High-mobility-group proteins P1, I and Y as substrates of the M-phase-specific p34cdc2/cyclincdc13 kinase

All dividing cells entering the M phase of the cell cycle undergo the transient activation of an M-phase-specific histone H1 kinase which was recently shown to be constituted of at least two subunits, p34cdc2 and cyclincdc13. The DNA-binding high-mobility-group (HMG) proteins 1, 2, 14, 17, I, Y and an HMG-like protein, P1, were investigated as potential substrates of H1 kinase. Among these HMG proteins, P1 and HMG I and Y are excellent substrates of the M-phase-specific kinase obtained from both meiotic starfish oocytes and mitotic sea urchin eggs. Anticyclin immunoprecipitates, extracts purified on specific p34cdc2-binding p13suc1-Sepharose and affinity-purified H1 kinase display strong HMG I, Y and P1 phosphorylating activities, demonstrating that the p34cdc2/cyclincdc13 complex is the active kinase phosphorylating these HMG proteins. HMG I and P1 phosphorylation is competitively inhibited by a peptide mimicking the consensus phosphorylation sequence of H1 kinase. HMG I, Y and P1 all possess the consensus sequence for phosphorylation by the p34cdc2/cyclincdc13 kinase (Ser/Thr-Pro-Xaa-Lys/Arg). HMG I is phosphorylated in vivo at M phase on the same sites phosphorylated in vitro by H1 kinase. P1 is phosphorylated by H1 kinase on sites different from the sites of phosphorylation by casein kinase II. The three thermolytic phosphopeptides of P1 phosphorylated in vitro by purified H1 kinase are all present in thermolytic peptide maps of P1 phosphorylated in vivo in proliferating HeLa cells. These phosphopeptides are absent in nonproliferating cells. These results demonstrate that the DNA-binding proteins HMG I, Y and P1 are natural substrates for the M-phase-specific protein kinase. The phosphorylation of these proteins by p34cdc2/cyclincdc13 may represent a crucial event in the intense chromatin condensation occurring as cells transit from the G2 to the M phase of the cell cycle.     

Phosphorylation of rap1GAP in vivo and by cAMP-dependent kinase and the cell cycle p34cdc2 kinase in vitro.

rap1GAP is a GTPase activating protein that specifically stimulates the GTP hydrolytic rate of the ras-related protein p21rap1.rap1GAP undergoes post-translational modification that causes a substantial change in its mobility on sodium dodecyl sulfate-polyacrylamide gels. At least part of this modification is due to the phosphorylation. Expression of a rap1GAP cDNA in insect cells labeled with 32Pi resulted in high level incorporation of radioactivity into serine residues of the expressed protein. Purified rap1GAP was phosphorylated in vitro by cAMP-dependent kinase and the cell cycle p34cdc2 kinase. The molar ratio of incorporated phosphate/rap1GAP was approximately 3 by cAMP-dependent kinase and 2 by p34cdc2. The sites of phosphorylation by both kinases were localized to a 100-residue segment contained in the carboxyl-terminal region of the predicted primary structure of rap1GAP. Highly favorable recognition sequences for the two kinases are contained within this fragment and are proposed as the sites of phosphorylation. Treatment of SK-MEL-3 cells with dibutyryl cAMP promoted phosphorylation of rap1GAP in vivo. Based on the results of comparative phosphopeptide mapping the sites of phosphorylation in vivo and in vitro are identical.     

p34cdc2 is physically associated with and phosphorylated by a cdc2-specific tyrosine kinase.

The mammalian homologue of the yeast cdc2 gene encodes a 34-kilodalton serine/threonine kinase that is a subunit of M phase-promoting factor. Recent studies have shown that p34cdc2 is also a major tyrosine-phosphorylated protein in HeLa cells and that its phosphotyrosine content is cell cycle regulated and related to its kinase activity. Here, we show that cdc2 is physically associated with and phosphorylated in vitro by a highly specific tyrosine kinase. Tyrosine phosphorylation of cdc2 in vitro occurs at tyrosine 15, the same site that is phosphorylated in vivo. The association between the two kinases takes place in the cytosolic compartment and involves cyclin B-associated cdc2. Evidence is presented that a substantial fraction of cytosolic cdc2 is hypophosphorylated, whereas nuclear cdc2 is hyperphosphorylated. Finally, we show that the tyrosine kinase associated with cdc2 may be a 67-kilodalton protein and is distinct from src, abl, fms, and other previously reported tyrosine kinases.     

Human Wee1 kinase inhibits cell division by phosphorylating p34cdc2 exclusively on Tyr15.

