Eukaryotic Cell-Cycle Control Paul Nurse ICRF Cell-Cycle Control Group, Department of Biochemistry, University of Oxford, South Parks Road, Oxford 0x1 3QU, U.K.
P)
D
r
The subject of my lecture is how the eukaryotic cell cycle is controlled. This is a topic which has seen real advances in the last few years, based to a large extent on the extensive genetic analysis of the yeast cell cycle developed over the previous decade. In this lecture I shall focus on the contribution made by my own laboratory working with the fission yeast Schizosaccharomyces pombe. This simple unicellular eukaryote has only 4-5 times the DNA content of Escherichia coli, but despite this simplicity its cell cycle exhibits most of the important features found in more complicated multicellular organisms. Fission yeast is also very convenient for both classical and molecular genetic analysis, advantages which have led to a powerful strategy for investigating the cell cycle. First, mutants, altered in cell-cycle behaviour, are isolated which either are unable to proceed through the cell cycle or are altered in its regulation. These mutants define gene functions required for proper progression through the cell cycle and for its control. Secondly, these genes are cloned from a gene library by selecting for a gene that complements the mutant function, exploiting the high frequency of transformation possible with fission yeast [l]. Thirdly, these cloned genes are investigated biochemically. They are sequenced, expressed in E. coli, the gene product purified and antibodies raised to provide tools which can be used to investigate the roles of the molecules during the cell cycle. In this way the study can proceed, initially from an abstract description of the problem based on identification of the genes involved, to a concrete molecular description of the components and processes making up the cell cycle. The strength of this approach is that nothing need be known about the nature of the molecules involved at the beginning of the work, but this information will gradually emerge as the study proceeds. Two major events are common to all eukaryotic cell cycles, S-phase when the chromosomes are
A
Delivered on I 8 July I99 I at the University of Manchester
r)
t
PROFESSOR P. N U R S E
replicated and M-phase when the replicated chromosomes are segregated to the two daughter cells. Both of these events must normally occur during each cell cycle to ensure that the two daughter cells receive a full complement of chromosomes. For this reason the central problems in cellcycle control are understanding what regulates the onset of S-phase and M-phase, and what co-ordi-
Abbreviations used: cdc, cell division cycle mutants; cdc2.5'".temperature-sensitive cdc2.5 gene; MPF, maturation promoting factor.
239
I992
Biochemical Society Transactions
240
nates these two events such that they always occur in the correct order. Investigation of these controls in fission yeast has been greatly helped by the isolation of two types of cell-cycle mutant. The first type are called cdc (for cell division cycle) mutants which are unable to complete an essential step in the cell cycle and as a consequence do not divide [2, 31. They are not defective in cellular growth and so the non-dividing, growing cells become much enlarged. Hecause the cdc genes defined by these mutants are essential for cell viability, the mutants must be conditional, that is the mutant phenotype is only expressed in a particular restrictive condition. llsually temperature-sensitive mutants are isolated which fail to complete the cell cycle at high temperature. Over 60 cdc-like genes have now been identified encoding functions required for essential steps in the cell cycle. The second type of mutant is not arrested in cell-cycle progress, but is advanced prematurely through the cell cycle to undergo cell division at a reduced cell size [4, 51. This is because cells usually divide at a constant size and if the cell cycle is shortened they d o not have sufficient time to reach this size, and thus divide smaller. These wee mutants define genes which encode functions acting as major rate-limiting steps for progression through the cell cycle. They are much rarer than the cdc mutants and only four genes