Subcellular mRNA Localization Regulates Ribosome Biogenesis in Migrating Cells - PubMed (original) (raw)
Subcellular mRNA Localization Regulates Ribosome Biogenesis in Migrating Cells
Maria Dermit et al. Dev Cell. 2020.
Abstract
Translation of ribosomal protein-coding mRNAs (RP-mRNAs) constitutes a key step in ribosome biogenesis, but the mechanisms that modulate RP-mRNA translation in coordination with other cellular processes are poorly defined. Here, we show that subcellular localization of RP-mRNAs acts as a key regulator of their translation during cell migration. As cells migrate into their surroundings, RP-mRNAs localize to the actin-rich cell protrusions. This localization is mediated by La-related protein 6 (LARP6), an RNA-binding protein that is enriched in protrusions. Protrusions act as hotspots of translation for RP-mRNAs, enhancing RP synthesis, ribosome biogenesis, and the overall protein synthesis in migratory cells. In human breast carcinomas, epithelial-to-mesenchymal transition (EMT) upregulates LARP6 expression to enhance protein synthesis and support invasive growth. Our findings reveal LARP6-mediated mRNA localization as a key regulator of ribosome biogenesis during cell migration and demonstrate a role for this process in cancer progression downstream of EMT.
Keywords: EMT; LARP6; La-related proteins; RNA localization; cancer; invasion; protrusion; ribosomal proteins; ribosome biogenesis.
Copyright © 2020 The Authors. Published by Elsevier Inc. All rights reserved.
Conflict of interest statement
Declaration of Interests The authors declare no competing interests.
Figures
Graphical abstract
Figure 1
RP-mRNAs Localize to Protrusions of All Migratory Cells (A) Schematic representation of transwell-based protrusion versus cell-body analysis experiments. (B) Panel of normal and malignant cell lines from diverse tissues of origin, chosen for transwell-based profiling. (C) RP-mRNAs are ubiquitously enriched in protrusions. Transcriptome distributions between protrusion and cell-body fractions were measured by RNA-seq in the panel of cell lines outlined in (B). Log2 of protrusion/cell body RNA ratio values for each cell line (Dataset S1) was plotted, with RP-mRNAs highlighted in green. ∗MDA- MB231 data were obtained from Mardakheh et al. (2015). All other cell lines were measured from a single matching protrusion and cell-body biological replicate. (D) Validation of RP-mRNA localization to protrusions by RNA-FISH. Representative RNA-FISH images of protrusions and cell bodies of MDA-MB231 cells, stained with probes against the indicated mRNAs (green). Cell boundaries (dashed lines) were defined by co-staining of the cells with anti-tubulin antibody or CellTracker. The filters (gray) were visualized by transmitted light microscopy. (E) Quantification of protrusion to cell-body RNA-FISH ratio values from experiments shown in (D). A total of 6–10 large field of view images from 2 independent experiments were quantified per each probe. (F) Schematic representation of the experimental setting for RNA-FISH imaging of cells invading through 3D collagen-I-matrix. Cells were seeded on the top collagen-I gels and allowed to invade into the matrix for 48 h, before fixation, staining, and confocal imaging of the invaded cells. (G) RP-mRNAs localize to the protrusions of MDA-MB231 cells in 3D. Representative RNA-FISH images of MDA-MB231 cells invading through collagen-I as described in (F), stained with probes against mRNAs (green). Cell boundaries (dashed lines) were defined by co-staining with anti-tubulin antibody. (H) Quantification of the polarization index (PI) values (Park et al., 2012) for the experiments shown in (G), as a measure of displacement of mRNAs away from the cell body. Each data point represents the PI value for a single quantified cell. A total of 22 cells from 2 independent experiments were quantified per each probe. All scale bars, 10 μm.
