Protein tyrosine phosphatase PTPRJ is negatively regulated by microRNA-328


R. Iuliano, Dipartimento di Medicina Sperimentale e Clinica, Università‘Magna Graecia’, viale Europa, 88100 Catanzaro, Italy
Fax: +39 0961 3694090
Tel: +39 0961 3695182
F. Paduano, Dipartimento di Medicina Sperimentale e Clinica, Università‘Magna Graecia’, viale Europa, 88100 Catanzaro, Italy
Fax: +39 0961 3694090
Tel: +39 0961 3694106


Expression of PTPRJ, which is a ubiquitous receptor-type protein tyrosine phosphatase, is significantly reduced in a vast majority of human epithelial cancers and cancer cell lines (i.e. colon, lung, thyroid, mammary and pancreatic tumours). A possible role for microRNAs (miRNAs) in the negative regulation of PTPRJ expression has never been investigated. In this study, we show that overexpression of microRNA-328 (miR-328) decreases PTPRJ expression in HeLa and SKBr3 cells. Further investigations demonstrate that miR-328 acts directly on the 3′UTR of PTPRJ, resulting in reduced mRNA levels. Luciferase assay and site-specific mutagenesis were used to identify a functional miRNA response element in the 3′UTR of PTPRJ. Expression of miR-328 significantly enhances cell proliferation in HeLa and SKBr3 cells, similar to the effects of downregulation of PTPRJ with small interfering RNA. Additionally, in HeLa cells, the proliferative effect of miR-328 was not observed when PTPRJ was silenced with small interfering RNA; conversely, restoration of PTPRJ expression in miR-328-overexpressing cells abolished the proliferative activity of miR-328. In conclusion, we report the identification of miR-328 as an important player in the regulation of PTPRJ expression, and we propose that the interaction of miR-328 with PTPRJ is responsible for miR-328-dependent increase of epithelial cell proliferation.


carboxyfluorescein succinimidyl ester


Homo sapiens microRNA


loss of heterozygosity






microRNA response element


3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide


small interfering RNA


The tyrosine phosphatase PTPRJ (also called DEP-1, HPTPη or CD148) is a widely expressed transmembrane protein that plays a role in multiple signalling pathways [1]. Substrates of PTPRJ include proteins that belong to the family of tyrosine kinase receptors (EGFR, HGFR, PDGFR, RET and VEGFR-2) [2–6] as well as those involved in cell adhesion (c-Src, p120-catenin and ZO-1) [7–9]. Consequently, in several epithelium-derived cell lines, PTPRJ inhibits cell proliferation and promotes cell adhesion [10–13]. PTPRJ has been proposed to be a tumour suppressor owing to downregulation of its expression in cancer cell lines [10,12–14] and inhibition of tumorous growth upon restoration of normal expression levels [12–14]. Although expression of PTPRJ decreases significantly during the progression of human tumours [15,16], little is known about the mechanisms involved in the process. Loss of heterozygosity (LOH) was detected at the PTPRJ locus (11p11.2) in lung, breast, colon and thyroid carcinomas [17,18]; however, LOH and promoter methylation of PTPRJ [19,20] can only partly explain the association between tumour progression and reduced PTPRJ expression.

MicroRNAs (miRNAs) are tiny non-coding RNAs that downregulate gene expression by binding to the 3′UTRs of their targets resulting in the inhibition of protein translation or mRNA degradation [21]. Given their varied effects on gene regulation, miRNAs have been implicated in several diseases, including cancer. MiRNA signatures have been used in the classification of human tumours and in cancer prognosis predictions [22–24]. MiRNA expression levels are not only useful as biomarkers of disease progression, but it is now clear that they also play a causative role in cancer progression. Genetic and epigenetic variations of miRNAs have been detected in tumours [25,26], and polymorphisms in miRNA coding regions have been associated with cancer susceptibility [27]. Probing their mechanistic role in cancer, it appears that miRNAs can target either onco-genes or tumour suppressor genes, and act by blocking or favouring tumour progression as a consequence. The balance in targets between oncogenes and tumour suppressors by a certain miRNA defines its role in cancer development [24]. Therefore, it is crucial to identify cancer-related genes targeted by miRNAs.

