MicroRNA miR-24 enhances tumor invasion and metastasis by targeting PTPN9 and PTPRF to promote EGF signaling

MicroRNAs (miRNA) are non-coding single-stranded RNA molecules with 18–24 nucleotides, which are produced from endogenous transcripts of short stem-loop structures. Through binding to the 3′-untranslated region (3′UTR) of different target mRNAs (mRNAs), microRNAs can repress the translation of the target mRNAs (Seitz et al., 2003). As a new class of regulatory molecules, miRNAs have diverse functions in cell proliferation (Shatseva et al., 2011; Viticchiè et al., 2011; Yu et al., 2012), survival (Fang et al., 2012a; Ye et al., 2011), invasion (Deng et al., 2011; Luo et al., 2012; Siragam et al., 2012), cell differentiation (Goljanek-Whysall et al., 2012; Kahai et al., 2009; Rutnam and Yang, 2012b), morphogenesis (Wang et al., 2008a), tissue growth (Shan et al., 2009), angiogenesis (Fang et al., 2011; Lee et al., 2007; Smits et al., 2010; Zou et al., 2012), tumor development (Nohata et al., 2012; Volinia et al., 2006) and metastasis (Huang et al., 2008; Ma et al., 2007; Rutnam and Yang, 2012a). The largest functional group of miRNAs is the one involved in cancer development. Recent studies have shown that miR-24 can inhibit apoptosis of both cancer cells (Qin et al., 2010) and cardiomyocytes (Qian et al., 2011) and can accelerate cell proliferation (Zaidi et al., 2009). Upregulation of miR-24 has been observed in oral carcinoma (Lin et al., 2010) and is one of the most abundant miRNAs in cervical cancer (Wang et al., 2008b). It appears that miR-24 can repress expression of the tumor suppressor p16^INK4a (Lal et al., 2008). Overall however, the role of miR-24 in cancer cell invasion and metastasis is still not clear.

Epidermal growth factor receptor (EGFR) is a member of the receptor tyrosine kinases. It is expressed on the cell surface mediating a variety of signals associated with cell proliferation. It has been demonstrated that altered expression of EGFR could result in cancer development (Zhang et al., 2007). Furthermore, breast cancer patients with EGFR upregulation have a less favorable prognosis than those with EGFR-negative tumors (Koletsa et al., 2010). Phosphorylation of EGFR plays a crucial role in the biological functions of EGFR (Magkou et al., 2008). Protein-tyrosine phosphatases can remove the phosphate groups from the phosphorylated receptor tyrosine kinases (Tonks, 2006) and antagonize their activity (Zhu et al., 2008). It has been reported that tyrosine-protein phosphatase non-receptor type 9 (PTPN9) can dephosphorylate EGFR and inhibit growth and invasion of MBA-MD-231 cells (Yuan et al., 2010). The role of other PTPs in EGFR dephosphorylation is not clear. In this study, we demonstrated that miR-24 can activate EGFR phosphorylation resulting in enhanced tumor growth, tumor local invasion and metastasis, which resulted in decreased animal survival. Consistent with this was the activation of EGFR. The activation of EGFR activity was achieved by repressing expression of two phosphatases PTPN9 and receptor-type tyrosine-protein phosphatase F (PTPRF). We confirmed that both PTPN9 and PTPRF were the targets of miR-24.

To dissect the roles of miRNA in breast cancer development, we examined miRNA levels in breast cancer tissues and normal breast tissues. RNAs isolated from the paraffin-embedded tissues were subjected to real-time PCR analysis. The assays showed that miR-24 levels were significantly higher in breast tumor tissue than in the corresponding normal breast tissue from the same patient (Fig. 1A, P = 0.000136). We also isolated RNAs from frozen tumor tissues and the surrounding normal breast tissues. Real-time PCR analysis indicated that of the 23 pair specimens obtained from 23 patients, 15 pairs of samples showed higher levels of miR-24 in the tumors than in the normal tissues, while the other eight pairs showed lower levels of miR-24 in the tumor tissues (Fig. 1B). These experiments suggested that miR-24 may play a role in cancer development.

We generated a miR-24 expression construct (Fig. 1C) which was transfected into breast cancer cells, and then used cell sorting to isolate a stable pooled cell line. After confirming integration of the construct into the genome (supplementary material Fig. S1A, upper) and expression of miR-24 (supplementary material Fig. S1A, lower), we conducted cell invasion assays. Mock- and miR-24-transfected 4T1 cells were loaded onto a Transwell insert with serum-free Dulbecco’s modified Eagle’s medium (DMEM) and then incubated at 37°C for 24 hours. The invasive cells were stained and counted in six randomly selected fields under a light microscope. We found that cells transfected with miR-24 displayed higher levels of invasion than the mock-transfected cells (Fig. 1D; supplementary material Fig. S1B). To corroborate this result, we performed the invasion assay in a human breast cancer cell line MT-1 (Fig. 1E; supplementary material Fig. S1C) and found that expression of miR-24 promoted cancer cell invasion. The invasion assay was also performed in the presence of Matrigel and we confirmed that 4T1 cells transfected with miR-24 had higher invasive capacity than the mock-transfected cells (supplementary material Fig. S1D).

