microRNA-mediated messenger RNA deadenylation contributes to translational repression in mammalian cells

Abstract

Animal microRNAs (miRNAs) typically regulate gene expression by binding to partially complementary target sites in the 3' untranslated region (UTR) of messenger RNA (mRNA) reducing its translation and stability. They also commonly induce shortening of the mRNA 3' poly(A) tail, which contributes to their mRNA decay promoting function. The relationship between miRNA-mediated deadenylation and translational repression has been less clear. Using transfection of reporter constructs carrying three imperfectly matching let-7 target sites in the 3' UTR into mammalian cells we observe rapid target mRNA deadenylation that precedes measureable translational repression by endogenous let-7 miRNA. Depleting cells of the argonaute co-factors RCK or TNRC6A can impair let-7-mediated repression despite ongoing mRNA deadenylation, indicating that deadenylation alone is not sufficient to effect full repression. Nevertheless, the magnitude of translational repression by let-7 is diminished when the target reporter lacks a poly(A) tail. Employing an antisense strategy to block deadenylation of target mRNA with poly(A) tail also partially impairs translational repression. On the one hand, these experiments confirm that tail removal by deadenylation is not strictly required for translational repression. On the other hand they show directly that deadenylation can augment miRNA-mediated translational repression in mammalian cells beyond stimulating mRNA decay. Taken together with published work, these results suggest a dual role of deadenylation in miRNA function: it contributes to translational repression as well as mRNA decay and is thus critically involved in establishing the quantitatively appropriate physiological response to miRNAs.

Figure 1. mRNAs targeted by let-7 exhibit shortened poly(A) tails.

(A) Schematic of R-luc constructs carrying three let-7 target sites in their 3′ UTR (functional: R-luc-3xbulge; seed region mutated: R-luc-3xbulgemut [18]). (B) Let-7-dependent repression of R-luc protein and mRNA levels. HeLa cells were cotransfected with plasmids pR-luc-3xb or pR-luc-mut and pGL3 (F-luc) as control, and with an LNA anti-miR to let-7, miR-499 (non-specific; ns), or mock transfected (none) as indicated. Protein and RNA was extracted from cells at 6 and 24 hours after transfection. R-luc expression was normalised to F-luc expression for both protein and mRNA, measured by luciferase assay and qPCR, respectively. Repression by endogenous let-7 was calculated as the ratio of R-luc-mut expression to that of the corresponding R-luc-3xb. The bars represent averages of 2–3 measurements with standard error (24 hour protein) or range (all others). (C) Schematic of the LM-PAT assay and performance test. An LM-PAT assay for GAPDH mRNA was performed with HeLa cell total RNA (left panel). To mark the position of the shortest possible LM-PAT product, PCR was also carried out with cDNA primed with an anchor-(dT)₁₂VN oligonucleotide that clamps to the 3′UTR–poly(A) junction (labeled ‘TVN’ above lane). An RNA sample was also subjected to RH-Hs GAPDH oligonucleotide-mediated cleavage by RNAse H and high resolution northern blot analysis (right panel). Cleavage was further done in the presence of oligo(dT) to generate a 3′ UTR fragment without poly(A) tail. Position of size markers are indicated to the right of the panels (in base-pairs or nucleotides, respectively). (D) The poly(A) tail lengths of R-luc reporter and endogenous GAPDH (control) or HMGA2 mRNAs were measured by LM-PAT assay six hours after transfection. (E) P19 cells were induced to differentiate by treatment with retinoic acid and RNA was purified from cells at the times indicated. Let-7 and U6 RNA (loading control) expression were measured by northern blotting (10 µg total RNA per lane), while HMGA2 and GAPDH mRNA poly(A) tail lengths were measured by LM-PAT assay.

Figure 2. Let-7-triggered target mRNA deadenylation is rapid.

HeLa cells were transiently transfected with let-7 reporter plasmids as in Fig. 1 , followed by protein and RNA analysis at the indicated times. (A) Let-7-mediated repression on protein (squares) and mRNA (triangles) levels, calculated as in Fig. 1B. Results are averages of triplicate data and are comparable to results gained from two independent experiments. See Fig. S2 for raw expression data. (B) LM-PAT assays of R-luc-3xb and R-luc-mut mRNAs (GAPDH as control). (C) LM-PAT products were visualized by ethidium bromide stain (top) or Southern blot (bottom).

