The Hunt for Cyclin
2008; Cell Press; Volume: 134; Issue: 2 Linguagem: Inglês
10.1016/j.cell.2008.07.011
ISSN1097-4172
Autores Tópico(s)Cancer-related Molecular Pathways
ResumoIt is 25 years since Tim Hunt discovered cyclin, the oscillating protein that drives activation of cyclin-dependent kinases and entry into mitosis (Evans et al., 1983Evans T. Rosenthal E.T. Youngbloom J. Distel D. Hunt T. Cell. 1983; 33: 389-396Abstract Full Text PDF PubMed Scopus (953) Google Scholar). It is 25 years since Tim Hunt discovered cyclin, the oscillating protein that drives activation of cyclin-dependent kinases and entry into mitosis (Evans et al., 1983Evans T. Rosenthal E.T. Youngbloom J. Distel D. Hunt T. Cell. 1983; 33: 389-396Abstract Full Text PDF PubMed Scopus (953) Google Scholar). In the late summer of 1982, students at the Marine Biological Laboratory in Woods Hole, Massachusetts, were finishing their last experiments and getting ready to return home. On the rocks of the nearby harbor, the sea urchins were gravid but nearing the end of their spawning season. Tim Hunt was busily radiolabeling the spawned sea urchin eggs to monitor patterns of protein synthesis after fertilization. Looking at the autoradiogram of the protein gel, the result was striking (Figure 1). A specific protein, aptly called cyclin, would continuously accumulate and then precipitously disappear at cell division (Evans et al., 1983Evans T. Rosenthal E.T. Youngbloom J. Distel D. Hunt T. Cell. 1983; 33: 389-396Abstract Full Text PDF PubMed Scopus (953) Google Scholar). Could this oscillating protein be linked to the "clock in the cell"? Tim Hunt's classic experiment set in motion a series of discoveries that would explain the fundamental mechanism determining cell division and would garner Tim—together with Lee Hartwell and Paul Nurse—the 2001 Nobel Prize in Physiology or Medicine. By providing a clear model of the cell cycle engine, it became possible to monitor and understand the detailed molecular events of cell division within complex biological processes during development, tissue regeneration, and tumorigenesis. The experiment was devilishly simple but required careful execution and a biochemist's sense of which time points would show the important dynamic events. Tim's early training as a biochemist guided his approach. As he explains in a footnote to his Nobel Prize lecture (Hunt, 2002Hunt T. Biosci. Rep. 2002; 22: 465-486Crossref PubMed Scopus (30) Google Scholar), most embryologists did not take 10 min time points! In the early 1960s, Tim had become interested in protein translation. Cell-free reticulocyte lysates for translation in eukaryotic cells were new. Reticulocytes produce mostly globin mRNAs, and β-globin is the major translation product, so monitoring protein synthesis was straightforward with this new system, which attracted not only Tim but also his biochemist colleagues, Tony Hunter and Richard Jackson. Important experiments included investigating how the synthesis of hemoglobin was controlled by its physiological cofactors heme and iron. Although protein synthesis was known to be essential for many developmental and adaptive processes, it was not clear which steps during protein synthesis were the most strongly regulated. It was also unclear whether protein turnover or regulation of the steady-state accumulation of proteins would be driven by changes in protein synthesis or might also be strongly affected by protein degradation. The dynamics of protein accumulation, or even the extent of dynamic control in cell biology, was not well understood at that time. The idea that a protein could be triggered for rapid destruction at the moment chromosomes aligned on the metaphase spindle was unprecedented. Tim noted in his Nobel Prize lecture (Hunt, 2002Hunt T. Biosci. Rep. 2002; 22: 465-486Crossref PubMed Scopus (30) Google Scholar) that in solving major research problems, often "the clues and answers came from indirect attacks in unexpected quarters, rather than a full frontal assault." This comment is apropos of Tim's studies in protein translation, leading to key principles in cell division and protein destruction. So how indirect was the attack, and how unexpected the quarters? The "quarters" were unexpected because