The cytoskeleton is an internal network of protein filaments that gives cells their shape, enables them to move, and organizes nearly every major process happening inside them. Far from being a rigid scaffold, it is constantly being built up and torn down, rearranged in response to signals, and used as a highway system by molecular motors hauling cargo. Three main filament types make up the bulk of it, but the full picture includes motor proteins, crosslinkers, signaling hubs, and even connections that reach through the nuclear envelope into the cell’s command center. Understanding what the cytoskeleton does and how it works sheds light on everything from muscle contraction to Alzheimer’s disease to how cancer drugs kill dividing cells.
Three Filament Families, Three Different Jobs
The cytoskeleton is built from three major classes of protein filaments, each with distinct physical properties and cellular roles. Actin filaments (also called microfilaments) are the thinnest, at roughly 7 nanometers in diameter. They concentrate near the cell surface and are the primary drivers of cell shape changes and movement. Microtubules are hollow tubes about 25 nanometers across, built from a protein called tubulin. They radiate outward from a central organizing point near the nucleus and serve as the main tracks for long-distance transport inside the cell. Intermediate filaments fall between the other two in size, at about 10 nanometers, and their defining feature is mechanical toughness rather than dynamic remodeling.
What makes actin filaments and microtubules unusual among structural components is that they are not permanent. Both are assembled from individual protein subunits that snap together and fall apart in a constant cycle. This turnover is fueled by energy-carrying molecules: actin uses ATP, and tubulin uses a related molecule called GTP. The energy isn’t spent to build the filament per se but rather to make the filament unstable enough that the cell can tear it down quickly when needed. Intermediate filaments, by contrast, do not rely on this kind of energy-driven turnover. They assemble through a layered, hierarchical process that produces filaments capable of withstanding extreme stretching and large mechanical loads without breaking.
How Actin Filaments Grow and Shrink
An actin filament grows when individual actin subunits, each carrying a molecule of ATP, add onto one end. Shortly after joining the filament, that ATP gets split into ADP, which changes the subunit’s grip on its neighbors and ultimately makes the older parts of the filament less stable. This happens in two stages: first the ATP is cleaved, and then the leftover phosphate group is slowly released. The cleavage step actually locks the filament together tightly, while the phosphate release loosens it.
Measurements of individual filaments show that phosphate release inside the body of a filament happens randomly, with a half-life of about 102 seconds. At the actively growing tip, though, phosphate release is far faster, with a half-life under half a second when the filament starts shrinking from that end. This difference matters because it means a growing filament carries a temporary “cap” of freshly added, tightly bound subunits at its tip, while the older interior gradually becomes more prone to disassembly.
A helper protein called profilin speeds up actin assembly by loading subunits with fresh ATP and delivering them to the growing end. Profilin also accelerates disassembly of the filament once growth stops. The result is a filament that can elongate rapidly when the cell needs to push a membrane forward but can also be recycled quickly when the push is no longer needed.
Why Microtubules Switch Between Growing and Shrinking
Microtubules display a behavior called dynamic instability: a single microtubule can be growing steadily one moment and then abruptly switch to rapid shrinkage. The switch from growth to shrinkage is called catastrophe, and the reverse switch, from shrinkage back to growth, is called rescue. The standard explanation is that a cap of GTP-bound tubulin at the growing tip holds the structure together, and when that cap is lost, the microtubule falls apart.
That model is broadly correct, but reality is messier. Using a specially designed antibody that recognizes GTP-bound tubulin inside microtubules, researchers found that GTP-tubulin does sit at the growing tip as expected, but also turns up as isolated patches in older stretches of the microtubule. This suggests that GTP hydrolysis is sometimes incomplete during assembly, leaving behind “remnants” that may act as rescue points, allowing a shrinking microtubule to recover and resume growth.
This matters for the cell because microtubules need to be both exploratory and stable. A microtubule growing outward from the cell center can probe different directions, collapse if it doesn’t find a target, and try again. But when it does connect to something important, like a chromosome during cell division, it needs to be stabilized. The interplay between the GTP cap, interior remnants, and various stabilizing proteins gives cells fine-grained control over which microtubules survive and which are recycled.
