When talking about his earlier life and times, Sir Paul Nurse, President of the Royal Society and Nobel laureate, uses one word rather frequently. It is, somewhat surprisingly, failure. A clever grammar school boy from a working class family, it took him a while to get to grips with passing exams at school, and his route into university was more tortuous than most. He did an unexciting PhD, publishing two papers, which in the years since have garnered only five citations. After his postdoc, working on an untrendy subject, he failed to get any of the lectureships for which he applied and had to take a post funded by short-term money at the University of Sussex, to keep afloat. When that money ran out, he failed to get still more jobs and was on the brink of leaving the country for a laboratory in Germany when he was rescued by Walter Bodmer, Director of the ICRF.

In addition to this lexicon of failure, other facts fall out from past interviews with Paul. How did he get interested in science? In two ways: first, as a small, solitary boy, standing entranced in the street in his pyjamas as Laika the space dog and Sputnik 2 flew through the night sky over Wembley; subsequently, observing with fascination the life beneath his feet as he walked to and from school through the fields. Interviewers have covered other subjects as diverse as his Desert Island Discs (luxury: a telescope; favourite record: Billy Joel “Just the way you are”), his attempts to learn the recorder as an adult, and why he shaved off his moustache (“I thought it would make me look more vulnerable.”). In recent years, the bombshell of Paul’s discovery that his parents were, in fact, his grandparents, and his real mother was the person he had thought was his older sister, has added an extra level of tragic complexity to the story.

Paul Nurse’s baptism as a scientist came in a most unlikely place, just off the North Circular Road near the Hanger Lane gyratory in Willesden, North London, at the Guinness Brewery-owned Twyford Laboratories. He was 17 years old and had finished school a year early, leaving with good A levels, but crucially lacking the French O level he would need to get him into university in the 1960s. His school career had been almost comically inconsistent; he veered between the bottom and the top of the class, sometimes in the same subject, but his bête noire was definitely languages. Despite having a razor-sharp mind able to grasp almost anything as long as it has logical consistency, Paul struggled, and still struggles, with pronouncing unfamiliar words in English, let alone in French. Combined with a total ignorance of grammatical structure and an inability to associate the strange French words with either their English equivalents or the objects they described, he was a walking linguistic disaster. Accordingly, the original plan for his post–A level year was to stay on in the sixth form to work at his French and to redo an A level or two (an A and two B grades not being considered quite good enough). However, after a term of boredom and wanting to earn a bit of money to help with the family income, Paul dropped out. Resigning himself to taking French O level by himself (famously, he eventually failed it a grand total of six times), he got a job making biological media for a group studying Salmonella food poisoning at the Twyford Labs, while he reapplied to university.

Fortunately for the young Nurse, somebody high enough up in the hierarchy of a British university had the good sense to realise that incompetence in a foreign language was no barrier to becoming a good scientist. In 1967, after a journey on his motorbike up the new M1 motorway for a hastily convened interview, Paul started a degree in Biology at Birmingham University, thanks to John Jinks, head of the Genetics Department. Jinks’s own staunchly working-class background may have predisposed him towards bucking the Establishment’s desire for unnecessary qualifications. He cooked up a deal with a member of the French faculty to allow Paul to take and pass a semi-fictitious course in French in his first year, and, with the rules bent but not quite broken, Paul was free to start on his path to scientific glory.

The path certainly didn’t seem particularly glorious to begin with. Paul is remembered by Jack Cohen, one of his Birmingham lecturers, as being undistinguished, and Paul himself summarises his third-year undergraduate project, on the respiration rate of dividing fish eggs, as “a disaster, but a useful disaster,” falling at the last hurdle because a vital control had been omitted. Having started off with an interest in ecology and genetics, he abandoned them early on, the former because he “couldn’t bear to be muddy and wet and out of control” and the latter because of an excessive emphasis on cytogenetics, not the most exciting of topics. Instead, Paul took a classical biology degree, majoring in botany and zoology, and got a First, despite having missed all of the second-year zoology lectures in a valiant attempt, no doubt, to fail at something yet again.

The haphazard route by which he arrived at his first-class degree had provided Paul with two unusual mental resources, which give some clues to his subsequent outstanding success: he was very relaxed about failing, never regarding it as much of a problem, and his sense of intellectual excellence was entirely self-referential, making him highly competitive, but only with himself.

Because I failed at so many things so often, because I was in a mess at school ... it gave me a sort of internal discipline—you take less note of what other people think of you, what other people say, because you don’t get off on being praised about things. I had to be resilient inside. … I was constantly comparing me to me when I did well, and not with other people. I realise it’s very odd, but it’s really useful, because when I … failed examinations, I couldn’t get into university, I couldn’t get a job, when you put all that together, it was a constant low to medium level [of] failure about things. So when I got to difficult problems and I failed, I didn’t go into depression or anything. And when you get into research, it’s constant failure all the time, and I was perfectly trained for it.

So what does a self-competitive optimist with no fear of failure do for a PhD? Probably influenced at least slightly by the fish eggs debacle in his final year, Paul had already decided that there was no point slogging away in a lab unless you were working on something that really mattered, something that was important enough to stand a good risk of failure on a grand scale. Paul’s lecturers at Birmingham, traditional biologists all, had encouraged him to apply to botany departments for his PhD, and, although he had been offered a project at Cambridge, the topic, on the biochemistry of RNA, was not to his taste. His omnivorous reading habit had led Paul far beyond the mundane nuts-and-bolts problems that such a PhD project would address, into a theoretical land of conjecture and speculation about how biological entities are ordered; how does an organism build itself up from a single cell, and how does the resulting community of cells interact during a lifetime?