In this study, we report that microRNA-328 (miR-328) targets the 3′UTR of PTPRJ resulting in its decreased expression. Furthermore, PTPRJ downregulation could explain the proliferative effects observed in HeLa cells with miR-328 overexpression. Therefore, we propose that miR-328 is a negative regulator of PTPRJ expression, with a possible role in the development of human tumours.


The 3′UTR of PTPRJ was analysed by the bioinformatics algorithm mirtar [28] to identify possible interacting miRNAs. Several miRNAs were selected as likely interactors and were considered for further bioinformatic analyses with other computational algorithms (findtar and rnahybrid) [29,30]. Among the selected miRNAs, for miR-328, two miRNA response elements (MREs) were found with mirtar; four with findtar; and five with rnahybrid in the 3′UTR of PTPRJ (Table 1). For further analyses on MREs, we considered the sites picked with rnahybrid.

Table 1. Computational analysis with mirtar, findtar and rnahybrid of microRNA recognition elements (MREs) for miR-328 in the 3′UTR of PTPRJ mRNA. MREs identified by rnahybrid are indicated as MRE_1 to MRE_5 in the text.Thumbnail image of

In order to evaluate the reliability of computer prediction, we performed the transfection of synthetic miR-328 in two different cell lines (HeLa and SKBr3), and we then detected PTPRJ by western blot. We found a significant decrease of PTPRJ protein levels in miR-328 transfected cells compared with the control (Fig. 1A). As a positive control, a PTPRJ-specific small interfering RNA (siRNA) was used. To evaluate whether miR-328 affects PTPRJ expression by decreasing mRNA levels, we quantified, by qRT-PCR, mRNA levels of PTPRJ and its alternatively spliced shorter isoform, encoding an mRNA with a different 3′UTR (Fig. 1B). We found that miR-328 decreased mRNA levels of PTPRJ but not of the shorter isoform, indicating that miR-328 probably acts on the 3′UTR of PTPRJ mRNA. In addition, the PTPRJ siRNA oligonucleotide that binds a region on exon 6 shared by the two PTPRJ isoforms decreased mRNA levels of both isoforms (Fig. 1C). Thus, we demonstrated that miR-328 downregulates endogenous PTPRJ at both the mRNA and the protein levels.

Figure 1.

 PTPRJ expression is downregulated by miR-328. (A) Western blot analysis for PTPRJ detection in protein extracts from cells transfected transiently for 24 h with siRNA-PTPRJ, Lipofectamine or miR-328. Expression is normalized on γ-tubulin. (B) Schematic representation of the PTPRJ pre-mRNA (long and short forms) and location of miRNA and siRNA target sites on PTPRJ mRNA (Genbank accession no. NM_002843.3). (C) qRT-PCR with specific primers amplifying the two isoforms of the PTPRJ gene from cell lines transfected transiently for 24 h with Lipofectamine, siRNA-PTPRJ or miR-328. Values were normalized to HPRT RNA levels. *< 0.01.

