17.03.02 · mol-cell-bio / cell-organization

Cytoskeleton and contractile proteins

draft3 tiersLean: nonepending prereqs

Anchor (Master): Alberts et al., MBoC 7e; Howard, Mechanics of Motor Proteins and the Cytoskeleton (2001); Bray, Cell Movements, 2nd ed. (2001)

Intuition [Beginner]

A cell is not a bag of dissolved chemicals. It has an internal skeleton -- the cytoskeleton -- made of three types of protein filaments: actin microfilaments, microtubules, and intermediate filaments. These filaments give the cell its shape, enable it to move, position its organelles, and divide when it reproduces.

Actin microfilaments are the thinnest filaments (7 nm diameter), made of the protein actin. They form a dense mesh just beneath the plasma membrane (the cell cortex), providing mechanical support. When a cell crawls across a surface, actin polymerises at the leading edge, pushing the membrane forward. Actin also forms the contractile ring that pinches a dividing cell in two during cytokinesis.

Microtubules are hollow tubes (25 nm diameter) made of the protein tubulin. They radiate outward from a structure near the nucleus called the centrosome, forming a highway system along which motor proteins carry cargo. Two motor proteins walk on microtubules: kinesin moves toward the plus end (outward, away from the centrosome), and dynein moves toward the minus end (inward, toward the centrosome). During cell division, microtubules form the mitotic spindle that separates chromosomes.

Intermediate filaments (10 nm diameter) are the toughest filaments, providing mechanical strength. Unlike actin and microtubules, which are dynamic and constantly assembling and disassembling, intermediate filaments are more permanent structures. Keratin in epithelial cells, vimentin in fibroblasts, and nuclear lamins (lining the inside of the nuclear envelope) are examples.

Myosin is the motor protein that walks on actin filaments. In muscle cells, myosin II forms thick filaments that pull on actin thin filaments, generating the force of contraction. In non-muscle cells, various myosin types transport cargo along actin filaments or generate contractile forces for cell migration.

All three filament systems convert chemical energy (from ATP hydrolysis) into mechanical work. A single kinesin molecule can walk along a microtubule for hundreds of steps without falling off, carrying a vesicle from the cell centre to the periphery. This processive motion is made possible by the motor's two "feet" (motor domains): at least one foot is always bound to the microtubule.

Visual [Beginner]

The dynamic nature of cytoskeletal filaments is crucial. Actin and microtubules are polar: they have a fast-growing plus end and a slow-growing minus end. Polymerisation at the plus end and depolymerisation at the minus end can occur simultaneously, a phenomenon called treadmilling. This dynamic instability allows the cell to rapidly reorganise its cytoskeleton in response to signals.

Worked example [Beginner]

Kinesin-1 walks along a microtubule at approximately 800 nm/s. Each step moves the motor 8 nm (the distance between tubulin dimers along a single protofilament), and each step requires one ATP molecule.

Step 1. Calculate the stepping rate: $800 nm/s /8 nm/step = 100 steps/s$ . Kinesin takes 100 steps per second.

Step 2. Calculate the ATP consumption rate: one ATP per step = 100 ATP molecules per second per kinesin motor.

Step 3. Calculate the time to traverse a typical cell. A mammalian cell is roughly 20 $μ$ m = 20,000 nm in diameter. Time = 20,000 nm / 800 nm/s = 25 seconds. A single kinesin can transport a vesicle from the centrosome to the cell periphery in about 25 seconds.

Step 4. Force generation. A single kinesin motor generates a stall force of approximately 6-7 pN (piconewtons). To move against a viscous drag force in the cytoplasm, the motor must overcome a force of roughly $F = 6 π η r v$ (Stokes drag), where $η \approx 0.001 Pa s$ (cytoplasm viscosity, roughly that of water), $r \approx 50 nm$ (vesicle radius), and $v = 800 nm/s$ . This gives $F \approx 6 π \times 0.001 \times 50 \times 1 0^{- 9} \times 800 \times 1 0^{- 9} \approx 0.75 \times 1 0^{- 12} N = 0.75 pN$ . The motor's stall force of 6-7 pN is well above this drag, so kinesin can easily transport a vesicle through the cytoplasm.

What this tells us: the cytoskeleton is a nanoscale transport and force-generation system where individual protein motors convert chemical energy (ATP) into directed mechanical motion with high efficiency, and the forces they generate are well-matched to the viscous drag environment of the cell.

Check your understanding [Beginner]

Exercise (easy, multiple choice).

Which cytoskeletal filament is responsible for forming the mitotic spindle that separates chromosomes during cell division?

A. Actin microfilaments B. Intermediate filaments C. Microtubules D. All three contribute equally

Hint

The mitotic spindle is a bipolar structure that captures and pulls chromosomes. Which filament system radiates from the centrosome?

Answer

Option C. Microtubules form the mitotic spindle. The spindle has two poles (centrosomes) from which microtubules extend. Kinetochore microtubules attach to chromosomes at the kinetochore, polar microtubules overlap at the spindle midzone, and astral microtubules anchor the spindle to the cell cortex. Dynein and kinesin motor proteins generate the forces that align and separate the chromosomes.

Exercise (medium, short answer).

Explain the phenomenon of actin treadmilling. Why does it occur, and how does it contribute to cell motility?

Hint

Actin subunits (ATP-bound) add at one end while ADP-bound subunits fall off the other end. The critical concentration differs at the two ends.

Answer

Actin treadmilling occurs because the critical concentration for polymerisation (the monomer concentration at which assembly and disassembly are balanced) differs at the two ends of the filament. The plus (barbed) end has a lower critical concentration (~~0.1 $μ$ M) than the minus (pointed) end (~~0.6 $μ$ M). At intermediate monomer concentrations (0.1-0.6 $μ$ M), subunits add at the plus end while dissociating from the minus end, producing net flux of subunits through the filament without changing its length.

In cell motility, actin polymerises at the leading edge (plus end facing the membrane), pushing the membrane forward. Older subunits depolymerise at the minus end toward the cell interior. This creates a treadmill that drives lamellipodial protrusion, allowing the cell to crawl. The Arp2/3 complex nucleates new actin branches at the leading edge, amplifying the protrusive force.

Formal definition [Intermediate+]

Actin dynamics

G-actin (globular actin, 42 kDa) polymerises into F-actin (filamentous actin), a right-handed helix with 13 subunits per 6 turns, rising 2.75 nm per subunit and 36 nm per helical repeat. The polymerisation kinetics at each end follow:

\frac{d N}{d t} = k_{on} [G -actin] - k_{off}

where $k_{on}$ and $k_{off}$ differ for the plus and minus ends. At steady-state treadmilling, the flux of subunits through the filament is:

J_{treadmill} = k_{on}^{+} [G] - k_{off}^{+} = k_{off}^{-} - k_{on}^{-} [G]

The rate of treadmilling in vivo is enhanced by actin-binding proteins: profilin (replenishes ATP-G-actin), ADF/cofilin (accelerates depolymerisation at the minus end), and thymosin beta-4 (buffers the G-actin pool).

Microtubule dynamics

Microtubules are hollow cylinders of 13 protofilaments, each a linear polymer of alpha-beta tubulin heterodimers (8 nm per dimer). They exhibit dynamic instability: stochastic switching between a growing phase (polymerisation at 0.5-2 $μ$ m/min) and a shrinking phase (depolymerisation at 5-20 $μ$ m/min). The switch from growth to shrinkage is called catastrophe; the reverse is rescue.

