Hypoxia-Induced miR-210 Promotes Endothelial Cell Permeability and Angiogenesis via Exosomes in Pancreatic Ductal Adenocarcinoma

Background Exosomes have been proven to play important diagnostic, regulatory, or communication roles in tumorigenesis, tumor progression, or metastasis; in recent studies, lots of molecules, including miRNAs, were found to be aberrantly expressed in tumor exosomes and were correlated with tumor development. However, studies about the expression, relationship, or control mechanisms of miRNAs in exosomes in pancreatic ductal adenocarcinoma (PDAC) are scarce and urgently needed. The aim of this article was to identify and investigate abnormally expressed miRNAs in PDAC exosomes in vivo and in vitro. Methods Microarray studies were used to detect aberrantly expressed miRNAs in PDAC exosomes, and miR-210 expression in cells or exosomes was further analyzed by qRT-PCR. Bioinformatics analyses, dual-luciferase assays, WB, and other assays were utilized to explore the miRNA molecular mechanisms. The living cell coculture model and immunofluorescence analysis were employed to image the communication between PDAC cells and endothelial cells. Other biological experiments in the study include a real-time intravital imaging system, EdU, transwell, xenograft models, and so on. Results miR-210 is significantly expressed in PDAC exosomes and malignant cells. High miR-210 significantly facilitated tumor angiogenesis, cell invasion, and proliferation in PDAC cells. Further mechanistic detection revealed that miR-210 negatively regulated EFNA3 expression and participated in the PI3K/AKT/VEGFA or Wnt/Β-catenin/RHOA pathways, thus promoting tumor angiogenesis and cellular permeability. PDAC cells promote endothelial angiogenesis or permeability via miR-210 transmission by tumor exosomes. Exosomal miR-210 promotes PDAC progression in vivo. Further detection of PDAC plasma exosomal miR-210 suggests that exosomal miR-210 expression was high and significantly associated with vascular invasion and TNM stage and was an independent risk factor for PDAC overall survival. Conclusions PDAC cell-secreted exosomes could promote angiogenesis and cellular permeability of neighboring endothelial angiogenesis or permeability via miR-210 transmission. Exosomal miR-210 may play important roles in tumor biology and may be a useful prognostic marker in PDAC.


Background
Pancreatic ductal adenocarcinoma (PDAC) is currently one of the most malignant and lethal tumors worldwide [1]. In recent decades, PDAC has been the only tumor with an increasing mortality rate, despite marked improvements in prognosis with various other cancers due to tremendous advances in diagnosis and treatment techniques [2]. According to the latest report, the median survival time of PDAC patients is approximately 6 months and the 5-year survival rate is less than 8% [3]. Radical resection is still the most important treatment at present. However, nearly 80% of PDAC patients have been in the middle or advanced stage and have lost their opportunity for surgery at the frst admission and less than 20% of PDAC patients have the opportunity for radical surgery [4]. Even so, the 5-year survival rate after surgery is only approximately 25% [5]. Terefore, the key to the treatment of PDAC is to improve the early diagnosis rate and to fnd new therapeutic targets that inhibit invasion and metastasis.
Exosomes are cell-secreted, nanosized bilayer membrane extracellular vesicles with diameters ranging from 40-100 nm [6]. Exosomes can carry a diferent number of bioactive molecules, including proteins, DNA, various types of RNAs (such as miRNAs, mRNAs, lncRNAs, and circular RNAs), lipids, and metabolites according to the cells of origin via cellular communication [7][8][9]. Exosomes were initially viewed as waste products produced during diferent physiologic cell metabolisms [10]. However, increasing evidence has shown that exosomes play an important role in cellular communication by transferring exosomal molecules to recipient cells, thus resulting in a series of physiological changes, including promoting cell growth, proliferation, and migration [11,12]. Exosomes can be secreted by various cells, including tumor cells, and are commonly found in various human body fuids, including plasma and ascites [13]. In turn, various abnormal biological alterations of diferent cells (including tumor cells) can also be refected by detecting the expression of exosomal molecules (including miRNAs) in body fuids [14]. Recent studies have shown that tumor cells can secrete many more exosomes than normal cells, and abnormal exosomes can be easily detected in various body fuids, such as urine, saliva, ascites, or blood [9]. Many studies have shown that the abnormal expression of miRNA in tumor exosomes is closely related to various biological changes, such as angiogenesis, tumor invasion, and metastasis [14,15]. Te detection of abnormally expressed molecules in exosomes may play an important role in the early diagnosis, prognosis, or monitoring of the treatment of cancer and provide new therapeutic targets to treat tumors via exosomes. Terefore, the detection of exosome-based molecules in body fuids may provide novel methods for the early diagnosis and treatment of PDAC.
In this study, we identifed miR-210 in PDAC exosomes by miRNA microarray analysis, and the biological functions and detailed regulatory mechanisms were also detected. We also visualized miR-210 transmission between PDAC cells and endothelial cells. Finally, the expression of miR-210 in plasma exosomes was further detected and its relationship with clinical survival was analyzed.

