INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

The protein encoded by the RB gene, pRb, is a nuclear phosphoprotein that is involved in regulating progression through G₁ into S phase (reviewed in Refs. 1 and 2). It is hypophosphorylated in G₀ and is phosphorylated progressively as quiescent cells progress through G₁, reaching a hyperphosphorylated form at the G₁-S border. Hyperphosphorylation is thought to inactivate pRb, allowing cells to enter S phase. Microinjection of an excess of purified, hypophosphorylated pRb up to mid G₁ results in reversible cell cycle arrest (1). Thus, pRb may only be required up to a point some time in mid- or late G₁, and there is evidence that pRb forms part of the mechanism controlling the G₁ restriction point (reviewed in Ref. 2).

pRb is thought to be phosphorylated by a succession of cyclin-dependent kinases (cdks),¹ which are activated at different times during G₁: for example, cdk6 and cdk4 are activated in early or mid-G₁, followed by cdk2 in late G₁ (reviewed in Ref. 3). pRb then remains hyperphosphorylated throughout the rest of the cell cycle and is dephosphorylated to a hypophosphorylated state during mitosis, possibly by a type 1 serine-threonine phosphatase (4). During continuous proliferation, cells in early G₁ (G_1A) contain partially phosphorylated pRb (5), which becomes hyperphosphorylated during late G₁ (G_1B) (5, 6). The hypophosphorylated form is present in cells that are quiescent or arrested in early G₁ by the action of negative growth factors, such as -interferon (-IFN) (7, 8) or transforming growth factor- (9-11) and in terminally differentiated cells (reviewed in Ref. 12).

The pRb protein shares regions of homology with two other proteins, p107 and p130,² and there is evidence that all three can regulate cell proliferation (15-17). Like pRb, they are phosphoproteins that have consensus sites for phosphorylation by cdks. In T98G cells that are entering the cell cycle for the first time from G₀, p130 is phosphorylated to Form 3 during mid G₁, which coincides with the phosphorylation of pRb and the induction of cyclin D1 (13). Exit from the cell cycle caused by transforming growth factor- results in the accumulation of p130 Form 1, and exit caused by serum starvation results in the accumulation of Form 2 (14). p107 is not phosphorylated as cells enter G₁ from G₀ but is phosphorylated from S phase onward (18). Cyclin A-cdk2 and cyclin E-cdk2 associate with both p107 and p130 in vitro (13, 16, 19-21), but it is not clear whether cdk2 phosphorylates either in vivo. In contrast, there is evidence that p107 can be phosphorylated in cells in culture by cyclin D1-cdk4 (18).

The hypophosphorylated form of pRb will bind transcription factors, such as E2F, which regulate genes encoding proteins that are required for cell cycle progression (reviewed in Ref. 22). It is thought that this is also how p130 and p107 function (reviewed in Ref. 23). The E2F transcription factor is a heterodimer of an E2F protein together with a member of the DP family (22). The major E2F complex in human primary, CD34+ hemopoietic progenitor cells (24) and in quiescent primary T-cells (17), B-cells, and monocytes is composed of E2F-4, together with DP-1 bound to p130 (25), but in cycling cells, E2F-4 binds both pRb and p107 (26) or p107 alone (24). Indeed, it has been suggested that the presence of the p130-E2F-4-DP-1 complex defines G₀ (27), because this complex is not present in proliferating cells. Different members of the pRb family have some specificity for different E2Fs: E2F-1, -2, and -3 bind pRb rather than p107 (28, 29), whereas E2F-4 and -5 bind p130 and p107 in preference to pRb (17, 30-32). The functional consequences are that artificially elevating the level of E2F-1 preferentially overcomes cell cycle arrest by pRb, and E2F-4/DP-1 overcomes arrest caused by p130 (30, 33, 34). Furthermore, expression of E2F-1 (35, 36), E2F-2, or E2F-3 (37) alone or E2F-4 in combination with DP-1 (30, 34) causes quiescent cell lines to progress through G₁ into S phase.

Different genes are induced preferentially by different E2Fs (37); for example, the genes for dihydrofolate reductase and thymidine kinase by E2F-2, for cyclin A and cdc2 by both E2F-1 and -2, and for cdk2 by E2F-3. Thus individual genes with specific E2F binding sites in their promoters are regulated by specific E2Fs, and so regulation of individual E2F-DP combinations by different pRb family members provides a mechanism for turning the transcription of specific genes on or off at specific times during entry into the cell cycle. Indeed, pRb, p107, and p130 have been shown to bind individual E2Fs at different stages during entry into the first cell cycle (Ref. 26 and references therein), and so transcription controlled by specific E2Fs could be repressed at different times. Such coordinated activation of transcription is required for cell proliferation because transfection of either pRb, p107, or p130 will prevent progression into the cell cycle, which can be overcome by increased expression of individual E2Fs (17, 38-40). However, suppression of cell proliferation by p107 and pRb are not the same. There are two domains in p107, either of which will prevent proliferation (41): one domain binds E2F, and in this respect the mechanism of growth inhibition may be similar to that of pRb; the second growth-inhibitory domain of p107 binds cyclin A or E-cdk2, and this domain does not have a counterpart in pRb. pRb and p107 may also differ in that the c-Myc protein, which is important in driving proliferation, is inhibited by binding to p107 (42). In addition to regulating E2F activity, pRb also regulates the activity of all three classes of RNA polymerases (reviewed in Ref. 43).

