Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2017 Apr 6;169(2):326-337.e12.
doi: 10.1016/j.cell.2017.03.031.

Transcription Impacts the Efficiency of mRNA Translation via Co-transcriptional N6-adenosine Methylation

Boris Slobodin  1 , Ruiqi Han  2 , Vittorio Calderone  2 , Joachim A F Oude Vrielink  2 , Fabricio Loayza-Puch  2 , Ran Elkon  3 , Reuven Agami  4
Affiliations

Transcription Impacts the Efficiency of mRNA Translation via Co-transcriptional N6-adenosine Methylation

Boris Slobodin et al. Cell. .

Abstract

Transcription and translation are two main pillars of gene expression. Due to the different timings, spots of action, and mechanisms of regulation, these processes are mainly regarded as distinct and generally uncoupled, despite serving a common purpose. Here, we sought for a possible connection between transcription and translation. Employing an unbiased screen of multiple human promoters, we identified a positive effect of TATA box on translation and a general coupling between mRNA expression and translational efficiency. Using a CRISPR-Cas9-mediated approach, genome-wide analyses, and in vitro experiments, we show that the rate of transcription regulates the efficiency of translation. Furthermore, we demonstrate that m6A modification of mRNAs is co-transcriptional and depends upon the dynamics of the transcribing RNAPII. Suboptimal transcription rates lead to elevated m6A content, which may result in reduced translation. This study uncovers a general and widespread link between transcription and translation that is governed by epigenetic modification of mRNAs.

Keywords: N6-adenosine methylation; RNAPII; TATA; gene regulation; m(6)A; transcription; translation efficiency.

