• E-mail
• Print
• Comment
• Font Size
• Digg
• del.icio.us
• Discuss article

Stimulation of Quiescent Cells By Individual Polypeptide Growth Factors is Limited to One Cell Cycle

Posted on: Saturday, 13 November 2004, 03:00 CST

Fibroblast growth factor; Cell cycle; Cyclins; Senescence; Platelet-derived growth factor

Since little is known about the function of polypeptide growth factors as regulators of multiple cell cycles, we compared the ability of FGF1, PDGF-AB and serum to induce a second round of DNA synthesis in Swiss 3T3 cells previously exposed to either FGF1, PDGF- AB or serum during the first cell cycle using [^sup 14^C]- and [^sup 3^H]thymidine in a double labeling system to distinguish between the first and second cell cycles. Surprisingly, we observed that cells exposed to either FGF1 or PDGF-AB in the first cell cycle were unable to synthesize DNA in response to FGF1 or PDGF-AB in the second cell cycle; yet these cells responded well to serum as a second cycle mitogen. Interestingly, while cells exposed to either FGF1 or PDGF-AB in the second cycle displayed normal receptor- mediated signaling and expressed cyclin D and E, they, like senescent fibroblasts and endothelial cells, failed to express cyclin A, and the continuous exposure of cells to either FGF1 or PDGF-AB resulted in a decrease in the kinase activity of the cyclin E/cdk2 complex. In addition, an increased association of this complex was observed with p21 CIP in an FGF1-dependent manner as well as with p27 KIP in a PDGF-AB-dependent manner. Lastly, the downregulation of p21 expression using an antisense strategy was able to partially rescue the replicative response of Swiss 3T3 cells to FGF1 in the second cycle. These data suggest that (i) FGF1 and PDGF-AB may limit their mitogenic effect to a single cell cycle, (ii) entry into the second round of replication is serum dependent and (iii) the self-limiting nature of FGF1 and PDGF-AB correlates with the accumulation of the cdk inhibitors, p21 and p27, respectively.

Abbreviations. BCS Bovine calf serum. - cdk Cyclin-dependent kinase. - CIP cdk inhibitor protein. - DMEM Dulbecco's modified Eagle's medium. - FGF Fibroblast growth factor. - KIP CD kinase inhibitor protein. - PDGF Platelet-derived growth factor. - PGF Polypeptide growth factor. - Q Quiescence. - R Receptor. - Rb Retinoblastoma. - S Serum.

Introduction

In spite of impressive achievements in the understanding of the biochemical mechanisms underlying the proliferative effects of polypeptidc growth factors (PGF), a basic paradox of PGF activity remains unsolved and rarely addressed. Indeed, although individual PGFs are efficient stimulators of DNA synthesis when applied to quiescent cells, they fail to support the serial propagation of cells in vitro. Certain cell lines and strains can be propagated in defined media containing complex cocktails of various growth and survival factors but scrum is still a standard cell culture media supplement (Freshney, 2000). Moreover, although serum contains functional levels of PDGF as a result of platelet lysis, many normal diploid cell types such as endothelial cells (Maciag et al., 1981a) and keratinocytes (Maciag et al., 1981b) require the supplementation of serum with FGF for their serial propagation. A report described a continuous proliferation of L6 myoblast cells stimulated with FGF; however, those cells overexpressed FGFR (Shaoul et al., 1995). Since serum is a non-physiologic fluid of undefined and variable composition, it is unclear to what extent the results obtained in cell culture experiments could be applied to in vivo situations when individual PGFs are locally released either during development or inflammation in response to injury. This is a potentially significant issue since the clinical utility of PGFs as replicative agents of tissue repair is currently under intense investigation (Schratzberger et al., 2003).

To date, mechanistic studies of the proliferative effects of PGFs have been primarily limited to the first cell cycle following PGF stimulation of quiescent cells, and it has been firmly established that the presence of an extracellular PGF is required during the entire G^sub 0^ to G^sub 1^ transition period to initiate a maximal replicative response (Zhan et al., 1993). In the present study, we questioned whether an individual PGF is able to promote a second cell cycle. We were primarily interested in studying the effects of FGF1 in the second cell cycle, since FGF1 is known to be released in response to cell stress (Friesel and Maciag, 1999) and to regulate cell proliferation in support of physiological and pathological processes including inflammation, angiogenesis, restenosis, and wound healing (Friesel and Maciag, 1999; Mandinov et al., 2003). The effects of long-term cell stimulation with FGF1 were compared with the effects of another important and well-studied mitogen, PDGF-AB. We chose the Swiss 3T3 cell culture as the experimental system, since these cells are the prototypic mammalian standard for synchronous cell cycle studies as a result of their ability to rigorously establish a monolaycr G^sub 0^ state and their failure to display apoptotic behavior under conditions of serum deprivation (Richmond et al., 1984; Shipley and Ham, 1983). Indeed, our results suggest that FGF1 and PDGF-AB exhibit a self-limiting mitogenic activity, which restricts the target cell to a single cell cycle.

Materials and methods

Cell culture and analysis of DNA replication

Swiss 3T3 cells (ATCC; Manassas, VA) were maintained in Dulbecco's modified Eagle's medium (DMEM) (Hyclone; Logan, UT) supplemented with 10% bovine calf serum (BCS) (Hyclone) and antibiotics (GIBCO; Rockville, MD). Cellular quiescence was achieved by exposing confluent populations to DMEM containing 0.25% BCS for a period of 48 hours. Cells were stimulated by the addition of either 10% BCS, 10 ng/ml recombinant human FGF1 (Zhan et al., 1993) plus 10 U/ml heparin (Pharmacia & Upjohn; Kalamazoo, MI), or 5 ng/ml PDGF- AB (R and D Systems; Minneapolis, MN) in the presence of 0.25% BCS. When a change or withdrawal of growth factors was performed, the cell culture medium was removed, monolayers washed three times with DMEM containing 10 U/ml heparin and then supplemented with new medium.