the central chapter in Tim's story was set during his summer sojourns at the Woods Hole Marine Biological Laboratory, a summertime haven for mitosis gurus and other scientists attending the famed advanced summer courses. Tim was a Woods Hole regular, and he spent those summers examining the synthesis of proteins that might be important in the embryogenesis of invertebrates such as surf clams (Spisula solidissima) and sea urchins (Arbacia punctulata). These organisms were a historical favorite at Woods Hole because a trip to the nearby reefs or fishing docks provided an easily accessible source. Eggs and sperm are readily produced by these creatures in the laboratory (only a little filtered sea water is needed), and the embryos are beautifully transparent and well-suited for microscopy. The remarkable synchrony of egg development made biochemical and molecular characterization of the phases of cell division straightforward. Tim worked with Eric Rosenthal and Joan Ruderman to look at proteins synthesized in Spisula eggs and observed three prominent new protein species appearing around 40 min after fertilization (Rosenthal et al., 1980Rosenthal E.T. Hunt T. Ruderman J.V. Cell. 1980; 20: 487-494Abstract Full Text PDF PubMed Scopus (151) Google Scholar). Surf clam eggs are laid as G2-arrested oocytes, and fertilization begins as a synchronous cycle of mitotic exit, S phase, and mitosis. At the time, there were few examples of translational control, and these new proteins, encoded by maternal mRNAs, were intriguing. Were these key structural proteins or regulatory proteins, and were they important in cell division or in some longer-term developmental process? In 1971, Yosio Masui and Clement Markert identified Maturation Promoting Factor, or MPF, a protein activity that induced mitosis in oocytes from the frog Xenopus laevis, even in the presence of protein synthesis inhibitors. Later, MPF (often called Mitosis Promoting Factor) was shown to induce mitosis in all eukaryotic cells. The notion of a factor sufficient to induce mitosis in the absence of new protein synthesis was surprising. MPF activity could be extracted from a mature embryo and transferred by injection to naive oocytes over multiple cycles, suggesting that there might be some mechanism to amplify MPF after each transfer. Marc Kirschner, Mike Wu, and John Gerhart showed that MPF has self-amplifying properties, suggesting that an enzymatic activity was sufficient for mitosis (Gerhart et al., 1984Gerhart J. Wu M. Kirschner M. J. Cell Biol. 1984; 98: 1247-1255Crossref PubMed Scopus (380) Google Scholar). Then, MPF was linked to a protein kinase activity, suggesting that MPF might be a kinase. The other critical observation was that MPF activity oscillated during the cell cycle, suggesting that MPF might be linked to the "clock in the cell." The biochemical mechanism for this self-amplifying oscillator was unclear. But it was known from nonequilibrium dynamics that specific chemical reactions could oscillate by accumulating products that were highly spatially organized, only then to reverse the flux of the equilibrium to regenerate substrate. How such a mechanism could be achieved in a cell was an exciting mystery. In 1982, Tim was back at Woods Hole teaching the physiology summer course. His earlier work with Ruderman in clams showed that new proteins were synthesized after fertilization. Did similar proteins exist in the sea urchin, given their different physiology after fertilization? Unlike clams, sea urchin eggs are laid during G0 of the cell cycle, with mitosis resuming after a considerable time lag. Tim looked for differences in the patterns of protein synthesis after fertilization of Arbacia eggs. The control groups included eggs activated without sperm by a calcium ionophore. The eggs were incubated with radiolabeled methionine, and samples collected at 10 min intervals. Radiolabeled protein was precipitated and separated according to mass on protein resolving gels, and protein species were visualized with photographic film to see the radioactive bands (Figure 1; Evans et al., 1983Evans T. Rosenthal E.T. Youngbloom J. Distel D. Hunt T. Cell. 1983; 33: 389-396Abstract Full Text PDF PubMed Scopus (953) Google Scholar). Although several bands