Forty years on, Paul is still preoccupied with this question, which lies at the heart of understanding biology itself, and the length of elapsed time is indicative of the scale of the problem. Wisely, his younger self made a very logical decision to start small, studying one of the most important but probably the simplest developmental decisions a cell can make—how its structures are organised in space and time to enable it to divide into two. Understanding the molecular events that regulated this decision, dissecting the changes in enzyme and gene regulation that were required, might just be approachable at a single-cell level, even taking into account the rather limited molecular techniques then available.

After his uninspiring visit to Cambridge, Paul eventually settled on a PhD at the University of East Anglia (UEA) in Norwich. The Biological Sciences School at UEA had a lecturer, Tony Sims, who was just beginning a project on cell cycle changes in the enzymes involved in amino acid metabolism in the yeast Candida utilis, and this seemed to fit Paul’s criteria for fundamental cell cycle research very well. In addition to the suitability of the project, Norwich as a place was attractive to Paul, partly because his family originated from there and it was where he had been born, and partly because as a budding socialist, he was very taken by the idea of working at a new university, free from the Establishment constraints that he’d already had to fight against to even get a degree. UEA was less than 10 years old and had taken part enthusiastically in the period of student unrest sweeping through Europe in the late 1960s; the unfortunate Princess Margaret had had a blazing Union Jack thrown at her during a visit a couple of years before Paul’s arrival. Paul signed on and started work there in 1970.

His PhD work on amino acid metabolism, coupled to his bookworm tendencies, had led Paul into thinking a lot about metabolic pathways, and particularly the ways of controlling flux through a pathway. Metabolic pathways are the production lines of the cell, combining raw materials into more sophisticated molecules, which are then fed into the vast machinery required to maintain and drive the life of the cell. However, as for any production line, the rate of flow through the system must be controlled very carefully, in order to ensure that products are made in the right amounts, neither flooding nor starving the market, so to speak. Metabolic pathways can be regulated either at multiple points or in just a few places, depending on how they are set up, and it occurred to Paul that this sort of reasoning could be applied to cell cycle control. If the cell cycle were regulated in only a few places, then it should be possible to speed up the cycle drastically, or alternatively slow or stop it altogether, by messing with just one regulation point. To switch similes, if the cell cycle controls were like the sluice gates in a watermill, the speed at which the cell traversed the cycle, like the speed at which a mill wheel turns, could be entirely regulated by tinkering with just one item, the molecular equivalent of opening and closing a sluice. If this were true, then there should be cells that went into mitosis too early or went into mitosis late or not at all, and by finding the mutations that had caused the changes, the identity of the regulators could be established.

This theoretical daydreaming was all very well, but there was one big snag to it; much as he racked his brains, Paul was unable to think of how to approach the problem practically, how to find the mystery regulators of the cell cycle. Then, one night in 1972, staying late to do yet another boring experiment, he found the answer. It lay in a paper in the journal Proceedings of the National Academy of Sciences, written by an American called Leland Hartwell.

Using this method, Hartwell quickly isolated about a thousand mutants that were unable to grow at 36°C but were quite happy at normal room temperature, and began to categorise them based on their defects. Although some were too boring to study, he did find mutations that could be specifically attributed to defects in one of a number of cellular processes such as DNA, RNA, or protein manufacture, cell division, and cell growth. To prove the point that this genetic approach would work to identify the individual genes responsible for each mutation, Hartwell used his protein synthesis mutants to show that three were caused by specific defects in three particular proteins, and hence the genes that encoded them.

Using time-lapse video microscopy, Reid, Hartwell, and a second PhD student, Joe Culotti, then went one step further. By filming mutant cultures continuously, starting at the time at which the cells were shifted to the restrictive temperature, they found that some cells stopped in the cycle almost immediately, whereas others went around another cycle before arresting at the same point. This meant that mutations could act at one particular point in the cycle and that if cells happened to have got past that point before the temperature shift, they would continue until they ran up against the mutation point in the next cycle. Careful tracking of the cells able to do another round of cycling also threw up another interesting observation—mutations did not necessarily act at the point where the cells arrested in cycle but could exert their effects much earlier on.

It was Hartwell, Culotti, and Reid’s paper describing this work that Paul fell upon during his nocturnal trip to the UEA library, and to someone searching for a way to get a grip on an intractable problem, it was an astonishing moment: As Paul says, “It completely blew my head off.” The more Paul discovered about Hartwell’s work, the more he became convinced that it was the simplest, most direct way to get to the heart of what regulated the cell cycle. The mutants would lead straight to the genes of most importance, and then identifying the genes and working out what they did would put the framework in place to solve the whole problem. Genetics, not biochemistry, was clearly the way to go. Fired with evangelical zeal, Paul began working out the steps that would enable him to follow Hartwell into the maze.

Murdoch Mitchison, part of the “third of a ton of Biology Professors” born to the writer, poet, and activist Naomi Mitchison, and older brother of Av, David Lane’s mentor, was very definitely interested in cell division, having recently published an extremely influential book, The Biology of the Cell Cycle. Mitchison had shifted into the cell cycle field in the 1950s, after earlier work on sea urchins and red blood cell membranes, and had realised early on that fission yeast was a great model, perhaps better than budding yeast, because to divide, it simply got longer and longer until it split in two. This meant that position in the cell cycle could be measured by length, rather than the more subjective bud size. The fission yeast cell cycle also looked a lot more “mammalian” than that of budding yeast, because the timing of each part of its cycle was similar to that of higher eukaryotes, a fact that fueled Paul’s growing enthusiasm for his potential new laboratory workhorse.

