Altered mRNA transport and local translation in iNeurons with RNA binding protein knockdown

Abstract

Neurons rely on mRNA transport and local translation to facilitate rapid protein synthesis in processes far from the cell body. These processes allow precise spatial and temporal control of translation and are mediated by RNA binding proteins (RBPs), including those known to be associated with neurodegenerative diseases. Here, we use proteomics, transcriptomics, and microscopy to investigate the impact of RBP knockdown on mRNA transport and local translation in iPSC-derived neurons. We find thousands of transcripts enriched in neurites and that many of these transcripts are locally translated, possibly due to the shorter length of transcripts in neurites. Loss of frontotemporal dementia/amyotrophic lateral sclerosis (FTD/ALS)-associated RBPs TDP-43 and hnRNPA1 lead to distinct alterations in the neuritic proteome and transcriptome. TDP-43 knockdown (KD) leads to increased neuritic mRNA and translation. In contrast, hnRNPA1 leads to increased neuritic mRNA, but not translation, and more moderate effects on local mRNA profiles, possibly due to compensation by hnRNPA3. These results highlight the crucial role of FTD/ALS-associated RBPs in mRNA transport and local translation in neurons and the importance of these processes in neuron health and disease.

Figure 1:

Boyden chambers enable collection of a pure neuritic fraction. A. Schematic of Boyden chambers. Chambers enable separation of neurites (bottom) from whole cell (top) via a thin microporous membrane. B. Images of neurons expressing cytosolic mScarlet in approximately 1% of cells cultured in Boyden chamber. Images were captured as a z-stack and maximum intensity projected. Scale bar 50 μm. C. Western blot validation of neuron fractionation using anti-ACTB (magenta) as a whole cell marker and anti-NeuN (green) as a nuclear marker. n=3, fractionated into 3 neurite samples and 3 whole cell samples. Molecular weight markers are indicated on the right. D. Volcano plot of WT proteomics reveal more whole cell enriched proteins (log₂(fold change) < 0, left) than those enriched in neurites. Dashed horizontal line represents adjusted p-value = 0.05. Vertical dashed lines identify proteins with log₂(fold change) of 1 or −1. n=6 neurite samples, 5 whole cell samples.

Figure 2:

Thousands of transcripts are enriched in neurites. A. MA plot of transcripts identified in WT neuron fractions. Points above the horizontal dashed line are genes that are enriched in neurites while those below are enriched in the whole cell fraction. Y-axis indicates log₂(fold change) of a transcript’s relative abundance between neurites and whole cell fractions. An FDR threshold of 0.05 and a fold change threshold of log₂(1.5) were applied. Red, blue, and gray represent an increase, decrease, or no change respectively of a gene’s relative abundance. Top 20 genes with largest fold change are indicated by name. n=4 samples each neurite and whole cell fractions. B. KEGG pathway analysis of transcripts localized to neurites reveals enrichment in terms related to neurodegeneration, including ALS, Parkinson’s disease, and Huntington’s disease. C. Representative images of HCR FISH of three highly neuritically enriched transcripts, ATXN2, RPS27A, and UBQLN1. These genes were picked as they were highly differentially localized and are highly expressed in iNeurons. FISH signal in green, anti-Tau staining in magenta, and nuclear stain (Hoechst) in blue. Scale bar 20 μm. D. Quantification of FISH in C, illustrating increased transcripts found in neurites compared to soma, as expected from transcriptomics in A. Small purple/blue points represent the value per image, n=8 per well. Large dots represent the average of each image per well, n=4. Spots identified in cell bodies (grown from Hoechst signal) were divided by the number of nuclei per image. Spots identified in neurites (Tau staining) were divided by the number of nuclei in that image to pseudo-normalize to cell number. P-values from two-way ANOVA with corrections by Šídák’s multiple comparisons test.

Figure 3:

Neuritic transcripts are shorter and more likely to use the most proximal UTR. A. Isoform switch consequences in WT neurites compared to whole cell fraction. Neurites are more likely to have shorter isoforms. IR=intron retention, NMD=nonsense mediated decay. B. GO Cellular Component of NMD sensitive genes from IsoformSwitchAnalzyzeR shows cytoplasmic related terms, while GO Biological Process shows protein and RNA metabolism related terms. C. Expression of genes that switch from NMD-sensitive in neurites to NMD-insensitive in whole cell fractions in UPF1 KD neurons shows that the neuritic NMD-sensitive isoforms have increased expression after UPF1 KD while whole cell NMD-insensitive isoforms do not change, demonstrating they are indeed sensitive to NMD. D. Plot of 3’ UTR usage. Neurites are more likely to use the most proximal UTR (pink) than a more distal UTR (black). E. Density plot of UTR usage. Neurites use more proximal UTRs rather than more distal UTRs. F. Transite analysis of transcripts enriched in neuritic fraction (LFC > 5) identifies RBM8A as the highest scoring RBP enriched in the 3’ UTR of those transcripts. Motif rank plot shows distribution of RBP binding motif enrichment score (y-axis) ordered by RBP motif enrichment rank (x-axis). Color indicates significance of enrichment. Only a subset of significantly enriched RBP binding motifs after multiple testing correction is shown (adjusted p-value ≤ 0.05). G. Transite analysis of transcripts enriched in the whole cell fraction (LFC < −5) identifies SRSF7 as the highest scoring RBP enriched in the 3’ UTR of those transcripts. Motif rank plot shows distribution of RBP binding motif enrichment score (y-axis) ordered by RBP motif enrichment rank (x-axis). Color indicates significance of enrichment. Only a subset of significantly enriched RBP binding motifs after multiple testing correction is shown (adjusted p-value ≤ 0.05). H. RNA binding dynamics of SRSF7 and RBM8A from https://chronology.rna.snu.ac.kr show that SRSF7 binds to mRNAs early in the lifespan of the transcript (peak at 42.8 minutes) while RBM8A binds later (peak at 77.3 minutes).

Figure 4:

Identification of newly synthesized proteins identifies many neuritically localized mRNAs. A. Schematic of QuanCAT experiment. Neurons are treated with a 2-hour pulse of the methionine analog AHA as well as heavy arginine and lysine. 10% of the resulting lysate was kept as an input fraction, while the remaining lysate was bound to alkyne agarose beads using click chemistry. Following binding, the unbound proteins were retained as a flow through fraction. Beads were stringently washed and proteins were trypsin digested on beads prior to mass spec. B. Rank plot of the proteins containing heavy amino acids identified in neurons. Neurite enriched transcripts (Figure 2A) are indicated in purple. ACTB is the protein with the highest intensity, while several mitochondrial and ribosomal proteins are also identified as newly synthesized. n=3. C. KEGG pathway analysis of proteins containing heavy amino acids identify ALS and Parkinson’s disease as the highest enriched pathways. D. Top Cellular Component terms enriched in proteins that incorporate heavy amino acids include cytoskeletal and axon-related terms. E. Schematic of pSILAC experiment. Neurons were grown in Boyden chambers and exposed to a two hour pulse of heavy arginine and lysine prior to chamber scraping and cell lysis. Proteins were trypsin digested and mass spec performed on both fractions. F. Rank plot of proteins containing heavy amino acids identified in neurites. Neurite enriched transcripts (Figure 2A) are indicated in purple. MRPS28, a mitochondrial ribosomal protein, is the most highly abundantly translated protein detected. n=4 neurite samples. G. KEGG pathway analysis of proteins containing heavy amino acids in neurites identifies pathways found in the transcriptomics GO analysis (Figure 2B), including pathways of neurodegeneration, as well as terms related to axon growth, like axon guidance.

Figure 5:

FTD/ALS-associated RBP KD causes moderate changes to the neuritic proteome, potentially modulated by compensation. A. Knockdowns are performed by lentiviral transduction of sgRNA into iPSCs expressing dCas9-KRAB, preventing transcription by binding to the promoter of the gene of interest. B. Volcano plot of TDP-43 KD vs NT proteins in the neurites shows a small number of proteins that are increased in neurites (right, orange) but a larger number decreased (left, blue), including known cryptic-exon containing genes STMN2 and UNC13A. Vertical dotted lines indicate a log₂(fold change) of 1 or −1, horizontal dotted line indicates a p value of 0.05. n=6 samples each condition. C. Volcano plot of hnRNPA1 KD vs NT proteins in the neurites shows many proteins with increased abundance in neurites (right, orange) and a smaller number decreased (left, blue). The related hnRNP, hnRNPA3 is increased. Vertical dotted lines indicate a log₂(fold change) of 1 or −1, horizontal dotted line indicates a p value of 0.05. n=5 NT, n=6 KD. D. Volcano plot of TDP-43 KD vs NT proteins in the whole cell fraction shows more proteins that are increased in whole cell compared to the neurite fraction (right, orange) but still a larger number of proteins with decreased abundance (left, blue), including known cryptic-exon containing gene STMN2. Vertical dotted lines indicate a log₂(fold change) of 1 or −1, horizontal dotted line indicates a p value of 0.05. n=5 samples each condition. E. Volcano plot of hnRNPA1 KD vs NT proteins in the whole cell fraction shows fewer proteins with increased abundance than was observed in neurites (right, orange) and a similarly small number of proteins with decreased abundance (left, blue). Vertical dotted lines indicate a log₂(fold change) of 1 or −1, horizontal dotted line indicates a p value of 0.05. n=6 samples each condition. F. Representative images of immunofluorescence (IF) of hnRNPA1 and hnRNPA3 levels in NT and KD neurons. sgRNA expression is visualized by co-expressed cytosolic BFP signal, immunofluorescence for the RBP of interest in green, and DRAQ5 nuclear stain in magenta. Scale bar 100 μm. G. Quantification of hnRNPA1 IF microscopy shows increased hnRNPA3 IF signal in hnRNPA1 KD neurons, suggesting compensatory changes in RBP levels upon knockdown. n=12 wells (large blue dots), 4 images per well (small gray dots). p-values from one-way ANOVA (p-value <0.0001) corrected by Šídák’s multiple comparisons test. H. Quantification of hnRNPA3 IF microscopy shows a slight increase in hnRNPA1 IF signal in hnRNPA3 KD neurons, suggesting compensatory changes in RBP levels upon knockdown. n=12 wells (large blue dots), 4 images per well (small gray dots). p-values from one-way ANOVA (p-value <0.0001) corrected by Šídák’s multiple comparisons test.