Cell migration assays were performed to examine the role of miR-24. Mock- and miR-24-transfected 4T1 and MT-1 cells were cultured until 95% confluence. The monolayer of cells was wounded and the cultures were maintained in the presence of mytomycin to inhibit cell proliferation. Cell migration was photographed under a light microscope every 24 hours. The distances between the wounding center and the front of the migrating cells were measured for statistical analysis. The experiments indicated that expression of miR-24 promoted migration of MT-1 cells (Fig. 1F; supplementary material Fig. S1E) and 4T1 cells (supplementary material Fig. S1F).

We examined proliferation of 4T1 cells transfected with miR-24 or the control vector. The experiments showed that the miR-24-transfected cells had higher rates of proliferation than the vector-transfected cells (supplementary material Fig. S1G). Analysis of cell cycle progression indicated that more vector-transfected cells than miR-24-transfected cells were in the G1 phase suggesting a lower rate of proliferation of the former cells (supplementary material Fig. S1H). Consistent with this was the finding that miR-24-transfected cells formed more and larger colonies than the mock-transfected cells in the colony formation assays (supplementary material Fig. S2A). In sphere formation assays, we observed that the miR-24-transfected cells formed more and large spheres than the mock-transfected cells (supplementary material Fig. S2B). In low serum conditions, the miR-24-transfected cells showed increased adhesion and spreading relative to the mock-transfected cells (supplementary material Fig. S2C). The above characteristics of the miR-24-transfected cells were indicative of metastatic cancer cells. We thus performed drug-resistance experiments by culturing the mock- and miR-24-transfected MT-1 cells in the presence of Docetaxel, Doxorubicin and Epirubicin and found that the miR-24-transfected cells were more resistant to these drugs than the mock-transfected cells (supplementary material Fig. S2D).

To corroborate the role of miR-24 in mediating cell activities, we developed an antisense construct to suppress miR-24 function. In cell proliferation assays, we detected that transfection of the anti-miR-24 construct decreased cell proliferation after four days of culture (supplementary material Fig. S2E). We also performed invasion and migration assays in the MDA-MB-231 cells that expressed higher levels of miR-24 and found that expression of anti-miR-24 significantly decreased the invasive capacity (supplementary material Fig. S2F,G).

In order to confirm the function of miR-24, three transfected stable cell lines were injected into nude mice subcutaneously. After tumors formed, their sizes were measured regularly until the mice were sacrificed or up to 80 days. The experiments showed that tumors formed by the miR-24-cells (miR-24 tumors) grew faster than the tumors formed by the mock-transfected cells (mock tumors) or the anti-miR-24-transfected cells (anti-miR-24 tumors; Fig. 2A). During the assays, mice were sacrificed when the tumors reached the limit size set by the Animal Care Committee of Sunnybrook Research Institute. The survival curves demonstrated that miR-24 expression decreased survival rates (Fig. 2B).

The tumors were harvested, sectioned, subjected to Hematoxylin and Eosin (H&E) staining. In the mock tumors, a mixture of dead cells (showing nuclear condensation and fragmentation) and live cells were detected. Cell death was reduced in the miR-24 tumors but much higher in the Anti-miR-24 tumors than in the mock tumors (Fig. 2C). In the anti-miR-24 tumors, severe necrosis was detected in the center of the tumor. In addition, significant invasion was observed in the miR-24 tumors, showing the invasion of tumor cells into muscles. Tumor sections were also probed with anti-CD34 antibody to examine angiogenesis. We found that miR-24 expression increased tumor-associated vascularization, while anti-miR-24 expression inhibited vascularization (Fig. 2D).

We further performed tumor metastatic assays. Mock- and miR-24-transfected MT-1 cells were injected into NOD-SCID mice via tail vein, 15 in each group. Six weeks after the injection, three mice in the miR-24 group developed cachexia and four had lung metastasis in necroscopy. Typical metastatic lesions in the lungs are shown (supplementary material Fig. S2H). DNA was isolated from lung tissues and subjected to PCR to amplify the CMV promoter to indicate metastasis of the cancer cells. Expression of miR-24 promoted metastasis (Fig. 2E). Lung sections were obtained and subjected to H&E staining. Nodules were detected indicating metastasis of the tumor cells to the lungs (Fig. 2F). The miR-24-group showed multi-focal metastasis lesions in the lungs (supplementary material Fig. S2I).

