The human colon cancer cell line HCT116 and Caco-2 exhibiting wild-type TP53 expression, SW480 and DLD-1 of TP53 mutant cells, and non-transformed human colon fibroblasts CCD-18Co were obtained from the American Type Culture Collection (ATCC). Cancer cells were cultured at 37°C under 5% CO₂ in Dulbecco's Modified Eagle Medium (DMEM) (Sigma-Aldrich; Merck KGaA) supplemented with 10% fetal bovine serum (FBS) or Eagle's Minimum Essential Medium (EMEM) (ATCC) supplemented with 20% FBS, and fibroblasts were cultured at 37°C under 5% CO₂ in EMEM with 10% FBS. To investigate the effect of apelin-13 (Cayman Chemical Company) on recombinant human TGF-β1 (rhTGF-β1)(FUJIFILM Wako Pure Chemical Corporation)-induced mRNA and protein expression in fibroblasts, the cells were pretreated with apelin-13 at the concentration of 10 to 1,000 nM for 60 min and then stimulated with rhTGF-β1 for 24-48 h. ML221 (Sigma-Aldrich; Merck KGaA), the APJ antagonist, was used to examine the effect of APJ suppression on fibroblasts at concentrations of 0.1 to 10 µM.

Cells were cultured in DMEM with 10% FBS in 15-cm dishes (Tissue Culture Dish VTC-D150, AS ONE CORPORATION) for 48 h, washed with phosphate-buffered saline (PBS), replaced with DMEM without FBS, and cultured for another 48 h, after which the culture supernatant was collected, and sequential centrifugation was performed. The culture supernatant was first centrifuged at 300 × g for 10 min at 4°C, then at 2,000 × g for 10 min to precipitate the cells. The supernatant was then centrifuged at 10,000 × g for 30 min at 4°C, and then ultra-centrifuged at 100,000 × g for 70 min at 4°C to pellet the extracellular vesicles, which were then washed with a suspension in PBS. The final pellet was resuspended in 100 µl of PBS. The protein concentration was measured using a bicinchoninic acid (BCA) protein assay kit (Thermo Fisher Scientific, Inc.). Isolated exosomes were administered to fibroblasts culture supernatant at a concentration of 100 µg/ml. The morphology of the exosomes was captured using a transmission electron microscope (H-7600; Hitachi High-Technologies Corporation) after preparation as described here. A total of ~5 µl of sample was placed on Parafilm. Then, a carbon-coated 400 mesh copper grid was positioned on the top of the drop for 10 sec and washed by a droplet of distilled water. The grid was contrasted by adding a drop of 2% uranyl acetate on Parafilm and incubating the grid on the top of the drop for 10 sec. Excess liquid was removed by gently using an absorbent paper. After dying, the samples were used for observation. Furthermore, the expression of CD9, CD63 and ALIX, which are exosome markers, was confirmed by western blotting.

Non-contact co-cultures were performed using Transwell inserts with 0.4-µm pores (Corning, Inc.). Fibroblasts or cancer cells were seeded on a six-well (2×10⁵ cells per well) or 12-well (3×10⁴ cells per well) plate (Corning, Inc.) in EMEM or DMEM with 10% FBS, and cancer cells or fibroblasts were seeded on a Transwell insert in the same conditions. After 24-48 h incubation, serum-free EMEM was used as the co-culture medium, and each culture was incubated for 48 h.

CCD-18Co cells were transfected for 48-72 h at 37°C with small interfering (si)RNAs against APJ [Thermo Fisher Scientific, Inc.; #1; cat. no. s223458, 5′-CAG AUG CAC GAG AAA UCC ATT-3′ (sense) and 5′-UGG AUU UCU CGU GCA UCU GTT-3′ (antisense), #2; s1186, 5′-UGU GGG CUA CCU ACA CGU Att-3′ (sense) and 5′-UAC GUG UAG GUA GCC CAC Agg-3′ (antisense)], negative control siRNA (Thermo Fisher Scientific, Inc.; cat. no. 4390843), microRNA (miR)-5703 mimic (Thermo Fisher Scientific, Inc.; cat. no. 4464066), miR-5703 inhibitor (Thermo Fisher Scientific, Inc. cat. no. 4464084), miRNA mimic negative control (Thermo Fisher Scientific, Inc. cat. no. 4464058) and miRNA inhibitor negative control (Thermo Fisher Scientific, Inc. cat. no. 4464076) and SW480 cells were transfected with RAB27A siRNA [Thermo Fisher Scientific, Inc.; cat. no. s532296; 5′-CCA GUG UAC UUU ACC AAU Att-3′ (sense) and 5′-UAU UGG UAA AGU ACA CUG Gtc-3′ (antisense)] at a final concentration of 10 nM using the Lipofectamine^® RNAiMAX Reagent (Thermo Fisher Scientific, Inc.) according to the manufacturer's transfection protocol in six-well or 12-well plates (Corning, Inc.). In addition, as previously described (11), lentiviral GFP-IRES-short hairpin (sh)RNA vectors against TP53 (cat. nos. RHS4430-101161166; 101162286; 101168779; 99365289) were obtained from Thermo Fisher Scientific, Inc., and HCT116^{sh control} and HCT116^{sh p53} cells were generated. Cells with stable shRNA expression were cultured with 2 µg/ml puromycin (InvivoGen) after colony selection.

