First, a little history. Although there were earlier isolated papers about pombe, the modern experimental work stems from two people. The first was Urs Leupold who took up its genetics and published a long paper in 1950(l).He had various reasons for choosing pombe but an important one was that it is normally haploid in vegetative growth. I came along somewhat later and chose pombe as a good material for cell physiology and in particular for the analysis of growth during the cell cycle. It was a fairly large micro-organism which was easy to culture. It grew only in length, so volume calculations were simple and cells could be roughly positioned in the cell cycle by their length. It divided in two by a medial septum and did not have the unusual method of cell cycle growth which separates budding yeast from nearly all other cells. I did not choose it because it was a eukaryote since the distinction between eukaryotes and prokaryotes was not drawn until the early 60s. Although there are now many excellent microphotographs of pombe, there is some historical interest in the diagram that I drew for my first substantial paper on pornbe(’) in 1957 when I thought that pombe was a relatively unknown cell (Fig. 1). It shows the cell shape, the septum, and the division scars left by earlier septation. In the paper, I misnamed the septum as a ‘cell plate’ - an error which survived for many years in the literature. I also misnamed the nucleus as a ‘central vacuole’ but the nuclear arrangements in yeasts were controversial at the time. Work on pombe has expanded greatly since these early papers and a mark of its ‘coming of age’ is the very
1
1Optn
Fig. 1. (Original legend). Schizosaccharomyces pombe in phase contrast. Central vacuoles, lipid granules and bumps due to division scars are shown. Reproduced by permission of Experimental Cell Research from ref. 2.
recent publication of a multi-author monograph titled Molecular Biology of The Fission and references for most of what I say in this article can be found there. Despite its fashionable title, it contains more than what is strictly molecular biology, namely chapters on genetics, cytology, morphogenesis and growth. Today, cell biologists are probably most interested in the contributions made with pombe on the control of mitosis and DNA synthesis. This topic started when Paul Nurse learnt pombe genetics from Leupold in the early ’70s and brought it to Edinburgh as a tool for studying the cell cycle. From this came the isolation of cdc mutants which are blocked in their passage through the cell cycle at a high (restrictive) temperature and also wee mutants which are altered in size at division. In the case of the cdc mutants, we were following the pioneer footsteps of Lee Hartwell who first isolated and named these mutants in budding yeast. In the late ’ ~ O S ,Paul Nurse and his colleagues here, Peter Fantes, Pierre Thuriaux and Kim Nasmyth developed a model of the cell cycle controls which was as sophisticated as any at the time. The timing of mitosis was proposed to be controlled by a mechanism which monitored cell size just before mitosis and was activated when the cell reached a critical size. Although the nature of the mechanism was unknown, three genetic elements were important - cdc 2, cdc 25 and wee 1. A similar, though less well understood size control also operated on the initiation of the S period in the small wee 1 mutants. Some of us at any rate thought that the mitotic control might be the build-up or the run-down of a molecular species, a ‘mitogen’, which reached a critical concentration at a particular cell size. The ’80s saw the application to pombe of the new and powerful methods of molecular genetics and recombinant DNA technology. Many of the cell cycle genes were cloned and sequenced and their protein products identified. A number of groups have been involved in this research of which probably the most important are the laboratories of Paul Nurse and of David Beach working on the mitotic control (the G2-M transition) and of Mitsuhiro Yanagida(4) working on the events of mitosis itself. A current model of the mitotic control has the activation of cdc 2 as the initiating event of mitosis, with this activation being promoted by cdc 25 and being inhibited by wee 1. However the number of genes which affects this control has increased from these three to thirteen (Fig. 2). How far all these are an integral part of the control pathway or whether they are peripheral interacting elements remains to be determined. Further exciting developments of the last few years have shown that parts of the pombe mitotic control are a highly conserved mechanism which is not peculiar to yeast but also occurs in a wide range of eukaryotes including both plants and men. This has brought together the work on pombe and the biochemistry of cell cycle control in early embryos which had been following a parallel pathway, though without a comparable genetic element. Two important parts of this
Nutrients
i.\
Protein
' ? I
I
r 4 CDCI,
Fig. 2. Elements of mitotic control. Boxed symbols represent the products of the respective genes. Reproduced by permission of Academic Press from Fantes in Ref. 3.