Figure 3. let-7-mediated repression is relieved by depletion of RCK/p54 or TNRC6A despite target mRNA deadenylation.

HeLa cells were transfected with siRNAs targeting GW182, RCK, the non-specific miCXCR4 control (ns), or mock transfected (none) as indicated. (A) Residual mRNA levels were measured by qPCR between 24 and 48 hours and normalised against RPL13a. Bars represent the average expression of the relevant mRNA in three independent experiments with standard error. mRNA expression in cells transfected with ns was set at 1 (hence do not carry an associated error). (B) Extracts from cells harvested 24 or 48 hours after RCK or control siRNA transfection were analysed by western blotting and infrared fluorescent imaging using antibodies against RCK and tubulin. RCK protein level (normalised to tubulin loading control) was reduced to ∼25% (24 hours) and ∼11% (48 hours). * Cross-reacting non-specific band. (C, D) After 24 hours of knockdown, cells were transfected with let-7 reporter plasmids. R-luc activity (C) and mRNA levels (D) were measured 24 hours later and repression calculated as in Fig. 1. Derepression achieved by siRNA knockdown was calculated as 1 - (repression in siRNA treated cells/repression in “none” cells) and normalised to derepression by let-7 anti-miR seen in parallel transfections (set to 100% derepression, Fig. 1B. Bars represent averages of 4–10 (a) or 4–5 (b) independent experiments with standard error. (E) HeLa cells were transfected as above and total RNA purified 6 hours after plasmid transfection for LM-PAT assay of the R-luc and GAPDH mRNAs. Data representative of 3–4 independent experiments are shown. (F) After 30 hours of knockdown, cells were transfected with in vitro transcribed capped R-luc-mut mRNA with or without a poly(A) tail (expression of R-luc-mut with poly(A) tail in “none” set to 1.0 for each experimental set prior to averaging across experiments). Averaged results from 3–7 independent experiments are shown with standard error.

Figure 4. Full let-7-mediated translational repression of R-luc-3xb mRNA requires a poly(A) tail.

HeLa cells were transfected with capped reporter mRNAs and incubated for 16 hours. (A) Schematic of the variant R-luc-3xb and mut mRNAs. (B) R-luc activity from the R-luc-mut mRNAs co transfected with LNAs indicated (normalized to F-luc reference; expression from the cap&tail mRNA is set to 1.0 for each experiment prior to averaging across experiments). (C) Repression by let-7 was calculated as in Fig. 1B. Averaged results from 4–7 independent triplicate experiments are shown with standard error in B and C. (D, E) Multiple cell aliquots were co-transfected with R-luc reporter mRNAs and LNAs as indicated. Cells were harvested at different times to measure R-luc protein expression, which was not normalized to F-luc. Three time series of each kind (with duplicate measurements) were each scaled to total level of recovered R-luc protein and then averaged. Error bars represent standard error (except at two [range] and six hours [no error bars]). The functional half-life of an mRNA in cells is defined as the time required for half-maximal accumulation of R-luc activity , .

Figure 5. Blocking mRNA deadenylation impairs let-7-mediated translational repression.

(A) Schematic of the R-luc 3xb and mut cap&A₆₂ mRNAs indicating the site of ‘tail blocker’ (tb) binding. (B, C) HeLa cells were transfected with capped reporter mRNAs and incubated for 16 hours. (B) R-luc activity from the R-luc-mut-A₆₂ mRNAs cotransfected with tb or control (ns) LNA or no LNA (none). Normalized expression without LNA is set to 1.0. (C) Repression by let-7 was calculated as in Fig. 1. Averaged results from 4–7 independent triplicate experiments are shown with standard error in E and F. (D) Multiple cell aliquots were co-transfected with R-luc reporter mRNAs and LNAs as indicated (as for Fig. 4D,E). Cells were harvested at different times to measure R-luc protein expression, which was not normalized to F-luc. Two time series of this kind (with duplicate measurements) were each scaled to total level of recovered R-luc protein and then averaged. Error bars represent data range.