Cell-cycle transitions are controlled by CDKs (cyclin-dependent kinases), whose activation is usually associated with the binding of cyclins. RINGO/Speedy proteins can also bind to and activate CDKs, although they do not have amino acid sequence homology with cyclins. The RINGO/Speedy family members studied so far positively regulate cell-cycle progression. In the present paper, we report the biochemical and functional characterization of RINGO/Speedy E. We show that RINGO/Speedy E is a functionally distant member of this protein family that negatively affects cell-cycle progression. RINGO/Speedy E overexpression inhibits the meiotic progression in Xenopus oocytes as well as the proliferation of mammalian cells. RINGO/Speedy E can bind to endogenous CDK1 and CDK2 in both cellular systems. However, the RINGO/Speedy E-activated CDKs have different substrate specificity than the CDKs activated by other RINGO/Speedy proteins, which may account for their different effects on the cell cycle. Our results indicate that, although all RINGO/Speedy family members can activate CDKs, they may differently regulate cell-cycle progression.
Cell-cycle progression is controlled by a family of serine/threonine protein kinases named CDKs (cyclin-dependent kinases). CDKs are essentially inactive as monomers and their activation requires the binding of activating subunits named cyclins as well as the phosphorylation of a conserved residue in the T-loop of the kinase domain [1–4]. In addition to CDK activation, cyclins contribute to substrate-specificity determination as well as to the subcellular localization of CDKs [5–7]. CDKs can be negatively regulated by the phosphorylation of Thr-14 and Tyr-15 in the ATP-binding pocket [2,8].
Different CDKs have been implicated in the regulation of different cell-cycle stages . In particular, CDK1 bound to B-type cyclins, a complex that is usually referred to as MPF (M-phase- or maturation-promoting factor), plays a key role in M-phase entry and progression. CDK1–cyclin B complexes are formed before mitosis, but are maintained inactive owing to the phosphorylation of Thr-14 and Tyr-15 by the kinases Wee1 and Myt1. These inactive complexes are referred to as pre-MPF. Dephosphorylation of Thr-14 and Tyr-15 is carried out by phosphatases of the CDC25 family. MPF activation at the onset of mitosis results from both the inhibition of Wee1/Myt1 and the stimulation of CDC25, with phosphorylation playing an important role in both processes [8,10–12]. MPF itself is thought to phosphorylate Wee1, Myt1 and Cdc25, forming a positive-feedback loop mechanism, which contributes to MPF activation at mitotic entry. The regulation of MPF activity has been studied intensively in Xenopus oocytes [13–15]. Fully grown Xenopus oocytes are induced to enter meiosis by the hormone progesterone in a process called meiotic maturation. Progesterone stimulates several signal-transduction pathways leading to Myt1 inhibition, CDC25 stimulation and pre-MPF activation . In G2-arrested Xenopus oocytes, only 10–20% of CDK1 is associated with cyclin B forming the pre-MPF [17,18]. The remaining CDK1 is present as a monomer, and this subpopulation is thought to contribute to the activation of pre-MPF during the G2/M-phase progression of oocytes [13,19].
RINGO/Speedy proteins can activate some CDKs, although they do not share significant sequence homology with cyclins . The first member of this family was identified in Xenopus as a potent inducer of oocyte maturation [21,22]. Endogenous XRINGO (Xenopus RINGO)/Speedy accumulates in oocytes upon progesterone stimulation and contributes to meiotic progression, probably by CDK1 or CDK2 activation [21,23,24]. An orthologue named Spy1 or RINGO/Speedy A, has been proposed to participate in the G1–S transition [25,26] and in the DNA-damage-response pathway [27,28] in cultured mammalian cells. More recently, four additional mammalian RINGO/Speedy proteins have been identified [29,30]. These proteins can all bind to and activate CDK1 and CDK2, and some of them can also activate CDK5 .