We hypothesized that miR-328 interacts directly with the 3′UTR of PTPRJ mRNA to suppress PTPRJ expression. To test this hypothesis, the ability of miR-328 to regulate the 3′UTR of PTPRJ was evaluated by luciferase reporter assays. The region starting from nucleotide +6259 to nucleotide +7761 of the PTPRJ sequence (NM_002843.3) was cloned downstream of a reporter luciferase gene (pmirGLO 3′-WT) (Fig. 2A), as described in Materials and methods. Co-transfection of synthetic miR-328 and pmirGLO 3′-WT reduced the luciferase activity significantly in comparison with Lipofectamine and miR-342 (Fig. 2B), an miRNA that did not influence endogenous PTPRJ expression in HeLa cells (data not shown). However, co-transfection of miR-328 in cells transfected with pmirGLO empty vector did not influence luciferase activity significantly. These results prompted us to study recognition elements of miR-328 in the PTPRJ 3′UTR. Five possible MREs (MRE_1 to MRE_5, Fig. 2A) containing an miR-328 seeding region were predicted by computational analyses, as reported before (Table 1). To map interaction sites of miR-328 in the 3′UTR of PTPRJ, we generated a deletion mutant that includes the 3′UTR (pmirGLO 3′Δ) downstream of the SacI site and contains just two possible miR-328 binding sites (MRE_4 and MRE_5). Luciferase activity of that mutant was significantly affected by the co-transfection with miR-328, and the amount of luciferase reduction was not lower than that detected by the pmirGLO 3′UTR. Therefore, deletion of nucleotides +6259 to +7082 of the human PTPRJ 3′UTR containing three MREs (MRE_1 to MRE_3) did not decrease susceptibility to miR-328 inhibition (Fig. 2C), indicating that MRE_1 to MRE_3 were not functional. To better characterize the interaction, we generated two site-specific mutations in MRE_4 and MRE_5 (3′Δ-4 mut and 3′Δ-5 mut) affecting the possible remaining miR-328 binding sites in the 3′UTR of PTPRJ mRNA (Fig. 2D). When the MRE_5 site is mutated (pmirGLO 3′Δ-5 mut), only a slight decrease in the luciferase activity was observed. Mutation of the MRE_4 site (pmirGLO 3′Δ-4 mut) significantly abolished the repression of luciferase activity operated by miR-328. The same results were obtained with the double mutant 3′UTR (pmirGLO 3′Δ-4/5 mut), indicating that MRE_4 in position from +7174 to +7203 plays a key role in the miR-328 mediated repression (Fig. 2E). Taken together, these data indicate that miR-328 directly regulates PTPRJ expression mainly through interaction with the predicted MRE_4 in the 3′UTR of PTPRJ.

Figure 2.

PTPRJ is a direct target of miR-328. (A) Reporter plasmid containing five potential miR-328 binding sites within the 3′UTR of human PTPRJ gene (MRE_1 to MRE_5). (B) Luciferase assay of HeLa cells co-transfected with PTPRJ WT 3′UTR reporter plasmid (pmirGLO 3′-WT) or empty vector (pmirGLO) and synthetic miR-328 or miR-342. (C) Luciferase assay of HeLa cells co-transfected with 3′UTR reporter plasmid (pmirGLO 3′-WT) or deleted 3′UTR (pmirGLO 3′Δ) and synthetic miR-328. (D) Base pairing for comparison between mature miR-328 and wild-type (PTPRJ WT MRE) or mutant (PTPRJ mut MRE) putative target sites (MRE_4 and MRE_5) in the 3′UTR of PTPRJ mRNA. Mutations were generated on the potential target sequence of PTPRJ (bold letters with underlining) by site-directed mutagenesis. (E) Reporter repression of miR-328 in transfected HeLa cells with vector is indicated. Luciferase values were normalized to the double mutant (pmirGLO 3′Δ-4/5 mut). Significant differences are indicated by an asterisk, P < 0.01. All the experiments were performed in quadruplicate and repeated twice.