biochemistry are Maturation Promoting Factor and proteins called 'cyclins' which fluctuate during the cycle of early embryos. Maturation Promoting Factor has now been isolated and it is believed to have two subunits, one of which is a cyclin and the other is the gene ) . addition, there is a product of cdc 2 ( ~ 3 4 ' ~ ' ~ In homologue of cyclin in cdc 13 in pombe. A recent thoughtful review of the field is by Murray and Kirschner@). Standing back a little, I have to admire the very considerable achievements in understanding the machinery of mitotic control. A great deal of hard and well-directed work has been done and we now know a number of important gear wheels, of which we were ignorant ten years ago. It is also very encouraging to find what seems to be a broad universality in eukaryotic cells. There is no doubt that it is more complicated than we believed it to be in earlier years and perhaps this will also be true of many other aspects of cell regulation. A good deal of the control of mitosis seems to be embedded in a complex network of interacting proteins; there is, as yet, much less evidence for transcriptional controls of the process. So what, if anything, is missing from this train of gear wheels? The answer is the further gears that precede and follow the train. The ones that follow may be the easiest to identify since they lead to a known object - mitosis. The problem here is that mitosis is a physical event and not too much is yet known about its biochemistry. A number of proteins are involved but how they interact is far less clear. Here, Yanagida's isolation of mutations affecting mitosis in pombe will certainly prove to be enlightening. The preceding gears may be harder to identify. They may not exist in the compressed and invariant cycles of early embryos but they are surely there in pombe and most other growing eukaryotic cells. There must be a timing mechanism or trigger which sets the gear wheels going. The main clue to this in pombe is the size control which operates before mitosis. It is known to exist and to be affected by various things including nutrients but
its molecular basis is unknown. The future challenge will be not only to fit these two models together but also to incorporate the effects of pulses of heat and cycloheximide@) which stem from even earlier models of cycle control. Another similar challenge which is growing at the moment is understanding the molecular basis of the initiation of DNA synthesis at the G1/S transition. In pombe, this is an important control point in wee cells as well as in mammalian cells and bacteria. There are certainly going to be interesting developments here, probably in the near future. Since this is my article about my favourite cell, 1 have licence to expand a little on cell growth during the pombe cell cycle, which is my favourite topic though one much less fashionable than mitotic control. The strategy of a growing cell is to double all its components of growth but the tactics that it adopts varies with respect to the components. Total protein increases continuously in most cells whereas DNA has a restricted period of increase. Over the years, a lot of information has been collected about cell cycle growth in pombe both with single cells and synchronous cultures and more is known about this process in pombe than in most other eukaryotic cells. One important point is that for all the 18 components and properties that have been carefully studied, there are periodic patterns during the cell cycle. There is no definitive evidence of continuous exponential increase. I use the term periodic in a wide sense to cover what I have called a 'linear pattern' where there are linear segments of a growth curve interrupted by more or less sharp doublings in rate once per cycle. This is a subtle pattern to the observer of a growth curve and hard to prove without rate measurements by pulses but it may well make a considerable difference to the cell when a component of growth suddenly doubles its rate of production. We have very little information about the control of these periodic changes though there is evidence that in some cases they are dependent on cell size and in others that they are dependent on mitosis. However, in the majority of the cases that we have studied, periodicities continue at approximately cell cycle-length intervals after the main cell cycle events of DNA synthesis and mitosis have been blocked. This is an unexpected and surprising result; most people would predict that periodicities, at any rate in growing cells, would be dependent on the main cell cycle events. Our present view is that there are 'oscillatory' controls (using oscillator in the broadest sense to cover any periodic timing mechanism) which are closely entrained to the normal cell cycle but can free-run when the cycle is blocked. We are now investigating the nature of the entrainment signals but there is clearly much to be done before the molecular bases of the controls are understood. There are other intriguing problems in growth and morphogenesis. Cells of pombe grow mainly at their ends. Tip growth is in itself an interesting and largely unsolved problem of how the wall is laid down in a
