The present study was approved by the Institutional Review Boards of The Affiliated Hospital of Xuzhou Medical College. Glioma tissue samples, obtained from patients undergoing resection for glioma tumors at the Department of Neurosurgery of The Affiliated Hospital of Xuzhou Medical College, were histologically confirmed by experienced pathologists based on the World Health Organization (WHO) grading system (28). Non-cancerous brain tissues were collected from patients undergoing decompression surgery for intracranial traumas.

All procedures performed in the present study involving human participants were in accordance with the ethical standards of the Institutional and/or National Research Committee, and with the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. Written informed consent was obtained from all individual participants or their legally authorized representatives prior to enrollment in the study.

Human glioma cell lines U251 and U87 (Shanghai Cell Bank, Type Culture Collection Committee, Chinese Academy of Sciences) were cultured in Dulbecco's modified Eagle's medium and F-12 (DMEM/F-12) (Invitrogen, Carlsbad, CA, USA) supplemented with 10% fetal bovine serum (FBS) (Sijiqing, Hangzhou, China), followed by incubation at 37̊C in a humidified incubator with 5% CO₂.

The mammalian expression plasmid encoding full-length human Nedd4-1 was a gift from Dr Xuejun Jiang (Memorial Sloan-Kettering Cancer Center, New York, USA). The shNedd4-1 plasmid was constructed by Shanghai GenePhama, and the human Nedd4-1 residues 217–549 was a gift from Dr Nile Brose (Max Planck Institute, Göttingen, Germany). The Myc-tagged Rap2a plasmids, including wild-type (WT-Rap2a), dominant-active (DA-Rap2a) and dominant-negative (DN-Rap2a), were gifts from Dr Kenichi Kariya (University of the Ryukyus, Okinawa, Japan). DA-Rap2a and DN-Rap2a were constructed by mutating Gly 12 to Val and Ser 17 to Asn, respectively. Hemagglutinin (HA)-tagged ubiquitin and its mutants including K63R-ubiquitin and K48 R-ubiquitin were gifts from Shanghai Institute of Neuroscience, Chinese Academy of Sciences. The bacterial expression plasmid for the glutathione S-transferase (GST)-fusion protein of Ras-binding region of RalGDS was cloned by reverse transcription polymerase chain reaction (RT-PCR) using a human brain cDNA library as a template. The following antibodies were commercial products: rabbit polyclonal antibodies to Nedd4-1 and Rap2a (Santa Cruz Biotechnology, Inc., Santa Cruz, CA, USA), mouse anti-TNIK (BD Biosciences, Franklin Lakes, NJ, USA), mouse monoclonal antibodies P4D1 and FK1 to ubiquitin (Santa Cruz Biotechnology, Inc.), rabbit monoclonal antibodies to ubiquitin Apu2 and Apu3, and mouse monoclonal (9E10) antibody to Myc (Millipore, Billerica, MA, USA).

Total RNA was extracted from U251 and U87 cells using TRIzol reagent (Tiangen, Beijing, China). The RT-PCR protocol was performed according to the manufacturer's instructions using reverse transcription reagents (Takara, Dalian, China). Primers (Invitrogen, Shanghai, China) for Nedd4-1 were as follows: sense, 5′-ACTTTATCCATTACCGACAG-3′ and antisense, 5′-TGGTGGCTTCATCTTCTC-3′, yielding a 320-bp product. Primers for the internal standard β-actin were: sense, 5′-GCGCGGCTACAGCTTCAC-3′ and antisense, 5′-GGGGCCGGACTCGTCATA-3′, yielding a 500-bp product. PCR conditions were as follows: the first-strand DNA template was heated at 94̊C for 3 min before amplification, a total of 30 cycles (for Nedd4-1) or 23 cycles (for β-actin) were used including 94̊C for 30 sec (denaturation), 63̊C for 30 sec (annealing), 72̊C for 35 sec (extension), and 72̊C for 7 min (extension).

Standard western blotting was performed using antibodies against Nedd4-1, Rap2a, TNIK, Myc and β-actin. Cell extracts were prepared 48 h after transfection and subjected to 10–12% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), followed by transference to 0.45-µm pore size polyvinylidene fluoride membranes (Millipore) and blotted with the antibodies as indicated.

