CD45RO-CD8+ T cell-derived exosomes restrict estrogen-driven endometrial cancer development via the ERβ/miR-765/PLP2/Notch axis.

As one of the most prevalent gynecological malignancies, uterine corpus endometrial cancer (UCEC) brings about approximately 200000 diagnosed cases and 76000 female deaths all over the world annually 1 , 2 . With all the innovating diagnosis, drugs and therapies emerging, the incidence of UCEC is rising 3 , and the 5-year survival rate of advanced UCEC patients is below 20%, which reflects a very poor prognosis 4 . Hence, novel insights into the molecular mechanisms and genetic characteristics of UCEC progression are urgently demanded. It has been a well-established consensus that overdosed estrogen exposure is a risk factor for UCEC, particularly the endometroid type 5 , which represents 80% of total UCEC cases 6 . Encoded by ESR1 and ESR2 gene respectively, estrogen receptors alpha (ERα) and estrogen receptor beta (ERβ) are classical intranuclear estrogen receptors in normal endometrium and UCEC cells 7 , and regulate several cell processes, like cell growth, differentiation, etc. 8 . To date, the functions of ERα is widely accepted as a tumor-promoter in hormone-responsive cancer, while the controversial role of ERβ is yet to be further elucidated. MicroRNAs (miRNAs) refer to small non-coding RNAs modulating protein accumulation post-transcriptionally 9 through inhibition of messenger RNA (mRNA) translation or acceleration of mRNA degradation 10 . MiR-765 has been reported to play a part in the progression of several cancers, such as osteosarcoma 11 , clear cell renal cell carcinoma 12 , gastric cancer 13 , and breast cancer 14 . However, no attention has been paid to the biofunctions of miR-765 in UCEC so far. Cell-cell interaction in the tumor microenvironment has been shown to directly affect tumor progression and metastasis in many ways 15 . Apart from cancer cells, numerous cell types play their indispensable roles in the tumor microenvironment, such as bone marrow-derived inflammatory cells, lymphocytes, blood vessels, fibroblastic cells, etc 16 . It has been widely accepted that the infiltrates of CD8 + T cells into the cancer niche are linked to better prognosis for patients in a number of human malignancies including UCEC 17 . As for CD8 + T cells, in addition to the classical tumor-killing mechanisms, emerging evidence has suggested that they exert effects on tumor by delivering exosomes to the recipient tumor cells, thus inducing the release of specific cargos, such as such mRNA, miRNA, proteins, and lipids 18 . However, researches on CD8 + T cell-derived exosomal miRNAs in UCEC are still scarce. The current study is to investigate the role and crosstalk of miR-765, estrogen and CD8 + T cell-derived exosomes of tumor microenvironment in the pathogenesis of UCEC in vitro and in vivo , and identify potential biomarkers and reveal novel molecular mechanisms for UCEC.

Signed informed consents were obtained from all the enrolled participants. Cancer tissues and adjacent normal tissues were collected from 15 endometrial cancer patients (35‐55 years old) whose surgical pathology specimens were identified as endometrial adenocarcinomas at the Department of Gynecology from May 2018 to August 2019. All tissues were preserved at -80 ℃ until use. This research was approved by the Ethics Committee of Obstetrics and Gynecology Hospital of Fudan University. Primary endometrial epithelial cells (EECs) were isolated from benign endometrium which were rinsed with phosphate buffered saline (PBS), minced into small pieces, and incubated in collagenase Ⅳ and DNase I in water bath at 37 °C. The cell suspensions were then filtered through cell strainers to separate epithelial cells. Ishikawa, RL95-2, and KLE cell lines were obtained from the Cell Bank of the Chinese Academy of Sciences (Shanghai, China). All the cells were cultured in DMEM/F12 (Gibco, Auckland, NZ) supplemented with 1% penicillin-streptomycin (HyClone, Utah, USA) and 10% certified FBS-charcoal-stripped (Biological Industries, Israel) at 37 °C with 5% CO 2 . Ishikawa or KLE cells were treated with or without