The compound itself is a product of coordination chemistry, the branch of inorganic chemistry concerned with metal centers surrounded by organic ligands. Schiff bases, named for the nineteenth-century Italian chemist Hugo Schiff, are molecules containing an imine functional group, a carbon-nitrogen double bond formed by the condensation of an amine with an aldehyde or ketone. In this case, the ligand is 2-((allylimino)methyl)-6-methoxyphenol, derived from 3-methoxy-2-hydroxybenzaldehyde and allylamine. Three of these ligands coordinate a single cobalt ion in the +3 oxidation state, producing a mononuclear octahedral geometry in which six donor atoms surround the metal center in a three-dimensional arrangement resembling two pyramids joined base to base. This architecture is not merely aesthetic; the geometry, charge distribution, and lipophilicity of a metal complex all influence how it interacts with biological membranes, proteins, and nucleic acids, and researchers in this field have spent decades tuning these parameters to maximize anticancer activity while minimizing harm to healthy tissue.

The appeal of cobalt in this context is part of a broader renaissance in metal-based therapeutics. Cisplatin and its platinum cousins remain mainstays of chemotherapy for testicular, ovarian, lung, and other cancers, but their utility is constrained by severe side effects, including nephrotoxicity, ototoxicity, and peripheral neuropathy, as well as by the emergence of platinum resistance in many tumors. That has motivated medicinal inorganic chemists to explore other metal centers, including ruthenium, copper, gallium, and cobalt, each with distinct coordination preferences, redox behavior, and mechanisms of cellular uptake. Cobalt occupies an interesting position in biology: it is an essential trace element at the heart of vitamin B12, yet certain cobalt coordination complexes display potent cytotoxicity toward malignant cells. Previous work by some of the same authors and by other groups has shown that cobalt(III) Schiff base complexes can interfere with angiogenesis, induce oxidative stress, and disrupt signaling pathways in cancer cells, and the present study extends that line of inquiry into the cell cycle machinery and the apoptotic threshold of colorectal cancer.

Colorectal cancer itself provides a compelling backdrop for this work. It is among the most commonly diagnosed malignancies worldwide, with global burden projections indicating a substantial rise in new cases and deaths through 2040, driven in part by aging populations and shifting dietary and lifestyle patterns. At the molecular level, colorectal tumors are characterized by well-mapped disruptions of the pathways that govern proliferation and survival, including the Wnt, RAS, and p53 axes, as well as dysregulation of the BCL-2 family of proteins that controls mitochondrial apoptosis. Because the cell cycle engine and the apoptosis machinery are so central to malignant behavior, agents that can simultaneously brake the former and disable the latter are of particular interest, and this dual action is precisely what the Iranian research team set out to document.

Methodologically, the study followed a logical escalation from gross cytotoxicity to molecular mechanism. The researchers first used the MTT assay, a colorimetric test in which metabolically active cells convert a yellow tetrazolium dye into a purple formazan product, to quantify how many HT-29 cells survived exposure to increasing concentrations of [CoL3]. The resulting IC50 of 3.47 micromolar indicates that the compound kills half of the cells at a remarkably low concentration, a potency that compares favorably with many experimental metal complexes and suggests efficient cellular uptake or intracellular activation. The team then turned to annexin V-FITC/propidium iodide staining, a flow cytometry technique that exploits the fact that early apoptotic cells expose the phospholipid phosphatidylserine on their outer membrane while still excluding the DNA-binding dye propidium iodide, whereas late apoptotic and necrotic cells become permeable to both probes. This staining revealed significant increases in both early and late apoptotic populations among treated cells, establishing that the cytotoxicity observed in the MTT assay was not merely nonspecific poisoning but genuine engagement of the apoptotic program.

