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
. 2011 Feb;21(2):203-15.
doi: 10.1101/gr.116657.110. Epub 2010 Dec 22.

Deep annotation of Drosophila melanogaster microRNAs yields insights into their processing, modification, and emergence

Eugene Berezikov  1 , Nicolas Robine, Anastasia Samsonova, Jakub O Westholm, Ammar Naqvi, Jui-Hung Hung, Katsutomo Okamura, Qi Dai, Diane Bortolamiol-Becet, Raquel Martin, Yongjun Zhao, Phillip D Zamore, Gregory J Hannon, Marco A Marra, Zhiping Weng, Norbert Perrimon, Eric C Lai
Affiliations

Deep annotation of Drosophila melanogaster microRNAs yields insights into their processing, modification, and emergence

Eugene Berezikov et al. Genome Res. 2011 Feb.

Abstract

Since the initial annotation of miRNAs from cloned short RNAs by the Ambros, Tuschl, and Bartel groups in 2001, more than a hundred studies have sought to identify additional miRNAs in various species. We report here a meta-analysis of short RNA data from Drosophila melanogaster, aggregating published libraries with 76 data sets that we generated for the modENCODE project. In total, we began with more than 1 billion raw reads from 187 libraries comprising diverse developmental stages, specific tissue- and cell-types, mutant conditions, and/or Argonaute immunoprecipitations. We elucidated several features of known miRNA loci, including multiple phased byproducts of cropping and dicing, abundant alternative 5' termini of certain miRNAs, frequent 3' untemplated additions, and potential editing events. We also identified 49 novel genomic locations of miRNA production, and 61 additional candidate loci with limited evidence for miRNA biogenesis. Although these loci broaden the Drosophila miRNA catalog, this work supports the notion that a restricted set of cellular transcripts is competent to be specifically processed by the Drosha/Dicer-1 pathway. Unexpectedly, we detected miRNA production from coding and untranslated regions of mRNAs and found the phenomenon of miRNA production from the antisense strand of known loci to be common. Altogether, this study lays a comprehensive foundation for the study of miRNA diversity and evolution in a complex animal model.

