Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 16 (4), 673-95

Human tRNA-derived Small RNAs in the Global Regulation of RNA Silencing

Affiliations

Human tRNA-derived Small RNAs in the Global Regulation of RNA Silencing

Dirk Haussecker et al. RNA.

Abstract

Competition between mammalian RNAi-related gene silencing pathways is well documented. It is therefore important to identify all classes of small RNAs to determine their relationship with RNAi and how they affect each other functionally. Here, we identify two types of 5'-phosphate, 3'-hydroxylated human tRNA-derived small RNAs (tsRNAs). tsRNAs differ from microRNAs in being essentially restricted to the cytoplasm and in associating with Argonaute proteins, but not MOV10. The first type belongs to a previously predicted Dicer-dependent class of small RNAs that we find can modestly down-regulate target genes in trans. The 5' end of type II tsRNA was generated by RNaseZ cleavage downstream from a tRNA gene, while the 3' end resulted from transcription termination by RNA polymerase III. Consistent with their preferential association with the nonslicing Argonautes 3 and 4, canonical gene silencing activity was not observed for type II tsRNAs. The addition, however, of an oligonucleotide that was sense to the reporter gene, but antisense to an overexpressed version of the type II tsRNA, triggered robust, >80% gene silencing. This correlated with the redirection of the thus reconstituted fully duplexed double-stranded RNA into Argonaute 2, whereas Argonautes 3 and 4 were skewed toward less structured small RNAs, particularly single-strand RNAs. We observed that the modulation of tsRNA levels had minor effects on the abundance of microRNAs, but more pronounced changes in the silencing activities of both microRNAs and siRNAs. These findings support that tsRNAs are involved in the global control of small RNA silencing through differential Argonaute association, suggesting that small RNA-mediated gene regulation may be even more finely regulated than previously realized.