Since activation of EGFR is tightly associated with tumor invasion and metastasis, we analyzed levels of phosphorylated EGFR (pEGFR) and EGFR in both 4T1 and MT-1 cells and observed a dramatic induction of pEGFR levels in both cell types, while the levels of EGFR were similar, suggesting activation of the existing EGFR protein (Fig. 3A; supplementary material Fig. S3A). The cell lines expressing higher levels of miR-24 produced higher levels of pEGFR (supplementary material Fig. S3B). We further analyzed pEGFR level in the tumor sections and found that the miR-24 tumors expressed higher level of pEGFR than the mock tumors (Fig. 3B). The effect of miR-24 on pEGFR expression was confirmed by western blot analysis, which showed higher levels of pEGFR in the miR-24 tumor than in the mock tumor (Fig. 3C, left). In addition, the miR-24 tumor displayed higher levels of pErk than the mock tumor (Fig. 3C, right). Lung sections prepared from mock and miR-24 mice were also analyzed for pEGFR activity. A higher level of pEGFR activity was displayed in the miR-24 lung nodules than in the mock control (Fig. 3D). On the other hand, transfection with anti-miR-24 decreased pEGFR level (supplementary material Fig. S3C).

We employed the pEGFR inhibitor AG1478 to confirm the effect of pEGFR in mediating miR-24 functions. Mock- and miR-24-transfected MT-1 cells were treated with AG1478 followed by analysis of pEGFR levels. We detected downregulation of pEGFR levels in the presence of AG1478 compared with the untreated miR-24-cells (supplementary material Fig. S3D, left). In cell invasion assays, we found that treatment with AG1478 significantly decreased cell invasion in both the miR-24- and mock-transfected cells compared with the untreated controls (Fig. 3E; supplementary material Fig. S3E). In migration assays, we found that treatment with AG1478 significantly reduced the activities of cell migration in both the miR-24- and mock-transfected cells compared with the untreated controls (Fig. 3F; supplementary material Fig. S3F).

To test how miR-24 exerted metastasis-promoting activities, we analyzed potential targets of miR-24 using the software miRBase (http://www.mirbase.org/cgi-bin/mirna_entry.pl?acc = MI0000080). There were a great number of candidates potentially targeted by miR-24. We focused on PTPN9 and PTPRF, two protein tyrosine phosphatases. The potential binding sites appeared at nucleotides 2333–2355 for PTPN9 and nucleotides 6111–6132 for PTPRF (Fig. 4A). We examined the levels of these proteins in 4T1 and MT-1 cells stably transfected with miR-24. In both cell lines, miR-24 expression resulted in decreased levels of PTPN9 and PTPRF (Fig. 4B, left). Similar results were obtained in the miR-24 tumor (Fig. 4B, right). Consistent with this, transfection with anti-miR-24 increased PTPN9 and PTPFR levels (Fig. 4C). Treatment with AG1478 produced little effect on PTPN9 and PTPRF expression (supplementary material Fig. S3D, right).

To confirm the targeting of PTPN9 and PTPRF by miR-24, we performed luciferase activity assays. The luciferase reporter vector was engineered with a fragment of the 3′UTR of PTPN9 or PTPRF harboring the miR-24 target site, producing two luciferase reporter constructs Luc-PTPN9 and Luc-PTPRF (supplementary material Fig. S4A). In addition, the potential target sites were mutated as shown (Fig. 4A) producing two mutant constructs Luc-PTPN9-mut and Luc-PTPRF-mut (supplementary material Fig. S4A). Breast cancer cell line 4T1 was co-transfected with miR-24 and the luciferase reporter construct Luc-PTPN9 or Luc-PTPN9-mut. The luciferase reporter vector containing an unrelated sequence served as a negative control. The presence of the miR-24 target site reduced luciferase activities, an effect that was reduced when the target site was mutated (Fig. 4D, left). As well, co-transfection with miR-24 and Luc-PTPRF displayed significantly decreased luciferase activities compared with the control, an effect that was reversed when the target site was mutated (Fig. 4D, middle). As expected, a mixture of PTPN9 and PTPRF (50% of each) produced similar results in the luciferase activity assays (Fig. 4D, right). We also conducted the luciferase activity assays in a different cell line U343 and similar results were obtained (supplementary material Fig. S4B).

Examination of the target sequences indicated that the miR-24 target sites in the 3′UTRs of PTPN9 (supplementary material Fig. S4C) and PTPRF (supplementary material Fig. S4D) were highly conserved across different species. In all species obtained, only one nucleotide in PTPN9 and two nucleotides in PTPRF contained some deviations, while the rest were 100% homologous.

To confirm that PTPN9 and PTPRF were important in mediating the effects of miR-24 in regulating cell invasion and migration, we introduced siRNA to suppress their expression. Cell lysates prepared from MT-1 cells transfected with siRNAs against PTPN9, PTPRF or a control oligo were analyzed by western blotting to confirm appropriate RNA silencing as compared with the control (supplementary material Fig. S4E). We then analyzed EGFR expression and its activation using lysate from miR-24-transfected cells, which served as a control. Silencing of PTPN9 and PTPRF had no effect on EGFR expression but enhanced EGFR phosphorylation, reaching levels similar to those found in miR-24-transfected cells (Fig. 4E). Treatment with EGF also enhanced EGFR phosphorylation (supplementary material Fig. S4F).