For radioautographic studies, Swiss 3T3 cells were plated on glass coverslips. DNA synthesis, as a result of the first round of DNA replication after BCS or PGF stimulation, was assessed using [^sup 14^C]deoxythymidine (dT) incorporation (0.02 Ci/ml, 57.5 Ci/ mM, NEN; Boston, MA), while the second cycle of DNA synthesis was measured by the incorporation of a high concentration of [^sup 3^H]dT (1 Ci/ml, 2 Ci/mM, NEN). Radiolabeled cells were fixed and processed for radioautography as described (Polunovsky et al., 1983). For every time point, three radioautographic preparations were analyzed microscopically, and 500 cells in each sample were examined.

Immunoblot analysis and immunoprecipitation

Monolayers of Swiss 3T3 cells were used to prepare cell lysates for immunoblot analysis. Cells were washed with cold phosphate- buffered saline (PBS), scraped into cold PBS containing 1 mM sodium vanadate and collected by centrifugation. Cell pellets were lysed in 0.5 ml of cold lysis buffer (20 mM Tris-HCl, pH 7.5, containing 300 mM sucrose, 60 mM KCl, 15 mM NaCl, 5% glycerol, 2 mM EDTA, 1% Triton X-100, 1 mM PMSF, 2 g/ml aprotinin, 2 g/ml leupeptin, 0.2% deoxycholate, and 1 mM sodium vanadate), and the lysate was clarified by centrifugation at 4C. Protein concentration was quantified using Coomassie Plus Protein Assay Reagent (Pierce; Rockford, IL) and the DU 640 spectrophotometer (Beckman; Fullerton, CA) at an excitation wavelength of 595 nm. The lysate was mixed with an equal volume of SDS-PAGE sample buffer and heated at 95C for 10 minutes. Equal amounts of the protein lysate were resolved by 10% SDS-PAGE, transferred to nitrocellulose membranes (Hybond C) and blotted with an appropriate antibody. Immunoblot analysis utilized rabbit antibodies against phosphorylated Erk1 and Erk2 (Promega; Madison, WI), rabbit antibodies against Erk1 and Erk2 (Santa Cruz; Santa Cruz, CA), mouse monoclonal antibodies against cyclin D1 (Sigma; St. Louis, MO), rabbit antibodies against cyclin E (Upstate Biotechnology; Lake Placid, NY), rabbit antibodies against cyclin A (Santa Cruz), mouse monoclonal or rabbit antibodies against p21 CIP (Pharmingen; San Diego, CA), mouse monoclonal antibodies against p27 KIP (Transduction Laboratories; San Diego, CA) and rabbit antibodies against cdk2 and cdk4 (Santa Cruz). Proteins were visualized using a horseradish peroxidase-conjugated goat antibody against either rabbit or murine IgG (BioRad; Hercules, CA) with the ECL detection system (Amersham; Piscataway, NJ).

The tyrosine phosphorylation of FGFR1 and PDGFRβ was assessed by immunoprecipitation with either an affinity-purified polyclonal rabbit antibody against FGFR1, which had been raised against a synthetic peptide located at the carboxy-terminus of FGFR1 (Friesel and Dawid, 1991) or with a rabbit antibody against PDGFRβ (Upstate Biotechnology). Cell lysates were obtained as described above, agitated at 4C for 1 hour with either the anti- FGFR1 or anti-PDGFRβ antibody followed by the addition of protein A-Sepharose (Pharmacia; Peapack, NJ) and agitation at 4C for 1 hour. The antibody complexes were washed three times with lysis buffer; the immunoprecipitated proteins were eluted in SDS-PAGE sample buffer, resolved by 7.5% SDS-PAGE and immunoblotted with \an anti-phosphotyrosine monoclonal antibody (Upstate Biotechnology) as described above.

Immunoprecipitation and in vitro kinase assay

To evaluate the association between the cdk2/cyclin E complex and either p21 or p27, cell lysates were obtained as described above and rotated at 4C for 1-3 hours with a rabbit anti-cyclin E antibody conjugated to agarose (Santa Cruz). The antibody complexes were washed three times with lysis buffer, eluted in SDS-PAGE sample buffer, resolved by 12% SDS-PAGE and immunoblotted with either a rabbit anti-p21 CIP antibody (Pharmingen), a mouse monoclonal anti- p27 KIP antibody (Transduction Laboratories), or a rabbit anti-cdk2 antibody (Santa Cruz) as described above.

To determine the kinase activity of the cyclin E/cdk2 complex, cyclin E immunoprecipitates were obtained as described above and the complexes were washed three times with lysis buffer and twice with kinase buffer (50 mM THs-HCl, pH 7.5, containing 10 mM MgCl^sub 2^ and 1 mM dithiothreitol). The kinase assays were performed by mixing the respective immunoprecipitates with 5 g of histone H1 and 10 Ci of [γ-^sup 32^P]ATP (NEN) in 30 l of kinase buffer, incubation of the reaction mixture at 30C for 30 minutes and termination of the reaction with SDS-PAGE sample buffer. The reaction mixtures were resolved by 12% SDS-PAGE, transferred to nitrocellulose membranes (Hybond C), and the phosphorylation of histone H1 was analysed by radioautography using a phosphoimager (Typhoon 8600). The kinase activity was correlated with the level of immunoprecipitated cdk2, which was determined by cdk2 immunoblot analysis followed by quantification using version 5.1 of the program, ImageQuant.

Immunodepletion experiments

To evaluate the level of the cyclin E/cdk2 complex not associated with either p21 or p27, cell extracts were prepared as described above and subjected to immunodepletion using either normal rabbit IgG, a rabbit anti-p21 antibody (Pharmingen), normal mouse IgG, or mouse monoclonal anti-p27 antibody (Transduction Laboratories) followed by the addition of protein A-Sepharose (Pharmacia). Following depletion, supernatants were subjected to cyclin E immunoblot analysis as described above to detect the presence of non- depleted cyclin E.

Immunofluorescence confocal microscopy and antisense inhibition of p21 expression

Immunofluorescence was performed using cells plated on glass cover-slips and fixed with 4% formaldehyde in PBS at different time points after PGF and serum stimulation. Cells were preincubated for 1 hour in blocking buffer (BB: PBS containing 5% BSA, 0.1% Triton X- 100, 0.1% Tween 20), incubated with the first antibody (either mouse monoclonal anti-p21 CIP or mouse monoclonal anti-p27 KIP) diluted in BB (1 g/ ml) for 1 hour, washed with PBS, incubated for 30 minutes in BB containing a second FITC-conjugated antibody (Sigma) together with the DNA-specific stain, TO-PRO3 (Molecular Probes; Eugene, OR) and embedded in 50% glycerol. Cells were examined using the confocal microscope LTCS-SP (Leica; Heidelberg, Germany) at 488 and 633 nm excitation, and the images of median optical sections of cell nuclei were recorded.