To understand why the cells were dying and, crucially, why they were stalling before they died, the investigators analyzed cell cycle distribution by flow cytometry, which measures the amount of DNA in individual cells and thereby assigns them to the G0/G1, S, or G2/M phases. Treatment with [CoL3] produced arrest at both the G1/S and G2/M transition points, with statistical significance at p < 0.001 and p < 0.01 respectively. The G1/S checkpoint is the gate through which a cell commits to DNA replication, and its enforcement depends heavily on the cyclin D-CDK4 and cyclin D-CDK6 complexes, which phosphorylate the retinoblastoma protein and thereby release transcription factors needed for S-phase entry. Inhibiting CDK4 and CDK6 is a validated therapeutic strategy: palbociclib, ribociclib, and abemaciclib are approved CDK4/6 inhibitors that have transformed the treatment of hormone receptor-positive breast cancer, and there is active interest in extending this class to colorectal cancer, where cyclin-dependent kinase signaling is frequently dysregulated.

The molecular data in the new study align closely with that pharmacological logic. Real-time PCR measurements showed that [CoL3] significantly reduced the messenger RNA levels of CDK4 and CDK6, with mean relative expression values of 0.62 (p < 0.05) and 0.33 (p < 0.01) compared with untreated controls. Western blotting, which detects the proteins themselves, confirmed that the downregulation propagated to the protein level, with mean protein intensities of 0.48 for CDK4 and 0.79 for CDK6, both at p < 0.001. The concordance between transcript and protein suppression strengthens the interpretation that the compound acts upstream of these kinases rather than merely destabilizing their protein products as a secondary effect, and it provides a coherent mechanistic explanation for the observed G1 phase blockade. Meanwhile, the same treatment significantly downregulated BCL2 (p < 0.01), the prototypical anti-apoptotic guardian of the mitochondrial outer membrane, which helps explain how the arrested cells were subsequently funneled into apoptosis rather than simply pausing and recovering.

Adding a computational dimension to the experimental work, the researchers performed molecular docking analyses to model how [CoL3] would fit into the ATP-binding pockets of CDK4 and CDK6, the same sites occupied by clinically approved kinase inhibitors. The docking results demonstrated favorable spontaneous binding of the cobalt complex to both kinases, reinforcing the mechanistic involvement of the compound in cell cycle inhibition and suggesting a plausible physical basis for the kinase suppression observed in the cell-based assays. Docking studies of this kind cannot by themselves prove direct target engagement in living cells, and the authors appropriately frame the computational results as supporting rather than conclusive evidence, but the convergence of docking predictions, gene expression changes, protein-level downregulation, and cell cycle phenotypes creates an internally consistent mechanistic picture.

The study also builds on the group’s own prior work with the same compound. In an earlier publication in Biochimica et Biophysica Acta, the team reported that [CoL3] downregulates cyclooxygenase-2, encoded by the PTGS2 gene, in colorectal cancer cells by inducing a G-quadruplex structure in the gene’s promoter region, a four-stranded DNA conformation that can silence transcription. Together with the present findings, this suggests that the cobalt Schiff base may be a multi-target agent, capable of simultaneously dampening inflammatory signaling, cell cycle progression, and apoptotic resistance through distinct molecular interactions. Multi-target profiles are increasingly valued in oncology because cancers are heterogeneous diseases that often evade single-agent therapies, although they also raise the stakes for selectivity, since a compound that hits many targets in cancer cells may also hit important targets in healthy ones.

As with any cell line study, the distance between these results and the clinic remains substantial. HT-29 cells are a widely used but single model of colorectal cancer, and the effects of [CoL3] on normal colonic epithelial cells, on three-dimensional tumor organoids, and in animal models will need to be established before any consideration of human testing. Questions about pharmacokinetics, biodistribution, the stability of the octahedral complex in physiological conditions, and potential off-target toxicity all await answers. Nevertheless, the study offers a rigorous mechanistic demonstration that a rationally designed cobalt(III) Schiff base can arrest colorectal cancer cells at the G1/S and G2/M checkpoints and trigger early apoptosis by suppressing CDK4, CDK6, and BCL2, and it exemplifies the strategy of borrowing validated drug targets, in this case the kinases behind some of the most successful targeted therapies of the past decade, and attacking them with an entirely new chemical scaffold drawn from coordination chemistry. For a field searching for alternatives to platinum-based drugs, that is a result worth watching closely.

