Sister chromatids move apart during anaphase, the fourth phase of mitosis, after the spindle assembly checkpoint is satisfied and separase cleaves cohesin. At anaphase onset, the protease separase cuts the cohesin ring holding sister chromatids together, and poleward pulling by kinetochore microtubules plus spindle elongation drives each chromatid toward its opposite centrosome. Anaphase completes in roughly 6 to 10 minutes in typical human cultured cells, making it one of the shortest mitotic phases.
The separase trigger that starts anaphase
The physical separation of sister chromatids is not gradual. It is a switch. Throughout prometaphase and metaphase, sister chromatids stay glued at their centromeres by cohesin, a ring-shaped protein complex that topologically encircles both DNA duplexes. Cohesin is not removed everywhere at once; most of it is stripped from chromosome arms during prophase by phosphorylation, but a protected pool remains at centromeres. That centromeric pool is the actual lock.
The key is that separase, the enzyme that cleaves cohesin, is held inactive by securin until the spindle assembly checkpoint (SAC) clears. The SAC monitors kinetochore attachment. If even one kinetochore is unattached or lacks tension, the checkpoint generates the mitotic checkpoint complex, which inhibits the anaphase-promoting complex (APC/C). Once every kinetochore is properly attached to microtubules from opposite poles, the checkpoint silences, APC/C becomes active, and it tags securin for degradation by the proteasome.
Degrading securin frees separase. Separase then cleaves the cohesin subunit RAD21 at centromeres, opening the ring. The two chromatids, until now one duplicated chromosome, become independent chromosomes. In human cells this cleavage is the point of no return; once cohesin opens at centromeres, anaphase proceeds even if spindle attachments are later disturbed.
Microtubules supply the force, not the trigger. Kinetochore microtubules shorten at their plus ends while depolymerizing, pulling chromatids poleward. Meanwhile, interpolar microtubules slide past each other and push the poles apart. Both motions overlap in time.
How fast chromatids actually travel
Anaphase splits into two mechanically distinct subphases. Anaphase A is chromatid-to-pole movement; anaphase B is pole-to-pole spindle elongation. Anaphase A moves chromatids at roughly 1 to 3 micrometers per minute in mammalian cells. Anaphase B separates the poles at roughly 1 to 2 micrometers per minute, and the two contributions add.
Scale matters. A typical human metaphase chromosome is about 1 to 2 micrometers long, and the spindle is roughly 10 to 15 micrometers across. Chromatids must travel several micrometers to reach the poles. At 1 to 3 micrometers per minute, most chromatid-to-pole movement completes within a few minutes, consistent with the 6 to 10 minute anaphase observed in many cultured human cell lines.
| Mitotic phase | Main event | Approximate duration (cultured human cells) |
|---|---|---|
| Prophase | Chromatin condenses; centrosomes separate | 10 to 30 minutes |
| Prometaphase | Nuclear envelope breaks down; kinetochores attach | 10 to 20 minutes |
| Metaphase | Chromosomes align at the equator | 10 to 20 minutes |
| Anaphase | Sister chromatids separate and move to poles | 6 to 10 minutes |
| Telophase | Nuclear envelopes reform; chromosomes decondense | 5 to 15 minutes |
| Cytokinesis | Cytoplasm divides | Overlaps telophase; tens of minutes |
Durations vary widely by cell type, temperature, and organism. These values describe typical adherent human cell lines, not every tissue. Embryonic and plant cells can run much faster. Treat the numbers as ranges, not fixed constants.
| Feature | Anaphase A | Anaphase B |
|---|---|---|
| Direction of motion | Chromatids toward poles | Poles away from each other |
| Primary driver | Kinetochore microtubule shortening | Interpolar microtubule sliding and plus-end growth |
| Typical rate | 1 to 3 micrometers per minute | 1 to 2 micrometers per minute |
| Molecular motors | Kinetochore-associated motors and microtubule depolymerization | Bipolar kinesins and dynein at the spindle midzone |
| Timing | Begins at anaphase onset | Overlaps and often outlasts anaphase A |
Timeline from checkpoint to two nuclei
Anaphase is a brief window inside a longer sequence. The following timeline tracks the events that lead from a satisfied checkpoint to separated chromatid sets, using typical mammalian cell timing.
| Time from anaphase onset | Event |
|---|---|
| Roughly minus 2 to 5 minutes | Spindle assembly checkpoint satisfied; APC/C activated |
| 0 minutes | Securin degraded; separase cleaves centromeric cohesin |
| 0 to 1 minute | Sister chromatids visibly separate; anaphase A begins |
| 1 to 4 minutes | Chromatids move poleward at 1 to 3 micrometers per minute |
| 2 to 6 minutes | Anaphase B spindle elongation peaks |
| 5 to 10 minutes | Chromatids cluster near poles; anaphase ends |
| 6 to 15 minutes | Telophase begins; nuclear envelope reassembly starts |
The precise minutes are cell-type dependent. The ordering is not. Checkpoint satisfaction always precedes cohesin cleavage, and cohesin cleavage always precedes poleward chromatid motion.
