Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Jul 16;59(2):203-16.
doi: 10.1016/j.molcel.2015.05.033. Epub 2015 Jul 2.

Uridylation of RNA Hairpins by Tailor Confines the Emergence of MicroRNAs in Drosophila

Madalena M Reimão-Pinto  1 , Valentina Ignatova  1 , Thomas R Burkard  1 , Jui-Hung Hung  2 , Raphael A Manzenreither  1 , Ivica Sowemimo  1 , Veronika A Herzog  1 , Brian Reichholf  1 , Sara Fariña-Lopez  1 , Stefan L Ameres  3
Affiliations

Uridylation of RNA Hairpins by Tailor Confines the Emergence of MicroRNAs in Drosophila

Madalena M Reimão-Pinto et al. Mol Cell. .

Abstract

Uridylation of RNA species represents an emerging theme in post-transcriptional gene regulation. In the microRNA pathway, such modifications regulate small RNA biogenesis and stability in plants, worms, and mammals. Here, we report Tailor, an uridylyltransferase that is required for the majority of 3' end modifications of microRNAs in Drosophila and predominantly targets precursor hairpins. Uridylation modulates the characteristic two-nucleotide 3' overhang of microRNA hairpins, which regulates processing by Dicer-1 and destabilizes RNA hairpins. Tailor preferentially uridylates mirtron hairpins, thereby impeding the production of non-canonical microRNAs. Mirtron selectivity is explained by primary sequence specificity of Tailor, selecting substrates ending with a 3' guanosine. In contrast to mirtrons, conserved Drosophila precursor microRNAs are significantly depleted in 3' guanosine, thereby escaping regulatory uridylation. Our data support the hypothesis that evolutionary adaptation to Tailor-directed uridylation shapes the nucleotide composition of precursor microRNA 3' ends. Hence, hairpin uridylation may serve as a barrier for the de novo creation of microRNAs in Drosophila.

