All chemicals were purchased from Sigma-Aldrich (St. Louis, MO, USA), unless otherwise mentioned. Dulbecco’s modified Eagle’s medium (DMEM), penicillin-streptomycin, Dulbecco’s phosphate-buffered saline (PBS) and fetal bovine serum (FBS) were purchased from Invitrogen (Carlsbad, CA, USA).

The genotypes of the Saccharomyces cerevisiae strain MaV203 (Invitrogen) used in this study were as follows: MATα, leu2-3, 112, trp1-901, his3Δ200, ade2-101, gal4Δ, gal80Δ SPAL10::URA3, GAL1::lacZ, HIS3UAS GAL1::HIS3@LYS2, can1R and cyh2R. This strain contains a set of non-reverting autotrophic mutations. A blend of yeast extract, peptone, dextrose and adenine sulfate medium was used to grow the untransformed yeast strain. Synthetic dropout minimal medium (Clontech Laboratories, Inc., Victoria, Australia) with appropriate nutrients was used to separate the untransformed yeast strain from the transformed.

Human liver CDK2AP1 cDNA (Stratagene, La Jolla, CA, USA) was ligated between the EcoR I and Xho I sites of a pBD-GAL4 Cam phagemid vector (Stratagene), downstream of the GAL4 DNA-binding domain. A liver cDNA library was digested by an interference-resistant helper phage from the Cam Phagemid Vector kit (Stratagene) and transformed into the Escherichia coli (E. coli) XLOLR strain (Stratagene). The human CDK2AP1 cDNA and the UPP gene were cloned into the pGEX-4T-1 GST fusion expression vector (GE Healthcare, Piscataway, NJ, USA). As there was no commercially available antibody to detect the UPP, HA and FLAG protein tags were used to identify the expression of the UPP and p12^CDK2AP1. HA and FLAG mouse monoclonal antibodies (Zhongshan Golden-Bridge Biotechnology Co., Ltd., Beijing, China) were used to test the expression of HA-p12^CDK2AP1 and FLAG-UPP in eukaryotes.

A human cDNA library was screened using the pBD-GAL4 plasmid, which was constructed with human full CDK2AP1 cDNA. The cDNA library bacterial XL1-Blue MRF was digested by a phage. The two vectors were then cotransfected into the MaV203 S. cerevisiae strain. Subsequently, positive clones were verified by distinguishing the His⁺ phenotype from the Leu⁺ and Try⁺ phenotypes. The screening procedure fully complied with the manufacturer’s instructions. Positive interaction was confirmed by testing the dropout culture media, and then the positive plasmid was transformed into E. coli JM109 for sequencing.

The pGEX 4T-p12^CDK2AP1 and pGEX 4T-UPP plasmids were transformed into BL21 E. coli to express GST-p12^CDK2AP1 and GST-UPP fusion proteins. Isopropyl β-D-1-thiogalactopyranoside (200 μl, 1 mM) was added to the flasks and the cells were induced for 5 h. The cells were then harvested with PBS. Pelleted cells were lysed by sonication in PBS and Complete Protease Inhibitor Cocktail (Roche, Mannheim, Germany). Following centrifugation at 12,000 × g for 5 min, the supernatants were applied to Glutathione Sepharose 4B (GE Healthcare). Thrombin Protease (GE Healthcare) was added to the GST-UPP tube for 30 min. The beads were pelleted, washed three times in PBS, resuspended in Laemmli sample buffer and analyzed by 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE).

Human 293T cells and HeLa cells were cultured under 37°C, 5% CO₂ in DMEM supplemented with 10% FBS, penicillin (10,000 U/ml) and streptomycin (10,000 μg/ml). When the cells reached 50–60% confluence, Lipofectamine 2000 reagent (Life Technologies Corp., Rockville, MD, USA) was used to transfect the pcDNA 3.1-FLAG-UPP and pcDNA 3.1-HA-p12^CDK2AP1 plasmids according to the manufacturer’s instructions. The medium was changed after 5 h.

A Protein G Immunoprecipitation kit was used to perform the Co-IP assay according to the instructions provided with the kit. The 293T cells were lysed in IP buffer 48 h after transfection. The lysate was divided into two spin columns and the samples were analyzed by SDS-PAGE. Proteins were transferred onto nitrocellulose membranes (Invitrogen) and then incubated with primary antibodies. The anti-HA antibody (1:10,000/2 μl) was applied to the UPP membrane, while the anti-FLAG antibody (1:10,000/2 μl) was applied to the p12^CDK2AP1 membrane. After washing four times with TBST (20 mM Tris-HCl, pH 7.4; 137 mM NaCl and 0.1% (v/v) Tween 20), the membranes were incubated with horseradish peroxidase-conjugated goat anti-mouse secondary antibodies. The ECL Western Blotting Detection system (Amersham Biosciences, Piscataway, NJ, USA) was used to detect horseradish peroxidase on the immunoblots. The dried membranes were exposed to Carestream Kodak Biomax MR film for 2 min at room temperature.

