Regulation of cell proliferation by autocrine motility factor/phosphoglucose isomerase signaling.

Autocrine motility factor (AMF)/phosphoglucose isomerase (PGI; EC 5.3.1.9) is a housekeeping cytosolic enzyme that plays a key role in both glycolysis and gluconeogenesis pathways. AMF/PGI is also a multifunctional protein that displays cytokine properties, eliciting mitogenic, motogenic, and differentiation activities, and has been implicated in tumor progression and metastasis. Because little is known about AMF/PGI-dependent signaling in general and during tumorigenesis in particular, we sought to study its effect on the cell cycle. To elucidate the functional role of PGI, we stably transfected its cDNA into NIH/3T3 and BALB/c 3T3-A31 fibroblasts. Ectopic overexpression of PGI results in the acquisition of a transformed phenotype associated with an acceleration of G1 to S cell cycle transition. These were manifested by up-regulation of cyclin D1 expression and cyclin-dependent kinase activity and down-regulation of the cyclin-dependent kinase inhibitor p27Kip1. The reduced p27Kip1 protein expression level in PGI-overexpressing cells could be restored to control levels by treatment with proteasome inhibitor. PGI-overexpressing cells also exhibited elevated expression of Skp2 involved in p27Kip1 ubiquitination and elevation in the levels of retinoblastoma protein hyperphosphorylation. Thus, we may conclude that the overexpression of AMF/PGI enhances cell proliferation together with up-regulation of cyclin/cyclin-dependent kinase activities and down-regulation of p27Kip1, whereas the induction of 3T3 fibroblast transformation by PGI is regulated by the retinoblastoma protein pathway.

tumor-secreted cytokine originally identified by its ability to induce tumor cell migration via a unique cognate 78,000 kDa (gp78) seven-transmembrane glycoprotein receptor (autocrine motility factor receptor (AMFR)) (5-7). AMF/PGI is a secreted protein found in the conditioned medium of transformed cells (5,8) as well as in the serum and urine of cancer patients (9,10). Overexpression of AMF/PGI and its receptor have been found in a wide spectrum of malignancies and are associated with migration-dependent processes during cancer progression, metastasis (11)(12)(13), and angiogenesis (14).
Earlier studies have shown that AMF/PGI treatment of A31 fibroblasts stimulate cell growth and overexpression of PGI by NIH/3T3 cells, cell proliferation, and motility (15,16). These findings suggest that the AMF/PGI exhibits a growth factorlike activity, implying that this signaling pathway is linked to cell cycle augmentation. Perturbed control of the G 1 phase of the cell cycle is a critical step for cellular transformation and tumorigenesis (17) during which cells respond to signals by either advancing toward another division or withdrawing from the cycle into a resting state (17). Passage through the G 1 phase and entry into S phase are controlled by cyclin-dependent kinases (CDKs) that are sequentially regulated by cyclins D, E, and A (17). D-type cyclins bind to and activate CDK4 and CDK6 during G 1 phase (18) followed by activation of CDK2 in complex with cyclin E in the late G 1 phase, which is essential for initiation of the S phase. CDK2 also binds to cyclin A during S phase, playing a critical role in DNA replication (17). Specifically the INK4-type inhibitors, consisting of p16Ink4a, p15Ink4b, p18Ink4c, and p19Ink4d, regulate the activities of CDK4 and CDK6. On the other hand the Kip/Cip-type inhibitors inhibit the CDK2 type, comprised of p21 Cip1 , p27 Kip1 , and p57 Kip2 (19). The retinoblastoma protein (Rb) is a critical target protein that is phosphorylated via these cyclin-CDK complexes (20) and controls gene expression mediated by the E2F transcriptional regulators, which activate genes essential for transition from G 1 to S phase (20).
Herein we identify downstream targets of PGI whose activities may be important for the cell cycle. Moreover we provide evidence that 1) cyclin D1 is a significant G 1 to S transition control factor whose level is modulated by the PGI signaling pathway, 2) CDK inhibitory protein p27 Kip1 proteasome-dependent degradation is controlled by PGI, and 3) show that Rb regulates the induction of 3T3 fibroblasts transformation by AMF/PGI overexpression.