Immunohistochemistry was performed as previously described, with minor modifications (29). Paraffin-embedded tissue sections were obtained from the Department of Pathology, The Affiliated Hospital of Xuzhou Medical College, including 11 cases of WHO grade II astrocytomas, 12 cases of WHO grade III anaplastic astrocytomas, and 12 cases of WHO grade IV glioblastomas, with non-cancerous brain tissues as controls. Slices were incubated at 4̊C overnight with the anti-Nedd4-1 antibody (1:100 dilution; Abcam, Cambridge, MA, USA). Subsequent to microphotography, the percentages of glioma cells with positive staining were presented and the Nedd4-1 expression levels were thereby categorized.

Cell transfection was performed using PolyJet™ In Vitro DNA Transfection Reagent (SignaGen, Rockville, MD, USA) when U251 cells reached 90% confluence on 10-cm plates. Plasmids encoding shRNA-Nedd4-1, full-length Nedd4-1, HA-tagged ubiquitin, and Myc-tagged Rap2a and its mutants were transfected in the respective experiments. Forty-eight hours after transfection, the cells were harvested, rinsed with phosphate-buffered saline (PBS) and lysed in 1% SDS or 1% NP40 buffer.

Cell motility was examined by scratch assay as previously described, except for minor modifications (30). Twelve hours after transfection, an artificial gap was created on the confluent cell monolayer using a plastic pipette tip. Migrated cells were quantified at 48 h (magnification, ×200) in order to compare with the cell counts at base line with a computer-aided microscopy imaging system. All experiments were performed in triplicates.

The capabilities of cell migration were assessed by Transwell chamber assay. The invasion assay was performed as previously described with minor modifications (31). Glioma cells were harvested and resuspended in serum-free medium at a concentration of 1×10⁵ cells/ml, and 200 µl was added to the top chamber. The chambers were incubated for 24 h at 37̊C in an incubator with 5% CO₂. The noninvasive cells on the upper surface of the membrane were removed by wiping with cotton-tipped swabs. The 8-µm pore-size membrane was fixed and stained, and the population of the invaded cells was counted by light microscopy. All the experiments were performed in triplicates.

Affinity pull-down assay with GST-Nedd4-1. The pull-down assay was performed as previously described with minor modifications (24). Total proteins were extracted from a 1% NP-40 extract of U251 cells and separated by 12% SDS-PAGE followed by western blotting using anti-TNIK (1:500) and anti-Rap2a (1:500) antibodies.

U251 cells in 10-cm culture dishes were incubated with 1 mM dithiobis(succinimidyl propionate) (DSP) (Pierce, Rockford, IL, USA) for 3 min on ice. Excess DSP was quenched with 20 mM Tris-Cl (pH 7.5 at 4̊C) for 15 min at room temperature, and then the proteins were extracted with RIPA buffer [50 mM Tris-Cl (pH 8.0) at 4̊C, 5 mM EDTA, 150 mM NaCl, 1% NP-40, 0.5% sodium deoxycholate, 0.1% SDS, 0.2 mM phenylmethylsulfonyl fluoride (PMSF), 1 µg/ml aprotinin and 0.5 µg/ml leupeptin]. Nedd4-1 was immunoprecipitated with anti-Nedd4-1 polyclonal antibodies coupled to protein A-sepharose beads (GE Healthcare, Milwaukee, WI, USA) at 4̊C for 14 h. The beads were rinsed with RIPA buffer, and then the precipitated proteins were eluted and boiled in 2X loading buffer containing 50 mM DTT, and separated by SDS-PAGE followed by western blotting using anti-Nedd4-1, anti-TNIK and anti-Rap2 antibodies, respectively.

A total of 40 µg of GST fusion proteins of RalGDS encoding the Ras-binding domain (RBD) was purified and immobilized on GST-sepharose beads to pull down the GTP-bound form of Myc-Rap2a extracted from the U251 cells transfected with Myc-Rap2a, Nedd4-1 and HA-ubiquitin. After extensive washing at 4̊C overnight with a buffer containing 50 mM Tris-HCl (pH 7.5), 200 mM NaCl, 2.5 mM MgCl₂, 10% glycerol and 1% NP-40, the bound proteins were eluted by boiling in 2X loading buffer followed by SDS-PAGE and western blotting.

