Logo of emborepLink to Publisher's site
EMBO Rep. Feb 2011; 12(2): 172–177.
Published online Jan 7, 2011. doi:  10.1038/embor.2010.191
PMCID: PMC3049427
Scientific Reports

MicroRNA evolution by arm switching

Abstract

MicroRNAs (miRNAs) modulate transcript stability and translation. Functional mature miRNAs are processed from one or both arms of the hairpin precursor. The miR-100/10 family has undergone three independent evolutionary events that have switched the arm from which the functional miRNA is processed. The dominant miR-10 sequences in the insects Drosophila melanogaster and Tribolium castaneum are processed from opposite arms. However, the duplex produced by Dicer cleavage has an identical sequence in fly and beetle. Expression of the Tribolium miR-10 sequence in Drosophila S2 cells recapitulates the native beetle pattern. Thus, arm usage is encoded in the primary miRNA sequence, but outside the mature miRNA duplex. We show that the predicted messenger RNA targets and inferred function of sequences from opposite arms differ significantly. Arm switching is likely to be general, and provides a fundamental mechanism to evolve the function of a miRNA locus and target gene network.

Keywords: microRNA, evolution, Hox, Drosophila, Tribolium

Introduction

A key step in microRNA (miRNA) processing is the cleavage of the precursor stem–loop by the Dicer enzyme to produce an approximately 22-nucleotide RNA duplex (Lee et al, 2003). Data from cloning and expression studies have shown that for many miRNAs, one arm of the RNA duplex preferentially accumulates. This mature miRNA is often assumed to be the dominant functional product that is incorporated into the RNA-induced silencing complex to direct translational repression or transcriptional degradation of messenger RNA (mRNA) targets (Hutvagner, 2005). High-throughput sequencing can be used to quantify the production of miRNAs generated from each arm of the duplex in different species, tissues and times in development. These data show that some miRNA precursors are processed to produce a single dominant mature miRNA, with a small proportion of reads originating from the opposite arm—often called the miR* sequence. Previous models have suggested that the choice of dominant miRNA arm is determined by thermodynamic and structural properties of the processed duplex (Khvorova et al, 2003; Schwarz et al, 2003). Other miRNA precursors are processed to produce significant quantities of mature miRNAs from both arms. It has also been shown that miR* sequences might function to downregulate targeted mRNAs (Okamura et al, 2008). Recent studies have demonstrated that sequences from opposite arms of a precursor miRNA can be sorted into different Argonaute (Ago) complexes, in which the dominant arm directs translational repression (by means of Ago1) and the miR* sequence directs transcriptional degradation (by means of Ago2; Czech et al, 2009; Ghildiyal et al, 2009; Okamura et al, 2009).

Data from several large-scale miRNA sequencing studies have highlighted the possibility that the arm from which the dominant mature miRNA is processed can switch in different tissues and at different developmental times (Ro et al, 2007; Ruby et al, 2007; de Wit et al, 2009; Chiang et al, 2010). Although these studies have suggested that a change in arm usage provides a mechanism to evolve the function of a miRNA gene locus, the nature of the molecular events that lead to such changes has been largely unexamined. The potential functional consequences for evolutionary changes in the processing of the miRNA duplex are profound, as mature miRNAs from opposite arms might target different mRNAs. Here, we combine comparative genomics and deep sequencing to reconstruct the evolutionary history of changes in arm usage in the ancient miR-100/10 miRNA family. We use a cell culture-based assay to quantify miRNA abundance and determine that the sequences that direct this choice must be outside the miRNA duplex. Furthermore, we use computational prediction of miRNA targets in this model system to gain functional insights into the process of arm choice and evolution.