Dynamic instability is driven by GTP hydrolysis. Beta-tubulin binds GTP, and the GTP-bound dimer adds to the growing end. After incorporation, GTP is hydrolysed to GDP. A cap of GTP-bound tubulin at the plus end stabilises the microtubule. If the cap is lost (all terminal subunits are GDP-bound), the filament rapidly depolymerises. The critical concentration for assembly depends on GTP-tubulin concentration.

Motor protein mechanochemistry

Motor proteins convert ATP hydrolysis into directed motion along a filament. The stepping kinetics of kinesin-1 follow a tightly coupled cycle:

ATP binding to the leading head triggers neck-linker docking (a conformational change that swings the trailing head forward).
ATP hydrolysis occurs after the new leading head binds to the next tubulin binding site.
Phosphate release from the trailing head weakens its affinity for the microtubule.
ADP release from the new leading head strengthens its binding.

The result is a hand-over-hand walk: each step advances the motor by 8 nm (one tubulin dimer) and consumes one ATP. The processivity (average number of steps before dissociation) of kinesin-1 is ~100 steps, corresponding to a run length of ~800 nm.

The force-velocity relationship for kinesin is approximately linear:

v (F) = v_{m a x} (1 - \frac{F}{F _{stall}})

where $v_{m a x} \approx 800 nm/s$ and $F_{stall} \approx 6 - 7 pN$ . At stall force, the motor holds position but makes no net progress.

Intermediate filaments

Intermediate filaments assemble from coiled-coil dimers that form tetramers, which assemble end-to-end and laterally into 10 nm rope-like structures. Unlike actin and microtubules, they are not polar (no directionality) and have no associated motor proteins. Their primary function is mechanical: they resist tensile stress. The stress-strain relationship of a single intermediate filament shows strain hardening (stiffening under stretch), a property shared with many synthetic polymers.

Counterexamples to common slips

Microtubules are static scaffolds. Dynamic instability means individual microtubules switch between growth and rapid shrinkage every few minutes. The entire microtubule array in a cell turns over in 5-10 minutes.
Intermediate filaments are purely structural. Vimentin organises the Raf/MEK/ERK signaling cascade by scaffolding Raf kinase; lamin A/C regulates YAP/TAZ transcription factor localisation in mechanosensing.
Myosin only functions in muscle. The myosin superfamily has 40+ classes. Myosin V transports vesicles along actin, myosin VI moves toward the minus end (unique among myosins), and myosin I links membranes to actin networks.

Key theorem with proof [Intermediate+]

Theorem (Kinesin step size and ATP coupling). The step size of kinesin-1 along a microtubule is exactly one tubulin dimer (8 nm), and each step consumes exactly one ATP molecule. The mechanochemical coupling ratio is therefore 1 ATP per 8 nm per step.

Proof (by structural and biophysical analysis).

Structural argument: Kinesin's two motor domains (heads) bind to sequential binding sites along a single microtubule protofilament. The distance between adjacent binding sites on one protofilament is 8 nm (the length of one alpha-beta tubulin heterodimer). Cryo-EM structures of kinesin-microtubule complexes show the two heads separated by exactly 8 nm, with each head contacting a beta-tubulin subunit. This geometric constraint fixes the step size.

Biochemical argument: The ATPase rate of kinesin, measured in solution with microtubules, is approximately 100 ATP/s at saturating ATP concentration. The velocity, measured by single-molecule fluorescence microscopy, is approximately 800 nm/s at the same conditions. The ratio:

coupling ratio = \frac{v}{d \times k _{ATP}} = \frac{800 nm/s}{8 nm/step \times 100 ATP/s} = 1.0 ATP/step

This tight coupling (1.0 ATP per step) was confirmed by single-molecule optical-trap experiments that simultaneously measured displacement and nucleotide turnover. Deviations from 1:1 coupling occur only under high load, where ATP consumption without forward stepping can occur. Under physiological loads, the coupling ratio is $1.00 \pm 0.05$ . $□$

Worked example: force generation by myosin II in muscle

A single myosin II motor domain generates a working stroke of approximately 5-10 nm and a force of ~3-5 pN. In a skeletal muscle sarcomere:

Thick filaments contain ~300 myosin molecules each
Half of these (~150) face each half-sarcomere
A single myofibril has ~2000 sarcomeres in series
A muscle fibre has ~5000 myofibrils in parallel

Total cross-bridges active at any moment (assuming 50% duty ratio): $150 \times 2000 \times 5000 \times 0.5 = 7.5 \times 1 0^{8}$ myosin heads.

Total force (at 3 pN per head): $7.5 \times 1 0^{8} \times 3 \times 1 0^{- 12} N = 2.25 \times 1 0^{- 3} N = 2.25 mN$ per fibre.

This matches experimental measurements of single muscle fibre force (~2-4 mN), confirming that muscle force is the sum of many individual molecular motors working in parallel.

Bridge. The tight 1:1 mechanochemical coupling of kinesin builds toward the actin-myosin contractile machinery examined in 18.04.02 pending, where myosin II working strokes summate across sarcomeres to produce macroscopic force. The foundational reason the coupling ratio is exactly 1.0 is that each ATP hydrolysis event corresponds to one complete conformational cycle of the motor domain — this is exactly the mechanism that appears again in 17.08.01 during cytokinesis, where myosin II in the contractile ring generates the pinching force that divides the cell. The bridge is between the single-molecule biophysics of motor stepping and the collective force generation that drives cell division and muscle contraction, putting these together with the cytoskeletal filament polarity that determines motor directionality.

Exercises [Intermediate+]

Exercise 2 (easy, short answer).

Explain why microtubules exhibit dynamic instability while actin filaments do not (actin treadmills instead). What is the molecular basis for this difference?

Hint

Both use nucleotide hydrolysis (GTP for microtubules, ATP for actin). The key difference is whether the nucleotide state of the entire filament end (not just individual subunits) determines stability.

Answer

Dynamic instability in microtubules depends on the GTP cap: a region of GTP-bound tubulin at the plus end that stabilises the entire filament. If the cap is lost (all end subunits are GDP-bound), the entire filament rapidly depolymerises. This all-or-nothing behaviour produces the characteristic switch between growth and rapid shrinkage.

Actin filaments, by contrast, have ATP-actin at the plus end, but the hydrolysis of ATP to ADP occurs progressively along the filament without creating a catastrophic switch. Individual ADP-actin subunits at the minus end dissociate one at a time, producing smooth treadmilling rather than stochastic growth-collapse cycles. The structural basis is that GDP-tubulin has a curved conformation that destabilises lateral contacts between protofilaments, causing the entire end to peel apart ("ram's horn" structure). ADP-actin does not undergo this dramatic conformational change.

Exercise 3 (medium, numeric).

Taxol (paclitaxel) is a chemotherapy drug that stabilises microtubules by binding along their length, preventing depolymerisation. If a cell has microtubules with an average half-life of 5 minutes in the absence of taxol, and taxol reduces the catastrophe frequency by a factor of 10, calculate the new average microtubule lifetime. Assume the original lifetime is determined by the average time between catastrophes plus the regrowth time, and simplify by assuming the catastrophe rate is the rate-limiting step.

Hint

If the half-life is 5 minutes and catastrophe is rate-limiting, the catastrophe rate is $λ = ln (2) / t_{1/2}$ .

Answer

If catastrophe is rate-limiting, the lifetime is approximately $1/ λ$ where $λ$ is the catastrophe rate. With half-life $t_{1/2} = 5 min$ : $λ \approx ln (2) /5 \approx 0.139 min^{- 1}$ , giving average lifetime $\approx 1/ λ \approx 7.2 min$ .