Swiss 3T3 cells were transfected with either the pCEP4-asp21 expression plasmid (B. Vogelstein, Johns Hopkins University, Baltimore, MD) or an insert-less pCEP4 plasmid (Invitrogen; Grand Island, NY) using the Fugene system (Roche; Indianapolis, IN). Stable transfectants were selected in medium containing 200 g/ml hygromycin B (Roche). The expression of p21 in the antisense and control transfectants was evaluated using p21 immunoblot analysis as described above.

Adenoviral transduction of cells with cyclin A

Quiescent Swiss 3T3 cells were transduced with adenoviral constructs containing either human cyclin A or GFP (C. Vaziri, Boston University, Boston, MA) using polylysine (Sigma) as previously described (Arcasoy et al., 1997; Fasbender et al., 1997). The transductants were stimulated with either FGF1, PDGF-AB or 10% BCS. The expression of cyclin A was confirmed by immunoblot analysis using a rabbit anti-cyclin A antibody, which recognizes human but not murine cyclin A (Santa Cruz).

Results

Cells stimulated with either FGF1 or PDGF do not enter a second round of DNA synthesis

The ability of quiescent Swiss 3T3 cells to enter a second round of DNA replication in response to an individual PGF and/or BCS was analyzed by cell radioautography. In order to distinguish between the first and second rounds of DNA synthesis, the method of double isotopic labeling (Polunovsky et al., 1983) was utilized in which the first and second rounds of DNA synthesis were independently assessed by [^sup 14^C]- and [^sup 3^H]dT incorporation, respectively (Fig. 1A). Radioautographic analysis enabled a distinction between single-labeled and double-labeled nuclei due to the discriminatory presence of extranuclear cytosolic tracks ([^sup 14^C]dT) in the first cell cycle and the high intensity of intranuclear labeling ([^sup 3^H]dT) in the second cell cycle (Fig. 1B).

Quiescent Swiss 3T3 cell monolayers were exposed to either BCS, FGF1 or PDGF-AB during the first and second cell cycles as shown in Figure 1C, and about 50% of the cells exposed to scrum in the first and second cell cycles entered the S-phase of the second cell cycle. Indeed, this is consistent with the observation that Swiss 3T3 cell mitosis stimulated from quiescence by either an individual PGF or serum occurs between 28 and 30 hours after stimulation (data not shown). However, we were surprised to observe that no more than 11% of the cells exposed to either FGF1 or PDGF-AB were able to enter a second round of DNA synthesis (Fig. 1C). In contrast, the addition of serum to cells exposed to either FGF1 or PDGF-AB in the first cell cycle enabled them to enter the S-phase of the second cell cycle with an efficiency similar to that observed with cells continuously exposed to serum (Fig. 1C). Interestingly, Swiss 3T3 cells exposed to either FGF1 or PDGF-AB in the first cell cycle failed to initiate an efficient entry into the S-phase in the second cell cycle when the FGF1-pretreated cells were exposed to PDGF-AB and vice versa (Fig. 1C). Similarly, cells exposed to serum in the first cell cycle did not exhibit an efficient re-entry into the S- phase when they were exposed to FGF1 or PDGF-AB in the second cycle. These data suggest that despite their ability to induce the transition of quiescent cells to the S-phase, FGF1 and PDGF-AB are unable to efficiently induce a second round of DNA synthesis, and cells require the presence of serum for traversing the second cycle. Indeed, unlike serum, both PDGF-AB and FGF1 failed to support the continuous growth upon stimulation of Swiss 3T3 cells (data not shown).

Fig. 1. The proliferation of Swiss 3T3 cells continuously exposed to PGFs and serum. A, B. Experimental design for the quanlitation of DNA synthesis in the first and second cycles after mitogenic stimulation. Confluent cultures of Swiss 3T3 cells were incubated for 48 hours in DMEM containing 0.25% BCS to achieve quiescence (Q) and were stimulated with either 10% BCS (S), 10 ng/ml FGF1 and 10 U/ ml heparin (F), or 5 ng/ml PDGF-AB (P). Cells were pulsed with [^sup 14^C]dT for a period from 14 to 24 hours after stimulation and chased with dT-free medium containing the appropriate mitogens for a period from 24 to 38 hours after the initial stimulation. At this time, [^sup 3^H]dT was added, cells were fixed 48 hours after the initial stimulation with the appropriate mitogens and processed for radioautography. Cells exhibiting perinuclear tracks as a result of [^sup 14^C]dT incorporation into DNA in the first cell cycle as well as cells containing a strong intranuclear [^sup 3^H]dT signal representing DNA synthesis in the second cell cycle (Polunovsky et al., 1983), and cells exhibiting both characteristics were counted. C. Entry of mitogen-stimulated Swiss 3T3 cells into the second cell cycle. The data are reported as the percentage of cells (mean S.D.) entering the second round of DNA synthesis and were obtained by determining the ratio of the number of cells containing the double label ([^sup 14^C]- and [^sup 3^H]dT) to those containing [^sup 14^C]dT. Six independent experiments yielded consistent results, and the one with the lowest serum response (S/S) is presented.

Growth factor receptor signaling pathways are functional in cells undergoing a second cell cycle

The presence of cell surface FGF receptor (R) is restored in less than one hour after ligand withdrawal (Zhan et al., 1993), and PDGF- AB fails to initiate second cycle DNA synthesis in cells previously exposed to FGF1 during the first cell cycle (Fig. 1C), suggesting that the effect with FGF1 is not a result of FGFR downregulation. Therefore, we sought to identify the mechanism responsible for the attenuation of DNA synthesis in the second cell cycle. While there are presently four known FGFR genes, Swiss 3T3 cells only express FGFR1 at detectable levels (Maher, 1996). Indeed, anti- phosphotyrosine immunoblot analysis of FGFR1 immunoprecipitates demonstrated that at 42 hours after the addition of FGF1, the phosphorylation of FGFR1α and β isoforms as well as the signature FGFR1 substrate, p90/FRS2 (Friesel et al., 1989) is not decreased (Fig. 2A) compared to 6 hours of FGF1 treatment. Similar dynamics of the tyrosine phosphorylation of PDGFRβ were demonstrated for PDGF-AB-stimulated cells (Fig. 2B).

