Receptor protein tyrosine kinases (RPTKs) play important roles in the regulation of a variety of cellular processes including cell migration, proliferation, and protection from apoptosis. Here, we report the identification and characterization of a novel RPTK-like molecule that has a critical role in induction of tumorigenesis and metastasis and is termed Novel Oncogene with Kinase-domain (NOK). NOK contains a putative single transmembrane domain and a conserved intracellular tyrosine kinase domain that shares homology with members of the platelet-derived growth factor/fibroblast growth factor receptor superfamily. NOK was exclusively located in the cytoplasm. NOK mRNAs were detected in limited human organs and expressed with the highest abundance in the prostate. A variety of tumor cells also expressed the NOK mRNAs. We demonstrated that NIH3T3 and BaF3 cells could be strongly transformed by the expression of the NOK gene as examined by colony formation experiment. In addition, BaF3 cells with the stable expression of NOK induced rapid tumorigenesis in nude mice. Interestingly, these NOK-expressing tumor cells could promptly invade and spread into various distinct organs and form metastatic foci, eventually leading to the rapid death of these animals. Moreover, molecular mechanism studies indicated that NOK could concomitantly activate both MAP kinase and phosphatidylinositol 3′-kinases (PI3K) pathways in stable BaF3 cells. Thus, our results both in vitro and in vivo suggest that NOK is a novel oncogene with the capacity of promoting cell transformation, tumorigenesis, and metastasis.
Receptor protein tyrosine kinases (RPTKs) play important roles in diverse cellular and developmental processes, such as cell proliferation, differentiation, and survival (1, 2, 3) . The typical structure of RPTK is composed of a single transmembrane domain, a diverse ectodomain specific for ligand binding and an intracellular tyrosine kinase domain for activating downstream signaling cascades. RPTKs are often involved in mitogenic signaling, therefore, stringent regulation of RPTK expression is required for maintaining the normal cellular functions (4) . In contrast, aberrant expressions and activation of RPTK can cause numerous genetic disorders, including tumor formation (4 , 5) . At present, at least 18 RPTKs have been demonstrated to function as oncogenes. Well-known examples include fibroblast growth factor receptor (FGFR), epidermal growth factor receptor, platelet-derived growth factor receptor (PDGFR), and MET/Ron tyrosine kinase receptor (4) .
Generally, oncogenic transformation of RPTK is ascribed to the enhanced kinase activity resulting from either constitutive receptor dimerization or protein-structural alterations (2) . Inappropriate genetic alterations such as gene amplifications and chromosomal translocations can increase the expression of RPTK, leading to a higher concentration of receptors and triggering receptor dimerization (6 , 7) and then activation by autophosphorylation (1 , 2) . For instance, epidermal growth factor receptor and MET receptors were often amplified in human breast cancer and gastric carcinoma, respectively, whereas TPR-MET and FGFR2-FRAG1 fusion genes have been found in human and rat osteosarcoma cell lines, respectively (8 , 9) . In addition, hyperactivation of RPTKs can also result from alterations in protein structures by point mutations or splicing variants that lose the critical domains for receptor autoinhibition and then expose the kinase domain to substrates (6 , 9) . For example, activating point mutations have been identified in the MET gene, resulting in activation of the receptor without ligand binding, which contributed to development of papillary renal carcinoma and head and neck squamous cell carcinomas (10 , 11) .
In addition, aberrant expressions of certain isoforms of RPTK have frequently been documented and implicated in numerous human cancers (4 , 5) . Soluble FGFR3 lacking the entire transmembrane domain has been isolated from human osteosarcoma and breast cancer cells (12 , 13) . An in-frame deletion of 49 amino acids in the ectodomain of the Ron tyrosine kinase receptor led to constitutive receptor activation in human gastric cancer cell line (KATO-III; Ref. 14 ). Most recently, a novel splice variant of FGFR4 (ptd-FGFR4) was isolated from a human pituitary tumor (15, 16, 17) . Intriguingly, this ptd-FGFR4 was NH2-terminally truncated at the upstream of IgIIIc domain, resulting in an intracellular FGFR4 variant without 5′ signal peptide, IgI, and IgII. As a result of the truncation, this protein was exclusively retained in the cytoplasm compartment. ptd-FGFR4 was constitutively active when it was stably expressed in NIH3T3 cells and caused cellular transformation and tumor formation in nude mice. Moreover, selective expression of ptd-FGFR4 in transgenic mice recapitulated pituitary tumor progression in human. Another example of receptors with NH2-terminal truncation is the hepatocyte growth factor receptor (Met). NH2-terminal truncated Met has been implicated in the human malignant musculoskeletal tumors (18) .
