Core Promoter Elements: TATA, Initiator and CpG-Island Promoters

A promoter is where the pre-initiation complex assembles, but "the promoter" describes at least three architecturally distinct systems. A TATA-box promoter with a single dominant TSS behaves differently from a broad CpG-island promoter with dozens of start sites, and both differ from a TATA-less promoter relying on Initiator-like sequences.

Those differences are not cosmetic. They determine how much a promoter's output fluctuates between cells, how sensitive it is to distal enhancers, which factors you can expect to see bound, and — practically — which experimental perturbation will produce an interpretable result.

The enhancer side of this picture is covered in our pillar article on promoters, enhancers and 3D genome organization; this piece stays inside the core promoter.

Two cautions about the table. First, these are tendencies from genome-wide studies, not rules that predict a specific gene's behaviour; there are well-documented exceptions in every row. Second, the categories intergrade — many promoters combine a weak TATA-like element with a CpG island, and assigning them to a class is a judgement call that should be stated rather than assumed.

Dissection starts with knowing where transcription actually begins:

• • CAGE or 5′-end sequencing gives TSS clusters and their relative usage, which is more informative than a single annotated TSS.
• • ATAC-seq and histone marks indicate nucleosome depletion and active chromatin, but do not identify the TSS by themselves.
• • Conservation and motif scanning suggest candidate elements; they do not demonstrate function.

A frequent error is designing edits against an annotated TSS that is not the dominant start site in the cell type under study. Checking TSS usage in your cells takes an afternoon and prevents a misinterpreted experiment.

Three approaches, in increasing order of precision:

Deletion series. Progressive 5′ deletions of a promoter fragment in a reporter construct define the minimal region supporting expression. Fast, but conflates spacing, nucleosome positioning, and motif loss, and is performed outside native chromatin.

Motif mutation by editing. Single-base changes in a TATA box or Initiator convert the question from "which region" to "which bases". Because these elements are short, edits are compact and interpretable, and the effects are usually large when the motif is genuinely required.

Motif swap and spacing variants. Replacing one core element with another, or shifting it by a few bases, tests whether the requirement is sequence-specific or positional. This is the step that distinguishes "TATA box required" from "some element at approximately this position required".

All three are stronger when performed at the endogenous locus, which is why base editing of core motifs in a validated cell background is often the most publishable version of a promoter study.

Can tell you: whether a motif is required for promoter output in the tested cell type; the magnitude of that requirement; whether spacing or orientation matters; whether the promoter's output is enhancer-dependent in that context.

Cannot tell you: which enhancers act on it (that requires the enhancer-side work); whether the promoter is rate-limiting for the gene in vivo; how the promoter behaves across all cell states; whether a variant in the promoter is pathogenic (that requires allele-level evidence).

Promoter strength measured in a reporter is also not transferable across contexts: the same promoter can differ by an order of magnitude between cell types, and CpG-island promoters are particularly sensitive to the methylation state of the host cell.

The last row is specific to CpG-island promoters and is regularly missed: sequence-correct constructs can be silenced by DNA methylation, and treatment with a demethylating agent will reveal whether that is the mechanism.

EDITGENE builds promoter-edit series by base editing (motif disruption, swap, spacing variants), endogenous TSS reporter knock-ins, and isogenic promoter-variant panels, with TSS usage confirmed in the target cell background.

Tell us the gene, the cell type, and the element you suspect. Contact EDITGENE to request a project consultation or quote.

Promoter classes overlap and class assignments are approximate; expression behaviour depends on cell type, methylation state, and chromatin context, so conclusions from one background should not be generalized without testing.

Reporter-based promoter strength is not equivalent to endogenous promoter output; where the claim concerns native regulation, edits at the endogenous locus are required.

• • Kadonaga JT. Perspectives on the RNA polymerase II core promoter. WIREs Developmental Biology. 2012;1:40–51.
• • Haberle V, Stark A. Eukaryotic core promoters and the functional basis of transcription initiation. Nature Reviews Molecular Cell Biology. 2018;19:621–637.
• • Carninci P, et al. Genome-wide analysis of mammalian promoter architecture and evolution. Nature Genetics. 2006;38:626–635.
• • Vo Ngoc L, et al. The RNA polymerase II core promoter in Drosophila. Genetics. 2019;212:13–24.
• • Rheinbay E, et al. Analyses of non-coding somatic drivers in 2,658 cancer whole genomes. Nature. 2020;578:102–111.