The balance between particular E2Fs and pRb or p130 is clearly important in determining whether a cell proliferates, remains quiescent, or is driven to apoptosis. Cells driven by E2F-1 enter the cell cycle and then die later by apoptosis (44-46). E2F-2 and E2F-3 also cause cells to enter S phase, but neither causes apoptosis (37). Thus, E2F-driven entry into S phase and the induction of apoptosis are not linked; rather, induction of apoptosis is unique to E2F-1. This has lead to the suggestion that E2F-1 but not E2F-2 or E2F-3 specifically induces genes that trigger apoptosis and that a proliferating cell is therefore primed for apoptosis unless rescued by the presence of anti-apoptotic proteins, such as pRb (48) or p53 (37).

From what has been described above, it is clear that regulation of cell cycle entry is complex and involves a coordinated series of phosphorylations of key cell cycle proteins and the activation of a number of genes. Cell cycle exit is also complex, and this is illustrated by the way -IFN causes cell cycle arrest. -IFN, which is naturally secreted as a component of the host defense mechanisms of the body, has marked antiviral properties, but -IFN also has antiproliferative effects on certain cell types (49). -IFN was thought to exert its antiproliferative effect in part via the pRb pathway (5, 7, 50, 51). However, the cytostatic effect of -IFN may be mediated by a more complicated series of mechanisms that lead to orderly cell cycle arrest rather than triggering apoptosis. We know about some of the processes involved: cdk activity is decreased (51-55), which may be due in part to a decrease in cdc25A and cyclins D3, E, and H and an increase in p18 in Daudi B-cells (54, 56) and induction of the cdk inhibitors p15, p16, p21^Cip1, and p27^Kip1 in other cell types (57-60). The initial phosphorylation of pRb during G_1A is prevented (5), and free E2F binding to DNA is reduced (8, 62). The latter may be due to the induction by -IFN of p202 (63), a protein that sequesters E2F in a form that does not bind DNA. Other transcription factors that are involved in controlling cell proliferation are also down-regulated. For example, the activity and abundance of the Oct-1 and Oct-2 transcription factors is decreased (64), and -IFN also causes the depletion of the c-MYC protein (51, 62, 65), which is mediated in part by the double-stranded RNA-activated kinase, PKR (66).

Regulation of the expression and phosphorylation of pRb, p107, and p130 during the transition from G₀ through G₁ into S phase has been studied by a number of groups. However, the processes regulating entry into the cell cycle (G₀G₁S) may be different from those regulating cell cycle progression through G₁ in actively cycling cells (MG₁S) (67) and are poorly understood. This is particularly important because negative growth factors, such as -IFN, exert their effects on actively cycling cells. We showed previously that -IFN causes inhibition of pRb phosphorylation and that this only occurs in G₁ (5, 7). In the study presented here, we investigated whether p130 and p107 are regulated by -IFN and whether their regulation is also cell cycle-dependent. Our studies show that regulation of the phosphorylation of pRb and p107 is different from that of p130 during the cell cycle of actively proliferating cells and that each is regulated in a different manner by -IFN as cells exit the cell cycle.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Reagents-- Chemicals were obtained from Sigma unless stated otherwise.

Cell Culture-- Cell culture and T-cell isolation methods were described previously (5, 7, 52, 68), with additional details as described below. Primary T-cells were isolated from the peripheral blood of normal volunteers and were at least 80% pure, as determined by flow cytometry with CD3, CD4, CD8, CD14, CD19, and CD45. The purified T-cells were cultured at 1 × 10⁶ cells/ml in RPMI medium/10% FCS (Life Technologies, Inc.) and stimulated with 1 µg/ml phytohaemagglutinin (PHA) (Glaxo-Wellcome) for the times shown. Cells were maintained in PHA for 60 h, interleukin-2 (IL-2) (Murex) was added to a final concentration of 20 ng/ml, and culture was continued for 5-7 days. IL-2 was withdrawn by washing the cells three times in 50 ml of prewarmed RPMI medium/10% FCS per 2 × 10⁷ cells, and the cells were re-seeded in the same medium at 1 × 10⁶ cells/ml.

