A new study published in Molecular Systems Biology by Charlie George Barker, Sumana Sharma, Evangelia Petsalaki and colleagues at the European Molecular Biology Laboratory and collaborating institutions across Europe offers one of the most detailed maps yet of how this resistance is built. The team focused on ARID1A, a component of the SWI/SNF chromatin remodeling complex that is mutated in approximately 11.5 percent of melanomas. Unlike the hotspot mutations that activate BRAF, ARID1A mutations are scattered across the gene, a pattern typical of tumor suppressors whose loss confers a survival advantage in many different ways. Previous genome-wide CRISPR screens had already flagged ARID1A loss as a driver of resistance to MAPK pathway inhibitors, but the precise circuitry behind that resistance remained obscure.

To resolve it, the researchers engineered a drug-resistant version of the A375 melanoma cell line, a BRAF V600E model that is normally exquisitely sensitive to BRAF and MEK inhibitors, by knocking out ARID1A. They then treated both the parental and knockout lines with vemurafenib, trametinib, or the combination, harvesting cells after six hours, a time point chosen to capture the earliest adaptive signaling responses before cell death begins. From these samples they generated an unusually rich dataset: mass spectrometry quantified 8,139 proteins and 3,207 phosphosites, while RNA sequencing measured the expression of 14,376 genes. They supplemented this with functional kinomics using PamChip peptide microarrays, which measure the actual enzymatic activity of dozens of kinases in cell lysates, and with Luminex-based phosphoprotein assays to confirm that the drugs were genuinely suppressing their targets.

The analytical challenge was to make sense of four orthogonal data layers simultaneously. The team’s solution combined two computational tools. First, they applied multi-omics factor analysis, or MOFA, an unsupervised matrix factorization method that identifies latent factors explaining variation across all data modalities at once. This revealed three dominant sources of variation: two factors capturing adaptive responses to drug treatment, and a third reflecting the sustained reprogramming caused by ARID1A loss. Second, they fed the factor loadings into phuEGO, a network propagation method that places the most extreme genes and phosphosites into their functional interaction context, producing minimal signaling networks for each factor. By merging positive and negative weights, the approach generated upregulated and downregulated network modules that capture the opposing molecular programs at play in drug response and resistance.

The first factor revealed changes that occurred regardless of which drug was used, pointing to mechanisms common to all MAPK inhibition. The most striking finding was the coordinated decrease of negative feedback regulators of the pathway, including the DUSP phosphatases DUSP1, DUSP2 and DUSP4, along with SPRY1/2/4 and SPRED1/2. These molecules normally act as brakes on receptor tyrosine kinase and MAPK signaling, terminating signals after they fire. When their abundance drops, the brakes come off, growth factor signaling intensifies, and the effect of the inhibitors is blunted. At the same time, the functional kinomics data showed activation of several receptor-associated kinases, including PRKD1, FYN and IGF1R, even though the abundance of their mRNAs and proteins barely changed. This dissociation between expression and activity underscores why measuring phosphorylation and kinase function, not just gene expression, is essential for understanding drug response.

The second factor isolated changes specific to combination therapy, and here the phosphoproteomics told an unexpected story. The team detected altered phosphorylation on a series of DNA repair proteins. TP53BP1, which promotes non-homologous end joining of double-strand DNA breaks, was heavily phosphorylated on serine 1101, a site associated with ionizing radiation damage, while losing phosphorylation at several other residues. RIF1, a key TP53BP1 regulator, was modulated at a site close to a known inhibitory phosphorylation mark, and ULK1 serine 556, an ATM-dependent autophagy trigger, was strongly upregulated. Enrichment analysis of the downregulated network highlighted terms related to ATM-mediated repair protein phosphorylation and recruitment of repair proteins to double-strand breaks, alongside suppressed immune signaling through TNF-alpha, interferons and interleukins. Intriguingly, in this particular cell line the combination therapy did not kill more cells than vemurafenib alone, suggesting these DNA damage signatures may mark cellular stress responses rather than enhanced therapeutic efficacy.

