= Solution
\b[First establish whether X changes where <RNA polymerase II> initiates, rather than merely changing which initiated <RNAs> survive.] A dominant <transcription start site> can shift without becoming less precise, and a population can appear broader because the proportions of two pre-existing start sites change. Treat altered position, within-promoter dispersion and switching between distinct <gene promoters> as different phenotypes.
Use a dose and short time course of X with vehicle, biological replicates, washout and viability measurements. Map capped $5'$ ends by <CAGE> and independently map newly initiated <RNA> with <GRO-cap> or another suitably validated nascent-start assay. <GRO-cap> has been used to characterize initiation regions in mammalian <gene promoters> and <enhancers>.<Analysis of nascent RNA identifies a unified architecture of initiation regions at mammalian promoters and enhancers by Leighton J. Core and others>[Abstract and Methods: GRO-cap mapping of transcription initiation]{ref} Confirm selected $5'$ ends with an independent targeted method that distinguishes capped starts from processed <RNA> ends. Use external standards and control cap recovery; if X inhibits capping, a capped-end assay can lose genuine starts. Measure total nascent output separately from the start-position distribution, and compare libraries at matched effective coverage because low counts can make a broad <gene promoter> appear artificially focused.
Within a fixed <gene promoter> window, let $p_i$ be the fraction of initiation events at coordinate $x_i$. Quantify dominant position, central width and, if useful, <Shannon entropy>
$$
\mu=\sum_i p_i x_i,\qquad \sigma^2=\sum_i p_i(x_i-\mu)^2,\qquad H=-\sum_i p_i\log p_i.
$$
A one-base shift of a single occupied start changes $\mu$ while leaving $\sigma^2=H=0$; splitting equally between two positions increases dispersion and gives $H=\ln 2$. Compare complete distributions with sampling uncertainty rather than interpreting one summary alone. Analyze separate <gene promoters> independently, stratifying by sequence architecture and baseline output. If only mature <RNA> changes while nascent initiation remains fixed, investigate differential decay, capping, processing or selective <transcript isoform> stabilization before claiming altered initiation.
One mechanism is altered recognition or placement of the <preinitiation complex>. <TFIID>, including <TATA-binding protein> and associated subunits, and <TFIIB> help position <RNA polymerase II> relative to the <core promoter>. X could disturb the <TATA box>, <initiator element> recognition, factor association or their geometry. Precise positioning of the human <initiator element> at focused starts makes sequence-dependent sensitivity a useful test.<The human initiator is a distinct and abundant element that is precisely positioned in focused core promoters by Long Vo ngoc and others>[Abstract and Results: focused promoters and initiator positioning]{ref} Map factor occupancy, compare X-sensitive and insensitive <core promoters>, and use matched reporter constructs differing only in an <initiator element>, <TATA box> or their spacing. A reporter must have its $5'$ ends mapped: light output alone cannot reveal an altered <transcription start site>. Reconstitute initiation on naked <DNA> with purified factors and add X. Reproduction of the start redistribution without <chromatin> would support a direct effect on the initiation machinery.
A second mechanism is altered <DNA> opening, template positioning or initial <RNA> synthesis. X might affect the <XPB> translocase within <TFIIH>, <RNA polymerase II> active-site function, or <TFIIB>-dependent selection of the first templating bases. Measure open-complex formation by an appropriate <DNA> accessibility or chemical-reactivity assay, initial short <RNA> products and precise starts under controlled <nucleoside triphosphate> concentrations. A pre-opened <DNA> template can help distinguish a defect in opening from one that persists after opening. <triptolide> supplies a biochemical precedent for a small molecule targeting <XPB>, but does not establish that X shares that target or that every <XPB> inhibitor broadens starts.<XPB, a subunit of TFIIH, is a target of the natural product triptolide by Denis V. Titov and others>[Abstract and biochemical target-identification experiments]{ref} Mammalian initiation should be tested directly; extensive scanning rules established in a different organism cannot simply be assumed to apply.
A third mechanism is a change in the <chromatin> landscape or <DNA supercoiling>. Altered <nucleosome> positioning could expose cryptic <gene promoters>, while changed <histone acetylation> or <chromatin remodelling> could broaden the region accessible to the <preinitiation complex>. Map <chromatin> accessibility, <nucleosomes> and relevant <histone> modifications before and soon after X treatment. Compare naked and nucleosomal <DNA> templates carrying the same <core promoter>; mutate the proposed <chromatin remodelling> factor or use acute depletion to test whether it is required. For a proposed <topoisomerase> effect, compare relaxed and supercoiled templates and measure <DNA supercoiling> directly. Occupancy changes that follow initiation changes can be consequences, so temporal order matters.
A fourth route is indirect alteration of factor supply or activity: X might change <nucleoside triphosphate> pools, <phosphorylation> of initiation machinery, or the abundance of a limiting <transcription factor>. Measure candidate metabolites and <protein> modifications, test rapid versus delayed onset, and restore the proposed limiting input. Altered promoter-proximal pausing can change the recovery of nascent products without moving the original start; compare initiation-specific maps with downstream engaged-<RNA polymerase II> profiles. A change that disappears after controlling <RNA> lifetime or pausing is not evidence for loss of start-site selection.
Finally, identify the direct molecular target through binding or engagement assays and a resistant <mutant> that retains normal baseline function. Require the resistant version to rescue the start-position phenotype at an X concentration that affects the normal version; overexpression rescue alone can reflect sequestration of the compound. An acute genetic perturbation should phenocopy X, and purified biochemical assays should reproduce the relevant effect. \b[Convergent evidence from nascent starts, controlled templates and target-specific rescue distinguishes direct selection defects from secondary changes in <gene expression>.]
Back to article page