RINGO/Speedy E is a more distant and less characterized family member. We reported previously that RINGO/Speedy E (formerly Ringo1) was able to bind to and activate CDK1, CDK2 and CDK5 . In the present paper, we provide more detailed biochemical and functional characterization of this protein. We show that RINGO/Speedy E impairs cell-cycle progression in Xenopus oocytes and in mammalian cells. Moreover, RINGO/Speedy E activates endogenous CDKs, but provides different substrate specificity from the CDKs activated by other RINGO/Speedy proteins. Thus RINGO/Speedy family members may differently regulate cell-cycle progression by providing CDKs with different substrate specificities.
Constructs for expression in Xenopus oocytes, FTX4–HA (haemagglutinin)–CDK2 (Xenopus) wild-type and different mutants, FTX5–XRINGO/Speedy, FTX5–RINGO/Speedy E, FTX5–RINGO/Speedy A and FTX5–Myt1 (Xenopus) C-terminus, as well as constructs to express recombinant proteins in Escherichia coli, GST (glutathione transferase)–RINGO/Speedy E, GST–CDK1 (Xenopus) and GST–Myt1 (Xenopus) C-terminus, have been described previously [21,29,31]. A construct to express in E. coli GST–RINGO/Speedy A1 was prepared by cloning human RINGO/Speedy A1 into the pGEX-KG vector digested with EcoRI. For expression in mammalian cells with an N-terminal GFP (green fluorescent protein) tag, RINGO/Speedy E was subcloned into the pEGFP-C2 vector as an EcoRI fragment.
Xenopus oocyte manipulation and injection
Xenopus oocytes were sorted, injected with in vitro-transcribed mRNAs and lysed as described previously .
Cell culture and transfection
U2OS cells were maintained in DMEM (Dulbecco's modified Eagle's medium) supplemented with 10% fetal bovine serum. Cells were transiently transfected in suspension using FuGENE™ 6 reagent (Roche). In a typical experiment, transfection of U2OS cells in suspension with FuGENE™ 6 reagent resulted in more than 70% of transfection efficiency. Cellular lysates were prepared in buffer containing 150 mM Tris/HCl, pH 7.5, 1 mM EDTA, 1 mM EGTA, 10 mM NaF, 1% (v/v) Nonidet P40, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 10 μg/ml pepstatin, 0.2 mM PMSF, 0.1 mM sodium orthovanadate, 1 mM DTT (dithiothreitol) and 1 mM benzamidine.
Immunoblotting and immunoprecipitation
Primary antibodies against Xenopus CDK1 (3E1 monoclonal) and cyclin B2 (affinity-purified rabbit antisera) were provided by J. Gannon and T. Hunt (Cancer Research UK, Clare Hall Laboratories, South Mimms, U.K.). The rabbit antiserum 3297.1 was used to detect the MAPK (mitogen-activated protein kinase) Xp42/XMpk1 . Antibodies against p90RSK (p90 ribosomal S6 kinase) and Myc were purchased from Santa Cruz Biotechnology and the anti-HA antibody was from Roche. An anti-RINGO/Speedy E monoclonal antibody was generated by the CNIO Monoclonal Antibody Unit. The blots were developed using Alexa Fluor® 680- (Molecular Probes) or Li-Cor IRDye 800- (Rockland) labelled antibodies with the Odyssey Infrared Imaging System (Li-Cor).
For IP (immunoprecipitation), 5 μl of anti-Myc (Santa Cruz Biotechnology; sc-40 AC), 5 μl of anti-HA (Santa Cruz Biotechnology; sc-7392 AC) or 10 μl of Suc1/Cks (Upstate Biotechnology) antibody–agarose conjugates were incubated with the oocyte lysates for 2 h at 4 °C. IP of cellular lysates with a RINGO/Speedy E monoclonal antibody was carried out overnight at 4 °C followed by incubation with IgG–agarose at 4 °C for 1 h. The beads were washed three times in lysis buffer and analysed by immunoblotting, or washed further in HIK (histone H1 kinase) buffer and used for kinase assays .
In vitro kinase assays
For CDK1–RINGO/Speedy kinase assays, bacterially produced GST–CDK1 (2 μg) was pre-incubated in concentrated Xenopus egg extracts and recovered on glutathione–Sepharose beads, as described in . The beads were washed and incubated at 30 °C for 10 min with GST–RINGO/Speedy proteins (2 μg) and then incubated further for 15 min in kinase buffer containing histone H1 (3 μg) or GST–Myt1 C-terminus (2 μg) as substrates. Phosphorylated proteins were separated in SDS/15% PAGE and detected by autoradiography.
