Hypoxic exosomes facilitate bladder tumor growth and development through transferring long non-coding RNA-UCA1

Bladder cancer cell-derived exosomes were isolated from the conditioned media of 5637 cells cultured under normoxic or hypoxic conditions. The morphology of exosomes was analyzed by transmission electron microscopy (TEM). As shown in Fig. 1a, both normoxic and hypoxic exosomes showed typical lipid bilayer membrane-encapsulated nanoparticles and the size of these nanoparticles was 50–200 nm. Nanoparticle tracking analysis (NTA) exhibited that the average size of normoxic and hypoxic exosomes was 209 nm and 184 nm, respectively (Fig. 1b). There was a difference in the size distribution of exosomes between TEM and NTA analyses, and the main reason for such difference might be the aggregation behavior of exosomes in nanoparticle tracking analysis. Furthermore, we also characterized four exosomal protein markers of exosomes by western blotting analysis. As shown in Fig. 1c, the proteins of CD63, TSG101, Hsp70 and Hsp90 were positive expressed in normoxic or hypoxic exosomes. These results indicate that bladder cancer 5637 cells can secrete normoxic or hypoxic exosomes with common exosomal features.

To explore the biological roles of 5637 cell-derived normoxic and hypoxic exosomes, we next investigated the effects of normoxic and hypoxic exosomes on UMUC2 cell proliferation, migration and invasion. Compare with normoxic exosomes or PBS control, UMUC2 cells co-cultured with hypoxic exosomes showed higher cell viability (Fig. 2a). As shown in Fig. 2b, UMUC2 cells co-cultured with hypoxic exosomes exhibited higher motility than co-cultured with PBS control or normoxic exosomes. Similar to the cell migration results, enhanced the invasive potential of UMUC2 cells also occurred with 5637 cell-derived hypoxic exosomes (Fig. 2c). These results suggest that hypoxic bladder cancer cell-derived exosomes promote bladder cancer cell proliferation, migration and invasion in vitro.

To investigate normoxic or hypoxic exosomal RNAs internalized by bladder cancer cell lines, 5637 cell-derived normoxic and hypoxic exosomal RNAs were labeled with Exo-Red dyes and incubated with three bladder cancer cell lines. As shown in Fig. 3, the labelled exosomal RNAs could be internalized by 5637 cells. Similar results were observed when adding the labelled exosomal RNAs to UMUC2 or T24 cells (Fig. 3).

To further identify whether lncRNA-UCA1 is secreted in 5637 cell-derived normoxic and hypoxic exosomes, we first explored the existence pattern of lncRNA-UCA1 in exosomes. We designed primers to amplify the full-length transcript of UCA1 (Fig. 4a). Reverse transcription-PCR (RT-PCR) results showed that the full-length transcript of UCA1 (~1.4 kb) could be amplified from the normoxic and hypoxic exosomes (Fig. 4b). We also designed three primers for quantitative real-time PCR (qRT-PCR) to detect the expression levels of lncRNA-UCA1 in exosomes (Fig. 4a). According to the RT-PCR result, the UCA1–2 primers were used to detect exosomal lncRNA-UCA1 expression in our current study (Fig. 4c). We then determined whether lncRNA-UCA1 was indeed present within exosomes, which are provided a double-layer membrane against degradation by RNase. As expected, the expression levels of lncRNA-UCA1 in normoxic or hypoxic exosomes treated with RNase was similar to that in untreated control. Furthermore, the expression levels of lncRNA-UCA1 significantly decreased in normoxic or hypoxic exosomes treated with both RNase and Triton X-100 (Fig. 4d and e). These results indicate that the full-length transcript of UCA1 acts as an exosomal lncRNA transferred by bladder cancer cell-derived normoxic or hypoxic exosomes.

To further evaluate whether exosomal lncRNA-UCA1 was internalized by bladder cancer cells, we have also evaluated the relative UCA1 expression levels in different bladder cancer cell lines by RT-PCR and qRT-PCR. 5637 cells expressed relatively higher levels of UCA1, whereas UMUC2 cells expressed relatively lower levels of UCA1. Moreover, the expression levels of UCA1 could rarely be detected in UMUC2 cells by RT-PCR and qRT-PCR (Additional file 1: Figure S1a and b). Thus, UMUC2 cells were treated with different concentrations of 5637 cell-derived normoxic or hypoxic exosomes. The expressing levels of lncRNA-UCA1 in UMUC2 cells were upregulated in a dose-dependent manner compared with control cells treated with PBS (Fig. 4f and g). To clarify whether the expression levels of lncRNA-UCA1 in exosomes were induced by hypoxia, we detected intracellular or exosomal lncRNA-UCA1 expression levels by qRT-PCR. We found that hypoxia could induce the upregulation of lncRNA-UCA1 not only in bladder cancer cells but also in bladder cancer cell-derived exosomes (Fig. 4h and i). Collectively, these data indicate UCA1 may act as not only an intracellular hypoxia-responsive lncRNA but also an exosomal hypoxia-responsive lncRNA in bladder cancer.

