Molecular Motors: Kinesin and Dynein

A molecular motor is a protein that converts the chemical energy released by ATP hydrolysis into mechanical work — directed movement or force generation — rather than letting that energy dissipate as heat. Kinesins and dyneins are the two great families of motors that do this on microtubules, the polar, hollow, 25-nm cytoskeletal filaments built from α/β-tubulin dimers. Microtubule polarity is the key organizing fact: every microtubule has a fast-growing plus end (typically near the cell periphery or, in neurons, the axon terminal) and a slower minus end (typically anchored at the centrosome/microtubule-organizing center). This polarity gives the track a built-in "one-way sign" that motors can read. Kinesins are a large superfamily (over 45 genes in humans, grouped into ~14–17 kinesin families). Most conventional kinesins (kinesin-1, the founding member) walk toward the plus end, making them the default motors for outbound traffic — from the cell body to the cell surface, or from the Golgi toward the plasma membrane. A minority of kinesins (notably kinesin-14 family members like Ncd) walk toward the minus end. Cytoplasmic dynein, by contrast, is essentially the sole minus-end-directed motor in most cells (a second cytoplasmic dynein, dynein-2, is specialized for retrograde transport inside cilia). Dynein therefore handles inbound traffic: endosomes heading to lysosomes, retrograde axonal signals returning to the cell body, and the chromosome-to-pole pulling forces of mitosis. Both motors share a conceptual architecture inherited from convergent — not shared — evolutionary solutions to the same problem also solved by myosin on actin (see Molecular Motors: Myosin): a catalytic motor domain that binds the track and hydrolyzes ATP, a neck/stalk region that amplifies small chemical changes into large mechanical displacements, and a cargo-binding tail or accessory-chain complex that links the motor to vesicles, organelles, chromosomes, or other cytoskeletal elements. Kinesin's motor domain is structurally related to myosin's, despite low sequence identity — both belong to the P-loop NTPase superfamily — a striking case of mechanistic convergence. Dynein's motor domain, in contrast, is unrelated: it is a massive ring of six AAA+ ATPase domains, making dynein one of the largest and most mechanically complex motor proteins known.

Kinesin-1 structure. Conventional kinesin-1 is a homodimer: two identical heavy chains, each contributing a globular motor head (~350 amino acids, binds microtubule and ATP), a short neck linker, an extended coiled-coil stalk that dimerizes the two heavy chains, and a tail that binds two light chains and, through them, cargo (often via adaptor proteins and Rab GTPases on vesicle membranes). The two heads are the "feet"; the stalk is the "legs and body" holding them together.

The ATPase cycle and hand-over-hand stepping. Each kinesin head binds microtubule and ATP with a coupled, cyclical affinity: ATP binding causes the head's neck linker to dock forward along the head, throwing the partner (rear) head forward by about 16 nm so it can bind the next tubulin-binding site, 8 nm further along the protofilament. ADP release, phosphate release, and ADP-rebinding at the trailing head keep the cycle running so that the two heads alternate lead and trail roles — a strictly alternating, hand-over-hand gait, confirmed experimentally by fluorescently labeling single heads and watching them alternately advance 16 nm while cargo advances 8 nm. Because ATP hydrolysis at one head is coordinated with (and normally forbidden until) a specific state at the other head, kinesin-1 remains bound to the microtubule via at least one head at essentially all times — this property, called processivity, is why a single kinesin can carry a vesicle continuously for a micron or more without falling off, taking on average ~100 steps per ATP-hydrolysis cycle "run" before releasing.

Step size and speed. Kinesin-1 takes 8-nm steps — matching the 8-nm periodicity of the tubulin dimer — and moves at up to roughly 1 μm/s in vitro (somewhat slower under physiological cargo load), consuming approximately one ATP per 8-nm step, giving it a coupling efficiency close to unity.

Dynein structure. Cytoplasmic dynein is far larger and structurally distinct. Each of its two heavy chains (~500 kDa each) folds into a ring of six AAA+ domains (AAA1–AAA6), only some of which hydrolyze ATP (mainly AAA1, the primary catalytic site, with AAA2–AAA4 playing regulatory roles). Projecting from this ring is a long, antiparallel coiled-coil stalk tipped with a small microtubule-binding domain (MTBD) — so the ATP-hydrolyzing site and the track-binding site are separated by ~15 nm and communicate via a mechanical signal running the length of the stalk. A second appendage, the linker, runs across the face of the AAA+ ring and acts as dynein's mechanical lever: ATP hydrolysis in AAA1 causes the linker to bend and straighten, remodeling the ring and swinging the stalk, which is dynein's power stroke, mechanistically analogous in concept (though not in structure) to kinesin's neck-linker docking and myosin's lever-arm swing. Unlike kinesin-1, cytoplasmic dynein does not work efficiently alone; it is normally activated and made processive by binding to a large multi-subunit adaptor, dynactin, and a cargo-specific activating adaptor (e.g., BicD2, Hook proteins). This dynein–dynactin–adaptor complex recruits two dynein motors side by side, and only in this activated, "processive" configuration does dynein reliably walk long distances rather than shuffle and detach — a major reason dynein's regulation was harder to crack experimentally than kinesin's.