Cell-cycle analysis, proliferation and apoptosis assays
Cell-cycle distribution was analysed by flow cytometry. Cells were trypsinized and fixed in chilled 70% ethanol for at least 1 h. Cells were then incubated in PBS containing 0.2 mg/ml RNAse and stained with 25 μg/ml propidium iodide for 30 min. Flow-cytometric analysis was carried out using a FACScan system and CellQuest program.
For proliferation assays, U2OS cells were transfected with GFP or GFP–RINGO/Speedy E constructs and seeded in a 96-well plate at a density of 2000 cells/well. Proliferation was followed for 3 days using the Cell Proliferation kit, MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl-2H-tetrazolium bromide] (Roche Diagnostics). For colony-formation assays, cells were trypsinized 48 h after transfection, counted and seeded (200 cells) in 10-cm-diameter dishes. Colonies were visualized by staining with Crystal Violet and were counted 14 days after plating. Apoptotic cell death was evaluated by ELISA using the Cell Death Detection ELISAPLUS kit (Roche) according to the manufacturer's instructions.
Detection of GFP–RINGO/Speedy E
U2OS cells were transfected with the expression construct for GFP–RINGO/Speedy E and cultured on glass coverslips for 24 h. Cells were then fixed with 4% (v/v) formaldehyde, and subcellular localization of RINGO/Speedy E was analysed by fluorescence microscopy (Zeiss).
All results are expressed as means±S.E.M. for at least three independent experiments. Statistical analysis was performed using Student's t test for paired and unpaired data compared with control values using the statistic Prism software (GraphPad). P<0.05 was considered to be statistically significant.
Biochemical properties of RINGO/Speedy E differ from other RINGO/Speedy proteins
We examined the effect of RINGO/Speedy E expression on Xenopus oocyte maturation, the assay used to identify and characterize other RINGO/Speedy proteins. Expression of either XRINGO/Speedy or RINGO/Speedy A suffices to induce maturation of Xenopus oocytes in the absence of progesterone, whereas RINGO/Speedy B expression accelerates progesterone-induced maturation [21,22,25,30]. In contrast with those results, we found that expression of RINGO/Speedy E not only was unable to induce oocyte maturation on its own, but also strongly inhibited oocyte maturation induced by progesterone (Figure 1A). The lack of maturation was scored morphologically by the absence of a white spot on the animal pole of the oocyte and the presence of the germinal vesicle inside the oocyte (Figure 1A, left-hand panel). This was also confirmed biochemically, as MPF, MAPK and p90RSK were all inactive in the RINGO/Speedy E-expressing oocytes treated with progesterone, in contrast with progesterone-treated oocytes injected with water (Figure 1A, right-hand panel).
Characterization of RINGO/Speedy E expressed in Xenopus oocytes
RINGO/Speedy E can bind to and activate several CDKs . Previous studies indicated that binding of cyclins and XRINGO/Speedy to the CDKs involves residues in the PSTAIRE helix and the T-loop [29,33]. To study the interaction between RINGO/Speedy E and CDKs, we co-expressed in Xenopus oocytes RINGO/Speedy E with a collection of CDK2 mutants  and then evaluated the binding by co-IP experiments. Surprisingly, all the CDK2 mutants tested, including K33A, R50A and R150A that are impaired in binding to XRINGO/Speedy and cyclin A1 , were able to bind to RINGO/Speedy E as efficiently as the wild-type CDK2 (Figure 1B, IP Myc). For comparison, a similar experiment was carried out using RINGO/Speedy A. As expected, the CDK2 mutants that were impaired in binding to XRINGO/Speedy and cyclin A1 did not bind to RINGO/Speedy A (Figure 1C). We also tested the activation of the different CDK2 mutants by co-IP experiments. Consistent with the binding data, RINGO/Speedy E increased the activity of all the CDK2 mutants, except for the catalytically inactive K33A (Figure 1B, IP HA). We could not perform the same experiment with RINGO/Speedy A, because this protein induces oocyte maturation and the concomitant up-regulation of endogenous cyclins, making it difficult to elucidate the relative contribution of RINGO/Speedy A and cyclins to the H1K activity measured in CDK2 immunoprecipitates. Altogether, our results suggested that RINGO/Speedy E binds CDKs differently from other RINGO/Speedy proteins or cyclin A1.