Histone H1 was obtained from Upstate Biotechnology (Lake Placid, NY), and GST-Rb 769 was from Santa Cruz Biotechnology (Santa Cruz, CA).
Parental A31 cells were transfected with pcDNA3.1 zeo or pcDNA 3.1 zeo-PGI using LipofectAMINE according to the manufacturer's instructions (Invitrogen). Isolation of single clones of the stable transfectants was accomplished by adding 750 g/ml Zeocin (Invitrogen) to the culture medium. The A31 cell line, stably transfected with pcDNA3.1 zeo or pcDNA3.1 zeo-PGI, was designated A31zeo or A31PGI, respectively. All experiments were repeated at least three times, and results were confirmed using both clonal cell lines and the pooled cell population.
Cell Proliferation Assay-Cell proliferation assays were performed by seeding cells at a density of 1 ϫ 10 5 cells/well in 6-well plates. Cells were fed DMEM with 10% FBS every other day, and the number of cells were manually counted with a hemocytometer.
Cell Cycle Synchronization and DNA Content Analysis-Cell cycle phase distribution was determined by flow cytometry of propidium iodide-stained cells. Whole cell suspensions were washed in phosphatebuffered saline (PBS), fixed in 70% ethanol, stained in 50 g/ml propidium iodide, 1 mg/ml RNase, 0.1% Triton X-100, and analyzed with a BD Biosciences.
For cell synchronization, exponentially growing cells were treated for 18 h with 100 ng/ml nocodazole to induce G 2 -M arrest (22). Mitotic cells were collected by gentle pipetting and were reseeded into fresh DMEM with 10% FBS. At various times after plating, cells were collected, and their cell cycle distribution was determined by flow cytometry as described above.
Western Blot Analysis-The cells were lysed in lysis buffer (50 mM HEPES, pH 7.9, 0.1% Nonidet P-40, 10% glycerol, 1 mM dithiothreitol, 1% sodium deoxycholate, 250 mM NaCl, 5 mM EDTA, 50 mM NaF, 0.1 mM sodium vanadate, 1 mM phenylmethylsulfonyl fluoride, 50 mM ␤-glycerophosphate, and 0.1 mg/ml leupeptin) at 4°C. Cell lysates containing equal amounts of protein were separated by SDS-PAGE and transferred to a polyvinylidene fluoride membrane (MSI, Westborough, MA). The membranes were blocked with 5% nonfat dry milk in Trisbuffered saline with 0.05% Tween 20 (TBS-T) incubated with primary antibody for 2 h, washed 3 times for 15 min in TBS-T, incubated with the secondary horseradish peroxidase-conjugated antibody (Zymed Laboratories Inc., San Francisco, CA) for 1 h, and finally washed 3 times. The horseradish peroxidase activity was detected by an incubation of the membrane with enhanced chemiluminescence reagent (Amersham Biosciences). A Kodak imaging system determined the density of the bands.
RNA Isolation and Reverse Transcription (RT)-PCR-Total cellular RNA was isolated according to the manufacturer's instructions. Using TRIzol reagent (Invitrogen). RT-PCR analysis was performed as described (23). For cyclin D1 cDNA, 20 amplification cycles (93°C for 1 min, 55°C for 1 min, 72°C for 1 min) were performed with the following primers: mouse cyclin D1, 5Ј-CTGACACCAATCTCCTCAACGAC-3Ј (forward) and 5Ј-GCGGCCAGGTTCCACTTGAGC-3Ј (reverse); glyceraldehyde-3-phosphate dehydrogenase, 5Ј-CCATGCCATCACTGCCACC-CAGAA-3Ј (forward), and 5Ј-GTCCACCACCCTGTTGCTGTAGCCG-3Ј (reverse). In all cases, PCRs were shown to be in a linear range by performing parallel control PCRs with increasing template cDNA concentrations. To verify the absence of contamination of RNA samples with DNA, we performed the PCR on samples that were processed identically to the target samples but were not reverse-transcribed. After RT-PCR, samples were electrophoresed on a 1.5% agarose gel containing 10 g/ml ethidium bromide, and the intensity of the bands was measured by a Kodak Digital Science Image System 44°C.