Results And Discussion

Ancient duplications of miR-100 led to four subfamilies

Although the animal genomes studied so far contain hundreds of miRNA genes, deep sequencing and phylogenetic reconstruction suggest that miR-100 is the only miRNA shared by all bilateria; conservation in the cnidarian Nematostella vectensis suggests that it appeared near the origin of multicellularity, around 650 million years ago (Grimson et al, 2008; Wheeler et al, 2009). Early duplication of miR-100 gave rise to at least three other ancient and conserved miRNA subfamilies in animals: miR-125, miR-10 and miR-993, collectively called the miR-100/10 family in this study. We used covariance model approaches, alignment inspection and synteny to exhaustively identify miR-100/10 family homologues in representative sequenced genomes (supplementary Fig S1 online). miR-100 and miR-125 are clustered in most bilaterian genomes with another ancient miRNA, let-7 (Sempere et al, 2003). miR-10 is located in a conserved position within the Hox complex of animals. The Hox complex is an ancient and highly conserved cluster of related transcription factors that specify segment identity in almost all animals, and originated before the divergence of protostomes and deuterostomes. The conserved genomic organization of the four miR-100/10 subfamilies allowed us to date the three duplication events that gave rise to these miRNAs to before the expansion of the Eumetazoa. The most parsimonious interpretation of patterns of conservation and genomic organization suggests that miR-10 and miR-125 derive from independent duplications of the ancestral miR-100 before diversification of protostomes and deuterostomes, whereas miR-993 arose from a later duplication of miR-10, soon after the divergence of protostomes (Fig 1).

Figure 1
The evolution of the miR-100/10 family. miR-10 family members are located in two clusters. Boxes represent the hairpin precursor sequences, with the direction of transcription shown by arrowheads and the dominant mature sequence filled. The left-hand ...

Additional independent duplication events have expanded the miR-100/10 family. For example, multiple duplications of miR-10 have occurred in the lancelet Branchiostoma floridae and additional duplications of miR-10 and miR-993 can be found in the sequenced lophotrochozoan genomes (for example, Capitella teleta). Interestingly, these duplicates are located within Hox complexes. Whole-genome duplication events on the vertebrate lineage have also resulted in additional miR-100, miR-125 and miR-10 family members. Some miR-10 paralogues have been subsequently lost from duplicated Hox complexes, presumably through pseudogenization, facilitated by functional redundancy. In a few cases, the Hox complex structure has disintegrated, for example, in tunicates (Ciona intestinalis), nematodes (Caenorhabditis elegans) and planarians (Schmidtea mediterranea). In these genomes, the association of miR-100/10 family members with Hox genes is also lost.

Arm choice has evolved in the miR-100/10 family

We sequenced a library of small RNAs from Tribolium using the ABI SOLiD platform (Marco et al, 2010). We further validated the relative expression of mature miRNAs of the miR-100/10 family in Tribolium by using quantitative PCR (supplementary Fig S3 online), and combined these data with those from next-generation sequencing experiments (Ruby et al, 2007; Stark et al, 2007; Wheeler et al, 2009; Marco et al, 2010) and high-throughput cloning (Landgraf et al, 2007) in other species. The data show that most mature miRNAs from the miR-100/10 family originate from the 5′ arm of the precursor hairpin (Fig 2A). However, the dominant sequence from the Drosophila miR-10 locus originates from the 3′ arm (Schwarz et al, 2003; Ruby et al, 2007; Stark et al, 2007), as do all previously validated miR-993 sequences in D. melanogaster (Ruby et al, 2007), Locusta migratoria (Wei et al, 2009), Procambarus clarkii (Wheeler et al, 2009) and the Ciona miR-100 orthologue miR-1473 (Shi et al, 2009; Fig 1). This indicates that the miR-100/10 family has undergone a minimum of three independent arm switching events (Fig 1). Evidence from the evolution of the miR-100/10 family shows that arm usage is more labile than thought previously and that the function of some miRNAs is likely to have evolved through changes that control the production of the dominant arm.

Figure 2
Arm-specific miRNA processing in miR-100/10 family members. (A) Relative expression of mature miRNAs from 5′ and 3′ arms. Schematic of a precursor hairpin (centre). The relative abundance of the 5′ and 3′ arms of the four ...