Taxol reduces catastrophe frequency by 10x: $λ_{taxol} = 0.0139 min^{- 1}$ . New average lifetime: $1/0.0139 \approx 72 min$ .

In practice, taxol-stabilised microtubules can persist for hours, effectively freezing the mitotic spindle. This is why taxol kills dividing cells: it prevents the dynamic microtubule rearrangements required for chromosome segregation, activating the spindle assembly checkpoint and triggering apoptosis.

Exercise 4 (medium, numeric).

A kinesin motor is transporting a vesicle at 600 nm/s under a load force of 3 pN. Using the linear force-velocity relation with $v_{m a x} = 800 nm/s$ and $F_{stall} = 7 pN$ , predict the velocity at a load of 5 pN. Will the motor reach its destination 5 $μ$ m away under a 5 pN load before it dissociates (average run length ~800 nm)?

Hint

Use $v (F) = v_{m a x} (1 - F / F_{stall})$ . For the run-length question, compare the required distance to the average run length.

Answer

Check at $F = 3 pN$ : $v = 800 (1 - 3/7) = 800 \times 0.571 = 457 nm/s$ . The observed 600 nm/s suggests either a lower load or a slightly different force-velocity curve.

At $F = 5 pN$ : $v = 800 (1 - 5/7) = 800 \times 0.286 = 229 nm/s$ .

The average run length is ~800 nm, but the required distance is 5000 nm. The probability of a single motor reaching 5000 nm without dissociating is approximately $e^{- 5000/800} = e^{- 6.25} \approx 0.002$ . A single motor almost certainly will not make it. In practice, multiple kinesin motors cooperate on a single vesicle, dramatically increasing the effective run length.

Exercise 5 (medium, short answer).

Epilepsy-associated mutations in the gene encoding the microtubule plus-end binding protein doublecortin (DCX) cause a brain malformation called lissencephaly ("smooth brain"). Explain the connection between cytoskeletal function and brain folding.

Hint

Brain folding (gyrification) depends on the expansion of the cortical plate during development, which in turn depends on neuronal migration. How do neurons migrate?

Answer

During brain development, newborn neurons migrate from the ventricular zone to the cortical plate along radial glial fibres, which are bundles of microtubules and intermediate filaments. This migration depends on microtubule dynamics in the leading process and dynein-mediated nuclear translocation.

Doublecortin stabilises microtubules in the growing neuronal process and links microtubules to the actin cytoskeleton. Mutations in DCX disrupt microtubule stability, impairing neuronal migration. Neurons fail to reach their correct cortical position, producing a disorganized and underdeveloped cortex that lacks the normal folds (gyri) and sulci. The resulting "smooth brain" (lissencephaly) causes severe intellectual disability and epilepsy.

Exercise 6 (hard, short answer).

The drug cytochalasin D binds to the barbed (plus) end of actin filaments and blocks further polymerisation without promoting depolymerisation. Predict the effect on: (a) cell motility, (b) cytokinesis, and (c) the actin cortex. Explain the molecular mechanism for each.

Hint

Blocking the plus end prevents new actin assembly. Consider which cellular processes depend on actin polymerisation at the plus end.

Answer

(a) Cell motility stops. Lamellipodial protrusion depends on rapid actin polymerisation at the leading edge (plus ends facing the membrane). Cytochalasin D caps these plus ends, preventing the formation of new actin filaments that push the membrane forward. Existing filaments depolymerise from their minus ends, and the leading edge retracts.

(b) Cytokinesis fails. The contractile ring that pinches the cell in two during cytokinesis is made of actin and myosin II. Its assembly requires new actin polymerisation at the division plane. Cytochalasin D prevents contractile ring formation, producing a binucleate cell (the nucleus divides but the cytoplasm does not).

(c) The actin cortex disassembles. The cell cortex is a dynamic meshwork that constantly remodels through balanced polymerisation and depolymerisation. Blocking new polymerisation while depolymerisation continues causes the cortex to thin and eventually disappear. The cell rounds up and loses its shape, becoming more sensitive to mechanical stress. Membrane blebbing occurs as the weakened cortex detaches locally from the membrane.

Exercise 7 (hard, numeric).

A single myosin V motor (a processive actin-based motor involved in vesicle transport) has a step size of 36 nm (the pseudo-repeat of the actin helix), a stall force of 3 pN, and an unloaded velocity of 300 nm/s. Calculate: (a) the stepping rate, (b) the ATP consumption rate (assuming 1 ATP per step), and (c) the mechanical power output at half-stall force.

Hint

Power = Force x Velocity. At half-stall force, use the linear force-velocity relation to find the velocity.

Answer

(a) Stepping rate: $300 nm/s /36 nm/step = 8.33 steps/s$ .

(b) ATP consumption: $8.33 ATP/s$ (one per step).

Power: $P = F \times v = 1.5 \times 1 0^{- 12} \times 150 \times 1 0^{- 9} = 2.25 \times 1 0^{- 19} W = 0.225 zW$ (zeptowatts).

In units of $k_{B} T$ /s at $310 K$ : $P / (k_{B} T) = 2.25 \times 1 0^{- 19} / (4.28 \times 1 0^{- 21}) \approx 53 k_{B} T / s$ . The free energy from one ATP hydrolysis is ~ $20 k_{B} T$ , so at 8.33 ATP/s the input power is ~ $167 k_{B} T$ /s, giving an efficiency of $53/167 \approx 32%$ , consistent with experimental measurements for myosin V.

Exercise 8 (hard, short answer).

In axonal transport, both kinesin and dynein are simultaneously attached to the same cargo vesicle, producing "tug-of-war" dynamics. Describe the experimental evidence for this bidirectional transport and explain how the cell regulates the net direction of movement.

Hint

Single-particle tracking in axons shows bidirectional movement with frequent directional switches. How might adaptor proteins or post-translational modifications bias the outcome?

Answer

Single-particle tracking of fluorescently labelled vesicles in squid axons (by Brady, Sheetz, and others) and in cultured mammalian neurons shows that individual vesicles undergo bidirectional movement: short runs in one direction, pauses, then runs in the opposite direction. This produces a biased random walk with net anterograde (outward) or retrograde (inward) transport depending on the cargo type.

The tug-of-war model proposes that kinesin and dynein on the same cargo engage in a stochastic competition. When more kinesins are active, the vesicle moves anterograde; when dyneins dominate, it moves retrograde. Regulation occurs through several mechanisms:

Adaptor proteins: JIP1, BICD2, and other scaffolds preferentially recruit one motor type over the other.
Post-translational modifications: Phosphorylation of dynein intermediate chain reduces its processivity, biasing transport toward kinesin.
Microtubule modification: Acetylated and detyrosinated microtubules (enriched in axons) are preferred tracks for kinesin; tyrosinated microtubules favour dynein.
Cargo load: The number of each motor type recruited depends on the vesicle's membrane composition and associated proteins.

Diseases like amyotrophic lateral sclerosis (ALS) show impaired bidirectional transport, suggesting that the tug-of-war balance is critical for neuronal health.

Advanced results [Master]

Actin dynamics and regulation

The actin cytoskeleton is the primary engine of cell motility and morphogenesis. Its capacity for rapid reorganisation depends on a large family of actin-binding proteins (ABPs) that control every aspect of filament assembly, disassembly, and three-dimensional architecture.

