NEWS & VIEWS
For News & Views online, go to nature.com/newsandviews
CE LL D IVISIO N
A last-minute decision Robust quantitative analyses of asymmetric division in certain cells in flies identify the major molecular players and, most interestingly, define a simple equation to explain this complex cellular process. See Letter p.280 C AY E TA N O G O N Z A L E Z
T
he development of multicellular organisms requires the generation of many cell types from a single precursor. In some cases, this is achieved by asymmetric cell divisions, in which the two daughters of a dividing cell receive different complements of intracellular ‘fate’ factors. Evolution has devised different mechanisms for asymmetric cell division1, some of which are rather counterintuitive — as in the sensory organ precursor (SOP) cells of the fruit fly. When SOPs divide, vesicles called Sara endosomes, which are loaded with cell-fate molecules, become restricted primarily to one daughter. But near the end of cell division, Sara endosomes are still found in the centre of SOPs, seemingly running the risk of being partitioned equally between the two daughter cells. It is only when cell division is close to completion that they find their way to one daughter2. On page 280 of this issue, Derivery et al.3 explain how this last-minute decision works. The authors recorded thousands of SOP divisions using cells in which molecules and structures of interest were engineered to fluoresce. Interpreting these data was a hugely daunting task, for several reasons. To begin with, the speed of events and the size of the a
pllb
Sara endosome
structures involved are not far from the temporal- and spatial-resolution limits of advanced microscopy. Furthermore, living cells are not all identical in size, shape and orientation. Signal-to-background ratios are often suboptimal, and — let’s face it — biology is often noisy, being subject to random fluctuations that obscure the real data. One video will never provide a hint of how this system works, and even 1,000 might prove no more informative. Derivery et al. found a creative way to circumvent these problems, making use of a contractile ring structure that cleaves the SOPs in two. Rather than trying to derive information from each individual video, the authors used the position of the ring and the onset of contraction as spatial and temporal reference points from which to align them, superimposing every recording to generate a movie of the ‘average’ SOP division (see go.nature.com/ ccnyjs). This average cell is artificial, and does not exist in nature. Nonetheless, it is a feast of data, packed with precious, reproducible, spatio-temporal information about the cellular components involved in compartmentalizing Sara endosomes to one daughter. Time-resolved density plots of different components of the average SOP revealed two distinct phases of division. A structure called the central spindle, which contains
plla
many microtubules, is required for cell cleavage. During the first phase, the authors observed that both these microtubules and Sara endosomes, which move back and forth along the microtubules at a constant speed, are equally distributed between the two sides of the dividing cell (dubbed pIIa and pIIb). In the second phase, microtubule density is about 20% higher in the pIIb side, but Sara endosomes spend almost twice as much time in the pIIa side, and move beyond the end of microtubules more often in this direction. As a result, more than 80% of Sara endosomes end up in pIIa (Fig. 1). Next, Derivery and colleagues demonstrated that the activity of just three proteins can explain these observations. Throughout the process, Sara endosomes are transported along the microtubules by the protein Klp98A — a relative of a mammalian motor protein that moves cargo to one end of microtubules4. The asymmetry of the microtubules at the central spindle is brought about by enrichment of the protein Patronin on the pIIb side. Patronin then protects microtubules that are oriented in a certain direction from the depolymerizing activity of the protein Klp10A (ref. 5). By contrast, Klp10A disassembles the unprotected microtubules on the pIIa side. Remarkably, this complex intracellular choreography fits a rather simple equation — the
b
Microtubule
Cleavage plane
Figure 1 | A mechanism for asymmetric cell division. Derivery et al.3 have defined the mechanism by which vesicles called Sara endosomes are apportioned asymmetrically into two daughter cells (pIIa and pIIb) during the division of sensory-organ precursor cells in the fruit fly. a, Towards the end of cell division, a spindle structure based on microtubules forms in the centre of
the cell. Sara endosomes move evenly up and down the microtubules, and the microtubules themselves are arrayed symmetrically about the cleavage plane, down which the cell will split. b, As division progresses, some microtubules are disassembled. Levels of disassembly are higher on the pIIa side of the cell. Sara endosomes spend more time in pIIa, accumulating on this side.