RINGO/Speedy E activates endogenous CDK1 in Xenopus oocytes
Although progesterone-treated or untreated RINGO/Speedy E-expressing oocytes did not mature, according to morphological and biochemical criteria (Figure 1A), the total lysates from these oocytes contained significant levels of H1K activity (Figure 2A). The H1K activity detected in total lysates of mature oocytes originates mostly from active MPF. Interestingly, pre-MPF was not activated in the oocytes expressing RINGO/Speedy E, as indicated by the existence of a slow-migrating and Thr-14/Tyr-15-phosphorylated CDK1 band (Figure 1A, right-hand panel). The H1K activity in RINGO/Speedy E-expressing oocytes coimmunoprecipitates with RINGO/Speedy E (Figure 2A), suggesting that it originated from the direct activation of endogenous kinases other than MPF by RINGO/Speedy E.
RINGO/Speedy E activates monomeric CDK1 in Xenopus oocytes
Considering the ability of RINGO/Speedy E to bind to different CDKs, we predicted that CDKs could account for the observed H1K activity. Indeed, addition of the CDK inhibitor roscovitine inhibited the H1K activity in total lysates of RINGO/Speedy E-expressing oocytes as it did in oocytes treated with progesterone (Figure 2B). Moreover, the H1K activity of oocytes expressing RINGO/Speedy E was immunoprecipitated with CDK1 antibodies and was also detected in pull-downs with the CDK1-binding protein Suc1/Cks (Figure 2C). Interestingly, we detected binding of RINGO/Speedy E to endogenous CDK1 in the co-IP experiments, but the binding was restricted only to the faster-migrating CDK1 band, which is thought to correspond to cyclin-free, monomeric CDK1 non-phosphorylated on Thr-14 and Tyr-15 (Figure 2D). Immunoblotting confirmed that CDK1 bound to RINGO/Speedy E was not in a complex with cyclin B2 (Figure 2D). Altogether, our results strongly suggest that RINGO/Speedy E binds to and activates endogenous monomeric CDK1 in Xenopus oocytes.
Differential substrate specificity of CDK1–RINGO/Speedy E
The ability of RINGO/Speedy E to stimulate the kinase activity of CDK1 on histone H1 without inducing oocyte maturation was intriguing. There is evidence that RINGO/Speedy proteins might contribute to the substrate specificity of CDKs . Since different cyclins can modulate the substrate preferences of a given CDK [35–37], it is possible that RINGO/Speedy family members might also determine distinct specificities in the phosphorylation of CDK substrates. Therefore we hypothesized that the reason RINGO/Speedy E-activated CDK1 was not able to trigger Xenopus oocyte maturation could be that RINGO/Speedy E was impaired in targeting CDK1 to oocyte substrates relevant for meiotic maturation. To test this possibility, we followed the phosphorylation of Myt1 in oocytes expressing either XRINGO/Speedy or RINGO/Speedy E. Myt1 is an important regulator of the G2 arrest in Xenopus oocytes, which is known to be phosphorylated upon M-phase entry [31,38–40]. As expected, XRINGO/Speedy induced the phosphorylation of Myt1, which resulted in reduced electrophoretic mobility of the protein. The expression of RINGO/Speedy E also induced Myt1 phosphorylation, but much less efficiently than the phosphorylation induced by XRINGO/Speedy. Moreover, when H1K activity levels were compared with the degree of Myt1 phosphorylation at the 3 h time point, it seemed clear that RINGO/Speedy E induced the phosphorylation of Myt1 less efficiently than XRINGO/Speedy (Figure 3A).