PubMed Disclaimer

Figures

None
Graphical abstract
Figure 1
Figure 1
Post-transcriptional Modifications of Small RNAs in Drosophila (A) Identification of post-transcriptional modifications in small RNA libraries. Genome-matching (GM) reads map perfectly to the genome; prefix-matching (PM) reads contain non-genome-matching nucleotides at the 3′ end (red). (B) Abundance of genome-matching and prefix-matching miRNAs (in parts per million; ppm) in small RNA libraries generated from S2 cells (left) or adult male flies (right). S2 cell data represented as mean of three biological replicates ± SD. miRNAs with > 5% PM reads are indicated in black. (C) Classification of miRNAs according to Argonaute loading (miR and miR∗), location in the pre-miRNA hairpin (5p- and 3p-miRNA), and mechanism of biogenesis (canonical miRNAs and mirtrons). (D) Tailing analysis of small RNA deep sequencing datasets from S2 cells (left panel) and whole male flies (right panel). The fraction of tailed reads was calculated for each miRNA and represented as Tukey boxplots for all miRNAs (black boxes) or subsets of miRNAs as classified in (C) (white boxes). The number of miRNAs for each subset is indicated. p values (Mann-Whitney test) are indicated. (E) Non-genome-matching nucleotide additions to miRNAs consist mostly of uridine and adenine. The nucleotide composition of non-genome-matching additions was determined for each miRNA and averaged across all miRNAs. See also Figure S1.
Figure 2
Figure 2
The Cytoplasmic TNTase CG1091/Tailor Is Required for miRNA Uridylation and Normal Fertility in Flies (A) Domain organization of known and putative TNTases in flies based on InterPro database. The characteristic domain structure consists of a nucleotidyltransferase domain (red box) and a PAP/25A-associated domain (blue). Expression in S2 cells was based on modENCODE mRNA sequencing (Cherbas et al., 2011). (B) Post-transcriptional modification of miR-184-3p in S2 cells is detectable by high-resolution northern hybridization. Higher-molecular-weight bands of miR-184 (> 23 nt) correspond to prefix-matching reads (red) as evidenced by small RNA sequencing. (C) CG1091 is required for post-transcriptional modification of miR-184. Upon depletion of the indicated candidate TNTases in S2 cells by RNAi miR-184 was detected by high-resolution northern hybridization. Double-stranded RNA targeting green fluorescent protein (GFP) or luciferase (LUC), and untreated S2 cells served as controls. 2S rRNA served as loading control. (D) Schematic representation of CG1091 gene and RNA transcripts. CG1091-RB is predominantly expressed in S2 cells (see Figure S2C). The region targeted by dsRNA for depletion of CG1091 by RNAi in S2 cells, the location of a 7-bp deletion introduced by CRISPR/Cas9 genome editing in flies (CG1091c4-1/6), and a piggy-bac insertion into the coding sequence of CG1091 (CG1091f05717) are indicated. (E) RNAi in S2 cells and CG1091c4-1/6 in flies depleted CG1091 protein to levels that are not detectable by western blotting. Actin represents loading control. Asterisks indicate non-specific signal in lysates of male flies. (F) Depletion of CG1091 affects tailing of miRNAs in S2 cells (left) and in flies (right). The fraction at which each miRNA is modified is plotted. S2 cell data represented as mean of three independent biological replicates ± SD. miRNAs that show a significant depletion in post-transcriptional modifications are indicated in gray (p < 0.05, Student’s t test; FDR < 0.1; Benjamini and Hochberg, 1995). A > 2-fold reduction in tailing is indicated in red. (G) Pair-wise comparison of miRNA tailing status revealed a statistically significant decrease upon depletion of CG1091 in S2 cells and in flies. p values (Wilcoxon matched-pairs signed rank test) are indicated. (H) Depletion of CG1091 impacts miRNA uridylation. Abundance and composition of post-transcriptionally added nucleotides was determined for each miRNA upon CG1091 depletion, normalized to control samples (control dsRNA in S2 cells and w1118 in flies), and averaged across all miRNAs. (I) Upon CG1091 depletion, mirtrons are significantly more depleted in non-genome-matching nucleotide additions when compared to canonical miRNAs in both S2 cells and in flies. Change in tailing was determined for the indicated classes of miRNAs. p values (Mann-Whitney test) are indicated. (J and K) Depletion of CG1091 in flies affects male (J) and female (K) fertility. Fertility was determined in CG1091c4-1/c4-6 and CG1091c4-6/f05717 flies and compared to control (w1118) flies. p values (Student’s t test) are indicated. Data represented as mean ± SD. (L) In S2 cells, Myc-tagged CG1091 localizes to the cytoplasm. FLAG-Myc-tagged CG1091 expression was driven by an Actin5C promoter (pAFMW-CG1091) upon transient transfection in S2 cells, followed by immunostaining (anti-Myc) and imaging. Single color channel images show total inversions for DAPI and Myc-CG1091. Scale bar = 5 μm. See also Figure S2.