Proliferation of HeLa and SKBr3 cells is sensitive to PTPRJ levels modulated by vector transfection [10,31]. Therefore, to functionally link the PTPRJ/miR-328 interaction to cell proliferation, we tested the effects of miR-328 overexpression on HeLa cells by using the 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay. We generated an expression vector containing the genomic region flanking miR-328 (pcDNA3.1-miR-328) (Materials and methods). Upon transient transfection in HeLa cells, significant expression of miR-328 was detected by qRT-PCR (data not shown). Transfection with pcDNA3.1-miR-328 enhanced proliferation of HeLa cells in a dose-dependent manner (Fig. 3A). We also found that knockdown of PTPRJ expression with siRNA induced a significant increase in cell proliferation (Fig. 3B), similar to the effects of miR-328 overexpression. Importantly, miR-328 transfection in siRNA-treated cells did not further increase cell proliferation (Fig. 3B), indicating that the effect of miR-328 on proliferation was due, at least in part, to downregulation of PTPRJ expression. In addition, PTPRJ overexpression (Fig. S1) rescued the cell proliferation effect observed in HeLa cells treated with the miR-328 construct (Fig. 3C). To confirm the effect of miR-328 on cell proliferation, an in vitro cell proliferation assay was performed in HeLa cells by using the fluorescent dye carboxyfluorescein succinimidyl ester (CFSE), which labels the intracellular protein content. With each cell division, intracellular fluorescence decreases with increasing cell number, the intensity of which can be assayed by flow cytometry. Proliferation in HeLa cells increased with both pcDNA3.1-miR-328 and pcDNA3.1 + PTPRJ-siRNA compared with empty vector, confirming that miR-328-dependent induction of cell proliferation can also be achieved by repressing PTPRJ levels (Fig. 3D,E). The proliferative effect of miR-328 was also evaluated in SKBr3 cells. MiR-328 increased cell proliferation in transiently transfected SKBr3 cells as demonstrated by MTT and CFSE assays (Fig. S2).

Figure 3.

 MiR-328 increases proliferation in HeLa cells by repressing PTPRJ. (A) Proliferation of HeLa cells transfected with vectors is indicated. Cell proliferation was measured 48 h after seeding with the MTT assay and values were normalized to the empty vector. *P < 0.05 compared with pcDNA3.1 (250 ng), analysed by ANOVA followed by Dunnett’s test. (B) Proliferation of HeLa cells co-transfected with pcDNA3.1 (250 ng) + siRNA-PTPRJ (100 nm), pcDNA3.1-miR-328 (250 ng) and pcDNA3.1-miR-328 (250 ng) + siRNA-PTPRJ (100 nm). Cell proliferation was measured after 48 h by MTT assay and values were normalized to the Lipofectamine. *P < 0.05 compared with Lipofectamine, analysed by ANOVA followed by Dunnett’s test. (C) Proliferation of HeLa cells co-transfected with synthetic miR-328 (100 nm) + pCEFL (250 ng) and synthetic miR-328 (100 nm) + pCEFL-PTPRJ (250 ng). Cell proliferation was measured after 48 h by MTT assay, and values were normalized to the Lipofectamine. *P < 0.05 compared with Lipofectamine, analysed by ANOVA followed by Dunnett’s test. (D) Representative cell cycle progression of HeLa cells transiently transfected with pcDNA3.1, pcDNA3.1-miR-328 and pcDNA3.1 + siRNA-PTPRJ using CFSE. Cells were analysed at 0 and 24 h by flow cytometry. (E) Geometric mean intensities of CFSE staining in HeLa cells transiently transfected with pcDNA3.1, pcDNA3.1-miR-328 and pcDNA3.1 + siRNA-PTPRJ at 24 h. P values were generated by ANOVA followed by Bonferroni’s test for multiple comparisons (significant differences are indicated by an asterisk, P < 0.05).

To further investigate the link between PTPRJ and miR-328, we generated stable HeLa cell lines overexpressing miR-328. Expression of miR-328 was quantified by qRT-PCR and two clones, miR-328 clone-1 and miR-328 clone-2, showed 50- and 80-fold increases, respectively, in miR-328 expression compared with empty vector control (Fig. 4A). Western blot analysis of the HeLa miR-328 clones confirmed PTPRJ downregulation (Fig. 4B). Additionally, flow cytometry showed a reduction in membrane PTPRJ expression in the miR-328 stable clones compared with empty vector control (Fig. 4C). Finally, we aimed to elucidate the effects of miR-328 overexpression on proliferation rates in HeLa cells. This was performed by using the MTT assay, and the miR-328 stable clones showed an increase in cell proliferation compared with empty vector control (Fig. 4D). This was also confirmed by the CFSE assay: clearly, the two miR-328 clones displayed an increase in cell proliferation compared with empty vector control (Fig. 4E,F). PTPRJ is a negative regulator of cell migration [11]. To investigate whether miR-328 affects HeLa cell migration, we used the wound-healing assay. We found that cell migration was increased in both miR-328 stable clones compared with empty vector control (Fig. S3).

