High-grade serous ovarian cancer (HGSOC) is hallmarked by early onset of peritoneal dissemination, which distinguishes it from low-grade serous ovarian cancer (LGSOC). Here, we describe the aggressive nature of HGSOC ascitic tumor cells (ATCs) characterized by integrin α5high (ITGA5high) ATCs, which are prone to forming heterotypic spheroids with fibroblasts. We term these aggregates as metastatic units (MUs) in HGSOC for their advantageous metastatic capacity and active involvement in early peritoneal dissemination. Intriguingly, fibroblasts inside MUs support ATC survival and guide their peritoneal invasion before becoming essential components of the tumor stroma in newly formed metastases. Cancer-associated fibroblasts (CAFs) recruit ITGA5high ATCs to form MUs, which further sustain ATC ITGA5 expression by EGF secretion. Notably, LGSOC is largely devoid of CAFs and the resultant MUs, which might explain its metastatic delay. These findings identify a specialized MU architecture that amplifies the tumor–stroma interaction and promotes transcoelomic metastasis in HGSOC, providing the basis for stromal fibroblast-oriented interventions in hampering OC peritoneal propagation.

miR-9 mediates the downregulation of BRCA1 and impedes DNA damage repair in ovarian cancer. miR-9 may improve chemotherapeutic efficacy by increasing the sensitivity of cancer cells to DNA damage and may impact ovarian cancer therapy.

Quercetin (3,3',4',5,7-pentahydroxyflavone, Qu) is a promising cancer chemo-preventive agent for various cancers because it inhibits disease progression and promotes apoptotic cell death. In our previous study, we demonstrated that Qu could evoke ER stress to enhance drug cytotoxicity in ovarian cancer (OC). However, Qu-induced ER stress in OC is still poorly understood. Here, we demonstrated that Qu evoked ER stress to involve in mitochondria apoptosis pathway via the p-STAT3/Bcl-2 axis in OC cell lines and in primary OC cells. Unexpectedly, inhibition of ER stress did not reverse Qu-induced cell death. Further functional studies revealed that Qu-induced ER stress could activate protective autophagy concomitantly by activating the p-STAT3/Bcl-2 axis in this process. Moreover, the autophagy scavenger 3-MA was shown to enhance Qu's anticancer effects in an ovarian cancer mice xenograft model. These findings revealed a novel role of ER stress as a "double edge sword" participating in Qu-induced apoptosis of OC and might provide a new angle to consider in clinical studies of biological modifiers that may circumvent drug resistance in patients by targeting protective autophagy pathways.

Chemoresistance is a major unmet clinical obstacle in ovarian cancer treatment. Epigenetics plays a pivotal role in regulating the malignant phenotype, and has the potential in developing therapeutically valuable targets that improve the dismal outcome of this disease. Here we show that a series of transcription factors, including C/EBPβ, GCM1, and GATA1, could act as potential modulators of histone methylation in tumor cells. Of note, C/EBPβ, an independent prognostic factor for patients with ovarian cancer, mediates an important mechanism through which epigenetic enzyme modifies groups of functionally related genes in a context-dependent manner. By recruiting the methyltransferase DOT1L, C/EBPβ can maintain an open chromatin state by H3K79 methylation of multiple drug-resistance genes, thereby augmenting the chemoresistance of tumor cells. Therefore, we propose a new path against cancer epigenetics in which identifying and targeting the key regulators of epigenetics such as C/EBPβ may provide more precise therapeutic options in ovarian cancer.