Previous studies have suggested that arm choice is governed by the asymmetrical stability of the miR/miR* duplex resulting from Dicer cleavage (Schwarz et al, 2003; Hutvagner, 2005). This model implies a passive mechanism driven primarily by thermodynamic properties of the processed duplex. We have found that although the D. melanogaster and T. castaneum miR-10 sequences show opposite dominant arm usage, the miR/miR* duplexes are identical in sequence and, as such, have identical thermodynamic properties. This observation suggests that the thermodynamic model alone is unlikely to control the choice of arm. To examine the influence of the cellular environment on RNA stability and arm choice, we quantified mature miRNA levels in Drosophila S2 cells. We constructed and expressed plasmids to drive the expression of 300-nucleotide transcripts derived from the miRNA primary transcript centred around the miR-10 precursor from both Tribolium and Drosophila. Relative steady-state production of mature miRNA sequences from 5′ and 3′ arms of both sequences in Drosophila S2 cells was assayed by using TaqMan miRNA quantitative PCR assays (Fig 2B). These experiments show that the dominant D. melanogaster miR-10 sequence is processed from the 3′ arm, in accordance with the endogenous profile observed in high-throughput sequencing experiments. The profile of arm use from the T. castaneum miR-10 sequence in Drosophila S2 cells shows a near 2:1 dominance of the 5′ arm, similar to the pattern observed in Tribolium RNA deep-sequencing experiments. We further expressed constructs containing Tribolium and Drosophila miR-100 precursors in Drosophila S2 cells, and the endogenous profile of arm use was recapitulated in both cases. The observation that identical miRNA duplexes can have opposite arm usage profiles does not support a purely thermodynamic model of arm choice (Khvorova et al, 2003; Schwarz et al, 2003). These data support a model in which the sequence signals determining this choice are located outside this duplex (in the loop and/or flanking regions), and that control of arm use is regulated before Dicer processing (Fig 3).

Figure 3
A model of the differential processing of the miR-10 transcript. The precursor miRNA hairpin is excised from the primary miRNA transcript by Drosha. Dicer cleavage yields the mature miRNA duplex, and the dominant arm targets the RNA-induced silencing ...

We used evolutionary and expression analysis to identify examples of arm-switching events between both paralogous and orthologous members of the miR-100/10 family. Data also demonstrate that arm switches can occur in different tissues and across a developmental time course. Indeed, although the 5′ arms of miR-100 and miR-125 are dominant in mammals, some tissues have a profile with the 3′ mature sequence favoured (Landgraf et al, 2007). Many chicken miRNA families show switches in dominant arm use in a developmental time course (Glazov et al, 2008). Different mouse tissues have also been shown to have switches in dominant miRNA arm usage (Ro et al, 2007; Chiang et al, 2010). The miR-10 sequence does not seem to have significant tissue- or stage-specific arm switching in Drosophila or Tribolium (supplementary Fig S4 and Table S1 online). Nonetheless, these observations support our model that arm choice is specified outside the mature duplex.

Arm switching impacts mRNA target regulation

The sequences of mature miRNA products from 5′ and 3′ arms are different. Opposite arms might therefore regulate the expression of either distinct sets of mRNA transcripts or common targets through different sites in the target mRNA. We used two methods (seed-based and non-seed-based) to predict targets of the 5′ and 3′ arms of miR-10, miR-993, miR-100 and miR-125 in Drosophila untranslated regions (UTRs; Table 1). It is interesting to note that, in each case, the number of target sites predicted for the dominant miRNA arm was fewer than those for the other arm. This suggests a selective pressure against the occurrence of miRNA target sites. We found that target transcripts for each D. melanogaster 5′/3′ pair do not overlap more than expected by chance. The functional consequences of arm switching are therefore profound: homologous miRNA loci can generate distinct functions by regulating arm usage.