The siRNA sequences used were designed and purchased from Genepharma Co. (Shanghai, China), and were as follows: Forward, 5′-UCUUUUCCAAACUGAAUGCGC-3′ and reverse, 5′-GCAUUCAGUUUGGAAAAGACA-3′ for the UPP gene; and forward, 5′-AGGAGAUCAGACCCACGUATT-3′ and reverse 5′-UACGUGGGUCUGAUCUCCUTT-3′ for the p12^CDK2AP1 gene. The siRNA oligo was transfected into the 293T cell line with Lipofectamine 2000. After 24 h, the cells were harvested and divided into two centrifuge tubes. Total RNA from the cultured cells was isolated using TRIzol reagent (Life Technologies Corp., Carlsbad, CA, USA). The primers used to detect the expression levels were as follows: Forward, 5′-AACGGAACGGAATGC CAGATCCTA-3 and reverse, 5′-TTCAGAGCCAAGTGA ACCATGGGA-3′ for p12^CDK2AP1; and forward, 5′-TGGTGC ACATCGTGTAACCTGTCT-3′ and reverse, 5′-AAGCTT TCCTTCCTCAGCCACTCT-3′ for the UPP. Following cDNA synthesis, qPCR was performed using the Bio-Rad Real-Time PCR Detection system (Bio-Rad, Hercules, CA, USA) with the human housekeeping cDNA GAPDH as the control. Radio immunoprecipitation assay buffer cocktail (Dingsheng Biotechnological Co., Beijing, China) was added to the tube containing 293T cells. Anti-mouse HA monoclonal antibody was used to perform western blotting to test the expression of p12^CDK2AP1. The human 293T, A-549, HeLa, MCF-7, ACC-M, fibroblast and human mesenchymal stem cell lines were incubated to examine the endogenous expression of the p12^CDK2AP1 and UPP genes. The comparative threshold cycle (CT) method was used to analyze relative changes in gene expression.

The UPP-transfected 293T and HeLa cells were harvested at 12 and 24 h. Prior to flow cytometric examination, the cells were combined with propidium iodide (1 mg/ml) and RNase (50 μg/ml). The cell cycle was analyzed by flow cytometry (FACSCalibur; Becton-Dickinson, San Jose, CA, USA), and ModFit and CellQuest software was utilized. The plasmid pcDNA 3.1-FLAG-UPP was transfected into 293T cells. 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (20 μl, 5 mg/ml) was applied to the wells after incubation for 24 h. Dimethyl sulfoxide (Amersco, Solon, OH, USA) was added to each well to dissolve the formazan crystals for 10 min. The samples were then transferred to a plate reader (ELx808; BioTek, Shanghai, China) and the absorbance at 570 nm was measured.

The bisulfite sequencing PCR (BSP) method was used to examine the UPP gene promoter methylation in the 293T and HeLa cell lines. The methylation primers used to perform PCR were as follows: Forward, 5′-GTAGTG TGAGTGAGGGTTTTGATTT-3′ and reverse, 5′-CTAACA ACAAATTACCCCAACTTTC-3′. Streptozotocin buffer (Tris-HCl 10 mM, pH 8.0; EDTA 4 mM; NaCl 0.4 M), SDS and protease K were applied to harvested 293T and HeLa cells. DNA extraction followed the standard procedure. Na₂S₂O₅ (5 mol/l) and hydroquinone mixture were used to prepare HeLa and 293T cell DNA. Subsequently, PCR was performed with the conditions as follows: 94°C for 5 min, 94°C for 40 sec, 56°C for 40 sec and 72°C for 50 sec for 34 cycles, and then 72°C for 10 min. An E.Z.N.A. Cycle Pure kit (Omega Bio-Tek Inc., Norcross, GA, USA) was used to purify the PCR products. Subsequently, the PCR products were ligated into a pMD 18T vector (Takara Bio, Inc., Tokyo, Japan) for sequencing. BiQ Analyzer software was used to statistically analyze the results (21).

The animal experiment followed the Laboratory Animal Center Institutional Animal Care and Use committee guidelines of the Peking University Health Science Center. Nude mice (n=36; Balb/c nude, female, four weeks old) were divided into three groups at random. The HeLa cells which were stably transfected with the UPP gene by an Overexpression Lentivector (SBI, Mountain View, CA, USA) were screened by flow cytometry. HeLa-UPP cells were subcutaneously injected at a density of 1×10⁵ cells/ml into the armpits of the nude mice, with the HeLa cell line as the negative control and the blank plasmid-infected HeLa cells as the positive control. Once the tumors reached 3–5 mm, all the tumors were harvested. The weights and volumes of the tumors were then calculated at the same time.

All cell experiments were repeated three times. The data were calculated as the mean ± standard deviation and the statistical analyses were performed using SPSS Software, version 19.0 (SPSS, Inc., Chicago, IL, USA). The differences were considered to be statistically significant when P<0.05, as determined by Student’s t-test.

