1. CB201_12 Tim Mitchison Lecture 3 Force generation by polymerization dynamics Nucleation: controlling where and when polymers form

2. Force generation by the cytoskeleton One of the main functions of the actin and microtubule cytoskeletons, and their prokaryotic counterparts, is to generate force for cell motility in a spatially and temporally controlled manner

3. Force generation by the cytoskeleton One of the main functions of the actin and microtubule cytoskeletons, and their prokaryotic counterparts, is to generate force for cell motility in a spatially and temporally controlled manner Force from polymerization dynamics Eukaryotes and prokaryotes

4. Force generation by the cytoskeleton One of the main functions of the actin and microtubule cytoskeletons, and their prokaryotic counterparts, is to generate force for cell motility in a spatially and temporally controlled manner Force from polymerization dynamics Eukaryotes and prokaryotes ATPase motor proteins Only Eukaryotes

5. Polymerization dynamics can perform mechanical work by pushing or pulling Pushing by polymerization Leading edge protrusion (actin) Listeria motility (actin) Plasmid separation in bacteria (ParM) Pulling by depolymerization Chromosome movement in mitosis (microtubules)

6. Mechanical work requires enery dissipation Mechanical work performed = force x distance Total energy dissipated =  G per elementary step x number of steps taken Efficiency = work done/energy dissipated In general, the efficiency of converting chemical energy into mechanical work must be less than 100% if the process that does the work is to proceed unidirectionally – ie some heat must be dissipated to make the process irreversible. This law of thermodynamics was developed for steam engines but applies equally to biology The efficiency of biological motors can be quite high. Food  human  rowing Total efficiency = ~ 20% Food  ATP efficiency = ~40% Therefore, effecience of ATP  mechanical work in muscle = ~50% (Wikipedia)

7. Elementary steps + Actin filaments grow by ~2nm per subunit (Actin monomer is ~4nm long, filament has 2 strands) Kinesin moves 8nm per step

8. Elementary steps + Actin filaments grow by ~2nm per subunit Kinesin moves 8nm per step Each step is coupled to hydrolysis of 1 molecule of ATP to ADP + Pi This liberates ~8-12 kilocal per mol (= ~20kT per molecule) Bolzman constant ~4pN.nm

9. Elementary steps Kinesin moves 8nm per step Each step is coupled to hydrolysis of 1 molecule of ATP to ADP + Pi This liberates ~8-12 kilocal per mol (= ~20kT per molecule) Efficiency = 5pN.8nm/20kT = ~50% Forcedistance Chemical energy dissipated

10. Elementary steps + Actin filaments grow by ~2nm per subunit (4nm subunit, 2 stranded polymer) Kinesin moves 8nm per step Each step is coupled to hydrolysis of 1 molecule of ATP to ADP + Pi This liberates ~8-12 kilocal per mol (= ~20kT per molecule) How do we think about force generation from polymerization or depolymerization?

11. Microtubule polymerizing in a microfabricated box. The force from polymerization causes the microtubule to buckle. Polymerization slows as the force on the ends increases. Eventually a catastrophe occurs. M. Dogterom and coworkers Science 278:856(1997), J Cell Biol 161:1029(2003)

12. Microtubule polymerizing in a microfabricated box. The force from polymerization causes the microtubule to buckle. Polymerization slows as the force on the ends increases. Eventually a catastrophe occurs. M. Dogterom and coworkers Science 278:856(1997), J Cell Biol 161:1029(2003) How much force? Simple argument for maximum possible force: For every tubulin added, the microtubules grows 8/13nm Suppose the full energy of GTP hydrolysis is used to promote this reaction GTP -> GDP +  G = ~ -50 kJ/mol = 5x10 -4 /6x10 -23 J/microtubule Force = work/distance = ~ 10 -19 /0.5x10 -9 = ~2x10 -10 N = ~200pN