To evaluate the functional status of the Ras-MAPK pathway downstream of these PGF receptors, we assessed the phosphorylalion of Erk 1 and Erk 2. Indeed, immunoblot analysis of the phosphorylated forms of the MAP kinases (Fig. 2C) suggests that these kinases are functional 30 and 42 hours after the administration of either BCS, FGF1 or PDGF-AB. Interestingly, throughout the entire stimulation period, FGF1 was able to induce a significantly higher level of MAPK ph\osphorylation than either serum or PDGF (Fig. 2C). In addition, Northern blot analysis demonstrated the sustained expression of the transcripts encoding Jun, Myc and ornithine decarboxylase during the time periods in the second cell cycle when FGF1 failed to initiate an increase in DNA synthesis (data not shown). These data suggest that deficiencies in receptor tyrosine kinase-mediated signaling do not account for the failure of Swiss 3T3 cells to respond to individual PGFs as mitogens in the second cell cycle.

Cells stimulated with individual PGFs express cyclins D and E and exhibit an attenuated expression of cyclin A

Since FGFR1 and PDGFR signaling was not attenuated in the second cell cycle in Swiss 3T3 cells previously exposed to FGF1 and PDGF- AB, we examined the expression of the G^sub 1^ cyclin-dependent kinases and their respective cyclins. The stimulation of cells with either FGF1, PDGF-AB or BCS induced a sustained increase of cdk2 and cdk4 expression, and the expression of cdk2 and cdk4 at 24 and 42 hours after stimulation did not differ significantly between FGF1-, PDGF-AB- and serum-stimulated cells (data not shown).

The expression of the D cyclins usually begins near the mid- G^sub 1^ period of the cell cycle and initiates the signaling cascade that ultimately leads to DNA synthesis (Sherr and Roberts, 1999; Stevaux and Dyson 2002). Protein complexes composed of the D cyclins and cdk4 induce the phosphorylation of Rb and sequester p27 KIP and p21 CIP, which are negative regulators of cell proliferation (Sherr and Roberts, 1999; Stevaux and Dyson, 2002). As shown in Figure 3A, cyclin D1 immunoblot analysis suggests that the induction and expression of cyclin D1 appears to be unaltered during the time periods in the second cell cycle when FGF1 and PDGF-AB are not functional as PGFs. Additionally, cells stimulated with FGF1 demonstrated exaggerated levels of cyclin D1 expression in comparison to cells treated with either PDGF-AB or BCS (Fig. 3A). Because entry into the S-phase is promoted by the cyclin E/cdk2 complex (Sherr and Roberts, 1999; Stevaux and Dyson, 2002), we also examined cyclin E expression in the first and second cell cycles. As shown in Figure 3A, we observed that like cyclin D1, the induction and expression of cyclin E is unaffected following the administration of either BCS, FGF1, or PDGF-AB in the second cell cycle.

Fig. 2. Cell signaling events in the first and second cell cycles. FGFR1 (A) and PDGFRβ (B) tyrosine phosphorylation in response to FGF1 and PDGF-AB. Quiescent Swiss 3T3 cells were exposed to either 10 ng/ ml FGF1 and 10 U/ml heparin (A) or 5 ng/ml PDGF-AB (B) for the period of time indicated as described in the Materials and methods section. Cell lysates were immunoprecipitated with either an anti-FGFR1 (A) or PDGFRβ (B) antibody, the precipitate was resolved by 7.5% SDS-PAGE and subjected to phosphotyrosine immunoblot analysis. FGFR1α (filled arrowhead), FGFR1β (hollow arrowhead) and p90/FRS2 (thin arrow) are marked in (A). NC represents the negative control processed with non- immune IgG instead of anti-FGFR1 or anti-PDGFRβ antibody. C. Tyrosine phosphorylation of MAP kinasts in mitogen-stimulated Swiss 3T3 cells. Cell lysates were obtained from quiescent (Q) Swiss 3T3 cells after stimulation with either FGF1, PDGF-AB, or 10% BCS at the times indicated, and similar cell lysates were obtained from cells treated with either BCS, PDGF-AB, or FGF1 in the first cell cycle followed by exposure to either BCS, PDGF-AB, or FGF1 in the second cell cycle. Lysates were resolved by 10% SDS-PAGE and blotted with antibodies against phosphorylated Erk1 and Erk2 (top) or antibodies against the Erk1 and Erk2 polypeptides (bottom).

Because the regulation of the cell cycle by cyclin D1 and cyclin E appears not to be involved in the self-limiting function of FGF1 and PDGF-AB in the second cell cycle, we examined the expression of cyclin A, which is induced during the late stage of the G^sub 1^ period with continual expression throughout the remainder of the cell cycle (Sherr and Roberts, 1999). Cyclin A is also known to form a complex with cdk2 which is critical for cells to proceed through the S-phase of the cell cycle (Coverley et al., 2002; Hengstschlager et al., 1999; Sherr and Roberts, 1999). Interestingly, immunoblot analysis of cyclin A expression revealed that while cyclin A was expressed by Swiss 3T3 cells in response to serum in the second cycle, we observed a significant reduction in the level of cyclin A expression in cells exposed to either FGF1 or PDGF-AB in the second cell cycle (Fig. 3A). In addition, a lower-molecular-weight cyclin A- immunoreactive species was observed in the FGF1-treated cells after exposure to the mitogen for 24 hours but this band was not present in the PDGF-AB-treated samples (Fig. 3A). While serum was able to revert the loss of the cyclin A expression in PDGF-AB-treated cells in the second cell cycle after 6 and 18 hours of exposure, it was only able to attenuate this response in the FGF1-treated population at 6 hours (Fig. 3A). Lastly, while the restimulation of the FGF1- pre-treated cells with serum for 18 hours prevented the attenuation of the level of cyclin A expression, it also induced the exclusive appearance of the lower-molecular-weight cyclin A-immunoreactive species (Fig. 3A). These data suggest that a major difference between the signaling events established by serum, FGF1 and PDGF-AB in the second cell cycle may involve an attenuation of cyclin A expression by FGF1 and PDGF-AB.