estrogen (10 nM; Sigma-Aldrich, 50-28-2), an ERα inhibitor (MPP dihydrochloride hydrate, 10 μM; Sigma-Aldrich, 911295-24-4), an ERβ inhibitor (PHTPP, 10 μM; Sigma-Aldrich, 805239-56-9) or an ER inhibitor for both ERα and ERβ (Fulvestrant, 250 nM; Sigma-Aldrich, 129453-61-8), an inhibitor for Notch signaling (DAPT, 50 μM; Sigma-Aldrich, 208255-80-5), or an exosomes inhibitor (GW4869, 10 μM; Sigma-Aldrich, 6823-69-4) for 48 h, and then the transcriptional levels and protein levels of specific molecules or the bio-functions were detected. The GFP-labeled lentivirus vectors containing the has-miR-765 mimic lentivirus (miR-765 mimics) and the corresponding control miRNA lentivirus (miR-NC, negative control), the PLP2 over-expression lentivirus (OE-PLP2) and the corresponding control lentivirus (NC-PLP2) and the small interfering RNA si-PLP2 and the corresponding control si-NC were obtained from GeneChem (Shanghai, China). Transduction with the lentiviral vectors was conducted using transduction reagents and 8 mg/mL Polybrene (GeneChem) for 12 h. The siRNA transduction was conducted with Lipofectamine3000 Reagent according to the manufacturer's instructions (Invitrogen, Carlsbad, California, USA). After transfected into Ishikawa, RL95-2, and KLE cell lines, stable cell lines were then established, and the efficiency was confirmed by RT-qPCR. Cell Counting Kit‑8 (CCK‑8) reagent (Dojindo, Tokyo, Japan) was used to test cell viability according to the manufacturer's instructions. All the experiments were conducted at least in triplicate. All the transwell chambers (Corning Incorporated, Corning, NY, USA) were coated with 50 µL Matrigel (354480; BD Biosciences) and incubated at 37 °C for 1 h in advance. Next, 5 × 10 4 cells re-suspended in 200 µL FBS-free culture medium were seeded in the upper chamber. The bottom chamber was filled with 500 µL culture medium containing 20% FBS. After 24 h of incubation, a cotton swab was used to eliminate noninvasive cells in the upper chamber. Invasive cells were fixed in 4% paraformaldehyde, stained with 0.5% crystal violet, and imaged under an inverted microscope. Finally, invasive cells in five randomly selected fields were counted. A total of 2 × 10 6 cells/ml lymphocytes derived from peripheral blood of healthy human donors were cultured in RPMI 1640 complete medium (containing 50 mM 2-ME, 1 mM sodium pyruvate, 10 mM HEPES, and 10 U/mL IL-2). CD45RO + CD8 + or CD45RO - CD8 + T cells were purified by fluorescence activated cell sorting (FACS). Purified CD8 + T cells were then cultured in exosome-free RPMI 1640 complete medium for 48 h. Next, supernatants of cell culture medium were collected and then centrifuged at 800 × g for 5 min and followed by centrifugation of 2000 × g for 10 min to remove cellular fragments. Then, the medium was filtered through a 0.2 µm pore strainer (Syringe filter), and ultracentrifuged at 100000 × g for 2 h at 4 °C and then pelleted exosomes from CD45RO + CD8 + or CD45RO - CD8 + T cells were obtained. For transmission electron microscopy observations, exosome pellets were fixed in 4% paraformaldehyde at 4 °C for 1 h. Then, the pellets were loaded onto electron microscopy grids coated with formvar carbon, contrasted, and embedded in a mixture of uranyl acetate and methylcellulose. Sections were observed with a Philips Tecnai-10 transmission electron microscope operating at 80 kV (Phillips Electronic Instruments, Mahwah, NJ). Antibodies for flow cytometry assays were used for measurement of cell markers, which include FITC anti-mouse/human Ki-67 Antibody (151211, Biolegend, USA), APC anti-human CD8 antibody (980904, Biolegend, USA), PE anti-human CD28 antibody (302907, Biolegend, USA) and PE/Cyanine7 anti-human CD45RO Antibody (304229, Biolegend, USA). Matched immunoglobulin G (IgG) antibodies were used as isotype controls. Flow cytometry assays were performed using according to the manufacturer's instructions. Cell sorting was conducted with a Beckman CytoFLEX S flow cytometer (Beckman) using Becton CytExpert software. Data were analyzed using FlowJo V10 software. Total RNA was extracted by TRIzol (Invitrogen, Carlsbad, California, USA) from