FMRP regulates neuronal RNA granules containing stalled ribosomes, not where ribosomes stall
- Department of Neurology and Neurosurgery, Montreal Neurological Institute, McGill University, Canada
- Department of Biochemistry, McGill University, Canada
- Goodman Cancer Institute, McGill University, Canada
- Canadian Centre for Computational Genomics, McGill University, Canada
- Department of Anatomy and Cell Biology, McGill University, Canada
- Centre de Recherche en Biologie Structurale, McGill University, Canada
Figures
Figure 1
Characterization of sucrose gradient sedimentation.
(A) A summary of the protocol for isolating the pellet and fraction 5/6 from mouse and rat whole brain homogenate using sucrose gradient fractionation. (B) SDS-PAGE stained with Coomassie Brilliant Blue showing the distribution of proteins from each fraction of the sucrose gradient from mouse and rat brains. Equal volumes of resuspended ethanol precipitates (fractions 1–10) and resuspended pellet were used. (C) Average UV absorbance from fractions 1–10 and pellet from WT (Red) and FMR1 KO (blue) preparations (error bars are S.D., N=3). (D) Representative electron microscopy of WT and FMR1 KO ribosome clusters from fraction 5/6 and pellet. Scale bar is 200 nm. (E) Representative immunoblots of fractions (as defined in A) for WT and FMR1 KO mice. For each blot, 1/100th of the starter, 1/10th of fraction 5/6, and 1/20th of the pellet fraction were loaded. (F) Quantification of differences in enrichment between WT and FMR1 KO. Levels of proteins were determined from scans of immunoblots (see ‘Materials and methods’). Enrichment is defined by the ratio of level of protein (see ‘Materials and methods’) between fraction 5/6 and pellet. For S6, the ratio of WT enrichment to FMR1 KO enrichment was calculated for each blot and the average calculated (N=8). For all the other RBPs, the WT and FMR1 KO enrichment were normalized to the S6 enrichment for that preparation, and the normalized values were used to calculate the difference in enrichment between WT and FMR1 KO (UPF1, n=8, Stau2, n=8, PurA, n=3, G3BP1, n=3, ZBP1, N=3, hnRNPA2/B1, N=3). Error bars are S.D. A one-sample t-test against 1 was used to test significance, with Bonferroni correction for multiple tests. No value reached p<0.05. All raw blots used in (E) and (F) can be found in Figure 1—source data 1 and all other blots in Figure 1—source data 2.
-
Figure 1—source data 1
PDF file containing Ponceau and immunoblots for Figure 1E, indicating the relevant bands and treatments.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig1-data1-v1.pdf
-
Figure 1—source data 2
Zipped file containing a subset of full blots (Ponceau and immunoblots) used in Figure 1E and all blots used to generate data (text and Figure 1F).
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig1-data2-v1.zip
Figure 2
Loss of FMRP does not affect anisomycin-puromycin competition.
(A) A summary of the protocol for isolating the pellet and treating with puromycin. (B) Representative immunoblot stained with antibodies to puromycin (anti-puro) (top) to showcase the inhibition of puromycylation (100 uM) by anisomycin (100 uM) in liver polyribosomes (left) or comparing the pellet from rat, mouse WT, and mouse FMR1 KO (right). Bottom: corresponding membrane stained with Ponceau before immunoblotting. The liver experiment was replicated twice with similar results. (C) Quantification of the amount of puromycylation resistant to anisomycin (puro + aniso/puro) in rat (N=3), WT mouse (N=4), and FMR1 KO (N=4). All groups are insignificant from each other (one-way ANOVA, p>0.05). All raw blots shown in (B) can be found in Figure 2—source data 1 and all other blots in Figure 2—source data 2.
-
Figure 2—source data 1
PDF file containing original Ponceahand immunoblots for Figure 2B, indicating the relevant bands and treatments.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig2-data1-v1.pdf
-
Figure 2—source data 2
Zipped file containing all blots (Ponceau and immunoblots) used in Figure 2B.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig2-data2-v1.zip
Figure 3 with 3 supplements
Higher nuclease reduces size of RPFs in WT pellet.
