Cytokinesis is the physical splitting of one cell into two, and it is the final act of cell division. After a cell copies its DNA and separates its chromosomes, it still has to divide the cytoplasm, membrane, and organelles into two independent daughter cells. In animal cells, this happens through an inward-pinching groove called the cleavage furrow, powered by a belt of protein filaments that tightens like a drawstring. The process is more involved than that simple image suggests, though, because the cell has to choose precisely where to cut, build molecular machinery on the fly, deliver fresh membrane to the growing furrow, and then sever the last physical connection between the two daughters without damaging either one.
Choosing Where to Cut
The first challenge is spatial: the cell needs the furrow to form at the right spot, directly between the two sets of separated chromosomes. The signal comes from structures left behind by the mitotic spindle, particularly a dense bundle of overlapping microtubules between the chromosome masses called the central spindle. A protein complex called centralspindlin accumulates on these midzone microtubules and at the tips of other microtubules that reach toward the cell cortex. Centralspindlin then recruits a signaling protein called ECT2 to the equatorial region of the cell membrane. ECT2 is a guanine nucleotide exchange factor, which in plain terms means it flips a molecular switch: it activates a small signaling molecule called RhoA right at the equator.
This local burst of RhoA activity is what tells the cortex to start building a contractile ring in that specific band of membrane and nowhere else. Studies in human cells have shown that when ECT2 is removed or its ability to reach the membrane is disrupted, RhoA fails to accumulate at the equator and no contractile ring forms.1Developmental Cell. Spatiotemporal Control of Ect2 Localization and Function at the Equatorial Cortex Coordinates Cleavage Furrow Formation with Chromosome Segregation Both the central spindle microtubules and astral microtubules that reach toward the cortex contribute to positioning ECT2, creating a redundant system that makes the equatorial signal robust.2Journal of Cell Science. Centralspindlin regulates ECT2 and RhoA accumulation at the equatorial cortex during cytokinesis The concentration of RhoA is maintained as the furrow deepens because centralspindlin-ECT2 complexes on the central spindle stay in contact with the inward-folding membrane, continuously reinforcing the signal.
Assembling the Contractile Ring
Once RhoA is activated at the equator, it triggers two things simultaneously: the rapid assembly of actin filaments and the activation of myosin II motor proteins. Formin proteins, switched on by RhoA, nucleate and elongate new actin filaments. Myosin II, also activated downstream of RhoA, slides along those actin filaments in much the same way it does during muscle contraction. Together, the actin and myosin form a ring-shaped band that runs around the cell’s equator just beneath the plasma membrane.3Current Biology. Anillin Is a Scaffold Protein That Links RhoA, Actin, and Myosin during Cytokinesis
Holding this ring together requires scaffolding. A protein called anillin binds to RhoA, actin, and myosin simultaneously, keeping active myosin anchored in the equatorial plane. Without anillin, myosin drifts away from the furrow and the ring falls apart. You can think of anillin as the structural glue that prevents the contractile machinery from sliding out of position as it generates force.
How the Ring Squeezes Shut
The contractile ring does not simply tighten the way you might cinch a drawstring bag. As it constricts, it also has to shrink, shedding components as its circumference decreases. Modeling work in budding yeast has shown that two forces drive the constriction: myosin II motors pulling actin filaments past each other, and the controlled disassembly of actin filaments themselves, which shortens the ring from within.4PubMed Central. Actin depolymerization drives actomyosin ring contraction during budding yeast cytokinesis Cells that cannot depolymerize actin properly constrict much more slowly, suggesting that the breakdown of filaments is just as important as motor-driven sliding.
Meanwhile, the deepening furrow creates a practical problem: as the ring pulls the membrane inward, the cell needs more membrane surface area at the cleavage site than it had before. Recycling endosomes, compartments inside the cell that shuttle membrane between internal stores and the surface, deliver fresh membrane to the furrow.5PubMed Central. Endosomal recycling regulation during cytokinesis Without this membrane delivery, the furrow cannot complete ingression because the surface simply runs out of material to fold inward.
