In this study, the research team investigated the effects of quercetin on tumor stem cells derived from ECs, including spheroid formation, stemness maintenance, and malignant behavior. We also explored the potential mechanisms underlying the regulatory effects of quercetin, which may be achieved through the inhibition of the STAT3/JAK2 signaling pathway, possibly dependent on the presence of estrogen receptors.

To bolster the generalizability of our conclusions, we expanded our experimental model to include a panel of endometrial cancer cell lines with defined ERα status beyond the original EMN8 and EMN21 systems. This panel comprised ERα-positive lines (Ishikawa and ECC-1) and ERα-negative lines (HEC-1-A and KLE). Cancer stem cells (CSCs) were enriched from all lines and subsequently treated with quercetin (Supplementary Figure S2A). Analysis revealed that the ERα levels in CSCs derived from ERα-positive Ishikawa and ECC-1 cells were markedly lower than those in their respective parental cells (PCs) (Supplementary Figure S2B). Following quercetin treatment, these CSCs exhibited a significant reduction in the expression ratios of key stemness markers (Oct4 and Nanog) and in the phosphorylation levels of STAT3 and JAK2 (p-STAT3/STAT3 and p-JAK2/JAK2). In contrast, both the parental and CSC populations of the ERα-negative lines HEC-1-A and KLE consistently exhibited low basal ERα expression. Accordingly, quercetin treatment in these models did not elicit a significant inhibitory effect on the ratios of stemness markers or on the activation status of STAT3 and JAK2 (Supplementary Figure S2C, D). Building on these findings, a key priority for our future work will be to employ patient-derived organoids (PDOs) to enhance the clinical relevance of our research.

Moreover, to rigorously verify that the inhibitory effect of quercetin on endometrial cancer stem cells (CSCs) is dependent on ERα, we performed additional experiments using models with well-defined ERα status. Specifically, ERα was knocked down in ERα-positive Ishikawa-CSCs and ECC-1-CSCs (knockdown efficiency data are provided in Supplementary Figure S1-F). Results demonstrated that ERα knockdown significantly attenuated the suppressive effects of quercetin on CSC stemness markers (OCT4 and NANOG) as well as on the phosphorylation levels of STAT3 and JAK2 (p-STAT3/STAT3 and p-JAK2/JAK2). This effect was consistent with that observed following treatment with the ERα-specific inhibitor Fulvestrant (Supplementary Figure S3-A-C).

Conversely, we overexpressed ERα in ERα-negative HEC-1-A-CSCs and KLE-CSCs (overexpression efficiency data are shown in Supplementary Figure S1-G). Following ERα overexpression, quercetin treatment led to a marked enhancement in the inhibition of stemness markers (OCT4 and NANOG) and signaling activity (p-STAT3/STAT3 and p-JAK2/JAK2) in these CSCs (Supplementary Figure S3-D-F).

Furthermore, to better elucidate the kinetic relationship between ERα modulation and downstream signaling, we performed a time-course experiment in which EMN8/EMN21-derived CSCs were treated with quercetin and harvested at 0 h, 2 h, 6 h, and 12 h. Western blot analysis was used to monitor dynamic changes in ERα and key components of the JAK2/STAT3 pathway. The results demonstrate that the inhibitory effect of quercetin on the expression of stemness markers (OCT4 and NANOG) and the phosphorylation of STAT3 and JAK2 (p-STAT3/STAT3 and p-JAK2/JAK2) in CSCs strengthened progressively over time (Supplementary Figure S4).

It is worth noting that quercetin exhibits limited bioavailability. Previous pharmacokinetic studies reported that after oral administration in an aqueous suspension in rats, quercetin had an absolute bioavailability of 16% and a maximum plasma concentration (Cmax) of 2.01 µM. When formulated with ethanol and polyethylene glycol 200, its bioavailability increased to 27.5%, with a Cmax of 3.44 µM [24]. Although the concentrations used in our in vitroexperiments are higher than these reported plasma levels, such discrepancies are common in in vitrostudies due to factors such as protein binding, cellular uptake efficiency, and the absence of metabolic clearance. Moreover, we demonstrated that ERα-positive CSCs exhibit heightened sensitivity to quercetin. As shown in Fig. 5, ERα-overexpressing CSCs were effectively inhibited at lower concentrations, suggesting that ERα status could serve as a predictive biomarker for quercetin response. This supports the biological relevance of our chosen doses and underscores the potential for efficacy in sensitive cellular contexts.