Flow Cytometry-- Two color flow cytometry of DNA (stained with propidium iodide (PI)) and total cell protein (stained with fluorescein isothiocyanate (FITC)) (5, 69) was used routinely for cell cycle analysis. Analysis and electronic sorting was carried out using an Epics-Elite flow cytometer (Coulter Electronics). Doublets of cells in G₁ can contaminate the G₂-M fraction, and so they were excluded as far as possible by only including cells that were within linear gates for forward scatter versus forward scatter peak, for FITC staining versus FITC peak, and also for PI staining versus PI peak. Populations of sorted cells were centrifuged at 400 × g for 7 min and resuspended at 1 × 10⁶ cells/ml in fresh PI/FITC stain. At least 5 × 10⁵ cells were collected in each sorted fraction, and the yield was determined by manual counting using an improved Neubauer chamber. Approximately 5 × 10⁴ cells were re-analyzed to determine their purity, and this is indicated in the figure legends for Figs. 1, 3, and 4. In each case, cell cycle profiles were supplied to the reviewers (data not shown). The remainder were centrifuged at 400 × g for 7 min in microcentrifuge tubes, resuspended at 1 × 10⁷ cells/ml in SDS-PAGE sample buffer, and then boiled immediately for 10 min. As controls, fixed and PI/FITC-stained but unsorted cells were pelletted and boiled in SDS sample buffer. The SDS sample buffer contained protease and phosphatase inhibitors (NaF, -glycerophosphate, Na₃VO₄·14H₂O, EGTA, phenylmethylsulfonyl fluoride, Na₂P₂O₇·10H₂O, aprotinin, pepstatin A, leupeptin; see Ref. 52). Either 1 or 2 × 10⁵ cells were loaded in each lane of Western blots.

Immunoprecipitation, DNA Binding and Western Blotting-- Immunoprecipitations were carried out with 1 × 10⁷ Daudi cells lysed in 800 µl of modified immunoprecipitation buffer as described previously (52), except that the buffer contained 0.5% Nonidet P-40 rather than 1% Triton X-100. Solutions were kept on ice, and all manipulations were carried out in a 5 °C cold-room. The lysate was centrifuged for 2 min at 10,000 × g, the supernatant was transferred to a fresh tube, and the nuclear pellet was then extracted for 20 min with 80 µl of the same buffer containing 450 mM NaCl. After centrifugation for 2 min at 10,000 × g, the supernatants were pooled and preabsorbed for 1-2 h with 5 µg of rabbit IgG (DAKO Ltd., High Wycombe, UK) and 20 µl of protein A-agarose (Repligen, Cambridge, MA). Immunoprecipitation was carried out with 5 µg of anti-E2F-4 (C20) antibody and then with protein A-agarose or with 40 µg of anti-E2F-1 antibody (C20) coupled to Sepharose (Santa Cruz Biotechnology Inc.). One-third of each immunoprecipitate was electrophoresed per gel. These procedures were sufficient to extract all of the p130, pRb, and p107 from Daudi cells, and none was detected when SDS lysates of the residual nuclear pellets were subjected to Western blotting (data not shown).

Whole Cell Extracts and Phosphatase Treatment-- Whole cell extracts were prepared by resuspending 2 × 10⁷ cells in 100 µl of lysis buffer containing 20 mM HEPES, pH 7.8, 450 mM NaCl, 25% (v/v) glycerol, 0.2 mM EDTA, 0.5 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 µg/ml leupeptin, 0.5 µg/ml Sigma protease inhibitor, 1.0 µg/ml trypsin inhibitor, 0.5 µg/ml aprotinin, 40 µg/ml bestatin, and 2 mM diisofluorophosphate. This was subjected to three freeze (dry ice/ethanol)-thaw (37 °C) cycles. The samples were then centrifuged at 10,000 × g, 4 °C for 10 min, and the supernatant was removed, aliquoted, and stored at 70 °C. Protein concentration in the extracts was measured using the Bio-Rad protein assay. Where indicated, 4 µg of protein from whole cell extracts were incubated either with 100 or 400 units of  protein phosphatase (New England Biolabs) in 25 µl of the appropriate buffer with or without 50 mM NaF and 1 mM Na₃VO₄·14H₂O at 30 °C for the times shown. SDS sample buffer was then added to the samples, which were then boiled for 5 min, separated by SDS-PAGE, and Western blotted.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

p130 and p107 in Cells Arrested by -IFN-- The pRb protein is dephosphorylated when Daudi cells are arrested by -IFN (7), and this causes pRb to migrate faster on SDS-gels. The form of p130 that migrates slowest (Form 3) is also lost in cells cultured with -IFN and is replaced by Forms 1 and 2 (see, for example Fig. 1A and B, lanes 1 and 2). In order to determine whether the change in migration of p130 caused by -IFN can be accounted for by phosphorylation changes, we incubated whole cell extracts of proliferating Daudi cells in vitro with or without the dual-specificity  phosphatase. Phosphatase treatment (Fig. 1A) caused a shift from p130 Form 3 to a species that comigrated with Forms 1 and 2 (lanes 4-9), as well as a faster migrating form (Form 0) when a higher concentration of -phosphatase was used (lane 11). The shift was inhibited by the addition of phosphoatase inhibitors (lanes 3 and 12). Thus, changes in the migration of p130 in response to -IFN are consistent with its dephosphorylation.