In the current study, a novel protein with RPTK like structure [Novel Oncogene with Kinase-domain (NOK)] has been cloned and characterized. This molecule shares ∼30% amino acid identity with members of the FGF/PDGF receptor superfamily but seems to be separated much earlier during the revolution process. NOK is expressed in a number of organs. Studies both in vitro and in vivo demonstrated that NOK could act as an oncogene to transform NIH3T3 and BaF3 cells and to promote tumor formation. Importantly, NOK could also induce tumor cell invasion as well as metastasis in several distant organs. Thus, NOK-mediated oncogenesis may represent an interesting example of cancer development involving RPTK.
MATERIALS AND METHODS
Cell Culture, Transfection, and Establishment of Stable Cell Lines.
Murine embryo fibroblast cells NIH3T3, modified human embryonic kidney cells 293T, human monocytic cells U937, murine pre-B cells BaF3, and murine mast cells WEHI-3B cells were obtained from the Cell Culture Center of Basic Institute of Medical Sciences, Chinese Academy of Medical Sciences. Other cell lines including human hepatoma cells LO2, human epidermoid carcinoma cells A431, murine macrophage cells Ana-1, SV40 transformed monkey kidney cells Cos1 and Cos7, human ovary cancer cells Ho8910, human cervix carcinoma cells HeLa, human promyelocytic leukemia cells HL-60, and human chronic myelogenous leukemia cells K562 were from Cell Center of Institute of Biochemistry and Cell Biology, Chinese Academy of Sciences. BaF3 cells were grown in RPMI 1640 supplemented with 10% fetal bovine serum, 2 mm glutamine, 50 μm mercaptoethanol, 100 units/ml penicillin, 100 μg/ml streptomycin, and 10% WEHI-3B conditional medium. All cultured cells were grown in DMEM containing 10% calf serum, 100 units/ml penicillin, 100 μg/ml streptomycin, and 2 mm glutamine. Cells were plated on either 6-well dishes or 35-mm-diameter culture plates at a density of 5 × 104 cells/ml and incubated at 37°C for 2 days. About 2–4 μg of DNA were used for transfection into NIH3T3 cells with LipofectAMINE 2000 (Life Technologies, Inc.). BaF3 cells (about 5 × 106) were transfected using electroporation with 20 μg of plasmid DNAs using a BTX machine at 1500 uF and 235–240 V (t ∼ 35–40 ms). Transfected cells were first plated on 96-well dishes and selected in the presence of 800 μg/ml G418 for 10 days. Then, selected resistant clones were expanded in 10-cm culture dishes for additional analysis.
Isolation of Full-Length cDNA of NOK.
The full length of human NOK was isolated from RNAs prepared from human amygdala tissue using RNAzol extract kit (Life Technologies, Inc.) with specific primers corresponding to either the 5′ or 3′ end of the cDNA: 5′-atgggcatgacacggatgct-3′ and 5′-tcaaagcatgctatagttg-3′, respectively. Mouse homologue was obtained from colon tissue RNAs with specific primers derived from mouse NOK cDNA sequence: 5′-atgggagagaagggtcacct-3′ and 5′-tcaaaggacactgaagctat-3′. One μg of total RNA derived from respective organs was used as template for reverse transcription-PCR. The expected PCR products were subcloned into pGEM-T vector (Promega) and confirmed by DNA sequencing.
The full length of NOK with an influenza hemagglutin (HA) epitope tag was subcloned into the EcoRV and NotI sites of pCDNA3 to form pCDNA3-NOK. To construct the NOK/EGFP fusion gene, NOK was released with NheI and SalI and subcloned into pEFBos to form plasmid pEFBos-NOK. The extracellular domain of EPO receptor was subcloned into EcoRV and NotI sites of pCDNA3-EPOR, and Flag-tagged transmembrane and intracellular domain of NOK were inserted into the NotI and XbaI sites of pCDNA3-EPOR to obtain the plasmid pCDNA3-EPOR/NOK. All constructs were confirmed by sequencing.