In fission yeast, the M-phase inducing kinase, a complex of p34cdc2 and cyclin B, is maintained in an inhibited state during interphase due to the phosphorylation of Cdc2 at Tyr15. This phosphorylation is believed to be carried out primarily by the Wee1 kinase. In human cells the negative regulation of p34cdc2/cyclin B is more complex, in that Cdc2 is phosphorylated at two inhibitory sites, Thr14 and Tyr15. The identities of the kinases that phosphorylate these sites are unknown. Since fission yeast Wee1 kinase behaves as a dual-specificity kinase in vitro, a popular hypothesis is that a human Wee1 homolog might phosphorylate p34cdc2 at both sites. We report here that a human gene, identified as a possible Wee1 homologue, blocks cell division when overexpressed in HeLa cells. This demonstrates functional conservation of the Wee1 mitotic inhibitor. Contrary to the dual-specificity kinase hypothesis, purified human Wee1 phosphorylates p34cdc2 exclusively on Tyr15 in vitro; no Thr14 phosphorylation was detected. Human and fission yeast Wee1 also specifically phosphorylate synthetic peptides at sites equivalent to Tyr15. Mutation of a critical lysine codon (Lys114) believed to be essential for kinase activity abolished both the in vivo mitotic inhibitor function and in vitro kinase activities of human Wee1. These results conclusively prove that Wee1 kinases inhibit mitosis by directly phosphorylating p34cdc2 on Tyr15, and strongly indicate that human cells have independent kinase pathways directing the two inhibitor phosphorylations of p34cdc2.     

Fos-Jun heterodimers and Jun homodimers bend DNA in opposite orientations: implications for transcription factor cooperativity

Association of Fos and Jun with the AP-1 site results in a conformational change in the basic amino acid regions that constitute the DNA-binding domain. We show that Fos and Jun induce a corresponding alteration in the conformation of the DNA helix. Circular permutation analysis indicated that both Fos-Jun heterodimers and Jun homodimers induce flexure at the AP-1 site. Phasing analysis demonstrated that Fos-Jun heterodimers and Jun homodimers induce DNA bends that are directed in opposite orientations. Fos-Jun heterodimers bend DNA toward the major groove, whereas Jun homodimers bend DNA toward the minor groove. Fos and Jun peptides encompassing the dimerization and DNA-binding domains bend DNA in the same orientations as the full-length proteins. However, additional regions of both proteins influence the magnitude of the DNA bend angle. Thus, despite the amino acid sequence similarity in the basic region Fos-Jun heterodimers and Jun homodimers form topologically distinct DNA-protein complexes.     

Phosphorylation of p97(VCP) and p47 in vitro by p34cdc2 kinase.

The hexameric ATPase p97/yeast Cdc48p has been implicated in a number of cellular events that are regulated during mitosis, including homotypic membrane fusion, spindle pole body function, and ubiquitin-dependent protein degradation. p97/Cdc48p contains two conserved consensus p34cdc2 kinase phosphorylation sites within its second ATP binding domain. This domain is likely to play a role in stabilising the hexameric form of the protein. We therefore investigated whether p97 could be phosphorylated by p34cdc2 kinase in vitro, and whether phosphorylation might influence the oligomeric status of p97. Monomeric, but not hexameric, p97 was phosphorylated by p34cdc2 kinase, as was the p97-associated protein p47. However, phosphorylation by p34cdc2 kinase did not impair subsequent re-hexamerisation of p97, implying that the phosphorylated residue(s) are not critical for interaction between p97 monomers. Moreover, p97 within both interphase and mitotic cytosols was almost exclusively hexameric, suggesting that the activity of p97 is not regulated during mitosis by influencing the extent of oligomerisation.     

Mitogen-activated protein kinases phosphorylate nuclear lamins and display sequence specificity overlapping that of mitotic protein kinase p34cdc2.

Members of the mitogen-activated protein (MAP) kinase family are implicated in mediating entry of cells into the cell cycle, as well as passage through meiotic M phase. These kinases have attracted much interest because their activation involves phosphorylation on both tyrosine and threonine residues, but little is known about their physiological targets. In this study, two distinct members of the MAP kinase family (p44mpk and p42mapk) are shown to phosphorylate chicken lamin B2 at a single site identified as Ser16. Moreover, these MAP kinases cause depolymerization of in-vitro-assembled longitudinal lamin head-to-tail polymers. Ser16 was previously shown to be phosphorylated during mitosis in vivo, and to be a target of the mitotic protein kinase p34cdc2 in vitro. Accordingly, lamins were proposed to be direct in vivo substrates of p34cdc2. This proposal is supported by quantitative analyses indicating that lamin B2, when assayed in vitro, is a substantially better substrate for p34cdc2 than for MAP kinases. Nevertheless, a physiological role of MAP kinases in lamin phosphorylation is not excluded. The observation that members of the MAP kinase family display sequence specificities overlapping that of p34cdc2 raises the possibility that some of the purported substrates of p34cdc2 may actually be physiological substrates of MAP kinases.     