Cell Culture and Transfection.
Te cell lines involved in the study were authenticated by STR profling, and Hs766T and HUVECs were purchased from American Type Culture Collection (ATCC, Manassas, USA). Other PDAC cell lines, including Hs766T-L1, Hs766T-L2, and Hs766T-L3, are frst-, second-, and third-generation primary cells from nude mouse liver metastatic tissue of Hs766T, respectively, as described in our previous paper [16]. Cells were incubated in 5% CO 2 at 37°C and cultivated in RPMI 1640 (Gibco, USA) or DMEM (Gibco, USA) supplemented with 10% FBS (HyClone, USA). Transfections of plasmid and shRNA were performed using the transfection reagent Lipo 3000 (Invitrogen, USA) according to the manufacturer's instructions.

Cell Invasion Experiment.
Cell invasion was examined using Transwell assays. In brief, control or transfected PDAC cells resuspended in a FBS-free medium were cultured in Transwell migration chambers (8 μm polycarbonate flter) precoated with Matrigel (BD, USA). After 36 h, cells that failed to invade from the top of the membrane were removed with a cotton swab, and cells on the bottom of the membrane were fxed and then stained. Invaded cell images from fve randomly selected grid squares were counted.

EdU Incorporation Assay.
Te EdU assay was conducted with a keyFluor488 Click-iT EdU Detection Kit (KGA331-100, KeyGene, Nanjing, China) according to the manufacturer's instructions. In brief, control or transfected PDAC cells were incubated with EdU, fxed and permeabilized, and then incubated with Click-iT. After further incubation of the nucleus with DAPI (Beyotime, China), the slides were fnally imaged with a fuorescence microscope.

RNA Isolation and PCR Analysis.
Total RNA from PDAC cells or clinical samples was extracted by using TRIzol and TRIzol LS (Termo, USA) reagent according to the manufacturer's instructions. Te cDNA was generated with PrimeScript RT Reagent Kit, with gDNA Eraser (TaKaRa, Japan), or Mir-X miRNA qRT-PCR SYBR Kit (Clontech, Japan) according to the manufacturers' protocols. Real-time PCR was performed using SYBR Premix Ex Taq (TaKaRa, Japan) on a CFX96 Real-Time System (Bio-Rad, USA) based on the instructions.