P27Kip1 Stability in Vivo-Because p27 Kip1 accumulates in serumstarved and density-arrested cells (24), we examined the p27 Kip1 protein changes under these conditions. For the serum dependence assay, cells were grown to 50% confluency in DMEM with 10% FBS. Then the medium was changed to serum-free DMEM and maintained without FBS for 24 h. For cell-cell contact dependence assay, cells were grown to near confluency (95-100%) in DMEM with 10% FBS.
To evaluate the changes in p27 Kip1 protein stability, cultured cells were treated with cycloheximide (30 g/ml) for 2 or 4 h to block total cellular protein synthesis. The cells were incubated with proteasome inhibitors lactacystin (5 M) for an appropriate time in experiments designed to assay proteasome targeting of p27 Kip1 . Cells lysates prepared after each incubation period were analyzed by Western blotting with p27 Kip1 antibody as described above.
Immunofluorescence Microscopy-Cells were seeded on coverslips in 6-well plates at a density of 1 ϫ 10 4 cells/well in DMEM containing 10% FBS. The next day, the medium was changed to fresh DMEM with or without 10% FBS. After 36 h incubation, cells were fixed in 4% paraformaldehyde/PBS for 10 min and permeabilized with 0.1% Triton X-100/PBS for 15 min. The cells were blocked with 1.0% bovine serum albumin in PBS for 30 min at 4°C and then incubated with anti-p27 Kip1 antibody (1:100 dilution in 0.1% bovine serum albumin in PBS) at 4°C for 2 h. After 3 washes with 0.05% Triton X-100/PBS, cells were incubated for 1 h with fluorescein isothiocyanate-conjugated anti-rabbit IgG diluted 1:200 in 0.05% Triton X-100/PBS. After three washes with PBS, coverslips were then mounted on a glass slide with a drop of SlowFade reagent (Molecular Probes, Eugene, OR). Immunofluorescence was recorded with a Sony digital CCD camera (DXC-970MD) mounted on an Olympus BX40 microscope.
In Vitro CDK Assay-CDK2 and CDK4 kinase assays were per-

FIG. 1. Effect of PGI on cell growth.
A, expression of PGI in transfected clones. Cell lysates from parental, vector-only transfected, and PGI-transfected cells were subjected to Western blot analysis with anti-PGI antibody (top row) or anti-AMFR antibody (second row). The third row was probed with anti-actin antibody as a control. The secretion of PGI was analyzed by Western blotting of 50 g of protein from conditioned medium (bottom row) and probed with anti-PGI antibody. First lane, parental NIH/3T3; second lane, empty vector-transfected NIH/3T3; third through fifth lanes, PGI-overexpressing NIH/3T3 clone 1, 2, and 3, respectively; sixth lane, parental A31; seventh lane, empty vector-transfected A31; eighth through tenth lanes, PGI-overexpressing A31 clone 1, 2, and 3, respectively. B, growth properties of PGI-overexpressing NIH/3T3 and A31 cells. Cells were grown in medium containing 10% FBS, and cell numbers were then determined. Each bar represents the mean of triplicate determinations Ϯ S.D. Similar results were obtained in three independent experiments. Left panel: E, parental NIH/3T3; ‚, empty vector-transfected NIH/3T3; q, PGI-overexpressing NIH3T3 clone 1; OE, PGI-overexpressing NIH3T3 clone 2; f, PGI-overexpressing NIH3T3 clone 3. Right panel: E, parental A31; ‚, empty vector-transfected A31; q, PGI-overexpressing A31 clone1; OE, PGI-overexpressing A31 clone 2; f, PGI-overexpressing A31 clone 3. C, the cell cycle distribution of PGI-transfected cells. Parental, empty vector-transfected, and PGI-overexpressing cells were cultured in the presence of 10% serum for 3 days, and then their cell cycling profiles were analyzed by flow cytometry. Similar results were obtained in at least three independent experiments and from other PGI-overexpressing clones. y axis, cell number; x axis, DNA content.