Figure 2
Depletion of LARP Proteins Reveals a Role for LARP6 in RP-mRNAs Localization to Protrusions (A) Quantitative proteomics reveals protrusion-enriched RBPs. Left: volcano plot comparison of protein levels in protrusions relative to cell bodies, across 6 independent cell lines from Figure 1B. Log2 of protrusion/cell body protein ratio values from each cell line (Dataset S2) were used to calculate Benjamini-Hochberg corrected p values for protrusion enrichment and depletion, using a one-sample t test analysis. Protrusion-enriched “RNA-binding” proteins (FDR < 0.05), defined according to GOMF database, are marked in red. RIGHT: The list of individual protrusion-enriched RBPs marked on the volcano plot. (B) siRNA screening reveals LARP6 as a crucial regulator of RP-mRNA localization to protrusions. Representative RNA-FISH images of RPL34 mRNA in protrusions of MDA-MB231 cells (green) transfected with non-targeting (NT) control or indicated siRNAs. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. The transwell filters (gray) were visualized by transmitted light microscopy. (C) Quantification of RPL34 mRNA enrichment in protrusions from experiments shown in (B). A total of 5–10 large field of view images per condition, measured from 3 independent experiments, were quantified. p values were calculated using two-tailed homoscedastic t test. ∗∗∗p < 0.001. (D) Validation of LARP6 by 3 independent siRNAs. Representative RNA-FISH images of RPL34 mRNA in protrusions of MDA-MB231 cells (green) transfected with control or 3 independent LARP6 siRNAs. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. The transwell filters (gray) were visualized by transmitted light microscopy. (E) Quantification of RPL34 mRNA enrichment in protrusions from experiments shown in (D). A total of 5 large field of view images per condition, measured from 2 independent experiments, were quantified. p values were calculated using two-tailed homoscedastic t test. ∗∗p < 0.01. (F) LARP6 depletion prevents RP-mRNAs localization to protrusions of 3D invading cells. Representative RNA-FISH images of RPL34 mRNA distributions in NT- or LARP6 siRNA-transfected MDA-MB231 cells (green) invading through 3D collagen-I matrix, as described in Figure 1F. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. (G) Quantification of the polarization index values from experiments shown in (F) as a measure of displacement of mRNAs away from the cell body. Each data point represents the PI value for a single quantified cell. A total of 18 cells per condition from 2 independent experiments were quantified. p values were calculated using two-tailed, homoscedastic t test. ∗∗∗p < 0.001. (H) Depletion of LARP6 significantly reduces RP-mRNA levels in protrusions. MDA-MB231 cells transfected with NT control or LARP6 siRNAs were subjected to transwell fractionation followed by RNA-seq. Log2 of NT/LARP6 KD transcript read counts in the protrusion fractions from 2 independent experiments are plotted (Dataset S3), with RP-mRNAs marked in green. Arrow marks the direction of RP-mRNA shift, with the Benjamini-Hochberg-corrected p value of the shift reported next to it. (I) Depletion of LARP6 significantly increases RP-mRNA levels in cell bodies. Log2 of NT/LARP6 KD transcript read counts in cell-bodies of the cells described in (H) are plotted (Dataset S3), with RP-mRNAs marked in green. Arrow marks the direction of RP-mRNA shift, with the Benjamini-Hochberg corrected p value of the shift reported next to it. (J) LARP6 depletion induces mis-localization of RP-mRNAs from protrusions to cell bodies. 2D-annotation enrichment analysis (Cox and Mann, 2012) of data shown in (H) and (I). Each data point represents a functional category from GO and KEGG databases, with similar categories being highlighted in the same colors (Dataset S4). Upon LARP6 depletion, mRNAs coding for ribosomal and translation-related categories (green) change in an anti-correlative fashion in protrusions and cell-bodies, suggestive of mis-localization. Other significantly altered categories change in a correlative fashion, suggestive of expression change throughout the cell. All scale bars, 10 μm.