Myt1 is a poor substrate for CDK1–RINGO/Speedy E
Next, we performed in vitro kinase assays using recombinant CDK1 activated by either RINGO/Speedy E or RINGO/Speedy A, which is the mammalian family member most similar to XRINGO and is able to induce Xenopus oocyte maturation . In agreement with the results in oocytes, we found that the two RINGO/Speedy proteins provided CDK1 with different abilities to phosphorylate either histone H1 or Myt1 in vitro. We needed to use about four times more activity of CDK1–RINGO/Speedy E than CDK1-RINGO/Speedy A, based on histone H1 as a reference substrate, in order to obtain the same level of Myt1 phosphorylation (Figure 3B). Thus it is likely that the reduced efficiency of RINGO/Speedy E to target CDK1 to appropriate substrates accounts for its inability to induce oocyte maturation.
RINGO/Speedy E blocks oocyte maturation by sequestering CDK1
The activation of monomeric CDK1 has been shown to contribute to progesterone-induced MPF activation and oocyte maturation . As RINGO/Speedy E can bind to monomeric CDK1 in oocytes (Figure 2D), we postulated that the inhibition of oocyte maturation by RINGO/Speedy E could be due to the sequestration of endogenous monomeric CDK1. To test this possibility, we overexpressed CDK1 in the oocytes expressing RINGO/Speedy E. Interestingly, CDK1 overexpression was able to rescue the inhibitory effect of RINGO/Speedy E on progesterone-induced oocyte maturation. The same result was observed upon overexpression of the CDK1 mutant T161A, which cannot be phosphorylated in the T-loop. Moreover, the oocytes expressing RINGO/Speedy E together with either wild-type or T161A CDK1 matured significantly faster in response to progesterone stimulation than water-injected oocytes (Figure 4A). In contrast, the kinase-dead mutant CDK1 K33R was unable to rescue the inhibition by RINGO/Speedy E (Figure 4A). We confirmed that wild-type and the two mutants of CDK1 were all able to bind to RINGO/Speedy E with similar efficiencies in co-IP experiments (Figure 4B). We also found that the RINGO/Speedy E-triggered H1K activity in oocyte lysates was increased upon co-expression with either wild-type or T161A mutant CDK1, but not in the case of CDK1 K33R expression (Figure 4B). These results indicate that RINGO/Speedy E can activate CDK1 in the absence of the T-loop phosphorylation, as reported previously for XRINGO/Speedy and RINGO/Speedy A [24,34]. More importantly, the results strongly suggest that RINGO/Speedy E inhibits cell-cycle progression in Xenopus oocytes by sequestering CDK1.
CDK1 overexpression rescues the inhibition of oocyte maturation by RINGO/Speedy E
RINGO/Speedy E overexpression impairs cell-cycle progression of U2OS cells
Next, we investigated the effect of RINGO/Speedy E overexpression on the cell cycle of mammalian cells. GFP–RINGO/Speedy E overexpressed in U2OS cells was found to localize to the nucleus (Figure 5A) and was able to bind to endogenous CDK1 and CDK2 in co-IP experiments (Figure 5B). Interestingly, using flow-cytometric analysis we found that the overexpression of RINGO/Speedy E severely disturbed cell-cycle distribution of U2OS cells. Most notably, the G1 population was significantly reduced, while the G2/M population, and to a lesser extent the S-phase population, proportionally increased, suggesting that RINGO/Speedy E overexpression might impair G2- and/or M-phase progression (Figure 5C). We also noticed that the percentage of sub-G0 cells was higher in GFP–RINGO/Speedy E-expressing cells, suggesting that inhibition of cell-cycle progression by RINGO/Speedy E might eventually lead to apoptosis. This was confirmed by measuring apoptosis in RINGO/Speedy E-expressing cells directly (Figure 5D). Consistent with these observations, cells expressing RINGO/Speedy E had reduced proliferation ability (Figure 5E) and also produced significantly fewer colonies when compared with GFP-transfected cells (Fig. 5F). Taken together, these results indicate that RINGO/Speedy E overexpression interferes with cell-cycle progression in mammalian cells.
Effect of RINGO/Speedy E overexpression on cell-cycle progression of mammalian cells
CDK activity can be regulated by both cyclins and RINGO/Speedy proteins. Although the function of different members of the RINGO/Speedy family in cell-cycle regulation is just starting to be addressed, the RINGO/Speedy proteins characterized so far positively modulate progression through the meiotic and mitotic cell cycles. We have now found that one member of this family, RINGO/Speedy E, negatively affects cell-cycle progression, although it can activate both CDK1 and CDK2.