Figure 4.

 HeLa clones stably expressing miR-328 decrease PTPRJ expression and induce cell proliferation. (A) Relative miR-328 expression in two HeLa stable clones (miR-328 clone-1 and miR-328 clone-2) quantified by qRT-PCR. Values were normalized to pcDNA3.1 vector. *< 0.01. (B) Western blot analysis of PTPRJ expression in HeLa stable clones. Expression is normalized to γ-tubulin. Relative levels of PTPRJ were quantified by quantity one software (BioRad Laboratories). (C) Fluorescence-activated cell sorting analysis of membrane-bound PTPRJ expression in HeLa clones stably expressing miR-328 or empty vector. Grey peaks represent unlabelled cells; open peaks represent cells labelled with PTPRJ antibody. (D) Proliferation of HeLa stable clones measured after 24 and 48 h with the MTT assay. Values were normalized to the empty vector. *P < 0.05 compared with pcDNA3.1 at 24 h, analysed by ANOVA followed by Dunnett’s test. (E) Representative cell cycle progression of HeLa stable clones using CFSE. Cells were analysed at 0, 24 and 48 h by flow cytometry. (F) Geometric mean intensities of CFSE staining in HeLa stable clones at 24 and 48 h. *P < 0.05 compared with pcDNA3.1, analysed by ANOVA followed by Bonferroni’s test.


Several studies have addressed the function of PTPRJ in the regulation of cellular pathways and a number of PTPRJ-interacting proteins or substrates have been identified [1], underlining the role of PTPRJ as a negative regulator of cell proliferation in epithelial [10,12–14] and endothelial [32] cells. PTPRJ acts by downregulating receptor-type tyrosine kinase activity [2–6] or by inhibiting the pro-proliferative PI3K and Ras-dependent pathways [33,34]. To clarify the role of PTPRJ in vivo, knockout mice were generated resulting in viable mice that do not develop tumours spontaneously [35]. However, significant downregulation of PTPRJ expression has been observed in tumours [15,16]. Thus, it is important to study the mechanisms involved in the regulation of PTPRJ expression that is currently not very well understood. In general, regulation of the expression of protein tyrosine phosphatases is poorly characterized [36].

Post-transcriptional regulation of PTPRJ has been studied by Karagyozov et al. [37], who found a novel mechanism of translational regulation that was based on the use of different translational start sites. Our study provides an independent mechanism of control of PTPRJ expression which involves the 3′UTR. We used three independent bioinformatics algorithms to identify MREs in the 3′UTR of PTPRJ. Interestingly, MRE_4, which was the only MRE to be detected by all three algorithms, appeared to interact with miR-328.

Recently, a novel mechanism of action has been attributed to miR-328. In fact, miR-328 acts through a decoy activity that interferes with the function of regulatory proteins [38]. However, miR-328 is also able to repress gene expression through the typical miRNA mechanism of acting on the 3′UTR of target genes [38], as shown here.

In our study, miR-328 increased proliferation of HeLa and SKBr3 cells by means of a mechanism dependent, at least in part, on the downregulation of PTPRJ expression. The number of possible miR-328 gene targets in the cell lines analysed probably influenced the extent of PTPRJ downregulation. However, effects on cell proliferation would not necessarily have to correlate with the extent of PTPRJ downregulation as different cell lines could have varied sensitivities to expression of PTPRJ, or other targets of miR-328.

The list of validated targets of miR-328 includes CD44, ABCG2 and PIM-1 [38–40]. CD44 is involved in cell adhesion and could act synergistically with PTPRJ. Given that CD44 and PIM-1 are considered pro-oncogenic proteins, this could appear to contrast with the tumour suppressor role of PTPRJ. However, the oncogenic activity of PIM-1 is mainly restricted to haematological malignancies [41] in which, it should be emphasized, the role of PTPRJ is still unclear. Indeed, CD148/PTPRJ has been proposed to be a valuable marker of mantle cell lymphomas in an unbiased proteomic study [42].