Table 1
Predicted targets of mature miRNAs derived from the 5′ and 3′ arms of miR-100/10 family members

Moreover, miRNA arm switching is probably frequent. We observed two types of arm switching in the miR-100/10 family alone. First, the early duplication of miR-10 giving rise to miR-993 is associated with a switch of dominant arm in the resulting miR-993 family. In the general case, duplication of an ancestral precursor that generates mature sequences from both arms—with each arm having unique targets—might lead to partitioning of targets between paralogues by regulated arm choice (sub-functionalization). Alternatively, duplication of a miRNA with a single dominant product, followed by arm switching, generates a new function (neo-functionalization; Ruby et al, 2007). Indeed, de Wit et al (2009) recently demonstrated an arm-switching event in paralogous copies of the Caenorhabditis-specific miR-246 family. The second type of change in arm usage is seen within the miR-10 family, in which the dominant arm has switched in Drosophila. This event provides a mechanism to alter the targets and function of orthologous sequences in related organisms. In the general case, the change in miRNA arm use might be a near-complete switch—as in miR-10—or a more subtle tuning of steady-state levels of miRNAs. Comparison between high-throughput sequencing data suggests that over 10% (5/46) of orthologous miRNAs in D. melanogaster and T. castaneum have shown a pronounced switch in the arm from which the dominant mature sequence is produced (Marco et al, 2010). Further, more than 40% of one-to-one orthologous miRNAs show a tuning in arm use in which there is a difference of more than tenfold (Marco et al, 2010). Understanding the mechanisms that control the switching or tuning events is vital for correct experimental interpretation in model systems; for example, Drosophila miR-10 will be an excellent model for understanding the functional consequences of arm switching, but is not likely to be a good model for the conserved functions of miR-10 in other animals.

The concept of a single functional miRNA processed from a precursor is clearly an oversimplification. Functional sequences from both arms might be produced, with differential regulation of orthologous sequences in different organisms, and even of the same sequence in different tissues and stages. The sequences from opposite arms have distinct targets, possibly mediated by alternative Ago complexes (Czech et al, 2009; Ghildiyal et al, 2009; Okamura et al, 2009). The ability to regulate the expression of sequences from different arms will have a profound effect on the targets and functions of a subset of miRNAs.

Methods

miR-100/10 family homologue detection and alignment. Members of the miR-100, miR-10, miR-993 and miR-125 subfamilies and the let-7 family were retrieved from the miRBase database (Griffiths-Jones et al, 2008) and aligned with ClustalW (Larkin et al, 2007). A consensus secondary structure was predicted using RNAalifold from the ViennaRNA package (Bernhart et al, 2008). A covariance model was built from the alignment and searched against fully sequenced bilaterian genomes using INFERNAL 1.0 (Nawrocki et al, 2009). The genomes and assemblies searched were N. vectensis ( JGI 1.0), D. melanogaster (BDGP 5), T. castaneum (Baylor 2.0), C. teleta ( JGI 1.0), B. floridae (JGI 2.0), Homo sapiens (NCBI36), C. intestinalis ( JGI 2.0) and C. elegans (Wormbase WS200). Newly identified homologues were realigned to the model using cmalign from INFERNAL 1.0. The alignment, structure and model were refined by an iterative procedure of covariance model searches, realignment, secondary structure prediction and manual editing (RALEE; Griffiths-Jones, 2005). Subfamilies were classified by analysis of syntenic regions: genomic regions containing the Hox complex were manually annotated using BLASTX and BLASTP to identify orthologous Hox genes by position, sequence conservation of homeodomains and diagnostic amino-acid motifs (for example, the hexapeptide and UBDA motif). miR-10 and miR-993 homologues were classified by position in the Hox complex. miR-100 and miR-125 homologues were classified by proximity to members of the let-7 miRNA family.