have been identified which can influence dramatically the advancement of cells into division. These four genes make up a regulatory network which determines the duration of G2 and thus the cell-cycle timing of M-phase [6]. Study of both the cdc and wee mutants have revealed two controls, one acting in G1 regulating the onset of S-phase, and the second acting in G2 regulating the onset of M-phase [7]. The latter control consists of the regulatory network referred to above with the four gene functions acting together to determine when the cell undergoes M-phase. This is called the mitotic control. The other control is called start and is the point where cells become committed to the cell cycle. Mutants in two cdc genes arrest at start before the programme of events leading to S-phase have begun. As a consequence these cells are not yet committed to the cell cycle and are able to undergo alternative developmental pathways such as conjugation, if challenged to do so. One gene, cdc2', is required at both the start and mitotic controls and as a consequence is of special interest for cell-cycle control. This gene will be the major subject of my lecture. Genetic epistasis studies have demonstrated that the mitotic regulatory network consists of two
Volume 20
pathways, one inhibitory and the other activatory, which together regulate the cdc2+ gene product. The activatory pathway consists of cdc2.5' encoding an 80 kDa protein phosphatase required to activate the 34 kDa protein kinase encoded by cdc2+ [XI. The protein level of this p34""' kinase is unchanged throughout the cell cycle but its kinase activity rises to a peak during M-phase and is low during the other phases of the cycle 191. This increase in kinase activity brings about M-phase by phosphorylating key substrates required for the various events of mitosis [lo]. Activation of the p34'd'' kinase involves the ~ 8 0 ' ~phosphatase '~' [ 11] which removes a phosphate from a phosphorylated tyrosine residue located 15 residues from the Nterminus of ~34'~'' [12]. This Tyr-15 is found within the ATP-binding site of the protein kinase. ' When Tyr- 15 is phosphorylated the ~ 3 4 ' ~ 'kinase activity is low, presumably because of interference with ATP binding or utilization. If Tyr-15 is replaced by a phenylalanine which cannot be phosphorylated, then ~ 3 4 ' ~ 'is' not inhibited and cells enter M-phase very prematurely at a reduced size. This indicates that Tyr-15 phosphorylation is the major rate-limiting step determining the cellcycle timing of M-phase. In a temperature sensitive cdc2.5 (cdc2.5") mutant incubated at its restrictive temperature, Tyr- 15 phosphorylation is high, but this phosphorylation is rapidly reduced on shifting the mutant to its permissive temperature conditions, leading to the return of p8W"" phosphatase activity to the cell. In contrast overproducing p8W"" results in advancement of the cell into M-phase. ~ 3 4 ' ~ ' ' isolated from cdc2.5" cells held at the restrictive temperature has low kinase activity but this can be increased by treatment in vitro with purified tyrosine phosphatase. These studies indicate that the onset of M-phase is brought about by the simple biochemical switch of Tyr- 15 dephosphorylation which is controlled by the p80'dr2' phosphatase. pS4rdcLkinase activation also requires a gene product encoded by another cell-cycle gene cdcl3' [2]. This is a H-cyclin [ 131, similar in sequence to the cyclins identified by studies of the early cell cycles of marine invertebrates. A cdcl3" mutant held at high restrictive temperature arrests in G2 before M-phase, while incubation at a slightly lower temperature causes arrest in M-phase itself with condensed chromosomes. Thus cdcl3 encoding p5hrdr' appears to function at the boundary between G2 and M-phase. Experiments indicate that p56"'" is required for activation of the ~ 3 4 ' ~ ' ' kinase but there is no evidence that it is rate limiting for activation [9]. ~ 5 6 ' ~ ' 'is destroyed as cells com+
Eukaryotic Cell-Cycle Control