The Arp2/3 complex (seven-subunit protein complex containing actin-related proteins Arp2 and Arp3) is the principal nucleator of branched actin networks ^{[Mullins 1998]}. Arp2/3 binds to the side of an existing ("mother") filament and nucleates a new ("daughter") filament at a characteristic 70-degree angle, creating the dendritic branched meshwork that fills lamellipodia. Arp2/3 alone is a weak nucleator; it requires activation by WASP-family proteins (WASP in hematopoietic cells, WAVE/WASP family verprolin-homologous protein in most other cells). The signalling pathway runs: extracellular signal → Rac GTPase → WAVE complex → Arp2/3 activation. The result is spatially controlled nucleation at the leading edge of migrating cells. The dendritic nucleation model, proposed by Mullins, Heuser, and Pollard in 1998, quantitatively accounts for the assembly of the lamellipodial network: Arp2/3 nucleates branches at a rate proportional to the local filament density, capping protein terminates elongation within 1-2 seconds (limiting branch length to ~0.2 μm), and ADF/cofilin disassembles the network at the rear of the lamellipodium, recycling actin subunits for reuse.

Formins (diaphanous-related formins, DRFs) nucleate and processively elongate unbranched actin filaments ^{[Kovar 2006]}. The formin homology 2 (FH2) domain forms a dimeric ring that remains processively associated with the growing barbed end, allowing rapid addition of profilin-actin complexes while preventing capping protein from binding. Each FH2 domain steps along the growing end in a "stair-stepping" mechanism: one half of the dimer releases while the other remains bound, then the released half rebinds one subunit ahead. Formins build the long, parallel, unbranched bundles found in filopodia and stress fibers. The formin mDia1 elongates filaments at rates up to 100 subunits/s (0.27 μm/s), approximately ten times faster than spontaneous elongation at physiological actin concentrations. Formin activity is regulated by Rho GTPases through the autoinhibitory interaction between the N-terminal GTPase-binding domain and the C-terminal diaphanous-autoregulatory domain (DAD): binding of GTP-bound Rho releases autoinhibition and activates the formin.

ADF/cofilin (actin-depolymerising factor / cofilin) is the principal actin depolymerising factor ^{[Bamburg 1999]}. It binds cooperatively to ADP-F-actin (not ATP-F-actin), inducing a twist in the filament helix (from 167 degrees to 162 degrees per subunit) that weakens lateral contacts and promotes severing. Severing creates new ends: new barbed ends can serve as nucleation sites (if capped, the new plus end is a substrate for Arp2/3 branching), while new pointed ends depolymerise rapidly. The net effect is accelerated filament turnover: cofilin can increase the steady-state treadmilling rate by 20-25-fold compared to pure actin alone. In vivo, the actin turnover half-life in lamellipodia is approximately 20-30 seconds, far faster than the 30-minute half-life of pure actin filaments at the same concentration. Cofilin activity is regulated by LIM kinase (LIMK), which phosphorylates cofilin on Ser-3, blocking its actin-binding site. LIMK is itself activated by ROCK (Rho-associated kinase) and PAK (p21-activated kinase), linking cofilin regulation to Rho GTPase signalling.

Profilin and thymosin-β4 together buffer the cytoplasmic G-actin pool. Thymosin-β4 sequesters ATP-G-actin in a 1:1 complex that cannot polymerise, maintaining a reservoir of polymerisation-competent monomers. Profilin also binds ATP-G-actin but, unlike thymosin-β4, the profilin-actin complex can add to barbed ends (profilin cannot bind to the barbed end, so it dissociates upon incorporation). Profilin also catalyses nucleotide exchange on ADP-G-actin (replacing ADP with ATP), replenishing the ATP-G-actin pool produced by cofilin-mediated depolymerisation. The balance between thymosin-β4 sequestration and profilin delivery determines the effective concentration of polymerisation-competent actin at the barbed end.

The spatial regulation of actin assembly is orchestrated by the Rho family GTPases: Rho, Rac, and Cdc42 ^{[Ridley 2001]}. Rac activates the WAVE complex → Arp2/3 → branched lamellipodial networks. Cdc42 activates WASP → Arp2/3 → filopodial bundles. Rho activates formins (via mDia) and myosin II contractility (via ROCK → myosin light chain phosphorylation). These three pathways are mutually antagonistic: Rac and Cdc42 suppress Rho activity (and vice versa), creating bistable switches that define cell front (Rac/Cdc42 active, Rho off) versus cell rear (Rho active, Rac/Cdc42 off). The front-rear polarity required for directed cell migration emerges from this reciprocal inhibition plus positive feedback (Rac activates PAK, which further activates Rac through GEF recruitment).

Microtubule dynamics and motor proteins

The dynamic instability of microtubules was discovered by Mitchison and Kirschner in 1984 ^{[Mitchison & Kirschner 1984]}. Their observations, made by dark-field microscopy of microtubules assembled from purified tubulin, revealed that individual microtubules switch stochastically between prolonged phases of growth (at 0.5-2.0 μm/min, corresponding to 1-4 dimers/s) and rapid shrinkage (at 5-20 μm/min, 10-40 dimers/s). The transition from growth to shrinkage is termed catastrophe and occurs with a frequency of 0.2-0.6 min $^{- 1}$ in purified tubulin (lower in cells with MAPs). The reverse transition, rescue, occurs at 0.8-2.6 min $^{- 1}$ in vitro. The mean lifetime of a growing phase is 2-5 minutes; the mean lifetime of a shrinking phase is 0.2-0.5 minutes.

The GTP cap model explains dynamic instability at the molecular level. GTP-bound tubulin dimers add to the growing end. After incorporation, the β-tubulin GTP is hydrolysed with a half-time of approximately 2-5 seconds. If addition of new GTP-dimers outpaces hydrolysis, a cap of GTP-tubulin (estimated at 20-200 dimers, or 160-1600 nm) persists at the plus end and stabilises the lattice. If the growth rate slows (due to low tubulin concentration or resistance), the GTP cap shrinks. When the last GTP-dimer is hydrolysed, the GDP-tubulin lattice is exposed: GDP-tubulin adopts a curved conformation incompatible with the straight protofilament, and the protofilaments peel outward in a "ram's horn" configuration that rapidly propagates down the microtubule. The shrinking rate is 5-20 μm/min because the curved GDP-protofilaments splay apart cooperatively, with the strain energy of each hydrolysed dimer contributing to the unwinding.

Microtubule-associated proteins (MAPs) modulate dynamics in cells. Tau protein binds along the microtubule lattice, stabilising it by reducing the catastrophe frequency and increasing the rescue frequency. Hyperphosphorylation of tau (as occurs in Alzheimer disease) reduces its microtubule affinity, destabilising microtubules in axons and contributing to neurodegeneration. MAP2 is the dendritic MAP counterpart, organising microtubule bundles in dendrites. MAP4 is ubiquitously expressed and protects microtubules from catastrophe in interphase cells. XMAP215 (in Xenopus; ch-TOG in mammals) is a polymerase that accelerates growth by delivering tubulin dimers to the plus end through its TOG domains, increasing the growth rate by 5-10 fold. MCAK/Kif2C (kinesin-13 family) is a depolymerase that induces catastrophe by removing dimers from both ends, bending the protofilament to accelerate the ram's horn transition.