PubMed Disclaimer

Figures

Figure 1.
Figure 1.
Distinct and overlapping patterns of miRNA expression in different tissues and samples. (A) Graph shows those miRNAs that contribute more than 1% of miRNAs in aggregated sets of ovary, head, and S2 cell data totaling about 30–70 M reads specifically mapped to miRNAs. It is clear that many miRNAs are either strongly enriched or seemingly absent from one of the three sample types. (B–D) Venn diagrams that show the overlap in miRNAs detected in ovary, head, and S2 cells at various levels of expression. As the contribution of each miRNA decreases from 1% (B) to >0.01% (C), we observe increasing coexpression among these distinct tissue/cell types. When considering miRNA expression down to a single read in each library, we observe nearly complete coexpression. The few miRNAs that were not detected (4*) are either questionable as canonical miRNAs (miR-280 and miR-289) or were detected at only a few parts per million in the esoteric cell line OSS (miR-2280 and miR-2281).
Figure 2.
Figure 2.
Examples of miRNA loci exhibiting five phased species. (A) mir-277; (B) mir-965. The most abundant product is the miRNA (green) followed by its partner miRNA* species (red). The 5′ and 3′ ends of these RNAs dovetail with the abundant loop reads (yellow), as well as 5′ miRNA overlap (moR) and 3′ moR reads (blue). The convention of highlighting mature species green and star species red is continued in all subsequent figures.
Figure 3.
Figure 3.
5′ variability of Drosophila canonical miRNAs. These charts summarize data for 135 canonical miRNAs that generated more than 1000 reads and had exclusively unique genomic mappings. (A,B) The 5′ end precision of mature miRNA species was generally high for well-expressed species; however, select loci generated abundant secondary and/or tertiary 5′ isomiRs. (C,D) The 5′ end precision of miRNA* (star) species was less than for mature miRNAs; still, only a relatively select group of highly expressed star species exhibited abundant 5′ isomiRs. The full analysis is available in Supplemental Table S4.
Figure 4.
Figure 4.
Exemplary loci illustrating precision and variability in miRNA processing. (A) Most miRNAs, such as mir-184, exhibit precisely defined 5′ ends of both miRNA and star species. Since the mature strand of mir-184 is highly biased over its star species, there is one dominant miRNA-type regulatory species produced from this locus. (B) mir-193 is a locus exhibiting balanced accumulation of small RNAs from its hairpin arms. In addition, both 5p and 3p arms exhibit abundant secondary and even tertiary 5′ isomiR species. All of these accumulate in AGO1; therefore, mir-193 produces at least five substantial miRNA-type regulatory RNAs. Note that the 3p RNAs also accumulate in AGO2 as evidenced by their enrichment in a library prepared from small RNAs resistant to oxidization.
Figure 5.
Figure 5.
Examples of antisense transcription and processing across miRNA operons. (A) The top genomic strand of the mir-275/mir-305 locus is abundantly converted into mature miRNAs, but the bottom genomic strand also exhibits confident evidence for miRNA production across both miRNA hairpins. Primary numbers indicate reads matching precisely to the highlighted species; numbers in parentheses sum all other isomiRs matching that hairpin arm. (B) The distal end of the mir-972-979 cluster on the X chromosome overlaps Grip84, transcribed on the other strand. We detected confident miRNA production from the antisense strands of mir-979 and mir-978. This locus also bears a perfect tandem hairpin (sblock212157/mir-4966) that is subject to alternate Drosha and Dicer-1 cleavage to produce multiple 5′ isomiRs on both hairpin arms; multiple species were also detected in AGO1-IP libraries. The entire hairpin is duplicated; thus, all reads could map to either location.
Figure 6.
Figure 6.
Examples of novel miRNAs annotated in this study. (A) sblock6825/mir-4984 and (B) sblock66958/mir-4982 are novel miRNA loci that generate specific miRNA/miRNA* duplex species and had at least some reads in AGO1-IP libraries. mir-4982 approaches the lower limit for read accumulation needed for confident annotation. (C) sblock87333 is an example of a “candidate” miRNA locus that was not assigned a miRNA gene name at present. It exhibits heterogeneous 5p arm species (pink), and its dominant 3p arm is 20 nt in length, which is not typical for known miRNAs. Nevertheless, the 3p species clearly exhibit a preferred 5′ end, and several versions of the 3p species extending to 22 nt were present in head AGO1-IP data (GSM488489); one of the 5p reads would potentially pair with this duplex in an appropriate fashion. Therefore, this locus may eventually prove to be a genuine miRNA locus.
Figure 7.
Figure 7.
Examples of novel miRNAs generated from mRNAs. (A) A miRNA from the sense strand of the Nrx-1 coding region. This locus generates a specific miRNA/miRNA* duplex and exhibits some reads from head AGO1-IP data. Inspection of 12 species alignments indicates that the hairpin sequence evolves readily by codon wobbles, at a rate similar to the flanking nonhairpin codons. (B) A miRNA from the antisense strand of the Aph-4 coding region. In addition to specific miRNA/miRNA* duplex reads, this locus also generated a phased 5′ moR. (C) miRNA production from a primary-mRNA transcript in which the hairpin is produced from the pairing of intronic and exonic sequence of CG5953.
Figure 8.
Figure 8.
Example of a transitional miRNA locus, which exhibits signatures of both RNA degradation as well as Drosha/Dicer-1 processing across its precursor. Each read length has been plotted in a distinct color to emphasize the heterogeneity of cloned species mapping to the 3′ UTR of CG15102. The reads have been ordered on the y-axis with the most abundant individual species at the bottom. It can clearly be seen that a specific set of 21–22 nt reads are specifically made. These map to a typical pri-miRNA hairpin with a lower stem and a miRNA/miRNA* duplex region.
Figure 9.
Figure 9.
Patterns of 3′ untemplated additions in Drosophila miRNAs. (Left) Scenarios for 3′ untemplated additions to the pre-miRNA versus the mature miRNA/miRNA* species. Preferred addition to the pre-miRNA hairpin is expected to be reflected in a bias for modifications of 5p species relative to 3p species. (Right) The overall frequency of 3′ additions observed on Drosophila miRNAs are U > A > C > G. For U and A additions, t-test reveals a statistically significant preference for 3p additions, consistent with a preference for pre-miRNA modifications. C additions were much less frequent but also appeared to exhibit some 3p preference. Judging 5p U or A addition frequencies relative to G additions as background suggested that mature miRNA/miRNA* species are also subject to uridylation and adenylation. The full analysis is presented in Supplemental Table S7.
Figure 10.
Figure 10.
RNA editing in Drosophila small RNAs. We collected S2 and head small RNA reads with one or two mismatches to 3′ cis-NATs or miRNAs and tabulated the nature of their nucleotide changes. (A) Endo-siRNAs from 3′ cis-NATs exhibit a preponderance of A→G changes indicative of adenosine deamination. (B) In contrast, miRNA reads do not collectively exhibit enrichment for A→G changes. (C) miR-100 is a highly conserved miRNA with abundant A→G transition reads present in multiple libraries. The full analysis is presented in Supplemental Table S8.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Aboobaker AA, Tomancak P, Patel N, Rubin GM, Lai EC 2005. Drosophila microRNAs exhibit diverse spatial expression patterns during embryonic development. Proc Natl Acad Sci 102: 18017–18022 - PMC - PubMed
    1. Ameres SL, Horwich MD, Hung JH, Xu J, Ghildiyal M, Weng Z, Zamore PD 2010. Target RNA-directed trimming and tailing of small silencing RNAs. Science 328: 1534–1539 - PMC - PubMed
    1. Bartel DP, Chen CZ 2004. Micromanagers of gene expression: The potentially widespread influence of metazoan microRNAs. Nat Genet 5: 396–400 - PubMed
    1. Bender W 2008. MicroRNAs in the Drosophila bithorax complex. Genes Dev 22: 14–19 - PMC - PubMed
    1. Berezikov E, Guryev V, van de Belt J, Wienholds E, Plasterk RH, Cuppen E 2005. Phylogenetic shadowing and computational identification of human microRNA genes. Cell 120: 21–24 - PubMed

Publication types

MeSH terms