Luciferase
Reporter Assays. Diferent treated cells were transfected with pGL3 EFNA3 or pGL3 mut-EFNA3 plasmid (Sangon Biotech, China) and the internal control pRL-TK Renilla Luciferase Plasmid (Promega, USA) combined with miR-210 mimics in 96-well plates. After incubation, the cells in the plate were harvested and processed according to the manufacturer's protocol (Dual-Luciferase Reporter Assay System, E1910, USA). Te fnal frefy luciferase activity was normalized to Renilla luciferase activity.
2.6. Western Blot Analysis. WB analysis was performed as described previously [17]. Generally, total protein was extracted by RIPA Lysis Bufer (Termo, USA). Protein was measured, separated, and transferred to PVDF membranes (Millipore, USA) after further incubation with primary and secondary antibodies overnight at 4°C. Te fnal protein was visualized by the Super ECL Detection Kit (KeyGEN Bio-TECH, China). Te antibodies used in this study were as follows: anti-HIF-1α (1 : 1000, #36169S, Cell Signaling, USA), anti-EFNA3 (1 : 1000, ab153706, Abcam, USA), anti-AKT 2.7. Exosome Experiments. As described previously, PDAC cells were cultured with exosome-depleted FBS, and 10 ml of conditioned medium was used for exosomes extraction with a Total Exosome Isolation Kit (Termo, USA) according to the manufacturer's protocol. Te FLUOROCET Ultrasensitive Exosome Quantitation Assay Kit (SBI, USA) was used to ensure the same number of exosomes was used in each experiment. As discussed before, 5 × 108 exosomes were used for exosomal RNA extraction and 1 × 10 8 exosomes were used for exosome-stimulation experiment in vitro. Te size and morphology were further confrmed by transmission electron microscopy (TEM) and exosomal marker tests, please refer to the previous studies [16][17][18].  Table 1. Formal informed consent was signed by all patients before collecting blood samples.

Tube Formation Assay.
Te formation of capillary-like structures was assessed in a 24-well plate using growth factor-reduced Matrigel (BD, CA, USA). HUVECs (4 × 10 4 cells) were plated on top of Matrigel after transfection. After 24 h, the cells were visualized under a contrast phase microscope. Te total tube area was quantifed as the mean pixel density obtained from image analysis of three random microscopic felds using ImageJ software.

Animal Experiment.
Animal experiments in this study were approved by the Institutional Animal Care and Use Committee of Southwest Hospital, Chongqing, China. As previously described, 4-to 6-week-old male nude mice were used and anesthetized with 1% sodium pentobarbital. Ten, a median abdominal incision was made to expose the spleen and pancreas, and 5 × 10 6 PDAC cells (in 100 μl of PBS) were injected to the head of the pancreas. After replacing the pancreas and closing the abdomen with a continuous 4-0 braided silk suture, the mice were imaged by the IVIS Lumina II System (Caliper Life Science, USA) each week until the experiments were completed.
2.11. Statistical Analysis. All analyses were performed using SPSS 26.0 software (IBM, USA), and the graphs and diagrams were generated by GraphPad Prism 9 (LLC, USA). Experiments were performed in triplicates. χ 2 test was used to compare the correlation between clinical information and miR-210 expression. Student's t test or one-way analysis of variance (ANOVA) was used to analyze statistics between two groups or multiple groups, and a P value <0.05 was considered to be statistically signifcant.

Tumor-Derived Exosomes Promote Tumor Angiogenesis in
Recipient PDAC Cells. To clarify whether tumor exosomes have a stimulatory infuence on endothelial cells, three PDAC cell lines were used, and the invasive or metastatic ability of the three PDAC cell lines from high to low was Hs766T-L3, Hs766T-L2, and Hs766T (i.e., Hs 766T-L0), respectively [16]. HUVECs with PBS or exosomes, which were marked with Dil dye, were cocultured from Hs766T or Hs766T-L2 cells and found that exosomes from Hs766T-L2 or Hs766T cells signifcantly increased the angiogenesis of HUVECs, as many more tube-like structures were signifcantly investigated compared to the PBS group (Figures 1(a)-1(d)).

