article published yes Benjamin Zoller author Alexis Benichou author 3a67230c-5fc0-11ef-a673-de9a2ffadafe Thomas Gregor author Gašper Tkačik author 3D494DCA-F248-11E8-B48F-1D18A9856A870000-0002-6699-1455 GaTk department Transcription in 4D: the dynamic interplay between chromatin architecture and gene expression in developing pseudo-embryos project Eukaryotic gene regulation relies on stochastic yet controlled promoter switching, in which genes transition between transcriptionally active and inactive states. Despite the molecular complexity of this process, recent studies have revealed a surprising invariance of the “switching correlation time” (TC)—the characteristic decay time of the autocorrelation function of promoter activity fluctuations—across gene expression levels in multiple genes and organisms. A biophysically plausible explanation for this invariance has so far been lacking. Here, we show that this empirical constraint imposes stringent requirements on minimal yet realistic models of transcriptional regulation. Specifically, reproducing TC–invariance requires regulatory architectures with at least four internal states and nonequilibrium dynamics that break detailed balance. Using Bayesian inference on Drosophila gap gene expression data, we demonstrate that such models i) quantitatively reproduce the observed TC–invariance, ii) remain robust to parameter perturbations, and iii) maximize information transmission from transcription factor concentration to gene expression. Remarkably, the TC-invariant modulation strategy we identify as optimal closely parallels contemporary control-theoretic results on the modulation of stochastic switching systems. Taken together, our results suggest that eukaryotic transcriptional regulation operates in a nonequilibrium regime to balance precision, reaction-rate limitations, and energy dissipation, thereby achieving near-optimal information transmission under fundamental physical constraints. https://research-explorer.ista.ac.at/download/22363/22376/2026_PNAS_Zoller.pdf application/pdfno National Academy of Sciences2026 eng 0027-8424 1091-6490 4240696210.1073/pnas.2524855123 12328 Zoller, B., Benichou, A., Gregor, T., &#38; Tkačik, G. (2026). Invariant nonequilibrium dynamics in gene regulation optimize information flow. <i>Proceedings of the National Academy of Sciences of the United States of America</i>. National Academy of Sciences. <a href="https://doi.org/10.1073/pnas.2524855123">https://doi.org/10.1073/pnas.2524855123</a> Zoller, Benjamin, Alexis Benichou, Thomas Gregor, and Gašper Tkačik. “Invariant Nonequilibrium Dynamics in Gene Regulation Optimize Information Flow.” <i>Proceedings of the National Academy of Sciences of the United States of America</i>. National Academy of Sciences, 2026. <a href="https://doi.org/10.1073/pnas.2524855123">https://doi.org/10.1073/pnas.2524855123</a>. Zoller B, Benichou A, Gregor T, Tkačik G. 2026. Invariant nonequilibrium dynamics in gene regulation optimize information flow. Proceedings of the National Academy of Sciences of the United States of America. 123(28), e2524855123. B. Zoller, A. Benichou, T. Gregor, and G. Tkačik, “Invariant nonequilibrium dynamics in gene regulation optimize information flow,” <i>Proceedings of the National Academy of Sciences of the United States of America</i>, vol. 123, no. 28. National Academy of Sciences, 2026. Zoller B, Benichou A, Gregor T, Tkačik G. Invariant nonequilibrium dynamics in gene regulation optimize information flow. <i>Proceedings of the National Academy of Sciences of the United States of America</i>. 2026;123(28). doi:<a href="https://doi.org/10.1073/pnas.2524855123">10.1073/pnas.2524855123</a> Zoller, Benjamin, et al. “Invariant Nonequilibrium Dynamics in Gene Regulation Optimize Information Flow.” <i>Proceedings of the National Academy of Sciences of the United States of America</i>, vol. 123, no. 28, e2524855123, National Academy of Sciences, 2026, doi:<a href="https://doi.org/10.1073/pnas.2524855123">10.1073/pnas.2524855123</a>. B. Zoller, A. Benichou, T. Gregor, G. Tkačik, Proceedings of the National Academy of Sciences of the United States of America 123 (2026). 223632026-07-19T22:01:46Z2026-07-20T13:15:11Z