Molecular Motors and Intracellular Transport
Filaments on their own are structural elements. The cytoskeleton becomes a transport network because of motor proteins: molecular machines that walk along filaments carrying cargo. There are three major families. Myosins walk along actin filaments. Kinesins generally walk toward the growing (plus) end of microtubules, moving cargo outward from the cell center toward the periphery. Dyneins walk the other direction, hauling cargo inward.
In neurons, where the distance between the cell body and a distant synapse can be enormous by cellular standards, microtubule-based transport is essential. Kinesin and dynein motors drive the traffic of vesicles, mitochondria, and signaling molecules up and down the axon. Without this system, synapses would starve for the proteins and energy sources they need.
Myosin’s role in muscle contraction has been studied in extraordinary detail. In striated muscle, thick filaments made mostly of myosin slide past thin filaments made mostly of actin. Each myosin “head” attaches to actin, tilts a lever arm to generate a power stroke, detaches, and resets. Each cycle consumes one ATP molecule. High-resolution measurements show that the working stroke actually happens in two steps: a larger initial displacement of roughly 3 to 5 nanometers, followed by a smaller step of about 1 to 1.3 nanometers. The speed of that second step differs between fast and slow muscle types, which helps explain why different muscles contract at different speeds.
How Cells Crawl
When a cell migrates, whether it is a white blood cell chasing bacteria or a cancer cell invading tissue, the cytoskeleton does the heavy lifting. The front edge of a crawling cell extends a broad, flat protrusion called a lamellipodium. Inside it, a branched network of actin filaments pushes the membrane forward. The branching is created by a protein complex called Arp2/3, which nucleates new actin filaments off the sides of existing ones, generating the dense meshwork that drives the membrane outward.
Arp2/3-driven branched networks are not limited to cell migration. They also generate pushing forces for reshaping membranes around organelles and for moving organelles within the cell. But their role in the lamellipodium is the most visually dramatic: the cell’s leading edge advances because thousands of actin filaments are polymerizing simultaneously, each adding subunits at its tip and collectively producing enough force to deform the membrane.
The Cytoskeleton During Cell Division
Cell division places extreme demands on the cytoskeleton. When a cell enters mitosis, its microtubule network is completely reorganized into the mitotic spindle, a bipolar structure that captures chromosomes and pulls them apart. Each chromosome attaches to spindle microtubules through a protein structure called the kinetochore, which assembles on a specialized region of the chromosome’s DNA. The kinetochore does not just grip the microtubule passively; it maintains a dynamic attachment that can sense tension and correct errors in chromosome alignment.
After the chromosomes are separated, the cell physically pinches in two. This is accomplished by a contractile ring made of actin and myosin that assembles around the cell’s equator and tightens like a drawstring. So both major dynamic filament systems, microtubules and actin, play indispensable roles, with microtubules handling chromosome segregation and actin handling the final physical division.
Cilia and Flagella
Some of the most elegant cytoskeletal structures extend beyond the cell surface. Cilia and flagella are built around an internal scaffold called the axoneme, which typically consists of nine microtubule doublets arranged in a ring. In motile cilia, dynein motor proteins attached between adjacent doublets generate sliding forces that are converted into rhythmic beating or rotational motion. This is the mechanism behind the coordinated sweeping of mucus in your airways and the swimming of sperm cells.
Not all cilia beat. Primary cilia are immotile and instead function as signaling antennae. They lack the dynein motors found in motile cilia and instead concentrate receptors on their surface, shuttling signaling molecules between the ciliary compartment and the rest of the cell. Defects in primary cilia are linked to a range of developmental disorders collectively known as ciliopathies, affecting organs from the kidneys to the brain.
Connecting the Cytoskeleton to the Nucleus
The cytoskeleton does not stop at the nuclear envelope. A set of proteins called the LINC complex (short for Linker of Nucleoskeleton and Cytoskeleton) spans the double membrane of the nuclear envelope, physically coupling cytoplasmic filaments to the nuclear lamina, a meshwork of intermediate-filament-like proteins lining the inside of the nucleus. This connection means that mechanical forces applied to the cell surface can be transmitted all the way to the nucleus, potentially influencing gene activity.