Mitchison had thought a great deal in biochemical terms about how the cell cycle might be regulated, and The Biology of the Cell Cycle, although a dry and dusty read, was the first clearly to lay out two new and important concepts. First was the relationship between the cell cycle and cell growth, the idea that because cells can only divide if they are big enough, they must therefore have a way of sensing size and tying that into cell cycle control. Second was the recognition of the causal dependency of events in the cell cycle, that one event had to happen in order that another, subsequent one could occur. Furthermore, Mitchison proposed that the time it took for a cell to complete the different steps in the cycle could be determined by these dependencies, and also by external, master timing mechanisms. This idea was a forerunner of more advanced theories of checkpoint control, which underpin our modern understanding of the cell cycle.

In Bern, Urs Leupold, a brilliant but eccentric geneticist, spent hours with Paul to teach him the rudiments of genetic analysis, although Paul remembers some of his wilder ideas as quite “barkingly mad.” Paul took to the abstract, highly theoretical subject material like a duck to water; his great strength had always been that he was almost pathologically orderly in his thought processes, and the rigorous logic of theoretical genetics matched his mind in a very satisfactory way. As well as the long hours with Leupold, Paul began to work with the lab’s fission yeast strains and set up his first temperature-sensitive screens, looking on the replica plates grown at the restrictive temperature for cells that were getting too long because they couldn’t complete the cell cycle and split into two. By the time he moved back to Edinburgh, and in the months following, he had isolated about 30 different cell division cycle (cdc) mutants. The Mitchison lab’s expertise in physiological studies of the cell cycle meant that he could categorise the cdc mutants on the basis of where they were blocked, just as Hartwell’s lab had done with budding yeast, and Paul found mutants blocked for DNA synthesis, mitosis, and cell division, aided in subsequent work by Kim Nasmyth, then a new and “frighteningly bright” graduate student in the Mitchison lab.

The temperature-sensitive screens that Paul had been performing were proving to be very fruitful, but searching through thousands of colonies on agar plates looking for the small number that might have mutated was unexciting and long-winded. In a bid to speed matters up, Paul decided that a better approach might be to mutate the cells, grow them up at the restrictive temperature, and then separate them on the basis of their size; the mutant cells stuck in cycle would be much larger, and they should be easy to separate by spinning through a lactose gradient in a centrifuge because the largest cells would all collect at the bottom of the gradient. The big cells could then be recovered by growing on agar at a normal temperature, and then tested by temperature shift, as normal. In fact, the idea was a complete failure for what Paul wanted, because most of the mutant cells didn’t recover, and the normal cells swelled up too as a result of the stresses they were under. However, whilst fruitlessly looking for extra-long cells extracted from the bottom slice of the gradient, Paul noticed to his surprise that there were some strange-looking, small, round cells that had made it to the bottom slice because they had clumped into a ball, which to a centrifuge looks just the same as a big cell. He was about to throw the plate containing the tiny cells away but suddenly stopped. If cells that were too long were caused by a block in the cell cycle, then surely, cells that were too small might be caused by a cell cycle that ran too fast, dividing before the cells had grown enough? Long cells, he realised, were not the key to finding the rate-limiting control steps in the cell cycle, because any event necessary for the cell cycle would become rate-limiting if it were inhibited. However, if there were a mutant that speeded up an event rather than inhibited it, such a mutant would be in a bona fide rate-limiting gene, because speeding up a normal rate-limiting step would push the cells to cycle faster. It was the difference between getting a puncture in a bike tyre (necessary), which would cause a journey to be slowed, and pedaling faster (rate-limiting) to arrive earlier.

Paul’s 1975 Nature paper on the wee mutant phenotype, “Genetic Control of Cell Size at Cell Division in Yeast,” saw his arrival on the yeast cell cycle scene, and during the next few years in Edinburgh, he published another 15 papers characterising in detail the stable of cell cycle mutants that he had built up. At first, his work ploughed a very similar furrow to Hartwell’s, and Paul worried that he would always be running to catch up the budding yeast cell cycle community, which was much larger than the small band of fission yeast enthusiasts. However, Hartwell’s work ran into a slow patch, and without his input, budding yeast cell cycle work stalled for a few years, allowing the fission yeast community to draw level.

The cell cycle is split into four phases: G₁, S, G₂, and M. S phase is the time during which DNA is duplicated; M phase is mitosis, when the duplicated DNA is split and the cell divides into two; and G₁ and G₂ are the gaps on either side of S phase. The transitions between G₁ and S phase and G₂ and M phase are key regulatory points in the cell cycle, where go/no go decisions are made regarding DNA synthesis and cell division. Paul’s work in Edinburgh, done in collaboration with Peter Fantes, Kim Nasmyth, Pierre Thuriaux and others in the Mitchison lab, focused mainly on the G₂/M transition, which is the preeminent control point in fission yeast. By the late 1970s, they had shown that all of the wee mutations Paul had isolated were causing changes in just two genes, called wee1 and cdc2, and that these two genes were rate-limiting for the G₂/M transition, somehow telling the cell cycle that the cell was big enough to divide into two without stinting on the genetic and biochemical dowry of either daughter cell. Interestingly, the cdc2 gene had first been identified as a mutant in the elongated cell screen, and Paul and Pierre Thuriaux proposed in 1980 that cdc2 could be mutated either so that it lost its function, in which case, mitosis was blocked and cells elongated, or so that it became more active, in which case, cells divided too early and were too small. They further suggested that cdc2 and wee1 were working together in a network to regulate the G₂/M transition and hence the onset of mitosis.