Loss of expression of miR-328 is strongly correlated with blast crisis in patients afflicted with chronic myelogenous leukaemia [38]. In glioblastoma multiforme, miR-328 levels are lower compared with normal brain tissues [43]. Conversely, miR-328 levels are upregulated in patients with brain metastasis of lung cancer [44]. Expression levels of miR-328 are significantly higher in blood samples of melanoma patients compared with healthy individuals [45].

Studies on miR-328 discussed thus far reflect the well understood feature of miRNAs: they can act as onco-miRs (favouring tumour progression) or oncosuppressor miRs (blocking tumour progression) depending on the cellular context, as exemplified by miR-221 and miR-222 that behave as tumour suppressors in haematological malignancies and as oncogenes in other tumours [46]. In fact, the role of an miRNA in a pathological context is also dependent on the relative expression levels of its targets in the tissue under examination [47]. However, in epithelium-derived cells, miR-328 increases cell proliferation (our results) and cell migration [44], and this is in line with the opposing role played by PTPRJ in these processes [10–15,48,49].

In conclusion, we have identified a novel mechanism of regulation of PTPRJ expression in which the 3′UTR of PTPRJ mRNA is targeted by miR-328. This interaction is important for proliferation of epithelium-derived cells. It would be of interest to investigate in future studies the correlation between expression of PTPRJ and miR-328 in human tumours.

Materials and methods

Bioinformatic analysis

The algorithms used were mirtar, rnahybrid and findtar. We first screened the possible miRNAs interacting with the 3′UTR of PTPRJ by using mirtar [28], a bioinformatics algorithm that allows a selection of potential interacting miRNAs with a given 3′UTR sequence. Every miRNA candidate resulting from that analysis was further investigated with findtar [29] and rnahybrid [30] to determine the best possible miRNA candidate for the interaction with the 3′UTR of PTPRJ. The threshold value of minimum free energy for the identification of MREs located on PTPRJ 3′UTR was set at the default value (−16 kcal·mol−1) in the screening with mirtar and at −25 kcal·mol−1 in the analyses with findtar and rnahybrid.

Cells and transfections

HeLa (human cervix adenocarcinoma) and SKBr3 (human breast adenocarcinoma) cell lines were maintained in culture at 37 °C and 5% CO2 in DMEM supplemented with 10% fetal bovine serum, 100 U penicillin and 100 μg·mL−1 streptomycin sulfate. Transfections were made with Lipofectamine® 2000, following the manufacturer’s instructions (Invitrogen, Carlsbad, CA, USA). For western blot and qRT-PCR experiments, 4 × 105 cells were seeded in six-well plates and transfected with 50 pmol Pre-miRTM (Ambion, Life Technologies, Paisley, UK) or 100 pmol siRNA. For luciferase assays, 2 × 05 cells were seeded in 24-well plates and co-transfected with 10 pmol·mL−1 Pre-miRTM (Ambion) and 25 ng·mL−1 reporter plasmid (pmirGLO vector; Promega, Madison, WI, USA).

Western blot analysis

Cells were harvested after 48 h, washed with ice-cold NaCl/Pi and lysed in a buffer containing 50 mm Tris/HCl pH 7.5, 1% Nonidet P40, 150 mm NaCl, 25 mm NaF, 1 mm Na3VO4 and a complete mixture of protease inhibitors (Roche, Basel, Switzerland). Lysates were clarified by centrifugation at 10 000 g for 15 min at 4 °C. Protein concentration was determined by a modified Bradford assay (BioRad, Hercules, CA, USA). Fifty micrograms of proteins were subjected to electrophoresis carried out in an 8%–12% gradient gel (Invitrogen) and then transferred to nitrocellulose membranes (Hybond-C; Amersham Biosciences, Piscataway, NJ, USA). The membranes were blocked with 5% non-fat dry milk and then probed for 2 h with the appropriate primary antibodies. After incubation with specific horseradish peroxidase conjugated secondary antibodies, protein bands were revealed with an enhanced chemiluminescence system (Santa Cruz Biotechnology, Santa Cruz, CA, USA). Primary antibodies used were anti-human PTPRJ and anti-tubulin purchased from R&D Systems (Minneapolis, MN, USA) and Santa Cruz Biotechnology, respectively.