tissues and cells, and the concentration was quantified by a NanoDrop spectrophotometer (NanoDrop Technologies; Thermo Fisher Scientific, Inc.). The PrimeScript™ RT Reagent Kit (TaKaRa Biotechnology, Co., Ltd., Dalian, China) was used to reversely transcribe total RNA to cDNA. Next, SYBR Green PCR Master Mix (TaKaRa Biotechnology) was used to perform RT-qPCR. The miScript Reverse Transcription Kit (Qiagen GmbH, Hilden, Germany) and miScript SYBR Green PCR Kit (Qiagen GmbH) were used to measure miR-765 level for reverse transcription and RT-qPCR, respectively. U6 acted as the endogenous control. All reactions were processed on the Applied Biosystems 7500 Real-Time PCR System (Thermo Fisher Scientific, Inc.). Relative gene expression was analyzed using the 2 -ΔΔCt method. The primers were listed in Table S1 . Gene expression analysis was conducted by miRNA-seq and mRNA-seq for the conditions described in the relevant figures. For miRNA-seq, total miRNA was isolated using the miRNeasy Mini Kit (Qiagen GmbH, Hilden, Germany) according to the manufacturer's instructions. Quality of the total RNA was measured by the Agilent 2100 Bioanalyzer and samples with a RNA integrity number (RIN) higher than 7 were used for sequencing. cDNA libraries were generated using the NEBNext Multiplex Small RNA Library Prep Set for Illumina (New England Biolabs). Libraries were size selected using a 6% polyacrylamide gel and purified using the QIAQuick PCR Purification Kit (Qiagen GmbH). Purified libraries were normalized and pooled to create a double stranded cDNA library ready for sequencing. The samples were sequenced on the Illumina MiSeq to render 50 base pair single end reads. For mRNA-seq, treated cells were harvested for RNA extraction using TRIzol. Reagent and genomic DNA was removed using DNase I (Takara). The sequencing library was constructed after high-quality RNA was quantified and then sequenced with the Illumina HiSeq X Ten (2 × 150 bp read length). Treated cells were lysed in RIPA buffer supplemented with phenylmethylsulphonyl fluoride (Beyotime Institute of Biotechnology, Haimen, China). The BCA method was used to assess protein concentration. Proteins were separated by 12.5% SDS-PAGE and transferred to PVDF membrane and immunoblotted with the following antibodies: anti-PLP2 (1:1000, ab180131, Abcam, Cambridge, MA, USA), anti-E-cadherin (1:1000, Abcam, ab231303), anti-Vimentin (Abcam, ab92547), anti-GAPDH (1:1000, Abcam, ab181603), and anti β-actin (1:1000, Abcam, ab8226). After rinsed with TBST, the membranes were incubated with HRP-conjugated secondary antibody (1:10000; cat. no. A0208; Beyotime Institute of Biotechnology), developed with an enhanced chemiluminescence reagent (GE Healthcare Bio-Sciences, Pittsburgh, PA, USA) and visualized by Image Lab software (Bio-Rad Laboratories, Hercules, CA, USA). The experiments were performed in triplicate. UCEC cells were incubated with rabbit anti-human PLP2 antibody (1:200; ab180131, Abcam) overnight at 4 °C in a humid chamber. After washing three times with TBS (Beyotime, ST661), cells were incubated with dunkey anti-rabbit IgG H&L (Alexa Fluor® 488) preadsorbed antibody (1:400; 34206ES60, Yeasen, Shanghai, China). And cell nucleus was stained with 4′,6-diamidino-2-phenylindole (DAPI; Beyotime, C1006). Fluorescent in situ hybridization (FISH) for miR-765 in UCEC tissues using the RiboTM Fluorescent In situ Hybridization Kit (RiboBio, China). After fixing in a 4% (wt/vol) paraformaldehyde solution, samples were rinsed in 1 × PBS, permeabilized in 1 × PBS with 0.5% (vol/vol) Triton X-100 (10 min), washed in 1 × PBS with 0.1% (vol/vol) Tween-20 (1 min). Hybridization was carried out using the miR-765 FISH Probe Mix (RiboBio, China) in a humidified chamber at 37 °C for overnight. After RNA FISH, cells were fixed again for 5 min in 2% formaldehyde and subjected to immunofluorescence with anti-CD8 primary antibody (1:200; ab217344, Abcam) and fluorescent secondary antibody were sequentially. Immediately after DAPI stain solution was added, images were taken with immunofluorescence microscope (Leica, TCS SP5II). UCEC cells (1 × 10 7 ) stably expressing