Northern Blot and Reverse Transcription-PCR Analysis on mRNA Expression.
About 20 μg of total RNA from each tissue were loaded onto 1% agarose gel containing 0.22 m formaldehyde and electrophoresed within 1× MOP running buffer [0.02 m 3-(N-morpholino)propanesulfonic acid, 5 mm sodium acetate, and 1 mm EDTA] for 2–3 h at 90 V. After soaking the gel at 10 × SSC for 20 min, RNAs were transferred to nylon membrane by using Nytran transfer system (Schleicher & Schuell, Inc.) following the manual description. Transferred RNAs were immobilized with 254 nm of UV light for 4 min and then baked at 80°C for 1 h. About 2 × 105 cpm of [α-32P]dATP randomly labeled cDNA probe was added to fresh Express Hyb solution (Clontech), and hybridization was performed as described in the manual. The hybridized membrane was examined by a Molecular Dynamics PhosphorImager.
Total RNAs from 5 × 106 cells were extracted with 1 ml of Trizol (Life Technologies, Inc.) plus 200 μl of chloroform. The supernatant was precipitated with an equal volume of isoproponal. RNA pellet was resuspended in sterile water. About 1–2 μg of RNAs were used as templates for one-step reverse transcription-PCR analysis (Takara Biotechnology Co., Ltd) as described in the manual. The reaction mixture was first incubated at 50°C for 30 min for reverse transcription and then denatured at 94°C for 2 min before PCR cycles. Thirty cycles of PCR were conducted at the following conditions: 94°C for 30 s; 55°C for 30 s; and 72°C for 2 min. The reaction products were extended at 72°C for 10 min before storing at 4°C.
Western Blot Analysis.
Cells were lysed in 20 mm Tris-Cl (pH 7.4), 150 mm NaCl, 1 mm EDTA, 1% Triton X-100, 1 mm Na3VO4, 2.5 mm sodium PPI, 1 mm β-glycerolphosphate, 1 mm phenylmethylsulfonyl fluoride, 5 μg/ml aprotinin, and 5 μg/ml leupeptin (pH 7.5). Cell lysate was separated onto 10% SDS-PAGE and electro-transferred to nitrocellulose membrane (Hybond ECL; Amersham Biosciences) at 100 V for 1.5 h. The transferred membrane was blocked with 10% milk in Tris Buffered Saline Tween 20 (TBST) at 37°C for 1 h, then incubated with polyclonal or monoclonal antibodies, followed by horseradish peroxidase-conjugated secondary antibodies, and developed by enhanced chemiluminescence according to the manual description (ECL; Amersham Biosciences).
Anchorage-Independent Growth, Proliferation Assay, and in Vivo Tumorigenesis.
About 1 × 105 stable BaF3 or transfected NIH3T3 cells were resuspended in 5 ml of 0.4% top agar in DMEM supplemented with 10% calf serum and 400 μg/ml G418, which was layered over 5 ml of 0.7% bottom agar dissolved in DMEM with 10% calf serum plus 400 μg/ml G418 in a 60-mm culture dish (19) . After 2 weeks, the anchorage-independent colonies were stained with a complete DMEM medium with 0.25 mg/ml iodonitrotetrazonium for 2 days.
Stable BaF3 cells (1 × 105) were added to each well of a 96-well plate and starved at a condition without WEHI-3B conditional medium (RPMI 1640 plus 1% fetal bovine serum). During each time course (days 0, 1, 2, and 3), 1 μCi of [3H]thymidine was added to each well, and incubation continued for 6 h before harvesting. Triplicate was performed for each time point. To assay the EPO effect on chimeric stable BaF3 cells, different concentrations of EPO (0.001, 0.01, 0.1, 1, 5, or 10 units/ml) were added to each well and cultured for 2 days before the addition of 1 μCi of [3H]thymidine. Triplicates were performed for each EPO concentration. After 6 h of stimulation, cells were washed with 1× PBS and resuspended in 150 μl of 5% trichloroacetic acid. Cell pellet was lysed in 150 μl of 0.5 n NaOH/0.5% SDS, collected into a 96-well scintillation plate, and counted.
