During dorsal patterning of the Drosophila embryo, nuclei activate BMP target genes within minutes by integrating BMP signaling over time rather than reading instantaneous levels, and the homeodomain transcription factor Zen lowers the signaling threshold required for activation. The molecular implementation of this integration-based decoding strategy is unresolved: it is unknown whether Zen physically associates with BMP target promoters, what transcriptional machinery or co-factors it engages, and which spatially restrictive factors (Brinker is a candidate) prevent activation in lateral nuclei that receive sufficient integrated BMP.

Background

Morphogen gradients are concentration fields of diffusing signaling molecules that instruct cells in a developing tissue to adopt different fates according to the signal level they experience. Most classic models, such as the interpretation of the Bicoid gradient along the anterior-posterior axis of the Drosophila embryo, assume the input is roughly stable while cells decide. The dorsal patterning of the early Drosophila melanogaster embryo is different: the BMP (bone morphogenetic protein) ligands Dpp and Screw are expressed broadly across the dorsal half of the syncytial blastoderm, producing initially diffuse pathway activity that, over tens of minutes during nuclear cycle 14 (the last syncytial cycle before gastrulation), is sharpened into a narrow peak at the dorsal midline by a source-sink mechanism. In this mechanism the secreted inhibitor Sog (Short gastrulation, a Chordin homolog) binds BMP ligands laterally and shuttles them toward the midline, where the protease Tolloid cleaves Sog and releases the ligands. Nuclei must interpret this rapidly evolving gradient quickly to specify dorsal ectoderm and the amnioserosa (an extraembryonic tissue), via nested expression domains of BMP target genes such as ush, doc2, tup and c15. SMAD proteins are the intracellular transducers of BMP signaling: in Drosophila, Mad and its co-factor Medea (the homolog of mammalian SMAD4) accumulate in nuclei in proportion to pathway activity and bind BMP target promoters. Brantley, Janssen, Chao, Vergassola, Blythe and Di Talia (bioRxiv 2025; published in Current Biology 36:2853-2865, 2026) developed the first endogenous live reporter of BMP pathway activity in this system (functional GFP-tagged Medea) and combined it with MS2-MCP reporters of nascent transcription, in which stem-loop arrays inserted at endogenous target loci are visualized by a fluorescent RNA-binding protein so that active transcription sites can be tracked in living embryos. Their central findings were: (1) the probability and timing of target-gene activation are predicted far better by the time integral of nuclear BMP activity than by instantaneous levels; (2) the homeodomain transcription factor Zen, which is broadly expressed across the dorsal side early in nuclear cycle 14 under control of the maternal-to-zygotic-transition activator Zelda, is required for this rapid response — zen mutants show reduced sensitivity to BMP and delayed transcriptional onset despite higher pathway activity; (3) Zen does not act by changing chromatin accessibility (single-embryo ATAC-seq — an assay for transposase-accessible chromatin that reads out chromatin openness — is nearly unchanged in zen mutants) or Mad/Medea occupancy at target promoters, implying Zen acts at another step, plausibly recruitment or stabilization of the transcriptional machinery; and (4) in sog mutants, where the BMP field is expanded and flattened, the integrated input-output relationship still holds within about 45 microns of the dorsal midline (the region where Zen protein is present) but breaks down laterally: nuclei reach integrated BMP levels that would be sufficient in wild type yet fail to activate target genes, indicating additional spatially restrictive regulation, with the BMP-repressed transcriptional repressor Brinker (brinker is expressed where BMP signaling is low and represses BMP target promoters) proposed but untested as the responsible factor. The authors formalize the integration decoding as a first-passage-time problem: if promoter activation is a stochastic event with rate r(t) proportional to Mad/Medea concentration, the probability that activation has not yet occurred by time t decays as exp(-∫ r dt'), so the cumulative integral of signaling — and, through it, Zen's lowering of the activation barrier — naturally controls the timing of transcriptional onset. What remains unresolved is the molecular mechanism that implements this decoding: whether Zen directly binds BMP target promoters and what it recruits, and which factors restrict activation laterally in sog mutants.