The third factor, tied to ARID1A loss, produced the study’s most consequential insights. Although the knockout cells mounted transcriptional and proteomic drug responses that correlated strongly with those of parental cells, with transcriptomic correlations of 0.91 to 0.93, their signaling behavior diverged dramatically. ARID1A knockout cells sustained MAPK1/3 and JNK activity after treatment, indicating the drugs no longer fully silenced these pathways. The researchers traced this to a fundamentally altered baseline state. The knockout cells showed elevated abundance of several receptor tyrosine kinases, including EGFR and ROS1, along with integrins and CD44. Flow cytometry confirmed a drastic increase in EGFR on the cell surface, consistent with impaired receptor endocytosis in resistant melanoma cells. Elevated membrane receptors can sustain chronic signaling rather than transient, ligand-dependent pulses, providing a persistent growth signal that survives MAPK blockade.

Using random walk network propagation and maximum-flow analysis, the team showed that signals from EGFR and ROS1 converge on the transcription factor JUN through the adaptor proteins FYN, PRKD1, PTPN6 and NCK1. In parental cells, drug treatment activates PRKD1, which suppresses the JNK/c-Jun axis linked to apoptosis in BRAF-mutant melanoma. In ARID1A knockout cells, this regulatory logic is inverted: PRKD1 activation is suppressed, JNK inhibition is relieved, and JUN activity rises, a known driver of BRAF inhibitor resistance. The predictive power of the network approach was validated experimentally. When the researchers deleted EGFR using two independent guide RNAs, the drug-resistant phenotype of the knockout cells was abolished, and cell death exceeded even that of parental cells under combination therapy. Ephrin receptor signaling, also elevated in the knockout cells, emerged as an additional resistance route through NCK1 and related adaptors.

The study also connected ARID1A loss to immune evasion, with implications for immunotherapy. The knockout cells showed reduced activity of the RFX5, RFXAP, RFXANK and NFYC transcription factors, which drive MHC class II gene expression, and correspondingly reduced surface levels of HLA-DQ and HLA-DR. Analysis of 472 melanoma patients from The Cancer Genome Atlas confirmed the pattern: the 80 patients with ARID1A mutations or deletions had significantly lower predicted activity of these MHC regulators and elevated collagen and laminin expression, suggesting extracellular matrix remodeling that could physically impede T-cell infiltration. Together, the findings position PRKD1, JUN and NCK1 as key resistance nodes and highlight a sobering lesson about redundancy: with multiple receptors, including EGFR, ROS1 and FGFR1, feeding into rewired signaling, targeting any single protein is unlikely to succeed. Instead, the study’s integrative framework, which disentangles drug responses from baseline reprogramming and prioritizes experimentally testable vulnerabilities, points toward combination strategies aimed at JUN or immune pathway restoration, offering a systems-level blueprint for outmaneuvering one of cancer’s most adaptable escape artists.

Subject of Research: Multi-omics analysis of ARID1A-dependent resistance to BRAF and MEK inhibitor therapy in melanoma

Article Title: Integrative multi-omics defines melanoma drug response networks and ARID1A-dependent resistance mechanisms

Article References: Barker, C. G., Sharma, S., Santos, A. M., Nikolakopoulos, K.-S., Velentzas, A. D., Tormo-Garcia, C., Sharma, A., Völlmy, F. I., Minia, A., Pliaka, V., Clarke, J., Altelaar, M., Wright, G. J., Alexopoulos, L. G., Stravopodis, D. J., & Petsalaki, E. (2026). Integrative multi-omics defines melanoma drug response networks and ARID1A-dependent resistance mechanisms. Molecular Systems Biology, 22(5), 685-711. https://doi.org/10.1038/s44320-025-00183-5