Stable cell proliferation was also examined by 3-(4,5-dimethylthiazolyl-2)-2,5-diphenyltetrazolium bromide (MTT) assay using a standard protocol. About 1 × 105 stable cells (BaF3-p3, BaF3-NOK, and BaF3-EPOR/NOK) were plated onto each well of a 96-well plate and grew at starvation conditions (without serum and WEHI-3B conditional medium) for 0, 1, 2, and 3 days. Five h before the end of the incubation, 20 μl of MTT (5 mg/ml) solution were added to each well. Then, 200 μl of DMSO were added to each well and mixed to dissolve crystals. The reaction products were detected using a spectrometer at 570 nm. Each point represented the average from three independent experiments plus SDs.
Four- to 8-week-old athymic, Balb-c nude mice received s.c. injections of BaF3 control (wild type stably transfected with empty vector pcDNA3.0) or stable cells expressing NOK with cell number ≈ 1.0 × 107 into the right superior flanks. Each experiment was conducted in triplicate. Four weeks later, the mice were sacrificed. The tumor, liver, and spleen from each mouse were weighed, fixed with formalin, and stained with H&E. Also, various organs such as brain, lung, stomach, kidney, intestine, colon, and skeletal muscles were stained with H&E staining and analyzed for the presence of tumor cell infiltration.
Organs of nude mice were fixed in 10% formalin and then dehydrated gradually in alcohol. The tissues were embedded in paraffin and were sectioned in a thickness of 5 μm. The sections were first slightly overstained with hematoxylin for 4 min. After differentiating and destaining in acidic alcohol, the sections were blued in bicarbonate until nuclei stand out blue, and then stained with eosin for 2 min.
Identification, Cloning, and Analysis of the NOK Gene.
In a search for possible FGF receptor-like molecule, we identified a novel human RPTK (later named as NOK) and deposited in the National Center for Biotechnology Information database (GenBank accession no. AY563053). Specific primers were designed to amplify the full length of NOK cDNA from human amygdala tumor cells by reverse transcription-PCR. The mouse homologue of NOK was also cloned using RNAs prepared from mouse colon (GenBank accession no. AY563053). Sequence alignment indicated that human and mouse NOKs share 77.1% identity in their amino acid composition. The deduced human NOK protein is 422 amino acids in length with a single transmembrane helix and a putative kinase catalytic domain but lacks the extracellular domain. The NOK gene was mapped to human chromosome 12 (12p12–21.8–27.2) and was composed of at least 11 exons harboring within a ∼55-kb genomic region. Fig. 1A ⇓ shows that the coding sequence starts at exon 3 and extends to exon 11 translating a protein with a transmembrane domain of 23 amino acids at the NH2 terminus and an intracellular kinase domain of 270 amino acids at the protein COOH terminus. Analysis of protein phosphorylation sites using the NetPhos 2.0 program indicated that there were 10 tyrosine, 15 serine, and 10 threonine residues that might serve as the potential phosphorylation sites for intracellular protein kinases. These potential phosphorylation sites are well conserved between human and mouse NOK (Fig. 1B) ⇓ . Interestingly, NOK shares ∼30% and ∼20% amino acid identity with members of FGFR and PGDFR families, respectively. Phylogenetic analysis indicated that NOK was related to known members of FGFR and PGDFR and divided earlier from both groups during the evolution process. Thus NOK belongs to a distant member of FGFR/PDGFR family (Fig. 1C) ⇓ .
Analysis of the NOK Gene Promoter and Expression Patterns of NOK mRNAs in Normal Tissues and Cancer Cell Lines.
To search for the possible clue of NOK function, the expression pattern of human NOK mRNAs was examined using a membrane loaded with mRNAs from various tissues. NOK expressed at high levels in prostate and low levels in colon, placenta, brain, and skeletal muscle (Fig. 2A) ⇓ . The apparent size of NOK mRNAs in the gel was about 3 kb, which was close to the full cDNA length deposited in GenBank (2749 bp). Using specific primers derived from transmembrane region for 5′-RACE, we did not extend to any of the extracellular sequence of NOK, and no apparent open reading frame was identified in an extensive search of the genome sequence upstream of the known coding region of NOK.