MT-1 cells were treated with siRNAs against PTPN9, PTPRF, or the control oligo. The treated cells were harvested for invasion and migration assays. Silencing PTPN9 and PTPRF by treating the cells with siRNA against PTPN9 and PTPRF, respectively, promoted cell invasion (Fig. 4F; supplementary material Fig. S4G) and migration (Fig. 4G; supplementary material Fig. S4H).

To examine whether there was any correlation in the expression of miR-24 and its potential protein targets, we analyzed expression of pEGFR, PTPN9 and PTPRF in a number of cell lines including human breast carcinoma cells and found that expression of pEGFR was higher in the breast cancer cell lines MDA-MB468, MT-1, and MDA-MB231 than in the other cell lines tested (supplementary material Fig. S5A). Consistent with this was the lower levels of PTPN9 and PTPRF in MDA-MB468, MT-1 and MDA-MB231. Interestingly, the levels of miR-24 were also higher in these cell lines than the other cell lines (supplementary material Fig. S5B), suggesting a potential role of miR-24 in repressing PTPN9 and PTPRF expression in these cells, resulting in promotion of EGFR activity.

We then examined expression of pEGFR, PTPN9 and PTPRF in human breast carcinoma specimens. We found that expression of pEGFR was much higher in human breast carcinoma specimens than in the normal breast tissues (Fig. 5; supplementary material Fig. S5C). On the other hand, expression of PTPN9 and PTPRF was much weaker in the tumor tissues than in the normal tissue (Fig. 5A; supplementary material Fig. S5D,E). Double staining confirmed that expression of pEGFR was stronger than expression of PTPN9 and PTPRF in the tumor tissues, which was opposite to what was found in the normal tissues (supplementary material Fig. S6). The sections were probed for miR-24 expression using the miRCury LNA microRNA ISH Optimization Kit from Exiqon. We detected high levels of miR-24 expression in the cancer cells but not in the normal ductal structure of the breast (Fig. 5B).

To further confirm that PTPN9 and PTPRF played essential roles in mediating miR-24 effect, we conducted rescue experiments. The retroviral vector pBABE was engineered with the coding sequences of either PTPN9 or PTPRF (supplementary material Fig. S7A). MT-1 cells were transfected with PTPN9 combined with miR-24, PTPRF combined with miR-24, miR-24 alone or were mock transfected and then analyzed by western blotting. As expected, transfection with PTPN9 and PTPRF increased their expression respectively, which resulted in decreased pEGFR levels (Fig. 6A).

Invasion assays were performed on MT-1 cells transfected with PTPN9 combined with miR-24, PTPRF combined with miR-24, miR-24 alone or the mock construct. Enhanced cell invasion induced by miR-24 expression was abolished by transfection with PTPN9 and PTPRF (Fig. 6B; supplementary material Fig. S7B). The cells were also cultured to sub-confluence before being utilized in wound healing and migration assays. Enhanced cell migration by miR-24 expression was abolished by transfection with PTPN9 and PTPRF (Fig. 6C; supplementary material Fig. S7C). Ectopic expression of PTPN9 and PTPRF also abolished the effects of miR-24 on cell proliferation (supplementary material Fig. S7D) and colony formation (supplementary material Fig. S7E). The rescue effects of PTPN9 and PTPRF were also confirmed in another breast cancer cell line 4T1. Ectopic expression of PTPN9 and PTPRF reduced activation of pEGFR (supplementary material Fig. S7F), resulting in decreased cell invasion (supplementary material Fig. S7G), cell proliferation (supplementary material Fig. S8A), migration (supplementary material Fig. S8B) and colony formation (supplementary material Fig. S8C).

The MT-1 cells stably infected with pBABE-PTPN9 and pBABE-PTPRF were subjected to metastasis assays. After being sacrificed, the mice were examined for nodules in the lung. Lung tissues were harvested for DNA isolation followed by real-time PCR analysis to quantify ectopic DNA in the lung, which is a sign of metastasis. The experiments indicated that there were significantly higher levels of metastasis in the miR-24-treated mice than that in the mock-treated mice (Fig. 6D). Ectopic expression of PTPN9 or PTPRF decreased metastasis significantly. The tissues were also subjected to sectioning and H&E staining. We found nodules in the miR-24-treated mice but not in the mock-treated or in the PTPN9- and PTPRF-treated mice (Fig. 6E). The lung sections were immunostained for expression of PTPN9 and PTPRF to confirm metastasis of tumor cells (supplementary material Fig. S8D) and activation of pEGFR. We found that ectopic expression of PTPN9 and PTPRF abolished the increased expression of pEGFR by miR-24 (Fig. 6F).