Figure 3
Transcriptome-wide iCLIP Studies Reveal Direct Binding of LARP6 to RP-mRNAs (A) LARP6 is localized to cytoplasmic puncta that track microtubules. Representative IF images of LARP6 (red) and α-tubulin (green) in MDA-MB231 cells grown on collagen-coated slides. Nucleus was stained with NuclearMask (blue). (B) LARP6 puncta are enriched in protrusions. Representative IF images of LARP6 (red) in protrusions and cell bodies of MDA-MB231 cells. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. (C) Quantification of IF images from experiments shown in (B), revealing LARP6 enrichment in protrusions. A total of 11 large field of view images, measured from 2 independent experiments were quantified. (D) LARP6 co-localizes with RP-mRNAs in protrusions. Representative RNA-FISH and IF co-staining images of RPL34 mRNA (green) and LARP6 (red) in protrusions and cell bodies of MDA-MB231 cells. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. (E) Quantification of the % of co-localization of RPL34 mRNA with LARP6 in corresponding protrusion and cell-body images from experiments shown in (D). A total of 13 large field of view images from 2 independent experiments were quantified. Red lines connect values of protrusion and body from the corresponding images. p values were calculated using a two-tailed, homoscedastic t test. ∗∗∗p < 0.001. (F) Metaprofile plot of LARP6 iCLIP crosslink sites at the aligned annotated intergenic-5′UTR junctions (2,204 landmarks), showing preferential association with specific regions at the vicinity of TSS. (G) Metaprofile plot of LARP6 iCLIP crosslink sites at the aligned annotated 5′UTR-ORF junctions (4,122 landmarks), showing preferential association with the translation start site. (H) Metaprofile plot of LARP6 iCLIP crosslink sites at the aligned annotated 3′UTR-intergenic junctions (6,333 landmarks), showing association throughout the 3′UTR. (I) LARP6 mainly binds protein-coding transcripts. Pie chart showing the prevalence of coding versus non-coding RNAs among LARP6 binding targets (Dataset S6). (J) The KEGG category of ribosome (green), which is comprised all RP-mRNAs, is significantly enriched among LARP6-binding targets. Fisher’s exact test analysis (FDR < 0.02) of mRNA categories, which are significantly over-represented among the identified LARP6 targets. Each data point represents a functional category from KEGG database, with similar categories highlighted by the same colors (Dataset S7). (K) LARP6 interacts with RP-mRNAs via multiple regions. Distribution of LARP6-binding regions in RP-mRNAs. (L) An example genomic view of LARP6-specific binding sites after peak calling (gray tracks) in an RP-mRNA (RPLP2), along with read intensities for GFP and GFP-LARP6 iCLIP runs. Four distinct LARP6-binding sites are mapped to the RPLP2 locus: two mapping to the ORF region, one to RPLP2 3rd intron, which is annotated as SNORA52, and one to the 5′UTR. Inset: zoomed view of RPLP2 5′UTR showing the LARP6-binding site overlapping with the 5′TOP. Note that for most RP-mRNAs, annotation of TSS in Ensembl is further upstream of the more accurately annotated DBTSS (Suzuki et al., 2018). (M) Schematic representation of the MS2 reporter system for live-cell monitoring of 5′TOP mediated RNA localizations. (N) WT 5′TOP motif is sufficient for RP-mRNA localization to protrusions. Representative still images of the GFP-MCP signal in transwell protrusions of WT or MUT 5′TOP reporter engineered MDA-MB231 cells described in (M), following induction of reporter expression with 2 μg/mL doxycycline for 12 h. GFP-MCP exhibits a punctate pattern in protrusions of WT 5′TOP reporter expressing cells, indicative of association with mRNA particles, as opposed to a diffuse pattern in protrusion of MUT 5′TOP reporter expressing cells. (O) Quantification of mRNA particles in protrusions of WT-5’TOP versus MUT-5’TOP reporter expressing cells from experiments shown in (N). A total of 25 (WT) and 28 (MUT) time-lapse videos (3 s at 0.2-s intervals) from 2 independent experiments were quantified. The number of discrete particles identified at every frame image were quantified and normalized to the protrusion area to determine mRNA molecule density. The p value was calculated using a two-tailed, homoscedastic t test. All scale bars, 10 μm.