Quantitative real-time PCR

RNA extraction was performed with miRNeasy Mini Kit™ (Qiagen, Valencia, CA, USA), following the manufacturer’s instructions, and total RNA was quantified with a spectrophotometer. Total RNA samples (250 ng) were subjected to the reaction of reverse-transcription using the High Capacity RNA-to-cDNA Kit (Applied Biosystems, Foster City, CA, USA), following the manufacturer’s instructions. Five hundred nanolitres of cDNAs were amplified by real-time PCR with Promega SYBR green kit and 5 pmol of primers in a total volume of 25 μL.

The primers used were as follows: 5′-GTATTATCATTGGTGGCTTGTTC-3′ (forward for long and short form of PTPRJ), 5′-CATCTCCGTGGTGGTGAC-3′ (reverse for long form of PTPRJ) and 5′-AGGCAGGTGTTCAAATCATCC-3′ (reverse for short form of PTPRJ). Specific oligonucleotides used for hypoxanthine phosphoribosyltransferase (normalization control) amplification were reported by Vandesompele et al. [50].

Real-time PCR reactions were carried out in a BioRad iQ™5 apparatus with the following conditions: initial denaturation step at 95 °C for 3 min, followed by 40 cycles of 10 s at 95 °C and 1 min at 57 °C. Specificity of PCR products was checked by melting curve analysis and gel electrophoresis. Efficiencies of real-time PCRs were calculated by constructing curves with cycle threshold (Ct) obtained from amplifications of serial tenfold dilutions of a cDNA sample, resulting in higher than 90%. For miR-328 quantification, 50 ng of total RNA was subjected to reverse-transcription with miRCURY LNA™ Universal cDNA synthesis kit (Exiqon, Vedbaek, Denmark), following the manufacturer’s instructions. One hundredth of the cDNA was amplified by real-time PCR with SYBR® Green master mix 2X (Exiqon) and miRCURY LNA™ hsa-miR-328 specific PCR primer set (Exiqon) or miRCURY LNA™ U6 snRNA specific PCR primer set (Exiqon), following the manufacturer’s instructions. Reactions were performed in a total volume of 20 μL in a BioRad IQ™5 apparatus.

Plasmids and mutagenesis

The 3′UTR of PTPRJ containing the predicted sites for miR-328/PTPRJ interaction (the region from +6259 to +7761 of PTPRJ) was amplified by PCR using genomic DNA as the template, Pfu polymerase and the following primers: forward 5′-CATGTTTAAACCTTCTCAAATGGAAATTGCA-3′ and reverse 5′- GCCCTCGAGCAACCAACAGCAATACTCTGTA-3′. The amplified product was purified and then cloned in pmirGLO vector (Promega) at PmeI and XhoI restriction sites, downstream of the firefly luciferase reporter gene.

A deletion mutant (Delta) was generated by exciding from pmirGLO-PTPRJ the PmeI-SacI fragment and re-ligating the plasmid after blunting of the SacI digested end. Site-specific mutants of Delta plasmid were generated by PCR mutagenesis. Primers used were forward PmeI, 5′-CATGTTTAAACCTCAAATTGGAAGAGTCC-3′; reverse XhoI, 5′- GCCCTCGAGCAACCAACAGCAATACTCT-GTA-3′; forward mut7440, 5′-GAGGGGCGGAAATAAG-CTATGCAGC-3′; reverse mut7440, 5′-GCTGCATAGCTTATTTCCGCCCCTC-3′; forward mut7174: 5′-GGTGTCAATGAAAAAACAAAAGACTGGTG-3′; reverse mut7174, 5′-CACCAGTCTTTTGTTTTTTCATTGACACC-3′.