View larger version (48K): [in this window] [in a new window]   Fig. 1.   The phosphorylation state of p130 in continuously cycling cells. A, changes in p130 phosphorylation alter its migration in SDS-PAGE. Daudi whole cell extract from 1 × 106 cells was incubated with 100 units of -phosphatase without (0-60 min, lanes 4-9) or with (60 min, lane 3) sodium fluoride and sodium orthovanadate. The figure shows a Western blot for p130 and the positions of p130 Forms 0, 1, 2, and 3 are shown. An extract was also incubated for 60 min with a higher concentration of -phosphatase (400 units) without (lane 11) or with (lane 12) sodium fluoride and sodium orthovanadate. Lanes 13 and 10 are lysates incubated with or without sodium fluoride and sodium orthovanadate in the absence of -phosphatase. Proliferating and cells cultured for 24 h with -IFN and lysed directly in SDS buffer (lanes 1 and 2, respectively) are shown for comparison. Lanes 1-9 and 10-13 are from separate gels. B, the phosphorylation state of p130 in proliferating cells. Proliferating Daudi cells were fixed and stained for DNA and total cell protein, and cells in G1A and G1B, S, and G2-M, or in G1 and G2-M were isolated by flow cytometry as described under "Materials and Methods." The purity of the sorted fractions was 99.6, 98.8, 89, 77, 99.5, and 95%, respectively. p130 in G1A, G1B, S, and G2-M was visualized by Western blotting (lanes 5-8) For comparison, total cell lysates of quiescent primary T-cells and Daudi cells 24 h ±-IFN run on the same gel are in lanes 11 and lanes 9 and 10 respectively. Lane 12 is proliferating Daudi in G2-M (95% G2-M, 5% G1), and 5% loading of cells in G1 (99.5% pure) is shown in lane 13. Daudi cells cultured for 39 h ±-IFN and lysed directly (lanes 1 and 2) or fixed in 70% EtOH and stained with PI and FITC prior to boiling in SDS sample buffer (lanes 3 and 4) are shown.

Phosphorylation of p130 and pRb during Cell Cycle Entry from G₀-- p130 is present in Forms 1 and 2 when Daudi cells cultured with 0.5% FCS exit the cell cycle (see Fig. 5B, lane 3). The two-color cell cycle profile of these cells coincides with proliferating cells in G₁, and so they cannot be used to determine when p130 is phosphorylated to Form 3 as cells enter the cell cycle from G₀. Human primary T-cells isolated from peripheral blood are in G₀; they are small cells with 2 N DNA content and low protein and RNA content, and both protein and RNA content increase as cells enter G_1A from G₀ (see Ref. 5). Therefore, we used these cells to determine when p130 and pRb are phosphorylated as cells enter G₁ from G₀. Two color flow cytometry showed that T-cells enter G_1A at about 12-14 h after stimulation with PHA (Fig. 2A), at which time p130 Form 3 was detectable (Fig. 2B, lanes 4 and 5). Hyperphosphorylated pRb was also detectable by Western blotting at the same time (Fig. 2B, lanes 14 and 15). We have shown previously that p130 is also phosphorylated to Form 3 from Forms 1 and 2 as human primary CD34+ hemopoietic progenitor cells enter G_1A from G₀ (24). Thus, our data with two primary cell types are consistent with a model whereby p130 becomes inactivated at or before the G₀ to G_1A transition and thereafter remains inactive throughout the cell cycle of actively cycling cells.

View larger version (35K): [in this window] [in a new window]   Fig. 2.   Phosphorylation of p130 and pRb in primary T-cells during G0G1. A, two color flow cytometry. Human primary T-cells were stimulated with PHA for the times shown, fixed, and stained for DNA (PI) and total cell protein (FITC) The flow cytometry profiles are shown, and the percentages of cells in G0 and G1 are indicated below each panel. B, Western blots of p130 and pRb. The phosphorylation states of pRb and p130 in the primary T-cells stimulated for 0-20 h (lanes 1-8 and 11-18) were determined by Western blotting. Daudi cells cultured without (lanes 9 and 10) or with (lanes 19 and 20) -IFN for 48 h are shown. pRb(p) and pRb indicate hyper- and hypophosphorylated forms of pRb, respectively.

Regulation of p130 and pRb by -IFN-- Next, we investigated how p130 is regulated during cell cycle arrest caused by -IFN. At 24 h after the addition of -IFN, p130 Form 3 is partially depleted and Forms 1 and 2 are detectable. Loss of Form 3 occurs only in G₁ (Fig. 3A, lane 3) and not in S (lane 4) or G₂-M phases (lane 5). This is similar to pRb, which we showed previously is only dephosphorylated to a hypophosphorylated form in G₁ (5). After 36-48 h with -IFN, p130 Form 3 is replaced completely by Forms 1 and 2 (Fig. 3B, lane 2). However, this occurs at a time when hyperphosphorylated pRb is still present (compare Fig. 3B, lanes 2 and 6; see also Fig. 2B, lanes 10 and 20). In order to determine whether p130 Form 3 is depleted in phases other than G₁, we sorted Daudi cells that had been cultured with -IFN for about 40 h. Two color flow cytometry of DNA versus protein content showed that at this time cell cycle arrest in G₁ was not yet complete and that there was still a significant number of cells in mid to late S and G₂-M (data not shown). The cells in G₁ and S+G₂-M were isolated by flow cytometric sorting and, as shown in Fig. 3B, p130 was present as Forms 1 and 2 in both S+G₂-M, as well as in G₁ (lanes 3 and 4), whereas pRb was hypophosphorylated in G₁ and hyperphosphorylated in S+G₂-M (lanes 7 and 8). In comparison, both p130 and pRb were fully phosphorylated in cycling cells in S+G₂-M (Fig. 3B, lanes 1 and 5). Thus, in response to -IFN, the initial dephosphorylation of pRb and p130 both occur in G₁. However, thereafter, the phosphorylation state of p130 but not of pRb is regulated in S+G₂-M.