The kinesin superfamily (KIFs) encompasses 45 genes in mammals, organised into 14 families by sequence homology ^{[Vale 2003]}. Kinesin-1 (KIF5) is the canonical plus-end-directed transport motor: two N-terminal motor domains, a coiled-coil stalk for dimerisation, and a tail domain that binds cargo through adaptor proteins. Kinesin-2 (KIF3A/KIF3B heterodimer or KIF17 homodimer) drives anterograde intraflagellar transport (IFT) in cilia and flagella. Kinesin-5 (Eg5/KIF11) is a bipolar tetramer that slides antiparallel microtubules apart during spindle assembly — it has motor domains at both ends of the stalk, so it can simultaneously walk on two microtubules oriented in opposite directions. Kinesin-6 (MKLP1/KIF23) localises to the central spindle midzone during cytokinesis and is essential for midbody formation. Kinesin-7 (CENP-E/KIF10) is a kinetochore-associated motor that transports chromosomes along microtubules toward the plus end during congression. Kinesin-13 (MCAK/KIF2) is unusual in that it has motor domains in the middle of the polypeptide rather than at the N-terminus; it does not walk but instead diffuses to microtubule ends and catalyses depolymerisation.

Cytoplasmic dynein is the sole minus-end-directed microtubule motor in most cells ^{[Vallee 1988]}. Its structure is far more complex than kinesin: the motor domain is a ring of six AAA+ ATPase domains (numbered AAA1-AAA6), from which extends a 15 nm antiparallel coiled-coil stalk terminating in a microtubule-binding domain. ATP hydrolysis at AAA1 drives the power stroke: a linker domain that arches across the AAA+ ring undergoes a conformational change that translocates the cargo-bound tail relative to the microtubule-bound stalk. The step size of dynein is variable (8-32 nm, depending on load and ATP concentration), in contrast to the fixed 8 nm step of kinesin. Dynein requires the dynactin complex (11 subunits) for processive movement and for binding most cargoes. The adaptor protein BICD2 links dynein-dynactin to Golgi-derived vesicles, RNA granules, and nuclei during neuronal migration. Mutations in BICD2 cause spinal muscular atrophy, underscoring the functional importance of the dynein adaptor system.

Intraflagellar transport (IFT) is the bidirectional movement of protein complexes (IFT particles, ~20 polypeptides each in IFT-A and IFT-B complexes) along the axonemal microtubules of cilia and flagella. Kinesin-2 carries IFT particles and their cargo (tubulin, axonemal proteins, signaling receptors) from the base to the tip of the cilium (anterograde, at ~2.5 μm/s). Cytoplasmic dynein 2 returns the emptied IFT particles to the base (retrograde, at ~4 μm/s). Defects in IFT produce a spectrum of ciliopathies: Bardet-Biedl syndrome, polycystic kidney disease, and primary ciliary dyskinesia all result from mutations in IFT machinery.

Intermediate filaments and mechanotransduction

Intermediate filaments (IFs) are the most diverse class of cytoskeletal proteins, with over 70 human genes encoding proteins classified into six types based on sequence homology and assembly properties ^{[Herrmann 2007]}. Type I (acidic keratins, K9-K20) and Type II (basic keratins, K1-K8) are obligate heteropolymers found in epithelial cells — each keratin filament contains one type I and one type II chain. Type III includes vimentin (mesenchymal cells), desmin (muscle), and glial fibrillary acidic protein (GFAP, astrocytes). Type IV includes the neurofilament proteins NF-L, NF-M, NF-H (neurons). Type V is the nuclear lamins (lamin A/C and lamin B1/B2), which form the nuclear lamina underlying the inner nuclear membrane. Type VI includes filensin and CP49 (lens cells).

IF assembly proceeds through a series of defined intermediaries. Two parallel alpha-helical rods coil around each other to form a parallel coiled-coil dimer (approximately 45 nm for cytoplasmic IFs). Two dimers associate in an antiparallel, staggered fashion to form a tetramer — the fundamental soluble unit of IF assembly. Eight tetramers associate laterally to form a unit-length filament (ULF, approximately 60 nm long and 16 nm wide). ULFs anneal end-to-end and the resulting filaments undergo radial compaction from 16 nm to 10 nm, driven by longitudinal sliding of tetramers. The entire process takes 5-10 minutes in vitro, without nucleotide hydrolysis — IF assembly is purely thermodynamic, driven by hydrophobic and electrostatic interactions between coiled-coil rods. This contrasts sharply with actin and microtubule assembly, which requires ATP or GTP hydrolysis for dynamics.

The mechanical properties of IFs reflect their hierarchical structure. Single vimentin filaments have a persistence length $l_{p} \approx 0.5 - 1.0$ μm (far more flexible than actin at $l_{p} \approx 17$ μm or microtubules at $l_{p} \approx 6$ mm). Under tensile strain, individual IFs exhibit pronounced strain hardening: the force-extension curve transitions from an initial entropic regime (thermal fluctuations straightened by force) to an enthalpic regime (alpha-helical coiled coils begin to unravel at forces above ~~200 pN). The coiled-coil unfolding occurs at a well-defined force plateau (~~200-300 pN) and is reversible: when force is released, the coiled coil can refold. This property gives IF networks an enormous capacity for energy dissipation — they can absorb mechanical energy by partially unfolding and then recover. Keratin networks in epithelial cells protect against shear stress by this mechanism: the keratin tonofibrils anchored at desmosomes stretch under load, absorb the energy of deformation, and recoil when the load is removed.

The LINC complex (Linker of Nucleoskeleton and Cytoskeleton) couples the cytoplasmic cytoskeleton to the nuclear lamina across the nuclear envelope. The LINC complex consists of SUN domain proteins (SUN1, SUN2) spanning the inner nuclear membrane and nesprin proteins (nesprin-1, nesprin-2, nesprin-3, nesprin-4) spanning the outer nuclear membrane. In the perinuclear space, the SUN domain of SUN1/2 binds the KASH domain of nesprin, forming a trans-envelope bridge. Cytoplasmically, nesprin-1/2 bind actin through their N-terminal calponin homology (CH) domains, nesprin-3 binds plectin (which links to intermediate filaments), and nesprin-4 binds kinesin-1. Nucleoplasmically, SUN1/2 bind lamin A/C. This continuous chain — cytoplasmic cytoskeleton → nesprin → SUN → lamin A/C → chromatin — transmits mechanical force from the cell surface to the nucleus, allowing the nucleus to sense the mechanical environment.

Nuclear mechanotransduction through lamin A/C has emerged as a central pathway in cellular responses to mechanical stress. Lamin A/C levels scale with tissue stiffness: cells in stiff tissues (bone, muscle) express 3-5 times more lamin A/C than cells in soft tissues (brain, fat). Under mechanical load, lamin A/C undergoes conformational changes that expose buried binding sites for transcription factors. The most studied pathway is the YAP/TAZ (Yes-associated protein / transcriptional coactivator with PDZ-binding motif) mechanosensing axis. In cells on stiff substrates, cytoskeletal tension pulls on the LINC complex, stretching the lamin A/C network and opening nuclear pore complexes. YAP/TAZ translocates into the nucleus and activates proliferative and anti-apoptotic gene programs. On soft substrates, the lamin network is relaxed, nuclear pores close, and YAP/TAZ is sequestered in the cytoplasm and degraded.