Reading the track's polarity. Both motors "know" which way is plus and which is minus because the motor domain's contact surface with the microtubule is asymmetric, matching the asymmetric, polar arrangement of α- and β-tubulin subunits within each protofilament. The motor domain's docking orientation on the tubulin lattice, not any separate sensing mechanism, is what fixes directionality — a fact confirmed by engineering chimeric and mutant kinesins whose neck-linker orientation could be reversed, which reversed their direction of travel.

Reconstituting movement in a test tube (Vale, Reese & Sheetz, 1985; Vale, Schnapp, Reese & Sheetz, 1985). Ron Vale, Michael Sheetz, and colleagues, studying fast axonal transport in extruded squid giant axon cytoplasm, found that isolated organelles continued to move along microtubule tracks even outside an intact cell, and — crucially — that microtubules alone plus a soluble cytoplasmic fraction and ATP were sufficient to drive movement of latex beads coated with the fraction's proteins. This led to the isolation of a novel ATPase, which Vale named kinesin, distinct from myosin and dynein, that could by itself propel beads along microtubules in a defined direction. This was the first direct proof that a single protein species, using ATP, could generate processive, directional movement along a microtubule — establishing the entire concept of a microtubule motor protein.

The in vitro motility assay. The definitive assay format, developed by Sheetz, Vale, and James Spudich's approach borrowed from myosin/actin work, attaches purified motor to a glass coverslip and flows fluorescently labeled microtubules over it; ATP-dependent gliding of microtubules across the motor-coated surface, observed by video-enhanced light microscopy, demonstrated unambiguously that the purified motor alone (no other cytoplasmic factors) is sufficient for force generation, and that the direction of gliding reveals the motor's polarity preference (kinesin glides microtubules with the minus end leading — equivalent to kinesin itself walking toward the plus end). This assay became the standard tool for dissecting every subsequent question about motor mechanism, and it directly demonstrated that dynein (isolated originally from cilia by Ian Gibbons in 1965, and later shown to exist as a distinct cytoplasmic form by Vallee and colleagues in 1987) drives microtubule gliding in the opposite direction from kinesin — the first direct biochemical proof that the cell has two oppositely-directed microtubule motor systems.

Single-molecule optical trapping (Svoboda, Schmidt, Schnapp & Block, 1993). Using laser optical tweezers to hold a bead attached to a single kinesin molecule, Steven Block's group measured the actual step size and force output of one motor molecule in real time. They found kinesin advances in discrete 8-nanometer steps and stalls under a load of roughly 5–6 piconewtons, matching the 8-nm tubulin dimer repeat and confirming that kinesin moves processively as a single molecule rather than requiring cooperative teams — a landmark demonstration that single-molecule biophysics could resolve mechanochemical events invisible to bulk biochemistry.

Hand-over-hand vs. inchworm (Yildiz, Selvin, and colleagues, 2003–2004). A long-standing debate was whether kinesin's two heads step symmetrically (hand-over-hand, like a person walking) or whether one head always leads (inchworm, like a caterpillar). Using a super-resolution fluorescence technique (FIONA — fluorescence imaging with one-nanometer accuracy), Ahmet Yildiz and Paul Selvin labeled a single head of a kinesin dimer with a single fluorophore and tracked its position as the motor walked. The label alternated between positions displaced by 16 nm and 0 nm (i.e., the labeled head moved 16 nm, then stayed put for the next step while the other head moved), proving unambiguously that kinesin walks hand-over-hand, with each head alternately becoming the leading, then the trailing partner — settling a debate that pure bulk kinetics could not resolve and validating the neck-linker docking model of the mechanochemical cycle.

Cryo-EM and crystal structures of dynein's AAA+ ring (Carter, Vale, and others, 2011 onward). Structural biology revealed dynein's linker element bending across the AAA+ ring in the ADP/phosphate-bound "pre-powerstroke" state and straightening after phosphate release — directly visualizing the conformational change proposed to be dynein's power stroke, and explaining why dynein's stalk-mediated, allosteric communication between a distant ATPase site and the microtubule-binding domain is structurally so different from kinesin's directly-fused motor/track interface.

Example 1 — how fast can kinesin deliver cargo down a giant axon? A motor neuron axon in the human sciatic nerve runs from the spinal cord to the foot, roughly 1 meter. Fast axonal transport carried by kinesin-1 proceeds at an average measured rate of about 1–2 μm/s in vivo (somewhat below kinesin's unloaded in vitro maximum of ~1 μm/s to ~0.8 μm/s under typical cargo load, but bulk "fast axonal transport" rates measured by radiolabel pulse-chase are often reported around 50–400 mm/day = roughly 0.6–4.6 μm/s, reflecting saturating runs and multiple motors). Take a representative fast transport rate of 200 mm/day.

