How a TF binds the genome is controlled by both DBDs and IDRs, but it remains difficult to predict the relative contributions of these two regions. We demonstrated that very short activation domains can control the fraction of TF molecules bound to chromatin, sometimes accounting for more binding than the DBD. Mutations that increase activation domain strength in reporter assays increase the fraction of molecules bound to chromatin. Conversely, mutations that decrease activation domain strength decrease the fraction of molecules bound to chromatin. This trend holds for allelic series of four acidic activation domains, for three structurally diverse DBDs, and for synthetic and full-length TFs.

SMT has revealed many new dimensions of TF nuclear dynamics and has been instrumental for our discovery. The technique has shown that TF-chromatin interactions are generally short-lived (Brouwer and Lenstra, 2019; Lionnet and Wu, 2021; Liu et al., 2014), that coactivator chromatin binding is long-lived (Ferrie et al., 2024), and that IDRs can make larger contributions to chromatin binding than DBDs (Chen et al., 2022). Here, we demonstrate that minimal activation domains (39–60AA) can control the fraction of molecules bound to chromatin. Our results are consistent with an SMT study of an activation domain deletion in p53 showing reduced association with dense chromatin (Mazzocca et al., 2023).

There are three hypotheses for how increasing activation domain strength causes a higher fraction of TF molecules bound to chromatin: (1) activation domains could directly bind DNA, (2) activation domains could modulate DBD interactions with DNA, or (3) activation domains could bind to coactivators that are bound to chromatin, tethering the TF to chromatin.

Our short acidic activation domains are unlikely to bind DNA directly due to electrostatic repulsion: both the activation domains and DNA are negatively charged and repel each other. While protein net charge can play a large role in nuclear diffusion (Xiang et al., 2020), we found no correlation between net charge and fractions bound in factors we tested (Figure 3—figure supplement 4).

It is also possible that our activation domains are interacting with the DBD to modulate DNA binding, an emerging theme in TF biology (Baughman et al., 2022; Bentley et al., 2023; Bjarnason et al., 2024; He et al., 2019; Krois et al., 2018; Sun et al., 2021). For our results, intramolecular interactions are not the main drivers of chromatin binding because the phenotype is consistent across three structurally distinct DBDs.

Our PAPA experiments support the hypothesis that protein-protein interactions between activation domains and coactivators explain the positive correlation between activation domain strength and the fraction of TF molecules bound to chromatin. The overall chromatin-bound fraction of SNAP-tagged synthetic TFs was higher than that of the Halo-tagged versions, perhaps reflecting differences in physical properties of the two tags, such as volume and charge (Figure 3C vs Figure 5C–E, violet curves). TFs reactivated by PAPA from p300 had a higher chromatin-bound fraction than those directly reactivated by violet light (DR), indicating the expected correlation between p300 binding and chromatin binding (Figure 5). Compared to WT, the two superactive HIF1α alleles showed stronger binding to p300 by PAPA and a more pronounced shift toward chromatin-bound molecules when comparing PAPA to DR trajectories. Curiously, while the bound fraction of PAPA trajectories increased with activation domain strength, DR trajectories, if anything, showed the opposite trend. One possible explanation is that a large fraction of TF molecules bind p300, and each round of PAPA followed by re-bleaching depletes these complexes from subsequent rounds of direct reactivation—with more interaction resulting in greater depletion. This work marks the first time that PAPA has been applied to measure TF-coactivator interactions in vivo, which we anticipate will be a powerful new experimental approach to study transcription. Together with the PAPA experiments, the pharmacological perturbations support our hypothesis that activation domain-coactivator interactions strongly influence TF binding to chromatin (Figure 8). We cannot distinguish whether the coactivator directly binds DNA, binds a nucleosome, or binds another TF bound to DNA, but we suspect all three binding modes are involved.

Model for how activation domains contribute to chromatin binding by interactions with coactivators.

(A) A solitary transcription factor (TF) diffusing in the nucleoplasm, a lone wolf. (B) A TF bound to a coactivator with its activation domain diffusing through the nucleoplasm. (C) Three TFs bound simultaneously to a coactivator diffusing in the nucleoplasm as a complex or wolfpack. (D) A TF nonspecifically bound to random DNA. (E) A TF specifically bound with its DBD to a cognate DNA motif at the reporter locus. (F) TFs specifically bound to motifs at the reporter locus and simultaneously bound to a coactivator. (G) Proposed model for how the activation domain tethers the synthetic TF to chromatin through the coactivator. This situation occurs at active loci where the coactivator has been recruited by the orange and blue TFs.

In the CUT&RUN data, stronger activation domains cause higher occupancy at the reporter locus, but the vast majority of other genomic binding occurs at other active loci. Consequently, we believe our activation domain-mediated chromatin binding is distinct from the ability of long IDRs to direct TFs to specific genomic loci (Brodsky et al., 2020; Gera et al., 2022; Goldman et al., 2023; Hurieva et al., 2024; Mindel et al., 2024a). Instead, we believe the SMT is observing the same underlying biological process that causes activation domains to enhance promoter occupancy in single-molecule footprinting experiments (Doughty et al., 2024) and long TF residence times on promoters (Schaepe et al., 2025). It is remarkable that short activation domains can have such a large effect on binding.

Genome binding is an emergent property of full-length TFs, with contributions from the DBD, activation domains, and their interactions. In textbook models, the DBD binds DNA first, and then the activation domain recruits the coactivators. We speculate that recruitment can occur in both directions: sometimes the first event will be the activation domain binding to a coactivator that is already engaged at an active locus, and then the DBD will bind the local DNA (Figure 8). Different activation domains bind distinct coactivators (Alerasool et al., 2022), but only some of these interactions, which can have affinities of 10 nanomolar (Berlow et al., 2017; De Guzman et al., 2004), will be strong enough to tether TFs to chromatin.

Our results also challenge the traditional interpretation of activation domain strength. Previously, we had assumed that all our synthetic TFs had similar DNA occupancy because they have the same DBD. Our results here show that strong activation domains increase occupancy at a reporter locus. As a result, the higher reporter activity is a convolution of more binding and more coactivator recruitment (Doughty et al., 2024). This finding motivates the need for new assays for activation domain function that can separate binding from reporter activity.

In this work, we focused on the fraction of molecules bound to chromatin. We also see some changes in the distribution of inferred diffusion constants of mobile molecules, some of which are bimodal. After norstictic acid treatment, the slow-moving population shrank and the fast-moving population grew. The slow-moving population is also prominent in the PAPA data. We speculate that the slow-moving population might be TFs bound to coactivators exploring the nucleus in clusters or ‘wolf packs’ (Figure 8C), as we previously hypothesized (Staller, 2022a).