plete M-phase and is likely to be responsible for the inactivation of the ~ 3 4 ' ~ 'kinase ' which occurs at this time. T h e ~ 5 6 ' ~ ' "level then builds up again as cells proceed through the cell cycle, with p5h'dr" forming a stable complex with ~ 3 4 ' ~ 'Sufficient ~. ~ 5 6 ' ~ ' "is accumulated early in the cell cycle to activate pWd", and so p56""" accumulation is not rate limiting for this process. As a consequence over-producing ~56'~'" does not advance cells prematurely into M-phase. I Iaving dealt with the biochemical switch bringing about M-phase I now want to consider the problem of how M-phase is coupled to the previous S-phase and S-phase to the previous M-phase, the controls which ensure that these two events always occur in the correct order. If S-phase is blocked in fission yeast then the subsequent M-phase is also blocked. This coupling works through activation of the p34""' protein kinase. Mutants which are altered in Tyr- 15 phosphorylation such as cdc2F15 or cells do not block containing high levels of pXO"'", M-phase when S-phase is incomplete [ 141. This suggests that complete replication of the chromosomes sends a signal which prevents Tyr-15 dephosphorylation, thus blocking onset of M-phase and coupling it to S-phase. Hut if the cell does not require Tyr- 15 dephosphorylation for M-phase, or if it is overloaded with the phosphatase then this control does not work. As a consequence the cell is unable to detect that 1)NA replication is incomplete and is not blocked from entering M-phase. Such a situation is lethal for the cell because the nucleus is unable to divide correctly. Interestingly cells arrested before start do not try to enter M-phase in these circumstances, suggesting that commitment to the cell cycle is required before such a premature M-phase can take place. The other major coupling in the cell cycle is the dependency of S-phase upon the previous M-phase, which is necessary to maintain the ploidy level of the cell constant. This coupling control also appears to involve ~ 3 4 ' ~ ' ' .Mutants have been isolated which undergo an extra round of S-phase in the absence of M-phase, leading to an increase in ploidy 151. These mutations are thermosensitive cdc2" alleles which on incubation at a high temperature result in degradation of most of the ~34'~'' present in the cell. When the mutants are returned to the permissive temperature, those cells originally in G2 do not proceed to M-phase but undergo S-phase instead. This suggests that the heat treatment destroys most of mitotic control form present ' after heat in the G2 cells and the ~ 3 4 ' ~ 'generated treatment is in the form required for the start
control. If this proposal is correct then it implies that ~ 3 4 ' ~ ' 'provides a memory for the cell of whether it is in G1 or in G2, and this memory is lost when ~ 3 4 ' ~is' ' destroyed. The above experiments indicate that ~ 3 4 ' ~ ' ' has a very central role in the coupling between S-phase and M-phase. Mutants of cdc2 can enter M-phase if S-phase is not complete, and in certain circumstances other cdc2 mutants can enter S-phase when M-phase is not complete. An explanation for this behaviour is that completion of S-phase converts the start form into the mitotic control form and completion of M-phase converts the mitotic control form into the start form, but both of these requirements can be bypassed by the appropriate cdc2 mutants. With this view, the cell cycle can be considered as a p34'd'' cycle, and it is the latter which underlies the basic cyclic nature of the cell cycle. The central role of ~34""' for controlling the cell-cycle control in fission yeast has been shown to apply also to other eukaryotes. Two experimental approaches have been important for this demonstration. In the first, human homologues of cdc2+ have been isolated by transforming a human cDNA library into a cdc" mutant and selecting for cDNAs that can complement the defective cdc2" function, allowing growth at the restrictive tempenture [ 161. This complementation approach identifies genes which are functionally equivalent to the mutated gene rather than genes which are simply structurally similar. The human CDC2 gene is remarkably similar to the fission yeast cdc2' gene despite 1000 million years of evolutionary divergence. They both encode 34 kI)a proteins which are 63Yo identical in sequence. The fact that both control points involving cdc2' can be perfectly rescued by the human CDC2 gene suggests that very similar controls must be present