Figures

FIGURE 1.
FIGURE 1.
Small RNA Northern blot screen reveals a population of tRNA-derived 21–22-nt small RNAs that are 5′-phosphorylated and 3′-hydroxylated. (A) Northern blot screen candidate sequences. T4 RNA ligase-sensitive small RNAs in bold, except for known microRNA miR-20/cand22, which is indicated by an asterisk (*); tRNA 3′ “CCA” motif in italics; number of sequence hits in parentheses (out of 8554). (B) Northern blot screen examples of ∼21–22-nt T4 RNA ligase-sensitive small RNAs (293 cell RNA). Ligase-sensitive small RNAs evidenced by either disappearance and/or band shift (arrows). Cand22: 5′-phosphorylated, 3′-hydroxylated miR-20 (positive control); Cand14, Cand23, Cand33, and Cand35: type I tsRNA examples; Cand45: type II tsRNA. Initial genomic annotation of the small RNAs shown below the blots; subsequent manual blast revealed Cand35 and Cand45 to be derived from predicted tRNAs (in parentheses); still no perfect match could be identified for Cand14, Cand23, and Cand33. TAP: tobacco acid pyrophosphatase; ligase: T4 RNA ligase; −: untreated; +: treated; M: Decade (Ambion) RNA size marker. (C) tsRNAs are 5′-phosphorylated and 3′-hydroxylated (Northern blot of diagnostic enzyme treatments). Two hundred ninety-three cell RNA was treated with the following enzymes (potential activities described in parentheses), and enzyme susceptibility of the 21–22-nt small RNAs of interest deduced by their shift in gel mobility and/or disappearance in the Northern blot: (1) buffer; (2) T4 polynucleotide kinase (PNK) +ATP (5′ phosphorylation of 5′-OH and 3′ dephosphorylation); (3) T4 PNK, then Terminator (degrades 5′ monophosphorylated, unstructured RNAs; PNK-dependent RNA removal would indicate 5′-OH RNAs); (4) Terminator; (5) T4 RNA ligase +ATP (for 5′P-3′OH RNAs: intramolecular circularization; trans-ligation of RNAs containing either of these modifications); (6) TAP (hydrolyzes phosphoric acid anhydride bonds in triphosphorylated and capped RNAs, leaving 5′ monophosphate), then T4 RNA ligase (TAP-dependent T4 RNA ligation would indicate 5′ cap or 5′ triphosphate); (7) TAP; (8) T4 RNA ligase, no ATP + activated 3′-adapter oligo (adapter ligation would indicate 3′OH); (9) T4 PNK, then T4 RNA ligase, no ATP + activated 3′-adapter oligo (PNK-dependent adapter ligation would indicate either 3′-P, or 2′–3′ cyclic phosphate); (10) 3′-phosphatase-negative T4 PNK, then T4 RNA ligase, no ATP + activated 3′-adapter oligo (would confirm that a reaction in treatment “9” was dependent on 3′ dephosphorylation by T4 PNK); (11) polyA polymerase (PAP; adds polyA to 3′-hydroxyl RNAs); (12) buffer (same as 1). Blots were stripped and rehybridized with the indicated probes. Arrows indicate 21–22-nt RNAs of interest; HERV: human endogenous retroviral element.
FIGURE 2.
FIGURE 2.
RNaseZ-dependent, Dicer-independent cand45 biogenesis. (A) Cand45 expression is unchanged in a HCT116-derived cell line that contains a mutation in the Dicer helicase domain. Most (e.g., miR-20 and miR-21), but not all (e.g., let-7a) microRNAs are down-regulated in this cell line. wt: Parental HCT Dicer wild-type cell line; Dcr mut: HCT-derived Dicer helicase mutant cell line; T4: test for T4 RNA ligase sensitivity (−: untreated; +: treated). (B) In vitro RNaseZ/P processing of cand45 tRNA. A radioactively labeled, in vitro transcribed cand45 precursor tRNA was treated with buffer alone (“mock”), recombinant human RNaseZ and/or purified human RNaseP. Arrows indicate that RNA was treated sequentially with stated conditions. Reaction products (schematic for predicted fragments shown on the right) were visualized on a polyacrylamide gel. M: Decade (Ambion) RNA size marker. (C) Model for tsRNA biogenesis: RNA polymerase III (Pol III) generates a precursor tRNA (1). The 5′ leader and 3′ trailers are removed by RNaseP (2) and Z (3), respectively. The mature tRNA is then exported into the cytoplasm (4). There, Dicer recognizes some, potentially misfolded tRNAs to produce Type I tsRNAs (5). The small RNA produced by nuclear RNaseZ cleavage and Pol III termination is a Type II tsRNA. Based on the near-exclusive cytoplasmic localization of type II tsRNAs, it is possible that a cytoplasmic pool of RNaseZ is responsible for the processing into type II tsRNAs of immature tRNAs have evaded nuclear quality control (data not shown).
FIGURE 3.
FIGURE 3.
tsRNAs localize to the cytoplasm (Northern blot analysis of nuclear-cytoplasmic RNA fractionation). Sno38b and U6 snRNA serve as nuclear markers. Equal amounts of nuclear and cytoplasmic RNA were loaded; blots were stripped and rehybridized. N: nuclear RNA fraction; C: cytoplasmic RNA fraction; arrow: T4 RNA ligase (T4)-sensitive RNA of interest (−: untreated; +: treated); M: Decade (Ambion) RNA size marker.
FIGURE 4.
FIGURE 4.
tsRNA Argonaute coimmunoprecipitations. (A) FLAG-Argonautes and FLAG-MOV10 were expressed at similar levels in 293 cells (Western blot); actin was used as a loading control. (B,C) Northern blot analysis of RNA coimmunoprecipitations with FLAG epitope-tagged Gfp (negative control), human Argonautes 1–4 (A1–A4), Mov10 (M10), and either HDAg (B), or TRBP (C). Blots were stripped and rehybridized. (B) tsRNAs associate with human Argonautes 1–4, but not MOV10. (C) Small RNAs that are generated from a cand45 tRNA expression system in which cand45 had been replaced with the (arbitrary) sequences “targ1” and “targ2” preferentially associate with FLAG-Argonaute 3 and 4. Input: RNA isolated from 10% lysate used per immunoprecipitation; IP: immunoprecipitated RNA; M: Decade (Ambion) RNA size marker.
FIGURE 5.
FIGURE 5.