XRINGO/Speedy was first identified as a potent inducer of the meiotic maturation in Xenopus oocytes, and some of the mammalian RINGO/Speedy proteins were also shown to positively regulate Xenopus oocyte maturation [25,30]. In contrast, RINGO/Speedy E overexpression inhibits progesterone-induced maturation. Intriguingly, although CDK1 activation is usually associated with oocyte maturation, and RINGO/Speedy E can apparently activate the endogenous CDK1 in oocytes, the RINGO/Speedy E-expressing oocytes did not mature. We found that RINGO/Speedy E-activated CDK1 phosphorylates Myt1, an important regulator of pre-MPF activation in oocytes, less efficiently than other RINGO/Speedy proteins. This suggested that the reduced ability of RINGO/Speedy E to target CDK1 to certain substrates in oocytes might account for its inability to induce oocyte maturation. These results strongly argue that RINGO/Speedy proteins contribute to the substrate specificity of CDKs and that different RINGO/Speedy proteins could have diverse biological functions by providing CDKs with different substrate preferences, as cyclins do [35–37].
Overexpression of CDK1 was able to rescue the RINGO/Speedy E inhibition of progesterone-induced maturation, suggesting that sequestration of endogenous CDK1 is a likely mechanism by which overexpression of this protein interferes with oocyte maturation. This prediction is consistent with the observed interaction between RINGO/Speedy E and endogenous monomeric CDK1, and with the ability of kinase-dead CDK1 or anti-CDK1 antibodies to inhibit progesterone-induced maturation . One of the proteins that can activate monomeric CDK1 in oocytes is XRINGO/Speedy itself, which is up-regulated during the meiotic maturation . Therefore it is attractive to speculate that RINGO/Speedy E might inhibit oocyte maturation by competing with the endogenous XRINGO/Speedy for binding to the monomeric CDK1. This possibility would be consistent with the observation that RINGO/Speedy E binds to CDK1 more efficiently than XRINGO/Speedy . Therefore, if the two proteins are available for binding to endogenous CDK1, RINGO/Speedy E will probably out-compete XRINGO/Speedy.
The negative effect of RINGO/Speedy E on the meiotic cell cycle of Xenopus oocytes suggested that RINGO/Speedy E could also negatively regulate the mammalian cell cycle. Indeed, we have shown that RINGO/Speedy E overexpression impairs the proliferation of mammalian cells, which is just the opposite effect reported for RINGO/Speedy A overexpression . We also found that the overexpressed RINGO/Speedy E interacts with endogenous CDK1 and CDK2 in mammalian cells and induces the accumulation of cells in the G2- and M-phases of the cell cycle. Binding of RINGO/Speedy E to CDK1 may interfere with its role in the regulation of the G2–M transition and account for the negative effect of RINGO/Speedy E in mammalian cells, as in Xenopus oocytes. Alternatively, the phosphorylation of particular proteins by CDK1–RINGO/Speedy E might lead to slower progression through the G2- and M-phases of the cell cycle.
RINGO/Speedy proteins are strong candidates to regulate cell-cycle transitions through CDK1 and CDK2 activation. Current evidence indicates that XRINGO/Speedy and RINGO/Speedy A can positively regulate cell-cycle progression. In contrast, RINGO/Speedy E negatively affects cell-cycle progression, although it can potentially activate CDKs. Future work should investigate whether RINGO/Speedy E provides a general mechanism of negative cell-cycle control and how different RINGO/Speedy proteins co-operate with cyclins to regulate cell-cycle progression through CDK activation.
We thank Giovanna Roncador (CNIO Monoclonal Antibody Unit) for RINGO/Speedy E antibody generation, and Esther Seco for technical support. This work was funded by the Spanish Ministerio de Educacion y Ciencia (grant BFU2004-03566).
green fluorescent protein
histone H1 kinase
mitogen-activated protein kinase
M-phase- or maturation-promoting factor
p90 ribosomal S6 kinase