A corepressor is a protein that helps shut down gene activity, but it cannot find or bind to DNA on its own. Instead, it works as a partner: a DNA-binding protein called a repressor latches onto a specific stretch of DNA, then recruits the corepressor to do the heavy lifting of silencing the gene. Think of the repressor as the address label and the corepressor as the package of tools that actually gets the job done once it arrives.
This distinction matters because corepressors are not passive players. They bring enzymatic machinery that physically remodels the structure around DNA, making genes harder for the cell to read. They sit at the center of processes from immune cell development to hormone signaling to cancer.
How Corepressors Differ From Repressors
Repressors are sequence-specific DNA-binding proteins. They recognize a particular genetic address and park themselves there. Corepressors lack that ability. They contain no characterized DNA-binding domains and do not contact DNA directly. Instead, they arrive at the scene through protein-to-protein interactions with the repressor already sitting on DNA. Some proteins blur this line. A protein called NKAP, for instance, has been found at DNA sites but has no known DNA-binding domain, leaving researchers debating whether it is a true repressor or a corepressor that reaches DNA indirectly through a partner.
The functional split is clean, though: the repressor chooses which gene to silence, and the corepressor provides the molecular equipment to carry out the silencing.
The Silencing Machinery Inside a Corepressor Complex
Corepressors rarely work alone. They assemble into large multi-protein complexes, and the core tool they carry is an enzyme that strips chemical tags called acetyl groups from histone proteins. Histones are the spools that DNA wraps around, and when those spools carry acetyl tags, the DNA is loosely wound and easy for the cell to read. Removing those tags causes the DNA to wind tighter, compacting the chromatin and effectively locking the gene shut.
One well-studied example is the NuRD complex. It contains the enzyme HDAC1 (a histone deacetylase) and gets recruited to target genes through interactions with DNA-binding transcription factors. During the repression phase, NuRD strips acetyl groups from histones and remodels chromatin into a closed, repressive structure. Another major complex pairs the corepressor NCoR with HDAC3, along with subunits called GPS2 and TBL1. GPS2 and TBL1 bind cooperatively to a region of NCoR to form a three-part scaffold, and this scaffold indirectly links to HDAC3 through a specialized protein domain that also activates the enzyme’s latent deacetylase activity. In other words, the corepressor complex doesn’t just deliver the enzyme; it switches it on.
Not all corepressors rely solely on histone deacetylation. The Groucho/TLE family of corepressors, found across species from fruit flies to humans, appears to silence genes primarily through chromatin compaction, physically squeezing the DNA packaging so tightly that the gene-reading machinery cannot access it. Histone deacetylation plays a supporting role in Groucho-mediated repression, but compaction is the main event.
Major Corepressors and Where They Act
Two of the best-characterized corepressors in humans are NCoR (nuclear receptor corepressor) and SMRT (silencing mediator for retinoid and thyroid hormone receptor). Both were originally discovered through their interactions with nuclear receptors, a family of proteins that detect hormones and other signals inside the cell. Unliganded thyroid hormone receptors and retinoid receptors, for example, sit on DNA and actively repress their target genes by holding onto NCoR or SMRT. When the appropriate hormone arrives, the receptor changes shape, releasing the corepressor and picking up a coactivator instead, flipping the gene from “off” to “on.”
The progesterone receptor follows a slightly different pattern. NCoR and SMRT do not associate with it when no hormone is present. Instead, certain drug-like compounds that bind the receptor can promote corepressor recruitment, suppressing gene activation. This is one reason the same receptor can behave differently depending on which molecule occupies it.
Groucho/TLE corepressors operate in entirely different pathways. They are tethered to gene-control regions by DNA-binding proteins such as TCF, HES, and RUNX. In Wnt signaling, Groucho/TLE sits on Wnt-responsive enhancers through TCF and keeps them silent. When a Wnt signal activates a protein called beta-catenin, that protein overcomes Groucho-dependent repression and switches the gene on. A similar mechanism operates in Notch signaling, where Groucho/TLE binds HES repressors to keep Notch target genes quiet until the pathway is activated. These roles make Groucho/TLE corepressors central gatekeepers in embryonic development and tissue patterning.
The Corepressor-to-Coactivator Switch
Gene activation often requires more than just adding a coactivator. The corepressor already parked at the gene has to be actively removed first. This exchange is a regulated, deliberate process. For nuclear receptor target genes, proteins called TBL1 and TBLR1 mediate the swap: when a hormone binds its receptor, TBLR1 selectively triggers the removal of NCoR and SMRT, clearing the way for coactivators to take their place. This exchange mechanism is not unique to hormone receptors. It also operates for signaling proteins like c-Jun and NF-kappaB, suggesting it is a widespread strategy cells use whenever a signal-dependent gene needs to flip from silent to active.
For Groucho/TLE, the inactivation route is different. An enzyme called UBR5 tags Groucho with ubiquitin, marking it for removal or inactivation so beta-catenin can access TCF and turn on Wnt target genes. This is so central to Wnt signaling that eliminating Groucho by mutation makes certain other components of the pathway largely unnecessary.
Corepressor Dysfunction in Cancer
Because corepressors control which genes stay silent, their malfunction can unleash genes that drive cell growth. Disrupted NCoR and SMRT activity has been observed across many cancer types and leukemias.
- Breast cancer: NCoR levels are downregulated in invasive ductal carcinomas, and pathway analysis has linked loss of NCoR and SMRT complexes to alterations in luminal A breast tumors.
- Prostate cancer: Reduced NCoR/SMRT levels may contribute to hormone-independent activation of the androgen receptor, driving cancer progression even without the hormone signal that would normally be required.
- Leukemia: Acute promyelocytic leukemia can be caused by abnormal fusion proteins that hijack NCoR/SMRT-mediated repression. A chromosomal rearrangement involved in roughly 15% of acute myeloid leukemia cases generates a fusion protein that co-opts the corepressor machinery to silence genes that would otherwise restrain cell growth.
- Brain tumors: Somatic mutations in GPS2, a subunit of the NCoR-HDAC3 complex, have been identified in medulloblastoma, an aggressive pediatric brain tumor.
Overproduction of HDAC3, the deacetylase enzyme at the heart of NCoR complexes, has been observed in colon, lung, prostate, and breast cancers, correlating with poor survival and prognosis. This has made components of corepressor complexes active targets for drug development, particularly inhibitors that block histone deacetylase activity.
Built-In Controls on Corepressor Activity
Cells do not leave corepressor complexes running unchecked. One elegant control mechanism involves turning the corepressor’s own enzyme against itself. The acetyltransferase p300 can acetylate HDAC1, the deacetylase inside the NuRD complex. Once acetylated, HDAC1 loses its ability to strip acetyl groups from histones. It also inhibits its partner enzyme HDAC2, effectively shutting down the entire complex’s repressive power. This mechanism plays out during blood cell development: when immature blood cells are pushed to differentiate, acetylated HDAC1 accumulates at certain gene promoters, weakening the NuRD complex’s grip and allowing genes needed for maturation to turn on. The cell essentially uses the same chemical tag (acetylation) both as the thing the corepressor removes and as the switch that disables the corepressor itself.