Module 4: Motor Proteins

All cytoskeletal motors share a common architecture: a motor (catalytic) domain that binds the track and hydrolyzes ATP, a lever arm or stalkthat amplifies small catalytic-domain conformational changes into large mechanical steps, and a tail that binds cargo (often via light chains or adaptor proteins).

• Myosin superfamily: ~40 classes, sharing a ~80 kDa head (Rayment et al. 1993 first atomic structure, Science). Myosin II forms filaments (muscle sarcomere); myosin V is a processive dimeric cargo carrier; myosin VI is unique in walking toward the minus end of actin (Wells et al. 1999 Nature).
• Kinesin superfamily (KIFs): 14 families, >45 members in humans. Classified by motor position: N-kinesins (plus-end-directed, N-terminal motor), M-kinesins (depolymerases, central motor, e.g., MCAK/kinesin-13), and C-kinesins (minus-end-directed, C-terminal motor, e.g., kinesin-14/Ncd).
• Dynein superfamily: AAA+-family minus-end microtubule motors. Cytoplasmic dynein-1 handles retrograde cargo and mitosis; cytoplasmic dynein-2 (IFT dynein) powers retrograde intraflagellar transport; axonemal dyneins (outer and inner arms) drive ciliary/flagellar beating.

All three superfamilies evolved from a common ancestral P-loop NTPase (the G-protein fold in kinesin/myosin and the AAA+ fold in dynein). The mechanical work per ATP is set by the hydrolysis free energy (~20 k_BT under cellular conditions) and the coupling efficiency, which is typically 40-60% for kinesin and myosin.

Kinesin-1 (KIF5) is a heterotetramer: two heavy chains (~120 kDa each) that form the dimeric catalytic head plus a coiled-coil stalk, and two light chains (~60 kDa) at the tail that bind cargo. The two heavy-chain heads are linked by a short flexible neck linker (~14 residues). Each head binds ATP, hydrolyzes it, and releases ADP + P_i. Hydrolysis in one head is kinetically gated by the state of the other head—the key feature that underlies processivity.

Structural crystallography (Kull, Sablin, Lau, Fletterick & Vale 1996 Nature; Sindelar 2007 cryo-EM) revealed that kinesin and myosin share the same catalytic fold—a remarkable finding because the two motors walk on entirely different polymers and evolved independently. The “switch I” and “switch II” loops sense nucleotide state and drive conformational changes that reorient the neck linker.

Kinesin-1 is processive: a single dimer takes ~100 steps (often >500 under load-free conditions) before dissociating, translocating ~1 μm on average per encounter with a microtubule. Because each step is 8 nm and lasts ~10 ms at saturating ATP, the unloaded velocity is roughly 800 nm/s. Load, ATP concentration, and the state of the cargo adaptor all modulate velocity.

Before 2004, two alternative mechanisms for processive kinesin motion competed: a hand-over-hand (alternating heads swap leading/trailing roles) model and an inchworm model (one head always leads, the other follows). Yildiz, Tomishige, Vale & Selvin (2004, Science) resolved the question using FIONA (Fluorescence Imaging with One-Nanometer Accuracy): a Cy3-labeled head of a single kinesin dimer showed alternating steps of 16.6 nm and 0 nm, with the average head displacement per motor step being 8 nm. That is the precise fingerprint of hand-over-hand: the labeled head alternates between not moving (because the other head takes the step) and moving by twice the motor step (because it swings 16 nm from the trailing to the leading position).

The hand-over-hand cycle is coordinated by internal strain:

1. Rear head (ADP-bound, weakly attached) releases ADP and binds ATP as it swings forward.
2. ATP binding triggers neck-linker docking; the head throws forward by ~16 nm.
3. The new leading head binds the next tubulin dimer; strain in the neck linker inhibits premature ATP binding in the now-rear head.
4. Rear head hydrolyzes ATP, weakens its microtubule binding, and releases P_i.
5. Cycle repeats with head identities swapped.

This internal strain-gating ensures that only one head at a time is in a weak-binding state (ADP)—kinetically enforcing processivity. Mutations that disrupt neck-linker docking (e.g., proline substitutions) abolish processivity but not single-step motion.

Under opposing load \(F\), a motor slows. Rice, Purcell (Stanford biophysics, 1977) and later Hill established the phenomenological fit

Microscopically, load affects each transition in the chemomechanical cycle differently. The Bell approximation (Bell 1978) gives each rate an exponential dependence on force,\(k(F) = k_0 \exp(-F\delta/k_B T)\), with\(\delta\) the transition state distance. For kinesin, the rate-limiting strain-gated transition has \(\delta \approx 1\) nm, so at stall (\(F \approx 6\) pN) the transition is slowed by a factor of\(\exp(6\cdot 1/4.11) \approx 4\).