definite form outside the plasma membrane. But this of course is not a problem unique to pombe, since tip growth occurs in many fungi and in a more general form for all cells with structures outside the membrane. But there are some particular problems in pombe about the sites of wall growth. In some strains, at least, growth in length starts early in the cycle at one end only. Later, the second end starts to grow, usually more slowly, and this transition appears to be related to cell size. At about the time of mitosis, elongation stops and a little later, wall growth starts again when the quite complex septum grows inwards in the centre of the cell. When the cell divides and forms two new ends, one might well expect that wall growth which has been forming the new ends would continue there. But in most cases this does not happen and wall growth starts again at the opposite (old) ends. We know that wall growth is associated with the presence of actin but we know nothing about the mechanisms that shift the positions of wall growth. Then again, what is it that determines the central position of the nucleus? There is good evidence that the microtubules of the cytoskeleton are involved in holding the nucleus in position since the nucleus can wander if the microtubules are disorganised but it is not clear how they know what the right position is, especially in a cell growing at one end only. There is a point about variation which is made in several chapters in the recent book(3). Synchronous cultures average out cell cycle events over millions of cells. Examining individual cells, however, reveal quite real differences. In spite of size controls, cells do not divide at a constant size. What is more, they can show marked differences in patterns of growth. We do not know why, though it is worth remembering that one of the cell cycle control models, ‘transition probability’, does take one kind of variability into account(’). I have largely concentrated in this article on the cell cycle work on pombe both because of my own interests and because I believe it to be the most important recent work with this yeast. But it is not of course the only side ofpombe that has been worked on and other interesting lines of research, especially in genetics, can be found in the recent
I should finish by trying to answer a question that is sometimes put. Why work on pombe when the ‘other’ budding yeast Saccharomyces cerevisiae has a better background of physiology, biochemistry and genetics, both classical and molecular, and is of much greater economic importance? In the cell cycle field, the answer is that the work onpombe is well advanced, and in total, probably more so than in budding yeast. Furthermore, there are real and interesting differences which can be illuminating and which are not entirely unexpected, considering the evolutionary divergence of the two yeasts. The cycle controls of the wild-type cells are different, as also is their morphology and mode of division. They differ in chromosome number and centromere structure and also in the relations of vegetative growth and the sexual process. A final difference to note in their cell cycles is the quite different pattern of activity and transcription of DNA ligase@). There is clearly good reason to continue work on both yeasts.
References 1 LEUPOLD, U . (1950). Die Vererbung von Homothallie und Heterothallie bei Schizosaccharomyces pombe. C.R. Trav. Lab. Carlsberg, Ser. Physiol. 24, 381 -480. 2 MITCHISON, J . M . (1957). The growth of single cells. 1. Schizosaccharomyces pomhe. Exp. Cell Res. 13, 244-262. 3 Molecular Biology o f t h e Fission Yeast (eds. A . Nasim, P. Young and B . F. Johnson). 1989. Academic Press, San Diego. 4 YANACIDA, M. (1989). Gene products required for chromosome separation. In The Cell Cycle. J. Cell Sci. Suppl. 12, 213-229. 5 MURRAY, A. W. AND KIRSCHNER, M . W. (1989). Dominoes and clocks: the union of two views of the cell cycle. Science 246, 614-621. 6 POLANSHEK, M . M. (1977). Effects of heat shock and cycloheximide on growth and division of the fission yeast Schizosaccharomycespombe. J. Cell Sci. 23, 1-23. 7 BROOKS,R. F. (1981). Variability in the cell cycle and the control of proliferation. In The Cell Cycle (ed. P. C. L . John), pp. 35-61. Cambridge University Press. Cambridge. 8 WHITE,J . H. M . . BARKER, D. G., NURSE,P. A N D JOHNSTON, L. H . (1986). Periodic transcription as a means of regulating gene expression during the cell cycle: contrasting modes of expression of D N A ligase in budding and fission yeast. EMBO J. 5, 1705-1709.