The buzz going round about G₁/S and Start meant that there was not a lot of interest in G₂/M control, and Paul, by now looking for a lectureship somewhere as a prelude to leaving Edinburgh, made a pragmatic decision to do a little bit of opportunistic work on the G₁/S transition in fission yeast, in order to look more attractive to prospective employers. He and Yvonne Bissett, a technician in the Mitchison lab, set up a quick experiment to see whether any of their temperature-sensitive cdc mutants had anything to do with Start. The assay was simple, based on a normal lifestyle decision made by yeast cells, which divide happily in conditions of optimal nutrition, but stop dividing and mate (conjugate) when they detect problems with food supplies. The point at which cells make the decision to divide or to conjugate is Start, and once a cell has passed Start and is committed to divide, it cannot conjugate. Therefore, at the restrictive temperature, if a cdc mutant were needed for Start, cells would arrest in G₁, would not be able to commit to entering S phase, but would still be able to conjugate. If the cdc mutant were irrelevant to Start, cells would arrest at some other point in the cycle and no conjugation would occur. The readout of the assay was to select for cells that could conjugate, which Paul and Yvonne did by rigging the medium on the agar plates they used such that only conjugated cells would stay alive.

The “double block point” got the paper into Nature. What Paul had realised was that cdc2 might be not only the rate-limiting regulator for G₂/M, but also the key to Start control at G₁/S. Viewed in this light, the data made perfect sense; when shifted to the restrictive temperature, cdc2 mutant cells could either arrest in G₂, in which case they’d be unable to undergo conjugation because they were in the wrong bit of the cell cycle; or just before Start, in G₁, where they could conjugate with ease. This would appear as a dribble of surviving colonies at the end of the assay, just as Paul and Yvonne had seen.

Paul again: “That paper is one of my favourites because to get to that conclusion was not intuitive—it wasn’t how people were thinking about it; it wasn’t how I was thinking about it. To make the interpretation, to say, ‘Well, I can test that if I do this, but the chances of this working are zero.’ To do that, and then to get the result.… I don’t talk about it so much these days but it’s one of the three or four things I’m most pleased with.”

The Nature paper, “Gene Required in G₁ for Commitment to Cell Cycle and in G₂ for Control of Mitosis in Fission Yeast,” was published in August 1981, by which time Paul had moved on from Edinburgh, to a three-year position at the University of Sussex in Brighton. Despite his attempts to up his trendiness factor, he had been unable to get a lectureship anywhere and had had to settle for a short-term fellowship. Although not very secure in terms of job prospects, the location of Paul’s new lab was in one way ideal; the School of Biological Sciences at Sussex had expertise in what he was planning next—molecular biology.

Not everyone was convinced of the potential of molecular biology. Distaste for molecular biology and its purveyors was quite common amongst the more intellectually fastidious geneticists, who viewed it as a rather grubby, menial discipline, where one could get so caught up in the minutiae of getting tricky new methodologies to work that there was no time to think about more important concerns. To some extent, they had a point, because biological problems had a tendency to be parked while techniques were honed to perfection, but those geneticists willing to engage with molecular biology reaped a large reward in terms of data.

The divide is well illustrated in a Nature News and Views article that Paul wrote in 1980, at the end of his time in Edinburgh, entitled, “Cell Cycle Control—Both Deterministic and Probabilistic?” The article marks a transition from a theoretical to a real universe in cell cycle theory and points out to the theoreticians that although their work had laid down the important concepts underpinning cell cycle research, it was time to get their hands dirty and start figuring out what was really going on. It was almost a description of Paul’s own journey from theory to practice:

I had all the thinking that those guys had, and all the theory stuff, I was comfortable with that, but I was also very practical. And that was really helpful, because most people who came after that had no idea of the earlier stuff—it sunk without trace. But it was very important for me to make the right research decisions, so it actually influenced me very positively, not because I took it seriously, because I didn’t, but because it made me think about the research problems and the issues round them much, much more thoroughly than I would have done otherwise. I owe a lot to it.

There were two aspects to the problem: firstly, the transformation protocol itself, and secondly, getting DNA into a form that the fission yeast transcription machinery would be able to recognise and transcribe into protein. Paul was soon joined by David Beach, an escapee from a nearby lab, and the two of them split the work by category, Paul working on the problem of softening up the yeast cells so that they were ready to take up foreign DNA, and David trying to adapt the plasmid vectors that worked in budding yeast to make them suitable for fission yeast. The two got on well together because Beach was very effective and had a lot of experience of molecular genetics already. Vectors started to appear for testing, but Paul’s end of the experiment was not going so well. He had got his technician to make up a set of solutions according to Jean Beggs’s protocol for budding yeast, and very unexpectedly, they seemed to work very well for transforming fission yeast too. This was wonderful news, and Paul took the technique with him on a visit to another lab, eager to spread the word. Somewhat embarrassingly, however, when he tried to repeat his success, he couldn’t get a single transformed colony and returned home to Brighton with his tail between his legs. In Brighton, transformation continued, mysteriously, to work. For a time it seemed as though there might be a geographical restriction on fission yeast molecular biology, until Paul realised that the crucial solution in the protocol had been made up incorrectly, with one ingredient, sorbitol, missing. Sorbitol turns out to inhibit transformation in fission yeast, so by complete chance, Paul had hit on exactly the right modification to Beggs’s recipe. As an example of fortune favouring the slightly incompetent, this story can hardly be bettered, because by honest labour alone, it would almost certainly have taken a really long time to get to the same place.