in vertebrate cells. This has been confirmed for the mitotic control by the second experimental approach linking the yeast work to other eukaryotes. In Xenopus frogs the oocyte is arrested in G2 before meiotic M-phase, and maturation into an egg requires release from this arrest, and induction of M-phase. Maturation promoting factor or MPF has been purified which induces oocyte entry into M-phase. Purified MPF contains both ~ 3 4 ' ~ '['171 and H-cyclin, showing that this biochemical approach has identified the same proteins as the genetical approach used in fission yeast. The molecular mechanism by which ~ 3 4 ' ~ 'is' regulated at onset of M-phase is remarkably similar in vertebrate cells and fission yeast [ 181. In mouse
I992
24 I
Biochemical Society Transactions
242
cells the same Tyr- 15 is phosphorylated during G2, although the adjacent threonine, Thr-14, is also phosphorylated. Non-phosphorylatable mutants in these two residues have been constructed in the human gene and the two single and the double mutants have been expressed in Xenopus extracts. These extracts proceed in vitro towards M-phase and the endogenous ~ 3 4 ‘ ~ ‘kinase ’ is activated after a certain delay. This occurs at the same time as nuclei in the extracts begin to exhibit M-phase features. The wild type and two single mutant human genes are activated on schedule with the endogenous kinase, but the double mutant is activated prematurely to a higher level of activity. This indicates that phosphorylation of either of the Thr14 and Tyr-15 residues in the ATP-binding site restrains kinase activation during the G2 of vertebrate cells. The same double mutant also becomes activated even if DNA replication is blocked, which is not the case for the wild type or the two single mutants. Therefore in these circumstances cyclin levels are not limiting for the timing of ~ 3 4 ‘ kinase ~ ‘ ~ activation, and as a consequence M-phase onset is determined by dephosphorylation of these two residues in the ATP-binding site. This control is obviously very similar to that described in fission yeast, but has the additional restraining phosphorylation of Thr- 14, presumably providing extra control. It appears then that p34‘d‘’ plays an important role in controlling the cell cycle in all eukaryotic cells [ lO]. This is now quite clear for the onset of M-phase, where even details of the molecular mechanism of regulation are very similar. However, the control acting over onset of S-phase is not yet fully understood either in yeast or in vertebrate cells, and it is possible that there are differences in hou this control is regulated in different eukaryotes, although it is likely to involve ~34‘~‘’or a closely related homologue. It is this area of cell-cycle control which requires more work at the present time,
Volume 20
but given the rapid advances which continue to be made, the molecular basis of the start control and the role played by ~ 3 4 ‘ ~ ‘ ’should soon be unravelled. 1. Beach, I). & Nurse, 1’. (19x1) Nature (1,ondon) 200, 140- 142 2. Nurse, I’., Thuriaux, 1’. & Nasmyth, K. (1076) Mol. Gen. Genet. 146, 167- 17X 3. Nasmyth, K. & Nurse, 1’. (1981) Mol. (;en. Genet. 182, 119-124 4. Nurse. 1’. (1975) Nature (London) 256, 547-551 5. Nurse, 1’. & Thuriaux, 1’. (19x0) Genetics 96. 627-637 h. Kussell, 1’. & Nurse, 1’. (1987) Cell (Cambridge, 7. Nurse, 1’. & Hissett, Y. (10x1) Nature (London) 292, 5 5x- 5 00 8. Russell, 1’. & Nurse, 1’. (10x6) Cell (Cambridge, Mass.) 45, 145-153 9. Moreno, S., Hayles, J. & Nurse. 1’. (19x9) Cell (Cambridge, Mass.) 58, 3hl-372 10. Moreno, S. & Nurse. 1’. (1000) Cell (Cambridge, Mass.) 61, 540-551 11. Gould. K.. Moreno, S.. Tonks, N. & Nurse, 1’. (1000) Science 250, 1573-1 576 12. Gould. K. I,. & Nurse, 1’. (19x0) Nature (Imndon) 342, 30-4s 13. Hagan, I., Hayles, J. & Nurse, 1’. ( I OXX) J. Cell Sci. 91, 587-595 14. Enoch, T. & Nurse. 1’. (1900) Cell (Cambridge. M R S S . )60, 665-673 IS. Hroek. I).. Hartlett, K.. Crawford, K. Pr Nurse. 1’. (1001) Nature (Idondon)349, 388-303 16. Lee, M. G. Pr Nurse, 1’. (1087) Nature (London) 327. 31-35 17. Gautier. J., Norbury, C.. Imhka, M.. Nurse. 1’. & Mallet-, J. (19x8) Cell (Carnbridgc, Mass.) 54, 43 3 -4 39 18. Norbury, C., Hlow, J. & Nurse, 1’. (1991) EMHO J. 10, 3 32 1-3 320 10. Nurse, 1’. (1090) Nature (1,ondon) 344, 503-508 Keceived 10 Llecernber 190 1
