SOX9 C-terminus links enhancer occupancy to chromatin opening in adenoma cells
A new preprint places the SOX9 C-terminus at the center of a mechanism that converts transcription factor binding into chromatin remodeling. The work identifies an intrinsically disordered C-terminal module that recruits canonical BAF (cBAF) complexes through SMARCB1, allowing enhancer regions to shift into a more accessible state during APC-mutant adenoma formation. In practical terms, SOX9 is doing more than sitting on DNA; it is helping define which regulatory regions can be opened and maintained in a stem-like transcriptional program.
The study addresses a persistent problem in gene regulation: many transcription factors bind their genomic targets, yet only some of those sites are converted into functional regulatory hubs. The authors focus on a cell-state transition in which APC-mutant cells acquire stem-like properties, and they show that SOX9 participates in that transition by recruiting remodeling machinery rather than acting as a passive DNA-binding factor. The result is a direct bridge between sequence-specific occupancy and local chromatin accessibility.
One sentence captures the biological consequence cleanly: remove SOX9, and the enhancer program collapses.
Acute SOX9 loss strips cBAF from most co-bound enhancers
Using acute, reversible SOX9 degradation, the investigators found that displacement of SOX9 caused cBAF to leave roughly 80% of co-bound enhancers. The effect was stronger on cBAF than on other related complexes, including pBAF and ncBAF, pointing to a degree of specificity in the recruitment logic. Accessibility at these sites fell in parallel, and the stem-like cell identity associated with the enhancer program also collapsed. When SOX9 was re-expressed, both accessibility and cell-state features returned.
That on-off reversibility matters because it argues against a one-time developmental scar or an indirect adaptation. The enhancer state appears to depend on continuous SOX9 presence, implying that the remodeling complex is actively maintained rather than permanently installed. In a tumor context, that kind of dependency can create a vulnerability: if the recruiter disappears, the regulatory architecture begins to fail quickly.
The preprint also reports that a C-terminal truncation of SOX9 preserved DNA binding but disrupted cBAF engagement and chromatin opening. This separation of functions is important. It means the protein’s targeting domain and its remodeling interface are not the same thing, and enhancer activation requires both. A transcription factor can still recognize the genome yet fail to change cell state if it cannot enlist the right chromatin machinery.
Remodeler recruitment emerges as the limiting effector function
To test whether cBAF recruitment itself was the critical step, the team enforced SMARCB1 docking to truncated SOX9. Partial restoration followed: cBAF recruitment improved, accessibility increased, stem-like transcription recovered to some extent, and organoid self-renewal was supported again. The rescue was partial rather than complete, which fits a model in which the recruitment interface is a limiting effector function but not the sole determinant of the full SOX9 program.
That experiment does more than confirm mechanism. It shows that the failure of truncated SOX9 is not simply loss of DNA recognition or a nonspecific structural defect. The missing function is the ability to recruit the remodeler in a productive way. In a translational sense, that narrows the biologic question to a usable one: which transcription factor surfaces are capable of handing off enhancer control to chromatin remodelers?
The authors also examined the isolated SOX9 C-terminus and found that it supports higher-order cBAF assembly. Even without the rest of the protein, this region carried enough information to assemble the remodeling complex. That finding points to the C-terminus as an autonomous interaction module, not just a passive tail attached to a DNA-binding scaffold.
A transferable module can redirect remodeling when fused to dCas9
The most striking extension of the work is the COMET construct, in which the isolated SOX9 C-terminus was fused to dCas9. In this configuration, the module redirected endogenous cBAF to targeted genomic loci and opened chromatin at those sites. The result is a proof of principle that disordered transcription factor regions can be repurposed as modular instructions for chromatin remodeling.
Mechanistically, this is a useful inversion of the usual transcription factor story. Instead of asking whether a factor binds a locus, the study asks whether the factor can recruit the complex that changes the locus’s physical state. The answer here is yes, at least for the SOX9 C-terminus in the experimental systems described. A disordered region, often treated as structurally vague, turns out to encode a precise functional command: bring cBAF here.
The work also has implications for enhancer biology in cancer. APC-mutant adenoma formation depends on a stem-like accessibility program, and SOX9 appears to help sustain that program by coordinating with cBAF. If enhancer activation is driven by such transferability of remodeling instructions, then transcription factor domains may become as biologically informative as the motifs they bind.
Why this matters for enhancer control and cancer biology
For cancer biology, the appeal of this mechanism is its specificity. The study does not merely show that SOX9 is associated with open chromatin; it identifies the part of SOX9 that is required to recruit the remodeling machinery responsible for that openness. The distinction helps explain how a single factor can shape cell state without globally altering all chromatin. A selective recruitment module can concentrate remodeling power at a limited set of enhancers and thereby stabilize one transcriptional identity over another.
The data also suggest a broader principle: intrinsically disordered transcription factor regions may encode transferable interaction motifs that can be co-opted to direct chromatin remodelers. In this case, the SOX9 C-terminus behaves like a portable enhancer-opening device. When the module is intact, cBAF engagement is strong enough to sustain accessibility and stem-like transcription; when it is lost, binding alone is not enough to preserve the regulatory state.
Because this is a preprint and has not undergone peer review, the claims should be read as mechanistic and provisional rather than definitive. Still, the experimental arc is unusually coherent: acute depletion, domain truncation, rescue by enforced docking, and locus targeting through a synthetic fusion all converge on the same conclusion. The next step is not whether SOX9 binds DNA, but how its C-terminal chemistry is read by remodeling complexes in different cellular settings.
The unanswered question is how broadly this logic applies beyond APC-mutant adenoma models and whether other disordered transcription factor tails encode comparable recruitment instructions for chromatin remodelers.