Subject of Research: Anticancer activity of a cobalt(III) Schiff base complex against colorectal cancer cells via CDK4/6 and BCL2 inhibition

Article Title: CRC cell cycle arrest and early apoptosis induction by mononuclear octahedral cobalt(III) Schiff base through the inhibition of CDK4/6 and BCL2

Article References: Moshtaghian, A., Colagar, A. H., Khaleghian, A., & Zahedi, T. (2026). CRC cell cycle arrest and early apoptosis induction by mononuclear octahedral cobalt(III) Schiff base through the inhibition of CDK4/6 and BCL2. Medical Oncology, 43(10), Article 265. https://doi.org/10.1007/s12032-026-03389-1

Image Credits: AI Generated

DOI: 10.1007/s12032-026-03389-1

Keywords: colorectal cancer, cobalt(III) Schiff base, CDK4, CDK6, BCL2, apoptosis, cell cycle arrest, HT-29, metal-based anticancer agents, molecular docking, flow cytometry, Western blotting

The research team, led by Arun Kumar Trivedi of the Division of Cancer Biology at CSIR-CDRI, with Samradhi Singh and Swati Srivastava as co-first authors, approached the question with an unbiased lens. Rather than testing a predetermined molecular pathway, they first treated the aggressive triple-negative breast cancer cell line MDA-MB-231 with ormeloxifene and then ran the protein contents of those cells through two-dimensional gel electrophoresis, a classic but powerful proteomic technique that separates proteins first by their electrical charge and then by their molecular weight. Differentially regulated protein spots were subsequently identified by MALDI-TOF/TOF mass spectrometry, allowing the team to build a catalog of the molecular players whose abundance shifted in response to the drug.

What emerged from that catalog was strikingly coherent. When the researchers performed functional categorization and pathway-level integration of the differentially regulated proteins, the changes did not scatter across unrelated cellular processes. Instead, they converged on a coordinated cellular response characteristic of stress-induced cellular senescence, the programmed state of stable cell-cycle arrest that healthy tissues use to quarantine damaged or potentially malignant cells. In other words, the proteome of the treated cells was telling a consistent story: ormeloxifene was not merely poisoning the cells or slowing them down temporarily, it was reprogramming them toward a senescent identity.

To confirm that impression, the team deployed a battery of complementary assays. MTT viability tests showed that ormeloxifene inhibited the growth of both MCF7 cells, an estrogen receptor-positive breast cancer line, and the triple-negative MDA-MB-231 line, with half-maximal inhibitory concentrations ranging from 7 to 9 micromolar. Colony formation assays demonstrated that the drug sharply reduced the ability of surviving cells to establish new colonies, with a notably stronger effect in MCF7 cells. Flow cytometric cell-cycle analysis then revealed the mechanistic core of the growth arrest: in MCF7 cells, ormeloxifene induced a G0/G1 arrest by decreasing the levels of CDK4 and CDK6, the cyclin-dependent kinases that normally drive cells through the restriction point of the cell cycle, while simultaneously increasing the CDK inhibitors p21Cip1 and p27Kip1, the molecular brakes that clamp down on those kinases.

The senescence evidence itself came from two independent histochemical stains. Senescence-associated beta-galactosidase staining, the traditional workhorse marker of senescent cells, showed a time-dependent increase in staining after ormeloxifene treatment. Sudan Black B staining, a more recently validated method that detects lipofuscin, the oxidized pigment granules that accumulate specifically in senescent cells, corroborated the result. Critically, the team also observed increased expression of established senescence markers by immunoblotting, and they describe the senescence as irreversible, meaning the cells did not resume proliferation even after prolonged exposure. That irreversibility matters enormously in cancer biology, because a reversible arrest can be escaped by tumor evolution, whereas a permanent one cannot.