Total RNA or miRNAs were extracted and reverse transcribed as previously described (12). RNA was extracted from cells or nanovesicles using an RNeasy Mini Kit (cat. no. 74106; Qiagen GmbH) for mRNA and a miRNeasy Mini Kit (cat. no. 217084; Qiagen GmbH) for miRNAs according to the manufacturer's instructions. qPCR was performed using THUNDERBIRD qPCR Master Mix (Toyobo Life Science) on the QuantStudio 6 Flex (Applied Biosystems; Thermo Fisher Scientific, Inc.). The thermocycling conditions of the qPCR reaction were: 20 sec of initial denaturation at 95°C; followed by 40-60 cycles of 1 sec at 95°C for denaturation, and 20 sec at 60°C for annealing and extension. The mRNA expression was quantified using TaqMan gene expression assays (Applied Biosystems; Thermo Fisher Scientific, Inc.), and data were normalized to beta-2-microglobulin (B2M) expression levels. The miRNAs were reverse transcribed using miRNA-specific primers (Applied Biosystems; Thermo Fisher Scientific, Inc.), and miRNA expression was quantified using TaqMan gene expression assays. Data were normalized to the expression of spike-in syn-cel-miR-39 (Qiagen). The 2^−ΔΔCq method was used to analyze the relative gene expression as previously described (23). The used primers are listed in Table SI.

Protein extracts were prepared as previously described (11). Cells were lysed with radio-immunoprecipitation assay (RIPA) buffer containing protease inhibitor and phosphatase inhibitor cocktail (Nacalai Tesque, Inc.), and the proteins were collected on ice. The protein concentration was measured by the BCA protein assay kit. Equal amounts of protein (10 µg) were applied to each lane of 8% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) gels and run. The proteins were then transferred onto polyvinylidene difluoride (PVDF) membranes and blocked with 4% milk in 0.1% Tris-buffered saline with Tween-20 (TBST) at room temperature for 1 h. Then, each membrane was subjected to overnight incubation at 4°C with the appropriate primary antibodies for β-actin (1:1,000; cat. no. 4970; Cell Signaling Technology, Inc.), TP53 (1:200; cat. no. sc-126; Santa Cruz Biotechnology, Inc.), APJ (1:1,000; cat. no. ab84296; Abcam or 1 µg/ml, cat. no. 702069; Thermo Fisher Scientific, Inc.), α-SMA (1:1,000; cat. no. ab5694; Abcam), phosphorylated Smad2/3 (1:1,000; cat. no. 8828; Cell Signaling Technology, Inc.) and Smad2/3 (1:1,000; cat. no. 3102, Cell Signaling Technology, Inc.). After incubation, the membranes were incubated with horseradish peroxidase (HRP)-conjugated secondary antibody (1:3,000; cat. no. NA934-1ML; Cytiva) at room temperature for 60 min. Finally, the membranes were reacted with detection reagents (Super Signal West Pico PLUS Chemiluminescent Substrate; Thermo Fisher Scientific, Inc.) and exposed for an appropriate time to visualize the protein bands using the ChemiDoc MP (Bio-Rad Laboratories, Inc.). Primary antibodies used for western blotting are listed in Table SII.

Cell proliferation and viability were analyzed in 12-well plates (3×10⁴ cells per well). Water-soluble tetrazolium (WST) assays were performed using an SF cell counting reagent (Nacalai Tesque, Inc.). A total of 100 µl reagent was added to 1,000 µl culture medium per well. After 30 min of incubation at 37°C, the absorbance was measured at 450 nm with 600 nm used as a reference with a microplate reader (Thermo Fisher Scientific, Inc.).

