Identification and characterization of stem-bulge RNAs in Drosophila melanogaster

Abstract

Non-coding Y RNAs and stem-bulge RNAs are homologous small RNAs in vertebrates and nematodes, respectively. They share a conserved function in the replication of chromosomal DNA in these two groups of organisms. However, functional homologues have not been found in insects, despite their common early evolutionary history. Here, we describe the identification and functional characterization of two sbRNAs in Drosophila melanogaster, termed Dm1 and Dm2. The genes coding for these two RNAs were identified by a computational search in the genome of D. melanogaster for conserved sequence motifs present in nematode sbRNAs. The predicted secondary structures of Dm1 and Dm2 partially resemble nematode sbRNAs and show stability in molecular dynamics simulations. Both RNAs are phylogenetically closer related to nematode sbRNAs than to vertebrate Y RNAs. Dm1, but not Dm2 sbRNA is abundantly expressed in D. melanogaster S2 cells and adult flies. Only Dm1, but not Dm2 sbRNA can functionally replace Y RNAs in a human cell-free DNA replication initiation system. Therefore, Dm1 is the first functional sbRNA described in insects, allowing future investigations into the physiological roles of sbRNAs in the genetically tractable model organism D. melanogaster.

Keywords: DNA replication; Y RNA; non-coding RNAs; sbRNA.

Figure 1.

Predicted secondary structures of two putative Drosophila melanogaster sbRNAs. An evolutionarily conserved trinucleotide domain present in stem-bulge and Y RNAs that is functionally essential for DNA replication activity is highlighted in red/blue, a variable region in yellow, and a second conserved sequence present in sbRNAs in orange.

Figure 2.

Parameters of computational molecular dynamics simulations. Root mean square deviation (A) and Radius of gyration (B) obtained from the MD simulation from different Y RNAs and sbRNAs. 3D representations of D. melanogaster Dm1 and Dm2 RNAs, Y3 RNA from C. griseus, H. sapiens and M. musculus and Y5 RNA from M. mulatta were used for this analysis (see Figure S5-S10, corresponding .pdb files).

Figure 3.

Parameters of computational molecular dynamics simulations. Root mean square fluctuation (rmsf) of C1ʹ from D. melanogaster Dm1 (A) and Dm2 (B) RNAs. Red boxes highlight the rmsf of the GUG triplet and blue boxes of the CAC triplet.

Figure 4.

Global Phylogenetic Tree of Y RNAs and sbRNAs. sbRNA family (Blue). Y5 RNA family (Red). Y4 RNA family (Purple). Y3 RNA family (Orange). Y1 RNA family (Green). Gene species are presented by: Dm, Drosophila melanogaster; Bm, Bombyx mori; Ce, Caenorhabditis elegans; h, Homo sapiens; Ch, Cricetulus griseus; m, Mus musculus; Rn, Rattus norvegicus, Cp, Cavia porcellus; Oc, Oryctolagus cuniculus; Pt, Pan troglodytes; Mm, Macaca mulatta; Cf, Canis familiaris; Bt, Bos taurus; Md, Monodelphis domestica; z, Danio rerio; La, Loxodonta africana; x, Xenopus laevis; Xt, Xenopus tropicalis. This analysis involved 62 nucleotide sequences. There were a total of 149 positions in the final dataset. The bootstrap consensus tree inferred from 10,000 replicates is taken to represent the evolutionary history of the taxa analyzed [32]. Phylogenetic analyses were conducted in MEGA X [25] and the tree was drawn in Figtree v1.4.2.

Figure 5.

Expression of Dm1 and Dm2 sbRNAs, relative to 5S rRNA. Total RNA was prepared from Drosophila melanogaster S2 cells and adult flies, and quantified by qRT-PCR. Percentages of relative expression are shown as mean values ± standard deviations of n = 4 independent experiments.

Figure 6.

DNA replication initiation function. (A) Synthesis of Dm1 and Dm2 sbRNAs. The indicated RNAs were synthesised by in vitro transcription and visualised after denaturing polyacrylamide gel electrophoresis. Molecular sizes of single-stranded DNA oligonucleotide markers are indicated. (B) RNA-dependent initiation of chromosomal DNA replication in a human cell-free system. Human late G1 phase template nuclei were incubated in the absence or presence of control or Y RNA-depleted human cytosolic extracts supplemented with 100 ng of purified Dm1 or Dm2 sbRNAs, as indicated. Percentages of replicating template nuclei are shown for each reaction as mean values ± standard deviations of n = 5 independent experiments. Brackets indicate results of t-tests (unpaired, two-tailed with unequal variance) of control against the experimental samples (*** p = 0.007; n.s., not significant).