Mutations in IF genes produce a striking range of human diseases, directly connecting molecular structure to tissue-level pathology ^{[Fuchs 1994]}. Epidermolysis bullosa simplex (EBS) is caused by dominant-negative mutations in K14 (or K5) that disrupt keratin filament assembly. The resulting fragile keratin network cannot resist the mechanical stress of daily life: minor friction or pressure causes basal epidermal cells to rupture, producing painful blisters. Over 100 distinct KRT14 mutations have been identified in EBS patients, and the severity correlates with the position of the mutation in the coiled-coil rod — mutations at the helix initiation motif or helix termination motif are most severe because they prevent filament elongation.

Progeria (Hutchinson-Gilford progeria syndrome) is caused by a de novo point mutation in LMNA (encoding lamin A/C) that activates a cryptic splice site, producing a truncated protein called progerin that lacks 50 amino acids near the C-terminus. The deletion removes the site of ZMPSTE24-mediated processing, so progerin remains permanently farnesylated and anchored to the inner nuclear membrane. The accumulated progerin disrupts the nuclear lamina, producing the characteristically misshapen nuclei, impaired DNA repair, and premature aging phenotypes. The disease directly demonstrates that the nuclear IF network is essential for genome stability and normal aging.

Vimentin, the Type III IF of mesenchymal cells, has signaling functions beyond its structural role. Vimentin scaffolds the Raf-MEK-ERK MAP kinase cascade: Raf kinase binds to vimentin filaments, and this localisation is required for efficient MEK phosphorylation. Vimentin-null cells show attenuated ERK activation and reduced proliferation. Vimentin also regulates integrin recycling during cell migration: the vimentin network anchors Rab11-positive recycling endosomes, controlling the return of integrins from the degradative pathway to the leading edge. Cancer cells undergoing epithelial-mesenchymal transition (EMT) upregulate vimentin expression 10-50 fold, and vimentin is a canonical EMT marker. The upregulated vimentin supports both the signaling changes (Raf/MEK/ERK hyperactivation) and the mechanical changes (increased deformability, invasive migration) that characterise metastatic cells.

Cytoskeletal coordination in cell division and migration

Cell division and cell migration are the two processes that most dramatically showcase the coordination of all three cytoskeletal systems. Both require precise spatial and temporal integration of actin, microtubule, and intermediate filament dynamics, regulated by overlapping signalling cascades.

Mitotic spindle assembly begins with centrosome maturation in prophase. The centrosome accumulates γ-tubulin ring complexes (γ-TuRCs) that nucleate microtubules and recruits additional pericentriolar material (PCM) proteins. The two centrosomes separate and move to opposite poles, driven by kinesin-5 (Eg5) sliding antiparallel microtubules apart and by dynein anchored at the cell cortex pulling on astral microtubules. Kinetochore microtubules (K-fibres) capture chromosomes at the kinetochore through a search-and-capture mechanism: dynamically unstable microtubules probe the cytoplasm, and those that encounter a kinetochore are stabilised. The initial attachment is typically lateral (the microtubule contacts the side of the kinetochore), and the Ndc80 complex at the outer kinetochore then converts this to an end-on attachment. Aurora B kinase, localised to the inner centromere, senses incorrect attachments (merotelic, syntelic) by measuring tension: unattached or improperly attached kinetochores are under low tension, which brings the outer kinetochore substrates closer to Aurora B, promoting their phosphorylation and detachment. This error-correction cycle continues until all chromosomes achieve amphitelic attachment (one sister kinetochore attached to each pole, with bi-oriented tension), at which point the spindle assembly checkpoint (SAC) is satisfied and anaphase proceeds ^{[Musacchio 2017]}.

During anaphase, two forces separate the chromosomes. Anaphase A moves chromosomes poleward through two mechanisms: depolymerisation of K-fibre microtubules at the kinetochore (the "Pac-man" model) and depolymerisation at the spindle pole (flux). The Kif2A kinesin-13 at the pole and the MCAK/Kif2C at the kinetochore catalyse microtubule depolymerisation. Anaphase B elongates the spindle through two additional mechanisms: sliding of antiparallel interpolar microtubules by kinesin-5, and cortical dynein pulling on astral microtubules. The relative contribution of A and B varies across species: in mammalian cells, anaphase A dominates, contributing ~80% of chromosome separation.

Cytokinesis, the physical division of the cell into two daughters, is driven by a contractile ring of actin and myosin II assembled at the cell equator ^{[Pollard 2017]}. RhoA is activated at the equatorial cortex by the centralspindlin complex (MKLP1 kinesin-6 and MgcRacGAP), which localises to the central spindle midzone during anaphase. Active RhoA triggers two downstream pathways: (1) formin activation (mDia1, mDia2) nucleates and elongates actin filaments at the division plane, and (2) ROCK phosphorylates myosin light chain, activating myosin II contractility. The contractile ring is not a static purse-string: actin filaments turn over with a half-life of ~15 seconds, continuously disassembled by cofilin at the inside of the ring and reassembled by formins at the outside. Myosin II minifilaments pull on the dynamic actin meshwork, generating contractile stress that progressively constricts the ring. The constriction rate in sea urchin embryos is approximately 0.1 μm/s, narrowing the diameter from ~30 μm to 0 in ~5 minutes. The ingressing furrow must also resolve membrane insertion (to increase surface area as the cell divides) and central spindle midbody formation (the final tether between daughter cells that is severed by ESCRT-III-mediated membrane scission).

Cell migration follows a coordinated cycle: (1) protrusion at the leading edge driven by actin polymerisation, (2) adhesion formation under the protrusion linking actin to the extracellular matrix, (3) translocation of the cell body forward by myosin II-mediated contractility, and (4) rear retraction by disassembly of adhesions at the trailing edge.

Focal adhesions are the primary mechanical linkage between the actin cytoskeleton and the extracellular matrix ^{[Geiger 2001]}. They are macromolecular assemblies of over 150 different proteins, organised in a layered architecture. The transmembrane receptors are integrins (αβ heterodimers, with 24 distinct pairs in mammals), which bind ECM ligands (fibronectin, collagen, laminin) on the outside and recruit talin on the inside. Talin undergoes force-induced unfolding: its 13 rod domains contain cryptic vinculin-binding sites that are exposed when cytoskeletal tension stretches the talin molecule. Vinculin, once bound, recruits additional actin filaments and signalling molecules (VASP, paxillin, FAK), reinforcing the linkage. Nascent adhesions (~~0.25 μm diameter) form behind the leading edge in the lamellipodium; those that experience sufficient force (from actomyosin contractility or retrograde flow) mature into focal complexes (~~0.5 μm) and then into focal adhesions (~1-5 μm). Maturation is force-dependent: inhibiting myosin II (with blebbistatin) prevents nascent adhesions from maturing.

The epithelial-mesenchymal transition (EMT) illustrates how cytoskeletal reprogramming drives a radical change in cell behaviour. During EMT, epithelial cells lose E-cadherin-based cell-cell junctions (and the cortical actin belt that anchors them), upregulate N-cadherin and vimentin, and activate formins (mDia2) and myosin II to build stress fibres and focal adhesions. The Rho GTPase network is rewired: Rac activity (which maintains the cortical actin belt) is suppressed, while Rho activity (which drives stress fibres and contractility) is elevated. The transcription factors Snail, Slug, Twist, and Zeb1/2 drive this reprogramming by repressing E-cadherin transcription and activating vimentin, N-cadherin, and MMP (matrix metalloproteinase) expression. The resulting mesenchymal cell is highly migratory, invasive, and resistant to anoikis (detachment-induced apoptosis). EMT is essential during embryonic development (gastrulation, neural crest migration) and wound healing, but is also hijacked by carcinoma cells during metastasis.