Fig. 3. Expression of the cyclins, p21 and p27 in mitogen- stimulated Swiss 3T3 cells. A. Cyclin expression. Cell lysates obtained from quiescent (Q) Swiss 3T3 cells after stimulation with either FGF1 (top panel), PDGF-AB (middle panel), or 10% BCS (lower panel) at the times indicated in the first and second cell cycles were resolved by 10% SDS-PAGE and subjected to either cyclin D1, cyclin E, or cyclin A immunoblot analysis. B. Expression of p21 and p27. The experimental procedure is identical to that described in (A) except that the lysates were analyzed by p21 (upper panel) and p27 (lower panel) immunoblotting.

To determine whether the restoration of cyclin A could enable either FGF1 - or PDGF-AB-stimulated cells to enter the second round of replication, we transduced Swiss 3T3 cells with an adenoviral vector encoding human cyclin A, which has been previously shown to be fully functional in murine cells (Guo et al., 2000). Immunoblot analysis of the transductants revealed a high level of human cyclin A expression but double dT-labeling experiments failed to demonstrate any significant rescue of the second cycle of DNA synthesis (data not shown).

Cells stimulated with an individual PGF display an intranuclear accumulation of the CIP/KIP inhibitors in the second cycle

Because the expression of cyclin A failed to rescue the ability of FGF1 and PDGF-AB to induce DNA synthesis in the second cell cycle, and p21 CIP and p27 KIP are well recognized as repressors of cell proliferation, which are able to prevent entry into the S- phase of the cell cycle as a result of their ability to bind and inhibit cyclin E/cdk complexes (Sherr and Roberts, 1999; Stevaux and Dyson, 2002), we examined their expression levels in the second cell cycle in response to PGFs. Immunoblot analysis revealed that FGF1 was able to induce and sustain the expression of p21 at 15 hours after stimulation, and this intensity of p21 induction was not observed with either serum- or PDGF-AB-stimulated cells (Fig. 3B). In comparison with Swiss 3T3 cells treated with either BCS or FGF1, PDGF-AB-stimulated cells demonstrated a significant increase in the expression of p27 at 24 and 42 hours after stimulation (Fig. 3B). Since nuclear localization of both p21 and p27 is requisite for their anti-proliferative effects (Blagosklonny, 2002), we utilized immunofluorescence staining to examine the intracellular locale of p21 and p27, and indeed, confocal microscopy studies demonstrate that unlike serum- and PDGF-AB-treated cells, Swiss 3T3 cells stimulated with FGF1 exhibit a dramatic increase of intranuclear p21 during the second cell cycle (Fig. 4A). In contrast, similar analysis of p27 expression revealed an increase in intranuclear p27 in the second cell cycle of Swiss 3T3 cells stimulated with PDGF-AB but not with FGF1 or BCS (Fig. 4B). Interestingly, the strong enhancement of nuclear p21 and p27 levels in the second cycle was also observed when cells were stimulated respectively with FGF1 or PDGF-AB following incubation with serum during the first cell cycle (Fig. 4).

FGF1 and PDGF-AB induce an increase in the association of the cyclin E/cdk2 complex with CIP/KIP and inhibition of its activity in the second cell cycle

In the late G^sub 1^ phase of the cell cycle, the cyclin E/cdk2 complex indirectly activates the transcription of the cyclin A gene through the specific phosphorylation of the Rb protein (Stevaux and Dyson, 2002). Because cyclin A expression appeared to be downregulated during the second cycle in response to either FGF1 or PDGF-AB, we isolated the cyclin E/ cdk2 complex and analyzed its kinase activity in Swiss 3T3 cells stimulated with either serum or an individual PGF. Interestingly, while BCS-stimulated cells maintained cyclin E/cdk2 kinase activity at high levels after 20 and 42 hours post-stimulation, its activity in FGF1- and PDGF-AB- stimulated cells significantly decreased by 42 hours to a level similar to that observed in quiescent cells (Fig. 5A). A similar decrease of cyclin E/cdk2 activity was observed when cells were incubated with FGF1 or PDGF-AB following preincubation with serum during the first cycle (Fig. 5A). In addition, immunoprecipitation with anti-cyclin E antibodies followed by immunoblot analysis using either an anti-cdk2, anti-p21 or anti-p27 antibody demonstrated that cells stimulated with PDGF-AB but not w\ith either BCS or FGF1 exhibited a significant association of p27 with the cyclin E/cdk2 complex 42 hours after stimulation (Fig. 5B). While some increase can be observed in cells stimulated with BCS for 42 hours, this increase may have been the result of contact inhibition. Similarly, cells stimulated with FGF1, but not with PDGF-AB or serum, demonstrated a dramatic increase of p21 association with cyclin E/ cdk2 42 hours after stimulation (Fig. 5B).

We also evaluated the fraction of cyclin E associated with p21 using an immunodepletion strategy. Extracts of Swiss 3T3 cells treated with either FGF1, PDGF-AB, or BCS were depleted by immunoprecipitation with either an anti-p21 antibody or normal rabbit IgG, and the level of cyclin E remaining in the extracts was evaluated by cyclin E immunoblot analysis. We observed that the depiction of p21 reduced the level of cyclin E in extracts from Swiss 3T3 cells stimulated with FGF1 (Fig. 5C). Similar results were also obtained using a p27 depletion strategy in cells treated with PDGF-AB (Fig. 5C). These data suggest that the content of cyclin E not associated with either p21 or p27 may be at a level insufficient to enable the transition from the G^sub 1^- to S-phase of the second cycle.

