Exosomal lncRNA 91H is associated with poor development in colorectal cancer by modifying HNRNPK expression

Serum samples from 232 CRC patients and 50 healthy people who underwent a routine health checkup and showed no disease at Nanjing First Hospital Affiliated to Nanjing Medical University, between 2011 and 2014 were collected in this study. Among serum samples, 68 were patients who were diagnosed as CRC without treatment, 56 were patients who had just underwent surgery without chemotherapy or radiotherapy and 108 were other patients who had underwent operation for more than half a year. Clinical information of all patients were collected including age, sex, TNM stage, tumor grade and family history of disease.

Cell lines HCT-8, HCT-116 and FHC (normal human intestinal epithelial cell line) were directly obtained from Shanghai Cell Collection, Chinese Academy of Sciences on March, 2017. All above cell lines were maintained in DMEM (Hyclone, USA) with 10% fetal FBS (Gibco, MA, USA) and cultured at 37 °C with 5% CO₂. All the tumor cell lines had been tested by short tandem repeat (STR) method in ABI 3500 Genetic Analyzer (USA Life 3500).

0.5 mg/ml exosomes spotted onto a glow-discharged copper grid on the filter paper were dried for 10 min by the infrared lamp. Then exosomes were stained in a drop of 1% aqueous solution of phosphotungstic acid for 6 min and dried for 25 min by the infrared lamp. Finally, exosomes were examined under transmission electron microscopy at 100 keV.

Exosomes samples were lysed using a lysis buffer. Then, the protein concentration was quantified using a BCA Protein Assay Kit (Beyotime, Shanghai, China). Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and western blot analyses were performed according to the standard procedures. Primary antibodies used included anti-HNRNPK (G. Dreyfuss, UPENN, Philadelphia, PA), anti-CD63 (sc5275, Santa Cruz Biotechnology, Santa Cruz, CA, USA) and anti-β-actin (1: 2000, ab8227, Abcam, Cambridge, UK). Secondary antibodies were goat anti-rabbit HRP secondary antibody (System Biosciences, Mountain View, CA).

LncRNA 91H RNAs in vitro were transcribed using the Biotin RNA Labeling Mix (Roche) and SP6 RNA polymerase (Ambion) and purified by RNA Clean & Concentrator™-5 (Zymo Research). Then cell lysates were freshly prepared using Protea Prep Zwitterionic Cell Lysis Kit, Mass Spec Grade (Protea^®) with Anti-RNase, Protease/Phosphatase Inhibitor Cocktail, Panobinostat and Methylstat supplemented in the lysisbuffer. The BcMag™ Monomer avidin Magnetic Beads (Bioclone) were firstly prepared according to manufacturer’s instructions and then immediately subjected to RNA (20 µg) capture in RNA capture buffer for 25 min at room temperature with agitation. The RNA-captured beads were washed with NT2 buffer and incubated with 30 mg cell lysates diluted in NT2 buffer supplemented with 50 U/ml RNase OUT™, 50 U/ml Superase1·IN™, 2 mM dithiothreitol, 30 mM EDTA and Heparin 0.02 mg/ml at 4 °C for 2 h with rotation. Washing with NT2 buffer, NT2-high salt buffer (500 mM NaCl), NT2-high salt buffer (1 M NaCl), NT2-KSCN buffer (750 mM KSCN) for twice and PBS (once) for 5 min at 4 °C and eluted by 2 mM D-biotin in PBS. The eluted protein complexes were denatured, reduced, alkylated and digested with immobilized trypsin (Promega) for MS analysis at MD Anderson Cancer Center Proteomics Facility.

RNA of cell lines, serum and exosome samples were extracted with E.Z.N.A. Total RNA Kit (omega, USA) following the manufacturers’ protocol. To avoid the possible DNA contamination, RNA was digested with RNase free DNaseI and cleaned with the RNA easy MinElute Cleanup Kit (Qiagen).