accumulated after fertilization, one of them precipitously disappeared close to the time of cell cleavage. A careful plot of cyclin accumulation clearly showed that the disappearance of cyclin occurred at mitosis. The almost inexorable and what later proved to be correct conclusion was that "cyclin" underwent continuous synthesis (further demonstrated by pulse-chase experiments), followed by specific proteolysis at mitosis. The simple observation of an oscillating protein peaking at mitosis, even with no clear sense of the mechanism, suggested the link to a cell cycle oscillator. Tim then verified the existence of cyclins in the California sea urchin, Lytechinus pictus, and in Spisula. Once synthesized, cyclin could be destroyed in mitotic eggs even in the presence of protein synthesis inhibitors, further supporting the notion that its rapid decrease required proteolysis. Interestingly, drugs that blocked cell division, including agents that inhibited tubulin, slowed the disappearance of cyclins, suggesting a mechanism to slow cyclin destruction when cells were blocked in mitosis. The now-famous explanation is that the tubulin inhibitors activated the spindle assembly checkpoint to block cyclin destruction. Tim noted in his Nobel lecture that the same day he obtained the cyclin result, he had an important conversation with John Gerhart about MPF activation. Using frog oocytes, Gerhart, Wu, and Kirschner discovered that activation of MPF during meiosis II (but not meiosis I) required new protein synthesis (Gerhart et al., 1984Gerhart J. Wu M. Kirschner M. J. Cell Biol. 1984; 98: 1247-1255Crossref PubMed Scopus (380) Google Scholar). The implication was that between meiosis I and II, something must have happened to inactivate MPF. The possibility that proteolytic destruction of cyclin could explain MPF inactivation between meiosis I and II and the idea that cyclin could somehow be linked to MPF was very appealing, but far ahead of its time. As later experiments in other organisms would demonstrate, observing cyclin oscillations is not simple. For example, in the syncytial embryo of Drosophila, the oscillations of cyclins occur only locally near the nuclei (Edgar et al., 1994Edgar B.A. Sprenger F. Duronio R.J. Leopold P. O'Farrell P.H. Genes Dev. 1994; 8: 440-452Crossref PubMed Scopus (255) Google Scholar). So the use of the fertilized sea urchin embryo was a fortunate choice, and it made the discovery of cyclins possible. Later characterization of cyclins would require more sophisticated tools to clone the cyclin genes and to produce high quality antibodies for immunoblots. Tim's focus then turned to cloning cyclin, an ambitious undertaking in those days, and he recruited two superb young molecular biologists, Jon Pines and Jeremy Minshull. To clone cyclin B, Jeremy and Tim used a technique called hybrid arrest of translation (Minshull et al., 1989Minshull J. Blow J.J. Hunt T. Cell. 1989; 56: 947-956Abstract Full Text PDF PubMed Scopus (286) Google Scholar). Joan Ruderman's lab had cloned cyclin A (Swenson et al., 1986Swenson K.I. Farrell K.M. Ruderman J.V. Cell. 1986; 47: 861-870Abstract Full Text PDF PubMed Scopus (337) Google Scholar), and comparison of the two sequences revealed the cyclin box, which later proved to be the domain that binds to and activates the cyclin-dependent kinase Cdc2. Crystal structures of cyclin with its kinases would eventually rationalize how cyclin binds to and activates its kinases. Amazing experiments followed. Swenson and Ruderman showed that injection of in vitro transcribed cyclin A was sufficient to induce maturation of frog oocytes and activation of MPF, proving that cyclin is sufficient for MPF activity (Swenson et al., 1986Swenson K.I. Farrell K.M. Ruderman J.V. Cell. 1986; 47: 861-870Abstract Full Text PDF PubMed Scopus (337) Google Scholar). There was some confusion over how cyclin might be linked to MPF in oocytes, in part because it was not yet clear that there were two major MPF activities, both capable of activating Cdc2 kinase. One activity was binding of cyclin, and the other was tyrosine dephosphorylation of Cdc2 by the Cdc25 phosphatase. The dephosphorylation pathway is critical in meiosis I, so no new cyclin synthesis is required to activate MPF in oocytes. After meiosis I, however, cyclin is destroyed and must be resynthesized in meiosis II, making synthesis of cyclin B essential. Cloning of cyclin B also led to the famous experiments by Murray and Kirschner showing that the addition of in vitro transcribed cyclin B1 mRNA to frog egg extracts (cleared of their endogenous mRNAs by nuclease treatment) was sufficient to induce mitosis, demonstrating amazingly that cyclin synthesis was the central driver of mitosis (Murray and Kirschner, 1989Murray A.W. Kirschner M.W. Nature. 