(A) A summary of the protocol for the RPF procedure. (B) Size distribution of normalized footprint reads from the WT pellet fraction under standard or low nuclease treatment. (C) Representative image for read coverage for WT pellet fraction either with standard or low nuclease treatment UTR, untranslated region; CDS, coding sequence for both RPFs and RNA-seq libraries. (D) Representative image for the number of read extremities (shading) for each read length (Y-axis) based on the distance from start (left) to stop (right) with the 5’ end (top) and 3’ end (bottom) for the WT pellet fraction with either standard or low nuclease treatment. (E) Representative image for the periodicity statistics for RPFs in different regions of the mRNA for standard or low nuclease treatment. Although the representative images above included only one replicate, similar results were observed across all three replicates.
-
Figure 3—source data 1
List of genes with high ratios of long reads/short reads in WT pellet RPFs from high nuclease-treated sample.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig3-data1-v1.xlsx
Figure 3—figure supplement 1
Higher nuclease reduces size of RPFs in FMR1 KO pellet.
(A) Size distribution of normalized footprint reads from the FMR1 KO pellet fraction under standard or low nuclease treatment. (B) Representative image for read coverage for FMR1 KO pellet fraction either with standard or low nuclease treatment. UTR, untranslated region; CDS, coding sequence. (C) Representative image for the number of read extremities (shading) for each read length (Y-axis) based on the distance from start (left) to stop (right) with the 5’ end (top) and 3’ end (bottom) for the FMR1 KO pellet fraction with either standard or low nuclease treatment. (D) Representative image for the periodicity statistics for each read coverage RPFs. Though the representative images above only included one replicate for the result, similar results were seen in all three replicates.
Figure 3—figure supplement 2
Higher nuclease reduces size of RPFs in FMR1 KO fraction 5/6.
(A) Size distribution of normalized footprint reads from the WT pellet fraction under standard or low nuclease treatment. (B) Representative image for read coverage for WT pellet fraction either with standard or low nuclease treatment. UTR, untranslated region; CDS, coding sequence. (C) Representative image for the number of read extremities (shading) for each read length (Y-axis) based on the distance from start (left) to stop (right) with the 5’ end (top) and 3’ end (bottom) for the WT fraction 5/6 fraction with either standard or low nuclease treatment. (D) Representative image for the periodicity statistics for each read coverage RPFs. Though the representative images above only included one replicate for the result, similar results were seen in all three replicates.
Figure 3—figure supplement 3
Higher nuclease reduces size of RPFs in WT fraction 5/6.
(A) Size distribution of normalized footprint reads from the WT fraction 5/6 fraction under standard or low nuclease treatment. (B) Representative image for read coverage for WT fraction 5/6 f raction either with standard or low nuclease treatment. UTR, untranslated region; CDS, coding sequence. (C) Representative image for the number of read extremities (shading) for each read length (Y-axis) based on the distance from start (left) to stop (right) with the 5’ end (top) and 3’ end (bottom) for the WT fraction 5/6 fraction with either standard or low nuclease treatment. (D) Representative image for the periodicity statistics for each read coverage RPFs. Though the representative images above only included one replicate for the result, similar results were seen in all three replicates.
Figure 4 with 2 supplements
High magnesium buffer does not affect ribosome structure.
Composite cryo-EM maps of class 1 (A) and class 2 (B) 80S ribosomes found in the pellet after purification in high magnesium buffer and RNase I treatment. The top panels show a side view of the two classes of ribosome particles contained in the sample. The bottom panels show top views of the same cryo-EM maps. The 40S and 60S subunits are shown as transparent densities to facilitate visualization of the positions of the tRNA molecules in each class.
-
Figure 4—source data 1
Data acquisition, reconstruction, and refinement parameters and data deposition codes for the cryo-EM dataset.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig4-data1-v1.docx
Figure 4—figure supplement 1
Single-particle analysis image processing workflow.