Data are presented as mean ± standard error of mean (SEM). Statistical significance was determined using Student's t-tests for comparisons between two groups and ANOVA for intergroup comparisons. P-values of <0.05 were considered statistically significant (P<0.05).

To investigate the possible effect in glioma pathogenesis, we determined Nedd4-1 protein expression in clinical glioma tissues and non-cancerous brain samples. A representative immunoblot is shown in Fig. 1A. Analysis of the band density revealed that Nedd4-1 was significantly upregulated in the glioma tissues vs. that observed in the non-cancerous samples (Fig. 1B). We further verified Nedd4-1 expression by immunohistochemistry. As shown in Fig. 1C, Nedd4-1 immunoreactivity was predominant in the cytoplasm. As shown in Fig. 1D, the percentage of Nedd4-1-positive cells reached 57.35±2.02% (WHO grade II) and 75.82±3.06% (WHO grade III–IV) in glioma tissues vs. the percentage of 22.30±2.17% in non-cancerous brain samples. Thus, Nedd4-1 expression was significantly correlated with glioma malignancy. Fig. 1 shows significant upregulation of Nedd4-1 in glioma tissues vs. non-cancerous brain samples. These results indicated the potential involvement of Nedd4-1 in the pathogenesis of human glioma.

Given the overexpression of Nedd4-1 in glioma in contrast to non-cancerous tissues, we employed a gain-of-function strategy to assess the effect of Nedd4-1 on glioma cell migration. As shown in Fig. 2A and B, exogenous shNedd4-1 and Nedd4-1 were added to interfere with or overexpress Nedd4-1 in the U251 and U87 cell lines. The effect of Nedd4-1 on cell migration was assessed by scratch assay. It was shown that the migratory capacity of glioma U251 cells was significantly reduced upon downregulation of Nedd4-1 as compared with the scramble group. Conversely, the scratch in the Nedd4-1-overexpressing group was rapidly replenished compared with the vector group (Fig. 2C and D). We also performed cell migration assays using Transwell chambers to investigate the migratory capacities of the U251 and U87 cells. Likewise, the shNedd4-1 group showed significantly reduced migration compared with that noted in the scramble group (Fig. 2E and F), whereas the migratory behavior in the Nedd4-1-overexpressing group was evidently enhanced compared with that noted in the vector group in both the U251 and U87 cell lines (Fig. 2G and H). Collectively, we demonstrated that either the downregulation or overexpression of Nedd4-1 exerted relevant causality on the migratory capacity of both cell lines.

In addition to the regulation of cell migration by Nedd4-1, we investigated the effect of Nedd4-1 on cell invasion with Matrigel-precoated Transwell chambers. The results indicated that the invasion of U251 and U87 cells was significantly diminished by the downregulation of Nedd4-1 vs. the scramble group subsequent to 24 h of incubation (Fig. 3A and B), while the invasion in the Nedd4-1-overexpressing group was markedly enhanced vs. that noted in the empty vector group in both cell lines (Fig. 3C and D). Thus, both downregulation and overexpression of Nedd4-1 regulated the invasive capacities of the cell lines.

To verify the role of Nedd4-1 targets in glioma cells, we selected the Nedd4-1 WW domains (32) (residue 217–547) as an affinity matrix and U251 cell lysate (1% NP40) to pull down TNIK and Rap2a. The results exhibited the presence of TNIK and Rap2a in the affinity-purified complex, confirming the binding of Nedd4-1 to either of the substrates (Fig. 4A). We further identified that endogenous Nedd4-1, TNIK and Rap2a constituted a complex (Fig. 4B). U251 cell lysate was incubated with thiol-cleavable cross-linker prior to immunoprecipitation with an anti-Nedd4-1 antibody. Under this condition, TNIK and Rap2a were coimmunoprecipitated with Nedd4-1.