PubMed Disclaimer

Figures

None
Graphical abstract
Figure 1
Figure 1
A Screen for Examination of the Relationship between Transcription and Translation (A) Schematics of the barcoded polysomal profiling (BPP) approach. (B) A typical segregation of Rluc mRNAs. (C) Control Rluc transcript with 5′ TOP sequence exhibits rapid shift to non-translating fractions upon inhibition of mTORC1. (D) Control Rluc transcript shifts to denser fractions following splicing. The bar diagram below represents normalized relative Rluc protein expression assessed by luciferase assay in two separate clones. See also Figure S1.
Figure 2
Figure 2
Presence of the TATA Element in Promoters Enhances TE (A) Pro-Lib vectors encoding for TATA-containing promoters result in Rluc mRNAs shifted to the denser fractions of the gradient compared to other transcripts (e.g., under SPTBN1 promoter). (B) Pro-Lib vectors supplied with an artificial TATA yield Rluc mRNAs that are shifted to the denser fractions of the gradient compared to the parental TATA-less promoters. (C) Northern blot analysis of Rluc mRNAs; quantification reflects relative Rluc/Fluc ratio. (D) Relative levels of Rluc mRNAs and proteins produced by promoters with or without TATA element; note the super-induction of the protein expression. Data are represented as mean ± SEM, n = 4. (E) Relative levels of Rluc mRNAs and proteins produced by SV40 promoter upon mutagenesis of TATA; data are represented as mean ± SEM, n = 3. (F) Mutagenesis of SV40 promoter-derived TATA results in less efficient translation of Rluc mRNA as detected by BPP; data are represented as mean ± SEM of a representative gradient, n = 2. (G) Schematics of the CRISPR-Cas9-mediated mutagenesis of the endogenous TATA (in pink square) of c-Myc. Below: characterization of the different c-Myc alleles of one of the isolated clones, including relative abundances. (H) Mutagenesis of the c-Myc TATA results in lowered mRNA levels (left chart) and further reduced c-Myc activity (right chart); data are represented as mean ± SEM of n = 3. (I) Western blot analysis of c-Myc protein in the TATA-mutated clone and WT MCF7 cells; see Figure S2I for the uncropped blot. (J) Polysomal profilings of c-Myc mRNAs isolated from the clone with mutated TATA and WT MCF7 cells; data are represented as mean ± SEM of a characteristic gradient; n = 3. See also Figures S2G–S2K for the characterization of an additional clone.
Figure 3
Figure 3
Levels of mRNAs Positively Correlate with TE but Do Not Dictate It (A) Reporter mRNAs from the Pro-Lib screen were separated into groups with relatively high or low TE (derived from the polysomal profiles, see Figure S4) and compared with their expression levels (estimated from read counts observed for each vector over all fractions). p values calculated using Wilcoxon’s test. (B) Same comparison as described in (A) was performed after exclusion of mRNAs transcribed from promoters with artificial TATA element. (C) Levels of mRNA and protein resulting from the induced TRex-Rluc gene were measured and plotted relatively to the non-induced condition; data are represented as mean ± SEM of n = 3. (D) BPP examination of either induced or non-induced TRex-Rluc mRNAs; data are represented as mean ± SEM of a characteristic gradient; n > 5, see also Figure S5A. (E) Upper panels: paired RNA-seq and Ribo-seq datasets from different human cell lines were examined for relationship between mRNA expression level and TE (calculated as the (log2) ratio between densities of ribosome footprint and RNA-seq reads). Lower panels: comparisons between the 10% of genes with lowest and highest expression levels are presented; p values calculated using Wilcoxon’s test. (F) MCF7 cells were transfected with fold-wise amounts of in vitro transcribed (using HeLa nuclear extract) Rluc mRNA followed by measurement of RLuc activity after 18 hr; data are represented as mean ± SEM of n = 3. (G) Levels of Rluc mRNA and protein were compared between two populations of MCF7 cells expressing near single or multiple integrated copies of Lenti-Rluc unit. Data are represented as mean ± SEM of n = 3. (H) BPP analysis of Rluc transcripts described in (G); data are represented as mean ± SEM of a characteristic gradient, n = 3.
Figure 4
Figure 4
TE is Affected by the Rate of Transcription (A) Upper panels: positive genome-wide correlations between translational efficacies and rates of transcription in BJ and MCF7 cells. Lower panels: direct comparisons between 10% of genes with lowest and highest transcription rates. GRO-seq data were from (Korkmaz et al., 2016, Léveillé et al., 2015); Ribo-seq and RNA-seq data were from (Loayza-Puch et al., 2013); p values were calculated using Wilcoxon’s test. (B) Expression levels of Rluc mRNA and protein were examined following parallel induction of the TRex-Rluc gene and CPT treatment for 7 hr; data are represented as mean ± SEM of n = 3. (C) Lysates of the two populations described in (B) were subjected to BPP procedure; data are represented as mean ± SEM of a characteristic gradient, see also Figure S5D; n = 3. (D) Upper panels: genome-wide correlations between rates of transcription and TE changes after treatment with CPT. Lower