Identification and functional characterization of tRNA-derived RNA fragments (tRFs) in respiratory syncytial virus infection

Abstract

The discovery of small noncoding RNAs (sncRNAs) with regulatory functions is a recent breakthrough in biology. Among sncRNAs, microRNA (miRNA), derived from host or virus, has emerged as elements with high importance in control of viral replication and host responses. However, the expression pattern and functional aspects of other types of sncRNAs, following viral infection, are unexplored. In order to define expression patterns of sncRNAs, as well as to discover novel regulatory sncRNAs in response to viral infection, we applied deep sequencing to cells infected with human respiratory syncytial virus (RSV), the most common cause of bronchiolitis and pneumonia in babies. RSV infection leads to abundant production of transfer RNA (tRNA)-derived RNA Fragments (tRFs) that are ~30 nucleotides (nts) long and correspond to the 5'-half of mature tRNAs. At least one tRF, which is derived from tRNA-Glu-CTC, represses target mRNA in the cytoplasm and promotes RSV replication. This demonstrates that this tRF is not a random by-product of tRNA degradation but a functional molecule. The biogenesis of this tRF is also specific, as it is mediated by the endonuclease angiogenin (ANG), not by other nucleases. In summary, our study presents novel information on the induction of a functional tRF by viral infection.

Figure 1

Pipeline of analyses of Illumina high-throughput sequencing data. Flowchart of the sequencing data analyses is depicted. miRNA, microRNA; ncRNA, noncoding RNA; rRNA, ribosomal RNA; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA.

Figure 2

Classification of sncRNAs. Human genome-derived sncRNAs were sorted according to their origins, and their relative abundance (calculated from read numbers) was depicted in pie charts (left two charts). tRNA-derived sncRNAs were divided into non-tRFs and tRFs (middle two charts). tRFs were further sorted into tRF-1, 3, and 5 series (bar graphs on the right). The definition of tRFs and the subtypes (-1, -3, and -5 series) are described in the text and depicted in Supplementary Figure S2. miRNA, microRNA; sncRNA, small noncoding RNA; rRNA, ribosomal RNA; RSV, respiratory syncytial virus; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA.

Figure 3

Experimental validation and characterization of tRF-5s. (a) Sequence alignment of tRF-5s with their parental mature tRNAs and northern probes. The length of each sequence is indicated in parentheses. CCA sequence that is post-transcriptionally added to the 3′-end of tRNA is indicated in brackets. (b) Total RNA from indicated treatments in A549 cells was loaded to a denaturing polyacrylamide gel for northern hybridization using probes indicated in a; 5S rRNA is shown for equal loading. The positions of tRF-5 and mature tRNA are indicated on the right; molecular size markers are indicated on the left. The blot was exposed for ≤8 hours. Data are representative of 2–3 independent experiments. (c) Primary cultured SAE cells were mock- or RSV-infected at MOI of 3 for various lengths of time as indicated. All other experimental conditions were similar to those described in b, except that 4 µg of total RNAs were used for northern hybridization. (d) A549 cells were transfected with purified viral RNAs (lane 2), or infected with naive or UV-inactivated RSV for 24 hours (lanes 4 and 6). Cells treated with Lipofectamine 2000 (lane 1) or mock-infected (lanes 3 and 5) were used as negative controls. Northern hybridization was done as described in b. (e) Total, nuclear, or cytoplasmic RNAs from mock- or RSV-infected A549 cells were subjected to northern hybridization. All other descriptions are the same as in b. Data are representative of 2–3 independent experiments. EtBr, ethidium bromide; hMPV, human metapneumovirus; MOI, multiplicity of infection; nt, nucleotide; rRNA, ribosomal RNA; RSV, respiratory syncytial virus; SAE, small alveolar epithelial cell; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA; UV, ultraviolet.