Image Credits: AI Generated

DOI: 10.1038/s44320-025-00183-5

Keywords: melanoma, ARID1A, drug resistance, BRAF inhibitors, MEK inhibitors, MAPK signaling, multi-omics, phosphoproteomics, EGFR, JUN, immune evasion, network biology

The first of the highlighted studies tackles a long-standing obstacle in cancer biology: the difficulty of studying tumors in their native spatial context. Goldschmidt and colleagues applied state-of-the-art spatial transcriptomics to formalin-fixed paraffin-embedded biopsy samples from nine dogs diagnosed with COSCC. Spatial transcriptomics is a technique that measures gene expression while preserving the physical location of each cell within the tissue, allowing researchers to see not just which genes are active but where that activity occurs. This matters because tumors are not uniform masses; they contain distinct compartments, including normal epithelium, pre-malignant dysplastic epithelium, the tumor microenvironment, and regions of surface and deep tumor, each of which may behave differently and respond differently to therapy.

The spatial analysis revealed striking molecular patterns. Gene signatures associated with EGFR, the epidermal growth factor receptor, a well-known driver of cell proliferation in many human cancers, and with epithelial-to-mesenchymal transition, a process by which cancer cells acquire invasive, migratory properties, were increased and shared between dysplastic epithelium and tumor samples. This suggests that the molecular changes driving malignancy begin early, before full transformation. Within the cancer samples themselves, the researchers found enrichment of KRAS gene sets, along with the transcripts KRT17 and SPP1, markers often associated with aggressive tumor behavior. Perhaps most intriguing were the findings in the tumor microenvironment, defined in the study as peritumoral mesenchymal stromal tissue infiltrated with leukocytes. There, the team detected signals associated with an increased presence of tumor-associated macrophages, immune cells that tumors frequently co-opt to suppress anti-cancer immunity, and induction of IL10, an immunosuppressive signaling molecule. Enrichment of ICOS and CTLA4 was also noted, and these signals correlated with CD4-positive T cell transcripts in the microenvironment, hinting at a complex immune landscape that could be exploited therapeutically.

The second and third studies demonstrate how such molecular insights can be translated directly into new treatments. Katt and co-workers established several new COSCC cell lines using a combination of patient-derived xenografts and tissue-cultured cell lines created directly from COSCC biopsies. Patient-derived models are prized in oncology research because they preserve more of the biology of the original tumor than long-established laboratory lines. When the team investigated these new models, they uncovered evidence of RAS activation in every one of them. RAS is a signaling protein that, when switched on, drives cells to proliferate uncontrollably, and it sits upstream of MEK1/2, a pair of serine/threonine kinases that relay the growth signal deeper into the cell.

That molecular logic pointed straight at a drug. Trametinib is a small-molecule inhibitor of MEK1/2, already used in human medicine for cancers with activated RAS pathway signaling. The researchers screened trametinib against other cytotoxic and targeted agents using their novel cell lines and found, as predicted, that it exerted anti-tumor activity both in vitro and in a canine xenograft model. These preclinical results were compelling enough to justify testing the drug in pet dogs with COSCC. In a preliminary study of four dogs, objective responses were documented in two of the treated patients. A subsequent dose-escalating study of twenty dogs with COSCC confirmed that trametinib was well tolerated and showed a thirty percent objective response rate, with a further twenty-five percent of dogs exhibiting stable disease, meaning their tumors neither grew nor shrank significantly.

Crucially, the correlative analyses in that trial revealed a biomarker that could transform how dogs are selected for treatment. Dogs whose tumors carried BRAF p.V595E mutations responded significantly better to trametinib than dogs with BRAF wild-type tumors. This alignment with trametinib’s known mechanism of action in activated RAS signaling illustrates the essence of precision oncology: rather than treating every patient identically, clinicians can profile a tumor’s mutations and predict which patients are most likely to benefit. Together, the cell line work and the clinical trials illustrate a rational pathway for introducing new targeted therapies in COSCC, moving systematically from bench to patient-side, a journey that in human oncology often takes many years but here was compressed into a remarkably short arc.