Figure 3
Figure 3
Tailor-Dependent Uridylation Impacts Mature miRNA Levels and Prevents Mirtron Accumulation (A) Greater than 40% of all miRNAs are significantly changed in abundance upon Tailor depletion in S2 cells (p < 0.05, Student’s t test; FDR < 0.1; Benjamini and Hochberg, 1995). Data represented as mean ± SD. (B) Mirtrons significantly increase in abundance upon depletion of Tailor. p values (Mann-Whitney test) are indicated. (C) Changes in miRNA abundances correlate with post-transcriptional modifications. The fraction at which miRNAs are tailed in control-dsRNA-treated S2 cells is indicated for miRNAs that do, or do not, change significantly in abundance upon depletion of Tailor. p value (Mann-Whitney test) is indicated. (D) Changes in miRNA abundance correlate with Tailor-directed tailing. Changes in miRNA tailing between Tailor-depleted and control-dsRNA-treated S2 cells is shown for miRNAs that do, or do not, change significantly in abundance upon depletion of Tailor. p value (Mann-Whitney test) is indicated. See also Figure S3.
Figure 4
Figure 4
Tailor Regulates miRNA Abundance by Changing the Accuracy of Pre-miRNA 3′ Overhangs in Gain-of-Function Experiments (A) Expression of Tailor in S2 cells (CG1091OE) changes the abundance and tailing status of mature miRNAs when compared to untreated S2 cells in northern hybridization experiments. MicroRNAs originating from the 3p or 5p arm of pre-miRNAs are indicated. 2S rRNA represents loading control. (B) Quantification of three independent biological replicates of (A). Relative abundance of each miRNA in CG1091OE compared to untreated S2 cells is shown. p values (Student’s t test) are indicated. Data presented as mean ± SD. (C) Tailor modifies pre-miRNAs. The indicated pre-miRNAs from CG1091OE or untreated S2 cells were detected by northern hybridization. Dcr-1 was depleted by RNAi to increase pre-miRNA signals. (D and F) Tailor changes the accuracy of pre-miRNA 3′ ends. Sanger sequencing results of the indicated number of clones report the 3′ end of pre-miRNA-mapping reads. Genome-matching (GM, black) and prefix-matching (PM, red) reads are depicted. The fraction of reads mapping to the predicted 2-nt 3′ overhang is shown. (E and G) Efficient Dcr-1-directed pre-miRNA processing requires accurate 2-nt 3′ overhangs. In vitro dicing assays employed SBP-tagged affinity-purified Dcr-1 and the indicated 5′ radiolabelled synthetic pre-miRNAs. Data presented as mean ± SD. p values (Student’s t test) are indicated. See Figure S4 for primary data. See also Figure S4.
Figure 5
Figure 5
Pre-miRNAs, Particularly Mirtron Hairpins, Are Physiological Substrates for Tailor-Directed Destabilization through Tailing (A) Mapping results of high-throughput sequencing of 40- to 100-nt RNAs from S2 cells. (B) Abundance of genome-matching and prefix-matching pre-miRNAs in S2 cells. Data presented as mean of three independent biological replicates ± SD. Pre-miRNAs with highest fraction of prefix-matching reads (> 5%) are indicated in black. (C) Mirtrons are significantly more tailed compared to canonical pre-miRNAs. The fraction at which pre-miRNAs are tailed is shown for all miRNAs and the indicated classes of pre-miRNAs. The number of pre-miRNAs in each group is indicated. p value (Mann-Whitney test) is indicated. (D) Pre-miRNA tails consist mostly of uridine and adenine. The nucleotide composition of non-genome-matching additions was determined for each pre-miRNA and averaged across all pre-miRNAs. (E) Depletion of Tailor by RNAi affects tailing of pre-miRNAs in S2 cells. The fraction at which each pre-miRNA is modified is plotted. miRNAs that show > 2-fold reduction in tailing are indicated in black. (F) Depletion of Tailor by RNAi in S2 cells impacts pre-miRNA tailing. p values (Wilcoxon matched-pairs signed rank test) are indicated. (G) Depletion of Tailor by RNAi impedes addition of uridine, but not any other nucleotides, to pre-miRNAs. Abundance and composition of nucleotide additions was determined for each pre-miRNA, normalized to control treatment, and averaged across all pre-miRNAs. (H) Mirtron-hairpin uridylation requires Tailor. Change in tailing upon Tailor depletion in S2 cells by RNAi was determined for classified pre-miRNAs and normalized to control treatment. p value is indicated (Mann-Whitney test). (I) Tailor-directed uridylation destabilizes pre-miRNAs. Upon Tailor depletion, the levels of pre-miRNAs that are tailed by Tailor under unperturbed conditions increase significantly. p values are indicated (Mann-Whitney test). (J) Uridylation inhibits pre-miRNA dicing. In vitro dicing assays employed SBP-tagged affinity-purified Dcr-1 and the indicated 5′ radiolabelled synthetic pre-miRNA. (K) Quantification of data shown in (J). Data presented as mean ± SD. p values are indicated (Student’s t test). See also Figure S5 and Table S1.
Figure 6
Figure 6