Figure 4

tRF5-GluCTC exhibits miRNA-/siRNA-like trans-silencing capacity. (a) Sequence alignment of Pp-anti_GluCTC_WT with three mutant luciferase sensor plasmids. Mutated nts are in bold and italic. The sequence of tRF5-GluCTC inhibitor and mimic are also present. (b) A549 cells in hexaplicate were co-transfected with a firefly (Pp) luciferase reporter plasmid (cognate “Pp-anti_GluCTC_WT ” plasmid or “Pp-vector ” plasmid, 0.1 µg/well of 24-well plate), a plasmid expressing renilla (Rr) luciferase, and 120 nmol/l antisense oligos (“anti-GluCTC ” or “anti-control”). After 2 hours post-transfection, cells were mock- or RSV-infected, then harvested at 6 hours post-infection to measure luciferase activities. In all experiments, Pp luciferase values were normalized to renilla (Rr) luciferase values. Values at y-axis (Pp/Rr) are a representative of three independent experiments and are expressed as mean ± SE. **On the second white bar (RSV-infected and Pp-anti_GluCTC transfected) denotes P value <0.01, relative to the second black bar (mock-infected and Pp-anti_GluCTC transfected). (c) A549 cells were treated as described in b and total RNA was subjected to northern hybridization as described in Figure 3b (left panel). Densitometric analysis of tRF5-GluCTC band intensity, quantified using VisionWorksLS image acquisition and analysis software from UVP (Upland, CA) is shown after normalization to 5S rRNA. Data are summarized from three independent experiments (right panel). **P < 0.01, relative to second white bar. (d) A549 cells in hexaplicate were transfected with indicated Pp luciferase reporter plasmids. At 24 hours post-transfection, cells were mock- or RSV-infected, then harvested at 15 hours post-infection to measure luciferase activities. All other descriptions are the same as in b. *P < 0.05, **P < 0.01, respectively, relative to second white bar. (e) A549 cells in hexaplicate were transfected with indicated tRF-mimic oligos. Luciferase plasmids, Pp-anti_GluCTC_WT or _Mut3 (0.1 µg/well), and a plasmid expressing Rr luciferase, were co-transfected. At 40 hours post-transfection, cells were lysed for luciferase assays. Pp values were first normalized by Rr values, and then the Pp/Rr values of “Pp-anti_GluCTC_WT ” were normalized to those of “Pp-anti_GluCTC_Mut3”, yielding relative Pp/Rr values (y-axis). **Denotes P value <0.01, relative to control-mimic (black bars) in respective concentration. All other descriptions are the same as in b. miRNA, microRNA; nts, nucleotides; ORF, open-reading frame; RSV, respiratory syncytial virus; siRNA, small-interfering RNA; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA; WT, wild-type.

Figure 5

tRF5-GluCTC promotes RSV replication. (a–d). Knockdown of tRF5-GluCTC inhibits RSV replication. Antisense oligos transfection and RSV infection were performed as described in Figure 4b. At 15 hours post-infection (as in a and c) or indicated times post-infection (as in b), viral infectious particles were measured by immunostaining (as shown in a). Viral gene transcription and protein synthesis were investigated by real-time PCR (as shown in b) and western blot (as shown in c), respectively. The total protein synthesis was also assayed for the groups as indicated (as shown in d). (e,f) A549 cells were transfected with 50 nmol/l of indicated mimic oligos. At 6 hours post-transfection, cells were mock- or RSV-infected. At 15 hours post-infection, viral titration and viral gene expression were assayed, respectively. Data are representative of 2–3 independent experiments. **P < 0.01, relative to the black bar. pfu, plaque-forming unit; p.i., post-infection; RSV, respiratory syncytial virus; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA.

Figure 6

Cleavage sites in tRNAs to produce tRF5-GluCTC. (a) Alignment of tRF5-GluCTC isoforms to their parental mature tRNA with read numbers listed in parentheses. According to the read numbers of major isoforms (cloned reads >5,000; listed in bold), cleavage sites are designated by downward arrows. The anticodon is highlighted in gray. (b) Major cleavage sites and the anticodon are designated in the tRNA cloverleaf structure. tRF, tRNA-derived RNA fragment; tRNA, transfer RNA.

Figure 7

tRF5-GluCTC is generated by angiogenin. (a) A549 cells were transfected with 120 nmol/l of siRNA against indicated proteins or control siRNA as a negative control. At 40 hours post-transfection, the cells were mock- or RSV-infected for 6 hours. Total RNAs were then subjected to northern hybridization as described in Figure 3b (top panel); 5S rRNA expression is shown for equal loading (bottom panel). (b) Densitometric analysis of band intensities of tRF5-GluCTC in a. Band intensity was quantified using VisionWorksLS image acquisition and analysis software from UVP. Band intensity after 5S rRNA normalization are presented as mean ± SE of arbitrary units. *P < 0.05, relative to the first white bar. (c) The suppression of target mRNAs by each siRNA was confirmed by qRT-PCR measurement. The values are the relative expression levels of each mRNA, with the value of control siRNA set as one. ANG, angiogenin; qRT-PCR, quantitative reverse transcription-PCR; RSV, respiratory syncytial virus; siRNA, small-interfering RNA; tRF, tRNA-derived RNA fragment; tRNA, transfer RNA.