PCR amplification was performed using Pfu polymerase (Promega) and pmirGLO-Delta mutant as the template. Mutants were all cloned at PmeI and XhoI sites of pmirGLO. The 75 nucleotide precursor hairpin sequences of miR-328 and 100 nucleotides of genomic sequences flanking each side of the hairpin sequence were amplified by PCR using the following primers: forward, 5′-CACGAATTCCCTTGTCGAAGTCCTCCTGTGTAG-3′; reverse, 5′- CACCTCGAGGACTCCTTGCCCCATTCTCTGCG-3′.

To generate the miR-328 expression construct, the 467-nucleotide DNA fragment obtained was digested with EcoRI and XhoI and inserted into the EcoRI-XhoI sites of the pCDNA3.1 expression vector (Invitrogen). All plasmids were checked by sequence analysis.

Luciferase assay

HeLa transfected cells were lysed in a passive lysis buffer (Promega). One hundred microlitres of luciferase assay reagent 2 were added to 20 μL of lysate, and samples were analysed in a luminometer (firefly reading). After the addition of Stop and Glo reagent (Promega), the samples were again analysed in the luminometer (Renilla reading). Values were expressed as the ratio of firefly/Renilla readings.

Cell proliferation assay

HeLa cells were plated in 96-well plates at a density of 2 × 103 cells per well in 100 μL of medium and were transfected with siRNA and/or plasmid DNA. Each condition was plated in octuplicate. Cell proliferation rates were determined 48 h post-transfection with the MTT assay (Sigma-Aldrich, St Louis, MO, USA). Absorbance of the formazan dye that is formed was measured at a wavelength of 570 nm, using a Microplate Reader (BioRad); absorbance is proportional to the number of metabolically active cells in culture. For CFSE staining, cells were harvested and washed twice with NaCl/Pi, followed by labelling with 5 mm CFSE for 5 min at room temperature. Residual CFSE was removed by washing twice with NaCl/Pi, and cells were seeded in six-well plates containing growth medium. CFSE fluorescence was measured by fluorescence-activated cell sorting analysis, and data were analysed with flowjo software (Tree Star, Ashland, OR, USA).

Flow cytometry analysis

Cells were centrifuged at 400 g for 3 min, the supernatant was discarded and the cells were re-suspended in 500 μL of NaCl/Pi. Briefly, cells were blocked with 1% BSA in NaCl/Pi, washed once and incubated for 20 min with PTPRJ primary antibody (R&D), followed by incubation with phycoerythrin-conjugated secondary antibody. Samples were analysed with FACSCan instruments (Becton Dickinson, Franklin Lakes, NJ, USA). Data were analysed with flowjo software (Tree Star).

Generation of stable cell clones

HeLa cells were transfected using Lipofectamine 2000 (Invitrogen). Cells were plated into selection medium containing 0.4 mg·mL−1 G418 (Sigma Aldrich) 1 day post-transfection. After 25 days of selection, individual G418-resistant colonies were isolated, expanded and analysed for miR-328 expression levels by qRT-PCR.

Cell migration assay

Cell migration assay was performed using the scratch wound-healing method on HeLa cells that were 80–90% confluent in a six-well plate format. Cell migration was monitored by microscopy. Images acquired for each sample were then quantitatively analysed. For each image, distances between the points of the scratch were measured. By comparing the images from t = 0 to t = 12 h, which was the last time point, the distance of each scratch closure was obtained by measuring the distances using imagej software (NIH Image, Bethesda, MD, USA).

Statistical analyses

Results are expressed as the mean ± standard deviation of three separate experiments. The significance of differences between groups was evaluated by using Student’s t test or one-way analysis of variance (ANOVA), followed by Dunnett’s test/Bonferroni’s test for multiple comparisons. Analysis was conducted using the graphpad prism software (San Diego, CA, USA), and differences were considered significant if P < 0.05.


The present study was supported by a grant from Ministero dell’Istruzione, dell’Università e della Ricerca (MIUR) to CP and RI (No. 2007F7T537_003) and a grant from the Associazione Italiana Ricerca Cancro (AIRC) to FT. FP was a recipient of an FIRC (Fondazione Italiana Ricerca Cancro) fellowship.