formed as described previously (18). Briefly, 500 g of protein extracts were immunoprecipitated with 2 g of the anti-CDK2 or anti-CDK4 antibodies for 60 min at 4°C. Immunoprecipitated proteins were collected on protein A-Sepharose® 6MB (Amersham Biosciences). Kinase reactions were performed for 30 min at 30°C in kinase assay buffer (50 mM HEPES, pH 7.2, 10 mM MgCl 2 , 2.5 mM EGTA, 0.1 mM NaF, and 0.1 mM Na 3 VO 4 ) and contained 20 M [␥-32 P]ATP at a specific activity of 10 Ci/mmol and 1 g of histone H1 for CDK2 assays or 1 g of GST-Rb 769 for CDK4 assays reactions. Reaction products were resolved by SDS-PAGE. The gels were stained in Coomassie Blue, dried, and exposed to film. A Kodak imaging system determined the density of the bands. Normal rabbit immunoglobulin G (Sigma) was used as a negative control in immunoprecipitation experiments.
Statistical Analysis-Associations between the variables were tested by Student's t test or Fisher's exact test. All statistical differences were deemed significant at the level of p Ͻ 0.05.

RESULTS
PGI Shortens the G 1 Interval-We have established two different types of murine fibroblast based cell lines (NIH/3T3 and A31) stably transfected and overexpressing AMF/PGI. Three clones exhibiting high level expression of PGI in both NIH/3T3 and A31 cells were selected. The ratios of PGI expression of each clone were 5 to 6 compared with the empty vector-transfected control (Fig. 1A). PGI secretion was restricted to the PGI-transfected cells and could not be detected in the conditioned medium of the parental and empty vector-transfected cells (Fig. 1A). All AMF/PGI-overexpressing clones grew ϳ2fold faster than those of parental or empty vector-transfected control cells (Fig. 1B). The effect of ectopic expression of PGI on cell cycle distribution was determined next. There was no spe-cific accumulation of the PGI-overexpressing cells at any cell cycle phase relative to the control cells (Fig. 1C). The percentage of G 1 phase cells was slightly lower, and the proportion of the S phase cells was respectively higher in the PGI-overexpressing cells, but this did not change significantly. To study the effect of PGI on cell cycle progression, the length of G 1 phase was determined. Cells were treated with the mitotic inhibitor nocodazole for 18 h followed by shaking and replating in nocodazole-free medium. The cells were collected at the indicate time points after the mitotic shake, and the DNA profiles are shown in Fig. 2. At 8 h after replating, PGI-overexpressing cells shifted toward S phase (increased DNA content) as compared with parental and empty vector-transfected control cells ( Fig. 2A, bottom panel, arrow). The presence of S phase in control cells could not be detected prior to 10 h after cells were replated (top and second panels). Similar results were obtained from two other PGI-overexpressing NIH/3T3 clones, parental A31 cells, empty vector-transfected A31 cells, and three PGI-overexpressing A31 clones (Fig. 2B). These results indicate that PGI-overexpressing cells pass through G 1 phase at a shorter rate than control cells.
The Expressions of Cell Cycle-related Proteins in PGI Overexpressing Cells-To further understand the effect of PGI on cell cycle regulation, we examined the expression levels of cell cycle regulators during an exponential cell growth phase and searched for any possible variations in the expression of cell cycle components at different time points. We have found that the level of cyclin D1 increased 3-4-fold in each of the PGI- overexpressing NIH/3T3 and A31 cell clones as compared with the respective control cells (Fig. 3A). The levels of cyclin A, cyclin E, and p53 remained essentially unchanged in the PGIoverexpressing cells as compared with the controlled cells. For a more sensitive analysis, we performed RT-PCR analyses of the expression levels of cyclin D1 and confirmed that cyclin D1 mRNA was increased 3-fold in the PGI-overexpressing cells as compared with their normal counterparts (Fig. 3B). Furthermore, in contrast to p21 Cip1 , the level of p27 Kip1 was decreased in each PGI-overexpressing NIH/3T3 and A31 clones (Fig. 3A) by ϳ50 -60% relative to the control cells. The above-mentioned pattern of the cell cycle components change in expression in the PGI-overexpressing cells was the same, at 2 and 3 days of culture.