13. Microtubule polymerizing in a microfabricated box. The force from polymerization causes the microtubule to buckle. Polymerization slows as the force on the ends increases. Eventually a catastrophe occurs. M. Dogterom and coworkers Science 278:856(1997), J Cell Biol 161:1029(2003) How much force? Simple argument for maximum possible force: For every tubulin added, the microtubules grows 8/13nm Suppose the full energy of GTP hydrolysis is used to promote this reaction GTP -> GDP +  G = ~ -50 kJ/mol = 5x10 -4 /6x10 -23 J/microtubule Force = work/distance = ~ 10 -19 /0.5x10 -9 = ~2x10 -10 N = ~200pN Force can be estimated since we know the bending ridigity of the microtubule, and can thus estimate the force required to buckle it Measured force ~5pN per microtubule (similar to the force exterted by a single motor molecule) Not as efficient as a motor protein, but still substantial force on the molecular scale

14. Actin polymerization force pushes the front of motile cells forward

15. How do cells control where and when cytoskeleton polymers accumulate? Bacterium Neutrophil Chemotaxis Phagocytosis High density of actin filaments

16. Neutrophil chasing S aureus in a drop of blood David Rogers 1950s

17. How might cells control where and when cytoskeleton polymers accumulate? Neutrophil detects a bacterium seconds Signal (bacterial cell wall) Receptor in plasma membrane Signaling pathway Cytoskeleton reorganization

18. How might cells control where and when cytoskeleton polymers accumulate? Neutrophil detects a bacterium seconds Signal (bacterial cell wall) Receptor in plasma membrane Signaling pathway Cytoskeleton reorganization What kind of processes might work for this at the level of cytoskeleton filaments?

19. Many proteins binds to cytoskeleton filaments and control their behavior in cells Bundling Cross- linking Capping Gel-forming Depolymerizing, Severing Nucleating Moving Monomer binding, Monomer sequestering

20. Many proteins binds to cytoskeleton filaments and control their behavior in cells Bundling Cross- linking Capping Gel-forming Depolymerizing, Severing Nucleating Moving Monomer binding, Monomer sequestering

21. Nucleation is slow, elongation is fast Nucleating a new filament is slow. Each incoming subunit makes only a subset of the favorable bonds Elongating an existing filament is fast. Each incoming subunit makes all favorable bonds The observation that elongating an existing filament is (much) faster than starting a new one is termed the kinetic barrier to nucleation. The physical chemistry of polymer nucleation is similar to crystallization from a saturated solution or freezing of a supercooled liquid. In each case self- assembly can be nucleated by a pre-existing fragment of the polymer/crystal

22. Origin of the kinetic barrier to nucleation. 1) Condensation models (Oosawa-type models) Break one bond. Fast Break 2 bonds. Fast Break 3 bonds. Slow Diffusion controlled Break 3 bonds. Slow “minimal seed” with n subunits

23. Origin of the kinetic barrier to nucleation. 1) Condensation models (Oosawa-type models) - Requires multi-stranded polymer - Does not require conformational change of monomer (similar models work for crystallization) - Elongation rate is proportional to the concentration of the subunit. - Nucleation rate depends on concentration of subunit by a power law. Break one bond. Fast Break 2 bonds. Fast Break 3 bonds. Slow Diffusion controlled Break 3 bonds. Slow “minimal seed” with n subunits

24. Origin of the kinetic barrier to nucleation. 1) Condensation models (Oosawa-type models) Break one bond. Fast Break 2 bonds. Fast Break 3 bonds. Slow Diffusion controlled “minimal seed” with n subunits Break 3 bonds. Slow Assume rapid equilibrium Rate of formation of new filaments = concentration of ( n - 1)mers x rate that they turn into filaments n-1 monomers ( n - 1)mer Assume rapid equilibrium up until minimal seed. Then: [( n - 1)mer] ~ K d [monomer] n-1 ; nucleation rate ~ K d [monomer] n-1 x k[monomer] ~ K’[monomer] n N = 3-4 for actin Tobacman LS, Korn ED. J Biol Chem. 1983 258:3207-14.