Perhaps the most clinically tantalizing result, however, came from a combination experiment. The researchers found that ormeloxifene enhanced the sensitivity of MCF7 cells to tamoxifen, the gold-standard endocrine therapy for estrogen receptor-positive breast cancer that has been in use since the 1970s. Tamoxifen resistance remains one of the most stubborn problems in breast oncology, and any agent that can resensitize tumors to the drug would be of immediate translational interest. The authors suggest that the combination of ormeloxifene and tamoxifen may represent a potential therapeutic strategy for tamoxifen-resistant breast tumors, a hypothesis that will now need to be tested in more sophisticated models and, eventually, in patients.

The safety profile argument for repurposing ormeloxifene is unusually strong. The drug acts as a selective estrogen receptor modulator with mild estrogenic activity in bone and anti-estrogenic activity in uterine and breast tissue, a tissue-selective profile that mirrors the pharmacology of drugs like raloxifene. It has been used as a reversible, non-hormonal, once-weekly oral contraceptive in India for decades, and a phase II study published as far back as 1989 by Misra and colleagues evaluated centchroman as a non-steroidal anti-cancer agent in advanced breast cancer. Subsequent laboratory work has shown antiproliferative effects in head and neck cancer through modulation of the PI3K/mTOR pathway, efficient inhibition of ovarian cancer growth, suppression of desmoplasia and enhancement of gemcitabine sensitivity in pancreatic cancer, and G0/G1 arrest with ERK-mediated apoptosis in chronic myeloid leukemia cells. The new study adds senescence induction to this growing mechanistic portfolio.

Senescence as an anticancer mechanism has a complicated history. On one hand, inducing senescent cells in a tumor is a validated therapeutic concept, and chemotherapy itself frequently works partly by triggering senescence. On the other hand, senescent cells are not inert: they can secrete inflammatory and growth-promoting factors, and chemotherapy-induced senescence has been described as an adaptive mechanism that can drive resistance and tumor heterogeneity. The authors of the current study are careful to frame their findings as evidence that ormeloxifene restrains tumor progression by promoting irreversible senescence, and the irreversibility claim is central to that framing, since a senescent state that cannot be exited is far less likely to feed tumor regrowth than a transient one.

Technically, the study is a textbook example of how classical proteomics can still open doors in the era of genomics. Two-dimensional gel electrophoresis combined with MALDI-TOF/TOF mass spectrometry is labor-intensive and lower throughput than modern shotgun proteomics, but it excels at revealing shifts in protein abundance and isoforms in a visually interpretable format, and it allowed the team to detect the senescence signature before committing to targeted validation. The downstream validation cascade, spanning MTT assays, clonogenic survival, flow cytometry, dual senescence staining and Western blotting of cell-cycle regulators, follows the standard logic of discovery-driven cancer pharmacology: find the phenotype first, then confirm it with orthogonal methods.

For now, the work remains at the cell-culture stage, conducted in two breast cancer cell lines representing the two major clinical subtypes, hormone receptor-positive and triple-negative. The IC50 values in the low micromolar range are within the reach of achievable plasma concentrations for many oral agents, and the drug’s decades of human use as a contraceptive mean that its pharmacokinetics, dosing and safety in healthy women are already well characterized, potentially shortening the path to clinical evaluation considerably. The authors acknowledge support from the CSIR-National Laboratories Scheme and the DBT-BioCARe program, and the study was conducted at CSIR-CDRI, the institution where centchroman was originally developed. If the senescence mechanism holds up in animal models and combination studies with tamoxifen, a fifty-year-old contraceptive pill could find itself reborn as a senescence-inducing partner drug for breast cancers that have stopped responding to endocrine therapy, a repurposing story that few in the field would have predicted even a decade ago.

