Cytokinesis

Why cytokinesis matters

• It is the last, irreversible commitment of cell division. Mitosis can, in principle, be undone before anaphase; once the furrow closes and abscission cuts the bridge, two independent cells exist. Every one of the roughly 30 trillion cells in your body traces back through an unbroken chain of successful cytokineses from a single zygote.
• Failure makes tetraploid, cancer-prone cells. A cell that finishes mitosis but skips cytokinesis is binucleate and tetraploid, carrying doubled chromosomes and doubled centrosomes. On its next attempt it tends to build a multipolar spindle and scatter chromosomes into aneuploid daughters. Theodor Boveri linked abnormal chromosome number to cancer in 1914, and induced tetraploidy is now known to accelerate tumorigenesis when p53 surveillance is lost.
• Asymmetric cytokinesis builds tissues. By coupling the furrow to spindle position, cells can divide unequally — placing the cleavage plane off-center to give daughters different sizes and different inherited fate determinants. This is how stem cells self-renew while producing a differentiating sister, and how the first cleavages of a worm or fly embryo lay down the body axes.
• Controlled failure is a developmental tool. Cardiomyocytes and hepatocytes deliberately become polyploid as they mature; megakaryocytes undergo endomitosis, doubling their genome up to 128N with no cytokinesis, to build the cytoplasm they later shed as platelets. Skipping cytokinesis on purpose is a normal, programmed choice in these lineages.
• It is a drug target. Because dividing cells depend on it, cytokinesis machinery is exploited therapeutically. Aurora-B inhibitors (e.g. barasertib) and Polo-like-kinase inhibitors force cytokinesis failure and polyploidy in tumor cells, and the natural product blebbistatin — a specific non-muscle myosin II inhibitor — is the standard laboratory tool for blocking furrow contraction.
• The midbody is more than debris. The dense midbody left at the center of the bridge, once dismissed as leftover material, is now known to be inherited asymmetrically, to accumulate on cancer and stem cells, and to signal cell fate — a "midbody remnant" that one daughter often engulfs and that can influence proliferation and differentiation.

How cytokinesis works, step by step

In an animal cell, cytokinesis begins the moment anaphase starts and cyclin-B/Cdk1 activity falls. The trigger is spatial: the mitotic spindle itself tells the cortex where to divide. Two microtubule signals converge on the equator. The central spindle — antiparallel microtubule bundles left between the separating chromosomes, organized by the centralspindlin complex (the kinesin MKLP1 plus the GTPase-activating protein MgcRacGAP) together with the bundler PRC1 — recruits the RhoA guanine-nucleotide exchange factor ECT2 to the midplane. At the same time, dynamic astral microtubules radiating from the two spindle poles suppress cortical contractility near the poles. The net effect is a sharp equatorial band of active, GTP-loaded RhoA, kept crisp by Aurora-B kinase of the chromosomal passenger complex.

RhoA-GTP has two decisive effectors. Through the formin mDia2 (DIAPH3) it nucleates and elongates unbranched actin filaments — not the Arp2/3-branched networks of a lamellipodium but long, linear filaments suited to a contractile band. Through Rho-kinase (ROCK) it phosphorylates the regulatory light chain of non-muscle myosin II and inhibits myosin phosphatase, switching the motor on. Actin and bipolar myosin II filaments, crosslinked and tethered to the membrane by anillin and ringed by septin filaments, assemble into the contractile ring — a band only ~0.1–0.2 µm thick lining the equatorial cortex.

Myosin II motors then slide the antiparallel actin filaments past one another, and the ring constricts. As it tightens it drags the attached plasma membrane inward, forming the cleavage furrow that deepens around the cell like a drawstring closing a purse. Crucially, actin does not pile up as the ring shrinks — it disassembles continuously, so the ring loses material as its circumference falls. The furrow narrows the cell's waist until only a thin intercellular bridge, ~1–2 µm across and stuffed with residual antiparallel microtubules, connects the two daughters. At the center of that bridge sits the dense midbody, or Flemming body.

The last act is abscission, and it is not done by actomyosin. The midbody protein CEP55 recruits ALIX and TSG101, which in turn recruit ESCRT-III subunits (CHMP4B and partners). ESCRT-III polymerizes into helical filaments that spiral out from the midbody and constrict the membrane neck to a scission point, while the AAA-ATPase spastin severs the microtubule bundle. The bridge is cut — usually 1 to 3 hours after the furrow first ingressed — and two fully independent cells finally exist. A surveillance system, the Aurora-B–dependent NoCut/abscission checkpoint, delays this cut if chromatin is still trapped in the bridge, preventing the cell from guillotining its own DNA.

Plant cells solve the same geometric problem in the opposite direction. Locked inside a rigid cellulose wall, they cannot pinch inward. After anaphase a phragmoplast — a bipolar array of microtubules and actin — assembles between the daughter nuclei and guides Golgi- and trans-Golgi-derived vesicles to the midplane, where the plant-specific syntaxin KNOLLE drives their fusion into a flat membranous disc, the cell plate. The plate matures by depositing callose then cellulose and expands centrifugally until its edges fuse with the parental plasma membrane at a site pre-marked, before mitosis, by the preprophase band of microtubules. Same outcome — two walled-off cells — reached by building a wall outward rather than pinching a membrane inward.