1989; 339: 275-280Crossref PubMed Scopus (837) Google Scholar). Minshull, Blow, and Hunt had cloned two Xenopus B-type cyclins and found that antisense inactivation of both was required to block mitosis. So, one cyclin (A or B) was sufficient to induce mitosis, but multiple endogenous cyclin isoforms appeared to collaborate to induce mitosis in embryos. These three mitotic cyclins (A, B1, B2) were only a beginning, and indeed Tim's lab later demonstrated the presence and importance of multiple forms of cyclin B. The story of cyclin's connection to MPF is described in Paul Nurse's Nobel Prize lecture. Cdc2 was discovered as the kinase component of MPF in two ways: biochemical purification of MPF activity by Fred Lohka and Jim Maller and binding of MPF to the cyclin-Cdc2 regulator Suc1 by Bill Dunphy, David Beach, and John Newport. Connecting MPF to the fission yeast Cdc2 kinase (called Cdc28 in budding yeast, and now called Cdk1 by many) linked MPF to important genetic studies on the role of Cdc2 kinase in mitosis. The classic studies of Paul Russell and Paul Nurse (Nurse, 2002Nurse P.M. Biosci. Rep. 2002; 22: 487-499Crossref PubMed Scopus (99) Google Scholar) revealed the importance of Cdc2 in mitosis and of the Wee1-Cdc25 pathway in regulating tyrosine phosphorylation of Cdc2 in mitosis. Later work clarified that Cdc25 removed phosphotyrosine from Cdc2 to activate the kinase and that Wee1 could phosphorylate Cdc2 during S phase or after DNA damage. In addition to controlling mitotic entry, tyrosine phosphorylation of Cdc2 is the central mechanism blocking mitotic entry after DNA damage (Figure 2). Direct binding of cyclin to the Cdc2 kinase was demonstrated by the copurification of cyclin with Cdc2 in starfish by Marcel Doree's group. In those early days, the sufficiency of cyclin for inducing mitosis was a stunning result. However, later studies showed that multiple mitotic cyclins contributed to the events in mitosis. Considerable work explored the specific roles of different mitotic cyclins and showed that organisms balance their use of mitotic cyclins differently. More recently, knockout studies in mice suggest even more complexity in how cyclins, even the mitotic cyclins, are used in higher eukaryotes (Lee and Sicinski, 2006Lee Y.M. Sicinski P. Cell Cycle. 2006; 5: 2110-2114Crossref PubMed Scopus (74) Google Scholar). Knockout of specific mitotic cyclins and cyclin-dependent kinases in mice reveals a surprising developmental plasticity in the complement of cyclin genes needed. Possibly the most far-reaching idea to come out of Tim's discovery of cyclin oscillations was cyclin degradation by proteolysis. Work from Michael Glotzer, Murray, and Kirschner revealed that cyclin B was ubiquitinated and thus destroyed by ubiquitin-dependent proteolysis (Glotzer et al., 1991Glotzer M. Murray A.W. Kirschner M.W. Nature. 1991; 349: 132-138Crossref PubMed Scopus (1851) Google Scholar). Ubiquitination was known to be a major mechanism for the destruction of cytoplasmic proteins, but the discovery of cyclin as a substrate of the ubiquitin-proteasome system was an exciting link given the physiological importance of the cyclin substrate. The notion that cyclin destruction was triggered at a very specific cell cycle transition and was closely linked to chromosome segregation opened up the idea that a host of cellular events might also trigger protein destruction. But what was the mechanism of cyclin B destruction, and what was the proteolytic regulator? Studies from many groups (reviewed in Peters, 2006Peters J.M. Natl. Rev. 2006; 7: 644-656Crossref Scopus (963) Google Scholar) identified the Anaphase Promoting Complex or Cyclosome (APC/C) as the critical E3 ubiquitin ligase for cyclin. The APC/C was one of the first E3s identified and helped to define the molecular architecture of the RING finger class of E3 ligases. But why was cyclin only destroyed in mitosis? It turned out that phosphorylation of APC/C during mitosis activated the destruction of cyclins (Peters, 2006Peters J.M. Natl. Rev. 2006; 7: 644-656Crossref Scopus (963) Google Scholar). What was the signal for cyclin destruction? The classic paper by Murray, Kirschner, and Mark Solomon showed that cyclin had an N-terminal destruction signal (Murray et al., 1989Murray A.W. Solomon M.J. Kirschner M.W. Nature. 