The Midbody and the Final Cut
After the ring constricts fully, the two daughter cells are not yet independent. They remain connected by a thin intercellular bridge packed with bundled microtubules and a dense protein structure at its center called the midbody, sometimes referred to as the Flemming body. The midbody acts as a signaling platform that organizes the last step of division: abscission, the physical severing of the membrane bridge.
At the midbody, components like MKLP1 (part of centralspindlin) interact with the small regulatory protein Arf6, forming a complex that links the microtubule bundle to the overlying membrane.6PubMed Central. Structural basis for Arf6-MKLP1 complex formation on the Flemming body responsible for cytokinesis This connection helps maintain the bridge’s architecture while the abscission machinery assembles.
The actual membrane cut is carried out by a set of proteins called the ESCRT-III complex. These proteins form spiral-like filaments that constrict the membrane tube from the inside, eventually pinching it off. ESCRT-III is, in essence, a transplantable membrane-cutting machine that the cell deploys wherever it needs to sever a membrane tube, whether during cytokinesis, viral budding, or nuclear envelope repair.7PubMed Central. ESCRT-dependent control of membrane remodelling during cell division Once ESCRT-III completes its work, the two daughter cells are fully separated.
A Safety Checkpoint Before Abscission
Cells do not rush into abscission blindly. If chromosomes are still tangled in the cleavage plane, or if the spindle midzone is defective, cutting the bridge could shear DNA and kill both daughters. To prevent this, cells have an abscission checkpoint called the NoCut pathway. Research in yeast demonstrated that when midzone defects are detected, the Aurora kinase Ipl1 and anillin-related proteins Boi1 and Boi2 actively delay abscission until the problem is resolved.8Cell. The NoCut Pathway Links Completion of Cytokinesis to Spindle Midzone Function to Prevent Damage Cells that lacked this checkpoint suffered frequent chromosome breakage and died at much higher rates. In human cells, a similar Aurora B-dependent checkpoint operates at the midbody, ensuring that lagging chromosomes clear the bridge before the membrane is severed.
Signaling the Start of the Whole Process
The timing of cytokinesis is tightly linked to the earlier stages of mitosis through a master regulator called Cdk1 (cyclin-dependent kinase 1), paired with its partner cyclin B1. During early mitosis, Cdk1 activity is high and actively prevents premature assembly of the cytokinesis machinery. For example, Cdk1 phosphorylates the motor protein MKlp2, which is needed to relocate a key regulatory complex called the chromosomal passenger complex (CPC) from the chromosomes to the central spindle. As long as Cdk1 is active, MKlp2 cannot bind microtubules or interact with the CPC. When the cell transitions into anaphase and Cdk1 activity drops, those inhibitory marks are removed, allowing MKlp2 to bind the CPC and ferry it to the central spindle where it is needed for furrow specification.9PubMed. Cdk1 coordinates timely activation of MKlp2 kinesin with relocation of the chromosome passenger complex for cytokinesis This mechanism ensures that cytokinesis cannot begin until chromosome separation is well underway.