Wound healing assay was performed to assess the migratory ability of fibroblasts. Cells seeded on a six-well plate (2×10⁵ cells per well) were incubated for 24 h; thereafter, the medium was changed to serum-free, ML221 was applied, and cells were scratched using the tip of the 200-µl sterile pipette tube. In the same manner, cells were seeded on a six-well plate (2×10⁵ cells per well), were transfected with siRNA against APJ, and incubated for 24 h. Next, the medium was changed and incubated for another 24 h, after which cells were scratched the same as aforementioned. The wound reduction rates were calculated by measuring the area of the wound after 0 and 24 h in 3 microscopic fields (magnification, ×200; BZ-X700 all-in-one fluorescence microscope; Keyence Corporation), which were randomly selected.

Fibroblasts were seeded on a 35-mm dish (Matsunami Glass Ind., Ltd.) (2×10⁵ cells per dish) and were transfected with siRNA against APJ 72 h before the experiment or were treated with ML221 (10 µM) 24 h prior to the experiment. Cells were fixed with 100% methanol at room temperature for 15 min. After cells were blocked with 1% bovine serum albumin (Nacalai Tesque, Inc.), the primary antibody reaction against anti-α-SMA (1:400; cat. no. ab5694; Abcam) was performed at room temperature for 1 h. Following a secondary antibody reaction using Alexa Fluor Plus 488 (1:1,000; cat. no. 4412; Cell Signaling Technology, Inc.) at room temperature for 1 h, nuclear staining was performed using Hoechst (1:200; cat. no. 346-07951; Dojindo Laboratories, Inc.) at room temperature for 10 min. The images were analyzed using a fluorescence microscope (BZ-X700 all-in-one fluorescence microscope; Keyence Corporation). The antibody used for fluorescent immunostaining is listed in Table SII.

Immunohistochemical staining for p53 and APJ in colorectal cancers with submucosal invasion was performed using endoscopically resected specimens at the Osaka University Hospital between April 2015 and August 2020. The paraffin-embedded tumor tissues were cut into 4-µm sections. The sections were dewaxed with xylene, dehydrated with descending ethanol series, activated with antigen retrieval citrate buffer, and incubated with 3% H₂O₂ at room temperature for 10 min. The slides were blocked with serum-free phosphate buffer including casein at room temperature for 20 min and incubated with the primary antibodies at 4°C overnight. The primary antibodies were as follows: anti-P53 (1:100; cat. no. sc-126; Santa Cruz Biotechnology, Inc.) or anti-APJ (1:100; cat. no. ABD43; Sigma-Aldrich; Merck KGaA). The sections were incubated with HRP-conjugated secondary antibody (no dilution; cat. no. K5007; Agilent Technologies, Inc.) at room temperature for 20 min and counterstained with hematoxylin at room temperature for 15 sec. The slides were captured using a light microscope (VS200; Olympus Corporation). The expression of p53 in cancer cells, of APJ in fibroblasts from the tumor stroma, and of α-SMA in the stroma with HALO software were quantified (Indica Labs, Inc.). p53 expression was assessed using a previously described scoring method (24). The p53 expression levels were evaluated according to the positive rate of p53 staining in tumor cells, and p53 positivity was defined as more than 10% of p53 positive staining. APJ expression levels were assessed by calculating the H-score using the software. The area of α-SMA staining in the stroma was calculated by the aforementioned software. The antibodies used for immunohistochemistry in each case are listed in Table SII.

HCT116 cells (2×10⁵) were suspended in 200 µl of PBS and co-inoculated with negative controltransfected CCD-18Co cells (2×10⁵) and CCD-18Co cells with suppressed APJ expression (2×10⁵) subcutaneously into the left and right flank of 5-6-week-old male BALB/c nude mice (Charles River, Yokohama, Japan), respectively (each group n=10). A total of 20 mice were used. All mice were housed in a specific pathogen free condition at a constant temperature of 23±1.5°C with 45±15% humidity under a 12/12-h light/dark cycle and were fed with MFG (Oriental Yeast Co., Ltd.) with free access to water. In addition, gene suppression using siRNA in the xenograft experiments was performed as previously described. At the time of subcutaneous implantation, the mice were anesthetized with medetomidine hydrochloride (0.75 mg/kg), midazolam (4 mg/kg) and butorphanol tartrate (5 mg/kg). Tumor volume was measured three times a week and calculated using the formula: [tumor length x (tumor width)²]/2. Humane endpoints were set at a tumor volume of 2,000 mm³ and irreversible wasting. The duration of the experiment was 21 days. The last measurement of the tumor volume was obtained on the 21st day, followed by euthanasia by CO₂ administration with a fill rate of 30-70% of the chamber volume per min and death was verified by checking for the heartbeat and breath cessation. All mice used were euthanized.