We finally investigated the signaling pathway mediating the effects of miR-24. Cell lysates from miR-24- or mock-transfected MT-1 cell were analyzed by western blotting and probed for expression of MMP2, MMP11, ADAM15, TIMP2 and TIMP3. We detected upregulation of MMP2, MMP11 and ADAM15, but downregulation of TIMP2 in the cells (Fig. 7A). Expression of TIMP3 was not affected. To corroborate these results, we transfected MDA-MB-231 cells with anti-miR-24 and mock constructs, followed by western blot analysis of the expression of MMP2, MMP11, ADAM15, TIMP2 and TIMP3. Opposite to the above results, we detected downregulation of MMP2, MMP11 and ADAM15 but upregulation of TIMP2 in the cells (Fig. 7B).

To confirm that the up- and downregulation of these proteins was the consequence of miR-24 targeting PTPN9 and PTPRF and activation of pEGFR, we treated the mock- and miR-24-transfected MT-1 cells with 2.0 µM AG1478. Expression of the above proteins was analyzed by western blot analysis. We detected downregulation of MMP2, MMP11 and ADAM15 but upregulation of TIMP2 in the miR-24-transfected cells (supplementary material Fig. S9A). We confirmed that expression of pEGFR, MMP2, MMP11 and ADAM15 was upregulated while expression of PTPN9, PTPRF and TIMP2 was downregulated in the miR-24 tumors as compared with the mock tumors (supplementary material Fig. S9B). Upregulation of MMP2, MMP11, AMAD15 and pErk, while downregulation of TIMP2 was also detected in human breast carcinoma tissues compared with the normal duct structures in the breast (supplementary material Fig. S9C).

To understand the signaling pathway by which miR-24 functioned, we produced a figure to show how miR-24 possibly acts in breast cancer invasion and metastasis (Fig. 7C). In brief, miR-24 targeted PTPN9 and PTPRF, and downregulated expression of PTPN9 and PTPRF, which in turn activated EGFR (pEGFR). Increased pEGFR activity promoted expression of MMP2 and MMP11, which partly downregulated the expression of TIMP2 and upregulated the expression of ADAM15. As a consequence, increased expression of MMP2 and MMP11 enhanced breast cancer cell migration, invasion and metastasis. Thus, miR-24 represses PTPN9 and PTPRF, which coordinately regulate EGFR to mediate miR-24 functions. In an accompanying paper, we demonstrate that miR-17, producing miR-17-5p and miR-17-3p, represses PTEN, GalNT7 and vimentin. These targets, although functioning in different signal pathways, can mediate miR-17 functions coordinately, leading to the development of hepatocellular carcinoma (see Shan et al., 2013 in this issue of Journal of Cell Science).

Our study demonstrated that miR-24 was highly expressed in human breast carcinoma. This suggested a role of miR-24 in cancer development/progression, which was consistent with previous studies showing that miR-24 was highly expressed in patients with gastric cancer and lung cancer (Xie et al., 2011), cervical cancer (Wang et al., 2008b) and oral carcinoma (Lin et al., 2010). MiR-24 has also been reported to target p16, a tumor suppressor (Lal et al., 2008). We found that expression of miR-24 promoted tumor growth, invasion and metastasis by upregulating EGFR activity due to its suppressive effect on PTPN9 and PTPRF, two phosphatases of EGFR. Rescue experiments expressing these phosphatases reversed the effects of miR-24 on tumor cell invasion and metastasis. Currently reported miR-24 functions include promotion of cell proliferation (Zaidi et al., 2009) and inhibition of apoptosis in cancer cells (Qin et al., 2010) and cardiomyocytes (Qian et al., 2011). This may be due to its capacity in repressing a great number of potential genes (368 candidate targets being predicted at http://pictar.mdc-berlin.de/cgi-bin/PicTar_vertebrate.cgi?action = Search%20for%20targets%20of%20a%20miRNA&name2 = hsa-miR-24). In addition, miR-24 has been reported to target some ‘seedless’ 3′UTRs (Lal et al., 2009). Initially reported ten years ago (Lagos-Quintana et al., 2001), the roles of miR-24 are only starting to be recognized in a variety of cells and tissues.

To dissect the signaling pathway associated with miR-24 promotion of tumor growth, invasion and metastasis, we examined expression of a number of oncogenes and tumor suppressors and found that EGFR was strongly activated in the miR-24 expressing cells and tumors. As a cell surface receptor, EGFR transitions from an inactive state to an active state by autophosphorylation of its tyrosine residues (Swanton et al., 2008). Phosphorylation of EGFR results in activation of downstream MARK, ERK, Akt and JNK signaling pathways leading to DNA synthesis, cell proliferation and cell migration (Schmidt-Ullrich et al., 1997; Sørensen et al., 2006). Thus, phosphorylated EGFR plays important roles in tumor invasion and metastasis (Arora and Simpson, 2008; Gong et al., 2008). Over-activation of EGFR including constant activation and mutations is often associated with cancer development and progression (Lynch et al., 2004; Walker et al., 2009). Our results showing that EGFR was strongly activated upon miR-24 overexpression appeared to be a consequence of miR-24 repressing PTPN9 and PTPRF expression. Consistent with these results were the activation of Erk, ADAM15, MMP2 and MMP11 and downregulation of TIMP2 levels, which supported a role of EGFR in mediating miR-24 activities.