Having saved himself from the dreadful fate of becoming a budding yeast person, Paul could now get to grips with the problem of cloning fission yeast cdc genes, with the highest priority being cdc2, the dual regulator of Start and the G₂/M transition. He, Beach, and Barbara Durkacz, a new postdoc in the lab, decided to use a method published in 1980 by Kim Nasmyth, who had turned to the dark side following his PhD with Murdoch Mitchison and was now working on budding yeast as a postdoc in Ben Hall’s lab in Seattle.

Back in Sussex, using a fission yeast genomic library, a plasmid carrying cdc2 was identified by complementation cloning, but to make the paper a bit cuter, Paul and his colleagues decided to see whether they could redo the experiment, but this time trying to complement mutant cdc2 using a library from budding yeast. In other words, did cdc2 have a budding yeast homologue, able to have the same effect on the cell cycle? A budding yeast library was duly screened, and sequences were pulled out that could, indeed, complement the fission yeast cdc2 mutation. This was a great result because budding yeast and fission yeast are only very distantly related in evolutionary time, having diverged maybe 400 million years ago; finding a gene that worked in both species was pretty amazing. But what was the gene? Because cdc2 is important at Start, it seemed a good idea to obtain plasmids carrying Start regulatory genes from Steve Reed and test them out individually in the complementation assay. Most surprisingly, the one that worked was Nasmyth and Reed’s original CDC28 plasmid: cdc2 and CDC28 were functionally homologous, functionally the same.

Beach, Durkacz, and Nurse published their work in the Christmas 1982 issue of Nature, and in the final paragraph, they made a very prescient remark: if two yeasts separated by nearly half a billion years of evolution were using the same mechanism for regulating Start, might it not be possible that the mechanism was so evolutionarily ancient, so entrenched in a cell’s life cycle, that control of the analogous Restriction Point in mammalian cells might involve a functional equivalent of the cdc2 and CDC28 gene products? In that single sentence lay the seeds of a future Nobel Prize.

Getting fission yeast molecular genetics going had been a real coup for Paul and his small lab, and his rapid reemergence in the pages of Nature following his voluntary publishing shortfall should have been a good indicator to the scientific hierarchy that here was a future star in the academic firmament. Paul was enthusiastic, incredibly bright, charismatic, and working on a really important problem, and he should have walked into a job. But again, just as in Edinburgh, he went for many interviews but always came in as the second or third choice. In a profession that should be a meritocracy, he was hampered by the nonscientific prejudices of the hiring committees he faced. He had never worked in the secular cathedrals of genetics or molecular biology, did not have an Oxbridge degree, and did not have sufficiently influential mentors to drop a good word in the right ear. Furthermore, he was not doing the right kind of work, perched as he was on the fence between genetics and molecular biology at a time before the combined weight of fellow converts caused the fence to collapse. In short, he was a risk: he might be as scientifically gifted as he appeared, but he didn’t have a thoroughbred lineage, and nobody was prepared to commit themselves.

Driving all of these intense discussions about cell cycle theory, fission yeast lore, and technical troubleshooting was the day-to-day work of the lab, and here Paul’s mania for order and tidiness ruled. His own assessment of this is that he is “quite well organised in a shambolic sort of way,” but others are more illuminating. Sergio again:

And then, there was Paul’s legendary appetite for data, starting with his eagerness for more of it getting in the way of laboratory good manners. Tony Carr remembers that after one key experiment, “I came in in the morning and he’d already been through all the plates, decoded them from my lab books and worked out what the result was. After that I used to hide my lab books on the top shelf where he couldn’t reach them.” Jacky Hayles, still working with Paul 30 years on, now as joint head of his lab at the London Research Institute, talks about his attention to detail and his ability to see things in experiments that no one else had noticed, although “he can be very irritating when you see it yourself and you want to tell him, but he wants to tell you first.”

The next few years of the Nurse lab read like a Greatest Hits album for cell cycle research. Paul, despite his brief flirtation with G₁/S and Start, was still focussed on G₂/M control and what cdc2 was doing to regulate it, and answers were now coming thick and fast. Viesturs Simanis, a postdoc who arrived in 1985 from David Lane’s lab at Imperial College equipped with the ability to get almost any experiment, however difficult, to work, showed that the cdc2 gene encoded a protein kinase, able to put phosphate groups on other proteins and thereby regulate their activity, and Sergio Moreno went on to show that the kinase activity varied during the cell cycle and peaked just before mitosis, as might be expected for a rate-limiting regulator of G₂/M.

Russell and Nurse’s two Cell papers on cdc25 and wee1 in 1986 and 1987, and Simanis and Nurse’s on cdc2 in 1986, together with the constellation of papers surrounding them, were the vivid proofs of the theories that Paul had formulated back in Norwich as a PhD student. In the intervening years, Paul, with the help of his colleagues, had worked out how to find rate-limiting steps in the cell cycle, had gone out and isolated mutants in those steps, had taught a whole field how to do molecular genetics, and now knew the DNA sequences of the mutants and could see exactly how they worked together. The whole saga was an amazing tour de force, especially given the stunning lack of interest shown by the rest of the scientific world. cdc2, the master kinase that was responsible for flipping the on/off switch for the G₂/M transition, was itself decorated with phosphate groups (phosphorylated) by the wee1 proteins, also kinases, and this phosphate decoration inactivated it. When cdc2 needed to be switched on, cdc25, which encoded a phosphatase (a protein able to remove phosphate groups), undid wee1’s work, and cdc2 protein became active. It was beautifully simple. But did it matter anywhere other than yeast? Paul, pretty much alone in the field as usual, thought that it did.