The fourth study extends the comparative approach to feline patients and to immunotherapy-adjacent territory. Signal transducer and activator of transcription 3, or STAT3, is a transcription factor that is hyperactivated in many cancers, including FOSCC, where it drives tumor cell proliferation, immunosuppression, and drug resistance. A prior phase 0 trial in humans with head and neck squamous cell carcinoma had shown that a STAT3 decoy, an engineered molecule that binds and sequesters the transcription factor’s activation machinery, could downregulate STAT3 target genes. Building on that human data, Grandis and colleagues evaluated a systemically administered linear STAT3 decoy in multiple preclinical murine models of head and neck squamous cell carcinoma and then undertook a phase 1 clinical trial of twenty cats diagnosed with FOSCC.

The feline trial produced results that are encouraging for one of the most aggressive malignancies in companion animal medicine. A thirty-five percent disease control rate was documented, comprising two partial responses and five cases of stable disease. Importantly, the favorable responses were accompanied by measurable biological changes: shifts in peripheral blood immune parameters and increased intratumoral expression of PDCD1, the gene encoding the PD-1 immune checkpoint protein. That last finding is particularly significant because PD-1 expression within tumors is a marker of T cell exhaustion, the state in which anti-tumor immune cells become dysfunctional. Its increase in responding patients suggests the STAT3 decoy may be relieving one layer of immunosuppression and potentially sensitizing tumors to checkpoint blockade, a class of drugs that releases the brakes on anti-tumor immunity. Beyond its clinical promise for cats, the study exemplifies the power of a comparative approach in cancer research, in which findings in one species inform and accelerate progress in others, with the aim of benefiting multiple species simultaneously.

Weaving these threads together, Atherton’s editorial argues that veterinary medicine is now entering the age of immune oncology, heralded in part by the roll-out of checkpoint blockade. The identification of CTLA4 upregulation within COSCC by the spatial transcriptomics study, alongside the prognostic implications of increased PDCD1 expression in FOSCC patients treated with the STAT3 decoy, provides impetus to investigate these molecules as mediators of T cell exhaustion. That, in turn, may set the scene for future trials of anti-CTLA-4 or anti-PD-1 therapy in oral squamous cell carcinoma patients of both species. Meanwhile, the early but encouraging clinical data from dogs treated with trametinib offers cautious optimism that this targeted therapy could be added to the currently limited arsenal of useful anti-cancer drugs for COSCC.

Ongoing studies are expected to determine the optimal way to incorporate trametinib into treatment regimens. The editorial notes several plausible roles: the drug could be applied in a neo-adjuvant setting, given before definitive surgery or radiation with the aim of improving the tolerability and outcomes of conventional treatment, or it could serve as an additional palliative option for more advanced COSCC cases where curative intent is no longer feasible. What makes this moment notable is not any single result but the demonstration that the full precision-oncology workflow, from spatially resolved molecular profiling through patient-derived models to biomarker-guided clinical trials, now functions in veterinary patients. For the dogs and cats facing these cancers, and for the humans whose head and neck cancers share many of the same molecular vulnerabilities, the comparative road ahead looks considerably brighter than it did only a few years ago.

Subject of Research: Recent advances in the diagnosis and targeted therapy of canine and feline oral squamous cell carcinoma

Article Title: An editorial spotlight on recent progress in veterinary oncology

Article References: Atherton, M. J. (2025). An editorial spotlight on recent progress in veterinary oncology. Veterinary Oncology, 2(1), Article 30. https://doi.org/10.1186/s44356-025-00047-y

Image Credits: AI Generated

DOI: 10.1186/s44356-025-00047-y

Keywords: veterinary oncology, oral squamous cell carcinoma, spatial transcriptomics, trametinib, MEK inhibitor, STAT3 decoy, tumor microenvironment, checkpoint blockade, BRAF mutation, comparative oncology, precision medicine, immunotherapy