Velocity also depends on [ATP] via Michaelis-Menten kinetics:\(v([\text{ATP}]) = v_{\max} [\text{ATP}]/(K_m + [\text{ATP}])\) with\(K_m \approx 60\) μM for kinesin-1. At physiological [ATP] (~1 mM) the motor is near \(v_{\max}\); ATP depletion during metabolic stress slows intracellular transport proportionally.

Myosins move on actin filaments (plus-end-directed for most, with the lone exception of myosin VI). Their architecture is: a motor head (catalytic domain) attached to a lever arm stabilized by one or more light chains (usually an essential light chain [ELC] and a regulatory light chain [RLC]) wound around an IQ motif. A conformational change in the catalytic domain upon P_i release is amplified by the lever arm into a ~10 nm power stroke for myosin II or ~25 nm for myosin V.

• Myosin II: skeletal, cardiac, smooth-muscle and non-muscle isoforms; assemble into bipolar thick filaments; each motor has a low duty ratio (~5%—most of the time detached from actin) so ensembles of motors in a sarcomere can generate continuous force. Atomic structure: Rayment et al. 1993 Science; Dominguez et al. 1998.
• Myosin V: processive cargo carrier in melanosome and vesicle transport; 6 IQ repeats in the lever arm give a 36 nm step matching the half-helical pitch of F-actin; processive like kinesin (De La Cruz, Wells, Rosenfeld, Ostap & Sweeney 1999 PNAS).
• Myosin VI: the unique “backwards” motor. Walks toward the minus end of actin via a ~180\(^\circ\)insertion that inverts the lever arm direction (Wells et al. 1999 Nature); hauls cargo into the cell interior.
• Myosin VII: responsible for cargo transport in stereocilia; mutations cause Usher syndrome (deaf-blindness).
• Myosin XV: stereocilia tip motor; mutations cause non-syndromic deafness.

The lever-arm mechanism was confirmed in an elegant experiment by Uyeda et al. 1996 (PNAS) and Purcell et al. 2002 by engineering myosins with shortened or lengthened lever arms: the step size scaled linearly with lever-arm length, exactly as predicted.

Cytoplasmic dynein-1 is a ~1.4 MDa homodimer, each monomer containing a ~500 kDa heavy chain with six AAA+ domains arranged in a ring, a microtubule-binding domain (MTBD) at the tip of a coiled-coil stalk, and a tail that dimerizes and binds adaptors. AAA1 is the primary ATP hydrolysis site that powers the powerstroke; AAA3 and AAA4 act as regulatory ATP sites (Bhabha et al. 2014; Schmidt, Zalyte, Urnavicius & Carter 2015). The linker domain swings across the AAA+ ring in response to ATP hydrolysis, driving a ~8–20 nm step (Reck-Peterson et al. 2006; Reck-Peterson, Redwine, Vale & Carter 2018 Nat. Rev. Mol. Cell Biol.).

Unlike kinesin, dynein on its own is non-processive. Processive motility requires assembly with dynactin (a ~1 MDa multi-subunit filamentous complex nucleated on the actin-related protein Arp1) and a cargo adaptor (BICD2, Hook3, ninein, Rab7-RILP, Spindly). This dynein-dynactin-adaptor (DDA) complex takes ~100 steps of variable 8-16 nm size at ~300 nm/s (McKenney, Huynh, Tanenbaum, Bhabha & Vale 2014; Schlager 2014). The minimal stall force of mammalian cytoplasmic dynein is ~1 pN per motor, considerably lower than kinesin’s 6 pN.

Axonemal dyneins (outer and inner arms in motile cilia) are multi-headed complexes that drive the sliding of adjacent doublet microtubules relative to one another. Linker proteins (nexin/DRC) convert that sliding into the bending wave that constitutes the ciliary beat. Defects in outer dynein arm assembly (DNAH5, DNAI1, DNAAF1-3, CCDC39, CCDC40) cause primary ciliary dyskinesia / Kartagener syndrome.