Tailor Is a Bona Fide TUTase with Unique Targeting Properties (A) Immunopurified Tailor exhibits TNTase activity. FLAG-tagged Tailor was expressed in S2 cells, immunopurified, and incubated with a 22-nt, 5′ radiolabelled RNA in the presence of the indicated ribonucleotide triphosphate. (B) Tailor-directed nucleotide transfer requires Mg2+. Tailing reactions were performed using a 5′ radiolabeled substrate RNA described in (A) in the presence of rNTPs. Addition of EDTA inhibited the tailing reaction, whereas excess Mg2+ rescued the activity. (C–G) High-throughput biochemical characterization of Tailor-directed RNA tailing. A 37-nt RNA substrate containing four random nucleotides at the 3′ end was subjected to in vitro tailing reactions using immunopurified Tailor for 2 or 5 min, in the presence of rNTP, followed by 3′ adaptor ligation and high-throughput sequencing. (C) Validation of high-throughput tailing assay. Product of in vitro tailing reactions was resolved by denaturing PAGE (left) or analyzed by high-throughput sequencing (right). Sequencing reveals high selectivity for UTP incorporation despite even concentration of rNTPs in the tailing reaction. (D) Tailor-directed tailing efficiency is not influenced by substrate secondary structures. Substrates were binned according to secondary structure stability (effective free energy; EFE) and analyzed for the fraction at which each substrate was tailed, averaged across two time points (2 and 5 min). Median (white line), inner quartile range (IQR; dark gray area), and 1.5 IQR of lower and upper quartile (light gray area) are shown. No statistically significant difference was detected. (E) Substrate secondary structures impact Tailor processivity. Representation as described in (D), but showing for each substrate the mean length of the tail added to each substrate. Stars indicate significantly different tail length of individual EFE groups (Mann-Whitney test). (F) Tailor exhibits primary sequence specificity. The nucleotide identity of the four random 3′ nucleotides (irrespective of their position) in the indicated substrate RNAs is shown. (G) Tailor exhibits specificity for the 3′-terminal nucleotide. Cumulative distributions of the relative fraction tailed for the indicated substrate RNAs is shown. See also Figure S6.
Figure 7
Figure 7
Tailor Targets Non-conserved Hairpins Ending in 3′G to Confine miRNA Production (A) Tailor directs uridylation of miRNAs ending in 3′G. The nucleotide identity preceding non-genome-matching nucleotides of mature miRNAs (left panel) and pre-miRNAs (middle panel) in S2 cells and mature miRNAs in 0- to 5-day-old male whole flies (right panel) is shown. The number of analyzed (pre-) miRNAs is indicated. (B) Efficient Tailor-directed tailing of the mirtron-hairpin miR-1003 requires a 3′G. 5′ radiolabeled pre-miR-1003 containing the indicated 3′-terminal nucleotide was incubated with immunopurified Tailor for the indicated time, followed by denaturing gel electrophoresis and phosphorimaging. (C) Quantification of three independent replicates of the experiment shown in (B). Data presented as mean ± SD. (D) Conserved pre-miRNAs, but not recently evolved hairpins, are depleted in 3′ guanosine. Nucleotide frequency at the 3′ terminal position of high-confidence Drosophila melanogaster pre-miRNAs annotated in mirBase as well as the overall nucleotide frequency in the fly genome is depicted. (E) Relative nucleotide frequency of the data shown in (D). p values (binomial test) are indicated. (F) Non-conserved pre-miRNAs are more efficiently tailed compared to conserved hairpins. Fraction tailed for pre-miRNAs in S2 cells is shown. p value is indicated (Mann-Whitney test). (G) miRNAs that were detected by pre-miRNA—but not mature miRNA—cloning in S2 cells are enriched in non-conserved hairpins. Black bars represent conserved, and white bars non-conserved, pre-miRNAs. The number of hairpins in each dataset is indicated. (H) Non-conserved miRNAs are significantly upregulated upon depletion of Tailor in S2 cells when compared to conserved miRNAs. p value is indicated (Mann-Whitney test). See also Figure S7.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Ameres S.L., Zamore P.D. Diversifying microRNA sequence and function. Nat. Rev. Mol. Cell Biol. 2013;14:475–488. - PubMed
    1. Ameres S.L., Horwich M.D., Hung J.H., Xu J., Ghildiyal M., Weng Z., Zamore P.D. Target RNA-directed trimming and tailing of small silencing RNAs. Science. 2010;328:1534–1539. - PMC - PubMed
    1. Babiarz J.E., Ruby J.G., Wang Y., Bartel D.P., Blelloch R. Mouse ES cells express endogenous shRNAs, siRNAs, and other Microprocessor-independent, Dicer-dependent small RNAs. Genes Dev. 2008;22:2773–2785. - PMC - PubMed
    1. Bartel D.P. MicroRNAs: target recognition and regulatory functions. Cell. 2009;136:215–233. - PMC - PubMed
    1. Benjamini Y., Hochberg Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B. 1995;57:289–300.

Publication types

MeSH terms

Associated data