Compare this to diffusion. For a vesicle of diffusion coefficient D ≈ 10⁻¹⁰ cm²/s in cytoplasm, the mean time to diffuse a distance L scales as t ≈ L²/2D. For L = 1 m = 100 cm:

This enormous gap — 5 days by motor transport versus over a million years by diffusion — is the quantitative reason directional motors are not a luxury but an absolute requirement for large, polarized cells; without kinesin, a neuron literally could not resupply its distal axon or synapse within an organismal lifetime.

Example 2 — a mitotic motor phenotype and how to read it. A researcher generates a cell line with a temperature-sensitive mutation in the dynein heavy chain and observes, at the restrictive temperature, that chromosomes congress to the metaphase plate abnormally, that centrosomes fail to separate properly early in mitosis, and that some chromosomes lag near the spindle poles rather than aligning. Kinesin motors are unaffected. How do we reason through this? Cytoplasmic dynein, anchored at the kinetochore and cell cortex, normally: (1) helps pull duplicated centrosomes apart along microtubules to establish the bipolar spindle, (2) transports the checkpoint proteins away from properly attached kinetochores (silencing the spindle assembly checkpoint), and (3) contributes minus-end-directed forces that help align and segregate chromosomes. Loss of dynein function would be predicted to cause: incomplete centrosome separation (weak spindle bipolarity), persistent checkpoint signaling and mitotic delay (because checkpoint proteins are not stripped from kinetochores), and lagging chromosomes at anaphase (misattachment not corrected). This is exactly the phenotype pattern seen in real experiments using dynein inhibitors (e.g., the small molecule ciliobrevin) or dynein heavy-chain depletion by RNAi — providing a way to distinguish a dynein-specific defect from a general microtubule-destabilizing drug (which would show broadly disorganized, not specifically bipolar-defective, spindles) or a kinesin-5 (Eg5) inhibitor phenotype (which causes monopolar spindles from failure of centrosome separation via a distinct plus-end-directed sliding mechanism, but does not affect checkpoint silencing the same way).

: "All kinesins walk toward the plus end." | Most do, but several kinesin families (e.g., kinesin-14, including Ncd and HSET) are minus-end-directed, and kinesin-13 family members are non-motile microtubule depolymerases rather than transporters at all. | | "One motor protein pulls the cargo alone, unaided." | Many transport events involve competing/cooperating teams of kinesins and dyneins on the same cargo simultaneously (a tug-of-war), with net direction determined by which motor's activity dominates at a given moment, regulated by adaptor proteins. | | "Microtubule motors only matter for vesicle transport." | They also generate mitotic spindle forces, position the nucleus and Golgi, drive ciliary and flagellar beating (via axonemal dynein), and transport mRNA–protein complexes and signaling endosomes. |

Ronald D. Vale's discovery of kinesin grew out of a deceptively simple observation in squid giant axon cytoplasm: organelles kept gliding along microtubule tracks even in a cell-free extract, meaning the motor had to be a discrete, isolatable protein rather than some diffuse property of the living cytoplasm. Working alongside Michael Sheetz and Thomas Reese at the Marine Biological Laboratory in Woods Hole in 1985, Vale purified the responsible ATPase and showed it alone could push microtubules across a glass surface — a moment that transformed intracellular transport from a black box into a biochemical, reconstitutable system. Vale went on to found a career defining the entire kinesin superfamily's diversity of structure and function, later sharing the 2012 Albert Lasker Basic Medical Research Award (with Michael Sheetz and James Spudich) for the discovery of cytoskeletal motor proteins.

Neurodegenerative disease. Mutations in kinesin motor genes (e.g., KIF5A) and in genes encoding dynein or its regulators cause inherited neuropathies and are implicated in amyotrophic lateral sclerosis (ALS) and Charcot-Marie-Tooth disease, where defective axonal transport of mitochondria, mRNAs, and neurotrophic-factor-containing endosomes is thought to underlie the selective vulnerability of long motor and sensory axons.

Ciliopathies. Axonemal dynein defects cause primary ciliary dyskinesia, producing chronic respiratory infections (from immotile airway cilia failing to clear mucus), infertility (immotile sperm flagella), and, in about half of cases, situs inversus (because motile cilia at the embryonic node use dynein-driven beating to establish left-right body asymmetry).

Cancer chemotherapy. The mitotic kinesin Eg5 (kinesin-5), required for bipolar spindle assembly, is a drug target (e.g., ispinesib) developed as an alternative to microtubule-poisons like taxanes, aiming to block cell division specifically during mitosis with potentially fewer off-target neurotoxic effects than drugs that disrupt microtubules generally.

Viral hijacking. Several viruses, including herpesviruses and HIV, exploit dynein-dynactin machinery to travel retrogradely along axons or toward the nucleus after cell entry, making motor-adaptor interactions a target of interest for antiviral strategies.

Biotechnology. Kinesin's well-characterized in vitro motility has been engineered into synthetic nanotransport devices, and dynein/dynactin reconstitution systems are used to screen drugs that stabilize or disrupt specific motor-cargo interactions.

Ask the tutor: How do motor adaptor proteins decide which cargo gets a kinesin, a dynein, or both — and what happens during the resulting "tug-of-war"?