Up-regulated microRNA-143 in cancer stem cells differentiation promotes prostate cancer cells metastasis by modulating FNDC3B expression

The cell lines used in this study included the human prostate cancer cells PC-3, PC-3-M and LNCaP and SV40-transformed kidney cell line 293 T (ATCC, Manassas, VA). The prostate cancer cells were cultured in RPMI-1640 medium and 293 T were cultured in DMEM (Gibco, Invitrogen) supplemented with 10% FBS (Hyclone). Cells were grown to 90% confluence, trypsinized, and plated at a density of 1,000 cells/ml in serum-free DMEM/F12 medium (Gibco, Invitrogen) containing 20 ng/ml epidermal growth factor (EGF, R and D Systems, MN), 5 μg/ml insulin, 0.4% bovine serum albumin (Sigma, St. Louis, MO), and 2% B27 (Invitrogen, CA) in 10 cm² culture dishes. To propagate spheres in vitro, spheres were collected by gentle centrifugation, dissociated to single cells, and cultured to generate the next generation of spheres [7]. Cells were grown in a humidified atmosphere of 5% CO₂ at 37°C.

Total RNA samples were analyzed by Chipscreen (Chipscreen Biosciences, Ltd., Shenzhen, China) for miRNA microarray experiments. Procedures were performed as described in detail (http://​www.​chipscreen.​com). Each miRNA microarray chip contained 1,199 probes, including 703 identified human miRNAs and 393 predicted miRNAs. The arrays were scanned with a GenePix Pro 6.0 (Axon, Ltd), and images were analyzed using DMVS (Chipscreen Biosciences, Ltd.). Microarray data for each sample were normalized to the median. All data is MIAME compliant and that the raw data has been deposited in a MIAME compliant database (GEO, accession ID: GSE44069).

The transfection of PC-3, PC-3-M, and LNCaP cells with 50 nM miR-143 inhibitor or negative control (NC) (GenePharma, Shanghai, China) was performed with lipofectin 2000 (Invitrogen) according to the manufacturer’s protocol for 48 hours. The sequence of the miR-143 inhibitor and NC are listed in Table 1

Lentiviruses containing GFP-miR-143 inhibitor or GFP-negative control miRNA vector were purchased from Genepharma. PC-3-M cells were pre-seeded in a 6-well plate overnight and infected with 200 μl of virus. 24 hours after addition of viruses, infected cells were selected by adding 20 ng/ml puromycin to growth medium for 4–5 passages. Stable cell lines were verified by qRT-PCR and fluorescence microscope.

The transwell assay was done by using transwell chamber consisting of 8 mm membrane filter inserts (Corning). The spheres were collected by gentle centrifugation and trypsinized to single cells in Trypsin-EDTA solution as well as the adherent cells. The cells were suspended in serum-free RPMI-1640 medium and counted by automated cell counter (Countstar). Then 5 × 10⁴ cells were added to the upper chamber, whereas lower chamber were filled with complete medium. After 12 hours of incubation, the cells in the upper chamber were carefully removed with a cotton swab, and the cells that had migrated through the membrane to the lower surface were fixed with 90% methanol and stained with 0.1% crystal violet. The invasion assay was performed according to a similar method, except that Matrigel (BD Biosciences) was pre-coated on the membrane of the upper chamber. The cell count was done under the microscope (200×).

One day before scratching, the cells transfected with miR-143 inhibitor or NC were seeded into 6-well plates to almost total confluence in 24 hours. An artificial homogenous wound was created onto the monolayer with a sterile 10 μl tip. After scratching, the cells were washed with serum-free medium. Images of cells migrating into the wound were captured at 0 and 24 hours by inverted microscope (100×).

All of the animal care and experimental procedures were approved by the Institutional Animal Care and Use Committee of Sun Yat-sen University. Male BALB/c athymic nude mice (4–6 weeks old) were purchased from the Experimental Animal Center of Guangdong province and housed in SPF barrier facilities under a 12 h light/dark cycle. PC-3-M cells that stably expressed miR-143 inhibitor (5 × 10⁶) or NC (5 × 10⁶) were suspended in 100 μl PBS and subcutaneously injected into the right side of the posterior flank of the mice. Ten mice were used in each group. Tumor development and metastases were monitored with a bioluminescence imaging system (ZKKS-MulAurora PI1024) on day 30 post-injection. Mice were photographed under 470/30 nm illumination, and images of the emitted fluorescence were acquired at 545/60 nm. Luminescence data were gathered over the maximum exposure period without pixel saturation (0.5–12 seconds).