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Phosphorylation state of p130 during cell cycle arrest by -IFN. A, Western blot of p130 in Daudi cells cultured for 24 h with -IFN in G1 (lane 3; purity, 93%), S (lane 4; purity, 77%), G2-M (lane 5; purity, 87%) phases of the cell cycle, or a total cell lysate (lane 2). Proliferating cell lysate is shown in lane 1. B, Western blots of p130 (lanes 1-4) and pRb (lanes 5-8) in Daudi cells cultured with -IFN for 39 h are shown as follows: lanes 2 and 6, total lysate; lanes 3 and 7, G1 (purity, 98%); lanes 4 and 8, S+G2-M (purity, 66%). Proliferating cells in S+G2-M (lanes 1 and 5) are shown for comparison. (Note that p130 (lane 2) is the same as that shown in Fig. 1B, lane 12)

Regulation of p107 by -IFN-- We fixed and sorted actively cycling Daudi cells into various cell cycle phases (Fig. 4A) and analyzed the phosphorylation state of p107 by Western blotting (18). The fixing and sorting procedures did not alter the migration of p107 (Fig. 4B, lanes 1 and 2 (fixed), lanes 3 and 4 (not fixed)), and in different cell cycle phases, p107 is hypophosphorylated in G_1A (Fig. 4B, lane 5), is present as hypo- and hyperphosphorylated forms in G_1B (lane 6), and is hyperphosphorylated throughout the remainder of the cell cycle (lanes 7-9). These data are consistent with p107 being phosphorylated in late G₁ in cycling cells, and the timing is the same for primary T-cells entering the cell cycle for the first time from G₀ (data not shown). It has been suggested that there is a significant amount of hypophosphorylated p107 in S phase when fibroblasts are stimulated to enter the cell cycle for the first time after serum starvation (18). In order to determine whether p107 becomes hypophosphorylated in S phase in continuously cycling Daudi cells, we isolated proliferating cells in both early and late S phase (Fig. 4A), and in both cases, p107 was predominantly hyperphosphorylated (Fig. 4B, lanes 7 and 8). Thus, at least in proliferating Daudi cells, p107 only becomes hypophosphorylated in G_1A and is predominantly hyperphosphorylated from G_1B throughout the remainder of the cell cycle.

View larger version (24K): [in this window] [in a new window]   Fig. 4.   p107 in proliferating and -IFN-arrested cells. A, flow cytometric analysis of proliferating Daudi cells sorted into the cell cycle phases shown. The purity is indicated next to each profile. B, p107 Western blots of the sorted samples shown in A: G1A (lane 5), G1B (lane 6), early S (lane 7), late S+G2-M (lane 8), and G2-M (lane 9). Total lysates of Daudi cells cultured 39 h with -IFN or without and lysed directly (lanes 3 and 4) or fixed in 70% EtOH prior to lysis (lanes 1 and 2) The positions of hyper- and hypophosphorylated p107 (p107(p) and p107, respectively) are shown C, p107 Western blots of total cell lysates of Daudi cells cultured with (lanes 2, 4, and 6) or without (lanes 1, 3, and 5) -IFN for 24, 48, and 72 h. A tubulin blot of samples cultured for 72 h ±-IFN is shown below the corresponding p107 blot.

The Phosphorylation State of p130, pRb, and p107 during Cell Cycle Exit-- We have shown above that the phosphorylation state of p130 and pRb are differentially regulated in response to -IFN. In order to determine whether this is peculiar to -IFN or whether the same occurs in response to other signals to exit the cell cycle, we analyzed p130 and pRb in proliferating primary T-cells after IL-2 withdrawal, in Daudi cells exposed to a suboptimal serum concentration, and in HL60 myeloid cells cultured with RA. The results are shown in Fig. 5, A, B, and C, respectively. We observed two responses: (i) p130 Form 3 is lost before pRb is dephosphorylated in T-cells and Daudi cells withdrawn from growth factors (Fig. 5, A and B, lane 3); and (ii) p130 Form 3 and hyperphosphorylated pRb are depleted coordinately in HL60 cells between days 3 and 4 after RA treatment (Fig. 5C, lanes 2 and 3), at a time when the cells arrest in G₁. Therefore, -IFN and growth factor withdrawal have the same effect, causing depletion of p130 Form 3 before dephosphorylation of pRb, whereas we have not detected depletion of p130 Form 3 before pRb in HL60 cells cultured with RA. As also shown in Fig. 5C, p107 is dephosphorylated (lane 2) and then depleted (lane 3) from HL60 cells between days 3 and 4 with RA, which is similar to the regulation of p107 in Daudi cells by -IFN. p107, as well as p130 and pRb, is fully phosphorylated in proliferating cells cultured for 4 days without RA (Fig. 5C, lane 4).