panels: direct comparison of the effect on TE between the 10% of most highly and lowly transcribed genes; p values were calculated using Wilcoxon’s test. (E) Polysomal profilings of mRNAs after CPT treatment (red) show reduced TE compared to untreated cells (green). Columns represent the relative mRNA levels as detected by qRT-PCR; ALG8 was used as a control gene. Data are presented as mean ± SEM of three technical measurements of a characteristic graident; n = 3. See also Figure S6C. (F) Cells with barcoded TRex-IRES-Rluc cassette (schematics) were induced for 18 hr and subjected for examination of Rluc mRNA and protein levels relative to the non-induced cells; data are presented as mean ± SEM of n = 3. (G) Same cells as in (F) were treated in a similar way and subjected to BPP procedure. Data are presented as mean ± SEM of three measurements of a characteristic gradient; n = 3. (H) Pro-Lib vectors with SZT2 promoters, with or without the TATA element, were supplemented with IRES-encoding sequence as shown on the schematics. Cells expressing these constructs were subjected to quantification of mRNA and protein levels; data are presented as mean ± SEM of n = 3.
Figure 5
Figure 5
Enhancement of m6A Modification of mRNAs upon Attenuated Transcription (A) Cells bearing the TRex-Rluc gene were induced for 18 hr and subjected to the MeRIP procedure, together with untreated cells. Recovery efficiencies of regions spanning the Rluc mRNA were examined, normalized to the input levels and compared; AHNAK was used as a control methylated mRNA. Data are represented as mean ± SEM of n = 3. (B) Rluc mRNAs were transcribed in vitro using HeLa extract as detailed, resolved on agarose gels and quantified (left chart, n = 3) and further subjected to the relative comparison of their m6A contents (right chart, n = 2). (C) Total RNA isolated from MCF7 cells treated with CPT for 5 hr was enriched for polyA+ population and quantified (left chart, n = 3) and further subjected to the relative comparison of their m6A contents (right chart, n = 3). (D) mRNAs presented in Figure 4E were tested for recovery after MeRIP procedure (striped columns) relatively to their RNA levels (plain columns) using qRT-PCR. Data are represented as mean ± SEM of n = 2. See also Figures S6D and S6E. (E) Relative levels of polyA+ RNA populations (plain columns) in cells expressing different RNAPII mutants, and their relative m6A levels (striped columns) were examined; n = 2. (F) MCF7 cells were treated with CPT for 5 hr and subjected to immuno-precipitation of RNAPII, followed by detection of METTL3 protein using western blot. A characteristic blot is presented, n = 4. In the input panel, GAPDH was probed on a separate blot.
Figure 6
Figure 6
Effect of the m6A Modifications on Translation (A) Rluc mRNAs were transcribed in vitro using HeLa extract in the indicated conditions, and transfected into MCF7 cells in fold-wise amounts; RLuc enzymatic activity was measured after 24 hr. The trend lines are represented by the dashed lines; n = 3. (B) MCF7 cells were treated with CPT for 5 hr and subjected to polysomal profiling with subsequent measurement of the average levels of total RNA in the collected fractions. Data are presented as mean ± SEM, n = 3. (C) Polysomal segregation of total RNA from cells expressing “normal speed” or “slow” RNAPII; data are presented as mean ± SEM of n = 3. See also Figure S6H. (D) TRex-Rluc gene was induced for 18 hr in MCF7 cells transfected with the detailed siRNAs, or left uninduced. Expression levels of RLuc protein were assessed; data are represented as mean ± SEM of n = 3. (E) Cells bearing the TRex-Rluc gene were transfected with the indicated siRNAs, grown for 56 hr and subjected to BPP. Upper panel: non-induced Rluc; lower panel: Rluc expression was induced for 24 hr prior to harvesting; data are represented as mean ± SEM of three measurements of a characteristic gradient; n = 3. (F) Results of the MeRIP experiment presented in Figure 5A were normalized to calculate the total relative recovery of Rluc mRNAs by the different regions. Error bars represent mean ± SEM of n = 3. (G) Rluc mRNAs were in vitro transcribed using T7 polymerase in the presence of the indicated proportions of m6A nucleoside and in vitro translated using indicated amounts of mRNA as input. Data are represented as mean ± SEM of n = 4. (H) Model of the transcription-dependent regulation of TE. Slow or paused RNAPII results in the enhanced interaction with MTC (step 1), which leads to enhanced deposition of m6A on mRNAs (step 2), negatively affecting translation efficiency (step 3).
Figure S1
Figure S1