Identifcation of miR-210 in PDAC Exosomes.
Many miRNAs in exosomes were found to play variable important roles in tumor biology [15,19]. First, exosomal miRNA expression in the PDAC cell lines Hs766T and Hs766T-L2 to screen for possible useful tumor-related miRNAs were profled. Te top 30 diferentially expressed miRNAs are shown in Figure 2(a). Of these diferentially expressed miRNAs, it was noted that miR-210-3p (miR-210) was elevated in tumor exosomes, with a fold-change of 37.9 (Figure 2(a)). Ten, the high expression levels of miR-210 were further confrmed by qRT-PCR in exosomes from Hs766T-L2 cells (Figure 2(b)). Te cellular and exosomal miR-210 expression in Hs766T (the parent cell), Hs766T-L1, Hs766T-L2, and Hs766T-L3 (the daughter cell) cells was further detected. As the malignancy of PDAC cells increased, the expression level of miR-210 was also increased successively in both PDAC cells and their exosomes (Figure 2(c)). As previous studies reported that miR-210 was associated with hypoxia [20,21], then HUVECs in normal or hypoxic conditions (oxygen content: 0.1%) were cultured. Te results showed that after 24 h or 48 h, cellular or exosomal miR-210 was signifcantly increased under hypoxic conditions (Figures 2(d) and 2(e)). HIF-1α expression was also increased under hypoxic conditions ( Figure 2(f )). Tus, it was speculated that miR-210 may have an important function in PDAC. Ten, its upregulation (miR-210 lentivirus) or downregulation (miR-210 shRNA) in HUVECs was further confrmed (Figure 2(g)). Together, we screened miR-210 in tumor exosomes and further confrmed that miR-210 was highly expressed in malignant PDAC cells.

Promotion of Tumor Angiogenesis, Cell Invasion, and
Proliferation in HUVECs by miR-210. As miR-210 was highly expressed in malignant cells, we wondered whether miR-210 could promote angiogenesis, cell proliferation, or invasion. As expected, tube formation assays showed that miR-210 increased angiogenesis, while miR-210 shRNA decreased the angiogenesis of HUVECs (Figures 3(a) and 3(b)). Transwell invasion assays also revealed that miR-210 promoted cell invasion in HUVECs (Figures 3(c) and 3(d)). Furthermore, EdU assays showed that miR-210 overexpression increased cell proliferation, but the miR-210 inhibitor decreased cell proliferation in HUVECs (Figures 3(e) and 3(f )). Together, the results mentioned above suggest that miR-210 could promote angiogenesis, invasion, and proliferation.

Regulation of EFNA3 Expression by miR-210.
To further determine the detailed molecular mechanism of miR-210, bioinformatics tools were used to analyze its possible molecular target gene. It was noted that EFNA3, a classic tumor suppressor, has 7mer-m8 molecular binding sites for miR-210 (Figure 4(a)), and previous studies have also reported that EFNA3 can be regulated by miR-210 in the nervous system [22]. Terefore, the regulation of EFNA3 by miR-210 was further verifed. In the dual-luciferase reporter assays, wild-type or mutated EFNA3 3′UTR frefy luciferase plasmids, in which the miR-210 binding sites were deleted, were constructed. Te results showed that cotransfection of wild-type pGL3 EFNA3 and miR-210 inhibited Rluc expression, and this inhibition was even more obvious in the 200 nM miR-210 group than in the 70 nM miR-210 group (Figure 4(b)). Te qRT-PCR results showed that EFNA3 was signifcantly inhibited by miR-210 and was overexpressed when treated with miR-210 shRNA (Figure 4(c)). Furthermore, WB assays also suggested that EFNA3 protein was signifcantly reverse regulated by miR-210 transfection, which showed similar results (Figure 4(d)).

Imaging Exosomal miR-210 Communication between PDAC and Endothelial Cells. Ten, exosome communication between PDAC cells and endothelial cells was further explored. First, vascular endothelial cells were labeled with
Calcein-AMin vivo, and then Dil-labeled PDAC exosomes were added to the culture medium of PDAC cells for coculturing. Finally, dotted, red PDAC exosomal signals in the cytoplasm of HUVECs were found, as shown in Figure 6(b). To further confrm the miR-210 communication between PDAC and HUVECs via exosomes in vivo, HUVECs with the red fuorescence signal (-cy5)-tagged miR-210 were frst transfected. After two days of culture, the exosomes from the conditioned culture medium were extracted and labeled with a green fuorescence signal (-dio) for further cocultivation with HUVECs for 24 h. Interestingly, it was found that the red fuorescence signal (from miR-210) and green dotted-fuorescence signal (from exosome) were obviously mixed in the cytoplasm of HUVECs (Figure 6(c)). Together, the data mentioned above confrm the transmission of miR-210 between PDAC cells and HUVECs via exosomes.