The National Center for Biotechnology Information Gene Expression Omnibus (NCBI GEO) database and miR-DB database (http://mirdb.org) were used in the present study. To analyze APJ expression in colon fibroblasts, NCBI GEO database (accession no. GSE46824; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE46824) was used. NCBI GEO database (accession no. GSE120012; https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE120012) and miR-DB database were used to analyze miR-5703.

Data were expressed as the mean ± standard deviation. The two-tailed, paired or unpaired Student's t-test was used to compare two groups. A one-way analysis of variance (ANOVA) with Tukey's post hoc test was performed to analyze the differences among multiple groups. P<0.05 was considered to indicate a statistically significant difference. The statistical analyses were performed using the JMP^® Pro15 software (SAS Institute Inc.).

The present study was approved (approval no. 20061) by the Ethics Committee of Osaka University Graduate School of Medicine (Osaka, Japan) for using resected human samples, and written informed consent was obtained from all patients who provided resected specimens. The study design was per the principles of the Declaration of Helsinki (October 2013). The animal experimental procedures were performed per the Osaka University guidelines for animal experiments and were approved (approval no. 30-015-077) by the Animal Care and Use Committee of Osaka University Graduate School of Medicine (Osaka, Japan).

First, non-contact cell co-culture experiments were conducted using a Transwell insert for fibroblasts (CCD-18Co). To investigate the involvement of p53 inactivation in cancer cells with APJ expression in fibroblasts, p53-suppressed colorectal cancer cells, HCT116^{sh p53} cells, were established. Successful inhibition of TP53 expression in HCT116^{sh p53} cells was confirmed by RT-qPCR and western blotting (Figs. 1A and S1A). The expression level of APJ in CCD-18Co cells was decreased along with increased expression of α-SMA when co-cultured with HCT116^{sh p53} cells compared with co-cultured with HCT116^{sh control} cells or CCD-18Co cells alone in western blotting (Figs. 1B and S1B). The mRNA levels of TGFβ1 and VEGF-A expression in fibroblasts significantly increased when co-cultured with HCT116^{sh p53} cells (Fig. 1C). In addition, the expression levels of APJ were decreased and α-SMA were increased in fibroblasts when co-cultured with TP53-mutant colon cancer cells such as SW480 or DLD-1 cells, in contrast to TP53-wild colon cancer cells such as HCT116 or Caco-2 (Figs. 1D and S1C and D). The mRNA expression levels of TGFβ1 and VEGF-A were significantly increased when co-cultured with SW480 and DLD-1 cells as observed in HCT116^{sh p53} cells, which were also significantly increased than those co-cultured with HCT116 and Caco-2 cells (Fig. 1E).

Next, APJ expression was examined in fibroblasts from human colorectal cancer tissues. APJ expression levels in cancer stromal fibroblasts were evaluated based on the p53 status of colorectal cancer cells. Immunohistochemical staining was performed for p53, APJ and α-SMA in endoscopically resected cancers with submucosal invasion. The patient and lesion characteristics are summarized in Table SIII. The APJ staining score in cancer stromal fibroblasts was significantly lower in p53-positive (suggesting p53 mutation) colorectal cancers than in p53-negative (suggesting wild-type p53) colorectal cancers. The APJ staining score was assessed separately in the left colon (descending colon, sigmoid colon, rectum) and right colon (cecum, ascending colon, transverse colon), but there were no significant differences (Fig. S9). The relationship between clinicopathological characteristics and p53 status of endoscopic-resected cases is listed in Table SIV. Next, the included cases were divided into two groups according to their median APJ staining score (high or low). The area of α-SMA staining in each of these groups was then estimated, which showed a trend towards being larger in the low score group (Fig. 1F). mRNA expression datasets of fibroblasts extracted from fresh surgical specimens of colorectal carcinoma (CAF group) and normal colonic mucosa (normal colonic fibroblasts, NCF group) were obtained using NCBI GEO database (accession no. GSE46824). The datasets revealed that APJ expression was significantly lower in the CAF group than in the NCF group (Fig. 1G). These results indicated that TP53-inactivated colorectal cancer cells inhibit APJ expression in fibroblasts and induce fibroblast modification with myofibroblast-like properties.