The ability of FGF1 to limit entry into a second cell cycle may involve the function of p21

Since these studies implicated the function of p21 as a repressor of DNA synthesis in FGF1-treated Swiss 3T3 cells in the second cell cycle, cells expressing an antisense p21 construct were evaluated. Immunoblot analysis of p21 levels in the p21 antisense Swiss 3T3 cell transfectants stimulated with FGF1 for 42 hours demonstrated a 60% inhibition of p21 protein expression (data not shown). In addition, double dT labeling radioautography demonstrated that the rate of entry into the S-phase of the second cell cycle in response to FGF1 was 20.2 2.3% in the p21 antisense transfectants and 13.0 1.6% in insert-less vector transfectants. The observation that the expression of the p21 antisense construct did not result in a more pronounced stimulation of DNA synthesis in the second cycle may be the result of residual levels of the p21 polypeptide in the FGF1- stimulated antisense transfectants. Indeed, Swiss 3T3 cells pretreated with FGF1 for 48 hours required nearly 14 hours of exposure to serum to induce DNA synthesis (data not shown), which is quite similar to the period of time exhibited by quiescent cells to enter the S-phase in the first cell cycle. Therefore, we suggest that this may be the result of the time required for p21 degradation upon stimulation with serum (Fig. 3A). However, in antisense experiments, the increase of the rate of replication after entering the second cell cycle was significant suggesting that p21 may be involved in the failure of FGF1-treated cells to traverse the second cell cycle.

Discussion

Our data suggest that FGF1 and PDGF-AB are able to limit their mitogenic function to a single cell cycle as a result of the p21- or p27-mediated inhibition of the cyclin E/cdk2 complex activity in the second cell cycle. It is significant that the increase in the nuclear levels of p21 in FGF1-stimulated cells and of p27 in PDGF- AB-stimulated cells occurs in the second cycle, since the nuclear localization of these proteins is crucial for their inhibitory activities (Blagosklonny, 2002).

While serum is able to efficiently promote the initiation of second cycle DNA synthesis in cells exposed to either FGF1 or PDGF- AB in the first cell cycle, the functional component(s) present in serum remains unknown. Because PDGF-AB does not initiate the induction of DNA synthesis in the second cell cycle in Swiss 3T3 cells exposed to FGF1 in the first cell cycle, it is unlikely that serum-derived PDGF is the functional serum constituent. Likewise, co- stimulation of Swiss 3T3 cells with FGF1 and insulin (10 g/ml) was also unable to induce an efficient entry into the S-phase of the second cell cycle (data not shown), which suggests that it is also unlikely that serum-derived insulin is the functional mitogenic component present in serum. However, the observation that FGF1 and PDGF are able to limit their mitogenic activity to a single cell cycle and that serum rescues this response explains the requirement of serum for the serial propagation of a wide variety of diverse human cell types in vitro (Maciag et al., 1981a, b). Since serum is a non-physiologic fluid, the identity of the factors present in serum that are responsible for the rescue of cells in the second cell cycle may be noteworthy.

Fig. 4. Intracellular localization of p21 (A) and p27 (B) in mitogen-stimulated cells. A. Swiss 3T3 cells after stimulation with either BCS, PDGF-AB or FGF1 were fixed with 4% formaldehyde, incubated with a mouse anti-p21 antibody, washed with PBS and co- stained with a FITC-conjugated anti-mouse IgG antibody and the DNA- specific stain, TO-PRO3 (Molecular Probes). Photomicrographs were taken using a confocal microscope. a: Quiescence, b: 10% BCS, 18 hours; c: 10% BCS, 42 hours; d: FGF1, 18 hours, e: FGF1, 42 hours, f: PDGF-AB, 18 hours, g: PDGF-AB, 42 hours, h: 10% BCS, 24 hours followed by FGF1, 18 hours. Green: p21. Red: DNA; Yellow: overlay. B. Mitogen stimulation, fixation, staining, confocal microscopy and legend are similar to that described in (A) except that a mouse monoclonal anti-p27 antibody was used to stain the cells and "h" is 10% BCS, 24 hours followed by PDGF-AB, 18 hours. Green: p27; Red: DNA; Yellow: overlay.

An important biochemical characteristic of the cells continuously treated with either FGF1 or PDGF-AB is the attenuation of cyclin A expression and, at least in the case of FGF1 exposure, this attenuation is preceded by the appearance of a lower molecular weight immunoreactive form of cyclin A. Since we observed that cells treated with FGF1 in the first cell cycle followed by serum stimulation in the second cell cycle exclusively contain the lower molecular weight form of cyclin A yet remain competent to induce DNA synthesis, it is possible that the smaller molecular weight immunoreactive form of cyclin A may be functional.

Fig. 5. Analysis of the cyclin E/cdk2 complex. A. In vitro kinase activity of the cyclin E/cdk2 complex in mitogen-stimulated cells. Using an anti-cyclin E antibody, cyclin E/cdk2 complexes were immunoprecipitated from Swiss 3T3 cell lysates after stimulation with either FGF1, 10% BCS or PDGF-AB for 18 and 42 hours. Quiescent (Q) cells were used as a control. In vitro kinase activity was assessed by the incorporation of [γ-^sup 32^P]ATP into histone 1, the reaction product was resolved by 10% SDS-PAGE and analyzed using a Typhoon phosphor-imager. To examine the level of cdk2, the cyclin E immunoprecipitates were resolved by 10% SDS-PAGE and immunoblotted with an anti-cdk2 antibody. The intensities of phosphorylated histone H1 (kinase activity) and cdk2 bands were quantitated by densitometry, and the ratios of cdk2 kinase activity per cdk2 protein levels (mean S.D.) are presented. B. Co- precipitation of p21 and p27 with the cyclin E/cdk2 complex in mitogen-stimulated Swiss 3T3 cells. Swiss 3T3 cells were exposed to either FGF1, PDGF-AB or 10% BCS for 18 and 42 hours and the cyclin E/ cdk2 complexes were immunoprecipitated from cell lysates using an anti-cyclin E antibody. The immunoprecipitates were resolved by 10% SDS-PAGE and subjected to either p21, p27 or cdk2 immunoblot analysis. NC represents the negative control processed with non- immune rabbit IgG instead of the anti-cyclin E antibody and Q represents the signal obtained from quiescent cells. C. Immunodepletion (ID) of cyclin E in the lysates of mitogen- stimulated cells using antibodies to either p21 or p27. Swiss 3T3 cells lysates at 18 and 42 hours after cell stimulation with either FGF1, PDGF-AB or 10% BCS were immunodepleted using either a rabbit polyclonal anti-p21 antibody, a mouse monoclonal anti-p27 antibody, normal rabbit IgG, or normal mouse IgG, and the supernatants were resolved by 12% SDS-PAGE and subjected to cyclin E immunoblot analysis.