1 9 6 | NAT U R E | VO L 5 2 8 | 1 0 D E C E M B E R 2 0 1 5
© 2015 Macmillan Publishers Limited. All rights reserved
JOHN GUILLEMIN/BLOOMBERG VIA GETTY
NEWS & VIEWS RESEARCH extent of endosome asymmetry is inversely proportional to microtubule asymmetry and exponentially amplified by several factors, including inverse microtubule asymmetry. This equation readily accounts for the counterintuitive observation that microtubule-binding endosomes accumulate on the side of the cell that has fewer microtubules, and explains how a small difference in microtubule density can translate into a large bias in endosome segregation. The authors confirmed that the equation works over a wide range of induced levels of central-spindle asymmetry, and defined each of its parameters experimentally. Putting this equation to the test, Derivery and colleagues placed a Patronin-sequestering molecule on the pIIb side of a SOP, on a structuralprotein network called the cortex that lines the interior of the membrane. This molecular trap draws Patronin from the pIIb side of the central spindle, making pIIb more susceptible to microtubule disassembly by Klp10A. The authors found that the microtubule network became denser in pIIa, and that most Sara endosomes were delivered to pIIb, as predicted by the equation. And, consistent with their hypothesis, placing the Patronin trap on the pIIa cortex affected neither microtubule asymmetry nor the location of Sara endosomes. This study is filled with in vitro assays, mutant analyses, purpose-made computer code and more. Moreover, it is an excellent example of the power of using solidly quantified data to solve complex problems in molecular cell biology, applying equations when intuition will not do. Given the ubiquitous nature of its components, the mechanism identified by Derivery and colleagues might not be restricted to Sara endosomes in dividing SOPs. It could be applicable to any process in which a molecular cargo is moved unidirectionally by a motor protein along microtubules that are arranged as an asymmetric bundle. Moreover, the evolutionary conservation of the proteins involved points to the possibility that this mechanism operates in other species. As the authors point out, one case that merits close examination is that of the branched projections of neurons that receive stimuli from other neurons. These projections contain many antiparallel microtubules, and vesicle transport plays a fundamental part in their function. Future research will determine whether the authors’ mechanism can be generalized to this, or to any other process, in mammalian cells. But, for now, we can say that fruit flies have once again delivered insight into a basic biological mechanism that may well be applicable to other organisms. ■ Cayetano Gonzalez is at the Institute for Research in Biomedicine (IRB Barcelona), the Barcelona Institute of Science and Technology, 08028 Barcelona, Spain; and at the Catalan Institution for Research and Advanced
Studies (ICREA), Barcelona. e-mail: [email protected] 1. Knoblich, J. A. Nature Rev. Mol. Cell Biol. 2, 11–20 (2001). 2. Coumailleau, F., Fürthauer, M., Knoblich, J. A. &
González-Gaitán, M. Nature 458, 1051–1055 (2009). 3. Derivery, E. et al. Nature 528, 280–285 (2015). 4. Hoepfner, S. et al. Cell 121, 437–450 (2005). 5. Goodwin, S. S. & Vale, R. D. Cell 143, 263–274 (2010).
CATA LYSI S
The complexity of intimacy Catalysts that contain two types of active site split long hydrocarbon molecules into more-useful shorter ones. Research into controlling the nanoscale separation of the sites challenges accepted design rules for such catalysts. See Letter p.245 ROGER GLÄSER
T
he petroleum industry produces jet fuel and diesel using a catalytic process called hydrocracking, in which longchain hydrocarbons are broken into shorter, more-useful ones. More than 250 million tonnes of hydrocarbons are hydrocracked each year1 (Fig. 1). The catalysts comprise an acidic microporous silicate called a zeolite and a noble metal, mixed together with a binder material1. On page 245 of this issue, Zečević et al.2 report a major breakthrough in the design of hydrocracking catalysts, which greatly improves the selectivity of the process so that more of the desired products are formed. The results shatter the belief that the noble-metal sites and the acid sites must be as close together as possible for effective catalysis. The distance between the acid sites and the noble-metal sites of hydrocracking catalysts must be below a maximum limit defined by the
‘intimacy criterion’, which was first reported3 in 1962. The metal sites catalyse dehydrogenation reactions that convert the alkane reactants into unsaturated hydrocarbons called alkenes, which must then diffuse to an acid site to undergo isomerization and cracking (cleavage into shorter molecules; Fig. 2). If the metal sites are too far from the acid sites, then catalytic activity decreases — satisfying the intimacy criterion therefore overcomes diffusive limitations of the catalytic process. Previous studies of the intimacy criterion have examined inter-site distances only at or above the micrometre scale. Zečević et al. now report two catalysts in which platinum was controllably placed at nanometre-scale distances from the acid sites. The catalysts contain platinum particles of around 3 nanometres in diameter, with a narrow particle-size distribution, so that the metal was selectively supported on either the zeolite or the binder (γ-alumina); the overall platinum content in