Although tumor necrosis factor-related apoptosis-inducing ligand (TRAIL) has shown efficacy in a phase 2 clinical trial, development of resistance to TRAIL by tumor cells is a major roadblock. We investigated whether quercetin, a flavonoid, can sensitize human ovarian cancer cells to TRAIL. Results indicate that quercetin sensitized cancer cells to TRAIL. The quercetin induced expression of death receptor DR5 but did not affect expression of DR4 in cancer cells. The induction of DR5 was mediated through activation of JNK and through upregulation of a transcription factor CCAAT enhancer-binding protein homologous protein (CHOP); as silencing of these signaling molecules abrogated the effect of quercetin. Upregulation of DR5 was mediated through the generation of reactive oxygen species (ROS), as ROS scavengers reduced the effect of quercetin on JNK activation, CHOP upregulation, DR induction, TRAIL sensitization, downregulated the expression of cell survival proteins and upregulated the proapoptotic proteins. Furthermore, quercetin enhances TRAIL mediated inhibition of tumor growth of human SKOV-3 xenograft was associated with induction of apoptosis, activation of caspase-3, CHOP and DR5. Overall, our data suggest that quercetin enhances apoptotic death of ovarian cancer cells to TRAIL through upregulation of CHOP-induced DR5 expression following ROS mediated endoplasmic reticulum-stress.

There is an urgent need to make cisplatin (cDDP) more effective and less toxic in the treatment of ovarian cancer for its systemic side effects and high resistance rate. In this study, we investigated the effect of quercetin (Qu) pretreatment on the potentiation of cDDP in ovarian cancer. We found that Qu pretreatment significantly enhanced cDDP cytotoxicity in an ovarian cancer cell line and primary cancer cells. In addition, we demonstrated that Qu elicited obvious endoplasmic reticulum stress (ERS) and activated all three branches of ERS in ovarian cancer. Specific inhibitors of each ERS pathway, as well as the general ERS stabilizer tauroursodeoxycholic acid, notably diminished such enhancing effects. Furthermore, Qu notably suppressed STAT3 phosphorylation, leading to downregulation of the BCL-2 gene downstream of STAT3. Moreover, blocking ERS restored the protein levels of phosphorylated STAT3 as well as BCL-2 expression, thus abolishing the chemosensitization potency of Qu; these results revealed that Qu affected the STAT3 pathway to enhance cDDP cytotoxicity, and this effect involved ERS signaling. In a xenograft mouse model of ovarian cancer, Qu enhanced the antitumor effect of cDDP. Tumors from mice treated with cDDP in combination with Qu pretreatment had repressed STAT3 phosphorylation, lower BCL-2 and higher apoptosis levels compared with those from the other groups. Meanwhile, Qu markedly reduced the elevation of blood creatinine during cDDP intervention. These data indicate that Qu pretreatment potentiates the antitumor effects of cDDP in ovarian cancer while protecting the kidneys against damage. Therefore the strategy of Qu pretreatment may be beneficial in enhancing the therapeutic efficacy of cDDP against ovarian cancer.

Quercetin is proven to have anticancer effects for many cancers. However, the role of tumor suppressor p53 on quercetin’s radiosensitization and regulation of endoplasmic reticulum (ER) stress response in this process remains obscure. Here, quercetin exposure resulted in ER stress, prolonged DNA repair, and the expression of p53 protein; phosphorylation on serine 15 and 20 increased in combination with X-irradiation. Quercetin pretreatment could potentiate radiation-induced cell death. The combination of irradiation and quercetin treatment aggravated DNA damages and caused typical apoptotic cell death; as well the expression of Bax and p21 elevated and the expression of Bcl-2 decreased. Knocking down of p53 could reverse all the above effects under quercetin in combination with radiation. In addition, quercetin-induced radiosensitization was through stimulation of ATM phosphorylation. In human ovarian cancer xenograft model, combined treatment of quercetin and radiation significantly restrained the growth of tumors, accompanied with the activation of p53, CCAAT/enhancer-binding protein homologous protein, and γ-H2AX. Overall, these results indicated that quercetin acted as a promising radiosensitizer through p53-dependent ER stress signals.

The authors regret that in their original paper, the Masson's image of the control group in Fig. 7 J was incorrect as a result of an error during figure preparation. The corrected figure appears below.