Endometrial cancer (EC) is one of the most common gynecological cancers, with increasing incidence worldwide. The standard first-line chemotherapy combines paclitaxel and carboplatin, but a major obstacle to successful treatment is chemoresistance, the development of resistance by cancer cells that makes these drugs ineffective over time. Resistance to hormonal (endocrine) therapy with progesterone is an additional challenge in advanced and recurrent disease.
Improving treatment outcomes requires identifying the molecular drivers of drug resistance so they can be targeted pharmacologically. One known mechanism is epithelial-mesenchymal transition (EMT), a biological process in which cancer cells acquire a more invasive, stem-like, and drug-resistant phenotype. Wnt/beta-catenin signaling is a major activator of EMT and has been implicated in endometrial cancer chemoresistance and progesterone resistance.
The myosin family of molecular motor proteins, which move cargo within cells and regulate cellular architecture and mechanical forces, has recently been recognized as playing roles in cancer progression. A related family member, MYH9 (nonmuscle myosin IIA), was previously found to promote chemoresistance, proliferation, and metastasis in liver and nasopharyngeal cancer. Whether other myosin family members play similar roles specifically in endometrial cancer was unknown.
Natural compounds extracted from plants are a rich source of candidate anti-cancer agents. Sesamolin, a lignan extracted from sesame seeds (Sesamum indicum), has demonstrated anti-cancer activities in several tumor types, but its potential role in endometrial cancer had not been explored before this study.
Researchers began with a bioinformatics screen of 52 myosin family members using RNA sequencing data from 541 endometrial cancer patients and 35 normal controls from The Cancer Genome Atlas (TCGA) and the GEO database. They calculated a myosin activity score using the ssGSEA method and found that high overall myosin activity predicted poor survival in EC.
To identify the single most important myosin, two machine learning approaches were applied in combination. A random forest model ranked 15 differentially expressed myosins by their contribution to patient outcome prediction. A LASSO regression model independently validated prognostic myosins. These results were then integrated with a weighted gene coexpression network analysis (WGCNA), which groups genes that vary together across patients into functional modules. Only MYH14 appeared in the overlapping set across all four analytical approaches.
The clinical significance of MYH14 was confirmed by immunohistochemistry (IHC) on 118 endometrial cancer tissue sections and 17 adjacent normal controls, staining for MYH14 protein and correlating levels with patient survival and clinicopathological features. Multivariate Cox regression was used to test whether MYH14 is an independent prognostic factor controlling for age, stage, and grade.
Laboratory experiments used two endometrial cancer cell lines (Ishikawa and KLE). MYH14 was silenced using small interfering RNAs (siRNA) or overexpressed using plasmids. Drug sensitivity was measured using colony formation assays and cell viability assays with carboplatin, paclitaxel, and progesterone. Molecular docking and a Cellular Thermal Shift Assay (CETSA) confirmed that sesamolin physically binds to MYH14. Mouse xenograft models were used to test sesamolin's tumor-suppressing effects in vivo.
MYH14 was consistently upregulated in endometrial cancer compared to normal tissue across both TCGA and GEO datasets. Higher MYH14 expression correlated with higher tumor grade. Survival analysis confirmed that high MYH14 expression was associated with significantly shorter overall survival. Multivariate Cox analysis established MYH14 as an independent unfavorable prognostic indicator with a hazard ratio of approximately 1.5 to 1.6 (p less than 0.05) after controlling for age, stage, and grade.
Bioinformatics analysis showed that MYH14 expression positively correlated with a platinum drug resistance index (Spearman correlation with cisplatin: 0.189 and paclitaxel: 0.169, both p less than 0.001), pointing to MYH14 as a chemoresistance-promoting oncogene. When MYH14 was experimentally silenced with siRNA, cancer cells became significantly more sensitive to carboplatin, paclitaxel, and progesterone treatment. Silenced cells also showed increased DNA damage markers (gamma-H2AX) after drug treatment, confirming the drugs were doing more harm to the cancer cells when MYH14 was absent.