Phosphorylation of HMG-I by protein kinase C attenuates its binding affinity to the promoter regions of protein kinase C gamma and neurogranin/RC3 genes

A 20-kDa DNA-binding protein that binds the AT-rich sequences within the promoters of the brain-specific protein kinase C (PKC) gamma and neurogranin/RC3 genes has been characterized as chromosomal nonhistone high-mobility-group protein (HMG)-I. This protein is a substrate of PKC alpha, beta, gamma, and delta but is poorly phosphorylated by PKC epsilon and zeta. Two major (Ser44 and Ser64) and four minor phosphorylation sites have been identified. The extents of phosphorylation of Ser44 and Ser64 were 1:1, whereas those of the four minor sites all together were <30% of the major one. These PKC phosphorylation sites are distinct from those phosphorylated by cdc2 kinase, which phosphorylates Thr53 and Thr78. Phosphorylation of HMG-I by PKC resulted in a reduction of DNA-binding affinity by 28-fold as compared with 12-fold caused by the phosphorylation with cdc2 kinase. HMG-I could be additively phosphorylated by cdc2 kinase and PKC, and the resulting doubly phosphorylated protein exhibited a >100-fold reduction in binding affinity. The two cdc2 kinase phosphorylation sites of HMG-I are adjacent to the N terminus of two of the three predicted DNA-binding domains. In comparison, one of the major PKC phosphorylation sites, Ser64, is adjacent to the C terminus of the second DNA-binding domain, whereas Ser44 is located within the spanning region between the first and second DNA-binding domains. The current results suggest that phosphorylation of the mammalian HMG-I by PKC alone or in combination with cdc2 kinase provides an effective mechanism for the regulation of HMG-I function.     

The basic region of Fos mediates specific DNA binding.

The DNA-binding domains of the members of the Fos and Jun families of proteins consist of a basic region followed by a dimerizing segment with heptad repeats of leucine. Fos-Jun heterodimers and Jun alone, but not Fos alone, bind to the symmetrical sequences TGACTCA (AP-1 site) or TGACGTCA (cAMP response element or CRE). We set out to test the hypothesis that in the Fos-Jun heterodimer the basic region of Fos confers specific DNA-binding properties equivalent to the contribution of the basic region of Jun. Fos-Jun chimeric proteins were prepared consisting of the basic region of one protein joined to the leucine repeat of the other. Heterodimers with mixed Fos and Jun leucine repeat segments showed high affinity binding to the AP-1 site or CRE whether they contained two basic regions from Jun, two basic regions from Fos, or one from each source. Heterodimers with two Fos basic regions showed somewhat greater affinity for the CRE and AP-1 site than the heterodimer with two Jun basic regions. The DNA sequence specificity and the purine and phosphate DNA contact sites for each heterodimer were similar. We conclude that in the Fos-Jun heterodimer the basic region of Fos contributes specific DNA-binding properties equivalent to those of Jun. Our results support a model in which the Fos and Jun basic regions of the Fos-Jun heterodimer each interact with symmetrical DNA half sites.     

Two different protein kinases act on a different time schedule as glial filament kinases during mitosis.

Glial fibrillary acidic protein (GFAP) is a component of glial filaments specific to astroglia. We now report the spatial and temporal distributions of four phosphorylated sites in the GFAP molecule during mitosis of astroglial cells, determined by antibodies which can distinguish phosphorylated epitopes from non-phosphorylated-epitopes. Immunofluorescence microscopy showed that the Ser8 residues in the entire cytoplasmic glial filament system are initially phosphorylated when the cells enter mitosis. In cytokinesis, the phosphoSer8 residues become dephosphorylated, whereas Thr7, Ser13 and Ser34 in glial filaments at the cleavage furrow become the preferred sites of phosphorylation. The cdc2 kinase purified from mitotic cells can phosphorylate GFAP at Ser8 but not at Thr7, Ser13 or Ser34, in vitro. These results suggest that cdc2 kinase acts as a glial filament kinase only at the G2-M phase transition while other glial filament kinases are probably activated at the cleavage furrow before final separation of the daughter cells.     