The attenuation of cyclin A expression in the second cell cycle appears to be the indirect Rb-mediated result of the inhibition of cyclin E/cdk2 activity by p21 or p27 (Sherr and Roberts, 1999; Stevaux and Dyson, 2002). However, although cells stimulated with FGF1 in the second cycle following their serum stimulation in the first cycle fail to efficiently re-enter the S-phase, the cyclin A level is reduced in these cells not as dramatically as in cells continuously stimulated with FGF1 (Fig. 3A). Moreover, cyclin A expression is not sufficient to rescue DNA synthesis in the cells continuously stimulated with an individual PGF. This observation may be explained by reports that the failure of cyclin E/cdk2 to phosphorylate Rb results in an inhibition of the expression of genes encoding components of the replication machinery (Stevaux and Dyson, 2002).

An additional explanation for the self-limiting nature of FGFl and PDGF-AB may be the inability of cells to assemble functional replicative complexes on replication origins following the first mitosis (Coverley et al., 2002; Hengstschlager et al., 1999; Newlon, 1997), particularly the inability to load the hexameric minichromosome maintenance complex, which upon further activation acquires helicase activity (Lei and Tye, 2001). Indeed, the recent report that the kinase activity of cyclin E/cdk2 may be necessary for the assembly and function of the replication complex (Coverley et al., 2002) is consistent with this premise. Additionally, p21 may also interfere with DNA replication independently of cyclin E/cdk2 inhibition since p21 is known to associate with the proliferative cell nuclear antigen (PCNA) (Lu et al., 2002). In this situation, p21 would inhibit the ability of PCNA to bind and activate DNA polymerase δ. Interestingly, it has been shown that the knockout ofthe cyclin E gene does not interfere with continuous cell proliferation but prevents the exit of cells from the state of quiescence (Geng et al., 2003). In our system, the inactivation of the cyclin E/cdk2 complex in the second cell cycle is achieved not by cyclin E knockout but by the binding of p21 or p27 to cyclin E/ cdk2. Indeed, cells rescued with serum from the block in the second cell cycle achieved as a result of their continuous treatment with FGF1 exit from this state using kinetics similar to that observed for serum-stimulated quiescent cells. Thus the protracted transition of the cells to the S-phase could be required for loading of MCM proteins onto DNA replication origins (Geng et al., 2003).

It is interesting to note that the cells, which are proliferatively non-responsive to an individual PGF in the second cell cycle also exhibit biochemical characteristics strikingly similar to cells that have achieved a non-proliferative state as a result of serial propagation in vitro. Indeed, both non- proliferative states involve an unusual form of quiescence manifested by responsiveness to an exogenous PGF signal including the activation of receptor tyrosine kinases, the activation of Erk 1 and Erk 2, and the induction of immediate-early and mid-to-late G^sub 1^ genes (Garfinkel et al., 1996). Also, senescent human fibroblasts (Afshari et al., 1993; Lucibello et al., 1993) and endothelial cells (Andreeva, Prudovsky and Maciag, 2003, unpublished results) are characterized by the normal expression of cyclins D and E and the lack of cyclin A expression in vitro. Similar data have been reported by Murai et al. (2001), who showed that the inhibition of fibroblast proliferation by exogenous interleukin 1β involves the induction of a state which exhibits high levels of cyclins E and D and a downregulation of both cyclin A expression and cyclin E/cdk2 kinase activity as well as a dramatic upregulation of p21 expression. It is also interesting to note that these data correlate well with the IL1-dependent induction of human endothelial cell (Maier et al., 1990) and human smooth muscle cell (Hsu et al., 1999) senescence in vitro. However, unlike the non-proliferative character of the in vitro senescent phenotype, the state of quiescence achieved by an individual PGF in the second cell cycle is not terminal since it can be readily rescued by serum. While we do not know whether the quiescent phenotypes exhibited by these seemingly disparate cell culture systems are related, their common biochemical characteristics indicate this possibility.

Acknowledgement. The authors thank Sergei Akimov for expert technical assistance during the early stage of this study, Cyrus Vaziri, Boston Univ. School of Medicine, Boston, MA for the adenoviral construct containing cyclin A, Bert Vogelstein, Johns Hopkins Medical School, Baltimore, MD for the p21 antisense expression construct and Norma Albrecht for expert administrative assistance. This work was supported by NIH grants HL 35627, HL 54710 to I. Prudovsky, and RR 15555 to Robert Friesel and was submitted by V. Andreeva in partial fulfillment of the requirements for a doctorate of philosophy from the Engelhardt Institute of Molecular Biology, Moscow, Russia.

References

Afshari, C. A., Vojta, P. J., Annab, L. A., Futreal, P. A., Willard, T. B., Barrett, J. C., 1993. Investigation of the role of G1/S cell cycle mediators in cellular senescence. Exp. Cell Res. 209, 231-237.

Arcasoy, S. M., Latoche, J. D., Condor, M., Pitt, B. R., Pilewski, J. M., 1997. Polycations increase the efficiency of adenovirus-mediated gene transfer to epithelial and endothelial cells in vitro. Gene Ther. 4, 32-38.

Blagosklonny, M. V., 2002. Are p27 and p21 cytoplasmic oncoproteins? Cell Cycle 1, 391-393.

Coverley, D., Laman, H., Laskey, R. A., 2002. Distinct roles for cyclins E and A during DNA replication complex assembly and activation. Nat. Cell Biol. 4, 523-528.

Fasbender, A., Zabner, J., Chillon, M., Moninger, T. O., Puga, A. P., Davidson, B. L., Welsh, M. J., 1997. Complexes of adenovirus with polycationic polymers and cationic lipids increase the efficiency of gene transfer in vitro and in vivo. J. Biol. Chem. 272, 6479-6489.

Freshney, R. I., 2000. Culture of Animal Cells. Wiley-Liss, New York, NY, pp 105-106.

Friesel, R., Burgess, W. H., Maciag, T., 1989. Heparin-binding growth factor 1 stimulates tyrosine phosphorylation in NIH 3T3 cells. Mol. Cell. Biol. 9, 1857-1865.

Friesel, R., Dawid, I. B., 1991. cDNA cloning and developmental expression of fibroblast growth factor receptors from Xenopus laevis. Mol. Cell. Biol. 11, 2481-2488.