Subject of Research: Senescence induction by the selective estrogen receptor modulator ormeloxifene in breast cancer cells

Article Title: Ormeloxifene, a non-steroidal selective estrogen receptor modulator induces senescence in breast cancer cells

Article References: Singh, S., Srivastava, S., Siddiqui, S., Sharma, A., & Trivedi, A. K. (2026). Ormeloxifene, a non-steroidal selective estrogen receptor modulator induces senescence in breast cancer cells. Molecular Biology Reports, 53(1), Article 1665. https://doi.org/10.1007/s11033-026-12839-y

Image Credits: AI Generated

DOI: 10.1007/s11033-026-12839-y

Keywords: ormeloxifene, centchroman, breast cancer, cellular senescence, tamoxifen, CDK4, CDK6, p21Cip1, p27Kip1, drug repurposing, proteomics, selective estrogen receptor modulator

A new study published in BMC Genomics by Liang-Hao Ding of The University of Texas Health Science Center at Tyler and UT Southwestern Medical Center, together with Michael D. Story, Shoukath Sulthana, R Burke Squires, Elijah F. Edmondson, Michael M. Weil and colleagues, addresses this question at the level of whole genomes. The team performed whole genome sequencing on hepatocellular carcinoma (HCC), the most common form of primary liver cancer, that arose in mice after exposure to heavy-ion particle radiation or gamma-ray radiation, and compared the resulting mutational profiles with those of spontaneously arising liver tumors. Their goal was to identify genomic features that could serve as fingerprints of radiogenic cancer, and in particular of cancers caused by the HZE ions that dominate the deep-space radiation field.

The choice of experimental model is a critical part of the study’s design. Laboratory cancer experiments have traditionally relied on inbred mouse strains, which offer genetic uniformity but poorly reflect the genetic heterogeneity of human populations. In this work, the researchers used outbred mice, whose genomes carry a broad mosaic of genetic variants similar to the diversity seen in humans. This genetically diverse background provides a more realistic substrate for tumor development, allowing cancers to emerge through the interplay of random mutations, environmental insults and individual genetic susceptibility. An earlier study using the same model had already revealed a wide spectrum of cancer histotypes linked to different types of ionizing radiation, with hepatocellular carcinoma emerging as one of the most prevalent solid tumors following HZE ion exposure. That observation made radiation-associated HCC a natural target for deeper genomic interrogation.

The sequencing effort focused on three complementary layers of genomic information. The first was tumor mutation burden, or TMB, the total number of somatic mutations accumulated in a tumor’s genome. Mutation burden is a coarse but informative metric: radiation, particularly high linear energy transfer (LET) radiation, tends to inflict dense, clustered DNA damage that can elevate mutation counts above what spontaneous processes produce. The second layer was mutational signatures, the characteristic patterns of base substitutions and other mutation types that different mutagenic processes imprint on the genome. Just as forensic scientists match bullets to specific weapons, cancer geneticists decompose a tumor’s mutation catalogue into signatures attributable to specific causes, from ultraviolet light to tobacco smoke to defective DNA repair. The third layer comprised regions of hypermutation, localized stretches of the genome where mutations cluster at rates far exceeding the background, often marking sites of catastrophic or localized DNA damage and repair.

Across these layers, the analysis revealed both shared genomic features and distinct characteristics associated with the different radiation types. Radiation-associated tumors differed from spontaneous ones in their mutation burdens, their signature compositions and their hypermutation landscapes, and within the radiation-exposed group there were features that separated HZE-induced tumors from gamma-ray-induced ones. This distinction matters scientifically because gamma rays, as low-LET photons, produce sparse ionizations similar to X-rays, whereas HZE ions produce dense ionization tracks. If the genomic consequences of these two exposure types can be reliably separated, it opens the possibility of reading a tumor’s genome as a dosimeter of sorts, inferring not just that radiation played a role in its genesis but which kind of radiation was responsible.