To search for the NOK gene promoter, NOK cDNA sequence (GenBank accession no. NM_018423.1) was used to blast the human genome database. NOK cDNA is encoded by 11 exons that fall into the 27,615-kb Homo sapiens chromosome 12 genomic contig (GenBank accession no. NT_009714.16). A 100-kb genomic sequence immediate upstream of the 5′ end of NOK cDNA was used to analyze the potential promoter and first exon by the FirstEF program. The predicted result revealed that there was a putative polII promoter at the region of nucleotides 16,555–17,124 with high probability (P = 0.9932), and the first exon was identified to be a 104-bp fragment between nucleotide 17,055 and nucleotide 17,158 with consensus splicing donor and acceptor sites and high probability (P = 1.0000). In addition, FirstEF also identified a 200-bp region localized immediate upstream of the first exon to be a CpG island, which was an important signature of the 5′ region of many mammalian genes. This information positively indicates that the NOK gene has a well-defined promoter/enhancer element.
In addition, mouse NOK cDNAs were uncovered from total RNAs prepared from mouse colon. Northern blot indicated that mNOK transcripts were abundant in colon and small intestine but were not detected in mouse spleen, skeletal muscle, liver, kidney, heart, and brain (Fig. 2B) ⇓ . Reverse transcription-PCR analysis indicated that NOK mRNAs were present in transformed kidney cell lines (Cos-1 and 293T), in various tumor cells such as human hepatoma cells LO2, human cervix carcinoma cells HeLa, human ovary cancer cells Ho8910, and human chronic myelogenous leukemia cells K562, but not in other tumor cells such as macrophage/monocyte lineages (U937, Ana-1, and HL-60) and human epidermoid carcinoma (A431; Fig. 2C ⇓ ).
Sequence analysis revealed that NOK did not have a signal peptide at its 5′ end, indicating that NOK might not be a membrane protein (Fig. 1A) ⇓ . To define the cellular localization of NOK, the full length of NOK was fused to the NH2 terminus of EGFP to generate a chimeric protein of NOK/EGFP, and the fusion protein was transiently expressed in NIH3T3 cells. Compared with control, the NOK/EGFP protein was predominantly expressed in the cytoplasm (Fig. 2D) ⇓ , consistent with the observation that there was no signal peptide in the coding sequence of NOK.
NOK Could Promote Proliferation of BaF3 Cells Independent of Growth Factors.
We next addressed the question of whether NOK, a cytoplasmic protein tyrosine kinase, similar to previously reported ptd-FGFR4 that was also located in cytoplasm and observed in human pituitary tumor (16 , 17) , could promote cell proliferation. To assess the NOK effect on proliferation, two stable cell lines were established: one expressed the wild type of NOK and the other expressed the chimeric receptor in which the extracellular domain of mouse EPO receptor was fused to the 5′ of transmembrane and intracellular region of NOK (Fig. 3A) ⇓ . We found that both BaF3-NOK and BaF3-EPOR/NOK could sustain in a harsh condition of starvation in RPMI 1640 without adding any growth factor and serum for up to 2 weeks (Fig. 3B) ⇓ . Furthermore, using EPO as a stimulus, the level of cellular DNA synthesis in BaF3 cells expressing EPOR/NOK was significantly enhanced more than that of BaF3-p3 control. This increase of DNA synthesis reached a plateau when the EPO concentration in the media was greater than 0.01 units/ml (Fig. 3C) ⇓ . In addition, at this concentration, EPO could replace the conditional media of WEHI-3B for the long-term growth of BaF3-EPOR/NOK. To determine whether those stable cells could proliferate independent of growth factor under the starvation condition, we measured DNA replication rates using [3H]thymidine incorporation. In the absence of WEHI-3B conditional medium, both BaF3-NOK and BaF3-EPOR/NOK had higher DNA replication rates for at least 3 days as compared with BaF3-p3 cell control (Fig. 3D) ⇓ . To verify the proliferation effect induced by NOK oncogene in stable BaF3 cells, MTT cell proliferation assay was used to determine the populations of metabolically active cells at each time point (days 1, 2, and 3). Fig. 3E ⇓ shows that BaF3 cells stably expressing either NOK or EPOR/NOK could actively promote cell proliferation under a starvation condition for at least 3 days, whereas the control cell line BaF3-p3 failed to support cell growth in the absence of serum and IL-3. Taken together, these results indicated that overexpression of NOK intracellularly could promote cell proliferation, presumably by activation of downstream mitogenic signals, which might be an important feature of oncogenic protein tyrosine kinases.
Ectopic Expression of NOK in NIH3T3 and BaF3 Cells Induced Cell Transformation in Vitro and Tumorigenesis in Nude Mice.