LV-miR-765-NC, LV-miR-765, LV-PLP2-NC, LV-si-PLP2, were subcutaneously injected into the left flank area of 4‑week‑old nude mice (n = 8 mice/group). Tumor volumes were measured everyday (0.5 × length × width 2 ). Five weeks later, the mice were sacrificed and xenografts were assessed and weighed. Experiments on animals were approved by the Ethics Committee for Animal Experimentation of Obstetrics and Gynecology Hospital of Fudan University and strictly confirmed to the Institutional Guidelines for Use and Care of Laboratory Animals. Ishikawa cells were transfected with the reporter gene firefly luciferase (Ishikawa/luc) and treated in vitro either with control siRNA (NC) or PLP2 siRNA (siPLP2). 3 × 10 6 Ishikawa/luc + NC, Ishikawa/luc + siPLP2 were transplanted intravenously into 4 weeks old female nude mice. After 6 weeks, endpoint measurements were performed using NightOwl LB 981 systems (Berthold Technologies GmbH & Co. KG, Bad Wildbad, Germany). Mice were anesthetized and intravenously injected with 150 μL luciferin (27 mg/mL, 200 mg/kg) (Biosynth, Staad, Switzerland). Measurements started 20 min after luciferin injection. For all images an exposition time of 5 min was chosen. After completion of in vivo imaging animals were sacrificed. The intensity of the bioluminescence signal was color coded and overlayed with bright field picture. All analyses were conducted by SPSS 25 (IBM). Between two groups, a paired or unpaired t test with or without Wilcoxon matched‐pairs signed‐rank test or a Mann‐Whitney U test was used according to the analysis of normal distribution and variance homogeneity. Among multiple groups, a one-way ANOVA test with or without a Kruskal-Wallis test according to the analysis of normal distribution and variance homogeneity. The data were presented as mean ± standard error (SEM) or median and quartile for normally distributed or other data. Log‑rank test was performed and Kaplan‑Meier survival curves were plotted. P < 0.05 was considered to indicate a statistically significant difference.

To explore potential aberrant miRNAs in endometrial cancer, we analyzed gene microarray of cancer tissues (n = 4) and normal tissues (n = 4) (Figure 1 A). Differential miRNAs were shown in the volcano plot (Figure 1 B). Among these differentially expressed miRNAs, the top five up- and down-regulated (e.g., miR-765) miRNAs were verified by RT-qPCR (n = 15) (Figure 1 B-C). Compared with normal paracancerous tissues, there was a significant decrease of miR-765 in UCEC clinical specimens (n = 15) (Figure 1 D). Similarly, three kinds of UCEC cell lines (Ishikawa, RL95-2 and KLE cells) had the lower levels of miRNA-765 (Figure 1 E). To further evaluate the role of miR-765 in UCEC, miR-765-overexpressed Ishikawa and KLE cells were constructed by mimics transfection (Figure 1 F and Figure S1 A). As shown, miR-765 mimics led to reduced cell proliferation and Ki-67 expression in Ishikawa and KLE cells in vitro (Figure 1 G-H and Figure S1 B-C). In vivo , miR-765 mimics xenografts developed much slower and survived longer than those of NC group (Figure 1 I-J, and Figure S1 D-E). In addition, miR-765 mimics obviously down-regulated the expression of several EMT-related markers (e.g., COL3A1, FN1, CDH2, S100A4, MMP9, SNAIL and ZEB1 ) and up-regulated TJP1 expression in tumor lesions (Figure 1 K and Figure S1 F). These data suggest that miR-765 should be a tumor suppressor gene for inhibiting cell proliferation and EMT process of UCEC. To search potential downstream targets of miR-765, we conducted bioinformatics analysis and the results showed that miR-765 might regulate Proteolipid protein 2 (PLP2) indirectly through miRNAs-cluster effects (Figure 2 A). PLP2, a four-transmembrane domain protein located in the endoplasmic reticulum 19 , 20 , has been regarded as an oncogenic-inducer in several cancers including melanoma, osteosarcoma, breast cancer, hepatocellular carcinomas, and acute lymphoblastic leukemia 20 - 22 . Notably, a negative correlation between the expression of miR-765 and PLP2 was corroborated in 15 cancer tissues (r = -0.7596, P < 0.01) (Figure 2 B). RT-qPCR assays showed that miR-765 mimics inhibited