How Plant Cells Divide Differently
Plant cells have a rigid cell wall, so pinching inward the way animal cells do is not an option. Instead, plant cytokinesis builds a new wall from the inside out. After the chromosomes separate, a structure called the phragmoplast assembles in the middle of the cell. The phragmoplast is a scaffold of microtubules and actin filaments oriented with their growing ends facing the cell’s equator. Vesicles produced by the Golgi apparatus travel along these microtubules toward the center, where they fuse with one another to form a new membrane-bounded structure called the cell plate.10PubMed Central. Dynamics of phragmoplastin in living cells during cell plate formation and uncoupling of cell elongation from the plane of cell division
The cell plate starts as a disc of fused vesicles in the center and grows outward toward the existing cell wall. Early in its formation, it has a tubular, honeycomb-like structure as the vesicles merge in a star-shaped pattern sandwiched between the ends of phragmoplast microtubules.11PubMed. Cytokinesis and Building of the Cell Plate in Plants As more vesicle material is added and the plate matures, it flattens into a continuous sheet. Once the cell plate reaches and fuses with the parent cell’s plasma membrane at every edge, the two daughter cells are sealed off from each other, each with its own complete membrane and a shared new wall between them.12Trends in Plant Science. Vesicle dynamics during cell-plate formation in plants
Bacterial Cell Division and the Z Ring
Bacteria lack the actin-myosin machinery that animal cells use, but they face the same basic problem: splitting one cell into two. They solve it with FtsZ, a protein that is the evolutionary ancestor of tubulin, the building block of microtubules in animal and plant cells. FtsZ assembles into a ring structure, called the Z ring, at the future division site.13PubMed Central. Bacterial cytokinesis: From Z ring to divisome Another protein, FtsA, which belongs to the actin family, helps anchor the Z ring to the inner membrane and recruits a suite of downstream proteins that together form a multi-protein machine called the divisome.
The divisome’s main job is to build new cell wall (peptidoglycan) inward from the existing wall, progressively constructing a septum that splits the cell. Live imaging of bacterial division has revealed that FtsZ filaments treadmill around the ring, moving circumferentially along the division plane. This treadmilling motion drives the peptidoglycan-building enzymes along the ring, ensuring that new wall material is inserted evenly in increasingly smaller concentric rings until the cell is divided.14PubMed Central. Treadmilling by FtsZ filaments drives peptidoglycan synthesis and bacterial cell division Changing the speed of FtsZ treadmilling changes the rate of wall synthesis and, consequently, the rate of division itself.15PubMed Central. FtsZ dynamics in bacterial division: What, how, and why?
When Cytokinesis Fails
Because cytokinesis is the step that physically separates the two chromosome sets into individual cells, failure has serious consequences. If the cleavage furrow regresses or abscission does not complete, the result is a single cell with twice the normal chromosome number. Long-term imaging of mouse ovarian surface cells demonstrated that these tetraploid cells, arising from failed cytokinesis during otherwise normal bipolar division, subsequently mis-segregated chromosomes during later divisions, producing cells with abnormal chromosome counts.16PubMed Central. Tetraploid cells from cytokinesis failure induce aneuploidy and spontaneous transformation of mouse ovarian surface epithelial cells Those aneuploid cells went on to transform spontaneously, becoming cancerous. This chain of events, from a single failed cytokinesis to genome instability to tumor formation, is one reason researchers study cytokinesis not only as a basic cell biology question but as a cancer-relevant one.
Incomplete Cytokinesis on Purpose
Not every incomplete cytokinesis is a mistake. In certain tissues, cells deliberately stop the process partway through, leaving stable bridges called ring canals that connect groups of cells into a shared network. This happens in both the male and female germline across a wide range of species, from insects to humans.17PubMed Central. Structure and functions of stable intercellular bridges formed by incomplete cytokinesis during development In the fruit fly ovary, for example, a single egg cell and its 15 nurse cells remain connected through ring canals, allowing nutrients and regulatory molecules to flow freely between them.