To assay its transformation ability, NOK was transiently transfected into NIH3T3 cells, and then the transfected cells were plated into 0.4% soft agar. After 2 weeks of culture with 400 μg/ml G418, colonies were formed from NOK transfected NIH3T3 cells and measured as shown in Table 1 ⇓ . Ectopic expression of NOK in BaF3 cells also induced anchorage-independent growth and colony formation in soft agar. The transforming efficiency of BaF3-NOK seemed to be higher than that of BaF3-EPOR/NOK as quantitated by the number of colonies formed after 2 weeks of culture (Table 1) ⇓ . More impressively, s.c. inoculation of BaF3-NOK (≈1 × 107 cells) into nude mice induced tumor formation in vivo (Fig. 3F ⇓ ; Table 2 ⇓ ). As compared with BaF3-p3 control, the life spans of nude mice receiving BaF3-NOK cells were severely reduced (with an average ≈ 30 days). The tumors first appeared at the injection sites after 1 week of inoculation and then dramatically grew to an average of 5.09 ± 0.80 g at death (Table 2) ⇓ . The weights of spleens and livers of BaF3-NOK-injected mice were also significantly increased, indicating the potentially metastatic effect of NOK gene (Fig. 3G ⇓ ; Table 2 ⇓ ). Therefore, NOK overexpression at the intracellular compartment could promote tumor formation in vivo.
NOK Promoted Tumor Cell Invasion and Metastatic Progression.
To exam the possible metastatic effect of NOK, six nude mice were inoculated with BaF3-NOK stable cells and then sacrificed 4 weeks later. Tumor samples and various organ sections such as brain, liver, spleen, kidney, stomach, intestine, lung, skeletal muscle, and colon were prepared and stained with H&E. In the s.c. injection site, the stable BaF3-NOK cells behaved like a malignant tumor that did not have an envelope and could actively grow and penetrate into the adjacent skeletal muscle underneath and massively distributed within the inter-fiber compartments (Fig. 4, A and B) ⇓ . Moreover, staining of a variety of organ sections showed that the invasion of tumor cells was widely spread when nude mice were inoculated with BaF3-NOK cells. The metastatic tumor cells were prevalent in mouse liver, spleen, kidney, and skeletal muscle. In contrast, there were fewer invasive tumor cells and tissue distributions observed when BaF3-EPOR/NOK cells were inoculated (data not shown). The sizes of liver and spleen in BaF3-NOK-injected mice were abnormally enlarged compared with control (Fig. 3G ⇓ ; Table 2 ⇓ ). The average weight of liver in BaF3-NOK-inoculated mice increased 2.4-fold (3.09 ± 0.62 g versus 1.30 ± 0.25 g), whereas the average weight of spleen in BaF3-NOK mice increased even more severely to about 8.7-fold (0.78 ± 0.20 g versus 0.09 ± 0.02 g). In liver, the infiltration of tumor cells disrupted the plate arrangement of hepatocytes in the lobules (Fig. 4A) ⇓ . Under higher magnification, abnormal mitotic figures could be clearly identified in liver section (Fig. 4C) ⇓ . In kidney, the tumor cells were penetrated through arcuate veins, and then infiltrated and spread into the interspace of renal columns, implying that the spreading of tumor cells to distant organs might be directly through blood vessels (Fig. 4A) ⇓ . Although spleen is an unusual organ for tumor metastasis, these NOK-expressing cells frequently promoted the dissemination of transformed cells into spleen. These observations may suggest the preferential dissemination and/or the aggressive character of these tumor cells in vivo (Fig. 4A) ⇓ . In addition, the life span of mice that received injections of these cells was significantly shorter than that of control with an average survival time around 30 days. Thus, these results strongly indicate that the oncogenic properties of NOK are not only associated with cellular transformation and tumorigenesis, but also have a striking promoting effect on tumor metastasis of a number of distant organs, which leads to rapid animal death.
Both the MAP Kinase and Phosphatidylinositol 3′-Kinase (PI3K) Pathways Were Activated during NOK Oncogene Transformation.