25. Origin of the kinetic barrier to nucleation. 2) Conformational switch models Non-polymerizing conformation (normal form of subunit after folding) Polymerizing conformation (rare form of subunit) Seed catalyzes conformational change Slow, spontaneous conformational change + +

26. Origin of the kinetic barrier to nucleation. 2) Conformational switch models - Does not requires multi-stranded polymer (in principle) - Requires conformational change of monomer that is catalyzed by polymer - Nucleation rate is independent of elongation rate and can be very slow. Caspar DL, Namba K. (1990) Adv Biophys. 26:157-85; DePace et al 1998 Cell. 93:1241-52 More relevant to viral coat proteins and amyloid fibers Non-polymerizing conformation (normal form of subunit after folding) Polymerizing conformation (rare form of subunit) Seed catalyzes conformational change Slow, spontaneous conformational change + +

27. Nucleation factors in the cell The kinetic barrier to nucleation prevents polymerization of cytoskeleton subunits at random in the cell. The cell controls where polymers form using nucleating factors. Centrosome. Contains microtubule nucleating factor  -tubulin ring complex + + + + + + +

28. Nucleation factors in the cell The kinetic barrier to nucleation prevents polymerization of cytoskeleton subunits at random in the cell. The cell controls where polymers form using nucleating factors. Leading edge. Contains actin Nucleation + branching factor Arp2/3 complex Centrosome. Contains microtubule nucleating factor  -tubulin ring complex + + + + + + + These nucleating factors have the same fold as the filament subunit, suggesting a mechanism (templating) and an evolutionary origin. We now know other actin nucleating factors that are quite different in structure.

29. Evidence that centrosomes contain microtubule nucleating factors (cells imaged by fixation and immunofluorescence) Add nocodazole to depolymerize microtubules Wash out drug 5 min 20 min Brinkley BR.(1985). Annu Rev Cell Biol. 1:145-72.

30. Evidence that centrosomes contain microtubule nucleating factors (cells imaged by fixation and immunofluorescence) Add nocodazole to depolymerize microtubules Wash out drug 5 min 20 min Permeablize cells with non-ionic detergent Add tubulin, GTP Incubate at 37 o Brinkley BR.(1985). Annu Rev Cell Biol. 1:145-72.

31. Microtubule Organizing Centers (MTOCs): Centrosomes, centrioles, basal bodies (animals) and spindle pole bodies (fungi) Centrosome = Centriole + Peri-centriolar material (PCM) Centrioles PCM (fibrous)  -tubulin ring complex (nucleates MTs)

32. Discovery of  -tubulin Aspergillus (a mycelium forming fungus)  -tubulin mutant Select revertants  -tubulin,  -tubulin double mutant Defects in mitosis, nuclear transport Oakley and Oakley 1989. Nature 338:662-4.

33. Discovery of  -tubulin Aspergillus (a mycelium forming fungus)  -tubulin mutant Select revertants  -tubulin,  -tubulin double mutant  -tubulin knockout: no microtubules  -tubulin localizes to spindle pole bodies by immunofluorescence Defects in mitosis, nuclear transport Oakley and Oakley 1989. Nature 338:662-4.

34. Centrosomes, centrioles, basal bodies and spindle pole bodies Yeast spindle pole body forms on the nuclear envelope Wigge et al 1998 J Cell Biol. 141:967-77 Centrosome = Centriole + Peri-centriolar material (PCM) Animals Fungi

35.  -tubulin ring complex: the template model Agard 2001 Curr Opin Struct Biol.11:174-81 Agard 2011 Nat Rev Cell Mol Biol.12:709 Note  -tubulin has the same fold as tubulin, and the ring complex mimics a plus end

36. Actin nucleating complexes Arp2/3 complex Nucleates from the pointed (slow growing) end Nucleates from the side of a pre-existing filament Generates brnached networks Lammellipodia, Listeria comet tails, Endocytosis Formins Nucleate from the barbed (fast growing) end Remain at the growing end Generate long bundles Yeast actin cables, filopodia? Formin dimer