To investigate whether hypoxic exosome-mediated bladder cancer cell proliferation, migration and invasion are directly dependent on exosomal lncRNA-UCA1, we suppressed lncRNA-UCA1 expression in hypoxic 5637 cells by shRNA. The suppression efficiency of lncRNA-UCA1 shRNA in hypoxic 5637 cells was confirmed by qRT-PCR. In addition to affecting the expression levels of lncRNA-UCA1 in 5637 cells, lncRNA-UCA1 shRNA also inhibited the expression levels of lncRNA-UCA1 in 5637 cell-derived hypoxic exosomes (Fig. 5a). Furthermore, hypoxic control shRNA exosomes significantly promoted UMUC2 cell proliferation when compared to the PBS control group. Knockdown of lncRNA-UCA1 in hypoxic exosomes also led to a decrease in UMUC2 cell proliferation when compared to the hypoxic control shRNA exosomes group (Fig. 5b). Additionally, hypoxic control shRNA exosomes significantly increased UMUC2 cell mobility and invasion when compared to the PBS control group. Depletion of exosomal lncRNA-UCA1 in hypoxic exosomes decreased UMUC2 cell mobility and its depletion also downregulated the invasive potential of UMUC2 cells (Fig. 5c and d). Therefore, these results suggest that hypoxic exosomes regulate bladder cancer cell proliferation, migration and invasion in vitro, in part by exosomal lncRNA-UCA1.

To further investigate the role of exosomal lncRNA-UCA1 in tumor growth in vivo, we next established a xenograft model (Additional file 2: Figure S2). After five weeks, mice were sacrificed and the xenograft tumor formation in the right flank. Furthermore, enlarged ipsilateral axillary lymph nodes were observed and removed for pathological examination (Additional file 3: Figure S3a). As shown in Fig. 6a and c, hypoxic control shRNA exosomes substantially increased tumor size and weight. Furthermore, depletion of lncRNA-UCA1 in hypoxic exosomes could abrogate the tumor growth (Fig. 6d). However, no metastases in lymph nodes were detected by immunohistochemistry using hematoxylin and eosin stain (Additional file 3: Figure S3b). Moreover, the primary tumor tissues of nude mice injected with hypoxic control shRNA exosomes had an increased expression of lncRNA-UCA1, and knockdown of lncRNA-UCA1 in hypoxic exosomes significantly decreased the expression of lncRNA-UCA1 in primary tumor tissues (Fig. 6e). These results indicate that hypoxic exosomes promote tumor growth is dependent on exosomal lncRNA-UCA1 in vivo.

To elucidate the underlying mechanism of exosomal lncRNA-UCA1-mediated tumor growth and progression, 5637 cells were treated with hypoxic lncRNA-UCA1 shRNA exosomes or hypoxic control shRNA exosomes. As shown in Fig. 6f, the expressing levels of lncRNA-UCA1 in 5637 cells treated with hypoxic lncRNA-UCA1 shRNA exosomes were significantly downregulated. Western blotting analysis revealed the decreased expression levels of the proliferation marker Ki67 in hypoxic lncRNA-UCA1-depressing exosomes treated group (Fig. 6g). Furthermore, hypoxic lncRNA-UCA1-depressing exosomes also significantly increased the expression levels of E-cadherin, while markedly decreasing the expression levels of vimentin and MMP9 in vitro or in vivo (Fig. 6g and h). These findings indicate that hypoxic exosomal lncRNA-UCA1 promotes bladder tumor progression though EMT.