Because NOK had mitogenic effect as previously demonstrated by colony formation in soft agar and induced growth factor-independent growth in BaF3 cells, we speculated that overexpression of NOK might induce a series of cellular signaling cascades that were critical for cellular transformation and cancer development. Sequence analysis has indicated that the intracellular kinase domain of NOK shared 30∼34% identity with family members of FGFR/PDGFR, implying that NOK might share some common signaling cascades with FGFRs, such as RAS/MAPK pathway. In addition, the invasive and metastatic effects of tumor cells frequently resulted from a direct activation of PI3K pathway (7) . To assay the possible signaling processes that NOK might activate, cell lysates from BaF3-p3 and BaF3-NOK were prepared and analyzed by Western blot using either anti-p-extracellular signal-regulated kinase (ERK) (phosphorylated) or anti-ERK antibody. Fig. 5A ⇓ shows that stable expression of NOK strongly potentiated the phosphorylated forms of both ERK1 and ERK2, implying that RAS/MAPK pathway in BaF3-NOK cells was highly activated. To confirm this result and to see whether PI3K pathway was also activated, two kinase inhibitors were used: PD98059, which is a specific MEK inhibitor in MAPK pathway, and LY294002, which is a specific inhibitor of PI3K were added in a colony-forming assay. Results from Fig. 5, B and C ⇓ , show that both PD98059 and LY294002 dramatically inhibited colony formation of BaF3-NOK in soft agar, indicating that blocking either RAS/MAPK or PI3K pathway was sufficient to prevent cellular transformation. Previous studies have indicated that activation of PI3K pathway was not only associated with cellular transformation, but also promoted cancer related processes such as tumor cell adhesion, invasion, and metastasis (7) . Thus, the stimulation of both RAS/MAPK and PI3K pathways by oncogene NOK molecule may provide a mechanism for invasive growth and metastatic activity of the NOK-expressing cells.
Perturbation of RPTK function by aberrant expression, point mutation, or truncation has been documented frequently in promoting cellular transformation and tumorigenesis (4 , 21) . In this study, a novel FGFR like molecule, NOK, has been uncovered and shown to be a potent oncogene stimulating a series of cancer-related processes, not only for cellular transformation and tumorigenesis, but also for tumor cell invasion and metastasis. Thus, NOK is a new member of RPTK superfamily that has potential in regulation of cell proliferation, malignancy, and cancer development.
NOK is a naturally occurring form of an unknown protein with certain sequence homology with FGFR/PDGFR family members and resembles the structural features of ptd-FGFR4, which is a known oncogene. Missing of complete extracellular domain and cytoplasmic location are important features of these RPTKs. Both in vitro and in vivo functional studies in this report have indicated that NOK, like ptd-FGFR4, is a potent oncogene. We have shown that NOK-expressing cells gained over-proliferating potential to undergo growth factor-independent cell proliferation. In our nude mice model, NOK-expressing cells were highly tumorigenetic, invasive, and metastatic in a variety of distant organs. Consequently, the life spans of these tumor-carrying animals were significantly reduced. Compared with ptd-FGFR4 and many other known members of FGFR/PDGFR superfamily, NOK was apparently more malignant. Phylogenetic tree analysis indicated that NOK might be distantly related to the FGFR family because it divided earlier than PDGFR during the evolution process. Thus, NOK may be standing out as a distinct member of this super-RPTK family (Fig. 1C) ⇓ . It will be interesting to define the unique features of NOK action and detailed molecular mechanisms of NOK signal transduction. Our data indicated that similar to FGFRs, overexpression of NOK might trigger activations of both PI3K and MAP kinase (Fig. 5) ⇓ . The data presented in this paper strongly indicate that NOK may play certain roles in cancer development.
Tyrosine autophosphorylation on RPTK is a critical step to activate the catalytic activity of the kinase domain that may be governed by a cis-inhibition and trans-activation mechanism (6) . Using an immunoprecipitation experiment, we have observed that the fusion gene EPOR/NOK was constitutively active in promoting not only receptor autophosphorylation but also dimerization (data not shown). The results from growth factor-independent growth of both BaF3-NOK and BaF3-EPOR/NOK also strongly indicated that NOK might function in a constitutively active form in either the cytoplasm or cell membrane. Hubbard et al. (6) proposed that the equilibrium was present between the inactive versus active state in the receptor activation loop. For example, in both FGFR1 and the insulin receptor, the tyrosine residues in the activation loop of the undimerized receptor can occlude the substrate access to the active site, whereas receptor dimerization releases the autoinhibition by shaping a new receptor conformation by reposition the activation loop away from the active site resulting in receptor autophosphorylation (6 , 20) .