Paul’s conviction that cdc2 would be present in other species besides yeast had been bubbling under for some time, because he knew that entry into mitosis from G₂ was regulated just as tightly in higher eukaryotes as it was in fission yeast. In 1971, Yoshio Masui, working on frog oocytes, had discovered the existence of a mysterious activity, which he called Maturation Promoting Factor, or MPF. Frog oocytes arrest naturally in G₂ before fertilisation but can be forced to mature and go through mitosis by injection of MPF, contained in an extract made from mature egg cytoplasm. MPF activity turned out to be present in all of the higher eukaryotic cell types, including mammalian, that could be tested, and to be fundamentally important; it was a rate-limiting G₂/M regulator just like cdc2, but its true identity was still unknown by the early 1980s, despite strenuous attempts to purify it from the cytoplasmic soup in which it hid.

In 1985, Paul hired a new postdoc from Jean Beggs’s lab at Imperial College. Melanie Lee was looking for a second postdoc, and knowing that Viesturs Simanis, an old friend from Imperial, was having a great time in the Nurse lab, decided to apply there herself. The application process started in a pub near ICRF, where Melanie discovered Paul, “scruffy, little and very vibrant,” having a drink with a vision of male pulchritude “all in white, tall and Adonis-like,” who turned out to be Kim Nasmyth. Paul, much to her astonishment, said yes to Melanie almost immediately, and she started work shortly afterwards. Her first day in the lab featured a meeting with the boss to discuss possible projects, and she remembers Paul presenting her with two options: “He said, ‘You can do one of two projects. You can work on cdc10, which is fairly ordinary, or you can work on a project that’s high risk but justifies my presence in a cancer institute; that is, you clone human cdc2.’” Melanie, without hesitation, went for cdc2.

Melanie Lee and Freeway, 1987.

(Photograph courtesy of Melanie Lee.)

Paul had tried to sell the human cdc2 idea to other new lab entrants but had got no takers, because of the high likelihood that it would be a total bust. Of course, by Paul’s lights, this made it a fine project: “If I wasn’t doing things where I thought there was a reasonable chance of failure, I wouldn’t think we were doing anything important. I begin to realise I’m a bit odd in that way.” Spending years looking for a protein that might exist only in the mind of their boss was not a high priority for anyone hoping to get a good job on the back of their postdoc work. Besides, there was so much else to be done that it was easy to pick less-risky winners; other projects might require elegant genetics and tricky biochemistry, but as long as you were good at the bench and worked hard, the payoffs were almost guaranteed.

There was another factor too, that of her relative inexperience in the field. Unlike the rest of the lab, she came from a non-fission-yeast, non-cell-cycle background, and she lacked the knowledge to dispute Paul’s judgement that the project had a chance. As Paul says: “Melanie was the only one who said she’d have a go at it, because I said in my view this was the most interesting project in the lab. She was receptive to my advice, which was useful for a project like hers. I was suggesting she did quite bold things and she was prepared to trust me. I don’t think the others would have necessarily done that.… She was also extremely capable and stable, and really efficient and effective, which was necessary for this project.”

The human cdc2 project illustrates very well the hard slog and serial failures that accompany most scientific ventures. After many months of painfully slow nonprogress, it became very obvious to Melanie why Paul had emphasised the difficulty of the project and why nobody else had wanted to take it on. Very simply, there appeared to be no good way of even looking for a human cdc2 homologue by conventional methods, let alone finding something.

When searching for relatives of known proteins, there were at that time two standard ways to proceed: look for similarities at the DNA level and find the gene encoding the protein, or search for similarities at the protein level and then work back to the gene. Both methods were molecular fishing expeditions, in Melanie’s case using fission yeast baits dipped into enormous pools of human genes or proteins. To search by DNA homology, Melanie probed a human gene library with an antisense version of the fission yeast cdc2 gene sequence, hoping that the fission yeast antisense DNA would be able to pick out its human sense counterpart. To look for protein homology, she used an antibody able to recognise fission yeast cdc2, this time fishing in a human protein library, hoping that the antibody would recognise some shared structural component enabling it to hook out human cdc2.

Matters came to a head at a group meeting in the summer of 1986, about nine months after Melanie had started, when as usual, she had to present yet another litany of failure to the assembled lab. It was obvious that pursuing her current strategy was pointless, but what else was there? The only other method of finding homology was by functionality, as Beach, Nurse, and Durkacz had done when they had discovered that cdc2 and CDC28 were doing the same thing in fission and budding yeast, but how on earth could one do that between yeast and human? Leaving aside the sheer unlikeliness of a human gene being close enough in function to be able to rescue a fission yeast mutant, there was no way of doing the experiment, unless there was a human protein library that would work in fission yeast, which, surely, there wasn’t.