PLP2 transcription in Ishikawa and KLE cells (Figure 2 C). In addition, data of western blotting showed that miR-765 mimics markedly decreased the expression of PLP2 in Ishikawa and KLE cells (Figure 2 D and Figure S2 A). These results indicate that miR-765 negatively regulates PLP2 expression. Subsequently, miRNA sequencing was performed on NC and miR-765 mimics treated Ishikawa cells to screen the potential intermediate regulators between miR-765 and PLP2 (Figure 2 E). By means of a combined analysis of TargetScan ( http://www.targetscan.org/ ), PITA ( http://genie.weizmann.ac.il/pubs/mir07/mir07_data.html ), microT ( http://www.microrna.gr/microT ) databases and miRNA sequencing data, 4 miRNAs directly targeting PLP2 were screened out, including miR-3584-5p, miR-7-5p, miR-150-5p and miR-124-3p (Figure 2 F). We explored the specific mechanism and found that miR-765 regulated several transcription factors, which were closely related with transcription of miR-3584-5p, miR-7-5p, miR-150-5p and miR-124-3p ( Figure S2 B). Indeed, miR-765 mimics significantly enhanced the levels of these miRNAs in both Ishikawa and KLE cells (Figure 2 G and Figure S2 C). With the transfection of mimics of these miRNAs in Ishikawa and KLE cell lines, the level of PLP2 decreased consistently (Figure 2 H and Figure S2 D). Furthermore, we tested the effects of miRNA inhibitors on Ishikawa and KLE cells and found that miRNA inhibitors could effectively inhibit miRNA levels (Figure 2 I and Figure S2 E). Interestingly, the inhibitory effect of miR-765 mimics on the level of PLP2 in Ishikawa and KLE cell lines could be completely abrogated by miRNA inhibitors (Figure 2 J and Figure S2 F). These data suggest that the negative regulation of miR-765 on PLP2 should be dependent on the miRNAs cluster (e.g., miR-3584-5p, miR-7-5p, miR-150-5p and miR-124-3p) in UCEC cells. To explore the role of PLP2 in UCEC, we knocked down the expression of PLP2 and verified the efficiency by RT-qPCR and western blotting ( Figure S3 A-C). As shown, silencing PLP2 significantly suppressed the proliferation of Ishikawa and KLE cells (Figure 3 A), and induced an impaired EMT process, which was not only confirmed by downregulation of COL3, FN1, S100A4, SNAIL, ZEB1 as well as upregulation of TJP1 at the mRNA level (Figure 3 B), but also by downregulation of Vimentin and upregulation of E-cadherin at the protein level in Ishikawa and KLE cells (Figure 3 C and Figure S3 D-E). Additionally, absence of PLP2 inhibited the invasion of Ishikawa and KLE cells (Figure 3 D). In vivo, we also observed that PLP2 knockdown suppressed the metastasis of Ishikawa cells which were intravenously injected into mice and prolonged the survival of tumor-bearing mice (Figure 3 E-F). According to the TCGA database, in particular, low level of PLP2 was associated with a better prognosis in UCEC patients (Figure 3 G). Taken together, these results indicate that PLP2 should be a tumor-promoting protein by accelerating cell proliferation, EMT process, invasion and metastasis in UCEC. To further uncover the potential mechanism of PLP2 on EMT in UCEC cells, RNA sequencing was performed in NC and PLP2-overexpressed (OE-PLP2) Ishikawa and KLE cells. As depicted in Figure 4 A, the enrichment analysis of differential expression genes showed that EMT-related genes, NOTCH1 , NID and HES1 were significantly up-regulated in OE-PLP2 group. Next, we observed that PLP2 obviously promoted the activation of Notch signaling in Ishikawa and KLE cells (Figure 4 B-C and Figure S4 A-B). Additionally, DAPT, an inhibitor for Notch signaling pathway, could significantly restrict the stimulatory effect of OE-PLP2 on EMT-related molecules, cell invasion and proliferation in vitro (Figure 4 D-F and Figure S4 C-E). These data suggest that PLP2-meditaed EMT process and invasion of UCEC cells are dependent on the activation of Notch signaling pathway. In term of the vital driving factors of endometrial cancer, the effect of estrogen on miR-765/PLP2 levels was evaluated. As expected, stimulation with estrogen significantly decreased the expression of miR-765 in Ishikawa and KLE cells (Figure 5 A). However, these