Oral squamous cell carcinoma, or OSCC, is a formidable enemy in canine medicine. Oral tumors affect an estimated half a percent of dogs over their lifetimes and account for about six percent of the canine cancer burden, with OSCC making up roughly nineteen percent of all oral tumors. The disease is locally invasive, growing through multiple anatomical planes of the jaw and surrounding tissue, causing severe pain, difficulty eating, and in some cases pathological fractures of the bone. Up to twenty percent of affected dogs show signs of metastasis by the time they are diagnosed. Left untreated, it is deadly.

The current standard of care is blunt: wide-margin surgical excision. While surgery is highly effective when non-metastatic tumors are completely removed, with very low recurrence rates, it is also highly invasive and invariably disfiguring. Roughly thirty percent of dogs have difficulty eating after surgery, and larger operations bring longer healing times, greater postoperative pain, and higher risks of complications. Radiation is occasionally used for tumors that cannot be surgically removed or as an adjunct to surgery, but the fundamental treatment paradigm has not changed in decades, even as researchers have made major strides in understanding the molecular machinery driving the disease.

That machinery centers on the RAS signaling pathway. Recent studies have shown that canine OSCC displays highly activated RAS signaling compared with healthy gingival tissue, a finding confirmed by gene set enrichment analysis comparing tumor samples with normal tissue. RAS is a small GTPase that, when activated by extracellular signals binding to receptors on the cell surface, launches a cascade of molecular events that ultimately drive cell growth and survival. The most important arm of this cascade for tumor growth is the RAS-RAF-MEK-ERK pathway, which ends with ERK initiating the gene transcription programs that keep a tumor expanding. The question facing the Cornell team was whether this pathway could be pharmacologically attacked in a way that would benefit dogs whose tumors are driven by a variety of different upstream mutations.

The answer they settled on was trametinib, a small molecule that inhibits MEK1 and MEK2, the kinases sitting in the middle of the pathway. Trametinib binds to the non-phosphorylated form of MEK in an allosteric pocket near the ATP binding site, blocking its catalytic activity and its ability to activate ERK. Crystal structures show the drug does more than simply plug the pocket: it also prevents BRAF from binding to MEK and appears to stabilize MEK’s association with KSR, the kinase suppressor of RAS, further reducing MEK’s affinity for BRAF. Crucially, the team chose trametinib precisely because it works on wildtype, unmutated proteins. Since RAS itself is rarely mutated in canine OSCC and BRAF mutations appear in only about a third of tumors, drugs like sotorasib, which bind only mutant RAS, would have left most patients out in the cold. Trametinib, by contrast, targets the pathway regardless of what triggered its activation.

The drug also came with a reassuring safety pedigree. It is FDA-approved in humans for BRAF-mutant melanoma and non-small cell lung cancer, typically alongside the BRAF inhibitor dabrafenib. A phase I trial in eighteen dogs had already demonstrated an acceptable safety profile over nine weeks of treatment, and the European Medicines Agency reported the drug was well tolerated in beagles over thirteen weeks and fully eliminated within about seven days. Its high oral bioavailability in both species meant it could be given at home, avoiding the stress and expense of repeated hospital visits.

The trial itself was designed with a do-no-harm philosophy. Twenty dogs with histologically confirmed OSCC were enrolled between October 2023 and February 2025, after approval by Cornell’s Institutional Animal Care and Use Committee and with informed consent from owners. Dogs received daily oral trametinib at doses escalating from 0.015 to 0.035 milligrams per kilogram, with examinations every two weeks including caliper tumor measurements, complete blood counts, and serum chemistry panels. Full tumor staging with contrast-enhanced computed tomography was performed at the start, around day 28, and around day 56, and tumor volumes were measured from CT slices by a board-certified veterinary radiologist. Response was categorized using veterinary RECIST criteria, and any dog showing progressive disease was removed immediately and returned to standard care. Because the typical two-week wait before surgery would not be compromised, owners could try the drug without sacrificing the earliest opportunity for curative excision.