Different nuclear lamin proteins connect to the cytoskeleton in distinct ways. A-type lamins engage with both actin filaments and vimentin intermediate filaments through LINC complexes, affecting the cell’s overall stiffness and contractile force. B-type lamins preferentially interact with intermediate filaments and influence cytoplasmic stiffness. This selectivity means that the mechanical properties of a cell are tuned not only by the cytoskeleton itself but by how the nucleus is wired into it.
Sensing Mechanical Forces
Cells are not passive structures buffeted by their surroundings. They actively sense and respond to mechanical forces, a process called mechanotransduction. Much of this sensing happens at focal adhesions, sites where the actin cytoskeleton is linked to receptors called integrins that grip the extracellular matrix outside the cell. When mechanical stress is applied directly to integrins, the cell responds by stiffening, reinforcing the attachment and adjusting its internal tension. Stress applied to other surface receptors that are not involved in adhesion does not trigger this response, suggesting integrins serve as dedicated mechanoreceptors that channel forces into the cytoskeleton.
Through the LINC complex, these forces reach the nucleus, where they can influence stem cell fate and tissue development. Disrupting the connection between the cytoskeleton and the nucleus compromises a cell’s ability to respond appropriately to its mechanical environment, with consequences for bone remodeling, wound healing, and other processes that depend on cells sensing how stiff or stretched their surroundings are.
When Cytoskeletal Transport Fails in the Brain
The connection between the cytoskeleton and neurodegenerative disease is clearest in Alzheimer’s disease. A protein called tau normally binds to microtubules in axons and helps stabilize them. In Alzheimer’s, tau becomes abnormally phosphorylated, which weakens its grip on microtubules. The detached, heavily phosphorylated tau then aggregates into the neurofibrillary tangles that are a hallmark of the disease.
But the damage starts well before tangles form. Elevated tau phosphorylation reduces microtubule binding, freeing tau to drift along the axon rather than stabilizing the tracks that motor proteins depend on. Even more concerning, misregulated tau can physically block microtubule-based transport of vesicles and organelles, starving synapses of the supplies they need to function and increasing oxidative stress. Studies in fruit flies carrying human tau show that highly phosphorylated tau also sequesters the cell’s normal tau away from microtubules, compounding the damage and leading to visible disruption of the microtubule network, with misaligned and reduced numbers of intact microtubules in the axon.
Cancer Drugs That Target Microtubules
The dependence of cell division on microtubule dynamics makes them an attractive target for cancer therapy. Drugs that interfere with microtubule assembly or disassembly are among the most widely used classes of chemotherapy agents. These microtubule-targeting agents work by disrupting the normal function of the mitotic spindle, which activates a checkpoint that arrests the cell cycle and can trigger cell death. Even subtle alterations of microtubule dynamics, not necessarily complete destruction of the filaments, can be enough to engage this checkpoint and block proliferation.
Many of these drugs come from natural sources. Taxol (paclitaxel), originally derived from the bark of the Pacific yew tree, stabilizes microtubules and prevents their disassembly. Vincristine, from the periwinkle plant, does the opposite, preventing assembly. Both strategies are lethal to rapidly dividing cells because the mitotic spindle requires precisely controlled microtubule dynamics to function. Seven distinct binding sites on tubulin have been identified where different drugs can dock, offering multiple chemical strategies for disrupting microtubule behavior in cancer cells.
The Cytoskeleton in Plant Cells
Plant cells face a structural challenge that animal cells do not: they are encased in a rigid cell wall made largely of cellulose. The cytoskeleton plays a direct role in building that wall. Cellulose is synthesized by enzyme complexes embedded in the plasma membrane, and these complexes move along the membrane surface as they spin out cellulose fibers. The tracks they follow are set by cortical microtubules lying just beneath the membrane.