PGI Promotes Proteasome-dependent Degradation of p27 Kip1 Protein-Because of the fact that p27 Kip1 increases in serumstarved and contact-inhibited normal cells (24), we questioned the status of p27 Kip1 in PGI-transfected cells under diverse culture conditions. Cells were either grown to confluency or cultured as serum-starved for 24 h at 37°C. P27 Kip1 expression analysis revealed that in PGI-overexpressing cells it was significantly reduced (60 -70%) compared with the cells both in density-arrested and in serum-starved conditions (Fig. 4A). Next we examined whether the reduced p27 Kip1 protein level was due to change in protein stability by exposing cells to cycloheximide treatment. P27 Kip1 protein was rapidly degraded (80 -90% reduced) in cycloheximide-treated PGI-overexpressing NIH/3T3 and A31 cells, whereas it was more stable in control cells at 4 and 8 h (Fig. 4B). Altered regulation of p27 Kip1 protein stability was probably the major cause of its reduced protein expression in PGI (overexpressing) cells. Thus, we questioned whether p27 Kip1 degradation processed was dis- rupted by the 26 S protease system inhibitor in PGI-overexpressing cells. To address this, we used the 26 S proteasespecific inhibitor lactacystin and found that lactacystin treatment restores p27 Kip1 expression level in PGI-overexpressing cells and to that of the control cells (Fig. 4C). The F-box protein is the substrate-specific recognition component of Skp1-Cul1-F-box (SCF) ubiquitin-protein ligase complexes that is used to target specific proteins for degradation (25). Thus, we tested whether PGI alters the expression of Skp2, a member of the F-box family, leading to an increased ubiquitination and subsequent degradation of p27 Kip1 protein. As shown in Fig.  4D, Skp2 protein level was slightly elevated in PGI-overexpressing cells. The degradation of p27 Kip1 was partially impaired by the anti-PGI antibody treatment, suggesting that PGI-signaling inhibition might impact the degradation process of p27 Kip1 (data not shown). Up-regulation of Skp2 expression was inhibited by anti-PGI IgG (Fig. 4D).
Next, the subcellular localization of p27 Kip1 was determined by anti-p27 Kip1 indirect immunofluorescent staining. No differences in the distribution of p27 Kip1 localization were detected FIG. 5. Subcellular localization of p27 Kip1 in PGI-overexpressing cells. Cells were serum-starved for 36 h, fixed in 4% paraformaldehyde, and incubated with anti-p27 Kip1 antibody. In asynchronously growing parental, empty vector-transfected, and PGI-overexpressing NIH/3T3 and A31 cells, p27 Kip1 was detected in the nucleus and cytoplasm (data not shown). p27 Kip1 was expressed mainly in the nucleus during serum starvation in parental and empty vector-transfected cells. In contrast, no accumulation of p27 Kip1 was observed in the nucleus of PGI-overexpressing cells during serum starvation. a, parental NIH/3T3; b, empty vector-transfected NIH/3T3; c, PGI-overexpressing NIH/3T3 clone 1; d, parental A31; e, empty vector-transfected A31; f, PGI-overexpressing cells A31 clone 1. Similar results were obtained from other PGI-overexpressing clones.

FIG. 6. CDK activities and phosphorylation of the Rb protein. A, expression of CDK2 and CDK4 protein in PGI-overexpressing cells.