The results were remarkable in some patients. One dog achieved a complete response, with no tumor detectable by CT, and five achieved partial responses, giving an overall response rate of thirty percent. Five more dogs had stable disease, meaning fifty-five percent of the cohort derived clinical benefit. Some tumors responded with startling speed: two dogs’ tumors had shrunk to less than one cubic centimeter after just two weeks of treatment, and in five dogs the mean volume reduction at the first post-treatment evaluation was forty-five percent. In two responsive patients, bone window CT scans showed actual bone regrowth at the tumor site, visible at 28 and 56 days after treatment began, a sign of genuine tissue healing rather than mere tumor shrinkage. In some cases the visible portions of tumors became difficult or impossible to identify by external inspection.

Safety data were equally encouraging. No high-grade adverse events were detected in any dog, though about thirty percent of owners reported transient flatulence that resolved on its own. Bloodwork showed mostly grade 1 and grade 2 abnormalities that were typically no worse than pre-treatment values. The only two grade 3 events, elevated alkaline phosphatase in two dogs, were preceded by grade 2 liver enzyme elevations present before treatment began, suggesting preexisting liver pathology unrelated to the drug. No dog was removed from the study or had its dose modified because of side effects, and no neutropenia was observed. Notably, a prior phase I study had flagged hypertension and proteinuria in some dogs at similar doses, but the Cornell team did not observe owner-reported behaviors consistent with those conditions, though blood pressure and urinalysis were not directly monitored.

The most scientifically compelling finding emerged when tumors were stratified by BRAF mutational status. Seven dogs carried tumors harboring the BRAF p.V595E mutation, the canine equivalent of the human V600E mutation, and these responded dramatically better than BRAF wildtype tumors. Among BRAF-mutant cases, one dog had a complete response and four had partial responses, for a seventy-one percent objective response rate and a mean volume reduction of seventy-one percent in responding tumors. Wildtype tumors, by contrast, showed a mean reduction of only seven percent among responders, and total clinical benefit in that group was just eight percent. The difference was statistically significant at p equals 0.01. Because BRAF genotyping requires only simple PCR amplification and Sanger sequencing, the authors suggest it could readily be implemented by diagnostic laboratories to identify dogs most likely to benefit. The team recommends a starting dose of 0.030 milligrams per kilogram daily while cautioning that dose-response analyses were underpowered. Limitations remain, including the absence of an untreated control group, the lack of long-term recurrence and survival data, and the small cohort size, but the study establishes trametinib as a safe, orally available neoadjuvant option that can shrink tumors in about a third of dogs before surgery, and in the majority of those whose tumors are BRAF-mutant, with results visible in as little as two weeks.

Subject of Research: Clinical trial of the MEK inhibitor trametinib for treating oral squamous cell carcinoma in dogs

Article Title: Clinical application of the MEK inhibitor trametinib in dogs with oral squamous cell carcinoma

Article References: Katt, W. P., Balkman, C. E., Byron, M., Carney, P. C., Chrostek, E., Drozd, M. E., Duhamel, G. E., Fiani, N., Hume, K. R., King, A. L., Sylvester, S. R., Todd-Donato, A. B., Winokur, C. E., Wright, A. L., & Peralta, S. (2025). Clinical application of the MEK inhibitor trametinib in dogs with oral squamous cell carcinoma. Veterinary Oncology, 2(1), Article 31. https://doi.org/10.1186/s44356-025-00042-3

Image Credits: AI Generated

DOI: 10.1186/s44356-025-00042-3

Keywords: trametinib, MEK inhibitor, oral squamous cell carcinoma, canine cancer, BRAF mutation, RAS signaling, veterinary oncology, targeted therapy, clinical trial, neoadjuvant treatment, Cornell University, dog health