MYH14 knockdown also reduced cancer cell proliferation, migration, and invasion in wound healing and transwell assays, demonstrating that MYH14 promotes multiple hallmarks of malignancy beyond drug resistance. These functional results positioned MYH14 as a multifunctional oncogene in endometrial cancer.
Molecular docking analysis predicted that sesamolin from sesame seeds physically binds to MYH14 with higher docking energy than its metabolite sesamol. This prediction was confirmed experimentally using a Cellular Thermal Shift Assay (CETSA), which showed that sesamolin increased the thermal stability of the MYH14 protein in living cells, a hallmark of a direct drug-protein interaction.
In cell culture experiments, sesamolin treatment in a dose-dependent manner sensitized endometrial cancer cells to carboplatin and paclitaxel, increased DNA damage markers in drug-treated cells, restored sensitivity to progesterone treatment, and reduced cancer cell proliferation, migration, and invasion. These effects mirrored exactly what happened when MYH14 was silenced by siRNA, confirming that sesamolin works specifically through MYH14 inhibition.
In mouse xenograft experiments, oral sesamolin administration significantly reduced tumor size and weight compared to untreated controls, without causing significant body weight loss. This no-toxicity finding is important for assessing sesamolin's potential as a clinical candidate, as many chemotherapy agents cause significant systemic side effects. The in vivo data confirm that sesamolin's anti-tumor activity seen in cell lines translates to a living organism.
Sesamolin also worked synergistically with existing chemotherapy: when combined with carboplatin and paclitaxel in colony formation assays, the combination performed better than either agent alone. This suggests sesamolin could be used as a sensitizing agent alongside standard-of-care chemotherapy rather than as a replacement, making its clinical development more straightforward.
The mechanistic connection between MYH14 and drug resistance was traced through its interaction with MYH9. Co-immunoprecipitation confirmed that MYH14 and MYH9 physically bind to each other in endometrial cancer cells. MYH9 had previously been shown to block the enzyme GSK3-beta from attaching ubiquitin tags to beta-catenin, the key activator of the Wnt signaling pathway. Without ubiquitin tags, beta-catenin is not degraded and accumulates in the cell, driving Wnt target gene expression.
MYH14 amplifies this mechanism: it forms a complex with MYH9 that further blocks GSK3-beta from degrading beta-catenin. When MYH14 was silenced or blocked by sesamolin, GSK3-beta was released from inhibition, ubiquitin tagged and degraded beta-catenin normally, and Wnt signaling activity fell measurably by luciferase reporter assay. The result was reduced EMT, increased drug sensitivity, and decreased cell invasiveness.
Sesamolin acts as a protein-protein interaction (PPI) inhibitor: it binds MYH14 and physically disrupts the MYH14-MYH9 protein complex. This is a significant finding because PPI inhibition is an emerging drug development strategy that targets previously undruggable interactions between proteins rather than the active site of a single enzyme. Sesamolin provides a natural compound template for developing more potent MYH14-MYH9 disruptors.
This study establishes MYH14 as a newly identified oncogene in endometrial cancer, independently predicting poor patient survival and driving chemoresistance and endocrine therapy resistance through activation of the Wnt/beta-catenin signaling pathway. Machine learning was essential to this discovery, narrowing 52 myosin family members to the single most clinically relevant candidate through an intersection of random forest, LASSO, and network coexpression analysis.
Sesamolin, a natural lignan extracted from sesame seeds, directly inhibits MYH14, disrupts its interaction with MYH9, and thereby suppresses Wnt signaling and sensitizes endometrial cancer cells to existing therapies. Critically, sesamolin showed no toxicity in mouse models while achieving significant tumor suppression, suggesting it could be safely incorporated into multi-drug regimens.
Future studies should investigate the effect of sesamolin in more advanced preclinical models, including patient-derived organoids and immune-competent mouse models that better recapitulate the human tumor microenvironment. Clinical trials evaluating sesamolin as a chemotherapy sensitizer in patients with MYH14-high endometrial cancer would represent the next logical step in translating these findings to patient benefit.