J. M. Mitchison is at the Department of Zoology, West Mains Rd., Edinburgh EH9 3JT, UK.
1
1Optn
Fig. 1. (Original legend). Schizosaccharomyces pombe in phase contrast. Central vacuoles, lipid granules and bumps due to division scars are shown. Reproduced by permission of Experimental Cell Research from ref. 2.
recent publication of a multi-author monograph titled Molecular Biology of The Fission and references for most of what I say in this article can be found there. Despite its fashionable title, it contains more than what is strictly molecular biology, namely chapters on genetics, cytology, morphogenesis and growth. Today, cell biologists are probably most interested in the contributions made with pombe on the control of mitosis and DNA synthesis. This topic started when Paul Nurse learnt pombe genetics from Leupold in the early ’70s and brought it to Edinburgh as a tool for studying the cell cycle. From this came the isolation of cdc mutants which are blocked in their passage through the cell cycle at a high (restrictive) temperature and also wee mutants which are altered in size at division. In the case of the cdc mutants, we were following the pioneer footsteps of Lee Hartwell who first isolated and named these mutants in budding yeast. In the late ’ ~ O S ,Paul Nurse and his colleagues here, Peter Fantes, Pierre Thuriaux and Kim Nasmyth developed a model of the cell cycle controls which was as sophisticated as any at the time. The timing of mitosis was proposed to be controlled by a mechanism which monitored cell size just before mitosis and was activated when the cell reached a critical size. Although the nature of the mechanism was unknown, three genetic elements were important - cdc 2, cdc 25 and wee 1. A similar, though less well understood size control also operated on the initiation of the S period in the small wee 1 mutants. Some of us at any rate thought that the mitotic control might be the build-up or the run-down of a molecular species, a ‘mitogen’, which reached a critical concentration at a particular cell size. The ’80s saw the application to pombe of the new and powerful methods of molecular genetics and recombinant DNA technology. Many of the cell cycle genes were cloned and sequenced and their protein products identified. A number of groups have been involved in this research of which probably the most important are the laboratories of Paul Nurse and of David Beach working on the mitotic control (the G2-M transition) and of Mitsuhiro Yanagida(4) working on the events of mitosis itself. A current model of the mitotic control has the activation of cdc 2 as the initiating event of mitosis, with this activation being promoted by cdc 25 and being inhibited by wee 1. However the number of genes which affects this control has increased from these three to thirteen (Fig. 2). How far all these are an integral part of the control pathway or whether they are peripheral interacting elements remains to be determined. Further exciting developments of the last few years have shown that parts of the pombe mitotic control are a highly conserved mechanism which is not peculiar to yeast but also occurs in a wide range of eukaryotes including both plants and men. This has brought together the work on pombe and the biochemistry of cell cycle control in early embryos which had been following a parallel pathway, though without a comparable genetic element. Two important parts of this
Nutrients
i.\
Protein
' ? I
I
r 4 CDCI,
Fig. 2. Elements of mitotic control. Boxed symbols represent the products of the respective genes. Reproduced by permission of Academic Press from Fantes in Ref. 3.