Short RNA read mapping. Wild-type Tribolium were reared at 28°C in wholewheat flour supplemented with 1% yeast extract. Small RNA was extracted from mixed-sex adults using the miRVana miRNA isolation kit (Ambion). RNA was size selected (for less than 40 nucleotide) with a flashPAGE fractionator (Ambion). Purification was then performed with a flashPAGE reaction clean-up kit (Ambion). A library was constructed using the SOLiD Small RNA Expression Kit, according to the manufacturer's instructions (Ambion). To meet the sample quantity for SOLiD sequencing, the complementary DNA libraries were then further amplified using 15 cycles of PCR. The final products ranging from approximately 105 to 150 bp were purified and size selected. SOLiD sequencing was carried out at the Centre for Genomic Research at the University of Liverpool. Adaptor sequences were removed from the 3′ ends of reads, and reads were mapped to the Baylor release 3.0 of the Tribolium genome using Bowtie 0.11, with no mismatches allowed (Langmead et al, 2009). Reads mapping to more than one genomic position were discarded.

Prediction of target genes of miR/miR* pairs. We scanned the collection of 3′ UTRs in D. melanogaster (available at http://flybase.org/genomes/Drosophila_melanogaster/current/fasta) for canonical miRNA target sites (Bartel, 2009) and potential targets on the basis of RNA–RNA duplex stability, using mature miRNA sequences derived from the 5′ and 3′ arms of all four miR-100/10 family members (Enright et al, 2003). Where no miR* sequence is reported in the miRBase database, the miR* sequence was deduced from the hairpin structure. We then calculated the probability that 5′ and 3′ mature sequences (the miR/miR* duplex) share more common gene targets than expected by chance. This probability is given by the cumulative hypergeometric distribution (Sokal & Rohlf, 1995), and is equivalent to Fisher's exact test. As these probability values are highly dependent on the length and nucleotide composition of the 3′ UTRs, we estimated unbiased P-values by simulating, for each miR/miR* pair, sets of 1,000 synthetic miR/miR* pairs to generate a null distribution of probability values. Two independent sets of null distributions were generated. In the first, a synthetic mature miRNA was produced by shuffling the sequence of the known mature miRNA. Thereafter, a synthetic miR* was generated by taking the reverse complement of the synthetic miR sequence, adding a random two-nucleotide 3′ overhang sequence and introducing bulges in the duplex by random mutation of the synthetic miR* sequence, with a probability of 0.2 per nucleotide. These parameters were based on empirical observations of animal miRNAs from the miRBase database. In the second set, synthetic miR/miR* pairs were randomly generated.

Construction of miR-10- and miR-100-expression constructs and transfection of S2 cells. Expression constructs containing 250 bp flanking the miR-10 and miR-100 precursor sequences from D. melanogaster and T. castaneum were cloned into the pAc5.1 vector (Invitrogen) using the following primers: dme-miR-10-FP-KpnI 5′-TATATTTAGGGGTACCGCGATTGCCTAGCGGACTTCATTT-3′; dme-miR-10-RP-EcoRI 5′-TTAAGGGAATTCCACTTTTCCGCTTGCCATCAGCAACACTT-3′; tca-miR-10-FP-KpnI 5′-ATAAGGGGTACCTTTGTTGCAGGTTTTTCTCAAGA-3′; tca-miR-10-RP-EcoRI 5′-GGGGAATTCCGATTTATGTCTAAGGACAATCT-3′; dme-100-pac-FP-KpnI 5′-TTTGGGTACCCTTCGATTCATCGAGACACATTGAAGTTCACGA-3′; dme-100-pac-RP-EcoRI 5′-CAGGGAATTCCTTGACTCGACTTTTTATCCTTACTCCGCCATT-3′; tca-100-pac-FP-EcoRI 5′-CAGGGAATTCCCATGCTAAAAATATATATCTCCTGGACTGA-3′; tca-100-pac-RP-NotI 5′-AATTGCGGCCGCCTCAAACAGACCCTGCCACAAATCTAGTACTT-3′; sequence-verified constructs were transfected into S2 cells using 1.5 μg pAc5.1 plasmids and Effectene (Qiagen) according to the manufacturer's instructions. Small RNAs were extracted from both transfected and control S2 cells 48 h post-transfection, using the miRVana miRNA isolation kit (Applied Biosystems). Quantification of respective miRNA arm use (5′ and 3′) was carried out with TaqMan miRNA real-time PCR assays (Applied Biosystems). Samples from three independent transfections were assayed in triplicate. The U6 small RNA TaqMan assay (Ambion) was used for normalization.