Promotion of Tumor Progression via EFNA3 PI3K/AKT/ VEGFA or EFNA3/Wnt/Β-Catenin/RHOA by Exosomal miR-210.
Ten, whether exosomal miR-210 was biofunctional in PDAC was investigated. As shown in Figures 7(a) and 7(b), tube formation assays showed that exosomes with miR-210 overexpression or exosomes with miR-210 overexpression under hypoxic conditions signifcantly increased angiogenesis. Furthermore, we also found that the expression levels of p-AKT, VEFGA or Wnt3a, Catenin, and RhoA were elevated in exosomes with miR-210 overexpression or exosomes with miR-210 overexpression under hypoxic conditions, whereas EFNA3 expression was decreased (Figure 7(c)). Ten, a cell permeability assay was performed. As shown in Figure 7(d), when exosomes overexpressing miR-210 or exosomes overexpressing miR-210 were added under hypoxic conditions, the number of invaded Hs766T cells was much greater than that in the NC group.

Promotion of Tumor Progression In Vivo by miR-210.
Te results mentioned above suggest that exosomal miR-210 plays important roles in tumor progression in vitro. To further detect the roles of exosomal miR-210in vivo, an in situ PDAC model in nude mice was established. First, Hs766T cells with H-ExomiR-210 were cultured, and these cells were suspended in H-ExomiR-210 before the experiment. Ten, 10 6 cells were injected into the head of the pancreas of four-to six-weeks-old male nude mice. Te luciferase intensity was assessed each week after injection to confrm tumorigenesis in situ (Figure 8(a)). Te results showed that H-ExomiR-210 increased tumorigenesis, as intense luciferase signals compared to the NC mice, were detected.  RNA from plasma was extracted for further miR-210 detection. It was found that exosomal miR-210 was signifcantly associated with lymphatic invasion, vascular invasion, and TNM stage (Table 1). Subsequent survival analysis revealed that exosomal miR-210 and TNM stage were signifcantly related to poor survival rates in PDAC patients, and further multivariate analyses revealed that exosomal miR-210 expression was an independent risk factor for PDAC survival (Table 2).