Since EGFR was not a target of miR-24, we reasoned that miR-24 may have targeted the endogenous inhibitors for EGFR phosphorylation such as the protein tyrosine phosphatases (PTPs). Due to their critical roles in physiology and tissue development, a great number of PTPs were expressed in various cell types, and the activity of each PTP was tightly controlled. PTPN9 is a non-receptor type 9 phosphatase in the cytoplasm, playing a role in intracellular secretory vesicle fusion (Huynh et al., 2004). PTPN9 has been reported to inhibit EGFR signaling by dephosphorylating EGFR leading to impaired growth and invasion of breast cancer cells (Yuan et al., 2010). We showed in this study that miR-24 could promote EGFR functions by repressing PTPN9 expression. Our results confirmed the function of PTPN9 in dephosphorylating EGFR as reported (Yuan et al., 2010) and added an additional layer to the regulation of PTPN9 expression, namely the regulation of PTPN9 by miR-24. This regulation appears to be significant in breast carcinoma patients who express high levels of EGFR. We found that approximately two-thirds of breast cancer patients express high levels of miR-24 and that most of the breast carcinoma specimens displayed high levels of pEGFR. Whether or not these patients expressed high levels of activated EGFR awaits further investigation. In addition, our results demonstrated that in human breast cancer cell lines, both EGFR and miR-24 were highly expressed while PTPN9 was downregulated, suggesting a possible inhibitory role of miR-24 in PTPN9 expression while promoting EGFR activity (e.g. enhanced EGFR phosphorylation).

Furthermore, we demonstrated that PTPRF was another target of miR-24. PTPRF is a receptor-type F protein tyrosine phosphatase. Although the function of PTPRF is not clear, it has been reported that PTPRF plays a role in mesenchymal stem cell self-renewal (Song and Tuan, 2006). There have been no reports on the role of PTPRF in dephosphorylating EGFR. In this study, we showed a reverse correlation between pEGFR and PTPRF levels. These results suggested that PTPRF may dephosphorylate EGFR in these cells. In clinical specimens, we also detected an inverted relationship between PTPRF and pEGFR levels, supporting a role of PTPRF in EGFR dephosphorylation. Knockdown of PTPRF increased pEGFR levels leading to enhanced cell invasion and migration. Ectopic expression of PTPRF decreased pEGFR levels, cell invasion, migration and tumor metastasis to the lung. These results directly supported the role of PTPRF in the inhibition of EGFR activities. The precise manner by which PTPRF dephospohrylates EGFR, awaits further investigation.

The monoclonal antibodies against ERK2, pERK, and the polyclonal antibodies against EGFR, pEGFR, PTPN9, PTPRF, MMP2, MMP11 and ADAM15 were obtained from Santa Cruz Biotechnology. EGF, selective EGFR inhibitor AG1478, mytomycin, and the monoclonal antibody against β-actin used were obtained from Sigma. The polyclonal antibodies against TIMP2 and TIMP3 were obtained from Abcam. Horseradish peroxidase-conjugated goat anti-mouse IgG and horseradish peroxidase-conjugated goat anti-rabbit IgG were obtained from Bio-Rad. Immunoblotting was performed using the ECL Western Blot detection kit. Cell Proliferation Reagent WST-1 and High Pure PCR Template Preparation kits were obtained from Roche Applied Science. The MT-1 cells used in this study was originally described by Naundorf and co-workers (Naundorf et al., 1992), while 4T1 cells were described by Aslakson and Miller (Aslakson and Miller, 1992).

Plasmid miR-24 and control plasmid were generated which have GFP and selected antibiotics marker respectively. Both plasmids contain a Bluescript backbone, a CMV promoter driving green fluorescent protein (GFP) expression and a human H1 promoter driving pre-miR-24 or a non-related sequence serving as a control. The control plasmid was the same as the miR-24 construct except the pre-miR-24 sequence was replaced with a non-related sequence (5′-atacagtactgtgataactgaagtttttggaaaagctttagttattaa-3′), serving as a mock control. In order to analyze the function of miR-24, the two constructs were transfected into 4T1 cells. After cell sorting through monitoring GFP signal by FACS (fluorescence activated cell sorter), transfected stable cells were obtained and used in this study.