Interacting cellular partners of p12^CDK2AP1 were searched for by screening a human cDNA library using a bait construct with the full CDK2AP1 cDNA cloned in-frame through using a yeast hybrid system. Following screening, positive clones were identified by selecting with synthetic dropout minimal medium. A protein that activated yeast transcription was detected (Fig. 1A). The activating domain/library plasmids were confirmed by the cotransformation procedure, and the nucleotide sequence of the protein was obtained by cloning and sequencing. Through blasting the sequence in the NCBI BLAST database, a novel protein named UPP (BC006130) was identified. The UPP is encoded by the Loc93622 gene, which is located on 4q16.1, and includes 119 amino acids.

Considering the possibility of a false positive result in the yeast two-hybrid system, the interaction between p12^CDK2AP1 and the UPP was confirmed by GST-pull down. The prokaryotic expression plasmids pGEX-4T-p12^CDK2AP1 and pGEX-4T-UPP were constructed to express GST fusion proteins, and then the GST-p12^CDK2AP1 and GST-UPP proteins were purified by GST affinity chromatography (Fig. 1B and C). Subsequently, the GST tag of GST-UPP was cleaved by thrombin protease. The pull-down result supports the specific interaction of p12^CDK2AP1 with the UPP (Fig. 1D). To confirm this interaction in mammalian cells, the interaction of p12^CDK2AP1 and the UPP was examined using a Co-IP assay. The bound proteins were detected by western blotting using antibodies which were specific to HA and FLAG (1E and F). These results demonstrated that p12^CDK2AP1 interacts with the UPP in cells.

To analyze the endogenous expression of the UPP gene, non-neoplastic cell lines (293T, human mesenchymal stem cells and human fibroblast cells) and human cancer cell lines (ACC-M, A-549, MCF-7 and HeLa) were tested. The results of the qPCR demonstrated that the UPP and p12^CDK2AP1 were similarly expressed (Fig. 2A and B), that is, the UPP expression levels were higher in the non-neoplastic cell types than in cancer cell lines. Due to the high GC ratio (63%) of the UPP gene, the UPP gene was searched in the NCBI database and a CpG island of the gene was identified. The methylation of the UPP gene was investigated in the 293T and HeLa cell lines. The DNA of 293T and HeLa cells was extracted and amplified by BSP methods. The sequencing results demonstrated that there were 31 methylation sites in the UPP gene. In the DNA of the HeLa cells, all the methylation sites were methylated, while the 293T DNA was not methylated (Fig. 2C).

To investigate the association between CDK2AP1 and the UPP gene, siRNA and overexpression assays were performed. Following transfection of the p12^CDK2AP1 siRNA oligo, the western blot analysis demonstrated the silencing effect of the CDK2AP1 gene (Fig. 2D). The results indicated that the UPP expression levels were decreased when the CDK2AP1 gene was interfered with (Fig. 2E), while the UPP siRNA did not downregulate the expression levels of the CDK2AP1 gene (data not shown). Subsequently, the association between CDK2AP1 and the UPP was evaluated by overexpressing the two genes in the 293T cell line. The results revealed that overexpression of the CDK2AP1 gene significantly increases the expression levels of the UPP gene (Fig. 2F), while overexpression of the UPP gene does not affect the expression of CDK2AP1 (data not shown). Thus, the UPP may function as a downstream effector of CDK2AP1.

p12^CDK2AP1 mediates the cell cycle and affects cell vitality; thus, whether the UPP gene also shifts the cell cycle and is associated with cell vitality was investigated. The UPP and p12^CDK2AP1 genes were transfected into 293T and HeLa cells. After the genes had been transfected for 24 h, the flow cytometry data indicated that the percentage of cells in the G2/M phase was decreased significantly from 11.13% to 6.16% in the 293T cells (Fig. 3A) and from 12.15% to 7.65% in the HeLa cells (Fig. 3B). The cell growth curve demonstrated that overexpression of the UPP also suppresses cell proliferation (Fig. 3C and D). After transfection for 5 days, the cell vitality results indicated that overexpression of the UPP decreased cell survival (Fig. 3E and F).

To investigate the mechanisms of action of the UPP protein in vivo, a HeLa cell line stably transfected with the UPP gene (Hela-UPP) was constructed (Fig. 4A and B) and transplanted in nude mice as xenografts. After 62 days, the xenograft tumors were harvested and fixed (Fig. 4C and D), and the weight and volume of the tumors were measured. The results demonstrated that the mean weight of the tumors of the HeLa-UPP cells (0.38±0.07 g) was lower than that of the control cells (0.77±0.11 g; Fig. 4E), and the mean volume of the tumors in the HeLa-UPP cells (109.62±20.40 mm³) was also lower than that of the control group (228.62±76.58 mm³; Fig. 4F). These results indicated that the UPP effectively inhibits the growth of human tumor xenografts.