37. A pathogen provides a model for motility driven by actin polymerization - Listeria monocytogenes is a gram positive bacterium that infects us from contaminated food - Enters the cytoplasm of many cell types by breaking out of phagosomes - Nucleates actin filaments and forms a comet tail that propels it through the cytoplasm and into neighboring cells - Other pathogens (Shigella, pox virus) also move using actin comet tails “comet tail” of actin filaments Tilney and Portnoy (1989) J Cell Biol. 109:1597-608.

38. Listeria moving in cultured cell Julie Theriot ~1992 Phase contrast

39. Listeria provides a system for dissecting the molecular mechanisms underlying leading edge motility Identification of arp2/3 complex Listeria moving in cell extract fractionate cell extract by chromatography Purify a protein complex that nucleates actin polymerization on the Listeria surface Welch et al.(1997) Nature. 385:265-9

40. Listeria provides a system for dissecting the molecular mechanisms underlying leading edge motility Identification of arp2/3 complex Listeria moving in cell extract fractionate cell extract by chromatography Purify a protein complex that nucleates actin polymerization on the Listeria surface Listeria movement was later reconstituted using 7 proteins: Actin Arp2/3 complex (7 polypeptides) Profilin Cofilin Capping protein VASP + ActA on the bacterium surface Loisel et al.(1999). Nature. 401:613-6 Welch et al.(1997) Nature. 385:265-9

41. Arp2/3 structure Arp2 and Arp3 subunits have the same fold as actin

42. Arp2/3 in action Rhodamine actin TIRF microscopy Pollard and Kovar

43. Arp2/3 mechanism ActA, WASP etc. To nucleate, Arp2/3 must: 1)bind to the side of a pre- existing filament 2)recruiting an activating protein. The activating protein brings in the first subunit of the new polymer Arp2/3.

44. Arp2/3 mechanism This mechanism generates dendritic actin assemblies, as seen in the leading edge of motile cells by EM Pollard TD, Borisy GG. (2003) Cell. 112:453-65. ActA, WASP etc. To nucleate, Arp2/3 must: 1)bind to the side of a pre- existing filament 2)recruiting an activating protein. The activating protein brings in the first subunit of the new polymer Arp2/3.

45. How might cells control where and when cytoskeleton polymers accumulate? Neutrophil detects a bacterium seconds David Rogers 1950s

46. Activating proteins make Arp2/3 activity dependent on multiple inputs NWASP is activated by: Cdc42.GTP Phosphoinositol lipids Tyrosine phosphorylation WAVE is activated by: Rac.GTP Phosphoinositol lipids In both cases the WASP homolog acts as an AND gate for multiple biochemical signals These signals make Arp2/3 nucleation dependent on multiple signaling pathway inputs at the plasma membrane

47. fMLP GPCR G-protein coupled receptor Different GPCRs for different signals GDP Heterotrimeric GTPase (inactive GDP bound state) Leukocyte chemotactic signals are usually detected by GPCRs GTP GG G  Signals to the actin cytoskeleton Bacteria Human cells (eg leukocytes) Leukotriene B4 Chemokine – eg CCL2 etc Etc.

48. fMLP GDP Chemotactic receptors send multiple signals to the actin cytoskeleton GTP GG G  Actin polymerization at the leading edge WAVE Arp2/3 Myosin-II driven Contraction at the rear of the cell Rac

49. The actin cytoskeleton is polarized in motile cells Actin Myosin-II in a fibroblast cellActin RhoA in neutrophils

50. How does a neutrophil polarize? How are the multiple signaling outputs from chemotactic receptors spatially organized to promote polarization? Do different signals diffuse away from the receptor to different extents? Does the front of the cell inhibit the back (or vice versa) – and if so by chemical signals, or physical signals such as membrane tension? ? ?