Exosomal RNAs can be transferred from tumor cells to neighboring cells or cells at distant organs through entering the circulation [4]. Furthermore, hypoxia has been recognized as one of the hallmarks for solid tumors including bladder cancer [29]. Therefore, it is possible that exosomal lncRNA-UCA1 secreted by hypoxic bladder cancer cells can be detected in the circulation. To explore the circulating exosomal lncRNA-UCA1, we isolated exosomes from the serum of bladder cancer patients and matched healthy donors. TEM analysis revealed that the average size of isolated extracellular vesicles from bladder cancer patients or healthy individuals were exactly consistent with exosomes (50–200 nm in diameter, Fig. 7a). Moreover, western blotting analysis showed that the serum-derived exosomes from bladder cancer patients or healthy individuals were positive expressed for exosomal protein markers (Fig. 7b). To further identify lncRNA-UCA1 in serum-derived exosomes from bladder cancer patients or healthy individuals, we amplified the fragments of lncRNA-UCA1 in serum-derived exosomes. Indeed, the fragments of lncRNA-UCA1 from serum-derived exosomes of bladder cancer patients or healthy donors were determined by RT-PCR (Fig. 7c). We also found that the expression levels of exosomal lncRNA-UCA1 in the serum of bladder cancer patients were markedly higher than that in healthy control subjects, and the expression levels of exosomal lncRNA-UCA1 were normalized to ACTB (β-actin) or GAPDH (Fig. 7d and Additional file 4: Figure S4a). In addition, we used the ROC curve to evaluate the diagnostic value of exosomal lncRNA-UCA1 in serum. The ROC analysis demonstrated an area under curve (AUC) of 0.8783 (95% CI = 0.7926–0.964, P < 0.0001) for GAPDH better than ACTB (β-actin) (AUC: 0.7528, 95% CI = 0.6320–0.8736, P < 0.001) (Fig. 7e and Additional file 4: Figure S4b). The sensitivity and specificity were 80% and 83.33%, respectively (cut off value 1.603, GAPDH as an internal control). To further analyze the correlation between exosomal lncRNA-UCA1 and hypoxia marker HIF-1α in bladder cancer patients, HIF-1α expression was evaluated by immunohistochemical staining (Fig. 7f). As shown in Fig. 7g, exosomal lncRNA-UCA1 RNA levels in the serum of bladder cancer patients were positively correlated with HIF-1α expression. Altogether, these results indicate that exosomal lncRNA-UCA1 in serum may serve as a potential diagnostic biomarker for bladder cancer.

Human bladder cancer 5637, UMUC2 and T24 cells were grown in RPMI-1640 medium (Gibco, Grand Island, NY, USA) with 10% bovine calf serum and maintained at 37 °C under a humidified 5% CO₂ atmosphere. For hypoxic experiments, hypoxia was achieved using an oxygen control incubator (Heal Force, Shanghai, China), which was flushed with a mixture of 1% O₂, 94% N₂, and 5% CO₂. The cells were cultured for 24–72 h under hypoxic conditions.

Human serum samples were collected from the First Affiliated Hospital of Xi’an Jiaotong University (Xi’an, Shaanxi, P.R. China) and obtained informed consent from 30 patients diagnosed bladder cancer and 30 healthy volunteers without any malignancy. The study was approved by the Ethics Committee of the First Affiliated Hospital of Xi’an Jiaotong University. The clinical pathological characteristics of bladder cancer patients are listed in Additional file 5: Table S1.

Bladder cancer 5637 cells were cultured for 48–72 h in advanced RPMI-1640 medium (Gibco) without supplements under normoxic or hypoxic conditions. Exosomes were isolated from the supernatant of 5637 cells by differential centrifugations as previously described [45]. The media were harvested and centrifuged at 300×g for 10 min at 4 °C. The supernatant was further centrifuged at 16,500×g for 20 min at 4 °C and filtered through a 0.22 μm filter. Exosomes were then pelleted by ultracentrifugation at 120,000×g for 70 min at 4 °C. Exosome pellets were resuspended in 0.2 μm-filtered PBS.

Exosomes were isolated from the human serum samples by ExoQuick solution according to the manufacturer’s instructions (System Biosciences, SBI, Mountain View, CA). Briefly, serum samples were centrifuged at 3000×g for 15 min at 4 °C, and the serum samples (250 μl) were mixed with ExoQuick solution (63 μl). The mixtures were then refrigerated at 4 °C overnight and centrifuged at 1500×g for 30 min at 4 °C. Exosome pellets were resuspended in 0.2 μm-filtered PBS.

Exosomes were adsorbed to a 400 mesh carbon-coated copper grids and stained with phosphotungstic acid solution. Morphologies of the samples were observed by a JEOL JEM-100SX transmission electron microscope (JEOL Ltd., Tokyo, Japan).

The size distribution of exosomes was determined using a Delsa Nano Analyzer (DelsaNano, Beckman Coulter, Brea, CA, USA). The capture settings and analysis settings were performed manually according to the manufacturer’s instructions.