Malignant melanoma arises from melanocytes, the pigment-producing cells of the skin, and its incidence and mortality have been climbing steadily. Advanced disease carries a particularly grim outlook: melanomas are notoriously insensitive to conventional chemotherapy and radiotherapy, leaving patients with a five-year survival rate of roughly 32 to 52 percent. Targeted therapies against BRAF-mutant tumors and modern immunotherapies have improved outcomes, but resistance is common, and AKT signaling remains frequently activated in both BRAF-mutant and BRAF-wild-type melanomas. That makes an agent capable of damping AKT phosphorylation attractive precisely because it should not care which side of the BRAF divide a tumor sits on.

Bufalin itself is no newcomer to the pharmacopoeia. Extracted from the venom of toads used in the traditional medicine known as Chan Su, the steroid has been deployed clinically in China to treat tumor recurrence and metastasis, where it reportedly reduces the side effects of radiotherapy and chemotherapy while improving quality of life. Laboratory studies have already documented anti-tumor activity in lung, colorectal, breast and liver cancers. But in melanoma, prior work had focused mainly on apoptosis induction through caspase and mitochondrial pathways or on identifying individual binding targets. What remained unresolved was whether bufalin converges on p-AKT as a central node coordinating both cell death and cell cycle arrest — and whether this holds independent of BRAF status.

The new study set out to answer that question with a battery of assays in two BRAF-mutant melanoma cell lines, A375 and A2058, alongside human umbilical vein endothelial cells, the standard workhorse for modeling angiogenesis. In viability assays, bufalin inhibited melanoma cell growth in a dose- and time-dependent manner, with half-maximal inhibitory concentrations of 20 and 17.5 nanomolar at 24 and 48 hours in A375 cells, and 60 and 30 nanomolar in A2058 cells. Colony formation, a proxy for long-term proliferative capacity, fell significantly as bufalin concentrations rose. These are strikingly low concentrations — tens of nanomoles — underscoring the potency of the molecule.

Flow cytometry revealed how the growth suppression was achieved. Untreated cultures showed apoptotic fractions below 5 percent, but bufalin pushed apoptosis to 10 to 20 percent at 10 nanomolar and 30 to 40 percent at 20 nanomolar. At the same time, the proportion of cells trapped in the G2/M phase of the cell cycle rose markedly, indicating that bufalin does not merely kill cells but also halts their division machinery. The team then probed metastatic behavior: wound healing assays showed widened scratches after 12 hours of treatment, and transwell experiments demonstrated that fewer melanoma cells migrated through porous membranes or invaded through a Matrigel matrix at concentrations of 5 and 10 nanomolar. Migration and invasion, the twin engines of metastasis, were both throttled.

Crucially, the compound also targeted the tumor’s blood supply. In endothelial cells, bufalin inhibited proliferation with IC50 values of 40 nanomolar at 24 hours and 20 nanomolar at 48 hours, reduced colony formation, slowed wound closure, and curtailed migration through transwell chambers. Most dramatically, in Matrigel tube formation assays, the dendritic vascular networks that endothelial cells normally weave were fewer, fragmented and structurally destroyed at 10 nanomolar. Since tumors co-opt angiogenesis to secure nutrients and oxygen as they grow and spread, an agent that simultaneously attacks tumor cells and the endothelial cells feeding them addresses both sides of the tumor-vascular equation.

To identify the signaling pathway behind these effects, the researchers turned to network pharmacology, an interdisciplinary approach that fuses medicine, pharmacology and bioinformatics to map how a multi-target molecule engages a biological system. Cross-referencing predicted bufalin targets from the PharmMapper database with melanoma genes from MalaCards and angiogenesis genes from GENECLIP yielded 57 key shared targets. Gene Ontology analysis pointed to stimulus response and biological regulation as dominant biological processes, while Kyoto Encyclopedia of Genes and Genomes pathway enrichment ranked the PI3K-AKT pathway first among 20 candidate pathways. Protein-protein interaction network analysis identified MAPK, AKT1 and p53 as the most connected hub genes, with modular analysis again crowning PI3K-AKT as the most important signaling route.