View larger version (22K): [in this window] [in a new window]   Fig. 5.   Exit from the cell cycle. A, human primary T-cells proliferating in vitro were washed to remove IL-2 as described under "Materials and Methods." Samples were taken between 0 and 72 h after IL-2 withdrawal (lanes 1-5), and Western blots for p130 and pRb were done. B, proliferating Daudi cells were washed and cultured in RPMI/0.5% FCS for 4, 24, and 48 h (lanes 1-3). Western blots of p130 and pRb are shown. C, HL60 cells were cultured with RA, and samples were taken for p130, pRb, and p107 Western blots (lanes 1-3; days 2-4). At day 4, 72% of the cells in this experiment reduced nitro blue tetrazolium. Cells maintained for 4 days without RA (nitro blue tetrazolium-negative) are shown in lane 4.

E2F Complexes in -IFN Arrested Cells-- p130, pRb, and p107 each function at least in part by binding to E2F. The E2F proteins can be grouped into two classes, E2F-1, -2, and -3 and E2F-4 and -5, which have different specificities for binding members of the pRb family. In order to determine which forms of p130, pRb, and p107 associate with E2F during -IFN-mediated cell cycle arrest, we carried out coimmunoprecipitations with antibodies against one member of each class of E2F, namely E2F-1 and E2F-4. As shown in Fig. 6, hypophosphorylated pRb and predominantly p130 Form 1 were coimmunoprecipitated with E2F-4 (Fig. 6, A and B, lane 7) from lysates of Daudi cells cultured with -IFN for 24 h. We note that some of p130 Form 2 was also detectable on longer exposures but none of Form 3. Similar data were obtained for E2F-5 (not shown). Note that very little p130 or p107 and no detectable pRb were coimmunoprecipitated with E2F-4 from proliferating Daudi cells (Fig. 6, A and B, lane 6), and no p130, p107, or pRb was immunoprecipitated nonspecifically with rabbit IgG (Fig. 6, A and B, lanes 1 and 2). pRb coimmunoprecipitated with E2F-1 from -IFN treated cells (Fig. 6A, lane 9), and a small amount was detected in proliferating cells (Fig. 6A, lane 8). However, no p130 or p107 was detected coimmunoprecipitating with E2F-1 (Fig. 6B, lanes 8 and 9). Thus, pRb coimmunoprecipitates with both E2F-1 and -4 form lysates of Daudi cells cultured with -IFN, whereas p130 and p107 only bind E2F-4.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   E2F complexes with p130, pRb and p107. A and B, coimmunoprecipitation. Daudi cells were cultured with (lanes 2, 7, and 9) or without (lanes 1, 6, and 8) -IFN for 24 h, E2F-1 (lanes 8 and 9) and E2F-4 (lanes 6 and 7) were immunoprecipitated, and Western blots of pRb (A), p130, and p107 (B) were done. Control immunoprecipitations with rabbit IgG (lanes 1 and 2) and total lysates of cells cultured with -IFN for 0, 24, and 40 h (lanes 3-5) are shown. In B, the same blot was probed first for p107 and reprobed for p130. C, forms of p130 and E2F-4 that bind DNA. Lysates of 0.3-33 × 106 quiescent primary T-cells (lanes 2-5) or 0.17-33 × 106 cells cultured for 40 h +PHA (lanes 8-11 and 14) were incubated with and E2FWT DNA binding site coupled to beads as described under "Materials and Methods," and the proteins bound were analyzed by Western blotting. The Western blot was cut in half and probed for p130 (upper) and E2F-4 (lower). p130 and E2F-4 in 33 × 106 quiescent T-cells or in 3.3 or 33 × 106 T-cells +PHA that binds a mutant E2F DNA site are shown in lanes 1, 7 and 13, respectively. Lanes 6 and 12, total cell lysate of T-cells +PHA. The samples shown in lanes 1-11 and 12-14 are of two independent experiments. D, E2F-4 in Daudi cells +-IFN. Western blots of total cell lysates of CD34+ primary hemopoietic progenitor cells cultured for 0-4 days with SCF, IL-3, and IL-6 (lanes 1-3) and of Daudi cells cultured for 48 h ±-IFN (lanes 4 and 5) were probed for E2F-4. Lanes 6 and 7 are of E2F-4 in Daudi cells cultured for 24 h +-IFN that binds an E2FWT or E2FMUT DNA binding site, respectively. Lanes 1-5 and lanes 6 and 7 were run on separate gels. E, p130 and pRb in -IFN-arrested Daudi cells that bind DNA. The forms of pRb and p130 in extracts of Daudi cells cultured for 72 h +-IFN that bind E2FWT (lane 2) or E2FMUT (lane 1) DNA binding sites were determined by Western blotting. Total cell lysates of proliferating (lane 3) and cells cultured for 72 h +-IFN (lane 4) are shown for comparison. The blot was probed sequentially for pRb then p130.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Many proteins that are involved in controlling cell proliferation are regulated in susceptible cells by -IFN (reviewed in Ref. 70). The Daudi B-cell line has been used for a number of these studies because it is sensitive to growth arrest by -IFN. Proliferating Daudi cells cultured with -IFN arrest in G₁, and this is accompanied by the dephosphorylation of pRb to a hypophosphorylated form (5, 7, 50). Recently (62), it was shown that this causes pRb to complex with E2F-1 and E2F-4, and it was suggested that this results in repression of E2F-containing promoters, such as those in the cdc2 and c-myc genes. We showed previously (5) that in response to -IFN, pRb is dephosphorylated only in the G₁ phase of the cell cycle, and so it is only able to repress E2F activity once proliferating cells have entered G₁. pRb is the first member of a family of proteins that also includes p130 and p107, each of which can bind and repress E2Fs (3, 23). In the study presented here, we investigated whether -IFN regulates p130 and p107 and to what extent their regulation is also cell cycle-dependent.