Intermediate filament reorganization during mitosis is mediated by p34cdc2 phosphorylation of vimentin

As cells enter mitosis, the intermediate filament (IF) networks of interphase BHK-21 cells are depolymerized to form cytoplasmic aggregates of disassembled IFs, and the constituent IF proteins, vimentin and desmin are hyperphosphorylated at several specific sites. We have characterized one of two endogenous vimentin kinases from a particulate fraction of mitotic cell lysates. Through several purification steps, vimentin kinase activity copurifies with histone H1 kinase and both activities bind to p13suc1-Sepharose. The final enriched kinase preparation consists primarily of p34cdc2 and polypeptides of 65 and 110 kd. The purified kinase complex phosphorylates vimentin in vitro at a subset of sites phosphorylated in vivo during mitosis. Furthermore, phosphorylation of in vitro polymerized vimentin IFs by the purified kinase causes their disassembly. Therefore, vimentin is a substrate of p34cdc2 and phosphorylation of vimentin contributes to M phase reorganization of the IF network.     

cdc2 family kinases phosphorylate a human cell DNA replication factor, RPA, and activate DNA replication.

RPA is a single-stranded DNA binding protein complex purified from human cells and is essential for the initiation and elongation stages of SV40 DNA replication in vitro. In both human and yeast cells, the 34 kDa polypeptide subunit of RPA is phosphorylated in the S and G2 phases of the cell cycle and not in G1. One of the major RPA kinases present in extracts of human cells was purified and shown to be the cyclin B-cdc2 complex. This purified kinase, and a closely related cyclin A associated cdc2-like kinase, phosphorylated RPA p34 on a subset of the chymotryptic peptides that were phosphorylated in vivo at the G1-S transition. Two serines near the N-terminus of RPA p34 were identified as possible sites of phosphorylation by cdc2 kinase. These same serines were necessary for RPA phosphorylation in vivo. The purified cdc2 kinase stimulated SV40 DNA replication in vitro when added to G1 cell extracts. The kinase also stimulated unwinding at the origin of replication, one of the earliest steps in DNA replication requiring RPA, but only in the presence of an additional factor present in G1 cell extracts. Thus, one or more members of the cyclin-cdc2 kinase family may be required for the initiation and maintenance of S phase, in part due to their ability to phosphorylate and activate a cellular DNA replication factor, RPA.     

Elimination of cdc2 phosphorylation sites in the cdc25 phosphatase blocks initiation of M-phase.

The cdc25 phosphatase is a mitotic inducer that activates p34cdc2 at the G2/M transition by dephosphorylation of Tyr15 in p34cdc2. cdc25 itself is also regulated through periodic changes in its phosphorylation state. To elucidate the mechanism for induction of mitosis, phosphorylation of cdc25 has been investigated using recombinant proteins. cdc25 is phosphorylated by both cyclin A/p34cdc2 and cyclin B/p34cdc2 at similar sets of multiple sites in vitro. This phosphorylation retards its electrophoretical mobility and activates its ability to increase cyclin B/p34cdc2 kinase activity three- to fourfold in vitro, as found for endogenous Xenopus cdc25 in M-phase extracts. The threonine and serine residues followed by proline that are conserved between Xenopus and human cdc25 have been mutated. Both the triple mutation of Thr48, Thr67, and Thr138 and the quintuple mutation of these three threonine residues plus Ser205 and Ser285, almost completely abolish the shift in electrophoretic mobility of cdc25 after incubation with M-phase extracts or phosphorylation by p34cdc2. These mutations inhibit the activation of cdc25 by phosphorylation with p34cdc2 by 70 and 90%, respectively. At physiological concentrations these mutants cannot activate cyclin B/p34cdc2 in cdc25-immunodepleted oocyte extracts, suggesting that a positive feed-back loop between cdc2 and cdc25 is necessary for the full activation of cyclin B/p34cdc2 that induces abrupt entry into mitosis in vivo.     

The ionizing radiation-induced replication protein A phosphorylation response differs between ataxia telangiectasia and normal human cells.