The cryo-EM dataset obtained for the pellet fraction after purification in high magnesium buffer and RNase I treatment was subjected to the image processing workflow in the figure. The diagram displays the main image processing steps undertaken with this dataset and the two main ribosome populations that were found as a result of the image classification approaches. The resolutions for the consensus maps and each one of the subunits in the maps obtained through local refinement are also indicated.
Figure 4—figure supplement 2
Resolution analysis of the two major classes of 80S ribosomes in the granule fraction purified under high magnesium conditions.
Consensus cryo-EM maps (left panels) were initially calculated for class 1 (A) and class 2 (B) from the RNase I-treated pellet fraction purified under high magnesium buffer conditions. These maps were subsequently refined by local refinement by dividing the 80S ribosome into two major bodies, the 60S and the 40S particles. The top panels in (A) and (B) show the local resolution analysis of the 80S consensus and composite maps obtained using local refinement. Composite maps were obtained by aligning them and merging the 60S and 40S cryo-EM maps obtained by local refinement using the command 'vop add' in Chimera. Maps are colored according to their local resolution using the color coding indicated in the scale bars. The Fourier shell correlation (FSC) curves for the consensus maps and each one of the subunits after the maps were subjected to local refinement are shown for classes 1 and 2. For each class, we show the following FSC plots: ‘No mask’: the raw FSC calculated between two independent half-maps reconstructed from the data and no masking applied. ‘Loose’: FSC calculated after applying a loose soft solvent mask to both half maps. The loose mask is calculated by thresholding the density map at 50% of the maximum density value. The resulting volume is dilated to create a soft mask. Voxels in the mask within 25 angstroms of the thresholded region receive a mask value of 1.0. Voxels between 25 and 40 angstroms fall off with a soft cosine edge, and voxels outside 40 angstroms receive a value of 0.0. ‘Tight’: FSC calculated after applying a tight soft solvent mask to both half maps. The tight mask is calculated by the same procedure as the loose mask, except that the dilation distances are 6 angstroms for the value 1.0 distance and 12 angstroms for the value 0.0 distance. ‘Corrected’: FSC curve calculated using the tight mask with correction by noise substitution as described in Chen et al., 2013. In brief, the two half maps have their phases randomized beyond a certain resolution, then the tight mask is applied to both, and an FSC is calculated. This FSC is used along with the original FSC before phase randomization to compute the corrected FSC. This accounts for correlation effects induced by masking. The resolution at which phase randomization begins is the resolution at which the no-mask FSC drops below the FSC = 0.143 criterion. We used a FSC threshold of 0.143 to report the overall resolution of the maps. The ‘Viewing direction distribution’ plot in panels (A) and (B) shows the orientation distribution of the particles contributing to the cryo-EM map for each one of the classes. These plots are 2D histograms that show the number of particles with a viewing direction at a particular elevation/azimuth bin. The number of particles can be inferred by the color code scale bar to the right of each plot.
Figure 5 with 2 supplements
Gene set enrichment analysis WT and FMR1 KO RNA-seq and RPFs from pellet fraction.
(A) Description of RPF isolation and mapping. Results of selected GSEA sets significantly affected by the loss of FMRP from analysis of (B) RNA-seq, (C) abundance, and (D) occupancy. Increases in FMR1 KO vs WT are to the left and decreases to the right. All GSEA used an N of 3 for WT and FMR1 KO samples.
-
Figure 5—source data 1
RNA-seq and RKPM for all samples for all genes.
Occupancy and enrichment calculations for WT and FMR1 KO samples.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig5-data1-v1.xlsx
-
Figure 5—source data 2
DEG analyses comparing WT and FMR1 KO.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig5-data2-v1.xlsx
-
Figure 5—source data 3
GO analysis of WT and FMR1 KO.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig5-data3-v1.xlsx
-
Figure 5—source data 4
GSEA of differences between WT and FMR1 KO, including percentage of mRNAs matched that are FMRP targets.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig5-data4-v1.xlsx
Figure 5—figure supplement 1
Principal component analysis (PCA).