Equalized protein samples were subjected to Western blot analysis with antibodies to CDK2 and CDK4. No differences in expression of CDK2 and CDK4 were observed between PGI-overexpressing and control cells. The bottom row was probed with anti-actin antibody as a control. B, total CDK2 and CDK4 activities in the PGI-overexpressing NIH/3T3 and A31 versus control cells. Equal amounts (0.5 mg) of total proteins from the cell lysates were immunoprecipitated with 2 g of anti-CDK2 and anti-CDK4 antibodies. The in vitro immunocomplex kinase assays were performed with histone H1 or GST-Rb fusion protein as the substrate, as described under "Experimental Procedures." The reaction products were resolved on SDS-PAGE and exposed to x-ray film. C, the phosphorylation status of Rb in the PGI-overexpressing and control cells. First lane, parental NIH/3T3; second lane, empty vector-transfected NIH/3T3; third through fifth lanes, PGI-overexpressing NIH/3T3 clone 1, 2, and 3, respectively; sixth lane, parental A31; seventh lane, empty vector-transfected A31; eight through tenth lanes, PGI-overexpressing A31 clone 1, 2, and 3, respectively.
(data not shown) in asynchronously growing parental, empty vector-transfected, and PGI-overexpressing cells. Whereas (Fig. 5) serum starvation of parental and empty vector-transfected cells led to a clear increase of p27 Kip1 nuclear localization (Fig. 5, a, b, d, and e), no accumulation of p27 Kip1 was observed in the nucleus of PGI-overexpressing cells during serum starvation (Fig. 5, c and f).
Alterations of CDK2 and CDK4 Activity in PGI Overexpressing Cells-To continue the analysis we tested the level of CDK2 and CDK4 protein expression and found no differences between PGI-overexpressing and control cells (Fig. 6A). Next, we analyzed the CDK activities in both PGI-overexpressing and control cells. Cell lysates were immunoprecipitated with anti-CDK2 and anti-CDK4 antibodies, and the kinase activities of the immunocomplexes were determined with [␥-32 P]ATP and histone H1 or GST-Rb fusion protein as a substrate. Total kinase activity of CDK2 and CDK4 was elevated in each PGIoverexpressing NIH/3T3 and A31 clones compared with control cells (Fig. 6B). The ratios of CDK activity, determined by densitometry, were 1.9 -2.8 (CDK2) and 2.1-2.7 (CDK4), respectively, as compared with control cells. These results imply that PGI overexpression leads to up-regulation not only of cyclin D1 expression level but also of its associated kinase activities.
Cyclin D1/pRB/E2F Pathway Activated in PGI-transfected Cells-The Rb pathway is essential for the formation of numerous tumors (20), and hypophosphorylated Rb binds to a subset of E2F complexes, converting them to repressors that constrain expression of E2F target genes (26). Phosphorylation of Rb frees these E2Fs, enabling them to transactivate the same genes, a process initially triggered by the CDKs and then accelerated by the cyclin E-CDK2 complex (20). Thus, we questioned the status of Rb in PGI-overexpressing cells. The cells displayed an increase in the hyperphosphorylated form of Rb, which migrated in the gel more slowly than the hypophosphorylated form when compared with the parental and empty vector-transfected cells (Fig. 6C). This suggests that the Rb plays a role in the transformation induction in PGI overexpression cells. DISCUSSION PGI is a ubiquitous cytosolic enzyme that plays a key role in both glycolysis and gluconeogenesis pathways (1) and, therefore, a housekeeping gene transcribed in all cells. PGI is upregulated in a variety of human cancer cells (11)(12)(13) activated by ras (5) and myc (27) and, as a result, acts as a transforming agent (16). We examined the cell cycle-associated activities related to PGI overexpression in fibroblasts. The results presented here and summarized in Fig. 7 suggest that AMF/PGI is involved in the transition from G 1 to S phase during cell cycle progression, which might explain its role(s) in the modulation of tumor growth. Initially, PGI up-regulates cyclin D1 mRNA and protein expression levels. Of note, cyclin D1 is frequently overexpressed in human cancers, such as parathyroid adenoma, lymphoma, and breast cancer (28), and overexpression of cyclin D1 leads to a shortened duration of the G 1 phase and reduces serum dependence in fibroblasts (18, 29 -31). The activity of CDK4 was also found to be elevated in both the PGIoverexpressing NIH/3T3 and A31 cells. A similar up-regulation of CDK4 activity has been observed in NIH/3T3 cells by transformed ras (32) and myc (33), suggesting that an increase in cyclin D1-CDK4 activity may be a relatively common event in cellular transformation. Cyclin D1 promoter activity and mRNA levels are induced by various growth factors, including epidermal growth factor, platelet-derived growth factor, fibroblast growth factor, hepatocyte growth factor, keratinocyte growth factor, and insulin-like growth factor-1 (34 -40). These growth factors and their receptors activate Ras family mem-bers to mediate a signal transduction cascade of successive phosphorylation steps, leading to the activation of mitogenactivated protein kinase (41). It has been demonstrated that the Ras-mitogen-activated protein kinase pathway plays a pivotal role in cyclin D1 synthesis and assembly with CDK4, and nuclear retention of the enzymes depends on Ras-mitogenactivated protein kinase and phosphatidylinositol 3-kinase (PI3K)/Akt signaling (42). Glycogen synthase kinase-3␤ can phosphorylate cyclin D1 to trigger its nuclear export and proteasomal degradation (19). PI3K and Akt kinase negatively regulate glycogen synthase kinase-3␤ to enhance the stability of the cyclin D-dependent kinase (19).