One of the most striking findings concerns the gene Ctnnb1, which encodes beta-catenin, a central component of the Wnt signaling pathway that governs cell proliferation, differentiation and liver homeostasis. In human medicine, activating mutations in CTNNB1 are a well-known hallmark of non-viral, well-differentiated hepatocellular carcinoma, and they carry prognostic and diagnostic significance. In the mouse study, beta-catenin emerged as one of the most frequently mutated genes in the radiation-associated HCCs, mirroring its prominence in the human disease. By contrast, the spontaneous tumors showed a significantly lower frequency of Ctnnb1 mutation. This asymmetry suggests that radiation-driven liver carcinogenesis in this model converges on the same beta-catenin pathway that human non-viral HCC commonly exploits, while spontaneous tumors in these mice take other routes to malignancy. The convergence strengthens the translational relevance of the model, indicating that the radiation-induced mouse tumors recapitulate a molecular feature of the human cancers that mission planners ultimately care about.

The study also uncovered a chromosomal structural variant affecting the Cdk6 gene specifically in tumors from HZE-irradiated animals. CDK6 is a cyclin-dependent kinase that drives progression through the G1 phase of the cell cycle, and its dysregulation through amplification, translocation or altered regulation is a recognized mechanism of unchecked proliferation in several human cancers. Structural variants of this kind, which rearrange chromosomal segments rather than simply altering single DNA bases, are precisely the type of damage expected from the clustered, complex DNA breaks that dense ionization tracks produce. When a high-charge particle traverses the double helix, it can generate multiple breaks in close proximity, and the cell’s repair machinery, in stitching the fragments back together, can join the wrong ends, producing deletions, inversions, translocations and other rearrangements. Finding a Cdk6-affecting structural variant in the HZE group, and not in the comparison tumors, is consistent with this mechanistic picture and provides a candidate marker of heavy-ion exposure.

The broader implications of the work extend in two directions. For radiation biology, the results suggest that radiogenic hepatocellular carcinoma possesses unique genomic features, some of which are specifically tied to HZE radiation, offering a mechanistic window into how dense ionization tracks sculpt the cancer genome differently from sparse ionizations. For spaceflight risk assessment, the findings carry direct translational relevance. NASA’s cancer risk models for astronauts currently rest on thin empirical foundations, because epidemiological data for human HZE exposure are essentially limited to occupational cohorts such as atomic bomb survivors and radiation workers, whose exposures were predominantly low-LET. Experimental systems like the outbred mouse model used here, funded through NASA’s Specialized Center of Research on Radiation Carcinogenesis (NSCOR) program, help fill that gap by generating tumors under controlled, space-relevant exposure conditions and by supplying the genomic data needed to validate or refine risk extrapolations from mice to humans.

The technical infrastructure behind the study is also worth noting. Whole genome sequencing of multiple tumors, particularly from outbred animals whose reference-guided read alignment is complicated by genetic diversity, demands substantial computational resources. The authors acknowledge the support of the BioHPC facility at UT Southwestern and the Texas Advanced Computing Center, as well as the NASA Space Radiation Laboratory, where the radiation exposures were delivered. The work was funded by the NASA NSCOR programs under award NNX15AK13G, and all animal procedures were approved by the Colorado State University Institutional Animal Use and Care Committee. The article, published open access on 30 September 2026, was received in June 2025 and accepted in August 2026, reflecting the extended timeline typical of large-scale sequencing studies.

As human spaceflight moves from low Earth orbit toward the Moon and eventually Mars, the stakes of radiation carcinogenesis research rise accordingly. A round trip to Mars could expose crew members to doses and radiation qualities for which no human epidemiological data exist, and liver cancer, as this study shows, is among the malignancies that heavy-ion exposure promotes in a genetically diverse mammalian model. The mutational fingerprints described by Ding and colleagues, from elevated tumor mutation burdens and distinctive signatures to hypermutation regions, frequent beta-catenin mutation and HZE-specific structural variants affecting Cdk6, provide both a mechanistic foundation and a practical toolkit. If future studies confirm that these features reliably distinguish radiation-induced tumors, clinicians and researchers could one day examine a tumor’s genome and infer the radiation quality behind it, while mission designers could use the same knowledge to build more accurate cancer risk models and to develop shielding and countermeasure strategies grounded in the actual genomic consequences of the radiation astronauts will face.