(A) PCA of pellet RPFs from WT and FMR1 KO replicates. (B) PCA of pellet RPFs and RNA-seq samples from WT and FMR1 KO replicates.
Figure 5—figure supplement 2
Assessment of RPF abundance, occupancy, and enrichment in pellets of WT and FMRI- mice.
(A) Summary of protocol to generate RPF abundance, occupancy, and enrichment. (B) GO terms of the WT pellet (left) and FMR1 KO pellet (right) for abundance (top) and occupancy (bottom). For each graph, GO terms from the top 500 genes: biological function (top), cellular components (middle), and molecular function (bottom).
Figure 6
Comparison of putative stalled mRNAs vs total mRNAs in RPFs from the pellet of WT and FMR1 KO mice.
Comparison of mRNAs associated with ribosome resistant to initiation inhibitor mediated run-off (Shah et al., 2020) and FMRP-CLIPped mRNAs (Maurin et al., 2018; Darnell et al., 2011) to all other mRNAs. (A) WT pellet abundance, (B) FMR1 KO pellet abundance, (C) WT pellet occupancy, (D) FMR1 KO pellet occupancy, (E) WT enrichment, and (F) FMR1 KO enrichment. All analyses used an N of 3 for WT and FMR1 KO samples. As the N of the All mRNAs (13,079) were much larger than for the selected groups: Shah (185), Maurin (264), and Darnell (757) (Figure 6—source data 1) a random set from the total RNA group that matched the number in the selected group was generated for a two-tailed Welch’s t-test with Bonferroni correction for multiple t-tests. The median p-value from 10 random sets was used (Figure 6—source data 1). ****p<0.0001, ***p<0.001, **p<0.01, *p<0.05.
-
Figure 6—source data 1
Abundance, occupancy, and enrichment for WT and FMR1 KO samples matching the Shah, Maurin, and Darnell lists and subsets of these lists.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig6-data1-v1.xlsx
Figure 7 with 1 supplement
Comparison of putative stalled mRNAs between WT and FMR1 KO mice.
Comparison of mRNAs associated with ribosome resistance of initiation inhibitor run-off (Shah et al., 2020) and FMRP-CLIPped mRNAs (Maurin et al., 2018; Darnell et al., 2011) to all other mRNAs. The fold change between WT and FMR1 KO was calculated using DEG with an N of 3 for WT and FMR1 KO samples (see ‘Materials and methods’). These fold changes were then compared between the selected groups and all mRNas for (A) pellet abundance, (B) pellet occupancy, and (C) pellet enrichment. As the N of the All mRNAs (13,079) were much larger than for the selected groups: Shah (185), Maurin (264), and Darnell (757) (Figure 7—source data 1), a random set from the total RNA group that matched the number in the selected group was generated for a two-tailed Welch’s t-test with Bonferroni correction for multiple t-tests. The median p-value from 10 random sets was used (Figure 7—source data 1). ****, p<0.0001, ***p<0.001, **p<0.01, *p<0.05.
-
Figure 7—source data 1
Abundance, occupancy, and enrichment comparisons of the ratio between WT and FMR1 KO for mRNAs matching the Shah, Maurin, and Darnell Lists and subsets of these lists.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig7-data1-v1.xlsx
Figure 7—figure supplement 1
Occupancy analysis with subsets of FMRP targets and RNAs resistant to run-off.
The mRNAs found to be resistant to ribosome run-off (Shah et al., 2020) were divided into two sets consisting of mRNAs that are also FMRP targets (identified in Maurin et al., 2018 or Darnell et al., 2011) or not. The FMRP targets were divided into two sets consisting of mRNAs identified as one of the 200 most abundant mRNAs resistant to run-off (Shah et al., 2020) or not. Box and Whisker plots are shown with line for mean. Different from the total RNA group using two-tailed Welch’s t-test with Bonferroni correction for multiple t-tests. Differences between the two sets was evaluated with a two-tailed Welch’s t-test. (A) Occupancy in WT pellet. (B) Occupancy in FMR1 KO pellet. (C) Difference in occupancy between WT and FMR1 KO pellet. The Ns (number of mRNAs) used for all the comparisons can be found in Figure 7—source data 1.