Live imaging of fluorescently labeled cellulose synthase complexes showed them traveling at constant speeds along paths aligned with cortical microtubules. When researchers disrupted microtubule assembly, the pattern of cellulose deposition changed, confirming that the cytoskeleton guides where new cell wall material is laid down. A chemical probe called morlin, which interferes with cortical microtubule dynamics, similarly altered the movement and distribution of cellulose synthase complexes. Because the orientation of cellulose fibers determines how a plant cell can expand, the cytoskeleton effectively controls the direction of plant growth.
Bacteria Have Cytoskeletons Too
For decades, the cytoskeleton was considered a uniquely eukaryotic invention. That turned out to be wrong. Bacteria possess distant homologs of all three major eukaryotic filament types. The bacterial protein MreB is structurally similar to actin despite sharing little amino acid sequence similarity, and it assembles into dynamic filaments that help determine bacterial cell shape. ParM, another bacterial actin relative, segregates plasmid DNA during cell division. FtsZ, a tubulin homolog, forms a ring at the site where the bacterium will divide and helps recruit the machinery that pinches the cell in two. Bacteria even have a protein group, MinD-ParA, that appears to have no eukaryotic equivalent.
The existence of these proteins suggests that the cytoskeleton predates the split between bacteria and the organisms that eventually gave rise to animals, plants, and fungi. The bacterial actin-like cytoskeleton is thought to serve as a spatial organizer, positioning proteins and DNA complexes within the cell in a way that parallels how the eukaryotic cytoskeleton organizes vastly more complex cellular interiors.
Beyond the Big Three
The traditional picture of three filament types is an oversimplification. Septins, a family of GTP-binding proteins, form filaments and rings that function as diffusion barriers and scaffolds. In budding yeast, septins at the neck between a mother cell and its bud create a barrier that prevents membrane proteins from mixing freely between the two compartments, maintaining distinct plasma membrane identities. In mammalian neurons, analogous mechanisms may operate at dendritic branch points and spine necks, compartmentalizing signaling within different regions of a single neuron.
The cytoskeleton also depends heavily on crosslinker proteins that tie different filament systems together. Plakins are a family of giant modular proteins that act as molecular bridges between actin filaments, intermediate filaments, and microtubules, as well as connecting them to adhesion complexes at the cell surface. When this crosslinking fails, epithelial tissues lose their structural integrity, a pattern seen in certain skin-blistering diseases where the mechanical linkage between the cytoskeleton and cell-cell junctions is compromised.
Seeing the Cytoskeleton at Nanometer Scale
Much of what we know about cytoskeletal architecture was invisible until recently. Conventional light microscopy cannot resolve structures smaller than about 200 nanometers, which is larger than most individual cytoskeletal filaments. Super-resolution microscopy techniques, developed over the last two decades and recognized with a Nobel Prize in 2014, broke through this barrier. These methods allow researchers to image individual filaments and their associated proteins in three dimensions, in multiple colors, and even in living cells, with resolution measured in tens of nanometers rather than hundreds.
This technological leap has been especially productive for studying the cytoskeleton because so many of its important features exist right at the old resolution limit. The spacing of actin filaments in a stress fiber, the arrangement of proteins within a focal adhesion, the organization of axonemal components inside a cilium: all of these are now directly observable rather than inferred from indirect experiments. The ability to watch cytoskeletal dynamics in real time in living cells has shifted the field from static snapshots to something closer to watching the machinery in action.
Viruses and the Hijacked Highway
Viruses, which carry minimal equipment of their own, are masters at co-opting the host cell’s cytoskeleton. After entering a cell, many viruses ride microtubule-based motor transport to reach the nucleus, where they replicate. On the way out, newly assembled viral particles may hitch rides back to the cell surface along the same tracks. Some viruses manipulate actin dynamics to help them push through the cell membrane during budding, or even to propel themselves directly from one cell into an adjacent one, evading the immune system by never entering the extracellular space.
Studying how viruses interact with the cytoskeleton has been a two-way street. Researchers learn about viral strategies, but they also uncover details of normal cytoskeletal regulation that were not obvious from studying healthy cells. A virus that can selectively activate one motor protein over another, or redirect actin polymerization to a specific membrane site, reveals regulatory switches that the cell itself uses under different circumstances.