To directly address the concern about toxicity to normal endometrial tissue, we conducted additional experiments using the immortalized human endometrial epithelial cell line (hEEC). Treatment with quercetin at concentrations ranging from 0 to 100 µM for up to 72 h did not significantly reduce cell viability, as assessed by CCK-8 assay (Supplementary Figure S1D). These results confirm that the concentrations effective against CSCs are not cytotoxic to normal endometrial cells, indicating a favorable therapeutic window.

In summary, the combination of pharmacokinetic data from the literature, our findings on ERα-dependent sensitivity, and new toxicity assessments in normal cells provides a strong rationale for the concentrations used in this study. We are optimistic that formulation strategies or targeted delivery approaches may further enhance quercetin’s bioavailability and efficacy in future translational applications.

Many studies and clinical trials have shown the existence of cancer stem cells (CSCs) in certain types of tumors, including acute myeloid leukemia [25], head and neck squamous cell carcinoma [26], breast cancer [27], and others. Based on these studies, it is likely that the unlimited proliferation and metastasis of malignant cells are regulated and driven by these CSCs. Some researchers have suggested that the presence of CSCs may contribute to treatment failure in ECs [28]. For example, dysregulation of CSC signaling pathways, such as Hippo/YAP-1, Wnt/β-catenin, and Hedgehog (Hh), has been associated with tumor maintenance and resistance to therapy in esophageal cancer [29, 30]. In this study, we found that the addition of quercetin not only inhibited the formation of stem cell spheroids in a serum-free culture environment but also caused the disintegration of pre-formed stem cell spheroids. During this process, the expression levels of stemness-related factors also decreased, indicating that the inhibitory effect of quercetin on stem cell spheroids is achieved by inhibiting stemness rather than exerting cytotoxicity. There are various types of stem cell markers for tumor stem cells, such as cell surface markers differentiation cluster (CD) 271 and CD90, which have been routinely used to isolate cell subpopulations with stem-like properties from cancer cell banks, enhancing tumor-initiating ability, increasing recurrence rates, and increasing resistance to treatment [31, 32]. Another study suggested that transforming acidic coiled-coil protein 3 (TACC3) could be used to determine whether TACC3 can serve as a biomarker for the diagnosis and prognosis of ECs, as it can form spheres and display resistance to various drugs in vitro [33]. In this study, ALDH1A1, c-Myc, Nanog, and Oct4 were used as stemness markers, and their expression levels were positively correlated with the spheroid-forming ability of stem cells, demonstrating the feasibility of using them as stemness markers. However, in future research, more reported stemness markers still need to be introduced.

Quercetin is widely found in vegetables and fruits and belongs to the class of flavonoids. Studies have found that it has inhibitory effects on various tumor cells, including breast cancer, prostate cancer, liver cancer, esophageal cancer, ovarian cancer, and others [19]. Currently, both in vivo and in vitro studies have shown that quercetin has inhibitory effects on colon cancer. It not only inhibits the proliferation of colon cancer cells and induces apoptosis but also reduces the number of aberrant crypt foci in the colon [34]. The specific mechanisms underlying the anticancer effects of quercetin are not yet fully understood but may be achieved through the regulation of multiple molecules. In this study, we found that quercetin has an inhibitory effect on the stemness of tumor stem cells, and the expression of estrogen receptor alpha (ERα) is crucial for the sensitivity to quercetin. Overexpression of ERα allows relatively low concentrations of quercetin to inhibit the stemness of tumor stem cells. It is worth noting that overexpression of ERα does not affect stemness and cell behavior, indicating that the action of quercetin primarily relies on the expression of ERα. Estrogen can activate estrogen receptors and downstream signaling pathways, but we did not explore whether quercetin still exerts its tumor-suppressive effects in the presence of estrogen, which needs further investigation in future studies.

Additionally, while this study has not explored whether quercetin influences other signaling pathways such as PI3K/AKT, MAPK, or NF-κB, this limitation reflects our focused experimental design. From the outset, our primary objective was to rigorously test a specific hypothesis: that the ERα–JAK2–STAT3 axis serves as the central pathway through which quercetin suppresses cancer stem-cell-like properties. We believe that the collective evidence provided by genetic knockdown, overexpression, and pharmacological inhibition experiments strongly and sufficiently supports this conclusion.

Although a comprehensive investigation into additional pathways would be valuable, it lies beyond the current scope of our work. Our results consistently and specifically implicate the ERα–JAK2–STAT3 pathway in mediating quercetin’s effects. That said, we acknowledge that contributions from other mechanisms—such as PI3K/AKT or NF-κB—cannot be formally excluded and would represent a meaningful direction for future research.