Previously, we have shown that ectopic expression of PGI induced activation of the PI3K/Akt-signaling pathway (16). Indeed, the PI3K inhibitor LY294002 nearly abolishes up-regulation of cyclin D1 protein levels induced by PGI overexpression in NIH/3T3 and A31 cells (data not shown). Thus, it may be concluded that up-regulation of cyclin D1 could be attributed in part to the PGI activation of the PI3K/Akt pathway.
In PGI-overexpressing cells p27 Kip1 levels are decreased, FIG. 7. Model for cell cycle regulation by PGI through cyclin D1 up-regulation and p27 Kip1 degradation. First, PGI causes upregulation of activity of cyclin D1-CDK4 complexes that phosphorylates Rb. This phosphorylation allows the accumulation of E2Fs that activate the transcription of a large number of genes essential for DNA replication as well as further cell cycle progression. Moreover, cyclin D1-CDK4 complexes bind the Cip/Kip family, whose role is to inhibit cyclin E-CDK2 activity, followed by cyclin E-CDK2 activation. Both cyclin D-CDK4 and cyclin E-CDK2 then collaborate to sequentially phosphorylate Rb. Among the known E2F target genes is cyclin E, whose transcriptional up-regulation provides positive feedback to drive cells into S phase. Up-regulation of cyclin D1 levels by PGI signaling is explained by glycogen synthase kinase-3␤ (GSK-3␤) inactivation. Glycogen synthase kinase-3␤ negatively regulates cyclin D1 expression through transcriptional control of its gene and the control of its degradation by the proteasome. Upon PGI stimuli, the function of glycogen synthase kinase-3␤ is inhibited by its PI3K/Akt-mediated phosphorylation. Second, PGI deregulates Skp2 protein level followed by proteasomal degradation of p27 Kip1 cyclin-dependent kinase inhibitor. p27 Kip1 stability is regulated by the F-box protein Skp2, a component of the ubiquitin E3 ligase SCF Skp2 . Phosphorylation of p27 Kip1 on Thr-187 by active cyclin E-CDK2 complexes creates a binding site for Skp2. Ubiquitination of p27 Kip1 by SCF Skp2 results in degradation of p27 Kip1 by the proteasome and causes cell cycle progression (the SCF complex is composed of Skp1, Cul1, Rbx1, and Skp2). Finally, PGI up-regulate small GTPases RhoA and Rac1. The RhoA down-regulates the p27 Kip1 . The Rac1 stimulates transcription of the cyclin D1 promoter and Rb phosphorylation. Taken together, PGI activates the Rb pathway through small GTPases signal.