biochemistry are Maturation Promoting Factor and proteins called 'cyclins' which fluctuate during the cycle of early embryos. Maturation Promoting Factor has now been isolated and it is believed to have two subunits, one of which is a cyclin and the other is the gene ) . addition, there is a product of cdc 2 ( ~ 3 4 ' ~ ' ~ In homologue of cyclin in cdc 13 in pombe. A recent thoughtful review of the field is by Murray and Kirschner@). Standing back a little, I have to admire the very considerable achievements in understanding the machinery of mitotic control. A great deal of hard and well-directed work has been done and we now know a number of important gear wheels, of which we were ignorant ten years ago. It is also very encouraging to find what seems to be a broad universality in eukaryotic cells. There is no doubt that it is more complicated than we believed it to be in earlier years and perhaps this will also be true of many other aspects of cell regulation. A good deal of the control of mitosis seems to be embedded in a complex network of interacting proteins; there is, as yet, much less evidence for transcriptional controls of the process. So what, if anything, is missing from this train of gear wheels? The answer is the further gears that precede and follow the train. The ones that follow may be the easiest to identify since they lead to a known object - mitosis. The problem here is that mitosis is a physical event and not too much is yet known about its biochemistry. A number of proteins are involved but how they interact is far less clear. Here, Yanagida's isolation of mutations affecting mitosis in pombe will certainly prove to be enlightening. The preceding gears may be harder to identify. They may not exist in the compressed and invariant cycles of early embryos but they are surely there in pombe and most other growing eukaryotic cells. There must be a timing mechanism or trigger which sets the gear wheels going. The main clue to this in pombe is the size control which operates before mitosis. It is known to exist and to be affected by various things including nutrients but
its molecular basis is unknown. The future challenge will be not only to fit these two models together but also to incorporate the effects of pulses of heat and cycloheximide@) which stem from even earlier models of cycle control. Another similar challenge which is growing at the moment is understanding the molecular basis of the initiation of DNA synthesis at the G1/S transition. In pombe, this is an important control point in wee cells as well as in mammalian cells and bacteria. There are certainly going to be interesting developments here, probably in the near future. Since this is my article about my favourite cell, 1 have licence to expand a little on cell growth during the pombe cell cycle, which is my favourite topic though one much less fashionable than mitotic control. The strategy of a growing cell is to double all its components of growth but the tactics that it adopts varies with respect to the components. Total protein increases continuously in most cells whereas DNA has a restricted period of increase. Over the years, a lot of information has been collected about cell cycle growth in pombe both with single cells and synchronous cultures and more is known about this process in pombe than in most other eukaryotic cells. One important point is that for all the 18 components and properties that have been carefully studied, there are periodic patterns during the cell cycle. There is no definitive evidence of continuous exponential increase. I use the term periodic in a wide sense to cover what I have called a 'linear pattern' where there are linear segments of a growth curve interrupted by more or less sharp doublings in rate once per cycle. This is a subtle pattern to the observer of a growth curve and hard to prove without rate measurements by pulses but it may well make a considerable difference to the cell when a component of growth suddenly doubles its rate of production. We have very little information about the control of these periodic changes though there is evidence that in some cases they are dependent on cell size and in others that they are dependent on mitosis. However, in the majority of the cases that we have studied, periodicities continue at approximately cell cycle-length intervals after the main cell cycle events of DNA synthesis and mitosis have been blocked. This is an unexpected and surprising result; most people would predict that periodicities, at any rate in growing cells, would be dependent on the main cell cycle events. Our present view is that there are 'oscillatory' controls (using oscillator in the broadest sense to cover any periodic timing mechanism) which are closely entrained to the normal cell cycle but can free-run when the cycle is blocked. We are now investigating the nature of the entrainment signals but there is clearly much to be done before the molecular bases of the controls are understood. There are other intriguing problems in growth and morphogenesis. Cells of pombe grow mainly at their ends. Tip growth is in itself an interesting and largely unsolved problem of how the wall is laid down in a