Figure 8
Comparison of RPF peaks in the pellet of WT and FMR1 KO mice.
(A) Representation of how peaks of RPFs are selected. (B) Table of the number of peaks between replicates of WT pellet (N=3), FMR1 KO pellet (N=3), and combined (N=6), the percentage of peaks with FXS related motif and the enrichment of aspartate (Asp) and glutamate (Glu) in the peaks compared to non-peaks (Figure 8—source data 2). (C) RPF coverage of Tubb2b for the three replicates of WT and FMR1 KO. Shaded region is the open reading frame. Asterisks indicate consensus peaks (seen in all six samples with peaks within 6 bp).
-
Figure 8—source data 1
List of all peaks as well as motif counts.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig8-data1-v1.xlsx
-
Figure 8—source data 2
Amino acid enrichment of peaks.
- https://cdn.elifesciences.org/articles/106692/elife-106692-fig8-data2-v1.xlsx
Figure 9 with 1 supplement
RPM of hippocampal cultures derived from WT and FMR1 KO mice.
(A) Summary of the protocol for puromycylation HHT-Runoff and DHPG Reactivation on WT and FMR1 KO hippocampal culture. (B) Representative confocal images for puromycylated ribosomes with or without HHT runoff and DHPG reactivation. Circle denotes puromycin puncta. No visible staining was seen in the absence of puromycin. Scale bar shown below. (C) Quantification of RPM puncta density. Ns are neurites/cultures. WT (42/5); WT DHPG (54/5), FMR1 KO (41/4), and FMR1 KO DHPG (25/3). One-way ANOVA F(3,158) = 5.32, p<0.005 *, p<0.05 post-Hoc Tukey HSD test. Box and Whisker plot with line representing the median. (D) Quantification of size of RPM puncta. WT 189/5; WT DHPG 171/5, FMR1 KO (118/4), and FMR1 KO DHPG (48/3). Box and Whisker plot with line representing the median. One-way ANOVA showed no significance (p>0.05).
Figure 9—figure supplement 1
Effect of anisomycin and HHT on RPM of hippocampal cultures derived from WT and FMR1 KO mice.
(A) Representative confocal images for puromycylated ribosomes treated either with puromycin and anisomycin together or pre-treated with HHT for 15 min before treating with puromycin. Circle denotes puromycin puncta. No visible staining was seen in the absence of puromycin. Quantification of RPM puncta density of puncta >50 microns from the cell body compared to WT (see Figure 9) shown as box and whisker plots. There is no effect of adding anisomycin with puromycin or preincubation with HHT on the puncta density (B, D) or puncta size (C, E) in WT (B, C), or FMR1 KO (D, E). WT numbers are the same as Figure 9. Numbers are neurites/cultures. Puncta density WT (42/5); WT anisomycin (A) (27/4); WT homoharringtonine (H) (36/5), FMR1 KO (41/4), FMR1 KO anisomycin (A) (23/3), FMRP KO homoharringtonine (45,5). One-way ANOVA showed no significance for density or size (p>0.05). (F) For clarity, the data is also presented as mean ±S.D.