And then someone—Tony Carr, Viesturs Simanis, nobody can quite remember who—realised that there was. It was probably only the Nurse lab that had the combined knowledge and audacity to contemplate it, but there was a library that might, just might, work. It had been made by Hiroto Okayama in Paul Berg’s lab in Stanford, California, and funnily enough, Melanie had already been using it for her fruitless fishing experiments. The Okayama and Berg library contained pretty much every gene in the human genome cloned into a plasmid vector able to grow in Escherichia coli bacteria, just like many standard libraries from that time. However, the neat factor about it was that Okayama had worked out how to get the library to switch on, or express, all of its proteins in mammalian cells as well as E. coli, by topping and tailing the gene sequences with control elements that the fussy transcriptional machinery of mammalian cells could recognise and use.

Even if the library could be put into fission yeast, many further obstacles loomed, but as the lab discussed the idea with a mounting sense of excitement, the realisation flew around the room that although it would be bold to try it, it was not totally foolhardy. One by one, the requirements for getting the experiment on the road fell into place. There were two main issues. First was the question of what strain of yeast the library should be put into. Paul and Jacky had that one nailed instantly, because they knew that they had a cdc2 temperature-sensitive mutant strain, cdc2-33 leu1-32, which was totally watertight, stopping absolutely dead at the restrictive temperature. If the library went into that and any colonies grew at the higher temperature, then it had to be caused by an incoming gene rescuing the defective cdc2. Next was how to tell whether the library had made it into the yeast cells at all—the transformation procedure had to be maximally efficient to make sure that every human gene possible was represented. The way round that one was to use the Okayama and Berg plasmids to piggyback a second plasmid carrying a selectable marker, LEU2, into the cells. If the transformed cells were grown in medium lacking leucine, only those carrying the LEU2 plasmid would survive, because the cdc2 mutant strain that Paul and Jacky had suggested using had been modified so that it could not synthesise leucine on its own. The bare bones of the experiment were there: transform the library into cdc2-33 leu1-32 cells, along with the LEU2 plasmid, and plate at the permissive temperature in the absence of leucine; after a day, when any transformed cells would have had a chance to recover and start growing, shift to the restrictive temperature, and look for any colonies that could grow normally. Easy.

Like decorating, the secret of doing successful experiments lies in the time spent on preparation, and this experiment took a lot of forethought. The logistics were frightening. Normally, yeast transformations are done using about 10 million cells and ∼1 µg (a millionth of a gram) of plasmid DNA. To get enough colonies to ensure that every human gene had a decent chance of getting into a yeast cell, the procedure had to be amplified a hundred-fold; unfortunately, this was not just a matter of finding a big bucket to put all the cells in, but of doing a hundred separate transformations, because the protocol did not scale up efficiently. The hundred transformations then had to be spread onto hundreds of agar plates, which all had to be jammed into incubators, and then the massed plates had to be checked daily for growing colonies. It was just as well that Melanie was, in Paul’s words “about the most well-organized worker I’ve ever had,” and that Martin Goss, in Melanie’s opinion, was “perfect—he didn’t care whether it was a mad experiment or not, he still provided the technical help, and he was a good pair of hands at the bench, a jolly good technician.”

After a couple of small-scale runs, to maximise the transformation efficiency, Melanie and Martin ran their first big experiment, got nothing, and managed to really annoy the rest of the lab all at the same time; putting a vast number of coldish agar plates into the lab’s 36°C incubators turned out to cause such a drastic drop in temperature that everyone else’s temperature-sensitive mutants started growing again, wrecking a good few experiments. They tried again, being a bit more careful about the temperature issue. After two weeks, “just when you’re about to throw the plates out because you’re fed up with it,” they saw one or two rather pathetic colonies struggling along on the otherwise empty plates. By this time, Melanie and Martin were on a roll, so they set up another huge experiment and left the plates to cook whilst Melanie flew off to Banff in Alberta for that year’s annual yeast scientific beano, more formally known as the 13th International Conference on Yeast Genetics.

Back in London after the meeting, any vestiges of jet lag and alcohol poisoning dissipated instantly when Melanie looked at the latest experiment. This time, there were five colonies, and they looked really promising. There was an emergency lab meeting to decide what to do. Because fission yeast is very good at scrambling introduced DNA sequences, sometimes disassembling plasmids and stuffing them into its own genome, the most urgent task was to rescue the plasmid DNA from the cells before it vanished. The colonies were plated out on new agar to amplify them up, and when Paul came to take a look at them, he realised that they were really onto something, because whatever was rescuing the large cdc2 mutant cells had to have come from a library plasmid. Sergio Moreno, who had arrived in the lab only a few weeks previously, remembers: “Paul was not doing experiments at the time, all he was doing then was looking at the cells—he really liked to look at cells under the microscope. It was clear when he looked at [one of] the colonies that it was really growing. When he picked the colony and spread it he realised there were small cells that were the complementing ones mixed with large cells that were losing the plasmid. He was really excited about that and everybody in the lab was aware this was going to be very important.”

In the event, only two of the five colonies still had rescuable plasmid DNA in them, but it was enough. Melanie transformed the precious DNA back into the cdc2-33 leu1-32 cells, to see whether either plasmid would complement the cdc2 mutation, and to wild excitement, she hit the jackpot with both; at the restrictive temperature, cdc2-33 leu1-32 cells were able to cycle normally as long as the plasmids were present. Comparison of the two plasmids showed that they contained an identical DNA insert, and the insert was also present, although chewed up, in the three other original colonies from the screen.