Molecular docking then simulated how bufalin might engage AKT1 at the atomic level. The docking score of −3.956 suggested a favorable interaction, with the model showing bufalin forming a hydrogen bond with the 53rd asparagine residue of AKT1. Because docking is only a theoretical prediction, the team validated it experimentally. Western blotting showed that total PI3K and total AKT protein levels were essentially unchanged after bufalin treatment in both melanoma and endothelial cells, but phosphorylated AKT — detected at the Ser473 regulatory site — dropped significantly. The decisive test came from rescue experiments using SC-79, a small-molecule activator that drives AKT phosphorylation. Adding SC-79 restored p-AKT levels, and bufalin reversed that increase, demonstrating that the compound acts upstream to suppress AKT activation rather than merely masking its downstream readouts.

The in vivo evidence followed. In xenograft models, A375 melanoma cells were implanted subcutaneously into nude mice, and once tumors reached roughly 100 cubic millimeters, animals received daily intraperitoneal bufalin at 1.5 milligrams per kilogram for 12 days. Tumors in treated mice grew more slowly and ended the experiment significantly smaller and lighter than vehicle controls. Immunohistochemistry on excised tumor sections confirmed the mechanism in tissue: staining for phosphorylated AKT and for Ki-67, a canonical proliferation marker, was markedly reduced after treatment. Notably, the mice showed no obvious loss of body weight during the treatment period, suggesting the antitumor effects came without apparent systemic toxicity under the experimental conditions.

The authors argue that these results establish p-AKT as the key mechanistic node for bufalin’s action in melanoma, and the clinical implications are twofold. First, because p-AKT downregulation appears sufficient to trigger anti-tumor effects regardless of BRAF mutation status, bufalin could in principle work against a broad range of melanomas, including tumors resistant to BRAF and MEK inhibitors. Second, as a natural compound with an existing history of use in traditional Chinese medicine, bufalin could be repurposed as an adjunct therapy to enhance existing treatments or to overcome resistance. Earlier studies support this framing: bufalin has been shown to suppress hepatocellular carcinoma invasion via PI3K/AKT/HIF-1 signaling, to induce lung cancer apoptosis through PI3K/AKT inhibition, to reverse cisplatin resistance in gastric cancer by blocking AKT activation, and to enhance sorafenib’s anti-angiogenic effect through the PI3K-AKT-VEGF axis. The authors caution that much remains to be done: future work should test bufalin in combination with targeted agents and immunotherapies in preclinical models, and rigorously evaluate its safety and pharmacokinetics in melanoma patients. But the conceptual shift is clear — a single upstream event, the silencing of AKT phosphorylation, may be enough to orchestrate apoptosis, cell cycle arrest and angiogenesis blockade all at once, offering a new framework for developing toad venom’s most potent secret into a modern anti-melanoma drug.

Subject of Research: Bufalin's inhibition of melanoma progression and endothelial angiogenesis through suppression of AKT phosphorylation

Article Title: Bufalin Inhibits Melanoma Progression and Endothelial Angiogenic Phenotypes via AKT Signaling

Article References: Zhao, X. X., Li, M. X., & Qu, J. X. (2026). Bufalin Inhibits Melanoma Progression and Endothelial Angiogenic Phenotypes via AKT Signaling. Cancer Reports, 9(9), Article e70674. https://doi.org/10.1002/cnr2.70674

Image Credits: AI Generated

DOI: 10.1002/cnr2.70674

Keywords: bufalin, melanoma, AKT signaling, PI3K-AKT pathway, angiogenesis, toad venom, apoptosis, cell cycle arrest, network pharmacology, molecular docking, xenograft model, BRAF mutation