Therefore, abnormal expression of RPTK may shift or disrupt the equilibrium to a constitutively active direction. Studies indicate that activating point mutation in the activation loop releases receptor autoinhibition, leading the receptor to become constitutively active, which has accounted for numerous human cancers and genetic disorders (6) . The more active nature of NOK may suggest that the equilibrium may more often shift toward the active state for this novel oncogene. We speculate that this receptor-like oncogene lacking the almost complete extracellular domain may have an inherent advantage in forming a special protein structure resulting in opening up the otherwise occluded active site for substrate access. The receptor activation may further strongly activate downstream signal pathways, which may result in higher proliferation potential and the acceleration of malignant transformation and cancer development.
Studies have shown that tumor-related and mitogenic signaling pathways such as PI3K and RAS/MAPK are predominantly active in tumor cells with invasive and metastatic characters (21, 22, 23, 24) . In current study, stable BaF3 cells expressing either NOK or EPOR/NOK could sustain in a harsh growth condition without the supplement of any growth factor and serum for about 10 days, indicating that mitogenic signals were highly potentiated by the overexpression of NOK oncogene. There are multiple sites for tyrosine phosphorylation in NOK. Among them, seven are in the kinase domain and two are in the COOH-terminal tail. These residues may serve as docking sites for intracellular substrate binding to turn on downstream signaling cascades. The known examples can be found in the type I insulin-like growth factor receptor and plasmingen-related growth factor receptor (PRGFR), both of which play important roles in tumor cell invasion and metastasis (25 , 26) . In addition, there are also many serine and threonine residues presenting in the intracellular portion of NOK, which may serve as extra phosphorylation sites to potentiate the kinase molecule to a more hyperphosphorylated state. Using specific protein kinase inhibitors, we have shown that the activation of both PI3K and RAS/MAPK pathways was involved in cellular transformation induced by NOK (Fig. 5) ⇓ . Those pathways may be interrelated and interdependent because blocking either pathway is sufficient for the inhibition of colony formation. These results are consistent with other reports showing that concomitant activation of both PI3K and RAS/MAPK is required for tumor cell invasion and metastasis. The examples of such cases have been documented for Met (HGF/SF receptor) and type I insulin-like growth factor receptor (4 , 27) .
The most important finding of the current study is that NOK is capable of eliciting tumor cell invasion and metastasis. Probably due to this unique capability, the life span of the tumor-injected mice was significantly reduced, and the mice often died early accompanied by severe physiological and pathological abnormalities. The malignancy induced by NOK transformed BaF3 cells seemed to be very aggressive. It could penetrate locally and invaded to various distant organs such as the spleen, liver, kidney, and skeletal muscle within a short period of time. Pathological analysis on tissue sections revealed the manifest and prevalent mitotic figures within the invaded tumor cells. This analysis additionally highlighted the aggressive character of this novel oncogene during cancer development.
In conclusion, functional studies on NOK strongly indicate that this new member of the RPTK family may provide a novel model system for studies of RPTK-induced cellular transformation and alterations. This would be a critical step for understanding the transition from benignity to cancer and metastasis.
We thank Prof. Jiancun Hou from Peking Union Medical College for discussion, Drs. Quancai Cui and Hebin Song from Peking Union Medical College Hospital for photographs, Xiping Chen and Yiwei Lin from Peking Union Medical College for tissue section preparations, Yunfeng Zhang from Peking University Medical Center for animal assistance, and Drs. Susan Zong and Luhai Wang from Mount Sinai School of Medicine for providing technique information and staining chemical.
Grant support: National Key Basic Research Program of China Grants 2001CB510006 and 2002CB513007, Key Program Project of National Nature Science Foundation of China Grant 30030050, The 863 Program Grant 2001AA217141, 985 Project of Tsinghua University, and a special fund from Tsinghua Yuanxing Pharmaceuticals, Shenzhen, China.
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Requests for reprints: Xin-Yuan Fu, Department of Pathology, Yale University School of Medicine, New Haven, CT 06520-8023. E-mail: [email protected]; or Li Liu, Tsinghua Institute of Genome Research, Department of Biological Sciences and Biotechnology, Tsinghua University, Beijing 100084, China E-mail:
- Received July 14, 2003.
- Revision received February 17, 2004.
- Accepted March 1, 2004.
- ©2004 American Association for Cancer Research.