Figure 1 | A hydrocracking facility in Gdansk, Poland. 1 0 D E C E M B E R 2 0 1 5 | VO L 5 2 8 | NAT U R E | 1 9 7
© 2015 Macmillan Publishers Limited. All rights reserved
For News & Views online, go to nature.com/newsandviews
CE LL D IVISIO N
A last-minute decision Robust quantitative analyses of asymmetric division in certain cells in flies identify the major molecular players and, most interestingly, define a simple equation to explain this complex cellular process. See Letter p.280 C AY E TA N O G O N Z A L E Z
T
he development of multicellular organisms requires the generation of many cell types from a single precursor. In some cases, this is achieved by asymmetric cell divisions, in which the two daughters of a dividing cell receive different complements of intracellular ‘fate’ factors. Evolution has devised different mechanisms for asymmetric cell division1, some of which are rather counterintuitive — as in the sensory organ precursor (SOP) cells of the fruit fly. When SOPs divide, vesicles called Sara endosomes, which are loaded with cell-fate molecules, become restricted primarily to one daughter. But near the end of cell division, Sara endosomes are still found in the centre of SOPs, seemingly running the risk of being partitioned equally between the two daughter cells. It is only when cell division is close to completion that they find their way to one daughter2. On page 280 of this issue, Derivery et al.3 explain how this last-minute decision works. The authors recorded thousands of SOP divisions using cells in which molecules and structures of interest were engineered to fluoresce. Interpreting these data was a hugely daunting task, for several reasons. To begin with, the speed of events and the size of the a
pllb
Sara endosome
structures involved are not far from the temporal- and spatial-resolution limits of advanced microscopy. Furthermore, living cells are not all identical in size, shape and orientation. Signal-to-background ratios are often suboptimal, and — let’s face it — biology is often noisy, being subject to random fluctuations that obscure the real data. One video will never provide a hint of how this system works, and even 1,000 might prove no more informative. Derivery et al. found a creative way to circumvent these problems, making use of a contractile ring structure that cleaves the SOPs in two. Rather than trying to derive information from each individual video, the authors used the position of the ring and the onset of contraction as spatial and temporal reference points from which to align them, superimposing every recording to generate a movie of the ‘average’ SOP division (see go.nature.com/ ccnyjs). This average cell is artificial, and does not exist in nature. Nonetheless, it is a feast of data, packed with precious, reproducible, spatio-temporal information about the cellular components involved in compartmentalizing Sara endosomes to one daughter. Time-resolved density plots of different components of the average SOP revealed two distinct phases of division. A structure called the central spindle, which contains
plla
many microtubules, is required for cell cleavage. During the first phase, the authors observed that both these microtubules and Sara endosomes, which move back and forth along the microtubules at a constant speed, are equally distributed between the two sides of the dividing cell (dubbed pIIa and pIIb). In the second phase, microtubule density is about 20% higher in the pIIb side, but Sara endosomes spend almost twice as much time in the pIIa side, and move beyond the end of microtubules more often in this direction. As a result, more than 80% of Sara endosomes end up in pIIa (Fig. 1). Next, Derivery and colleagues demonstrated that the activity of just three proteins can explain these observations. Throughout the process, Sara endosomes are transported along the microtubules by the protein Klp98A — a relative of a mammalian motor protein that moves cargo to one end of microtubules4. The asymmetry of the microtubules at the central spindle is brought about by enrichment of the protein Patronin on the pIIb side. Patronin then protects microtubules that are oriented in a certain direction from the depolymerizing activity of the protein Klp10A (ref. 5). By contrast, Klp10A disassembles the unprotected microtubules on the pIIa side. Remarkably, this complex intracellular choreography fits a rather simple equation — the
b
Microtubule
Cleavage plane
Figure 1 | A mechanism for asymmetric cell division. Derivery et al.3 have defined the mechanism by which vesicles called Sara endosomes are apportioned asymmetrically into two daughter cells (pIIa and pIIb) during the division of sensory-organ precursor cells in the fruit fly. a, Towards the end of cell division, a spindle structure based on microtubules forms in the centre of
the cell. Sara endosomes move evenly up and down the microtubules, and the microtubules themselves are arrayed symmetrically about the cleavage plane, down which the cell will split. b, As division progresses, some microtubules are disassembled. Levels of disassembly are higher on the pIIa side of the cell. Sara endosomes spend more time in pIIa, accumulating on this side.
