Primary microRNA processing is functionally coupled to RNAP II transcription in vitro

Abstract

Previous studies in vivo reported that processing of primary microRNA (pri-miRNA) is coupled to transcription by RNA polymerase II (RNAP II) and can occur co-transcriptionally. Here we have established a robust in vivo system in which pri-miRNA is transcribed by RNAP II and processed to pre-miRNA in HeLa cell nuclear extracts. We show that both the kinetics and efficiency of pri-miRNA processing are dramatically enhanced in this system compared to that of the corresponding naked pri-miRNA. Moreover, this enhancement is general as it occurs with multiple pri-miRNAs. We also show that nascent pri-miRNA is efficiently processed before it is released from the DNA template. Together, our work directly demonstrates that transcription and pri-miRNA processing are functionally coupled and establishes the first in vivo model systems for this functional coupling and for co-transcriptional processing.

Figure 1. RNAP II txn/pri-miRNA processing system in vitro.

(a) Structure of the CMV DNA template encoding let-7a pri-miRNA used for txn/pri-miRNA processing. The sizes of 5′ flanking region, 3’ flanking region, and pre-let-7a are indicated. The thick line indicates the natural pri-miRNA sequences and the thin lines indicate the vector sequences. CMV-Δpre-let-7a is a mutant lacking the 72 nt pre-let-7a hairpin. (b) RNAP II txn/pri-miRNA processing was carried out for the indicated times using the CMV DNA template encoding let-7a pri-miRNA (lanes 1, 2), the CMV DNA template encoding let-7a Δpre-miRNA DNA template (lanes 3, 4) or no DNA template (lanes 5, 6). The asterisk indicates the 3′ flanking region, which lacks a cap and is therefore largely degraded during the reaction (lane 2). (c) RNAP II txn/pri-miRNA processing was carried using different amounts of MgCl₂ as indicated and the CMV DNA template encoding let-7a pri-miRNA. Size markers (in base pairs) and RNA species are indicated. Ori indicates the gel origin. RNAs were detected by phosphorimager and quantified using Quantity One software. Mean ± S.D. of three biological replicates are shown.

Figure 2. Identification of RNA species generated in the RNAP II txn/processing system.

(a) Predicted secondary structure of let-7a pri-miRNA showing the binding sites of the X, Y and Z DNA oligos used for oligonucleotide-directed RNAse H cleavage. (b) Schematic of let-7a pri-miRNA showing RNAse H cleavage sites in the 5′ flanking region, 3′ flanking region, and pre-let-7a loop using the X, Y and Z DNA oligos. (c) No oligo (lane 1) or the indicated oligos (lanes 2–7) were used for RNAse H cleavage of total RNA isolated from the txn/pri-miRNA processing system (lanes 2–4) or carried out directly in the txn/pri-miRNA processing system after completion of the reaction (at the 10 min time point). The RNAse H cleavage products or pri-miRNA processing products are indicated by an asterisk to the left of the gel lanes and are described in panel d from top to bottom for each gel lane in panel c.

Figure 3. Let-7a pri-miRNA processing is functionally coupled to RNAP II transcription.

(a,b) CMV DNA template encoding let-7a pri-miRNA (a) or the corresponding T7 let-7a pri-miRNA (b) was incubated in nuclear extract in the txn/pri-miRNA processing system for the indicated times. Markers (in base pairs) and RNA species are indicated. (c) RNAs were detected by phosphorimager and quantified using Quantity one software. The ratios of pre-let-7a signal at each point (lanes 2–5) to pri-let-7a signal at the start of the reaction (lane 1) were calculated and mean ± S.D. of three biological replicates are shown. The difference in pre/pri ratio between the two groups at each time point was analyzed by Student’s t-test.(d) CMV DNA template encoding let-7a pri-miRNA (lanes 1–4) or naked CMV let-7a pri-miRNA (lanes 5–8) was incubated in nuclear extract under txn/pri-miRNA reaction conditions for the indicated times. Markers (in base pairs) and RNA species are indicated. Ori marks the gel origin. (e) Same as c except that the data in panel d were quantitated. Mean ± S.D. of three biological replicates are shown, and the difference in pre/pri ratio between the two groups was examined using Student’s t-test.

Figure 4. Pri-let-7a miRNA processing occurs on an immobilized DNA template.

(a) Schematic of CMV pri-let-7a DNA template bound to streptavidin magnetic beads via biotin at the 3′ end. (b) CMV pri-let-7a DNA templates containing biotin or no biotin were incubated with streptavidin magnetic beads. DNA templates in bound (bd) and supernatant (sp) fractions were analyzed on a 1% agarose gel by staining with ethidium bromide. (c) Biotinylated CMV pri-let-7a DNA template was 3′ end-labeled, bound to streptavidin magnetic beads and incubated in nuclear extract at 30^o for 30 min. Bound and supernatant fractions were analyzed on an 8% denaturing polyacrylamide gel. (d) CMV pri-let-7a DNA templates containing biotin or no biotin were incubated in nuclear extracts under txn/processing conditions for the indicated times. Total RNA was isolated and fractionated on an 8% denaturing polyacrylamide gel. (e) Streptavidin magnetic beads were incubated with CMV pri-let-7a DNA templates containing biotin or no biotin. After washing, beads with or without immobilized DNA templates were incubated in nuclear extracts under txn/processing conditions for the times indicated. Total RNA from each sample was isolated and fractionated on an 8% denaturing polyacrylamide gel. (f) Immobilized CMV pri-let-7a DNA template was incubated under txn/processing conditions for the times indicated. The bound and supernatant fractions were separated, and the bound fraction was washed. Total RNA was then isolated from the bound and supernatant fractions and run on an 8% denaturing polyacrylamide gel.