Problem Statement

By what molecular mechanism do nuclei of the early Drosophila embryo decode the rapidly evolving BMP gradient through time-integrated signaling, and what bounds this decoding spatially? Concretely, determine: (1) the mechanism of Zen-dependent transcriptional competence — whether Zen physically associates with the promoters of BMP target genes (such as ush, doc2, tup and c15) early in nuclear cycle 14, and whether it promotes recruitment or stabilization of the transcriptional machinery (for example preloaded or paused RNA polymerase II or co-activators), thereby lowering the activation barrier posited by the first-passage-time model, given that Zen affects neither chromatin accessibility nor Mad/Medea occupancy; and identify the co-factors Zen interacts with in this process; and (2) the identity and mechanism of the spatially restrictive regulatory factors that prevent transcriptional activation in lateral nuclei more than about 45 microns from the dorsal midline of sog mutant embryos even though those nuclei accumulate integrated BMP signaling levels sufficient for activation in wild type — in particular whether the transcriptional repressor Brinker, or other repressors, modulates competence at BMP target promoters in this context. The two components above are logically linked, though each sub-question within them is independently answerable on its own; the common objective uniting them is a mechanistically specific explanation of integration-based decoding of a dynamic morphogen in this system.

References

  1. Brantley SE, Janssen J, Chao A, Vergassola M, Blythe SA, Di Talia S. Rapid transcriptional response to a dynamic morphogen by time integration. Current Biology 36(11):2853-2865.e6, 2026. DOI 10.1016/j.cub.2026.04.064. https://doi.org/10.1016/j.cub.2026.04.064
  2. Brantley SE, Janssen J, Chao A, Vergassola M, Blythe SA, Di Talia S. Rapid transcriptional response to a dynamic morphogen by time integration. bioRxiv 2025.06.09.658715, 2025. DOI 10.1101/2025.06.09.658715. https://doi.org/10.1101/2025.06.09.658715
  3. Shaikh R, Frank C, Larson NJ, Chen H-Y, Pradhan AG, Mendiola K, Hanjaya-Putra D, Umulis DM, Zartman JJ, Reeves GT, Li L. Biochemical principles of SMAD signaling across the animal kingdom. Biochemical Journal 483(8):1509-1534, 2026. DOI 10.1042/BCJ20250135. https://doi.org/10.1042/BCJ20250135
  4. Weevers SL, Repina N, Papasaikas P, Ferralli J, Smallwood S, Wittlieb J, Klimovich A, Tsiairis CD. Optogenetic perturbations reveal temporal integration of Wnt signaling during pattern formation. bioRxiv 2026.07.08.737164, 2026. DOI 10.64898/2026.07.08.737164. https://doi.org/10.64898/2026.07.08.737164
  5. Shimell M, Connacher R, Park S, Peterson AJ, Arora K, O'Connor MB. Shrew is an essential, recently evolved, Drosophila specific gene required for formation of the embryonic BMP activity gradient. Developmental Biology 534:9-22, 2026. DOI 10.1016/j.ydbio.2026.03.001. https://doi.org/10.1016/j.ydbio.2026.03.001
  6. Moore G, Revici R, Forbes Beadle L, Sutcliffe C, Birchenough H, Baldock C, Ashe HL. An avidity-driven mechanism of extracellular BMP regulation by Twisted gastrulation. Development 153(16):dev204584, 2026. DOI 10.1242/dev.204584. https://doi.org/10.1242/dev.204584
  7. Donoghue A, Mosby LS, Lago-Baldaia I, Hodgetts T, Ursu E, Erten Z, Hadjivasiliou Z, Fernandes VM. Dynamic integration of morphogen and juxtacrine signaling specifies cell fate at single-cell resolution. Current Biology 36(14):3580-3594.e5, 2026. DOI 10.1016/j.cub.2026.06.024. https://doi.org/10.1016/j.cub.2026.06.024