1 9 6 | NAT U R E | VO L 5 2 8 | 1 0 D E C E M B E R 2 0 1 5
© 2015 Macmillan Publishers Limited. All rights reserved
JOHN GUILLEMIN/BLOOMBERG VIA GETTY
NEWS & VIEWS RESEARCH extent of endosome asymmetry is inversely proportional to microtubule asymmetry and exponentially amplified by several factors, including inverse microtubule asymmetry. This equation readily accounts for the counterintuitive observation that microtubule-binding endosomes accumulate on the side of the cell that has fewer microtubules, and explains how a small difference in microtubule density can translate into a large bias in endosome segregation. The authors confirmed that the equation works over a wide range of induced levels of central-spindle asymmetry, and defined each of its parameters experimentally. Putting this equation to the test, Derivery and colleagues placed a Patronin-sequestering molecule on the pIIb side of a SOP, on a structuralprotein network called the cortex that lines the interior of the membrane. This molecular trap draws Patronin from the pIIb side of the central spindle, making pIIb more susceptible to microtubule disassembly by Klp10A. The authors found that the microtubule network became denser in pIIa, and that most Sara endosomes were delivered to pIIb, as predicted by the equation. And, consistent with their hypothesis, placing the Patronin trap on the pIIa cortex affected neither microtubule asymmetry nor the location of Sara endosomes. This study is filled with in vitro assays, mutant analyses, purpose-made computer code and more. Moreover, it is an excellent example of the power of using solidly quantified data to solve complex problems in molecular cell biology, applying equations when intuition will not do. Given the ubiquitous nature of its components, the mechanism identified by Derivery and colleagues might not be restricted to Sara endosomes in dividing SOPs. It could be applicable to any process in which a molecular cargo is moved unidirectionally by a motor protein along microtubules that are arranged as an asymmetric bundle. Moreover, the evolutionary conservation of the proteins involved points to the possibility that this mechanism operates in other species. As the authors point out, one case that merits close examination is that of the branched projections of neurons that receive stimuli from other neurons. These projections contain many antiparallel microtubules, and vesicle transport plays a fundamental part in their function. Future research will determine whether the authors’ mechanism can be generalized to this, or to any other process, in mammalian cells. But, for now, we can say that fruit flies have once again delivered insight into a basic biological mechanism that may well be applicable to other organisms. ■ Cayetano Gonzalez is at the Institute for Research in Biomedicine (IRB Barcelona), the Barcelona Institute of Science and Technology, 08028 Barcelona, Spain; and at the Catalan Institution for Research and Advanced
Studies (ICREA), Barcelona. e-mail: [email protected] 1. Knoblich, J. A. Nature Rev. Mol. Cell Biol. 2, 11–20 (2001). 2. Coumailleau, F., Fürthauer, M., Knoblich, J. A. &
González-Gaitán, M. Nature 458, 1051–1055 (2009). 3. Derivery, E. et al. Nature 528, 280–285 (2015). 4. Hoepfner, S. et al. Cell 121, 437–450 (2005). 5. Goodwin, S. S. & Vale, R. D. Cell 143, 263–274 (2010).
CATA LYSI S
The complexity of intimacy Catalysts that contain two types of active site split long hydrocarbon molecules into more-useful shorter ones. Research into controlling the nanoscale separation of the sites challenges accepted design rules for such catalysts. See Letter p.245 ROGER GLÄSER
T
he petroleum industry produces jet fuel and diesel using a catalytic process called hydrocracking, in which longchain hydrocarbons are broken into shorter, more-useful ones. More than 250 million tonnes of hydrocarbons are hydrocracked each year1 (Fig. 1). The catalysts comprise an acidic microporous silicate called a zeolite and a noble metal, mixed together with a binder material1. On page 245 of this issue, Zečević et al.2 report a major breakthrough in the design of hydrocracking catalysts, which greatly improves the selectivity of the process so that more of the desired products are formed. The results shatter the belief that the noble-metal sites and the acid sites must be as close together as possible for effective catalysis. The distance between the acid sites and the noble-metal sites of hydrocracking catalysts must be below a maximum limit defined by the
‘intimacy criterion’, which was first reported3 in 1962. The metal sites catalyse dehydrogenation reactions that convert the alkane reactants into unsaturated hydrocarbons called alkenes, which must then diffuse to an acid site to undergo isomerization and cracking (cleavage into shorter molecules; Fig. 2). If the metal sites are too far from the acid sites, then catalytic activity decreases — satisfying the intimacy criterion therefore overcomes diffusive limitations of the catalytic process. Previous studies of the intimacy criterion have examined inter-site distances only at or above the micrometre scale. Zečević et al. now report two catalysts in which platinum was controllably placed at nanometre-scale distances from the acid sites. The catalysts contain platinum particles of around 3 nanometres in diameter, with a narrow particle-size distribution, so that the metal was selectively supported on either the zeolite or the binder (γ-alumina); the overall platinum content in
Figure 1 | A hydrocracking facility in Gdansk, Poland. 1 0 D E C E M B E R 2 0 1 5 | VO L 5 2 8 | NAT U R E | 1 9 7
© 2015 Macmillan Publishers Limited. All rights reserved
