Sister chromatids are held together by a protein complex called cohesin, which forms a ring-shaped structure that physically encircles both copies of a duplicated chromosome. This molecular ring keeps the two identical DNA strands linked from the time they’re created during DNA replication until the cell is ready to pull them apart during division. Without cohesin, chromosomes would drift apart prematurely, and cells would end up with the wrong number of chromosomes.
How the Cohesin Ring Works
Cohesin is built from four core protein subunits that assemble into a structure resembling a closed loop. Two of these proteins, called SMC1 and SMC3, form long arm-like shapes that pair together. A third protein called SCC1 (also known as RAD21) bridges the ends of those arms, completing a triangular ring. A fourth protein, SCC3, attaches to the outside of this ring and helps regulate its behavior.
The prevailing model is that this ring works by simple topological entrapment: it literally encircles two strands of DNA, the way a carabiner clips around a rope. Opening the ring releases the DNA. This has been confirmed experimentally, since artificially cutting any part of the ring causes sister chromatids to fall apart. The ring’s interior is large enough to accommodate two DNA duplexes side by side, though the fit is snug enough to keep them from sliding out on their own.
When Cohesion Gets Established
Cohesin doesn’t wait until cell division to do its job. The complex is first loaded onto chromosomes during late mitosis and the G1 phase of the cell cycle, well before DNA replication begins. A loader protein called NIPBL (known as SCC2 in yeast) works with a partner protein to open the ring and thread it onto DNA.
The critical moment, though, is S phase, when the cell copies its DNA. As the replication machinery moves along a chromosome, cohesin that was already sitting on the unreplicated DNA gets repositioned to the replication fork. There, the ring captures the second, newly synthesized copy of the chromosome as it emerges. Research from human cells shows this capture happens before the new DNA strand is even fully stitched together, meaning cohesion is established almost immediately as replication proceeds. The interaction between cohesin, its loader, and the replication machinery peaks during mid-S phase.
One important detail: for the ring to grab the second DNA strand, that strand must initially be single-stranded, giving it enough flexibility to thread past the first strand already inside the ring. Once the single strand is converted to double-stranded DNA by normal replication, the entrapment becomes stable.
How Cohesin Is Removed in Stages
Cells don’t remove all their cohesin at once. Instead, the process happens in two distinct waves, and this staged removal is essential for accurate chromosome segregation.
During the early stages of mitosis (prophase), a protein called WAPL opens cohesin rings along the long arms of the chromosomes, stripping most of the cohesin away. By the time a cell reaches metaphase, when chromosomes line up at the center of the cell, the arms of sister chromatids are mostly free of cohesin. But cohesin at and near the centromere, the pinched middle region where the cell’s pulling machinery attaches, is specifically protected and stays in place.
This centromeric cohesin is shielded by a guardian protein called Shugoshin, which recruits a partner enzyme that counteracts the chemical signals WAPL uses to open the rings. Shugoshin also helps ensure that the pulling fibers from opposite sides of the cell attach correctly to each sister chromatid, creating the tension needed for an even split.
The Final Cut at Anaphase
The remaining centromeric cohesin is destroyed all at once when the cell transitions from metaphase to anaphase. A specialized enzyme called separase, a large protease in the same family as the enzymes involved in programmed cell death, cuts the SCC1 subunit of the cohesin ring. It targets a specific short amino acid sequence on SCC1, slicing the protein after a particular arginine residue. This opens the ring irreversibly, and the sister chromatids are free to be pulled to opposite ends of the cell.
Separase is kept inactive until the exact right moment by a checkpoint system that monitors whether all chromosomes are properly attached to the cell’s pulling machinery. Only when every chromosome passes this check is separase unleashed.
Cohesin Behaves Differently in Meiosis
During meiosis, the type of cell division that produces eggs and sperm, the cell swaps out one key cohesin component. Instead of SCC1, meiotic cells use a related protein called Rec8 as the ring’s bridge subunit. This swap has important functional consequences.
In the first meiotic division, homologous chromosomes (one from each parent) need to separate while sister chromatids stay together. Rec8-containing cohesin along the chromosome arms is cleaved by separase at the end of meiosis I, allowing homologs to pull apart. But Rec8 cohesin at the centromere is protected by Shugoshin, keeping sister chromatids attached through the transition into the second meiotic division. Only during meiosis II is the centromeric Rec8 finally cut, allowing sisters to separate. This two-step removal is what makes meiosis produce cells with half the normal chromosome number.
Dynamic Regulation on Chromosomes
Cohesin doesn’t simply sit still once loaded. It constantly cycles on and off chromosomes, regulated by competing proteins. NIPBL promotes loading and can also slide cohesin along the DNA. WAPL promotes removal by opening the ring’s exit gate. A third regulator, PDS5, acts as a brake, helping to stop cohesin at specific positions along chromosomes.
In mammalian cells, cohesin pauses at sites marked by a protein called CTCF, which acts like a roadblock. The presence of NIPBL and PDS5 at cohesin binding sites is largely mutually exclusive: where one is enriched, the other tends to be absent. This creates a system where some chromosomal locations are active loading zones while others are stable parking spots. Cohesin that can’t be stabilized at CTCF-marked sites tends to dissociate and reload elsewhere.
What Happens When Cohesin Goes Wrong
Because cohesin is so central to chromosome organization, mutations in its components or regulators cause a group of developmental disorders collectively called cohesinopathies. The best known is Cornelia de Lange syndrome, which affects growth, limb development, facial features, and cognitive function. More than half of cases result from mutations in the NIPBL gene, which encodes cohesin’s loader protein. Rarer cases involve mutations in SMC1A, SMC3, RAD21, or HDAC8, all of which contribute to cohesin’s structure or regulation.
These conditions aren’t caused by failed chromosome segregation. Instead, they result from cohesin’s other major job: regulating which genes are turned on or off during embryonic development. Cohesin helps organize DNA into loops that bring distant regulatory elements close to the genes they control. When this looping function is impaired, even subtly, the precise choreography of gene activation during early development goes awry.