Rap2a rather than TNIK was found to be ubiquitinated by Nedd4-1 in HEK cells, which decreased the activity of the GTP-bound form of Rap2a (24). To ascertain the potential effect of Nedd4-1 on Rap2a ubiquitination in glioma cells, expression plasmids for the full-length Nedd4-1, HA-tagged ubiquitin as well as Myc-tagged WT-Rap2a, DA-G12V Rap2a or DN-S17N Rap2a were co-transfected in the U251 cells, with lysates extracted under denaturing conditions (100̊C, 1% SDS) and Myc-tagged proteins immunoprecipitated from cell extracts using anti-Myc antibodies. As shown in Fig. 4C, immunoprecipitates from WT and DA-G12V instead of DN-S17N-expressing U251 cells showed a gradient of protein bands visible to an anti-ubiquitin antibody (P4D1) only in the presence of Nedd4-1 overexpression. These data demonstrated that Nedd4-1 selectively ubiquitinated WT-Rap2a and DA-Rap2a, but not DN-Rap2a.

To investigate the exact form of ubiquitination of Rap2a (i.e. mono-ubiquitination, di-ubiquitination or poly-ubiquitination), expression plasmids for full-length Nedd4-1, HA-tagged ubiquitin and Myc-tagged WT-Rap2a were co-transfected in U251 cells. Four types of anti-ubiquitin monoclonal antibodies which discriminate between mono-ubiquitinated and poly-ubiquitinated proteins were employed: P4D1 for poly-ubiquitin and mono-ubiquitin conjugates; FK1 only for poly-ubiquitin conjugates; Apu2 and Apu3 for K48-linked and K63-linked poly-ubiquitin chains, respectively. As shown in Fig. 4D, the escalation of ubiquitinated Rap2a was only identified by P4D1, indicating mono-ubiquitination of Rap2a by Nedd4-1. Apu3 recognized a single faint band representing ubiquitinated Rap2a by Nedd4-1 via K63-linked ubiquitination. No bands were determined by Apu2, exempting the Rap2a ubiquitination by Nedd4-1 from K48-linked polyubiquitin chain, which may lead to proteasome-dependent degradation. A total lysate of Neddd4-1-expressing U251 cells was immunoblotted using these antibodies as controls with blurred bands.

To further investigate the effect of Nedd4-1-mediated ubiquitination on Rap2a function in glioma U251 cells, we purified and immobilized a GST fusion of the Rap2a-binding region of RalGDS, which served to pull down active GTP-bound Rap2a. Nedd4-1 and Myc-tagged Rap2a were thereby co-transfected in the U251 cells, and the lysate was loaded onto GST-RalGDS or onto GST as a control. Only active GTP-bound Rap2a was pulled down, indicating that ubiquitination blocked Rap2a function by inhibiting the activity of the GTP-bound Rap2a (Fig. 4E and F).

To verify the role of Nedd4-1-mediated mono-ubiquitination of Rap2a in cell migration and invasion, expression plasmids for full-length Nedd4-1, HA-tagged WT-ubiquitin and its mutants (including HA-tagged K63R-ubiquitin, HA-tagged K48R-ubiquitin, HA-tagged ubiquitin and Myc-tagged WT-Rap2a) were co-transfected in the U251 and U87 cell lines, with MG-132 added to exclude the interference of multiple-ubiquitination in all groups. We assessed migration and invasion parameters in each group of both U251 and U87 cells. As shown in Fig. 5A and B, Nedd4-1 and Rap2a were overexpressed in the U251 and U87 cell lines. The Transwell assay results demonstrated that the migration of glioma cells was significantly enhanced in the WT-ubiquitin and K48R-ubiquitin groups vs. the vector and K63R-ubiquitin groups (Fig. 5C and D). The invasive capacity of glioma cells was also significantly increased in the WT-ubiquitin and K48R-ubiquitin groups vs. the vector and K63R-ubiquitin groups in the Matrigel Transwell assay (Fig. 5E and F). Furthermore, as illustrated (Fig. 5G), the results of immunoprecipitation revealed the Rap2a mono-ubiquitination of Nedd4-1 via K63-ubiquitin. These results indicated that Nedd4-1 may regulate the migration and invasion of glioma cells via the inhibition of GTP-Rap2a.