Subject of Research: Mutational profiles of radiation-induced hepatocellular carcinoma in outbred mice exposed to heavy-ion and gamma-ray radiation

Article Title: Mutational profiles of hepatocellular carcinoma that arose in outbred mice after heavy-ion particle and γ-ray radiation

Article References: Ding, L.-H., Story, M. D., Sulthana, S., Squires, R. B., Edmondson, E. F., & Weil, M. M. (2026). Mutational profiles of hepatocellular carcinoma that arose in outbred mice after heavy-ion particle and γ-ray radiation. BMC Genomics. https://doi.org/10.1186/s12864-026-13304-7

Image Credits: AI Generated

DOI: 10.1186/s12864-026-13304-7

Keywords: space radiation, HZE ions, hepatocellular carcinoma, whole genome sequencing, mutational signatures, tumor mutation burden, beta-catenin, Cdk6, outbred mice, radiation carcinogenesis, liver cancer, NASA

The biological target at the heart of the study is a pair of enzymes known as cyclin-dependent kinases 4 and 6, or CDK4 and CDK6. These kinases act as gatekeepers of the cell division cycle, driving the transition from the G1 phase, when cells grow and prepare, into S phase, when DNA is replicated. When the CDK4/6 machinery is dysregulated, cells can escape normal growth controls and proliferate uncontrollably, a hallmark of many malignancies. The importance of these enzymes is underscored by the clinical success of approved CDK4/6 inhibitors such as palbociclib, ribociclib and abemaciclib, which have transformed the treatment landscape for hormone receptor-positive breast cancer and are being explored in other tumor types, including lung cancer.

Yet the existing drugs are not the end of the story. Resistance mechanisms emerge, side effects limit some patients, and clinicians and researchers continue to call for chemically diverse scaffolds, meaning molecules with fundamentally different structural backbones, that can potently inhibit both CDK4 and CDK6 at the same time. Finding such dual inhibitors by traditional high-throughput screening is expensive and slow, which is precisely why the research team turned to computer-aided drug discovery. Their strategy was to build a computational funnel: a wide opening that could swallow an entire chemical library and progressively narrower stages that would discard weak candidates until only the most promising molecules remained.

The first stage of the funnel was ligand-based machine learning. Rather than requiring knowledge of the target protein’s three-dimensional structure, ligand-based models learn from the known activities of compounds that have already been tested. The researchers assembled curated datasets from the ChEMBL database, a public repository of bioactivity data, gathering 265 compounds with measured activity against CDK4 and 402 against CDK6. Each molecule was converted into a numerical representation called an extended-connectivity fingerprint, specifically the ECFP4 format, which encodes the local atomic environments of a molecule in a way that machine learning algorithms can digest. Fingerprint-based descriptors of this kind have become a workhorse of computational chemistry because they capture subtle structural features that often correlate with biological activity.

With the data prepared, the team benchmarked multiple machine learning algorithms to see which could best predict inhibitory potency. The winner was a Bayesian Ridge regressor, a statistical learning method that combines ridge regression with Bayesian regularization, making it robust against overfitting on modestly sized datasets. Trained on the ECFP4 fingerprints, the model achieved cross-validated coefficients of determination, R-squared values, of 0.731 plus or minus 0.022 for CDK4 and 0.721 plus or minus 0.070 for CDK6. In practical terms, this means the model could explain roughly seventy percent of the variance in compound potency from molecular structure alone, a level of predictive accuracy strong enough to make meaningful prioritization decisions. The consistency of performance across both targets was crucial, because the goal was a dual inhibitor, and a model that excelled at one kinase but stumbled on the other would have undermined the entire approach.

The trained machine learning filter was then deployed against a library of 22,823 compounds. The Bayesian Ridge model scored every molecule for its likely activity against CDK4 and CDK6, allowing the researchers to discard the vast majority of the library computationally before any expensive calculation was run. This is where the economics of the approach become compelling: instead of assaying or docking tens of thousands of compounds, the team could concentrate structural modeling resources on the small fraction that the statistical model deemed most promising. The survivors of the machine learning filter then entered the second stage of the funnel, molecular docking against both CDK4 and CDK6.