effects could be reversed by PHTPP (an ERβ inhibitor), and Fulvestrant (an ER inhibitor), but not MPP (an ERα inhibitor) (Figure 5 B-C). In contrast, estrogen up-regulated PLP2 expression, and the application of PHTPP and Fulvestrant rather than MPP could block this effect (Figure 5 D-I and Figure S5 A-C). Meanwhile, the up-regulation of PLP2 induced by estrogen could partly reversed by miR-765 mimics in Ishikawa and KLE cells (Figure 5 J-M and Figure S5 D-E). In treatment of advanced breast cancer, the selective estrogen receptor degrader fulvestrant, has been confirmed by improved efficacy and decreased side effects 23 . As shown, fulvestrant significantly down-regulated the expression of both ESR1 (coding gene of ERα) and ESR2 (coding gene of ERβ) in Ishikawa (ERα + , ERβ + ) and KLE (ERα - , ERβ + ) cells ( Figure S5 F-G). Subsequently, we further evaluated the possible treatment value of fulvestrant in UCEC, and found that fulvestrant alone up-regulated the expression of miR-765 ( Figure S5 H), but down-regulated the level of PLP2 ( Figure S5 I-K) in Ishikawa and KLE cells. Additionally, the knockdown of ERβ (si-ERβ) could partly abolished the regulatory effects of estrogen on miR-765 and PLP2 in Ishikawa and KLE cells ( Figure S5 L-M). These data demonstrate that estrogen regulates the miR-765 and PLP2 levels, and fulvestrant can restrict estrogen-driven regulation of miR-765 and PLP2 levels in UCEC cells in ERβ-dependent and -independent manners. Since CD8 + T cells function essentially in anti-tumor immunity, combined fluorescent in situ hybridization for miR-765 and immunofluorescent labeling for CD8 were carried out and the evidence of co-localization between miR-765 and CD8 + T cell was presented (Figure 6 A), revealing that miR-765 should be highly expressed in CD8 + T cells of UCEC lesions. To determine the possible source of miR-765 in CD8 + T cells, CD45RO + CD8 + T cells and CD45RO - CD8 + T cells were sorted by FACS ( Figure S6 A). Of note, RT-qPCR assay showed that a much higher level of miR-765 in CD45RO - CD8 + T cells was observed comparing with CD45RO + CD8 + T cells (Figure 6 B). To explore the source of miR-765, exosomes of CD45RO - CD8 + T cells and CD45RO + CD8 + T cells were purified and identified by TEM, flow cytometry and western blotting (abundant expression of CD63 and TSG101, and the absence of GRP78) respectively (Figure 6 C and Figure S6 B-F). Meanwhile, miRNA sequencing data of exosomal contents identified miR-765 as the most abundant miRNA in CD45RO - CD8 + T cell-derived exosomes (Figure 6 D). Interestingly, exosomes from CD45RO - CD8 + T cells significantly up-regulated miR-765 and down-regulated PLP2 levels in Ishikawa and KLE cells (Figure 6 E-G and Figure S6 G-I). However, the negative regulatory effect of exosomes on PLP2 levels in Ishikawa and KLE cells was completely abolished by GW4869 (an exosomes inhibitor) (Figure 6 H-I and Figure S6 J-K). Briefly, these results demonstrate that CD45RO - CD8 + T cell-derived exosomal miR-765 can suppress PLP2 expression in UCEC. Owing to the important roles of hormone-immune microenvironment on UCEC development, the dialogue between exosomes of CD45RO - CD8 + T cells and estrogen signaling in UCEC was evaluated. As illustrated in Figure 7 A-C, and Figure S7 A-C, exosomes of CD45RO - CD8 + T cells alone significantly up-regulated miR-765 but down-regulated PLP2 levels in Ishikawa and KLE cells. Additionally, exosomes partly or completely reversed estrogen-induced regulatory effects on miR-765 and PLP2 levels in Ishikawa and KLE cells. Subsequently, we observed that exosomes partly or completely abolished the stimulatory effects of estrogen on cell proliferation, Ki67 expression, EMT process and invasion of Ishikawa and KLE cells in vitro (Figure 7 D-I and Figure S7 D-I). Most noteworthy, treatment with exosomes alleviated the estrogen-induced tumor growth and poor prognosis of Ishikawa and KLE cells xenografts bearing mice (Figure 7 J-K and Figure S7 J-K). These data illustrate that CD45RO - CD8 + T-derived exosomes partially restricts estrogen-driven UCEC development via regulation of the miR-765/PLP2 axis.