Figure 4
LARP6-Dependent RP-mRNA Localization Enhances RP Synthesis and Ribosome Biogenesis (A) Schematic representation of the Ribopuro-FISH assay. A short pulse of puromycin results in labeling of nascent proteins. When emetine is present, puromycylated peptides remain associated to the ribosome. Detection of these peptides with anti-puromycin antibody visualizes cellular sites of active translation. Co-detection of a specific mRNA by RNA-FISH marks the fraction of mRNA associated with translation sites. (B) RP-mRNAs are associated with active sites of translation in protrusions. Representative Ribopuro-FISH images of RPL34 mRNA (green) and puromycin (red) in protrusions and cell bodies of MDA-MB231 cells at the indicated time points post protrusion induction. Cell boundaries (dashed lines) were defined from co-staining with anti-tubulin antibody. All scale bars, 10 μm. (C) Translation in protrusions relative to the cell bodies increases over time. Quantification of puromycin staining intensities in protrusions relative to cell bodies, from experiments shown in (B). A total of 7–10 large field of view images per condition, from 2 independent experiments, were quantified. (D) Association of RPL34 mRNAs with active sites of translation is higher in protrusions than cell bodies. Quantification of % RPL34 mRNA co-localization with puromycin in protrusions relative to the cell bodies from experiments shown in (B). A total of 6–10 large field of view images per condition were quantified as in (C). p values were calculated for each time-point relative to time zero, using a two-tailed, homoscedastic t test. ∗∗∗p < 0.001. (E) Schematic diagram of pulsed SILAC proteomics analysis of changes in protein translation rates induced by protrusion formation. Light (L) SILAC-labeled MDA-MB231 cells were grown overnight on top of two transwell filters without any media in the bottom chamber. The next day, media on top was changed to medium (M) or heavy (H) SILAC media, followed by addition of the same label media to the bottom chamber of one of the two transwells in order to open the pores to the cells. Cells were then allowed to form protrusions for 1, 2, 4, or 8 h, or left without protrusions for the same length of time as control. H/M ratios for each protein were determined by MS analysis of the whole cell lysates, as measurement of translation rate changes between open pore (with protrusions) and closed pore (without protrusions) conditions (Dataset S9). (F) Translation of RPs (green) is significantly increased after 4 h of protrusion formation. Log2 of H/M ratio values from 2 reciprocally labeled biological replicate experiments were plotted against each other (Dataset S9). Arrow marks the direction of shift in RPs, with Benjamini-Hochberg corrected p value of the shift reported next to it. (G) Translation of RPs (green) is significantly increased after 8 h of protrusion formation. Log2 of H/M ratio values from 2 reciprocally labeled biological replicate experiments were plotted against each other (Dataset S9). Arrow marks the direction of shift in RPs, with Benjamini-Hochberg corrected p value of the shift reported next to it. (H) 2D-annotation enrichment analysis of data shown in (F) and (G). Each data point represents a functional category from GO and KEGG databases, with similar categories highlighted with the same colors (Dataset S10). Translation of ribosomal and translation-related protein categories (green), as well as a number of RNA-metabolism-related protein categories (pink), is significantly enhanced following protrusion induction for 4 and 8 h. (I) Schematic representation of the experimental outline for pulsed SILAC mediated assessment of subcellular distributions of nascent proteins following protrusion induction. Absolute abundances of light (L)-, medium (M)-, and heavy (H)-labeled proteins in each subcellular compartment were measured by iBAQ, in presence or absence of protrusions, and used to calculate the % of labeled protein in each compartment. (J) Newly synthesized RPs accumulate in the nucleus. Box plot of the % of old and nascent RPs in the nuclear and cytosolic fractions of MDA-MB231 cells. Old RPs (L), nascent RPs synthesized under basal conditions without protrusions (M), and nascent RPs synthesized under protrusion-induced condition (H) were distinguished by their SILAC labeling state and separately quantified in each fraction within a single experiment (Dataset S11). Error bars are min-max range. Significance p values were calculated using a two-way t test analysis between L and M or H values. ∗∗∗p < 0.001. (K) Total RP levels are significantly increased upon long-term protrusion induction in NT control siRNA-treated