Discussion
In this study, we screened miR-210 via miRNA microarray analysis of PDAC exosomes. Further research revealed that miR-210 regulates EFNA3 expression and that miR-210 or EFNA3 participate in the PI3K/AKT/VEGFA or Wnt/ β-catenin/RHOA pathways. Also, miR-210 communication between PDAC cells and endothelial cells based on exosome transmission was proved. Finally, we also detected plasma exosomal miR-210 expression in clinical PDAC patients and found that high exosomal miR-210 expression is correlated with tumor progression and is an independent risk factor for PDAC survival.
Tere is no doubt that tumor angiogenesis is essential for tumor initiation, progression, and metastasis [23]. On the one hand, similar to nontumor tissues, tumor cells require angiogenesis to provide nutrients, oxygen, and metabolites; on the other hand, hematogenous tumor metastasis also depends on angiogenesis [24]. Tumors cannot grow without sufcient development of new blood vessels. Tus, targeting angiogenesis, especially endothelial cells, is one of the most important and common therapeutic targets in nearly all   [25]. Nevertheless, the therapy seems not efective enough in PDAC, as clinical trials of gemcitabine plus antiangiogenic therapy with bevacizumab or axitinib failed to reach their primary endpoint of OS, and the possible reasons have been explored to date [26]. Studying angiogenesis has been an interesting topic in PDAC. On the one hand, PDAC is a hypovascular, hyperstroma, and poorly perfused tumor, and its major pathological feature is high levels of fbrosis, called desmoplasia [27]. Desmoplasia in PDAC leads to vasculature collapse that hinders vessel formation and inhibits drug penetration and uptake. Tus, some studies suggest that angiogenesis is not active enough in PDAC [28]; on the other hand, PDAC is characterized by rapid growth, recurrence, or metastasis, which defnitely rely on an abundance of nutrients, especially the glucose supply itself [28,29]. Tis paradox has resulted in relatively few studies on PDAC angiogenesis. Recently, a study revealed long, "hairy"-like projections on the basal surface of microvessels in PDAC, which means vascularization in PDAC is characterized by a high microvascular density, impaired microvessel integrity, and poorly perfused vessels [30]. Consistently, our results showed that PDAC could promote both angiogenesis and vascular permeability via exosomes and miR-210. Tis suggests that PDAC may be characterized as a hypovascular tumor on a macroscopic view or also a hypervascular (aberrant vessels) tumor on a microscopic view. Te massive deviant, high-permeability microvessels may be part of the reasons why PDAC progresses or metastasizes rapidly. Tumor-derived exosomes (TEXs) have attracted increasing interest in diferent tumors, as exosomes have rich endogenous contents that refect the pathophysiology status and genic information of tumor cells [31]. Large amounts of DNA and diferent kinds of RNAs or proteins were found to be related to tumor initiation, progression, or prognosis; thus, TEXs are increasingly being recognized as potential tumor biomarkers. Recently, abundant miRNAs of TEXs have emerged in various tumors. A growing number of miRNAs have been shown to regulate nearly all physiological processes of tumor biology, including initiation, invasion, and metastasis [32,33]. In PDAC, many oncosuppressor miRNAs have been identifed in recent studies; for example, low miR-200 family expression was found to promote EMT, cell proliferation, and apoptosis [34]. Other similar tumor inhibitor miRNAs include miR-192, miR-26, and miR-34a [35]. Oncogenic miRNAs in TEXs were mainly focused, as the RNA levels in TEXs are relatively scarce, which is difcult for both to detect. Unlike detecting miRNAs in frozen tissue samples translationally, detecting miRNAs in TEXs has obvious advantages compared to detecting miRNAs in frozen tissue. First, exosomes can protect miRNAs from RNA degradation, and studies have shown that miRNAs in plasma exosomes are quite stable under diferent storage conditions. When exosomes were stored at −20°C, the overall amount of plasma miRNA was barely impacted for at least 5 years. Second, miRNAs in TEXs are easy to obtain in clinical practice (via blood drawn) compared with tumor or metastatic samples. Tird, miRNAs in TEX exosomes can be continuously detected both before and after surgery. Tus, studying miRNAs in TEXs may be a convenient but efective approach in clinical practice. In this study, miR-210 in PDAC TEXs was detected. It was also   found that the high level of exosomal miR-210 in PDAC was correlated with the progression and prognosis of PDAC, which suggests that the detection of exosomal miRNA in blood can be an expedient and valid approach to monitor the biological behavior of tumors.

Conclusions
In this study, we identifed oncogenic miR-210 in PDAC exosomes by miRNA microarray analysis. Ten, it was confrmed that miR-210 could regulate EFNA3 expression; further mechanistic analysis revealed that miR-210 could promote tumor progression via the PI3K/AKT/VEGFA or Wnt/Β-catenin/RHOA pathways. miR-210 can be transported from PDAC cells to endothelial cells via exosome communication in living conditions. Finally, high exosomal miR-210 expression was found to be associated with PDAC invasion and prognosis. Terefore, high exosomal miR-210 may be an important clinical marker and may play important regulatory and diagnostic roles in PDAC.

Data Availability
Te data used in this study are available on request from the corresponding author.

Conflicts of Interest
Te authors declare that they have no conficts of interest.