As an indispensable hormone for both gender, estrogen plays pivotal roles in many physiological processes, including metabolism, immune responses, bone or cardiovascular health, reproductive functions, etc. 24 . However, abnormal levels of estrogen in human body, which derived from either endogenous production or exogenous ingestion, may lead to benign diseases or even malignant cancers. For instance, high level of estrogen may lead to obesity 25 , endometriosis 26 , UCEC, breast cancer 27 , etc. While osteoporosis 27 and diabetes 28 should be attributed to the decrease of estrogen. Generally, estrogen plays its roles through estrogen receptors (ER). As the ligand-bound ERα promotes growth of hormone-responsive cancer, the ERβ levels and/or the ERβ/ERα ratio decreases during the tumorigenesis of some malignant cancers 29 . Moreover, ERβ has been reported to inhibit the proliferation of prostate and breast cancer, which indicates the tumor-suppressor role of ERβ 30 , 31 , Accumulated studies have uncovered the critical role of estrogen in the tumorigenesis and development of UCEC 32 . However, the role and mechanism of ERβ signaling in UCEC remain largely unknown. In our study, we observed estrogen reduced miR-765 levels and increased PLP2 levels, further contributing to the EMT process and metastasis of UCEC. Interestingly, PHTPP and Fulvestrant could partly reverse these effects, suggesting estrogen-mediated miR-765/PLP2 expression regulation and UCEC development should be achieved in ERβ-dependent and independent manners. Apart from the classic ERα and ERβ, several non-classic estrogen receptors and their variants have been unveiled to the public, such as G-protein-coupled estrogen receptor-1 (GPER-1) 33 , estrogen receptor-related receptors (ERRs) 34 , and ERα-36 35 . However, other mechanisms of Fulvestrant on blocking the estrogen-miR-765/PLP2 axis need to be further clarified. Despite that UCEC has been already categorized as an estrogen-motivated disease, it is also recognized as a genetic disease 36 . It has been reported that miR-765 has dual regulatory functions on several cancers. Here, we found that there was a negative correlation between miR-765 and PLP2 in UCEC patients. Further analysis showed that absence of miR-765 led to higher proliferation, EMT process, invasion and development of UCEC by upregulation of PLP2. Different from a previous report 12 , the negative regulatory effect of miR-765 on PLP2 expression was indirect in UCEC cells. Notably, other miRNAs (e.g., miR-150-5p and miR-124-3p) also involved in targeting PLP2 37 , 38 . Owing to the bioinformatics analysis and further analysis, it can be speculated that intermediate regulators should be involved in this process, including miR-3584-5p, miR-7-5p, miR-150-5p and miR-124-3p. However, the detailed molecular mechanism remains to be further explored. EMT, the process in which epithelial cells obtain mesenchymal features, is associated with tumor initiation, invasion, metastasis, and resistance to therapy in cancer 39 . Currently, the relationship between miRNAs and EMT is widely acknowledged in a variety of cancers. However, the role of miRNAs in the EMT of UCEC is poorly understood 40 . The only reports suggest that both miR-194 and miR-200c suppress BMI-1 directly, thus resulting in the inhibition of PTEN or transcription factors ZEB1 and ZEB2 to preserve the epithelial phenotype in UCEC 41 , 42 . As for the involvement of miR-765 in EMT, Lv et al reported that via activating the ERK, miR-765 promoted the progression and stimulate EMT process of osteosarcoma 43 . Notably, our data showed that both the upregulation of miR-765 and the succedent downregulation of PLP2 resulted in the suppression of EMT in UCEC, which was reflected by the