Friesel, R., Maciag, T., 1999. Fibroblast growth factor prototype release and fibroblast growth factor receptor signaling. Thromb. Haemost. 82, 748-754.

Garfinkel, S., Hu, X., Prudovsky, I. A., McMahon, G. A., Kapnik, E. M., McDowell, S. D., Maciag, T., 1996. FGF-1-dependent proliferative and migratory responses are impaired in senescent human umbilical vein endothelial cells and correlate with the inability to signal tyrosine phosphorylation of fibroblast growth factor receptor-1 substrates. J. Cell Biol. 134, 783-791.

Geng, Y., Yu, Q., Sicinska, E., Das, M., Schneider, J. E., Bhattacharya, S., Rideout, W. M., Bronson, R. T., Gardner, H., Sicinski, P., 2003. Cyclin E ablation in the mouse. Cell 114, 431- 443.

Guo, N., Faller, D. V., Vaziri, C., 2000. A novel DNA damage checkpoint involving post-transcriptional regulation of cyclin A expression. J. Biol. Chem. 275, 1715-1722.

Hengstschlager, M., Braun, K., Soucek, T., Miloloza, A., Hengstschlagcr-Ottnad, E., 1999. Cyclin-dependent kinases at the G1- S transition of the mammalian cell cycle. Mutat. Res. 436, 1-9.

Hsu, J. Y., HSU, M. Y., Sorger, T., Herlyn, M., Levine, E. M., 1999. Heparin/endothelial cell growth supplement regulates matrix gene expression and prolongs life span of vascular smooth muscle cells through modulation of interleukin-1. In Vitro Cell Dev. Biol. Anim. 35, 647-654.

Lei, M., Tye, B. K., 2001. Initiating DNA synthesis: from recruiting to activating the MCM complex. J. Cell Sci. 114, 1447- 1454.

Lu, X., Tan, C. K., Zhou, J. Q., You, M., Carastro, L. M., Downey, K. M., Antero G., 2002. Direct interaction of proliferating cell nuclear antigen with the small subunit of DNA polymerase delta. J. Biol. Chem. 277, 24340-24345.

Lucibello, F. C., Sewing, A., Brusselbach, S., Burger, C., Muller, R., 1993. Deregulation of cyclins D1 and E and suppression of cdk2 and cdk4 in senescent human fibroblasts. J. Cell Sci. 105, 123-133.

Maciag, T., Hoover, G. A., Stemerman, M. B., Weinstein, R., 1981a. Serial propagation of human endothelial cells in vitro. J. Cell Biol. 91, 420-426.

Maciag, T., Nemore, R. E., Weinstein, R., Gilchrest, B. A., 1981b. An endocrine approach to the control of epidermal growth: serum-free cultivation of human keratinocytes. Science 211, 1452- 1454.

Maher, P. A., 1996. Nuclear translocation of fibroblast growth factor (FGF) receptors in response to FGF-2. J. Cell Biol. 134, 529- 536.

Maier, J. A., Voulalas, P., Roeder, D., Maciag, T., 1990. Extension of the life-span of human endothelial cells by an interleukin-1 alpha antisense oligomer. Science 249, 1570-1574.

Mandinov, L., Mandinova, A., Kyurkchiev, S., Kyurkchiev, D., Kehayov, I., Kolev, V., Soldi, R., Bagala, C., deMuinck, E. D., Lindner, V., Post, M., Simons, M., Bellum, S., Prudovsky, I., Maciag, T., 2003. Copper chelation represses the vascular response to injury. Proc. Natl. Acad. Sci. USA 100, 6700-6705.

Murai, T., Nakagawa, Y., Maeda, H., Terada, K., 2001. Altered regulation of cell cycle machinery involved in interleukin-1- induced G (1) and G (2) phase growth arrest of A375S2 human melanoma cells. J. Biol. Chem. 276, 6797-6806.

Newlon, C. S., 1997. Putting it all together: building a prereplicative complex. Cell 91, 717-720.

Polunovsky, V. A., Setkov, N. A., Epifanova, O. I., 1983. Onset of DNA replication in nuclei of proliferating and resting NIH 3T3 fibroblasts following fusion. Exp. Cell Res. 146, 377-383.

Richmond, K. M., Riddle, P. N., Brooks, R. F., 1984. Apparent desensitization of Swiss 3T3 cells to the mitogens FGF and vasopressin. J. Cell. Physiol. 121, 547-557.

Schratzberger, P., Kirchmair, R., Vale, P. R., Losordo, D. W., 2003. Therapeutic angiogenesis by gene transfer in critical limb and myocardial ischemia. Curr. Pharm. Des. 9, 1041-1047.

Sherr, C. J., Roberts, J. M., 1999. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 13, 1501- 1512.

Shaoul, E., Reich-Slotkey, R., Berman, B., Ron, D., 1995. Fibroblast growth factor receptors display both common and distinct signaling pathways. Oncogene 10, 1553-1561.

Shipley, G. D., Ham, R. G., 1983. Multiplication of Swiss 3T3 cells in a serum-free medium. Exp. Cell Res. 146, 249-260.

Stevaux, O., Dyson, N. J., 2002. A revised picture of the E2F transcriptional network and RB function. Curr. Opin. Cell Biol. 14, 684-691.

Zhan, X., Hu, X., Friesel, R., Maciag, T., 1993. Long term growth factor exposure and differential tyrosine phosphorylation are required for DNA synthesis in BALB/c 3T3 cells. J. Biol. Chem. 268, 9611-9620.

Viktoria Andreeva, Igor Prudovsky1), Thomas Maciag2)

Center for Molecular Medicine, Maine Medical Center Research Institute, Scarborough, Maine, USA

Received December 24, 2003

Received in revised version April 30, 2004

Accepted May 14, 2004

1) Corresponding author: Dr. Igor Prudovsky, Center for Molecular Medicine, Maine Medical Center Research Institute, 81 Research Drive, Scarborough, Maine 04074, USA, e-mail: prudoi@mmc.org, Fax: ++207-885-8179.

2) The article is dedicated to the memory of Tom Maciag, scientist, friend and mentor.

Copyright Urban & Fischer Verlag Aug 2004

Source: European Journal of Cell Biology

More News in this Category