Animal vs plant cytokinesis

Feature	Animal (fungal) cytokinesis	Plant cytokinesis
Direction of build	Inward — membrane pinches from the cortex	Outward — new wall grows from the center
Core machine	Actomyosin contractile ring	Phragmoplast (microtubules + actin)
New boundary	Cleavage furrow → intercellular bridge	Cell plate → new cell wall
Force generator	Myosin II sliding actin filaments	Vesicle delivery + membrane fusion
Master regulator	RhoA GTPase (via ECT2)	ROP GTPases, MAP kinases, kinesins
Membrane fusion	Not the driving step (ring constriction is)	KNOLLE syntaxin fuses vesicles into the plate
Division-site memory	Set live by spindle position	Preprophase band marks the site pre-mitosis
Final separation	ESCRT-III abscission + spastin	Plate edges fuse with parental wall

Contractile ring vs abscission — two very different cuts

Property	Furrow ingression (contractile ring)	Abscission (final cut)
What moves the membrane	Actomyosin ring constriction	ESCRT-III filament constriction
Key proteins	RhoA, mDia2, ROCK, myosin II, anillin, septins	CEP55, ALIX, TSG101, CHMP4B, spastin
Structure formed	Cleavage furrow → thin bridge	Membrane scission at the midbody neck
Microtubules	Central spindle organizes the ring position	Bridge microtubules severed by spastin
Timescale	Minutes (couples to anaphase)	~1–3 hours after furrowing
Checkpoint	Spindle-position sensing of division plane	Aurora-B NoCut/abscission checkpoint
Failure outcome	Furrow regression → binucleate cell	Persistent bridge; regression → tetraploidy

Common misconceptions

• "Cytokinesis is just the end of mitosis." It is a mechanically and molecularly distinct process that merely overlaps mitosis in time. Mitosis segregates chromosomes with the spindle; cytokinesis divides the cytoplasm with actomyosin. The two can be uncoupled — a cell can complete mitosis and skip cytokinesis (becoming binucleate) or, in the syncytial early Drosophila embryo, run ~13 mitoses with no cytokinesis at all.
• "The contractile ring squeezes the whole cell shut." The ring only cinches the cell down to a thin bridge; it does not, and cannot, complete the separation. The bridge is severed later by a completely different machine, the ESCRT-III complex, in the process called abscission. Confusing furrow ingression with the final cut is the single most common error.
• "Plant cells divide by pinching in, just slower." Plants never furrow. Boxed in by a cellulose wall, they build the new boundary outward from the middle as a cell plate assembled on a phragmoplast. The machinery (KNOLLE syntaxin, vesicle fusion, ROP GTPases) has little in common with the animal contractile ring.
• "The ring uses the same actin as a crawling cell." The ring's actin is nucleated by formins (mDia2) into long unbranched filaments, not by the Arp2/3 complex that makes the branched networks of lamellipodia. Different nucleator, different architecture, different job.
• "The midbody is just leftover garbage." The midbody is a highly organized structure with over 400 associated proteins, it is inherited asymmetrically, it accumulates on stem and cancer cells, and its remnant carries fate-influencing signals. It is a signaling organelle, not debris.
• "The furrow position is decided by the chromosomes." The furrow is positioned by the spindle microtubules — the central spindle and the astral arrays — not directly by the chromatin. This is why you can push the spindle to one side and the furrow follows it, and why some divisions are deliberately asymmetric.

Famous experiments and history

• Walther Flemming (1878–1882). Using aniline dyes on salamander cells, Flemming coined "mitosis" and drew the earliest accurate pictures of a cell pinching in two. The dense body left in the bridge between daughters still bears his name — the Flemming body, or midbody.
• Marsland and the contractile-ring idea (1950s). Douglas Marsland's pressure experiments on dividing eggs, together with observations that the furrow behaves like a tightening band, led to the "contractile ring" hypothesis: a purse-string of contractile material at the equator.
• Thomas Schroeder's electron microscopy (1968–1972). Schroeder imaged sea-urchin and jellyfish eggs and directly visualized a thin band of aligned actin filaments — roughly 0.1–0.2 µm thick — lining the furrow cortex, and showed it thinned rather than thickened as constriction proceeded. This was the physical proof of the contractile ring.
• Ray Rappaport's torus experiment (1961). By pushing a glass bead into a sand-dollar egg to make it doughnut-shaped, Rappaport forced two spindles to share a cytoplasm and showed that a furrow forms wherever two sets of asters overlap — even where there are no chromosomes. This proved that the spindle microtubules, not the chromosomes, position the cleavage plane.
• C. elegans genetics of the ring. Genetic and RNAi screens in the worm one-cell embryo mapped the core furrow machinery — the centralspindlin subunits ZEN-4 (MKLP1) and CYK-4 (MgcRacGAP), the formin CYK-1, RhoA (RHO-1), ECT-2, and anillin — establishing the RhoA-centered pathway now known to be conserved in humans.
• ESCRT-III and abscission (2007). Work from the Sundquist and Gerlich labs showed that CEP55 recruits the ESCRT machinery (TSG101, ALIX, CHMP4B) to the midbody and that ESCRT-III filaments constrict the bridge to sever it — revealing that the same complex that buds vesicles and HIV particles also performs the final cut of cell division.