Setup of the Promoter Library (Pro-Lib) and Barcoded Polysomal Profiling (BPP), Related to Figure 1 (A) Schematics of the Pro-Lib vector. The reporter Rluc gene is flanked upstream by sequences encoding for putative promoters of interest (magenta colored) and downstream by unique 10-nt barcodes (also magenta colored). Transcription of the control Fluc gene is driven by HSV-TK promoter in all vectors. (B) Several Pro-Lib vectors with the indicated regions cloned upstream of Rluc gene were integrated into MCF7/FRT cells, grown separately and subjected to dual Rluc/Fluc assay. Note the uniform expression of the tested constructs that is substantially higher than the background signal (assessed by expression of vectors with no promoter or with random intron sequence cloned instead of a promoter) and milder than the expression of the strong SV40 promoter. Data are represented as mean ± SEM of n = 3. (C) Total RNA isolated from fractions of a typical BPP experiment were separated on 1.5% agarose gels and documented. Note the characteristic segregation of RNA as well as the relative enrichment of ribosomal RNAs in the polysomal fractions. (D) Total RNA concentrations in the different fractions were measured following lysis of cells in the presence or absence of EDTA and polysomal profiling procedure. Note the dramatic drop of the RNA content in the fractions 9-13 upon presence of EDTA in the lysis buffer. (E) BPP experiment identifies two Rluc transcripts (driven by SV40 and CD44 promoters) exhibiting super-sensitivity to the inhibition of mTORC1. Note the multiple barcodes for each promoter showing identical behavior. (F) The transcripts mentioned in (E) contain putative 5′TOP sequences (marked in blue). The TSS of the SV40 promoter was reported previously (Byrne et al., 1983), while the TSS of the CD44-driven construct was determined by 5′-RACE analysis in this study.
Figure S2
Figure S2
Effect of TATA Box on TE, Related to Figure 2 (A) Pro-Lib vectors encoding for promoters with TA-rich regions at their 3′ ends result in Rluc mRNAs shifted to the denser fractions of the gradient, compared to other transcripts (e.g., under SPTBN1 promoter). (B) Average standardized profiles of polysome fractions for the TATA+/TATA- paired constructs in Pro-Lib. (C) Promoter region of ASNSD1 gene yields better translated Rluc mRNAs when supplemented with an artificial TATA box at its 3′end. Cells expressing Rluc mRNAs from either native ASNSD1 promoter or with artificial TATA box were subjected to BPP and plotted for a direct comparison of TE (left panel). To test the robustness of the positive effect of TATA on TE, cells were subjected to various stress conditions, prior to BPP (right panels). Note that in all listed conditions the presence of TATA box enhances TE. (D) Point mutagenesis of the artificial TATA box reduces mRNA levels and prevents super-induction of the protein production. SZT2 and ASNSD1 promoters supplemented with an artificial TATA box were subjected to site-directed mutagenesis of the TATA box sequence (muTATA). Levels of the Rluc mRNAs and protein compared to the native promoters (lacking TATA) are shown; data are represented as mean ± SEM of n = 3. (E) Rluc mRNAs resulting from either native ASNSD1 promoter, one with the artificial TATA box (+TATA) or with a point-mutated TATA sequence (muTATA) were subjected to the BPP procedure. Note the shift toward denser fractions caused by the artificial addition of TATA box as well as the shift to the opposite direction upon mutation of the TATA sequence. (F) Screenshot of the DBTSS database (release 9.0) showing the distribution of TSS of c-Myc gene in MCF7 cells. Note that most of TSSs stem from a single TATA box-containing promoter (see zoom-in window). (G) Genotyping of the c-Myc proximal promoter region of the second clone bearing modified c-Myc TATA box. Numbers on the left represent relative abundances of the different alleles. (H) Lysates of the clone described in (G) were subjected to polysomal profiling and plotted against wild-type MCF7 cells. SEM represents three measurements of a characteristic gradient; n = 2. ∗∗∗p < 0.005, ∗∗∗∗p < 0.001. (I) Lysates of the two clones bearing modified c-Myc TATA boxes together with wild-type MCF7 cells (WT TATA) and cells transfected with vector encoding for c-Myc (Myc OE), were resolved on SDS-PAGE and probed for c-Myc and GAPDH proteins. Replicates on the blot represent biological repeats. (J) Relative quantification of the c-Myc bands presented in (I). (K) TE of a control gene (ALG8) was assessed by probing polysomal fractions described in Figure 2J with primers detecting ALG8 mRNA. SEM represents three measurements of a characteristic gradient; n = 3.
Figure S3
Figure S3
Analysis of Rluc mRNAs, Related to Figures 2, 3, 5, and 6 (A and B) Rluc mRNAs transcribed by ASNSD1 (A) or SZT2 (B) promoters either lacking or bearing an artificial TATA box were subjected to 5′RACE analysis in order to determine their TSSs. Reads were plotted in a quantitative manner using IGV software. Note the first ATG of the Rluc ORF and the precise focusing of the TSS by the TATA box. (C) Evaluation of the effect of the different 5′UTRs on TE. Rluc transcripts bearing the different 5′UTRs identified in A,B (see the numbered arrows) were transcribed in vitro using T7 RNA polymerase, polyadenylated and transfected into MCF7 cells in fold-wise amounts. Renilla luciferase activity was measured 24 hr later and plotted in a relative manner to draw the trend lines that represent the respective ratios of translation. The rightmost panel shows the absolute Rluc signals; n = 4. (D) Rluc mRNAs expressed from either induced or non-induced TRex-Rluc gene were subjected to 5′RACE analysis. Reads (numbers of sequenced colonies) were plotted in a quantitative manner on a scale ranging from TATA box (−243) to ATG (+1). (E) Rluc transcripts produced in vitro using HeLa extract in optimal conditions (1.5mM Mg++), upon high MgCl2 (3mM Mg++) or from promoter with mutated TATA element (muTATA) were subjected to 5′- and 3′-RACE analyses. The reads from both assays were analyzed and plotted in a quantitative manner using IGV software. (F) Transcripts described in (E) were subjected to analysis of the length of polyA-tails. (G) The TRex-Rluc cassette was induced for 24 hr in cells transfected with the indicated siRNAs. After isolation of total RNA, the 5′- and 3′ ends of the Rluc mRNAs were determined as detailed in (E).