Edge cases that bend the standard answer
The textbook sequence assumes all chromatids separate simultaneously. Biology is messier. One well-documented exception is premature sister chromatid separation, where cohesin fails early and chromatids drift apart before metaphase alignment. This produces aneuploid daughter cells and is a recognized mechanism in some cancer genomes. Separase overexpression or securin loss can drive this.
A second edge case is merotelic attachment, where one kinetochore binds microtubules from both poles. The SAC does not always detect this, so anaphase can begin with a lagging chromosome that trails behind the others. Lagging chromosomes are a common cause of micronuclei formation after mitosis. The chromosomes still move apart in anaphase, but one arrives late.
A counterintuitive professional caveat: anaphase looks like the end of chromosome segregation, but most segregation errors are decided earlier, at attachment. By the time chromatids visibly separate, the error is typically locked in. Microscopists who score anaphase figures for lagging chromosomes are reading a downstream symptom, not the root cause.
In meiosis, the situation differs. Meiosis I separates homologous chromosomes, not sister chromatids. Sister chromatids separate in meiosis II, in a division that resembles mitotic anaphase. This distinction trips up students and even some annotations.
Misconceptions worth correcting
The most common error is placing chromatid separation in metaphase. Metaphase is alignment; nothing separates yet. The transition happens at the metaphase-to-anaphase boundary, triggered by cohesin cleavage, which is why the answer is anaphase.
A second misconception treats the spindle as the sole force. The spindle pulls, but separation also requires releasing the cohesin lock. Without separase activity, chromatids cannot part even with perfect spindle attachment. Conversely, cohesin cleavage alone without spindle force leads to separation without proper poleward movement.
A third error confuses chromosomes and chromatids. Before anaphase, one duplicated chromosome contains two sister chromatids. After separation, each chromatid is called a chromosome. The chromosome count doubles at anaphase, even though the DNA content per nucleus halves only after cytokinesis.
Bulleted warnings for students, educators, and lab readers:
Do not say chromatids "split" — they are already separate molecules; the connection is released. Do not call anaphase the longest mitotic phase; it is among the shortest. Do not assume all cells show the same anaphase duration; embryonic cells can be far faster. Do not confuse meiotic anaphase I (homologs separate) with mitotic anaphase (sisters separate). Do not treat visible chromatid separation as proof that upstream attachment was correct.
Practical actions for accurate identification
If you are identifying mitotic phases on a slide or in a teaching lab, use morphology and position, not timing alone.
Confirm alignment first: a metaphase plate means chromatids have not parted. Look for a V-shaped configuration: chromatids pulled to opposite poles appear V-shaped as arms trail. Check the poles: two compact chromosome clusters moving apart indicate anaphase, not telophase. Distinguish telophase by nuclear envelope reformation and chromosome decondensation, not just pole separation. Score a minimum of 50 to 100 cells to estimate phase proportions reliably; small samples mislead. Flag lagging chromatids separately, because they signal attachment defects, not normal anaphase.
For questions phrased around exam thresholds, the reconciliation is straightforward. If the question asks during which phase chromatids move apart, answer anaphase. If it asks what triggers separation, answer separase cleavage of cohesin after checkpoint satisfaction. Both thresholds are satisfied by the same event at the metaphase-anaphase transition.
What this explanation cannot tell you
Anaphase timing and rates vary by organism, tissue, and temperature. The numbers here reflect common human cell lines and should be treated as representative ranges, not universal constants. Single-cell variability is substantial, and live imaging frequently shows outliers.
The molecular model, by contrast, is well established across eukaryotes: cohesin cleavage by separase releases sister chromatids, and spindle forces move them poleward. Newer work continues to refine motor contributions and checkpoint sensitivity, but the core sequence is stable.
This article is general educational information about cell biology, not a diagnostic or clinical resource. For questions about chromosome segregation errors in a specific patient or tissue sample, consult a qualified pathologist or clinical geneticist.