Exosomes were incubated with RNase A (Takara, Dalian, China, final concentration of 10 μg/ml) at 37 °C for 10 min followed by addition of RNase inhibitor (Takara, final concentration of 1 U/μl). Cellular and exosomal RNAs were isolated using the TRIzol reagent (Invitrogen, Life Technologies, Carlsbad, CA, USA). First-strand cDNA was synthesized with random primers using a PrimeScript™ RT reagent Kit with gDNA Eraser (Takara). Quantitative real-time PCR was carried out using a FastStart Essential DNA Green Master (Roche, Nutley, NJ, USA) on a CFX96 real-time PCR System (Bio-Rad, Hercules, CA, USA), and the results were normalized with ATCB (β-actin) or GAPDH as an internal control. Primers are listed in Additional file 6: Table S2.

Exosomes or cells were lysed in RIPA buffer containing a complete protease inhibitor tablet (Roche). The protein concentration of lysates was normalized according to Bradford relative protein quantification and proteins were separated by SDS-PAGE and transferred onto a nitrocellulose membrane. The membranes were incubated with CD63, E-cadherin, Ki67, MMP9, TSG101 (Abcam, Hong Kong, China), β-actin, Hsp70, Hsp90, Vimentin (Cell Signaling Technology, CST, Danvers, MA, USA) primary antibodies. Protein expression was assessed by ECL chemiluminescent regents and the intensity of the protein bands was quantified by densitometry using ImageJ software (National Institutes of Health).

Exosomal RNAs were labeled with Exo-GLOW™ kits (SBI) according to the manufacturer’s instructions. Exosomes were stained at 37 °C for 10 min in 1 × PBS containing 10 × Exo-Red dye and bladder cancer cells were incubated with the labeled exosomes at 37 °C for 6–24 h. Images were collected with a Nikon Eclipse Ti-S fluorescence microscope (Nikon, Tokyo, Japan).

For depletion of lncRNA-UCA1, 5637 cells were stable transfected with pRNAT-U6.1⁄UCA1 shRNA or pRNAT-U6.1⁄control shRNA plasmid. The following day, 5637 cells were cultured in advanced 1640 medium under hypoxic conditions for 24–48 h. The exosomes derived from 5637 cells were purified as described previously.

Cell proliferation assay was carried out using the Cell Count Kit (7Sea Pharmatech Co., Ltd., Shanghai, China). Briefly, UMUC2 cells (4000 cells/well) were seeded in 96-well plates and incubated with exosomes (5 μg/ml) for a total of 72 h. Absorbance was measured at 450 nm using a PerkinElmer Enspire plate reader (PerkinElmer, Waltham, MA, USA).

The invasion assay was carried out using 24-well Millicell chambers that were coated with Matrigel (BD Biosciences, San Jose, CA, USA). The migration assay was carried out in a similar fashion without the Matrigel coating. Cells were seeded in serum-free RPMI-1640 to the top chamber. Exosomes (10 μg/ml each well) were added to the bottom chambers and serum-free advanced 1640 medium with PBS as the control. After 24 h of incubation, UMUC2 cells on the upper membrane surface were wiped off using a cotton swab and the lower membrane surface was fixed with methanol, stained with crystal violet, and counted in five random fields.

Five-week-old female athymic nude mice were purchased from the Laboratory Animal Center of Xi’an Jiaotong University (Xi’an, Shaanxi, P.R. China). The 5637 cells (5 × 10⁶ cells per mouse) suspended in 200 μl of PBS were injected subcutaneously into the right flank of nude mice, and two weeks later, when the nude mice generate tumors with a size of 100 mm³, purified exosomes (10 μg) or PBS were then injected into the center of tumor sites. The tumor size and weight of nude mice were weighed and measured twice a week. After three weeks, the nude mice were sacrificed and their tumors tissues and lymph nodes were determined for histological examination.

Primary tumors were fixed in formalin and embedded in paraffin, and then cut at 4 μm thickness. The sections were incubated with primary antibodies (E-cadherin, HIF-1α, MMP9, Abcam) at 4 °C overnight. After washing with PBS, the sections were then incubated with Poly HRP-conjugated anti-mouse or anti-rabbit IgG for 30 min, followed by DAB. Finally, the sections were counterstained with hematoxylin. The staining intensity of HIF-1α was assessed by two independent pathologists as no staining = 0, weak staining = 1, moderate staining = 2, and strong staining = 3. Tumor cells in five random fields were scored based on the percentage of nuclei with positive staining cells (0–100%). An overall immunohistochemistry score was calculated by multiplying the intensity score with the percentage of positive staining tumor cells.

All statistical analyses were carried out using GraphPad Prism Software (GraphPad Software, Inc., San Diego, CA, USA). Statistical evaluations were determined using Student’s t-test (two-tailed) or Spearman correlation test. P value <0.05 was considered statistically significant. In vitro experiments were replicated at least three times.