P)
D
r
The subject of my lecture is how the eukaryotic cell cycle is controlled. This is a topic which has seen real advances in the last few years, based to a large extent on the extensive genetic analysis of the yeast cell cycle developed over the previous decade. In this lecture I shall focus on the contribution made by my own laboratory working with the fission yeast Schizosaccharomyces pombe. This simple unicellular eukaryote has only 4-5 times the DNA content of Escherichia coli, but despite this simplicity its cell cycle exhibits most of the important features found in more complicated multicellular organisms. Fission yeast is also very convenient for both classical and molecular genetic analysis, advantages which have led to a powerful strategy for investigating the cell cycle. First, mutants, altered in cell-cycle behaviour, are isolated which either are unable to proceed through the cell cycle or are altered in its regulation. These mutants define gene functions required for proper progression through the cell cycle and for its control. Secondly, these genes are cloned from a gene library by selecting for a gene that complements the mutant function, exploiting the high frequency of transformation possible with fission yeast [l]. Thirdly, these cloned genes are investigated biochemically. They are sequenced, expressed in E. coli, the gene product purified and antibodies raised to provide tools which can be used to investigate the roles of the molecules during the cell cycle. In this way the study can proceed, initially from an abstract description of the problem based on identification of the genes involved, to a concrete molecular description of the components and processes making up the cell cycle. The strength of this approach is that nothing need be known about the nature of the molecules involved at the beginning of the work, but this information will gradually emerge as the study proceeds. Two major events are common to all eukaryotic cell cycles, S-phase when the chromosomes are
A
Delivered on I 8 July I99 I at the University of Manchester
r)
t
PROFESSOR P. N U R S E
replicated and M-phase when the replicated chromosomes are segregated to the two daughter cells. Both of these events must normally occur during each cell cycle to ensure that the two daughter cells receive a full complement of chromosomes. For this reason the central problems in cellcycle control are understanding what regulates the onset of S-phase and M-phase, and what co-ordi-
Abbreviations used: cdc, cell division cycle mutants; cdc2.5'".temperature-sensitive cdc2.5 gene; MPF, maturation promoting factor.
239
I992
Biochemical Society Transactions
240
nates these two events such that they always occur in the correct order. Investigation of these controls in fission yeast has been greatly helped by the isolation of two types of cell-cycle mutant. The first type are called cdc (for cell division cycle) mutants which are unable to complete an essential step in the cell cycle and as a consequence do not divide [2, 31. They are not defective in cellular growth and so the non-dividing, growing cells become much enlarged. Hecause the cdc genes defined by these mutants are essential for cell viability, the mutants must be conditional, that is the mutant phenotype is only expressed in a particular restrictive condition. llsually temperature-sensitive mutants are isolated which fail to complete the cell cycle at high temperature. Over 60 cdc-like genes have now been identified encoding functions required for essential steps in the cell cycle. The second type of mutant is not arrested in cell-cycle progress, but is advanced prematurely through the cell cycle to undergo cell division at a reduced cell size [4, 51. This is because cells usually divide at a constant size and if the cell cycle is shortened they d o not have sufficient time to reach this size, and thus divide smaller. These wee mutants define genes which encode functions acting as major rate-limiting steps for progression through the cell cycle. They are much rarer than the cdc mutants and only four genes have been identified which can influence dramatically the advancement of cells into