which may lead to the inhibition of all CDKs activities (19). A similar decrease in the p27 Kip1 level has been found in cells transformed by the ras (43), myc (44,45), and v-Src oncogenes (46). In addition, down-regulation of p27 Kip1 is frequently observed in various human cancers, such as prostate, breast, non-small cell lung, and colon carcinomas (47,48). Nuclear, p27 Kip1 inhibits cyclin E-CDK2 activity, and export of p27 Kip1 to the cytoplasm is usually a prerequisite for its degradation. Relegated to the cytoplasm, p27 Kip1 is unable to control CDKs (49). Recently, it was reported that Akt promotes phosphorylation of p27 Kip1 on a threonine residue in its nuclear localization signal, impeding nuclear entry of p27 Kip1 (50). We found that PGI-transfected NIH/3T3 and A31 cells both expressed low levels of p27 Kip1 in the nuclei related to the control cells in a serum-deprived condition. It is possible that PGI regulates the nuclear export of p27 Kip1 to the cytoplasm through the PI3K/Akt-signaling pathway. The major regulatory machinery of p27 Kip1 protein levels is posttranslational ubiquitin-mediated proteolysis (25,51). We found that p27 Kip1 stability is decreased by PGI overexpression through the increase in ubiquitin-dependent degradation by the 26 S proteasome pathway. P27Kip1 degradation is by an SCF-type ubiquitin ligase complex (25), and Skp2 is a member of the F-box family of the specific substrate recognition subunit of SCF ubiquitin-protein ligase complexes. Expression of Skp2 was required for the ubiquitination and subsequent degradation of p27 Kip1 (52). Skp2 expression is found to be elevated in tumor cells (53,54), and the level of p27 Kip1 was reported to be inversely related to that of Skp2 in both squamous cell (55) and colorectal carcinomas (56). Here, we show that the level of Skp2 protein was increased in PGI-overexpressing cells. A decreased level of p27 Kip1 expression in PGI-overexpressing cells may be caused by increased expression of Skp2, which targets p27 Kip1 for degradation. The tumor suppressor, PTEN, was shown to regulate ubiquitin-dependent degradation of p27 Kip1 through the ubiquitin ligase SCF Skp2 (57) and to negatively control the PI3K-signaling pathway required for cell growth and survival (58). Skp2 may function as a critical component in the PTEN/ PI3K pathway for the regulation of SCF Skp2 and cell proliferation. Down-regulation of p27 Kip1 may be mediated by PGI through PTEN/PI3K/Akt-signaling pathways. AMFR is a RING finger-dependent ubiquitin protein ligase of the endoplasmic reticulum (59). AMFR mediates degradation of CD3-␦, a well characterized endoplasmic reticulum-associated degradation substrate, and specifically recruits MnUBC7, a ubiquitin-conjugating enzyme, implicated in endoplasmic reticulum-associated degradation (59). Therefore, AMFR is not a likely candidate of ubiquitin ligase for p27 Kip1 .
The important contribution of Ras aberrant activation in oncogenesis is well established (60). It interacts with a diverse spectrum of effectors. Moreover, it initiates a multiple cytoplasmic signaling cascades including up-regulation of PGI, cyclin D1, Rho GTPases, and down-regulation of p27 Kip1 expression (5,42). The Rho GTPases, RhoA, Rac1, and Cdc42 are well established regulators of cytoskeletal organization (61). The RhoA protein has been implicated in down-regulation of the p27 Kip1 (62). The Rac1 protein has been reported to stimulate transcription of the cyclin D1 promoter (63) and can stimulate Rb phosphorylation (64). Aberrant activation of Rho GTPases can also affect cell proliferation and transformation (42). In fact, we have recently demonstrated Rho GTPases RhoA and Rac1 to be up-regulated by PGI (65). Taken together, it appears that the Ras-signaling pathway regulates PGI to activate cyclin-CDK complexes and the Rb pathway.
In most human cancers inactivation of the regulatory pathways of the cell cycle, the Rb and p53 pathways occurs. Rb is a target of the cyclin D-and E-dependent kinases. In this study, we found an increase in the phosphorylation of Rb in PGIoverexpressing cells. Nevertheless overexpression of PGI may affect the cell cycle in multiple ways, all of which may contribute to transformation. PGI may transform fibroblasts through the CDK/Rb/E2Fs pathway. Now, it is pertinent to identify additional targets and examine their role in PGI regulation since the PGI-signaling pathway(s) may represent novel targets for cancer therapy.