Tables
Key resources table
| Reagent type (species) or resource | Designation | Source or reference | Identifiers | Additional information |
|---|---|---|---|---|
| Strain, strain background (mouse) | FMR1-/y | Jackson Laboratories, 003025 | Jackson Laboratories, 003025 | |
| Strain, strain background (mouse) | C57BL/6J | Jackson Laboratories, 000664 | Jackson Laboratories, 000664 | |
| Strain, strain background (rat) | Sprague–Dawley | Charles River Laboratories | RRID:MGI:5651135 | |
| Antibody | Anti-S6 (rabbit monoclonal) | Cell Signaling Technology #2217 | RRID:AB_331355 | IB (1:10,000) |
| Antibody | Anti-FMRP (rabbit polyclonal) | Cell Signaling Technology #4317 | RRID:AB_1903978 | IB (1:500) |
| Antibody | Anti-UPF1 (rabbit monoclonal) | Epitomics EP4682 | Abcam:ab133564 | IB (1:10,000) |
| Antibody | Anti-Stau2 (mouse monoclonal) | PMID:12140260 | MediMabs:MM0037-P | IB (1:1000) |
| Antibody | Anti-EEF2 (rabbit polyclonal) | Cell Signaling Technology #2332 | RRID:AB_10693546 | IB (1:1000) |
| Antibody | Anti-Pur-alpha (rabbit polyclonal) | Abcam #ab79936 | RRID:AB_2253242 | IB (1:000) |
| Antibody | Anti-ZBP1 (rabbit polyclonal) | Novus Biologicals #NBP1-76854 | RRID:AB_11018813 | IB(1:500) |
| Antibody | Anti-TIA1 (rabbit polyclonal) | ProteinTech #12133-2-AP | RRID:AB_2201427 | IB (1:1000) |
| Antibody | Anti-hnRNPA2B1 (mouse monoclonal) | Novus Biologicals #NB120-6102 | RRID:AB_790226 | IB 1:1000 |
| Antibody | Anti-ribophorin (mouse monoclonal) | Santa Cruz Biotechnology sc-48367 | RRID:AB_628221 | IB (1:200) |
| Antibody | Anti-puromycin (mouse monoclonal) | DSHB #PMY-2A4 | RRID:AB_2619605 | IB (1:1000) IF (1:1000) |
| Antibody | HRP-conjugated anti-rabbit (goat polyclonal) | Thermo Fisher Scientific #31460 | RRID:AB_228341 | IB (1:10,000) |
| Antibody | HRP-conjugated anti-mouse (goat polyclonal) | Thermo Fisher Scientific #31430 | RRID:AB_228307 | IB (1:10,000) |
| Antibody | Alexa Fluor 750-conjugated anti-mouse (goat polyclonal) | Thermo Fisher Scientific #A-21037 | RRID:AB_2535708 | IF (1:1000) |
| Peptide, recombinant protein | RNAse I | Epicentre | Epicentre:N6901K | |
| Peptide, recombinant protein | RNAse I | Thermo Fisher | Thermo Fisher Scientific:AM2294 | |
| Peptide, recombinant protein | SuperaseIN | Invitrogen | Invitrogen:AM2969 | |
| Software, algorithm | SPSS | SPSS | RRID:SCR_002865 | |
| Software, algorithm | ImageJ | Version 1.53A | RRID:SCR_003070 | |
| Software, algorithm | Cutadapt | Version 2.1 | RRID:SCR_011841 | |
| Software, algorithm | Bowtie2 | Version 2.3.5 | RRID:SCR_016368 | |
| Software, algorithm | Star Algner | Version 2.7.8a | RRID:SCR_004463 | |
| Software, algorithm | Picard | Version 2.26.6 | RRID:SCR_006525 | |
| Software, algorithm | SamTools | Version 1.19.2 | RRID:SCR_002105 | |
| Software, algorithm | edgeR | PMID:34557778 | RRID:SCR_012802 | |
| Software, algorithm | Bioconductor | PMID:34557778 | RRID:SCR_006442 | |
| Software, algorithm | cryoSPARC | Version 4.5 | RRID:SCR_016501 | |
| Software, algorithm | UCSF ChimeraX | PMID:32881101 | RRID:SCR_015872 | |
| Software, algorithm | Meme Suite | FIMO | RRID:SCR_001783 |
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
FMRP regulates neuronal RNA granules containing stalled ribosomes, not where ribosomes stall
eLife 14:RP106692.
https://doi.org/10.7554/eLife.106692.3