definite form outside the plasma membrane. But this of course is not a problem unique to pombe, since tip growth occurs in many fungi and in a more general form for all cells with structures outside the membrane. But there are some particular problems in pombe about the sites of wall growth. In some strains, at least, growth in length starts early in the cycle at one end only. Later, the second end starts to grow, usually more slowly, and this transition appears to be related to cell size. At about the time of mitosis, elongation stops and a little later, wall growth starts again when the quite complex septum grows inwards in the centre of the cell. When the cell divides and forms two new ends, one might well expect that wall growth which has been forming the new ends would continue there. But in most cases this does not happen and wall growth starts again at the opposite (old) ends. We know that wall growth is associated with the presence of actin but we know nothing about the mechanisms that shift the positions of wall growth. Then again, what is it that determines the central position of the nucleus? There is good evidence that the microtubules of the cytoskeleton are involved in holding the nucleus in position since the nucleus can wander if the microtubules are disorganised but it is not clear how they know what the right position is, especially in a cell growing at one end only. There is a point about variation which is made in several chapters in the recent book(3). Synchronous cultures average out cell cycle events over millions of cells. Examining individual cells, however, reveal quite real differences. In spite of size controls, cells do not divide at a constant size. What is more, they can show marked differences in patterns of growth. We do not know why, though it is worth remembering that one of the cell cycle control models, ‘transition probability’, does take one kind of variability into account(’). I have largely concentrated in this article on the cell cycle work on pombe both because of my own interests and because I believe it to be the most important recent work with this yeast. But it is not of course the only side ofpombe that has been worked on and other interesting lines of research, especially in genetics, can be found in the recent
I should finish by trying to answer a question that is sometimes put. Why work on pombe when the ‘other’ budding yeast Saccharomyces cerevisiae has a better background of physiology, biochemistry and genetics, both classical and molecular, and is of much greater economic importance? In the cell cycle field, the answer is that the work onpombe is well advanced, and in total, probably more so than in budding yeast. Furthermore, there are real and interesting differences which can be illuminating and which are not entirely unexpected, considering the evolutionary divergence of the two yeasts. The cycle controls of the wild-type cells are different, as also is their morphology and mode of division. They differ in chromosome number and centromere structure and also in the relations of vegetative growth and the sexual process. A final difference to note in their cell cycles is the quite different pattern of activity and transcription of DNA ligase@). There is clearly good reason to continue work on both yeasts.
References 1 LEUPOLD, U . (1950). Die Vererbung von Homothallie und Heterothallie bei Schizosaccharomyces pombe. C.R. Trav. Lab. Carlsberg, Ser. Physiol. 24, 381 -480. 2 MITCHISON, J . M . (1957). The growth of single cells. 1. Schizosaccharomyces pomhe. Exp. Cell Res. 13, 244-262. 3 Molecular Biology o f t h e Fission Yeast (eds. A . Nasim, P. Young and B . F. Johnson). 1989. Academic Press, San Diego. 4 YANACIDA, M. (1989). Gene products required for chromosome separation. In The Cell Cycle. J. Cell Sci. Suppl. 12, 213-229. 5 MURRAY, A. W. AND KIRSCHNER, M . W. (1989). Dominoes and clocks: the union of two views of the cell cycle. Science 246, 614-621. 6 POLANSHEK, M . M. (1977). Effects of heat shock and cycloheximide on growth and division of the fission yeast Schizosaccharomycespombe. J. Cell Sci. 23, 1-23. 7 BROOKS,R. F. (1981). Variability in the cell cycle and the control of proliferation. In The Cell Cycle (ed. P. C. L . John), pp. 35-61. Cambridge University Press. Cambridge. 8 WHITE,J . H. M . . BARKER, D. G., NURSE,P. A N D JOHNSTON, L. H . (1986). Periodic transcription as a means of regulating gene expression during the cell cycle: contrasting modes of expression of D N A ligase in budding and fission yeast. EMBO J. 5, 1705-1709.
J. M. Mitchison is at the Department of Zoology, West Mains Rd., Edinburgh EH9 3JT, UK.