Figure S4
Figure S4
Global TE Assessment of the Pro-Lib mRNAs, Related to Figure 3 Barcoded Rluc mRNAs with sufficient read coverages (> 10.000 reads over 15 polysomal fractions) in three independent BPP experiments were divided into two main groups by hierarchical clustering. One arm includes barcodes that showed peak in fractions 9-11 (lower TE, A-B) and the other arm includes barcodes whose polysomal profile peaked in fraction 11-13 (higher TE, C-D). Plots to the right of the heatmaps show the mean profile of the vectors assigned to each group over the three repeats; error bars = SD.
Figure S5
Figure S5
Transcription rates impacts translation, Related to Figures 3 and 4 (A) Segregation of the induced versus non-induced Rluc mRNAs on sucrose gradients. Cells expressing barcoded Rluc gene were treated for 21 hr to induce the expression of Rluc or left untreated, and subjected to BPP procedure. (B) Upper panels: test for correlation between mRNA expression and TE, as described in Figure 3E, was performed here on data from BJ cells in quiescent, senescent, and transformed states. Lower panels: direct comparison between 10% of mRNAs with the lowest and 10% with the highest abundance. (C) In MCF7 cells, a similar correlation was tested in p53-activated (by Nutlin3 treatment) versus the control cells. Both datasets are from (Loayza-Puch et al., 2013), representation as detailed for (B). (D) Cells with induced Rluc were treated with CPT (or DMSO) for 5 hr and subjected to BPP analysis.
Figure S6
Figure S6
Related to Figures 4 and 5 and 6 (A) Genes with high ribosomal occupancy in their 5′UTRs show overall reduced TE in both BJ and MCF7 cells; p values were calculated Wilcoxon’s test. (B) These genes do not display any clear correlation between their transcription rate and TE (showing two independent replicates), and were removed from the genome-wide analysis. (C) Treatment with CPT causes attenuation of translation. Transcripts showed in Figure 4E and additional mRNAs were detected in polysomal profiling experiments using primers against their 3′ UTRs. (D) Rluc mRNAs produced in cells stably expressing Lenti-Rluc cassette integrated at either single- or multiple-copy numbers, were subjected to MeRIP procedure and analyzed by qRT-PCR. In each experiment, the level of recovery from cells with multi-copy integration was calculated relatively to the recovery of the single-copy population; n = 4. (E) Rluc mRNAs transcribed from the promoter pairs with or without intact TATA box element were subjected to MeRIP procedure and analyzed by qRT-PCR; n = 2. (F) Efficiency of METLL14 and YTHDF1 siRNA-mediated knockdown. MCF7 cells were transfected with siRNAs targeting METTL14 or YTHDF1 genes or control non-targeting siRNA. After 56 hr, the cells were harvested, subjected to RNA isolation and tested for the levels of the respective mRNAs using qRT-PCR. (G) MCF7 cells transfected with the indicated siRNAs were subjected to polysomal profiling procedure and an average levels of RNA in the collected fractions were measured using NanoDrop and plotted; n = 2. (H) Cells expressing “normal speed” or “slow” mutants of RNAPII were subjected to polysomal profiling on sucrose gradients. Several mRNAs were detected using qRT-PCR to monitor changes in their TE; a characteristic profile is shown, n = 2.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Anders S., Pyl P.T., Huber W. HTSeq–a Python framework to work with high-throughput sequencing data. Bioinformatics. 2015;31:166–1699. - PMC - PubMed
    1. Byrne B.J., Davis M.S., Yamaguchi J., Bergsma D.J., Subramanian K.N. Definition of the simian virus 40 early promoter region and demonstration of a host range bias in the enhancement effect of the simian virus 40 72-base-pair repeat. Proc. Natl. Acad. Sci. USA. 1983;80:721–725. - PMC - PubMed
    1. Choi J., Ieong K.W., Demirci H., Chen J., Petrov A., Prabhakar A., O’Leary S.E., Dominissini D., Rechavi G., Soltis S.M. N(6)-methyladenosine in mRNA disrupts tRNA selection and translation-elongation dynamics. Nat. Struct. Mol. Biol. 2016;23:110–115. - PMC - PubMed
    1. Core L.J., Waterfall J.J., Lis J.T. Nascent RNA sequencing reveals widespread pausing and divergent initiation at human promoters. Science. 2008;322:1845–1848. - PMC - PubMed
    1. Dominissini D., Moshitch-Moshkovitz S., Schwartz S., Salmon-Divon M., Ungar L., Osenberg S., Cesarkas K., Jacob-Hirsch J., Amariglio N., Kupiec M. Topology of the human and mouse m6A RNA methylomes revealed by m6A-seq. Nature. 2012;485:201–206. - PubMed