cell cultures, no mice, no patients. That is not a flaw or an attempt to mislead — it is the normal first stage of modern drug discovery. Computational modelling allows scientists to narrow thousands of candidates down to a handful of the most promising ones, directing costly experiments exactly where the chances of success are greatest.
The road from these results to a pharmacy shelf is long and unpredictable. Cell-based experiments, animal studies and clinical trials all lie ahead. A molecule can fail to deliver at any of these stages. That is the reality of pharmacology.
Nevertheless, the significance of this work should not be understated. It sends several important signals.
First, it shows that compounds present in cannabis can theoretically act on several key oncogenic targets simultaneously — something that is difficult to achieve with a single synthetic drug.
Second, it provides grounds to believe that such action may be possible without significant toxicity.
Third, the study establishes a clear experimental agenda: scientists now know exactly which compounds to test, against which targets, and can verify whether laboratory results align with computational predictions.
FAQ: Cannabinol, THCV and THC — What Are These Molecules?
THCV (tetrahydrocannabivarin) is a structural “cousin” of THC with a shortened side chain. It is found in cannabis in relatively small quantities, does not produce significant psychoactive effects at typical doses, and is actively being studied for its potential anti-diabetic and neuroprotective properties.
Cannabinol (CBN) is an oxidation product of THC — it is what THC gradually turns into when the plant is stored for extended periods. CBN’s psychoactivity is significantly lower than that of THC. The compound is being studied as a sedative agent and potential antibacterial, but it remains in the shadow of more prominent cannabinoids in public discourse.
Δ⁹-THC is the best-known cannabinoid, responsible for cannabis’s psychoactive effects. At the same time, its anti-cancer activity is one of the most thoroughly documented aspects of cannabinoid pharmacology in preclinical research. In the form of dronabinol, THC is FDA-approved as a pharmaceutical — for nausea associated with chemotherapy and appetite loss in AIDS patients.
No. This study is computational in nature and represents only the first stage of development. Before clinical application can be considered, experiments on cell cultures and animal models are required, followed ultimately by human trials. Nevertheless, it is precisely studies like this that define the direction for the next steps.
Not at all. The molecular targets examined in this study — EGFR, BRAF V600E and TGF-β — are involved in other cancers as well. Scientists’ interest in cannabinoids extends well beyond dermatology: we have already covered a promising study on CBD in kidney cancer, which also points to encouraging anti-tumour potential of cannabinoids.
What Comes Next: Cream or Capsule?
The authors place particular emphasis on the potential of their candidates for topical formulations — ointments, creams, gels. This is logical: skin cancer is located precisely where external preparations can reach directly, bypassing the challenges of systemic bioavailability. The calculated skin permeability parameters suggest that the compounds can penetrate deeply enough to reach cells in the viable epidermis and dermis — without accumulating in the blood at dangerous concentrations.
The researchers also highlight the broader potential of the compounds identified. The same molecular targets — EGFR, BRAF V600E and TGF-β — are implicated not only in skin cancer but also in other malignancies, as well as in inflammatory and fibrotic diseases. In other words, should laboratory experiments confirm the computational predictions, the scope of application could prove wider than dermatological oncology.
In Conclusion
The history of medicine is full of examples of natural compounds becoming the basis of therapy — aspirin from willow bark, morphine from the opium poppy, taxol from the yew tree. Cannabis sativa, with its five hundred active molecules, is an obvious candidate to join that list.
The work of the Moroccan–Emirati team is neither a sensation nor a promise of a miracle cure. It is a conscientious scientific contribution to understanding what we are actually dealing with when we speak of the therapeutic potential of cannabis. Now the word passes to the laboratory.
Source: El Bouamri L. et al. Integrated Computational Investigation of Cannabis sativa Phytoconstituents as Putative Multi-Target Inhibitors in Skin Cancer. Pharmaceuticals, 2026, 19, 315.
Disclaimer
This article is intended for informational and educational purposes only. The data presented are based on the results of a computational (in silico) study and do not constitute clinical evidence of the efficacy or safety of the mentioned compounds in humans. None of the substances described have been approved as anti-cancer agents by the regulatory authorities of any country, except where explicitly stated in the text.
The Cannex editorial team does not provide medical advice and does not advocate self-treatment. If you or a loved one has been diagnosed with cancer, please consult a qualified oncologist. Any decisions relating to treatment must be made exclusively in consultation with a licensed medical professional.
The legal status of cannabis and its derivatives varies significantly between countries. The editorial team accepts no responsibility for actions taken by readers that violate the laws of their country of residence.