A luciferase reporter vector (pMir-Report; Ambion) was used to generate the luciferase constructs. A fragment of the 3′-untranslated region (3′UTR) of human PTPN9 or PTPRF was cloned by RT-PCR. Two primers, huPTPN9-SacI (5′-cccgagctcctctcctacgaacctcctacc-3′) and huPTPN9-MluI (5′-cccacgcgtggcaaatgctcatcaaacctcccc-3′), were synthesized to clone the 3′UTR fragment of PTPN9, while the 3′UTR fragment of PTPRF was cloned with the primers huPTPRF-SacI (5′-cccgagctcctaccgctcccctctcctccgccac-3′) and huPTPRF-MluI (5′-cccacgcgtggattcgcgaaaggaaaaatcagag-3′). The PCR products were digested with SacI and MluI and the fragments were inserted into a SacI- and MluI-opened pMir-Report Luciferase vector to obtain the luciferase constructs, Luc-PTPN9 and Luc-PTPRF, respectively. Mutations of the miR-24 binding sites were also created, which was derived by PCR approach, using two primers huPTPN9-SacI-mut (5′cccgagctcctctcctacgaacctcctacctgttggccagccttccttaaactaccctggacaccggactgcct-3′) and huPTPN9-MluI to generate a mutant construct Luc-PTPN9-mut, and two primers huPTPRF-SacI-mut (5′-cccgagctc ctaccgctcccctctcctccgccacccccgccgtggggctccggaggggacccagctcctgactgcca-3′) and huPTPRF-MluI to generate Luc-PTPRF-mut. To serve as a negative control, a non-related sequence was amplified from the coding sequence of the chicken versican G3 domain using two primers, chver10051SpeI and chver10350SacI (Shan et al., 2009). We do not expect any endogenous miRNA to bind to this fragment as it is in a coding region.

To generate PTPN9 expression construct, we amplified the PTPN9 coding sequence (access number NM_002833.2, 1782 base pairs) by PCR with pfu polymerase and two primers (Ptpn9-For, 5′-ccccgaattcgagcccgcgaccgcgccc-3′ and Ptpn9-Rev, 5′-ccccgtcgacttactgactctccacggc-3′). The PCR fragment was cloned at EcoRI and XhoI sites in a Myc-tagged pCDNA3.1 expression vector. After confirmation of the correct sequence, the 1782-bp HindIII and XbaI fragment was released from this plasmid and treated with Klenow enzyme to produce blunt end and sub-cloned at SnaBI of pBabe-puro retroviral vector.

To generate PTPRF expression construct, we performed the amplification with two primers (Ptprf-For, 5′-ccccggtaccatggcccctgagccagcc-3′ and Ptprf-Rev, 5′-ccccgcggccgcccgttgcatagtggtcaaa-3′). This 5694-bp fragment was cloned at BamHI and SacII sites in pBS SK vector. After confirmation of the correct sequence, the fragment was obtained with the restriction enzymes KpnI and SacII from pBS vector and sub-cloned at KpnI and SacII sites in pCDNA4/myc-HisB. An EcoRV and AgeI fragment from the PTPRF-pCDNA4/myc-HisB plasmid was released and treated with Klenow enzyme to produce blunt end and sub-cloned at the SnaBI of pBabe-puro retroviral vector.

The plasmids pBabe-PTPN9 and pBabe-PTPRF were transfected into Hek293T cells to produce retrovirus. Harvested retroviruses in the cultural media were mixed with 6 µg/ml of polybrene (Sigma) and added into culture dishes to infect cells. 24 hours after infection, cells were trypsinized and diluted into 96-well plates and selected with 2 µg/ml of Puromycin (Invitrogen). Individual clones were screened by western blotting with anti-Myc antibody (Sigma).

The method was as described previously (Deng et al., 2013).

The method was as described previously (Yang et al., 2012).

Harvested mock- and miR-24-transfected cells were mechanically and enzymatically dissociated in 0.05% Trypsin-EDTA at 37°C for 10 minutes, followed by resuspension in serum-free DMEM/F12 medium containing 10 ng/ml bFGF, 20 ng/ml EGF, L-glutamine (1∶400) and B27 supplement (1∶50). Single cell suspensions were plated in 60-mm Petri dishes at 1000 cells/well. After culturing for additional 3 days, sphere cells were collected by gentle centrifugation, dissociated to single cells in 0.05% trypsin-EDTA and then cultured to generate mammospheres of the next generation.

We performed two different cell migration assays. In the wound-healing experiment, the method was performed as described (Rutnam and Yang, 2012b). The other migration assay was performed with the modified chemotactic Boyden chamber motility assays, which was performed using 24-well cell culture plates and a 3 µm cell culture insert. The lower chamber of the Transwell was filled with 600 µl of 10% fetal bovine serum (FBS)/DMEM and incubated at 37°C for 2 hours. Mock- and miR-24-transfected cells (1×10⁵) in 100 µl serum-free DMEM were gently injected onto each filter insert (upper chamber) and then incubated at 37°C for 4 hours. The filter inserts were removed from the chambers, fixed with methanol for 5 minutes and stained with Harris’ Hemotoxylin for 20 minutes. Samples were subsequently washed, dried and mounted onto slides for analysis using a light microscope. Migrating cells were stained blue and were counted in six fields of views/membrane.

Cell invasion assay was performed with the modified chemotactic Boyden chamber invasion assays using 24-well cell culture plates and an 8 µm cell culture insert as described previously (Rutnam and Yang, 2012b).

The method was described previously (Deng et al., 2013).

Western blotting was performed as described previously (LaPierre et al., 2007).