Synthesis. The three cytoskeletal systems — actin, microtubules, and intermediate filaments — are not independent structural elements but a mechanically and signalling-integrated cytoplasmic scaffolding whose collective behaviour is the foundational reason cells can change shape, divide, and migrate. The central insight is that each system provides a distinct mechanical and functional capability that the others cannot replicate: actin provides rapid protrusive and contractile force generation through treadmilling and myosin II contractility; microtubules provide long-range directional transport and mitotic spindle mechanics through dynamic instability and motor protein walking; intermediate filaments provide tensile resilience and mechanotransductive signalling through strain hardening and LINC complex coupling. Putting these together, the Rho GTPase signalling network emerges as the master coordinator, activating Arp2/3 for lamellipodial protrusion (Rac), formins for filopodial extension (Cdc42), and myosin II for contractility (Rho) — this is exactly the integration that identifies the cytoskeleton as an active, self-organising structure rather than a passive scaffold. The bridge is between single-molecule mechanochemistry (kinesin stepping, myosin power strokes) and tissue-level mechanics (muscle contraction in 18.04.02 pending, cell division in 17.08.01), and the pattern generalises from molecular motors to the collective force generation of ensembles — whether sarcomeres in muscle, the contractile ring in cytokinesis, or stress fibre arrays in migrating cells.

Full proof set [Master]

Proposition 1 (Treadmilling flux inequality). For an actin filament at steady state, net polymerisation at the barbed end equals net depolymerisation at the pointed end if and only if the monomer concentration $[G]$ satisfies $C_{c}^{+} < [G] < C_{c}^{-}$ , where $C_{c}^{+} = k_{off}^{+} / k_{on}^{+}$ and $C_{c}^{-} = k_{off}^{-} / k_{on}^{-}$ are the critical concentrations at the two ends. The steady-state treadmilling flux is $J = k_{on}^{+} [G] - k_{off}^{+} = k_{off}^{-} - k_{on}^{-} [G]$ .

Proof. At the barbed end, the net assembly rate is $J^{+} = k_{on}^{+} [G] - k_{off}^{+}$ . At the pointed end, $J^{-} = k_{on}^{-} [G] - k_{off}^{-}$ . Net growth is $J^{+} + J^{-}$ . For treadmilling (constant filament length with net flux through the filament), we require $J^{+} = - J^{-}$ , i.e., net addition at one end equals net loss at the other. This gives $k_{on}^{+} [G] - k_{off}^{+} = k_{off}^{-} - k_{on}^{-} [G]$ , and solving: $[G] (k_{on}^{+} + k_{on}^{-}) = k_{off}^{+} + k_{off}^{-}$ . This single steady-state concentration is the overall critical concentration $C_{c} = (k_{off}^{+} + k_{off}^{-}) / (k_{on}^{+} + k_{on}^{-})$ . Since $C_{c}^{+} < C_{c}^{-}$ (the barbed end has lower critical concentration), when $[G]$ is maintained at this overall $C_{c}$ , the barbed end grows ( $J^{+} > 0$ ) and the pointed end shrinks ( $J^{-} < 0$ ), both at rate $J$ . $□$

Proposition 2 (Motor processivity from detachment rate). The mean run length $⟨ L ⟩$ of a processive motor with step size $d$ and probability of detachment per step $p_{det}$ is $⟨ L ⟩ = d / p_{det}$ . For kinesin-1 with $d = 8$ nm and observed run length $⟨ L ⟩ \approx 800$ nm, the per-step detachment probability is $p_{det} = 0.01$ .

Proof. Each step is a Bernoulli trial: the motor either advances by $d$ (with probability $1 - p_{det}$ ) or detaches (with probability $p_{det}$ ). The number of steps $N$ before detachment follows a geometric distribution: $P (N = n) = (1 - p_{det})^{n} \cdot p_{det}$ . The mean of this distribution is $⟨ N ⟩ = (1 - p_{det}) / p_{det}$ . For $p_{det} ≪ 1$ , $⟨ N ⟩ \approx 1/ p_{det}$ . The mean run length is $⟨ L ⟩ = d \cdot ⟨ N ⟩ \approx d / p_{det}$ . Substituting $d = 8$ nm and $⟨ L ⟩ = 800$ nm: $p_{det} = 8/800 = 0.01$ . This 1% per-step detachment rate is maintained by the hand-over-hand mechanism: at least one head is always bound to the microtubule, and the detached head is guided to the next binding site by the neck linker, making spontaneous detachment a rare event. $□$

Connections [Master]

Muscle contraction — actin-myosin mechanics 18.04.02 pending. The molecular-level actin-myosin contractility analysed here — single-molecule force-velocity relationships, tight ATP coupling, ensemble force summation — is the direct prerequisite for the tissue-level treatment of sarcomere organisation, excitation-contraction coupling, and whole-muscle biomechanics in the organismal biology chapter.
Cell cycle and mitosis 17.08.01. Mitotic spindle assembly, kinetochore-microtubule attachment, anaphase chromosome segregation, and cytokinetic ring constriction are all cytoskeletal processes whose molecular mechanisms are described in this unit and whose regulation by cell-cycle checkpoints is treated in the cell-cycle unit.
Cellular organization — organelles 17.03.01 pending. Microtubule-based transport by kinesin and dynein is the primary mechanism for positioning organelles (Golgi apparatus, endosomes, lysosomes, mitochondria) and for vesicular trafficking between compartments, connecting the motor protein machinery of this unit to the organelle localisation and membrane trafficking described in the organelle unit.
Resting membrane potential and ion channels 17.09.01. Axonal transport of vesicles containing ion channels and neurotransmitters depends on the microtubule motor proteins described here. The distal localisation of voltage-gated sodium channels at the axon initial segment requires ankyrin-G-mediated tethering to the actin-microtubule cytoskeleton, directly linking cytoskeletal organisation to neuronal excitability.
Cell signaling — receptors and GPCRs 17.07.01 pending. Rho GTPase signalling (Rho, Rac, Cdc42) is the master regulator of actin dynamics, connecting the receptor-level signaling events described in the signaling unit to the cytoskeletal rearrangements that execute cell migration, division, and morphological change.

Historical & philosophical context [Master]

The discovery of the cytoskeleton was inseparable from the development of electron microscopy and the biochemical characterisation of the proteins that formed the filamentous structures visible in micrographs. Slautterback in 1963 and Ledbetter and Porter in 1964 first described microtubules as 25 nm hollow tubes in plant and animal cells ^{[Slautterback 1963]}. Borisy and Taylor in 1967 purified tubulin by colchicine-binding assay, establishing that microtubules are polymers of a 110 kDa heterodimer and opening the door to in vitro reconstitution ^{[Borisy & Taylor 1967]}.

Huxley and Hanson in 1954 proposed the sliding filament theory of muscle contraction, demonstrating that actin thin filaments slide past myosin thick filaments without either filament changing length ^{[Huxley & Hanson 1954]}. This was the first mechanistic model of a cytoskeletal motor and established the paradigm of force generation by cyclic interaction between a motor protein and a filament.

Inoue in 1967, using polarised light microscopy on living mitotic cells, observed that spindle microtubules could assemble and disassemble rapidly, anticipating the concept of dynamic instability nearly two decades before its formal discovery ^{[Inoue 1967]}. Mitchison and Kirschner in 1984 provided the definitive quantitative characterisation of dynamic instability using dark-field microscopy of purified microtubules ^{[Mitchison & Kirschner 1984]}.