MDA-MB231 cells. Proteome changes between closed and open-pore (overnight) conditions in NT control siRNA-treated MDA-MB231 cells were quantified by TMT quantitative proteomics (Dataset S14). Log2 of NT siRNA open/close ratios from 2 biological replicate experiments were plotted against each other. The arrow marks the direction of shift in RP levels, with Benjamini-Hochberg corrected p value of the shift reported next to it. (L) Total RP levels are do not significantly change upon long-term protrusion induction in LARP6 siRNA-treated MDA-MB231 cells. Proteome changes between closed and open pore (overnight)-conditions in LARP6 siRNA-treated MDA-MB231 cells were quantified by TMT quantitative proteomics (Dataset S14) (n.s., not significant). Log2 of LARP6 siRNA open/close ratios from 2 biological replicate experiments were plotted against each other. (M) LARP6 depletion inhibits protrusion-induced enhancement of overall protein synthesis. Transwell seeded NT control and 2 independent LARP6 siRNA-treated MDA-MB231 cells were either prevented from protruding through pores (pores closed), or allowed to form protrusions (pores open) for 24 h, before labeling with OPP for 15 min. OPP was then visualized by Click-chemistry-mediated Alexa Fluor-488 labeling. Representative images of the cells from top of the filters are displayed. Cell boundaries (dash lines) were defined by anti-tubulin staining. Scale bars, 20 μm. (N) Quantification of normalized OPP staining levels from (M). A total of 15 large field of view images per condition from 2 independent experiments were quantified. p values were calculated using two-tailed, homoscedastic t test. n.s., non-significant; ∗∗∗p < 0.001.
Figure 5
LARP6 Is Important for Ribosome Biogenesis, 3D Invasion, and Proliferation of Migrating Cells (A) Schematic representation of SILAC proteome analysis following LARP6 depletion. Light (L) SILAC-labeled MDA-MB231 cells, transfected with NT control siRNA or 2 independent LARP6 siRNAs for 72 h, were lysed and mixed with H-labeled non-transfected MDA-MB231 lysates as reference. H/L ratio values in each mix was then used to calculate relative protein abundance changes. (B) LARP6 depletion significantly decreases total RP levels in MDA-MB231 cells. Changes in individual protein levels following LARP6 depletion with 2 independent siRNAs were quantified as described in (A) and plotted (Dataset S15). Benjamini-Hochberg corrected p value of decrease in RP (green) levels is reported on the graph. (C) 2D-annotation enrichment analysis of data shown in (B). Each data point represents a protein category inferred from GO and KEGG, and similar categories are highlighted by the same colors (Dataset S16). Categories of proteins comprised RPs (green), translation related (light pink), and RNA metabolism related (pink) are all significantly downregulated upon LARP6 depletion by 2 independent siRNAs. (D) LARP6 depletion results in accumulation of 5′ETS containing pre-rRNAs. RT-qPCR of 5′ETS pre-rRNA in MDA-MB231 cells transfected with NT control siRNA or 2 independent LARP6 siRNAs for 72 h. A specific probe against the 5′ETS region, along with a specific probe against GAPDH mRNA as loading control, were used to quantify –ΔΔCT values. Average values were calculated from 3 independent experiments, each performed in at least 3 technical replicates, per condition. Error bars are SD. p values were calculated using two-tailed, homoscedastic t test. ∗p < 0.05. (E) LARP6 depletion hampers the ability of MDA-MB231 cells to invade through 3D collagen. MDA-MB231 cells were treated with NT control siRNA or 2 independent siRNAs against LARP6 for 72 h before being subjected to 3D collagen-I Invasion assay. 5 × 5 tiled confocal images of fixed, Hoechst-stained cells (blue) at different migrated distances from the start point are displayed. Scale bars, 200 μm. (F) Quantification of invaded cell numbers from (E). Average values were calculated from 3–5 biological replicates per condition. Error bars are SD. p values were calculated using two-tailed, homoscedastic t test. ∗p < 0.05; ∗∗p < 0.01; ∗∗∗p < 0.001. (G) Long-term LARP6 depletion decreases MDA-MB231 proliferation. MDA-MB231 cells were transfected with NT control siRNA or 2 independent LARP6 siRNAs for 72 h, before reseeding to form colonies for a further 10 days prior to crystal violet staining. (H) Optical density of crystal-violet-stained colonies from experiments shown in (G) were measured by 570-nm absorbance (OD 570) after dye extraction. Average values were calculated from 3 independent experiments, each performed in 3 technical replicates. Error bars are SD. p values were calculated using two-tailed, homoscedastic t test. ∗∗∗p < 0.001.
