The story begins in 1969, when Robert Roeder and William Rutter first separated multiple forms of DNA-dependent RNA polymerase from eukaryotic cells. Over the following decades, biochemists painstakingly identified the general transcription factors, TFIID, TFIIB, TFIIE, TFIIF, and TFIIH, that guide polymerase II to promoters and position it precisely at the transcription start site. The TATA box, discovered in histone genes in the late 1970s, became the archetypal promoter element, and the crystal structure of the TATA-box-binding protein bound to DNA in 1993 revealed how this saddle-shaped factor bends the double helix nearly ninety degrees. But these early pictures were static snapshots of isolated parts. What the machinery does as a whole, and how it moves, remained opaque.

Cryo-electron microscopy changed everything. Where X-ray crystallography demanded rigid, homogenous crystals that giant, flexible assemblies rarely form, cryo-EM can capture heterogeneous, dynamic complexes frozen in vitreous ice, and computational classification can sort millions of particle images into distinct conformational states. Structures of the human pre-initiation complex, or PIC, the mega-assembly of polymerase II and general transcription factors on promoter DNA, progressed from low-resolution models to structures resolved at near-atomic detail. In 2021, researchers resolved the polymerase II pre-initiation complex at 2.9 angstroms, revealing the initial stages of DNA opening, and subsequent work captured the PIC in complex with the Mediator co-activator, with TFIIH, and even with the first nucleosome downstream of the promoter.

At the heart of the new understanding is TFIID, a massive 1.2-megadalton complex of the TATA-box-binding protein and thirteen TBP-associated factors, or TAFs. TFIID serves as the primary scaffold for promoter recognition. In its free state, it adopts a compact, autoinhibited conformation in which the lobes that will grip promoter DNA are folded inward. Binding of TBP to the TATA box triggers a dramatic rearrangement, and recent structures show that TFIID engages not only the TATA box but also downstream promoter elements, including the downstream core promoter region, or DPR, discovered through machine-learning analysis of human promoter sequences. Because the majority of human promoters lack a TATA box altogether, these TAF-mediated contacts with downstream DNA are essential for accurate initiation across most of the genome, and different TAF subunits are now known to dictate promoter selectivity and to integrate signals from sequence-specific activator proteins.

Superimposed on this scaffold is Mediator, the thirty-subunit co-activator that physically bridges enhancer-bound transcription factors to the promoter-bound machinery. Structural studies of the complete Mediator-PIC complex, including landmark human structures published in 2021, revealed how Mediator’s head, middle, and tail modules wrap around the polymerase and how its dissociable kinase module can regulate the interaction. The Pol II C-terminal domain, an intrinsically disordered tail of heptad repeats extending from the largest polymerase subunit, threads into the PIC-Mediator assembly, and its phosphorylation by the TFIIH-associated kinase module, through cyclin-dependent kinase 7, marks a critical checkpoint. Serine 5 phosphorylation of the CTD not only helps trigger promoter escape but also recruits the mRNA capping enzyme, coupling the earliest steps of transcription to RNA processing.

Perhaps the most striking theme to emerge from recent structural work is the role of chromatin itself. The +1 nucleosome, the first nucleosome downstream of the transcription start site, is no longer viewed as a mere obstacle. Structures of the +1 nucleosome-bound PIC-Mediator complex show that this histone package acts as a position-dependent regulator, physically orienting the pre-initiation complex, influencing transcription start-site selection, and even enforcing directionality. Epigenetic marks matter here: trimethylation of histone H3 lysine 4 anchors TFIID through the bromodomain of TAF3, while acetylation marks recruit BET-family factors that bridge TFIID and nucleosomes. The nucleosome, in other words, transmits epigenetic information directly into the transcription machinery, helping the PIC integrate genetic sequence and chromatin state into a single decision about where and in which direction to initiate.

The transition from initiation to elongation emerges as a choreographed, programmed process rather than a simple handoff. Before productive synthesis begins, polymerase II must melt the promoter DNA, a task performed primarily by the XPB translocase subunit of TFIIH, which pulls double-stranded DNA into the polymerase cleft to generate torsional stress. Early attempts at synthesis are often abortive: the enzyme repeatedly synthesizes and releases short RNAs of fewer than five nucleotides because the nascent transcript is thermodynamically unstable in the complex. During this phase, the enzyme scrunches downstream DNA into its active site, storing mechanical energy. Structural and single-molecule studies now delineate a three-step mechanism of promoter escape, in which steric pressure from accumulated template DNA and forward translocation of the RNA-DNA hybrid rupture the narrow channel formed by Pol II and TFIIB, causing general transcription factors to dissociate, the transcription bubble to collapse from roughly twenty-one to a stable eleven nucleotides, and the polymerase to commit to elongation, typically after transcribing about eight to twenty-three nucleotides.

Crucially, the review frames transcription initiation as a process governed by surface competition. Each stage of initiation involves factors that bind overlapping or competing surfaces on polymerase II. As CTD phosphorylation proceeds, inhibitory factors are released and regulatory factors are progressively exchanged: Mediator dissociates, DSIF and NELF bind to impose promoter-proximal pausing, and elongation factors await release signals. This model explains how the same binding surfaces can recruit different partners at different times, and how re-initiation, in which scaffold factors remain promoter-bound to launch successive rounds of polymerase loading, can generate the transcriptional bursting observed in single cells. It also rationalizes why paused polymerase can inhibit new initiation, creating feedback loops that shape gene expression dynamics.

The medical implications are substantial. TAF subunits such as TAF1 and TAF10 are required during development but dispensable in certain adult tissues, and PROTAC-mediated degradation of TAF1 has been shown to induce apoptosis in acute myeloid leukemia cells, making components of the initiation machinery emerging drug targets. As cryo-EM continues to push toward imaging these assemblies inside cells, and as single-molecule and nascent-RNA sequencing methods add temporal resolution, the field is converging on a unified view: transcription initiation is not a static lock-and-key event but a dynamic, chromatin-informed, programmed pipeline, and seeing it in atomic detail is finally telling us how life reads its own blueprint.

Subject of Research: Structural and mechanistic basis of transcription initiation by RNA polymerase II in eukaryotes

Article Title: The molecular basis of transcription initiation by RNA polymerase II

Article References: Chen, X., & Xu, Y. (2026). The molecular basis of transcription initiation by RNA polymerase II. Nature Reviews Molecular Cell Biology. https://doi.org/10.1038/s41580-026-01019-2

Image Credits: AI Generated

DOI: 10.1038/s41580-026-01019-2

Keywords: RNA polymerase II, transcription initiation, pre-initiation complex, TFIID, Mediator, cryo-electron microscopy, TATA box, promoter escape, plus-one nucleosome, CTD phosphorylation, TFIIH, gene regulation