RNA interference was carried out by using synthetic siRNAs, as described by earlier studies [17]. Two synthetic siRNAss (si-lncRNA 91H1 and si-lncRNA 91H2) corresponding to the lncRNA 91H RNA sequences 5′-GGCGUCAUUCUGAUGGGACTT-3′ and 5′-UUCAGGAGCUUAAGAUGCUTT-3′ were used and 5′-CGUGGGUGGAUGCAUGGAUTT-3′ was used as a negative control. The cells (5 × 10²) were grown on coverslips in six-well plates and transfected with 400 pmol of different siRNAs. After transfection, followed by lysed for total RNA isolation.

Total reverse transcriptions were performed as follows: 1ug RNA, 4 µl of buffer containing random hexamers and reverse transcriptase (RT; Takara Japan) were incubated for 15 min at 37 °C, for 5 s at 85 °C, and stored at 4 °C in a final volume of 20 µl. Real-time PCR amplifications were performed by using a QuantiTect Sybrgreen PCR Kit (Qiagen) with 2 µl of cDNA and 500 nM concentrations of the primers. The primers used for the lncRNA 91H were Fw 5′-CCTGAAACAGGGAGGTGGTG-3′ and Rev 5′-GTGCCTGGAGCCTGTCTAAC-3′. We selected β-actin, a housekeeping gene as an internal control: Fw: 5′-CTCCATCCTGGCCTCGCTGT-3′ and Rev: 5′-GCTGTCACCTTCACCGTTCC-3′. Relative quantification of gene expression changes was recorded after normalizing for 18S expression, computed by using the 2^−∆∆CT method (user manual 2, ABI 7500 System).

Infected HCT-116 cells were cultured in six-well plates for 48 h. Then wounds were created in confluent cells using a 20 μl pipette tip. Free-floating cells and debris were removed using PBS. Then the cells were divided into 3 groups: cancer exosomes, cancer exosomes with lncRNA 91H interference and negative control without exosomes and FBS-free medium was added with or without exosomes (2 μg/µl). Finally we observed the wound healing at 0, 48 h and photographed representative scrape lines using the light microscope. Each experiment was repeated in triplicate.

For migration assay, HCT-116 cells were divided into 3 groups: cancer exosomes, cancer exosomes with lncRNA 91H interference and negative control without exosomes. Then cells (1 × 10⁵) were seeded into the upper chamber of transwell plates with serum-free DMEM and exosomes (2 μg/µl), in a 24-well format with 6 mm diameters. 600 μl DMEM with 5% FBS were added to the bottom chamber. Culturing for 24 h, cells were fixed with methanol and stained with Wright’s stain. The top of the permeable membrane cells remained were removed using a cotton swab. At last the numbers of migrating cells were counted in five random fields under a light microscope and the average value of five fields was used.

For invasion assay, the top chambers were treated with 40 mg/ml basement membrane Matrigel (Becton–Dickinson, USA) at 37 °C for 30 min. As before, Cells treated with or without exosomes (1 × 10⁵) were added into the top chambers, the bottom chambers were filled with DMEM containing 10% FBS. After incubating for 24 h, the cells that invaded the reverse side of top chambers were fixed, stained, and counted using a light microscope. Each invasion assay was performed in three replicates.

The cut-off levels of exosomal 91H the receiver operating characteristic curve analysis (ROC). Comparison of continuous data was analyzed using an independent t test, and categorical data were analyzed by Chi square test. The impact of exosomal 91H expression and clinicopathologic parameters on relapse-free survival (RFS) was calculated by using a Cox proportional hazards model. Statistics with P value < 0.05 were considered as statistically significant. All these statistical calculations were performed using IBM SPSS 22.0 software (IBM, USA).

Exosomes in Fig. 1 were created by Gatan digitalmicrograph and Gels/blots were created by Syngene G:BOX. Photos of scratch wound and migration/invasion assays were all crested by Leica Application Suite. At last, we confirmed that all methods were performed in accordance with the relevant guidelines and regulations as described above.

