The mitotic spindle begins to form during prophase, the first stage of mitosis, when centrosomes separate and microtubules grow outward to start building the bipolar framework. It reaches its mature, fully bipolar form by metaphase, roughly 20–40 minutes after prophase onset in a typical human cell. Completion of the spindle assembly checkpoint then commits the cell to chromosome segregation.

That answer hides a scheduling surprise. The spindle's structural components are largely built during interphase, but the machine only becomes a functional segregation apparatus after nuclear envelope breakdown (NEBD) releases chromosomes into the cytoplasm. Understanding why requires separating what is assembled from what is activated.

Prophase: the first microtubule arrays

Prophase is when the mitotic spindle starts to form in the strict sense. Two centrosomes, each a pair of centrioles surrounded by pericentriolar material, begin to separate along the nuclear envelope. Microtubule nucleation from these centrosomes increases sharply, and the resulting arrays probe the cytoplasm.

The separation is driven by motor proteins pushing antiparallel microtubules apart, plus actin-based forces at the cell cortex. Aurora A kinase and PLK1 amplify centrosome maturation, recruiting gamma-tubulin ring complexes that cap microtubule minus ends. More nucleation sites mean more microtubules, and more microtubules mean a denser search-and-capture network.

Prophase lasts roughly 30–60 minutes in cultured human cells, though this varies widely by cell type and temperature. It ends with NEBD, the event that makes the spindle truly functional. Before NEBD, microtubules cannot access kinetochores, so the spindle is a scaffold without cargo.

The interphase groundwork most explainers skip

Here is the counterintuitive detail: much of the spindle's raw material is manufactured before mitosis even begins. Centrosome duplication starts in G1 and finishes in S phase, giving the cell exactly two centrosomes by G2. Duplication is semiconservative, with each daughter centriole templating a new one.

Microtubule nucleation capacity also rises through interphase. Gamma-tubulin accumulates at centrosomes, and microtubule stability changes as the cell approaches mitosis. By the G2/M transition, the cell has stockpiled the components but keeps them in a relatively inactive state.

This matters for the keyword question. If you ask when the spindle forms, the honest answer splits: its precursors assemble in interphase, its first arrays appear in prophase, and its mature bipolar architecture appears in prometaphase and metaphase. A single phase label oversimplifies a staged assembly process.

Prometaphase and metaphase: from arrays to a bipolar machine

Prometaphase begins at NEBD. Microtubules now contact kinetochores, protein structures on centromeric chromatin that bind up to roughly 20–25 microtubules per kinetochore in human cells. Initial attachments are often lateral and frequently incorrect, including attachments to the same pole from both sister kinetochores.

Error correction is active and fast. Aurora B kinase phosphorylates kinetochore substrates to destabilize low-tension attachments, forcing turnover until each sister kinetochore binds microtubules from opposite poles. This is the mechanical basis of amphitelic attachment, the only configuration that satisfies the spindle assembly checkpoint.

By metaphase, chromosomes align at the spindle equator, roughly equidistant from the two poles. The spindle is now fully bipolar, with kinetochore microtubules, interpolar microtubules, and astral microtubules all contributing distinct functions. Metaphase typically lasts about 20–40 minutes in human somatic cells.

Timeline of spindle assembly

Stage Approximate timing in human somatic cells Key spindle event
G1 Hours (variable) Centriole disengagement; licensing for duplication
S phase Hours (variable) Centrosome duplication completes
G2 2–4 hours Gamma-tubulin accumulates; nucleation capacity rises
Prophase 30–60 minutes Centrosome separation; first microtubule arrays
Prometaphase 10–30 minutes after NEBD Kinetochore capture; attachment error correction
Metaphase 20–40 minutes Bipolar spindle mature; chromosomes aligned
Anaphase onset After checkpoint satisfaction Spindle elongates; chromatids separate

These durations are approximate and vary by cell type, organism, and culture conditions. The ordering is conserved across eukaryotes, but absolute times are not.