Figure 6
Expression of LARP6 in Cancer Is Triggered by EMT and Acts to Enhance Protein Synthesis (A) Analysis of LARP6 expression in a panel of 33 human breast tumors by IHC. Three distinct patterns of LARP6 expression were detected among the tumor samples: “negative,” “weakly positive,” and “strongly positive.” Representative images for each category are shown. Scale bars, 50 μm. (B) LARP6 strongly positive tumors are significantly enriched among metaplastic carcinomas. Categorizing tumors based on their LARP6 IHC staining status as in (A) reveals a significant enrichment of LARP6 strongly positive tumors among metaplastic carcinomas (n = 7 out of 33). The p value was calculated using Fisher’s exact test. (C) Induction of EMT by human TGF-β1 upregulates LARP6. Left, morphology of MCF10AT cells following mock treatment or TGF-β1 (5 ng/mL) treatment for 7 days, reveals EMT induction. Scale bars, 50 μm. Right, immunoblot (IB) analysis of EMT markers (CDH1, ZEB1, and VIM) and LARP6, on the cells shown in left. GAPDH was used as loading control. (D) Quantification of changes in LARP6 and EMT marker proteins relative to GAPDH, from experiments shown in (C). IBs from 4 independent experiments as in (C) were quantified. Error bars are SD. p values were calculated using two-tailed, homoscedastic t test. ∗∗∗p < 0.001; ∗∗p < 0.01; ∗p < 0.05. (E) EMT enhances overall protein synthesis in a LARP6-dependent manner. MCF10AT parental and EMT pairs from (C) were treated with indicated siRNAs for 72 h before being subjected to OPP staining. (F) Quantification of OPP staining from experiments shown in (D). Normalized OPP averages were calculated from 7–11 field of view images from two independent experiments. Error bars are SD. p values were calculated using two-tailed, homoscedastic t test. n.s., non-significant; ∗p < 0.05.
Figure 7
Proposed Mechanism of Ribosome Biogenesis Regulation by LARP6-Dependent RP-mRNA Localization For a Figure360 author presentation of this figure, see
https://doi.org/10.1016/j.devcel.2020.10.006
. LARP6 binds RP-mRNAs and localizes them to the protrusive fronts of migrating mesenchymal-like cells, where their translation is enhanced due to the local enrichment of active translation machinery. Once translated, nascent RPs transport back to the nucleus to participate in ribosome biogenesis, leading to increased ribosome production and augmented overall protein synthesis.
References
- Bertrand E., Chartrand P., Schaefer M., Shenoy S.M., Singer R.H., Long R.M. Localization of ASH1 mRNA particles in living yeast. Mol. Cell. 1998;2:437–445. - PubMed
Publication types
MeSH terms
Substances
LinkOut - more resources
Full Text Sources
Other Literature Sources
Research Materials