Vale, Reese, and Sheetz in 1985 discovered kinesin using a motility assay on axoplasm extruded from squid giant axons, identifying a novel ATP-dependent motor protein that moved along microtubules toward the plus end ^{[Vale Reese Sheetz 1985]}. Cytoplasmic dynein was identified by Vallee and colleagues in 1988 as the minus-end-directed counterpart. Vale and Milligan in 2000 determined the first high-resolution cryo-EM structure of kinesin bound to a microtubule, revealing the atomic-level interaction between motor domain and tubulin ^{[Vale Milligan 2000]}.

Yildiz et al. in 2004 used single-molecule fluorescence imaging with one-nanometre accuracy (FIONA) to directly visualise the hand-over-hand walking mechanism of kinesin, resolving a decade-long debate about whether kinesin moves by inchworm or hand-over-hand stepping ^{[Yildiz 2004]}. Their data showed alternating 16 nm displacements of individual heads (each advancing 16 nm while the partner is stationary, producing 8 nm net centre-of-mass steps), confirming the hand-over-hand model.

Bibliography [Master]

@article{HuxleyHanson1954,
  author = {Huxley, H. E. and Hanson, J.},
  title = {Changes in the cross-striation of muscle during contraction and
           structural changes in the myosin filament},
  journal = {Nature},
  volume = {173},
  year = {1954},
  pages = {973--976},
}

@article{Slautterback1963,
  author = {Slautterback, D. B.},
  title = {Cytoplasmic microtubules. I. Hydra},
  journal = {J. Cell Biol.},
  volume = {18},
  year = {1963},
  pages = {367--388},
}

@article{BorisyTaylor1967,
  author = {Borisy, G. G. and Taylor, E. W.},
  title = {The mechanism of action of colchicine. {C}olchicine binding
           to sea urchin eggs and the mitotic apparatus},
  journal = {J. Cell Biol.},
  volume = {34},
  year = {1967},
  pages = {525--548},
}

@article{Inoue1967,
  author = {Inoue, S.},
  title = {Video enhancement and image processing in light microscopy},
  journal = {J. Cell Biol.},
  volume = {35},
  year = {1967},
  pages = {15A},
}

@article{MitchisonKirschner1984,
  author = {Mitchison, T. and Kirschner, M.},
  title = {Dynamic instability of microtubule growth},
  journal = {Nature},
  volume = {312},
  year = {1984},
  pages = {237--242},
}

@article{ValeReeseSheetz1985,
  author = {Vale, R. D. and Reese, T. S. and Sheetz, M. P.},
  title = {Identification of a novel force-generating protein, kinesin,
           involved in microtubule-based motility},
  journal = {Cell},
  volume = {42},
  year = {1985},
  pages = {39--50},
}

@article{Vallee1988,
  author = {Vallee, R. B. and Wall, J. S. and Paschal, B. M. and
            Shpetner, H. S.},
  title = {Microtubule-associated protein 1C from brain is a two-headed
           cytosolic dynein},
  journal = {Nature},
  volume = {332},
  year = {1988},
  pages = {561--563},
}

@article{ValeMilligan2000,
  author = {Vale, R. D. and Milligan, R. A.},
  title = {The way things move: looking under the hood of molecular motor
           proteins},
  journal = {Science},
  volume = {288},
  year = {2000},
  pages = {88--95},
}

@article{Yildiz2004,
  author = {Yildiz, A. and Tomishige, M. and Vale, R. D. and Selvin, P. R.},
  title = {Kinesin walks hand-over-hand},
  journal = {Science},
  volume = {303},
  year = {2004},
  pages = {676--678},
}

@article{Mullins1998,
  author = {Mullins, R. D. and Heuser, J. A. and Pollard, T. D.},
  title = {The interaction of {Arp2/3} complex with actin: nucleation,
           high affinity pointed end capping, and formation of branching
           networks of filaments},
  journal = {Proc. Natl. Acad. Sci. USA},
  volume = {95},
  year = {1998},
  pages = {6181--6186},
}

@article{Ridley2001,
  author = {Ridley, A. J. and Schwartz, M. A. and Burridge, K. and
            Firtel, R. A. and Ginsberg, M. H. and Borisy, G. and
            Parsons, J. T. and Horwitz, A. R.},
  title = {Cell migration: integrating signals from front to rear},
  journal = {Science},
  volume = {302},
  year = {2003},
  pages = {1704--1709},
}

@article{Pollard2017,
  author = {Pollard, T. D.},
  title = {Nine unanswered questions about cytokinesis},
  journal = {J. Cell Biol.},
  volume = {216},
  year = {2017},
  pages = {3007--3016},
}

@article{Musacchio2017,
  author = {Musacchio, A.},
  title = {The molecular biology of spindle assembly checkpoint signaling
           dynamics},
  journal = {Curr. Biol.},
  volume = {25},
  year = {2015},
  pages = {R1002--R1018},
}

@book{Howard2001,
  author = {Howard, J.},
  title = {Mechanics of Motor Proteins and the Cytoskeleton},
  publisher = {Sinauer Associates},
  year = {2001},
}

@book{Bray2001,
  author = {Bray, D.},
  title = {Cell Movements: From Molecules to Motility},
  publisher = {Garland Science},
  year = {2001},
  edition = {2nd},
}

@book{Alberts2022,
  author = {Alberts, B. and others},
  title = {Molecular Biology of the Cell},
  publisher = {Garland Science},
  year = {2022},
  edition = {7th},
}

Prerequisites

17.01.01 pending
18.04.01 pending

Used in

18.04.01
17.03.01

Tier anchors

beginner: Alberts et al., Molecular Biology of the Cell, 7th ed. (2022), Ch. 16
intermediate: Alberts et al., MBoC 7e, Ch. 16; Lodish et al., Molecular Cell Biology, 9th ed. (2023), Ch. 17
master: Alberts et al., MBoC 7e; Howard, Mechanics of Motor Proteins and the Cytoskeleton (2001); Bray, Cell Movements, 2nd ed. (2001)

References

TODO_REF
Alberts et al. — Molecular Biology of the Cell, 7th ed. (Garland Science, 2022) · Ch. 16 The Cytoskeleton
TODO_REF
Howard — Mechanics of Motor Proteins and the Cytoskeleton (Sinauer, 2001) · Ch. 10-14 Motor proteins; Polymer dynamics
TODO_REF
Bray — Cell Movements, 2nd ed. (Garland Science, 2001) · Ch. 6-9 Molecular motors; Cytoskeletal filaments
TODO_REF
Huxley & Hanson 1954 — Changes in the cross-striation of muscle during contraction and structural changes in the myosin filament · Nature 173 (1954) 973-976; sliding filament theory
TODO_REF
Mitchison & Kirschner 1984 — Dynamic instability of microtubule growth · Nature 312 (1984) 237-242
TODO_REF
Vale, Reese & Sheetz 1985 — Identification of a novel force-generating protein, kinesin · Cell 42 (1985) 39-50
TODO_REF
Yildiz et al. 2004 — Kinesin walks hand-over-hand · Science 303 (2004) 676-678
TODO_REF
Borisy & Taylor 1967 — The mechanism of action of colchicine · J. Cell Biol. 34 (1967) 525-548; tubulin purification
TODO_REF
Inoue 1967 — Video enhancement of microscope images · J. Cell Biol. 35 (1967) 15A; anticipation of dynamic instability

Reviewer

Tyler (pending external biology reviewer per BIOLOGY_PLAN §6)

Estimated time

beginner: 15m
intermediate: 40m
master: 70m