Supplementary information is available at EMBO reports online (http://www.emboreports.org).

Supplementary Material

Supplementary Information:

Acknowledgments

We thank Kevin Peterson for sharing data before publication, the University of Liverpool Centre for Genomic Research for SOLiD sequencing, Ana Kozomara for helping with miRNA expression data sets and Casey Bergman for helpful comments on the manuscript. This work was funded by the University of Manchester (fellowships to S.G.-J. and M.R.), the Wellcome Trust VIP scheme and the Biotechnology and Biological Sciences Research Council (BB/G011346/1).

Footnotes

The authors declare that they have no conflict of interest.

References

  • Bartel DP (2009) MicroRNAs: target recognition and regulatory functions. Cell 136: 215–233 [PMC free article] [PubMed]
  • Bernhart SH, Hofacker IL, Will S, Gruber AR, Stadler PF (2008) RNAalifold: improved consensus structure prediction for RNA alignments. BMC Bioinformatics 9: 474. [PMC free article] [PubMed]
  • Chiang HR et al. (2010) Mammalian microRNAs: experimental evaluation of novel and previously annotated genes. Genes Dev 24: 992–1009 [PMC free article] [PubMed]
  • Czech B, Zhou R, Erlich Y, Brennecke J, Binari R, Villalta C, Gordon A, Perrimon N, Hannon GJ (2009) Hierarchical rules for Argonaute loading in Drosophila. Mol Cell 36: 445–456 [PMC free article] [PubMed]
  • de Wit E, Linsen S, Cuppen E, Berezikov E (2009) Repertoire and evolution of miRNA genes in four divergent nematode species. Genome Res 19: 2064–2074 [PMC free article] [PubMed]
  • Enright AJ, John B, Gaul U, Tuschl T, Sander C, Marks DS (2003) MicroRNA targets in Drosophila. Genome Biol 5: R1. [PMC free article] [PubMed]
  • Ghildiyal M, Xu J, Seitz H, Weng Z, Zamore PD (2009) Sorting of Drosophila small silencing RNAs partitions microRNA* strands into the RNA interference pathway. RNA 16: 43–56 [PMC free article] [PubMed]
  • Glazov EA, Cottee PA, Barris WC, Moore RJ, Dalrymple BP, Tizard ML (2008) A microRNA catalog of the developing chicken embryo identified by a deep sequencing approach. Genome Res 18: 957–964 [PMC free article] [PubMed]
  • Griffiths-Jones S (2005) RALEE–RNA ALignment editor in Emacs. Bioinformatics 21: 257–259 [PubMed]
  • Griffiths-Jones S, Saini HK, van Dongen S, Enright AJ (2008) miRBase: tools for microRNA genomics. Nucleic Acids Res 36: D154–D158 [PMC free article] [PubMed]
  • Grimson A, Srivastava M, Fahey B, Woodcroft BJ, Chiang HR, King N, Degnan BM, Rokhsar DS, Bartel DP (2008) Early origins and evolution of microRNAs and Piwi-interacting RNAs in animals. Nature 455: 1193–1197 [PMC free article] [PubMed]
  • Hutvagner G (2005) Small RNA asymmetry in RNAi: function in RISC assembly and gene regulation. FEBS Lett 579: 5850–5857 [PubMed]
  • Khvorova A, Reynolds A, Jayasena SD (2003) Functional siRNAs and miRNAs exhibit strand bias. Cell 115: 209–216 [PubMed]
  • Landgraf P et al. (2007) A mammalian microRNA expression atlas based on small RNA library sequencing. Cell 129: 1401–1414 [PMC free article] [PubMed]