We determined first whether the phosphorylation states of pRb, p130, and p107 are regulated in different cell cycle phases during continuous proliferation. Second, we investigated whether dephosphorylation of each is regulated coordinately during -IFN-mediated cell cycle arrest. To answer these questions, we employed a rapid fixation method, followed by flow cytometric sorting and Western blotting, a protocol that we have used previously to analyze pRb (5). We have shown here that (i) in actively cycling Daudi cells, p130 is present predominantly as phosphorylated Form 3 in all cell cycle phases. p107 is hypophosphorylated in G_1A, becomes hyperphosphorylated in G_1B, and remains fully phosphorylated throughout the remainder of the cell cycle. Regulation of the phosphorylation state of p107 during G₁ is similar to pRb, except that pRb is partially phosphorylated in G_1A (not hypophosphorylated) and is hyperphosphorylated from G_1B onwards (5). (ii) In cells arrested with -IFN, p130 Form 3 is depleted, possibly by dephosphorylation, and principally Form 1, with some Form 2, predominates in every cell cycle phase. In contrast, pRb is dephosphorylated to a hypophosphorylated state only in G₁. p107 is dephosphorylated and then depleted. p130 Form 1 and hypophosphorylated p107 coimmunoprecipitate with E2F-4 from lysates of -IFN-treated Daudi cells. Hypophosphorylated pRb coimmunoprecipitates with both E2F-4 and E2F-1. These forms of p130 and pRb bind to an E2F^WT DNA binding site. The difference between the amounts of p130, p107, and pRb coimmunoprecipitated with E2F-4 from proliferating and -IFN-treated cells is most probably due to changes in the phosphorylation state of pRb, p130, and p107 rather than to gross changes in the phosphorylation state of E2F-4. Thus, -IFN may bring about cell cycle arrest not only through pRb but also by regulating the phosphorylation state of p130 and the abundance of p107.

Several studies have investigated the E2F complexes that are present as cells enter the cell cycle from a resting state (see Ref. 26 and references therein). Their function is presumed to be to induce the orderly transcription of specific genes, such as those encoding thymidine kinase and cdc2, which are required for progression into S phase and for controlling passage through the cell cycle (22). This may be true for entry into the first cell cycle. However, once in cycle the abundance of, for example, cdc2 changes only in line with the total amount of cellular protein (71-73).³ Therefore, regulation of E2F activity by the pRb family is likely to be different in cells entering the cell cycle for the first time and in those proliferating continuously. Cycling cells contained predominantly phosphorylated p130 Form 3 in all cell cycle phases from G_1A through to G₂-M. This form does not bind E2F complexes and was not isolated by binding to an E2F DNA binding site. Because p130 only binds E2Fs and DNA when it is in Forms 1 and 2, we conclude that p130 has no major role in controlling E2F function in Daudi cells during continuous proliferation. Our conclusions differ from those published recently that showed that a p130-E2F complex exists during G₁ (74). It is not clear whether there are differences because of the cell types or the methods used in the two studies. In addition, our data show that p130 is phosphorylated as quiescent primary CD34+ hemopoietic progenitors (24) and T-cells exit G₀ and enter G_1A. It has been suggested that p130 maintains cells in G₀ (14, 17, 37), and inactivation of p130 by phosphorylation upon entry into G_1A is most consistent with p130 having a role in allowing the transition from G₀G_1A rather than controlling progression through G₁.