Time-lapse imaging in fruit flies has shown that ring canals form through extensive remodeling of the midbody. Instead of being discarded after abscission as it would be in a normal division, the midbody ring stabilizes and is reinforced, eventually transforming into a permanent open channel. This midbody-to-ring canal transformation depends on a kinase called Citron kinase and has been observed not only in fruit fly females and males but also during mouse and Hydra spermatogenesis, suggesting it is an ancient and conserved mechanism.18Developmental Cell. Germline ring canals originate from stabilized midbody rings
Fungal Cytokinesis and the Double Requirement
Fungi occupy an interesting middle ground. Like animal cells, fission yeast and other fungi assemble an actomyosin ring to drive furrow ingression. But unlike animal cells, they also build a septum, a wall-like partition across the division plane, simultaneously with ring constriction. Neither the ring nor the septum alone is sufficient. The actomyosin ring is essential for initiating and guiding furrow formation, but the septum provides the structural material that completes the partition. How the cell coordinates these two processes in time and space remains an open question.19PubMed Central. Coordinating Septum Formation and the Actomyosin Ring during Cytokinesis in Schizosaccharomyces pombe
Dividing Inside a Tissue
Most textbook descriptions of cytokinesis imagine a cell dividing in isolation, floating freely in culture medium. Cells in a living tissue face a different reality: they are glued to their neighbors by junctions, and dividing without breaking the tissue apart requires careful coordination. In vertebrate epithelial tissue, the cell-to-cell junctions that maintain the tissue’s barrier function are not simply torn apart during division. Tight junctions remain intact throughout, keeping the tissue sealed even as the dividing cell changes shape dramatically.20PubMed Central. Maintenance of the Epithelial Barrier and Remodeling of Cell-Cell Junctions during Cytokinesis Adherens junctions at the cleavage furrow are stabilized by increased mechanical tension, preventing them from dissolving prematurely.
Forming new junctions between the two daughters is equally intricate. The dividing cell must disengage its adhesion complexes from the neighboring cell at the furrow site so the daughters can build a fresh junction between themselves. Septin proteins, which form filaments at the furrow, provide the mechanical force needed to overcome the pulling tension exerted by neighboring interphase cells.21Developmental Cell. Septin-Dependent Remodeling of Adherens Junctions Regulates Epithelial Cell Cytokinesis This makes epithelial cytokinesis a genuinely multicellular process: the neighbors actively participate, and the outcome depends on a balance of forces between the dividing cell and the cells surrounding it.22PubMed. Interplay between the dividing cell and its neighbors regulates adherens junction formation during cytokinesis in epithelial tissue
Ancient Roots in Archaea
The ESCRT-III proteins that sever the intercellular bridge during animal cell abscission are not a eukaryotic invention. Archaea, single-celled organisms that split from the lineage leading to complex life billions of years ago, use closely related proteins to divide. Structural studies have confirmed that the archaeal CdvB proteins are genuine ESCRT-III relatives, sharing the same ability to assemble into spiraling filaments that engage and remodel membranes.23PubMed Central. Molecular structure of the ESCRT-III-based archaeal CdvAB cell division machinery Physical modeling and live imaging of dividing archaea suggest that changes in the intrinsic curvature of these elastic filaments cause them to supercoil, driving ring constriction and deforming the membrane until the cell splits.24PubMed Central. Physical mechanisms of ESCRT-III-driven cell division
Particularly striking are the Asgard archaea, currently considered the closest prokaryotic relatives of eukaryotes. Their ESCRT-III proteins self-assemble into helical filaments that can bind and deform membrane vesicles made of eukaryotic-type lipids, which are chemically quite different from typical archaeal membranes.25PubMed Central. The Asgard archaeal ESCRT-III system forms helical filaments and remodels eukaryotic-like membranes This cross-compatibility suggests that the membrane-cutting toolkit used in the last step of human cell division may trace back to a common ancestor shared with archaea, making the ESCRT system one of the most ancient pieces of cell division machinery still in operation.
Watching Cytokinesis in Real Time
Much of what we know about the molecular architecture of the contractile ring comes from advances in live-cell imaging. Super-resolution fluorescence microscopy has pushed the boundaries of what can be resolved in living, dividing cells. In fission yeast, a technique called high-speed fluorescence photoactivation localization microscopy (FPALM) revealed that the precursor structures to the contractile ring, called cytokinesis nodes, are discrete structural units with defined protein compositions and specific spatial arrangements.26PubMed Central. Molecular organization of cytokinesis nodes and contractile rings by super-resolution fluorescence microscopy of live fission yeast These nodes coalesce into the ring, and imaging them at this resolution showed that the ring is not a simple homogeneous cable but a structured assembly with specific protein positions. The ability to watch these dynamics in living cells, rather than fixed snapshots, has reshaped models of how the ring contracts and disassembles, and continues to generate surprises about a process that textbooks once treated as straightforward.