Experiments using ML221 in CCD-18Co cells were performed to investigate the significance of APJ inhibition in colon fibroblast modification. The relative mRNA expression levels of TGFβ1, VEGF-A and ACTA2 were significantly increased in CCD-18Co cells with ML221 compared with control, and the western blotting analysis revealed increased α-SMA protein levels with ML221 (Figs. 2A and S2). α-SMA expression was next visualized using fluorescence immunostaining, and the expression of α-SMA was revealed to be significantly stronger in CCD-18Co cells with ML221 than in the control (Fig. 2B). A wound healing assay of CCD-18Co cells to examine the migration ability. The wound healing of CCD-18Co cells treated with ML221 was significantly faster than that of the control cells (Fig. 2C). Cell proliferation was also examined using a WST assay. Adding ML221 significantly increased the proliferation of CCD-18Co cells compared with the control in ML221 dose-dependent manner (Fig. 2D). Furthermore, the proliferation of colon cancer cells, which was increased by co-culture with CCD-18Co cells, was significantly increased by adding ML221 to CCD-18Co cells (Fig. 2E).

Next, APJ expression was suppressed in fibroblasts using APJ siRNA. The APJ mRNA and protein expression levels were successfully suppressed by siRNA with elevated α-SMA expression levels. The relative mRNA expression levels of TGFβ1, VEGF-A and ACTA2 were significantly increased with siRNA for APJ compared with the control (Figs. 3A and S3). α-SMA expression was also visualized using fluorescence immunostaining, and α-SMA was more strongly expressed in APJ-suppressed CCD-18Co cells than in the control (Fig. 3B). Wound healing assay revealed that the migration ability of APJ-suppressed CCD-18Co cells was significantly increased compared with that of the control (Fig. 3C). The WST assay revealed that the viability of APJ-suppressed CCD-18Co cells was significantly higher than that of the control (Fig. 3D).

Next, regarding fibroblast-mediated tumor growth, the importance of APJ expression of fibroblasts was investigated using TP53-wild cancer cells in xenograft experiments. Tumor volumes of co-implanted HCT116 and CCD-18Co cells were significantly higher than those of HCT116 cells alone, and APJ suppression in CCD-18Co cells promoted faster tumor growth than the control (Fig. 3E). These results suggest that suppression of APJ in fibroblasts induces modification of myofibroblast-like properties and accelerates tumor growth.

It has been reported that TGFβ1 induces α-SMA production via phosphorylated Sma- and Mad-related protein 2/3 (Smad2/3); the addition of apelin inhibits the induction of phosphorylated Smad2/3 by TGFβ1 (18,19). Next, the TGFβ-Smad pathway was evaluated in fibroblasts as a mechanism by which the apelin-APJ system can suppress the modification of fibroblasts into myofibroblasts-like properties. The mRNA expression levels of TGFβ1, VEGF-A and ACTA2 in CCD-18Co cells were increased by rhTGF-β1 stimulation. These increased expression levels were diminished significantly by the addition of recombinant human apelin-13 (Fig. 4A), suggesting the apelin-APJ system could suppress the TGFβ-induced fibroblasts modification. Western blot analysis revealed that phosphorylated Smad2/3 and α-SMA protein levels in CCD-18Co cells, which were increased by rhTGF-β1 stimulation, were decreased by adding recombinant human apelin-13 (Figs. 4B and S4). Increased expression of phosphorylated Smad2/3 was also found in CCD-18Co cells whose APJ expression was suppressed by ML221, siRNA or co-cultured cancer cells (Figs. 4C and S5A-D). These results suggested that the apelin-APJ system inhibits the TGFβ-Smad pathway in fibroblasts.