The characteristics of exosomes in the serum were observed under transmission electron microscope (TEM) which exhibited double layer membrane that ranged from 40 to 100 nm (Fig. 1a). In addition, CD63 which was the protein marker of exosome, was also used to confirm the evaluation of exosomes by western blot analysis (Fig. 1b).

By performing qRT-PCR, the expression of exosomal 91H in cell lines HCT-8, HCT-116 were 16.96 and 5.30-fold change higher than FHC with P < 0.05 (Fig. 2a). Then the expression of lncRNA 91H in exosomes was compared with other places of HCT-116 medium which demonstrated that the expression of lncRNA 91H in exosomes was 4.72-fold change higher than other places of cell medium (Fig. 2b, P < 0.05). At last, RNA interference was used in this study. By transfection for 48 h, lncRNA 91H was effectively silenced in HCT-116 cells and the exosomal 91H was also decreased in each siRNA group compared with negative control (Fig. 2c, P < 0.05).

The association between lncRNA 91H and exosomes were also detected in 68 CRC patients without treatment. The RT-PCR products of lncRNA 91H extracted from part of cancer patients’ exosomes were firstly electrophoresed on 2% agarose gel with the length of 271 kb. Then by qRT-PCR, we found that the expression of exosomal 91H in patients’ serum was significantly higher than healthy people (Fig. 2d, P < 0.05). Next the lncRNA 91H expression in exosomes and serum without exosomes was compared with the result that the expression levels of lncRNA 91H in exosomes was significantly higher than other places of patients’ serum (Fig. 2e, P < 0.05). At last, the exosomal 91H expression in patients with operation were detected which showed that exosomal 91H expression was apparently decreased after operation (Fig. 2f, P < 0.05).

We next investigated the molecular mechanism by which lncRNA 91 regulated tumor development. Pull down assays with cell lysate showed that heterogeneous nuclear ribonucleoprotein K (HNRNPK) protein might directly interact with lncRNA 91 by MS analysis (Fig. 3a). To further confirm the lncRNA 91H-protein interactions in vivo, RNA interference was used in this study which demonstrated that the HNRNPK protein expression significantly decreased compared with negative control (P < 0.05, Fig. 3b).

A scratch wound assay was firstly performed to evaluated the effect of exosomal 91H on cell motility. Cells treated with cancer exosomes resulted in enhancing motility of HCT-116 cells. Furthermore, compared with infected cells with negative control, the wound recovery was significantly lower in siRNA-infected cells (Fig. 4).

In addition, migration and invasion assays were used to further gauge the effect of exosomal 91H on CRC and normal FHC cells. The results demonstrated that the migration was significantly increased in infected CRC cells with cancer exosomes, compared with negative control cells (Fig. 5, P < 0.05) and the invasiveness was also enhanced in infected CRC cells with cancer exosomes, compared with other groups (Fig. 6, P < 0.05). These findings both indicated that exosomal 91H might be closely correlated with proliferation, migration, and invasion of CRC cell lines.

The exosomal 91H expression level in serum were determined by qRT-PCR for CRC patients. After half a year, the exosomal 91H expression level in 108 patients’ serum was 1.21 ± 1.24 which was higher than those who had just underwent operation (1.22 ± 0.94, P < 0.05). Furthermore, the optimal cut-off level of exosomal 91H expression for RFS was calculated by ROC curve analysis which divided the 108 CRC patients into a high 91H expression group (n = 57) with 91H expression ratio ≥ 0.85 and a low 91H expression group (n = 51) with 91H expression ratio < 0.85. Then a univariate analysis was conducted which showed that patients with higher exosomal 91H expression level usually had a higher risk in tumor relapse or metastasis than others (HR 5.13, 95% CI 1.23–21.35, P = 0.025). At last, multivariate analysis was performed in order to assess the independent prognostic factors with RFS. The results also indicated that patients with high 91H expression (HR 7.14, 95% CI 1.23–21.35, P = 0.040) were associated independently with RFS (Table 1).