downregulation of Vimentin and the upregulation of E-cadherin. Additionally, under the positive regulation of estrogen, PLP2 accelerates the invasion and metastasis of UCEC in vitro and in vivo. The Notch signaling pathway has been reported to participates in tumorigenesis, including regulation of EMT 44 . In current studies, the data of RNA sequencing and further analysis indicates that the stimulatory effect of PLP2 on EMT, invasion and development of UCEC should be dependent on the activation of Notch signaling pathway. Although the molecule mechanism of PLP2 on Notch signaling pathway still unclear, the evidence above provides novel insights into the associations among estrogen, miR-765/PLP2, Notch signaling and EMT in UCEC progression. One of the pivotal steps in the process of host anti-tumor immunity is the activation of CD8 + T cells, during which numerous membrane vesicular bodies termed exosomes are released 45 . It has been reported that T cells release exosomes containing miR-142-3p, miR-142-5p, and miR-155, which can be transferred in active form to β cells favoring apoptosis, thus contributes to Type 1 Diabetes Development 46 . Naohiro et al reported that cytotoxic miR-298-5p-embedded exosomes released by CD8 + T cells from normal mice could kill lesional mesenchymal cells in fibroblastic tumor 47 . A high level of infiltrating and circulating CD45RO + CD8 + T cells in metastases has been found to be favorable for the outcome of different tumors 48 , 49 . In current study, we observed CD8 + T cells and miR-765 were co-located in UCEC lesions. It is worthy to point out that CD45RO - naïve CD8 + T cells possessed higher level of miR-765 compared with CD45RO + memory CD8 + T cell subsets. Exosomes released by CD45RO - CD8 + T cells could up-regulated the levels of miR-765 in UCEC, and further decrease PLP2 expression, contributing to impaired cell proliferation, invasion, EMT process, and better outcomes of UCEC in vivo . In addition, these exosomes could partly or completely restrict estrogen-triggered UCEC development by regulating the miR-765/PLP2 axis, indicating that there should be a potential value of CD45RO - CD8 + T cell-released exosomes for treatment of UCEC. However, it is unclear whether there are other molecules from CD45RO - CD8 + T cell-released exosomes that can cooperate with miR-765 in the regulation of anti-estrogen effects in UCEC. Additionally, the regulatory factors for exosomes release from CD8 + T cell in local UCEC lesions need to be further explored.

Collectively, as shown in Figure 8 , our data in vitro and in vivo suggest that aberrant low level of miR-765 leads to high proliferation, EMT process, invasion and poor prognosis of UCEC by the activation of PLP2-Notch signaling pathway. As a typical risk factor of UCEC, estrogen/ERβ regulates the miR-765/PLP2 axis and further accelerates disease development. Most importantly, Fulvestrant can reverse estrogen-mediated miR-765/PLP2 expression regulation and UCEC development in ERβ-dependent and independent manners, further explaining the new mechanism of Fulvestrant on UCEC treatment. Additionally, CD45RO - CD8 + T cell-derived exosomes release high level of miR-765, and limit the tumor-promoting effects of estrogen on UCEC via regulation of the miR-765/PLP2 axis. Therefore, these findings provide evidence that the ERβ/miR-765/PLP2/Notch signaling axis regulates the progression of UCEC, and indicate a novel perspective on the anti-tumor mechanisms of CD8 + T cells and exosomes. In addition, the potential therapeutic value of CD45RO - CD8 + T cell-derived exosomes in UCEC should be emphasized due to the negative regulation of estrogen/miR-765/PLP2 axis-mediated disease development.

Supplementary figures. Click here for additional data file.