division. These four genes make up a regulatory network which determines the duration of G2 and thus the cell-cycle timing of M-phase [6]. Study of both the cdc and wee mutants have revealed two controls, one acting in G1 regulating the onset of S-phase, and the second acting in G2 regulating the onset of M-phase [7]. The latter control consists of the regulatory network referred to above with the four gene functions acting together to determine when the cell undergoes M-phase. This is called the mitotic control. The other control is called start and is the point where cells become committed to the cell cycle. Mutants in two cdc genes arrest at start before the programme of events leading to S-phase have begun. As a consequence these cells are not yet committed to the cell cycle and are able to undergo alternative developmental pathways such as conjugation, if challenged to do so. One gene, cdc2', is required at both the start and mitotic controls and as a consequence is of special interest for cell-cycle control. This gene will be the major subject of my lecture. Genetic epistasis studies have demonstrated that the mitotic regulatory network consists of two
Volume 20
pathways, one inhibitory and the other activatory, which together regulate the cdc2+ gene product. The activatory pathway consists of cdc2.5' encoding an 80 kDa protein phosphatase required to activate the 34 kDa protein kinase encoded by cdc2+ [XI. The protein level of this p34""' kinase is unchanged throughout the cell cycle but its kinase activity rises to a peak during M-phase and is low during the other phases of the cycle 191. This increase in kinase activity brings about M-phase by phosphorylating key substrates required for the various events of mitosis [lo]. Activation of the p34'd'' kinase involves the ~ 8 0 ' ~phosphatase '~' [ 11] which removes a phosphate from a phosphorylated tyrosine residue located 15 residues from the Nterminus of ~34'~'' [12]. This Tyr-15 is found within the ATP-binding site of the protein kinase. ' When Tyr- 15 is phosphorylated the ~ 3 4 ' ~ 'kinase activity is low, presumably because of interference with ATP binding or utilization. If Tyr-15 is replaced by a phenylalanine which cannot be phosphorylated, then ~ 3 4 ' ~ 'is' not inhibited and cells enter M-phase very prematurely at a reduced size. This indicates that Tyr-15 phosphorylation is the major rate-limiting step determining the cellcycle timing of M-phase. In a temperature sensitive cdc2.5 (cdc2.5") mutant incubated at its restrictive temperature, Tyr- 15 phosphorylation is high, but this phosphorylation is rapidly reduced on shifting the mutant to its permissive temperature conditions, leading to the return of p8W"" phosphatase activity to the cell. In contrast overproducing p8W"" results in advancement of the cell into M-phase. ~ 3 4 ' ~ ' ' isolated from cdc2.5" cells held at the restrictive temperature has low kinase activity but this can be increased by treatment in vitro with purified tyrosine phosphatase. These studies indicate that the onset of M-phase is brought about by the simple biochemical switch of Tyr- 15 dephosphorylation which is controlled by the p80'dr2' phosphatase. pS4rdcLkinase activation also requires a gene product encoded by another cell-cycle gene cdcl3' [2]. This is a H-cyclin [ 131, similar in sequence to the cyclins identified by studies of the early cell cycles of marine invertebrates. A cdcl3" mutant held at high restrictive temperature arrests in G2 before M-phase, while incubation at a slightly lower temperature causes arrest in M-phase itself with condensed chromosomes. Thus cdcl3 encoding p5hrdr' appears to function at the boundary between G2 and M-phase. Experiments indicate that p56"'" is required for activation of the ~ 3 4 ' ~ ' ' kinase but there is no evidence that it is rate limiting for activation [9]. ~ 5 6 ' ~ ' 'is destroyed as cells com+
Eukaryotic Cell-Cycle Control