Mock- and miR-24-transfected cells were washed and resuspended in cold PBS and incubated in ice-cold 70% ethanol for 3 hours. The cells were then centrifuged at 1500 rpm for 10 minutes and resuspended in propidium iodide (PI) master mix (40 mg/ml PI and 100 mg/ml RNase in PBS) at a density of 5×10⁵ cells/ml and incubated at 37°C for 30 minutes before analysis with flow cytometry.

Luciferase activity assay was performed using a dual-luciferase reporter system developed by Promega (E1960). This method was performed as described previously (Jeyapalan et al., 2011).

This was performed as previously described (Lee et al., 2009). The primers used as real-time PCR controls were human-U6RNAf and human-U6RNAr.

The method was described recently except that each group contained 10 mice (Fang et al., 2012b). At necroscopy, primary tumors, stromal tissues, lungs, liver, spine were dissected. Half of the organs were fixed in 10% formalin and the other half were frozen in liquid nitrogen for subsequent analysis. The animal experiments were performed according to the guideline approved by the Animal Care Committee at Sunnybrook Research Institute.

The mock- and miR-24-tansfected MT-1 cells were cultured in 10% FBS/RPMI 1640 media at 37°C with 5% CO₂. At 70% to 80% subconfluency, the cells were given fresh 10% FBS/RPMI 1640 media 24 hours before being introduced into the mice. Cell viability was determined and cells were suspended with greater than 95% viability without cell clumping. Four-week-old nude mice were injected in the tail vein with the above transfected cells (2×10⁵ cells in 150 µl 10% FBS/RPMI 1640 medium). Each group had 15 randomly selected mice. All mice were sacrificed 6 weeks after injection. At necroscopy, lungs, liver, spines were dissected. Half the organs were fixed in 10% formalin and the other half were frozen in liquid nitrogen for subsequent analysis.

Mouse lung tissues were homogenized and the genomic DNAs were isolated with High Pure PCR Template Preparation kit according to the manufacturer’s instructions. In order to estimate tumor burden, we extracted three samples from the above organs of each animal. Tumor burden for each individual tissue was measured using real-time PCR. Primers used were as follows: CMV forward (5′-gtcatcgctattaccatggtgatgcgg-3′) and CMV reverse (5′-agctctgcttatatagacctcccaccg-3′) for genotyping; β-actin forward (5′-ccggcatgtgcaaagccggcttcg-3′) and β–actin reverse (5′-ctcattgtagaaggtgtggtgcc-3′) for loading control.

Primary tumors or lungs were freshly excised and fixed in 10% formalin overnight, immersed in 70% ethanol, embedded in paraffin, and sectioned. The sections were subjected to Hematoxylin and Eosin (H&E) staining and immunohistochemistry and were de-paraffinized with xylene and ethanol and then boiled in a pressure cooker. After washing with Tris-buffered saline (TBS) containing 0.025% Triton X-100, the sections were blocked with 10% goat serum and incubated with primary antibody in TBS containing 10% goat serum albumin overnight. The sections were washed and labeled with biotinylated secondary antibody, followed by avidin-conjugated horseradish peroxidase provided by the Vectastain ABC kit (Vector, PK-4000). The slides were subsequently stained with DAB followed by Mayer’s Hematoxylin for counter staining and slide mounting.

For multiple antigen staining, an avidin/biotin blocking kit (SP-2001) was used to prevent the interaction of the second set of labeling reagents with the first set of labeling reagents. Briefly, the sections were incubated with Avidin solution for 15 minutes and then incubated in Biotin Solution for 15 minutes. After washing with Tris-buffered saline (TBS) containing 0.025% Triton X-100, the sections were blocked with 10% goat serum and incubated with primary antibody to second antigen and followed by each step as the first staining.

Breast specimens were collected at the time of mastectomy from previously untreated patients, fixed in 10% formalin overnight, immersed in 70% ethanol, embedded in paraffin and sectioned. Corresponding normal tissue was obtained from the same mastectomy specimen as the tumor, from a different quadrant of the breast. The tissue blocks were confirmed with examination under light microscopy by a pathologist and confirmed to be benign or malignant based on histology of the corresponding H&E stained slide. Sections were subjected to immuno-histochemistry probed with antibodies against EGFR, pEGFR, PTPN9 and PTPRF. The work was conducted following a protocol approved by the Institutional Review Board at the University of Iowa Carver College of Medicine and a protocol approved by the Affiliated People’s Hospital of Jiangsu University.

Experiments using freshly frozen clinical specimens were also conducted independently in the Affiliated People’s Hospital of Jiangsu University according to a protocol approved by the Clinical Research Ethics Committee at the Hospital. In brief, small RNAs were isolated from formalin-fixed, paraffin-embedded tissue blocks, using the Total Nucleic Acid Isolation Kit (Ambion) or from freshly frozen tissues. Quantification of miRNAs was performed with real-time PCR.

The results (means values ± s.d.) of all the experiments were subjected to statistical analysis using Student’s t-tests. The level of significance was set at P<0.01.