Paul was probably most affected:

I didn’t want to burn us, by just following nonsense. When the clones came up, when we showed it was due to a plasmid … what I remember is that every time Melanie did something, we discussed why it might not be saying what we hoped it would say. I then went through this psychologically peculiar time where I assumed it would fail at each step, but we just kept going. I was in this schizoid thing—I was keeping myself safe assuming it wouldn’t work, then I’d go home at night and think, “I’m just going to imagine it has worked, because I’ll go in tomorrow and it will be gone.”

Sequencing of the 2000 bp (base pairs) of DNA in the inserts carried on throughout the autumn, but by Christmas 1986, the work was finished and the complete sequence assembled. The good news was that the DNA sequence didn’t match either the fission yeast or the budding yeast cdc2 genes, and thus had to be something different, but depressingly, there were no obvious similarities between the new sequence and the yeast genes at the DNA level. But what about at the protein level? The yeast cdc2 and CDC28 proteins, although not completely similar, have one region that is invariant because it is needed for their activity as protein kinases. The region has the amino acid sequence proline–serine–threonine–alanine–isoleucine–arginine–glutamate, commonly abbreviated using the amino acid one-letter code as PSTAIRE. Melanie ran her new sequence through a programme that could decode the DNA, translating it to see what amino acids any protein made from it would contain, and with literally bated breath, she, Paul, and the lab stood over the old dot matrix printer as it churned out the result.

After that wonderful, climactic moment, there was still a fair amount of work to be done to show that the new sequence really did originate from a human source and to show that human cdc2 protein could be detected by the yeast cdc2 antibody (easier to do now that it was obvious what they were looking for), but the paper by Lee and Nurse, “Complementation Used to Clone a Human Homologue of the Fission Yeast Cell Cycle Control Gene cdc2,” was submitted to Nature in March and came out in early May 1987. The last sentence of the article was a vindication of Paul’s vision: “The identification of a cdc2-like function in human cells suggests that elements of the mechanism by which the cell cycle is controlled will probably be found in all eukaryotic cells.”

The appearance of such a major paper, the culmination of many years of speculation, hard work, boldness, and the odd bit of luck, should have been an occasion of great celebration for Paul and Melanie, but unfortunately, things didn’t quite turn out that way. The paper came out while both its authors were attending a conference in Heidelberg, and Paul was due to give a talk about its contents towards the end of the meeting. However, just before leaving Britain, Paul had picked up an ear infection, the pain of which was exacerbated by the flight from London to Frankfurt, and he ended up being admitted to the local hospital, spending most of the meeting there. He did manage to show up for his talk, but the import of what he was saying was overshadowed by his head being swathed in a huge white bandage. “I was in such pain I can’t even remember what I talked about. I had a horrible ear infection—it was unbelievably painful. The hospital had to puncture my eardrum and drain it. It was awful, awful. The worst pain.” Melanie fared little better, but for a nicer reason. She was in the first trimester of her first pregnancy and had awful morning sickness. Her summary of their triumph says it all: “I felt dreadful and Paul was really poorly.”

One of the first and most important molecules to fall to the newly unified field was MPF, Maturation Promoting Factor. In 1988, after almost two decades of effort, MPF was purified by Jim Maller’s lab in Denver and was shown to consist of two proteins, one of which was a kinase. Using antibodies raised against the PSTAIRE region of cdc2, two collaborative teams, the Nurse and Maller labs, and David Beach and John Newport’s labs, published back-to-back papers in Cell showing that the MPF kinase was, indeed, cdc2, as Paul had almost shown with Chris Ford back in Sussex in 1980. The primary mechanism regulating G₂/M in both yeast and higher eukaryotes was not just similar; it was identical.

In a particularly nice twist, MPF was also the means of uniting the work of two researchers who met in the early 1980s, hit it off instantly, and have kept the cell cycle world entertained with their energetic sparring ever since. When discussing science, Tim Hunt and Paul Nurse have the air of two small boys playing on a beach together, falling out, and making up, endlessly fascinated by the rock pools, the funny looking shells, and the task of building huge and sometimes precarious sandcastles. In 1982, Tim had accidentally stumbled into the cell cycle world through his work on protein degradation in sea urchin eggs, when he discovered cyclins, the keys to how the cell cycle turns. Cyclins must be made, degraded, and then made anew in order for cells to divide properly, and the reason for this became clear once MPF was purified. MPF comprises one molecule of Tim’s cyclin protein bound to one molecule of Paul’s cdc2 protein, and cdc2 cannot work without the assistance of cyclin; indeed, cdc2’s official biological name, CDK2, cyclin-dependent kinase 2, shows the defining nature of the interaction. Functional MPF is created when enough cyclin has built up to switch on cdc2 activity, thereby driving cells into mitosis. Cells cannot exit mitosis unless MPF activity is switched off, and for this to happen, cyclin must be degraded. At the end of mitosis, the cell divides, cyclin starts to build up again, and the process is repeated.

I think it’s a privilege to do what we do. I did make a pact with myself. I thought when I did a PhD, “Should I do something obviously useful, like work on malaria, for example, where contributing to our understanding, even in a small way, might be useful, or could I just indulge my own curiosity and work on whatever I like? And what I decided was, as long as I’m at the top of the tree I can do the latter. But if I slip from that then I will shift onto something more obviously useful.” And you’ll notice that I’ve had administrative, managerial posts for a long long time. People don’t believe me, but I don’t actually really enjoy it and I’m not even that good at it. I never remember half the things I’m supposed to remember or do. I do it because I feel I’m paying back a debt. So by spending half my time doing this other stuff, I justify what I do in the lab.

Nurse lab outing, ca. 1987.

(Photograph courtesy of Sergio Moreno.)

Web Resources