Molecular forces that shape the spindle

Spindle architecture emerges from force balance, not a blueprint. Kinesin-5 motors slide antiparallel interpolar microtubules apart, pushing poles away from each other. Dynein anchored at the cortex pulls astral microtubules outward, and minus-end-directed motors such as kinesin-14 pull poles together.

The spindle matrix, a poorly defined but real assembly of proteins including NuMA and dynein, contributes to pole focusing. Without proper focusing, spindles become splayed or multipolar, which can produce chromosome missegregation. Multipolar spindles are a hallmark of some cancer cells and a target of anti-mitotic drugs.

Microtubule dynamic instability sets the search-and-capture timescale. Individual microtubules grow and shrink at rates of roughly 10–30 micrometers per minute in many human cell types, though these rates vary. A cell roughly 20 micrometers across can be searched in minutes, not hours.

Misconceptions about spindle timing

A common misconception is that the spindle forms entirely during metaphase. In reality, prophase builds the first arrays, prometaphase builds bipolarity, and metaphase is when the mature structure is maintained and monitored. Metaphase is the end of assembly, not the beginning.

Another misconception is that the spindle is complete before NEBD. It is not functional for chromosome capture until the nuclear envelope opens. Some cells, including certain oocytes and plant cells, use acentrosomal or nuclear-envelope-associated microtubule nucleation, showing that the prophase-centrosome model is not universal.

A third error is treating the spindle as static. It is a steady-state structure with constant microtubule turnover, with half-lives measured in tens of seconds to a few minutes for many spindle microtubules. What looks solid under a microscope is dynamic at the molecular level.

What goes wrong when assembly fails

Spindle assembly errors trigger the spindle assembly checkpoint, which delays anaphase until every kinetochore is properly attached. The checkpoint is not a single protein but a signaling cascade built around Mad and Bub proteins. Its failure produces aneuploidy, a common feature of cancer cells.

Persistent kinetochore attachment errors can cause chromosome missegregation and aneuploid daughter cells. Multipolar spindles can produce fragmented nuclei and nonviable progeny. Prolonged checkpoint activation can drive cells into mitotic slippage or apoptosis. Premature anaphase onset before checkpoint satisfaction is a recognized aneuploidy mechanism. Spindle defects are a target of chemotherapy agents such as taxanes and vinca alkaloids.

These outcomes are well established in cell biology, but the specific frequency of each depends on cell type, genetic background, and dose of any perturbing agent.

Practical relevance of spindle timing

For researchers timing experiments, prophase is when centrosome separation and initial nucleation should be assayed. Prometaphase is the window for attachment dynamics, and metaphase is the window for steady-state spindle maintenance. Treating mitosis as a single time point loses this structure.

For students preparing for exams, remember the sequence: prophase starts the spindle, prometaphase captures chromosomes, metaphase aligns them. If a question asks during which phase the spindle forms, the accurate short answer is prophase, with the caveat that functional bipolarity is achieved later.

For clinicians, spindle poisons act largely during prometaphase and metaphase, not prophase. This timing underlies the rationale for prolonged exposure in some regimens, though dosing schedules are outside the scope of this article. Always consult current prescribing information and institutional protocols for clinical decisions.

Limits of the answer and what varies

The standard model describes a typical dividing human somatic cell in culture. Timing and even the sites of microtubule nucleation differ in oocytes, early embryos, plant cells, and many protists. Acentrosomal spindles in oocytes rely on Ran-GTP gradients and chromatin-mediated nucleation, so the prophase-centrosome account does not fully apply.

Some cells also use a nuclear envelope-associated spindle, and certain fungi use a spindle that forms inside the nucleus without full NEBD. The keyword question expects a standard eukaryotic answer, but the exceptions are real and mechanistically interesting. Precision matters here, and overgeneralizing from one model system is a common professional caveat.

Beyond these variations, the core sequence is robust: prophase initiates spindle assembly, prometaphase establishes bipolarity through kinetochore attachments, and metaphase maintains the mature apparatus under checkpoint surveillance. That sequence is the reliable takeaway for study and experimental design.