plete M-phase and is likely to be responsible for the inactivation of the ~ 3 4 ' ~ 'kinase ' which occurs at this time. T h e ~ 5 6 ' ~ ' "level then builds up again as cells proceed through the cell cycle, with p5h'dr" forming a stable complex with ~ 3 4 ' ~ 'Sufficient ~. ~ 5 6 ' ~ ' "is accumulated early in the cell cycle to activate pWd", and so p56""" accumulation is not rate limiting for this process. As a consequence over-producing ~56'~'" does not advance cells prematurely into M-phase. I Iaving dealt with the biochemical switch bringing about M-phase I now want to consider the problem of how M-phase is coupled to the previous S-phase and S-phase to the previous M-phase, the controls which ensure that these two events always occur in the correct order. If S-phase is blocked in fission yeast then the subsequent M-phase is also blocked. This coupling works through activation of the p34""' protein kinase. Mutants which are altered in Tyr- 15 phosphorylation such as cdc2F15 or cells do not block containing high levels of pXO"'", M-phase when S-phase is incomplete [ 141. This suggests that complete replication of the chromosomes sends a signal which prevents Tyr-15 dephosphorylation, thus blocking onset of M-phase and coupling it to S-phase. Hut if the cell does not require Tyr- 15 dephosphorylation for M-phase, or if it is overloaded with the phosphatase then this control does not work. As a consequence the cell is unable to detect that 1)NA replication is incomplete and is not blocked from entering M-phase. Such a situation is lethal for the cell because the nucleus is unable to divide correctly. Interestingly cells arrested before start do not try to enter M-phase in these circumstances, suggesting that commitment to the cell cycle is required before such a premature M-phase can take place. The other major coupling in the cell cycle is the dependency of S-phase upon the previous M-phase, which is necessary to maintain the ploidy level of the cell constant. This coupling control also appears to involve ~ 3 4 ' ~ ' ' .Mutants have been isolated which undergo an extra round of S-phase in the absence of M-phase, leading to an increase in ploidy 151. These mutations are thermosensitive cdc2" alleles which on incubation at a high temperature result in degradation of most of the ~34'~'' present in the cell. When the mutants are returned to the permissive temperature, those cells originally in G2 do not proceed to M-phase but undergo S-phase instead. This suggests that the heat treatment destroys most of mitotic control form present ' after heat in the G2 cells and the ~ 3 4 ' ~ 'generated treatment is in the form required for the start
control. If this proposal is correct then it implies that ~ 3 4 ' ~ ' 'provides a memory for the cell of whether it is in G1 or in G2, and this memory is lost when ~ 3 4 ' ~is' ' destroyed. The above experiments indicate that ~ 3 4 ' ~ ' ' has a very central role in the coupling between S-phase and M-phase. Mutants of cdc2 can enter M-phase if S-phase is not complete, and in certain circumstances other cdc2 mutants can enter S-phase when M-phase is not complete. An explanation for this behaviour is that completion of S-phase converts the start form into the mitotic control form and completion of M-phase converts the mitotic control form into the start form, but both of these requirements can be bypassed by the appropriate cdc2 mutants. With this view, the cell cycle can be considered as a p34'd'' cycle, and it is the latter which underlies the basic cyclic nature of the cell cycle. The central role of ~34""' for controlling the cell-cycle control in fission yeast has been shown to apply also to other eukaryotes. Two experimental approaches have been important for this demonstration. In the first, human homologues of cdc2+ have been isolated by transforming a human cDNA library into a cdc" mutant and selecting for cDNAs that can complement the defective cdc2" function, allowing growth at the restrictive tempenture [ 161. This complementation approach identifies genes which are functionally equivalent to the mutated gene rather than genes which are simply structurally similar. The human CDC2 gene is remarkably similar to the fission yeast cdc2' gene despite 1000 million years of evolutionary divergence. They both encode 34 kI)a proteins which are 63Yo identical in sequence. The fact that both control points involving cdc2' can be perfectly rescued by the human CDC2 gene suggests that very similar controls must be present in vertebrate cells. This has been confirmed for the mitotic control by the