Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters

Abstract

RNA polymerases are highly regulated molecular machines. We present a method (global run-on sequencing, GRO-seq) that maps the position, amount, and orientation of transcriptionally engaged RNA polymerases genome-wide. In this method, nuclear run-on RNA molecules are subjected to large-scale parallel sequencing and mapped to the genome. We show that peaks of promoter-proximal polymerase reside on approximately 30% of human genes, transcription extends beyond pre-messenger RNA 3' cleavage, and antisense transcription is prevalent. Additionally, most promoters have an engaged polymerase upstream and in an orientation opposite to the annotated gene. This divergent polymerase is associated with active genes but does not elongate effectively beyond the promoter. These results imply that the interplay between polymerases and regulators over broad promoter regions dictates the orientation and efficiency of productive transcription.

Fig. 1

Sample of GRO-seq data view on the University of California at Santa Cruz (UCSC) genome browser. A 2.5-Mb region on chromosome 5 showing GRO-seq reads aligned to the genome at 1-bp resolution, followed by an up-close view around the NPM1 gene. Pol II ChIP results (3) are shown in green; mappable regions, black; GRO-seq reads on the plus strand (left to right), red; GRO-seq reads on the minus strand (right to left), light blue; RefSeq gene annotations, dark blue.

Fig. 2

Alignment of GRO-seq reads to TSSs and 3′ ends. (A) GRO-seq reads aligned to Ref-seq TSSs in 10-bp windows in both sense (red) and antisense (blue) directions relative to the direction of gene transcription. (B) GRO-seq reads flanking the 3′ ends of genes. The sharp peak coincides with the new 5′ end created after cleavage at the poly-A site. Polymerase density extends considerably downstream before termination.

Fig. 3

Comparison of pausing with gene activity. Four classes of genes are found when comparing genes with a paused polymerase and transcription activity either by microarray or GRO-seq density in the downstream portions of genes. An example of each class is shown, with tracks shown in the UCSC genome browser as in Fig. 1. The gene names, pausing index, and P value, from top to bottom, respectively, are as follows: TRIO, 1.1, 0.62; FUS, 41, 2.8 × 10⁻⁴³; IZUMO1, 410, 7.6 × 10⁻³; and GALP (which has no reads and therefore no pausing index). The number of genes represented in each class is shown to the right.

Fig. 4

Correlation of promoter-proximal transcription patterns with gene activity. (A to D) Box plots (each showing the fifth, 25th, 50th, 75th, and 95th percentiles) that show the relationship of promoter-proximal (PP) sense peaks (red), divergent peaks (DP) (blue), pausing indices (green), and PP/DP ratios (orange) to the top, middle, and bottom deciles of gene activity. All deciles are significantly different from each other: P <10⁻⁹ for all comparisons except between the lowest and the middle deciles in (D) (P < 10⁻³). (E) ChIP profiles of Pol II and GRO-seq sense (S) and antisense (AS) strand reads aligned to TSSs. (F) ChIP profiles of H3ac and H3K4me2 and GRO-seq aligned to TSSs.