Potential miRNA targets were predicted and analyzed using two publicly available algorithms, including PicTar (http://pictar.mdc-berlin.de) and TargetScan (http://​www.​targetscan.​org/​). These searchable websites predict biological targets of miRNAs by searching for the presence of conserved 8mer and 7mer sites that match the seed region of miRNA and provide details 3′ UTR alignments with predicted sites. To decrease the number of false-positive results only putative target genes predicted by at these two programs was accepted.

The 3^′-UTR luci vector was constructed using the psi-CHECK2 luciferase reporter vector containing a fragment of the FNDC3B mRNA 3^′UTR, which carries a putative miR-143 complementary site. 293 T cells (5 × 10⁴ per well) were pre-seeded in a 24-well plate the day before transfection and transfected with 0.5 μg of the 3^′ UTR luci vector and 50 nM miR-143 mimics or negative control using Lipofectamine 2000 (Invitrogen). Assays were performed using the Dual-Luciferase Reporter Assay System (Promega) after 48 hours of transfection.

The total protein extracted from the samples was resolved on 10% sodium dodecyl sulfate–polyacrylamide gels and electrophoretically transferred to a polyvinylidene fluoride membrane. Blots were blocked with 5% skim milk followed by incubation with antibodies specific for either FNDC3B (Sigma) or GAPDH (Cell Signaling). The blots were then incubated with goat anti-rabbit or anti-mouse secondary antibody (Cell Signaling) and visualized using enhanced chemiluminescence.

Data were presented as the means ± SD from three separate experiments. The differences between groups were analyzed using Student’s t test when only two groups were compared or a one-way analysis of variance (ANOVA) when more than two groups were compared. The differences between groups of metastasis in vivo were analyzed using Chi-squared test (χ ² test). All of the statistical analyses were performed with SPSS 16.0. The difference was considered to be statistically significant at P <0.05.

First, to elucidate whether the PC-3 sphere cells turned into differentiated cells when sphere cells were digested into single cells for re-adherent culture (10% FBS-RPMI-1640 medium), we compared the expression levels of cancer stem cells markers, such as Oct4, Sox2 and Nanog by qRT-PCR. The expression of Oct4, Sox2 and Nanog were gradually decreased in re-adherent culture (Figure 1A). This suggested that PC-3 sphere cells had the cancer stem cells phenotype and differentiated in re-adherent culture. Second, to investigate whether miRNAs were differentially expressed in PC-3 spheres and adherent cells, we compared miRNA expression profiles using a miRNA microarray. We observed the increased expression of 25 miRNAs and decreased expression of 36 miRNAs in PC-3 sphere cells compared with adherent cells (Table 2). Third, to confirm our microarray data, qRT-PCR was performed to analyze the expression of the most significantly differentially expressed miRNAs (Table 2). The expression of miR-143 was down-regulated 8.4-fold in PC-3 sphere cells compared with adherent cells (Figure 1B). Next, we tested 10 miRNAs for which the expression levels were most changed during PC-3 sphere cells re-adherent culture on days 0, 2, and 4 by qRT-PCR. The expression of miR-143 was progressively increased during re-adherent culture, but no significant change was observed for the other 9 miRNAs (Figure 1C). Therefore, we selected miR-143 to further investigate its role in prostate cancer. These results suggested that miR-143 might play a regulatory role in prostate cancer stem cells differentiation in vitro.

To evaluate the metastatic mechanism of prostate cancer stem cells, we compared the migration capacity of PC-3 spheres and adherent cells with a transwell assay. Interestingly, less PC-3 sphere cells penetrated through the gel-membrane compared with adherent cells (Figure 2A). However, when we digested the sphere cells into single cells for re-adherent culture, the cells gradually showed increased migration capability and reached the level of adherent cells on the fourth day (Figure 2B). These data suggested that prostate cancer stem cells might exhibit lower metastatic ability but generate differentiated cells expressing a highly aggressive phenotype.