second experimental approach linking the yeast work to other eukaryotes. In Xenopus frogs the oocyte is arrested in G2 before meiotic M-phase, and maturation into an egg requires release from this arrest, and induction of M-phase. Maturation promoting factor or MPF has been purified which induces oocyte entry into M-phase. Purified MPF contains both ~ 3 4 ' ~ '['171 and H-cyclin, showing that this biochemical approach has identified the same proteins as the genetical approach used in fission yeast. The molecular mechanism by which ~ 3 4 ' ~ 'is' regulated at onset of M-phase is remarkably similar in vertebrate cells and fission yeast [ 181. In mouse
I992
24 I
Biochemical Society Transactions
242
cells the same Tyr- 15 is phosphorylated during G2, although the adjacent threonine, Thr-14, is also phosphorylated. Non-phosphorylatable mutants in these two residues have been constructed in the human gene and the two single and the double mutants have been expressed in Xenopus extracts. These extracts proceed in vitro towards M-phase and the endogenous ~ 3 4 ‘ ~ ‘kinase ’ is activated after a certain delay. This occurs at the same time as nuclei in the extracts begin to exhibit M-phase features. The wild type and two single mutant human genes are activated on schedule with the endogenous kinase, but the double mutant is activated prematurely to a higher level of activity. This indicates that phosphorylation of either of the Thr14 and Tyr-15 residues in the ATP-binding site restrains kinase activation during the G2 of vertebrate cells. The same double mutant also becomes activated even if DNA replication is blocked, which is not the case for the wild type or the two single mutants. Therefore in these circumstances cyclin levels are not limiting for the timing of ~ 3 4 ‘ kinase ~ ‘ ~ activation, and as a consequence M-phase onset is determined by dephosphorylation of these two residues in the ATP-binding site. This control is obviously very similar to that described in fission yeast, but has the additional restraining phosphorylation of Thr- 14, presumably providing extra control. It appears then that p34‘d‘’ plays an important role in controlling the cell cycle in all eukaryotic cells [ lO]. This is now quite clear for the onset of M-phase, where even details of the molecular mechanism of regulation are very similar. However, the control acting over onset of S-phase is not yet fully understood either in yeast or in vertebrate cells, and it is possible that there are differences in hou this control is regulated in different eukaryotes, although it is likely to involve ~34‘~‘’or a closely related homologue. It is this area of cell-cycle control which requires more work at the present time,
Volume 20
but given the rapid advances which continue to be made, the molecular basis of the start control and the role played by ~ 3 4 ‘ ~ ‘ ’should soon be unravelled. 1. Beach, I). & Nurse, 1’. (19x1) Nature (1,ondon) 200, 140- 142 2. Nurse, I’., Thuriaux, 1’. & Nasmyth, K. (1076) Mol. Gen. Genet. 146, 167- 17X 3. Nasmyth, K. & Nurse, 1’. (1981) Mol. (;en. Genet. 182, 119-124 4. Nurse. 1’. (1975) Nature (London) 256, 547-551 5. Nurse, 1’. & Thuriaux, 1’. (19x0) Genetics 96. 627-637 h. Kussell, 1’. & Nurse, 1’. (1987) Cell (Cambridge, 7. Nurse, 1’. & Hissett, Y. (10x1) Nature (London) 292, 5 5x- 5 00 8. Russell, 1’. & Nurse, 1’. (10x6) Cell (Cambridge, Mass.) 45, 145-153 9. Moreno, S., Hayles, J. & Nurse. 1’. (19x9) Cell (Cambridge, Mass.) 58, 3hl-372 10. Moreno, S. & Nurse. 1’. (1000) Cell (Cambridge, Mass.) 61, 540-551 11. Gould. K.. Moreno, S.. Tonks, N. & Nurse, 1’. (1000) Science 250, 1573-1 576 12. Gould. K. I,. & Nurse, 1’. (19x0) Nature (Imndon) 342, 30-4s 13. Hagan, I., Hayles, J. & Nurse, 1’. ( I OXX) J. Cell Sci. 91, 587-595 14. Enoch, T. & Nurse. 1’. (1900) Cell (Cambridge. M R S S . )60, 665-673 IS. Hroek. I).. Hartlett, K.. Crawford, K. Pr Nurse. 1’. (1001) Nature (Idondon)349, 388-303 16. Lee, M. G. Pr Nurse, 1’. (1087) Nature (London) 327. 31-35 17. Gautier. J., Norbury, C.. Imhka, M.. Nurse. 1’. & Mallet-, J. (19x8) Cell (Carnbridgc, Mass.) 54, 43 3 -4 39 18. Norbury, C., Hlow, J. & Nurse, 1’. (1991) EMHO J. 10, 3 32 1-3 320 10. Nurse, 1’. (1090) Nature (1,ondon) 344, 503-508 Keceived 10 Llecernber 190 1