Docking is a structure-based technique that uses three-dimensional models of the target proteins, typically derived from experimentally determined structures in the Protein Data Bank, to predict how a small molecule physically fits into the binding pocket of the enzyme. The docking calculations, performed at standard precision, evaluated the geometric complementarity and predicted binding energy of each candidate within the ATP-binding sites of both kinases. By requiring candidates to dock well against both CDK4 and CDK6, the researchers enforced dual-target affinity at the structural level, complementing the statistical predictions of the machine learning stage. This dual-target docking refinement whittled the prioritized list down to three candidate hits selected for biochemical evaluation in the laboratory.

The experimental results validated the computational strategy emphatically. Of the three candidates tested, HY-18,623 emerged as a potent dual inhibitor, with a half-maximal inhibitory concentration, IC50, of 3.5 nanomolar against CDK4 and 17.4 nanomolar against CDK6. Values in the low nanomolar range represent the kind of potency typically associated with advanced lead compounds, and achieving it against both kinases simultaneously with a single molecule is exactly what the field has been seeking. The result suggests that the computational funnel did not merely enrich for plausible binders but genuinely identified a molecule with drug-development potential.

To understand why HY-18,623 binds so tightly, the researchers turned to the third pillar of their workflow: molecular dynamics simulations. Docking provides a static snapshot of a predicted binding pose, but proteins and ligands in solution are constantly in motion, and a pose that looks perfect in a single frame may fall apart within nanoseconds. The team ran 200-nanosecond molecular dynamics simulations of the CDK4 and CDK6 complexes with HY-18,623, tracking the stability of the bound state over time using metrics such as root mean square deviation of the atomic positions. The simulations revealed that the compound maintains persistent hydrogen bonds with the so-called hinge residues of both kinases, Val96 in CDK4 and Val101 in CDK6. Hinge-region hydrogen bonding is a canonical anchor of kinase inhibitor binding, and its persistence throughout the simulations indicates a stable, high-affinity binding mode. Binding free energy analyses further quantified the favorable thermodynamics of the interaction, corroborating the experimental potency measurements.

Beyond the specific discovery of HY-18,623, the study offers a template for how modern drug discovery campaigns can be organized. The sequential integration of ligand-based machine learning, structure-based docking and dynamics-based validation creates a hierarchy of evidence, in which each stage interrogates the candidates from a different angle: statistical structure-activity relationships, physical fit within the binding site, and dynamic stability of the complex. The researchers have also made their data publicly available through a GitHub repository, supporting the open-science ethos that is increasingly shaping kinase drug discovery. For a field in which approved CDK4/6 inhibitors have already changed the course of breast cancer treatment, yet resistance and toxicity continue to motivate the search for alternatives, the arrival of a computationally discovered, nanomolar dual inhibitor with a structurally characterized binding mode is a development worth watching. HY-18,623 now stands as a promising lead candidate for further therapeutic development in oncology, and the workflow that produced it as a potentially reusable engine for discovering the next generation of kinase inhibitors.

Subject of Research: Computational virtual screening combining machine learning, docking and molecular dynamics to discover dual CDK4/6 inhibitors for cancer therapy

Article Title: Harnessing machine learning, docking and molecular dynamics for the virtual screening of compounds as CDK4/6 dual inhibitors

Article References: Wang, Y., Fang, L., Liu, Q., Zhang, Z., Xu, C., & Jiang, Y. (2026). Harnessing machine learning, docking and molecular dynamics for the virtual screening of compounds as CDK4/6 dual inhibitors. Molecular Diversity. https://doi.org/10.1007/s11030-026-11716-x

Image Credits: AI Generated

DOI: 10.1007/s11030-026-11716-x

Keywords: CDK4, CDK6, virtual screening, machine learning, molecular docking, molecular dynamics, drug discovery, kinase inhibitors, cancer, ChEMBL, ECFP4 fingerprints, HY-18,623