Investigation of Cand45-like small RNAs. (A) Selected Cand45-like candidate sequences (in red) with predicted RNaseZ cleavage sites (“Z”). Shown are the sequences of the tRNA 3′ ends and the RNA polymerase III termination region. (B) Cand45-like small RNA candidate expression analysis (Northern blot). Cand45-like small RNAs can be detected as discrete 21–28-nt small RNAs and are modulated by Argonaute overexpression. Gfp (negative control), Dicer, Ago1–Ago4, Mov10: transfected expression plasmids. Terminator treatment (“TER”; −: untreated; +: treated) was used to determine the 5′-phosphorylation status of cand45-like small RNA candidates. In these blots, the amount of “+/−” Terminator-treated Argonaute 4-associated RNA loaded was half that of the other samples. M: Decade (Ambion) RNA size marker.
FIGURE 6.
FIGURE 6.
Trans-silencing capacity of tsRNAs (dual luciferase assay). (A) Cand14-mediated trans-silencing. Dual luciferase assay with the Renilla luciferase reporter gene carrying a fully cand14-complementary target site in its 3′ UTR (psi-cand14). Addition of a cand14 antisense molecule increased Renilla luciferase expression, as expected, if cand14 had RNAi-like trans-silencing capacity. Likewise, overexpression of Ago2 enhanced cand14-mediated trans-silencing. The specificity of the de-repression with anti-cand14 was confirmed with three (antisense) control oligonucleotides (anti-con1, anti-con2, anti-con3). (B) Cand45 overexpression from a plasmid into which the genomic sequence of cand45 had been cloned (cand45-45; by Northern blot). Cand45-empty: cand45 cloning plasmid with only a cloning site between the cand45 RNaseZ cleavage site and the RNA PolIII terminator; cand45-con: cand45-derived expression plasmid in which the cand45 sequence in cand45-45 was replaced with an arbitrary control sequence; M: Decade (Ambion) RNA size marker. (C) Cand45 overexpression (cand45-45)-dependent, anti-cand45 oligo-induced trans-silencing in HCT116 cells. Dual luciferase assay with reporter gene carrying a fully cand45-complementary target site in the Renilla luciferase 3′ UTR (psi-cand45). Cand45-empty, cand45-con, cand45-45 as in C. (D) Confirmation of the specificity of the anti-cand45 induced trans-silencing effect in panel D by the use of three additional (antisense) control oligonucleotides (anti-con1, anti-con2, anti-con3). (E) Anti-cand45 induced trans-silencing is RNAi-related. Cand45 overexpression (cand45-45)-dependent, anti-cand45 (“anti-45”)-induced trans-silencing is enhanced by overexpression of slicing-competent Argonaute 2, but mitigated by overexpression of the nonslicing Argonautes 1, 3, and 4. (F) Predicted cand45:anti-cand45 duplex; A.U.: arbitrary units; error bars indicate standard deviation from n = 3 transfections.
FIGURE 7.
FIGURE 7.
Sense-induced trans-silencing due to preferential loading by Argonaute 2, but not by Argonautes 3 and 4 of the reconstituted double-stranded RNA. (A) Cand45-Argonaute coimmunoprecipitation before and after addition of cand45-complementary sense oligonucleotides (Northern blot; HCT116 cells). FLAG-Gfp and FLAG-Agos used in IPs indicated for each row; “Ago2, deltaPAZ” is a PAZ-deletion mutant of Ago2. Predicted structures of overexpressed cand45 and (2′-O-methyl) complementary oligonucleotides indicated with red (cand45) and black lines (sense); the green line marks the 2-nt 5′ extension of a cand45-derivative (“cand45 + 2”). Arrows indicated cand45 of interest, a double asterisk (**) indicates the results of cross-hybridization of the probe with the transfected sense oligonucleotides; and asterisk (*) marks an input that was incorrectly loaded (correct input requantitated based on separate experiment). For each input/IP pair, the knockdown efficiency is indicated below (“kd,” 100 = no knockdown; summary shown in Fig. 2C) showing correlation between Ago2 IP and silencing efficiencies. Blots were stripped and rehybridized with microRNAs let-7a and miR-20; input: RNA isolated from 10% lysate used per immunoprecipitation. (B) Cand45-AgoIP efficiencies (phosphorimage quantitations of A). Down arrows indicate instances where microRNA-Ago2 associations are reduced under conditions in which cand45-Ago2 associations are increased. (C) Summary of sense-induced trans-silencing results corresponding to the cand45-Ago IP experiment shown in Figure 2A,B (dual luciferase assay with reporter “psi-cand45wt 2x”). Color scheme as in B.
FIGURE 8.
FIGURE 8.
Increased tsRNA abundance correlates with reduction and increase in microRNA and siRNA efficacies, respectively. (A) tsRNA abundance can be modulated by varying serum concentrations or overexpressing the tRNA transcription factor Brf1 (Northern blot; corresponding U6snRNA-normalized phosphorimage quantitations shown below). Blots were stripped and rehybridized. Total RNA from 293 cells was harvested on day 4 after Brf1 transfections on days 0 and 2. For the serum experiments, HCT116 cells cultured for 5 d under 1.5% or 10% serum were chosen. (B, i) MicroRNA silencing capacity is reduced in the presence of increased tsRNA abundance (dual luciferase assay). The target sites of the Renilla luciferase reporters are indicated with “c.14” (cand14) on the X-axis; PM/MM: perfect match/translational reporters. Results are normalized to the Renilla/Firefly ratios of reporter plasmid with no predicted small RNA target site (“bantam”), with 1.5% serum and control pcDNA3∷empty set at 100 for each reporter. (B, ii) Silencing efficiency of translational psi-let-7aMM reporter as indicated by let-7a inhibition (dual luciferase assay). anti-Dharm miR-20: control microRNA hairpin inhibitor; anti-Dharm let7a: let7a hairpin inhibitor; let-7a antisense inhibitor. “anti-Dharm let-7a” where most apparent let-7a inhibition was observed was set = 100. (C) SiRNA silencing is improved in the presence of increased tsRNA abundance (real-time qRT-PCR). Three different siRNAs (si-1–si-3) targeting endogenously expressed RALY RNA were transfected at two concentrations, 500 pM and 50 nM and remaining RALY RNA levels normalized to actin measured.

Similar articles

See all similar articles

Cited by 236 PubMed Central articles

See all "Cited by" articles

Publication types

LinkOut - more resources

Feedback