1989; 339: 280-286Crossref PubMed Scopus (802) Google Scholar), leading to further characterization of the destruction (D) box, one of the earliest defined degradation signals (degrons). But how did the D-box function as a degradation signal? Many contributed here, but Yamano and Hunt added one explanation. Using cyclin B from fission yeast, they first showed that a protein fragment containing the D-box and ubiquitination site blocked cyclin degradation when expressed in fission yeast and inhibited cyclin B ubiquitination in Xenopus extracts (Yamano et al., 1998Yamano H. Tsurumi C. Gannon J. Hunt T. EMBO J. 1998; 17: 5670-5678Crossref PubMed Scopus (91) Google Scholar). They proposed that the D-box fragment could bind to and block a D-box receptor. To find the receptor, Yamano et al., 2004Yamano H. Gannon J. Mahbubani H. Hunt T. Mol. Cell. 2004; 13: 137-147Abstract Full Text Full Text PDF PubMed Scopus (106) Google Scholar used a recombinant purified protein with tandem D-boxes coupled to beads as an affinity matrix. The D-box matrix bound APC/C tightly when incubated in Xenopus extracts, but unexpectedly, this binding was independent of known activator proteins, arguing that the D-box receptor was within the core of the APC/C itself. Tim continued his studies of the major activities defining cell cycle stages. An S phase promoting activity (SPF) was known to be required to trigger DNA replication. Later work showed that SPF required the S phase cyclins E and A. What made S phase cyclins different from mitotic cyclins? Tim provided one explanation. Cyclins E and A were known to stimulate DNA replication in Xenopus extracts and to activate Cdk2, whereas cyclin B was known to induce mitosis and to bind to Cdc2 kinase. In vitro, the three kinases were quite similar in activity and phosphorylated similar substrates. So why didn't cyclin B induce DNA replication? As Tim's group showed, the key was that cyclin B is excluded from the nucleus by an active nuclear export sequence during S phase and only enters the nucleus at prophase to trigger mitosis. They hypothesized that without the ability to enter the nucleus, cyclin B could not gain access to substrates important for S phase. To redirect cyclin B, they removed the N-terminal nuclear export signal from cyclin B and replaced it with the nuclear localization signal of cyclin E's N-terminus. Amazingly, this retargeted cyclin B entered the nucleus and triggered DNA replication independent of cyclin E. Thus, nuclear exclusion is a major factor determining the selectivity of cyclin B for mitosis. This is very much the kind of experiment I associate with Tim. Simple in concept and beautifully executed, it makes an important point with a clear message. The discovery of cyclins by Tim Hunt 25 years ago is one of those rare scientific advances that changed the fields of biochemistry and molecular biology and provided key physiological insights into a fundamental cellular process. It is more than 50 years since the structure of DNA suggested the nature of semiconservative replication. Tim's discovery of cyclins helped provide the next insight to explain how the cycle of semiconservative replication and chromosome segregation were linked to cell division. That clarification greatly focuses the next generation of cell biologists on how cell growth and development instruct the cell cycle, where the cell cycle goes awry in cancer and other diseases, and hopefully how targeting the cell cycle clock may provide a strategy for eliminating cancer cells. The Hunt for CyclinPeter K. JacksonCellOctober 17, 2008In Brief(Cell 134, 199–202; July 25, 2008) Full-Text PDF Open Archive
Referência(s)