Motor biophysics was transformed by the invention of single-molecule assays that can observe a single motor walking, step by step:

• Optical trap (Svoboda, Schmidt, Schnapp & Block 1993 Nature): a focused laser beam traps a polystyrene bead attached to a single kinesin motor; step sizes resolved by interferometry. First direct demonstration of 8 nm kinesin steps and of stall forces.
• Three-bead assay (Finer, Simmons & Spudich 1994 Nature): two beads hold a single actin filament under tension; a third bead coated with myosin contacts the filament transiently. Captured the ~5 nm myosin-II power stroke and ~3-4 pN force.
• FIONA (Yildiz et al. 2003, 2004): fit a 2D Gaussian to the single-molecule fluorescence point-spread function to get ~1.5 nm localization precision. Established hand-over-hand kinesin walking and the 36 nm step of myosin V.
• Magnetic and atomic-force tweezers: apply torque or larger forces; used to probe myosin ensemble behavior and nucleic-acid motors.
• TIRF (total internal reflection fluorescence): images single motors walking on surface-attached microtubules, giving velocity and run-length distributions for thousands of individual motors.

These assays provide the raw data that the kinetic models in this module must reproduce—step size, dwell-time distributions, stall force, run length, load dependence.

A motor that is not carrying cargo does not need to hydrolyze ATP. Cells therefore tightly regulate motor activity through autoinhibition and cargo-induced activation:

• Kinesin-1 autoinhibition: the tail folds back and binds the motor head domain; kinesin light chains stabilize this inactive state. Cargo binding (e.g., JIP3, kinectin, TRAK1/2) or phosphorylation (JNK, CaMKII) releases the fold-back and activates the motor.
• Dynein autoinhibition “phi” conformation: the two motor heads pack against each other in a φ-shaped configuration. Dynactin plus a coiled-coil adaptor (BICD2, Hook3, Spindly) pulls the two heads apart and activates processive motility. Without an adaptor, dynein-dynactin alone is essentially static.
• Myosin-II RLC phosphorylation: myosin light-chain kinase (MLCK) phosphorylates the regulatory light chain at Ser19 to activate smooth/non-muscle myosin II; myosin phosphatase dephosphorylates and inactivates.
• Myosin V autoinhibition by the tail: Ca²⁺ binding, cargo binding, and Rab-GTP binding release the tail from the head in a gated mechanism.

Many cargos bind both kinesin and dynein, and the motors are opposed on the same microtubule. The balance of activation—phosphorylation state, Ca²⁺, adaptor identity—sets the net direction of transport. This is the substrate of the “tug-of-war” models in cell biology (Gross, Vershinin & Shubeita 2007; Hancock 2014).

Most intracellular cargos carry multiple motors of both polarities. When multiple kinesins cooperate on one cargo, they share the load: \(F_{\text{total}} = n\,F_{\text{single}}\)at saturation, but because the motors are not coordinated in time the effective force is smaller than the naive sum. Klumpp & Lipowsky (2005 PNAS) worked out the mean-field theory for \(n\) cooperating motors: the probability of having\(k\) bound at any moment is Poisson in the binding rate.

When opposing motors (kinesin + dynein) are on the same cargo, pauses, reversals, and stochastic switching emerge (Welte 2004 Curr. Biol.; Gross 2004). The steady state reflects the stochastic balance of binding/unbinding with mutual unloading by the opposing motor. Regulatory signals that shift this balance redirect intracellular traffic.

In axons, slow axonal transport of cytoskeletal elements (neurofilaments, tubulin) arises from an intermittent fast-transport mechanism: short kinesin runs interrupted by long pauses, averaging to 0.1–1 mm/day (Brown 2003). This is the substrate of Wallerian-like degeneration when transport fails.

Motor-protein disease mutations fall into three categories: loss of motor function, disruption of cargo binding, and disruption of autoinhibition/activation. A partial catalog:

Motor-targeted drugs are an emerging therapeutic area. Mavacamten (FDA-approved 2022 for obstructive hypertrophic cardiomyopathy) is a β-myosin allosteric inhibitor that stabilizes the super-relaxed state and reduces contractility. Monastrol and ispinesib target the mitotic kinesin Eg5 and arrest mitosis; several are in cancer clinical trials.

At cellular ATP and ADP concentrations, the hydrolysis free energy is\(\Delta G_{\text{ATP}} \approx -20\,k_B T \approx -50\text{ kJ/mol}\). A kinesin step takes this input and produces mechanical work:

This efficiency is remarkable for a molecular machine operating at 300 K in a viscous environment: comparable to the best man-made heat engines. It is made possible because kinesin does not work as a heat engine but as an isothermal chemical engine, avoiding the Carnot bound. The detailed balance of transitions and the free-energy profile along the reaction coordinate constrain how efficient the motor can be, as analyzed rigorously in the stochastic thermodynamics framework (Seifert 2012 Rep. Prog. Phys.).