  • Langmead B, Trapnell C, Pop M, Salzberg SL (2009) Ultrafast and memory-efficient alignment of short DNA sequences to the human genome. Genome Biol 10: R25. [PMC free article] [PubMed]
  • Larkin MA et al. (2007) Clustal W and Clustal X version 2.0. Bioinformatics 23: 2947–2948 [PubMed]
  • Lee Y et al. (2003) The nuclear RNase III Drosha initiates microRNA processing. Nature 425: 415–419 [PubMed]
  • Marco A, Hui JH, Ronshaugen M, Griffiths-Jones S (2010) Functional shifts in insect microRNA evolution. Genome Biol Evol 2: 686–696 [PMC free article] [PubMed]
  • Nawrocki EP, Kolbe DL, Eddy SR (2009) Infernal 1.0: inference of RNA alignments. Bioinformatics 25: 1335–1337 [PMC free article] [PubMed]
  • Okamura K, Phillips MD, Tyler DM, Duan H, Chou YT, Lai EC (2008) The regulatory activity of microRNA* species has substantial influence on microRNA and 3′ UTR evolution. Nat Struct Mol Biol 15: 354–363 [PMC free article] [PubMed]
  • Okamura K, Liu N, Lai EC (2009) Distinct mechanisms for microRNA strand selection by Drosophila Argonautes. Mol Cell 36: 431–444 [PMC free article] [PubMed]
  • Ro S, Park C, Young D, Sanders KM, Yan W (2007) Tissue-dependent paired expression of miRNAs. Nucleic Acids Res 35: 5944–5953 [PMC free article] [PubMed]
  • Ruby JG, Stark A, Johnston WK, Kellis M, Bartel DP, Lai EC (2007) Evolution, biogenesis, expression, and target predictions of a substantially expanded set of Drosophila microRNAs. Genome Res 17: 1850–1864 [PMC free article] [PubMed]
  • Schwarz DS, Hutvágner G, Du T, Xu Z, Aronin N, Zamore PD (2003) Asymmetry in the assembly of the RNAi enzyme complex. Cell 115: 199–208 [PubMed]
  • Sempere LF, Sokol NS, Dubrovsky EB, Berger EM, Ambros V (2003) Temporal regulation of microRNA expression in Drosophila melanogaster mediated by hormonal signals and broad-complex gene activity. Dev Biol 259: 9–18 [PubMed]
  • Shi W, Hendrix D, Levine M, Haley B (2009) A distinct class of small RNAs arises from pre-miRNA-proximal regions in a simple chordate. Nat Struct Mol Biol 16: 183–189 [PMC free article] [PubMed]
  • Sokal RR, Rohlf FJ (1995) Biometry: The Principles and Practice of Statistics in Biological Research, 3rd edn. New York: W. H. Freeman & Co
  • Stark A et al. (2007) Discovery of functional elements in 12 Drosophila genomes using evolutionary signatures. Nature 450: 219–232 [PMC free article] [PubMed]
  • Wei Y, Chen S, Yang P, Ma Z, Kang L (2009) Characterization and comparative profiling of the small RNA transcriptomes in two phases of locust. Genome Biol 10: R6. [PMC free article] [PubMed]
  • Wheeler BM, Heimberg AM, Moy VN, Sperling EA, Holstein TW, Heber S, Peterson KJ (2009) The deep evolution of metazoan microRNAs. Evol Dev 11: 50–68 [PubMed]

Articles from EMBO Reports are provided here courtesy of The European Molecular Biology Organization
PubReader format: click here to try

Formats:

Related citations in PubMed

See reviews...See all...

Cited by other articles in PMC

See all...

Links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...