MiR-143 expression and cells migration were increased during sphere cells differentiation. To evaluate whether miR-143 played an important role in the metastasis of prostate cancer cells, we decreased the expression of miR-143 in PC-3, PC-3-M and LNCaP cells by transient transfection with siRNA. MiRNA negative control-transfected cells, PC-3/NC, PC-3-M/NC and LNCaP/NC, were used as control groups. As shown in Figure 3A, the relative expression of miR-143 in PC-3, PC-3-M and LNCaP cells transfected with the miR-143 inhibitor was approximately 10-fold lower compared with cells transfected with NC. Migration and invasion assays were performed in vitro. Intriguingly, in wound-healing and transwell assays, cells migration was much slower and fewer cells penetrated through the gel-membrane when cells transfected with the miR-143-inhibitor compared with NC (Figure 3B,C). Moreover, cells invasion was markedly inhibited by the transfection with the miR-143 inhibitor (Figure 3D). These data strongly suggest that inhibition of miR-143 repressed the migration and invasion of prostate cancer cells in vitro.

To further explore the effect of miR-143 on prostate cancer metastasis in vivo, PC-3-M cells were stably transfected with NC or miR-143 inhibitor and subcutaneously injected into nude mice. As shown in Figure 4A, the relative expression of miR-143 in PC-3-M cells transfected with the miR-143 inhibitor was stably about 10-fold lower compared with cells transfected with NC at least 60 days. The development and metastasis of tumors in vivo were monitored by visualizing the bioluminescence emitted from the luciferase-tagged tumors on day 30. Mice injected with the miR-143 inhibitor PC-3-M cells developed fewer systemic metastasis (2/10 versus 8/10), especially fewer macroscopic and bioluminescent nodes in the liver (3/10 versus 9/10), compared with mice implanted with NC cells (Figure 4B,C, Table 3). Histological confirmations were made by H&E-stainning (Figure 4D). Mice injected with the miR-143 NC PC-3-M cells developed severe metastatic lesions in the ribs of nude mice, and only a litter bone and muscle were residual. However, no bone metastasis was seen in the mice implanted with miR-143 inhibitor cells. These findings suggested that the down-regulation of miR-143 inhibited prostate cancer cells metastasis in vivo.

To elucidate the mechanisms through which miR-143 induced prostate cancer cells metastasis, two publicly available algorithms were used to help predict miR-143 targets. Among the approximately 200 candidate genes, fibronectin type III domain containing 3B (FNDC3B) was a high-scoring candidate. FNDC3B is a member of the fibronectin family, which regulates cell motility, and is down-regulated in tumor cells with high metastatic potential [19]. Zhang et al. [20] reported that up-regulated miRNA-143 enhanced hepatocarcinoma metastasis by repressing FNDC3B expression. Thus, we focused on the possible regulation of FNDC3B by miR-143. As shown in Figure 5A, miR-143 was partially complementary to the FNDC3B mRNA 3^′ untranslated region (UTR) element. Consequently, a luciferase reporter assay was performed to verify whether miR-143 could directly target the FNDC3B 3^′-UTR. The luciferase activity of psi-CHECK2-FNDC3B was obviously decreased in the presence of miR-143, whereas no significant reduction was observed when cells were cotransfected with NC (Figure 5B). Moreover, we evaluated the effects of miR-143 on FNDC3B mRNA and protein levels in PC-3-M cells by RT-PCR and western blotting. FNDC3B mRNA levels showed no significant change whereas protein levels were increased in PC-3-M cells transfected with miR-143-inhibitor (Figure 5C). We also investigate the expression of miR-143 and FNDC3B in the tumor tissue of nude mice by RT-PCR and western blotting. As expected, the expression of miR-143 was significantly decreased whereas FNDC3B protein levels were enhanced in the tumor tissue of the group of miR-143-inhibitor (Figure 5D). Taken together, these findings suggested that miR-143 might modulate prostate cancer metastasis by targeting FNDC3B.