Cancer cell-derived exosomes and encapsulated miRNAs are crucial communication tools in the tumor microenvironment (12). Moreover, alterations in TP53 expression in donor cancer cells have been reported to modify their exosomal miRNA profiles and affect the expression of several genes in the surrounding recipient cells (25,26). Cancer cell-derived exosomes were used to explore the mechanisms underlying APJ suppression in colon fibroblasts. Exosomes were collected from HCT116^{sh control} and HCT116^{sh p53} cells culture supernatants by the ultracentrifugation method and the expression of CD9, CD63 and ALIX, which are exosome markers, was confirmed by western blotting. The size and morphology of isolated exosomes were determined by transmission electron microscopy (Figs. 5A and S6). Furthermore, no significant differences were observed in the protein levels of the pellets isolated from HCT116^{sh control} and HCT116^{sh p53} cell culture supernatants (Fig. 5B). CCD-18Co cells were stimulated with HCT116^{sh control}- or HCT116^{sh p53}-derived exosomes to confirm the effects of the exosomes. Western blotting demonstrated that APJ expression was decreased in CCD-18Co cells transfected with HCT116^{sh p53}-derived exosomes (Figs. 5C and S7A). In addition, the relative expression levels of TGFβ1, VEGF-A, and ACTA2 in mRNA were significantly increased with HCT116^{sh p53}-derived exosomes (Fig. 5D).

Next, exosomes were collected from the culture supernatant of SW480 cells by ultracentrifugation. When the exosomes were added to CCD-18Co cells, the protein expression of APJ in CCD-18Co cells was decreased with SW480-derived exosomes (Figs. 5E and S7B). The relative mRNA expression levels of TGFβ1 and ACTA2 were significantly increased in CCD-18Co cells with SW480-derived exosomes, as observed in HCT116^{sh p53}-derived exosomes (Fig. 5F). Exosome inhibition was evaluated using siRNA against RAB27A, a regulator of exosome secretion. APJ expression in CCD-18Co cells, suppressed by co-culture with SW480 cells, was restored by suppressing RAB27A in SW480 cells (Figs. 5G and S7C). The mRNA expression of TGFβ1, VEGF-A and ACTA2 in CCD-18Co cells, which were increased by being co-cultured with SW480 cells, were significantly decreased by suppressing RAB27A in SW480 cells (Fig. 5H). The WST assay revealed that the viability of CCD-18Co cells, which was increased by co-culturing with SW480 cells, was significantly decreased by the suppression of RAB27A in SW480 cells. Furthermore, the proliferation of SW480 cells, which was increased by co-culture with CCD-18Co cells, was reduced significantly by the suppression of RAB27A in SW480 cells (Fig. 5I).

As a mechanism of APJ suppression in fibroblasts, the role of miRNAs in cancer cell-derived exosomes was investigated. The list of miRNAs targeting the APJ gene was created using the miR-DB database and the miRNA profile in HCT116^{sh control} and HCT116^{sh p53}-derived exosomes was analyzed using the registered results of the miRNA microarray analysis (NCBI GEO accession no. GSE120012). A total of 55 miRNAs targeting the APJ gene were detected in HCT116^{sh control} and HCT116^{sh p53}-derived exosomes. Of the 55 miRNAs, 49 were found to be highly expressed in HCT116^{sh p53}-derived exosomes compared with those of HCT116^{sh control}. Among them, miR-5703 was identified as the miRNA with the highest expression in HCT116^{sh p53}-derived exosomes compared with those of HCT116^{sh control} (Fig. 6A).

Next, exosomes were collected from HCT116^{sh control} and HCT116^{sh p53} cell culture supernatants by ultracentrifugation and the expression ratio of miRNAs was validated using RT-qPCR analysis. The relative expression level of miR-5703 was significantly higher in HCT116^{sh p53}-derived exosomes than in HCT116^{sh control}-derived exosomes (Fig. 6B). The miRNA mimic was transfected into CCD-18Co cells to investigate the effect of the miRNA on APJ expression in fibroblasts; then, the APJ protein expression of CCD-18Co cells was suppressed and the mRNA expression levels of TGFβ1 and VEGF-A in CCD-18Co cells were significantly increased compared with the control (Figs. 6C and S8A). In addition, the WST assay was performed to investigate the viability of CCD-18Co cells treated with the miRNA mimic, which was significantly higher than that of the control (Fig. 6D). The effects of a specific miR-5703 inhibitor were also explored. Transfection of the miR-5703 inhibitor into CCD-18Co cells after co-culturing with SW480 cells restored the APJ expression in CCD-18Co cells, which was suppressed in the co-cultured SW480 cells (Figs. 6E and S8B). The WST assay revealed that the viability of CCD-18Co cells, which was increased by co-culturing with SW480 cells, was significantly decreased by transfection of the miR-5703 inhibitor into CCD-18Co cells. Finally, the effect of the miR-5703 inhibitor on fibroblast-mediated cancer cell growth was investigated. The miR-5703 inhibitor suppressed the proliferation of SW480 cells co-cultured with CCD-18Co cells (Fig. 6F).