Replication protein A (RPA), the trimeric single-stranded DNA-binding protein complex of eukaryotic cells, is important to DNA replication and repair. Phosphorylation of the p34 subunit of RPA is modulated by the cell cycle, occurring during S and G2 but not during G1. The function of phosphorylated p34 remains unknown. We show that RPA p34 phosphorylation is significantly induced by ionizing radiation. The phosphorylated form, p36, is similar if not identical to the phosphorylated S/G2 form. gamma-Irradiation-induced phosphorylation occurs without new protein synthesis and in cells in G1. Mutation of cdc2-type protein kinase phosphorylation sites in p34 eliminates the ionizing radiation response. The gamma-irradiation-induced phosphorylation of RPA p34 is delayed in cells from ataxia telangiectasia, a human inherited disease conferring DNA repair defects and early-onset tumorigenesis. UV-induced phosphorylation of RPA p34 occurs less rapidly than gamma-irradiation-induced phosphorylation but is kinetically similar between ataxia telangiectasia and normal cells. This is the first time that modification of a repair protein, RPA, has been linked with a DNA damage response and suggests that phosphorylation may play a role in regulating DNA repair pathways.     

Calcium/calmodulin-dependent phosphorylation and activation of human Cdc25-C at the G2/M phase transition in HeLa cells

The human tyrosine phosphatase (p54(cdc25-c)) is activated by phosphorylation at mitosis entry. The phosphorylated p54(cdc25-c) in turn activates the p34-cyclin B protein kinase and triggers mitosis. Although the active p34-cyclin B protein kinase can itself phosphorylate and activate p54(cdc25-c), we have investigated the possibility that other kinases may initially trigger the phosphorylation and activation of p54(cdc25-c). We have examined the effects of the calcium/calmodulin-dependent protein kinase (CaM kinase II) on p54(cdc25-c). Our in vitro experiments show that CaM kinase II can phosphorylate p54(cdc25-c) and increase its phosphatase activity by 2.5-3-fold. Treatment of a synchronous population of HeLa cells with KN-93 (a water-soluble inhibitor of CaM kinase II) or the microinjection of AC3-I (a specific peptide inhibitor of CaM kinase II) results in a cell cycle block in G2 phase. In the KN-93-arrested cells, p54(cdc25-c) is not phosphorylated, p34(cdc2) remains tyrosine phosphorylated, and there is no increase in histone H1 kinase activity. Our data suggest that a calcium-calmodulin-dependent step may be involved in the initial activation of p54(cdc25-c).     

Identification of phosphorylation sites in the repetitive carboxyl-terminal domain of the mouse RNA polymerase II largest subunit

The largest subunit of eukaryotic RNA polymerase II contains a carboxyl-terminal domain (CTD) which is comprised of repetitive heptapeptides with a consensus sequence Tyr1-Ser2-Pro3-Thr4-Ser5-Pro6-Ser7. We demonstrate here that the mouse CTD expressed in and purified from Escherichia coli can be phosphorylated in vitro by a p34cdc2/CDC28-containing CTD kinase from mouse ascites tumor cells. The product of this reaction, a phosphorylated form of the CTD, contains phosphoserine and phosphothreonine, but not phosphotyrosine. The same phosphoamino acid content is observed in the in vivo phosphorylated CTD from a mouse cell line. Synthetic peptides with naturally occurring non-consensus heptapeptide sequences can also be phosphorylated by CTD kinase in vitro. Phosphoamino acid analysis of these non-consensus heptapeptides together with direct sequencing of a phosphorylated heptapeptide reveals that serines (or threonines) at positions two and five are the sites phosphorylated by mouse CTD kinase. Thus, the -Ser(Thr)-Pro- motif common to p34cdc2/CDC28-containing protein kinases is the recognition site for mouse CTD kinase.     

Phosphorylation of neurofilament H subunit at the tail domain by CDC2 kinase dissociates the association to microtubules

We sought the mammalian neurofilament tail domain-specific kinase. Several well known kinases including cAMP-dependent protein kinase, protein kinase C, Ca(2+)-calmodulin-dependent protein kinase II, casein kinase I, and casein kinase II phosphorylated the high (NF-H) and middle molecular mass subunit (NF-M) of bovine neurofilaments, but they did not reduced the electrophoretic mobility of the dephosphorylated form of NF-M and NF-H by phosphorylation nor was the amount of phosphorylation increased by dephosphorylation of NF proteins, indicating that the phosphorylation sites by these kinases are not major in vivo phosphorylation sites at the tail domain. In contrast, cdc2 kinase phosphorylated specifically the dephosphorylated form of NF-H. 4 mol of phosphates were incorporated per mol of NF-H and this phosphorylation returned the electrophoretic mobility of the dephosphorylated form of NF-H to the position of the isolated, fully phosphorylated form of NF-H. Furthermore, the phosphorylation by cdc2 kinase dissociated the binding of dephosphorylated NF-H to microtubules. Phosphorylation sites were located at the carboxyl-terminal tail domain. The KSPXK motif, but not KSPXX, in the repetitive sequence was suggested to be the phosphorylation site by using synthetic peptides.