In contrast to p130, both pRb and p107 in actively cycling cells are dephosphorylated during G₁ (pRb partially) and are fully phosphorylated from the G₁-S border onwards. Thus, both could have a role to play in G₁. Recently, it was shown that p107-E2F complexes are present in the cytosol during G₁, whereas pRb-E2F complexes are nuclear (74). Thus, p107 in actively cycling Daudi cells may bind E2F (E2F-4 and -5) in a cytosolic form during G_1A, which is released as p107 becomes phosphorylated in G_1B. In contrast to p107, pRb can only be hypophosphorylated for a brief period during G_1A, because we detected multiple, partially phosphorylated forms in this subphase of the cell cycle. We have not detected these phosphorylated forms bound to E2F or to an E2F^WT DNA site, and it is possible therefore that other proteins in the cell (23, 75) are regulated by these forms of pRb, but this has yet to be shown. Once cells have progressed through G₁, the data presented here show that both p107 and pRb are predominantly hyperphosphorylated throughout S and G₂-M. Our data on pRb are consistent with dissociation of pRb and E2F during the G₁S phase transition reported previously (see Ref. 2). However, p107-E2F complexes have been detected by electrophoretic mobility shift assays in extracts of cells from late G₁ into S phase (21, 74, 76), and p107 has also been reported to become dephosphorylated again late in S phase (18). From the data presented here and from hydroxyurea block/release experiments (not shown), we would suggest that because most of the p107 is hyperphosphorylated and hence inactivated from G_1B through to G₂-M, p107 may only have a major role in binding E2F in actively dividing cells during G_1A.

In response to -IFN, p130 Form 3 in Daudi cells was completely depleted at a time when hyperphosphorylated pRb was not. Flow cytometric sorting experiments showed that p130 Form 3 was depleted in all cell cycle phases after culturing for 40 h with -IFN, whereas pRb was only dephosphorylated in G₁. Thus there is a clear difference in the mechanisms regulating these proteins in different cell cycle phases. Our data show that there is an initial response to -IFN during G₁ that leads to the dephosphorylation of pRb and loss of p130 Form 3, but later, p130 Form 3 is depleted during S+G₂-M. pRb may become dephosphorylated during mitosis (4) and, as certain cdks are inhibited by -IFN (51, 52, 55), pRb would remain hypophosphorylated in G₁. However, the effects of -IFN on members of the cdk family, such as cdk4 and cdk6, which are active during early to mid G₁ (3, 70) are as yet unknown. Such a mechanism does not explain why p130 Form 3 is depleted from all cell cycle phases. Our dephosphorylation experiments with -phosphatase would suggest that p130 Form 3 could be converted to Form 1 and/or Form 2 by dephosphorylation. However, we cannot rule out the possibility that p130 Form 3 could also be degraded. In continuously cycling cells, less p130 was detected in samples of cells sorted into S and G₂-M as compared with G₁, but it is not depleted in G₂-M, as reported for other cell types (14). There is a precedent for two mechanisms regulating the activity of one protein. The amount of phosphorylated STAT1 is regulated both by the ubiquitin-proteasome pathway (77) and by dephosphorylation by a tyrosine phosphatase (47), and it will be interesting to determine whether p130 is regulated in a similar manner.

In order to determine whether loss of p130 Form 3 preceding the dephosphorylation of pRb is peculiar to cell cycle arrest caused by -IFN, we analyzed what occurs during cell cycle exit for a number of hemopoietic cell types: by removing IL-2 from continuously proliferating primary T-cells, in Daudi cultured in low serum, and in HL60 cells during myeloid differentiation. In all cases, p130 Form 3 was depleted as cells arrested in G₁, consistent with a role for p130 in promoting exit from the cell cycle. In HL60 cells cultured with retinoic acid, p130 and pRb were both dephosphorylated at the same time, in agreement with experiments that suggest that cell cycle arrest may be mediated by sequestration of E2F by pRb and p130 (61).

From previous work, it was suggested that -IFN exerts its antiproliferative effects in part via the pRb pathway (5, 7, 50, 51). The cytostatic effect of -IFN may occur by preventing the initial phosphorylation of pRb during G_1A (5), by decreasing cdk activity (51, 52) and E2F binding (8, 62). Our data now show that -IFN also causes the accumulation of p130 Form 1 and the depletion of p107. Finally, our coimmunoprecipitation experiments with lysates of Daudi cells cultured with and without -IFN show that hypophosphorylated forms of pRb and p107, and p130 (principally Form 1) all associate with E2F-4, whereas E2F-1 only coimmunoprecipitates pRb. These data are in agreement with Ikeda et al. (61), who concluded that pRb may be a general regulator of E2Fs as cells exit the cell cycle, whereas p130 is more specific to E2F-4 and -5. Such complexes can also bind an E2F^WT DNA site in vitro, but given that the subcellular localization of each complex may be different (74), it remains to be determined precisely how pRb, p130, and p107 in -IFN-arrested cells control E2F: whether it is bound to a nuclear E2F DNA site or whether, for example, p130-E2F binding occurs in the cytosol. It has been shown by electrophoretic mobility shift assays that the principal E2F complex in extracts of primary hemopoietic cells in G₀ contains p130-E2F-4-DP-1 (17, 24, 25), which we have shown here contains p130 Forms 1 and 2 together with a hyperphosphorylated form of E2F-4. Thus, if p130-E2F-4-DP-1 defines cells in G₀ (61), our observations (Ref. 24 and this paper) would suggest that such a complex contains hyperphosphorylated E2F-4. In contrast, the p130-E2F-4 complex that exists in Daudi cells arrested by -IFN contains hypophosphorylated forms of E2F-4. Thus, by this criterion, -IFN causes Daudi cells to arrest in G₁ rather than in G₀, and this occurs in part by regulating the phosphorylation state of p130 rather than causing the hyperphosphorylation of E2F-4.