Figure 6:

RBP KD does not drastically alter mRNA localization but increases total neuritic mRNA. A. MA plot comparing TDP-43 KD to NT for the neuritic fraction. 98 genes show increased expression while 59 show decreased expression, including TDP-43 and STMN2. log₂(1.5) fold change cutoff. n=5 for each KD and NT. Genes that contain cryptic exons upon TDP-43 KD are highlighted in green. B. MA plot comparing hnRNPA1 KD to NT for the neuritic fraction. 25 genes show increased expression while 16 show decreased expression, including hnRNPA1. log₂(1.5) fold change cutoff. n=6 for each KD and NT. Genes that contain cryptic exons upon hnRNPA1 KD are highlighted in green. C. MA plot comparing TDP-43 KD to NT for the whole cell fraction. 179 genes show increased expression while 240 show decreased expression, including TDP-43 and STMN2. log₂(1.5) fold change cutoff. n=5 for each KD and NT. Genes that contain cryptic exons upon TDP-43 KD are highlighted in green. D. MA plot comparing hnRNPA1 KD to NT for the whole cell fraction. 21 genes show increased expression while 13 show decreased expression, including hnRNPA1. log₂(1.5) fold change cutoff. n=6 for each KD and NT. Genes that contain cryptic exons upon hnRNPA1 KD are highlighted in green. E. Representative images of polyA total mRNA FISH of NT and RBP KD neurons. mRNA is in magenta, anti-NEFL staining (axons) in green, and sgRNA in blue. Scale bar 20 μm. F. Quantification of polyA total mRNA FISH. TDP-43 and hnRNPA1 KD have increased neuritic mRNA compared to NT. Background signal is indicated by dotted line, from the non-fluorescent negative control. Neurites (NEFL) expressing sgRNA were used to mask FISH signal prior to intensity quantification. n=3 differentiations (replicates, means for each replicate indicated as large, outlined dots, with a line representing overall mean of condition), 12 wells per condition (mean of each well shown as a medium dot), 12 images per well (small gray dots). P-value from one-way ANOVA of replicate values (overall p < 0.0001), graph shows p-values corrected for multiple comparisons.

Figure 7:

RBP KD alters neuritic translation and neurite outgrowth. A. Schematic outlining fluorescent non-canonical amino acid tagging (FUNCAT). Cells are starved of methionine before adding methionine analog azidohomoalanine (AHA). AHA is incorporated into new proteins, which, after fixation, is detected by addition of a fluorophore through click chemistry. B. Representative FUNCAT images. sgRNA in blue, FUNCAT signal in red, nuclear stain (DRAQ5) in magenta. Scale bar 20 μm. C. Quantification of FUNCAT. TDP-43 has increased translation compared to NT. Background signal is indicated by dotted line (from methionine negative control, Met). Protein synthesis inhibitors anisomycin (Aniso) and cycloheximide (CHX) decrease FUNCAT signal. FUNCAT signal masked by sgRNA prior to intensity quantification. n=3 differentiations (replicates, means for each replicate indicated as large, outlined dots, overall mean indicated by line), 12 wells (medium dots), 12 images per well (small grey dots). p-value from one-way ANOVA of replicate values (overall p < 0.0001), graph shows significant p-value corrected for multiple comparisons. D. Representative micrographs of NT and RBP KD neurospheres 5 days after dox induction. Dashed lines indicate the extent of neurite outgrowth. sgRNA in blue, cytoplasmic mScarlet marker in magenta, and nuclear mNeonGreen in green. Scale bar 1 mm. E. Quantification of neurite diameter (red channel) divided by cell body diameter (green channel) indicates decreased neurite outgrowth in TDP-43 and hnRNPA1 KD. n ≥ 3 spheres per genotype per day. Significant days indicated by stars, **** p<0.0001, *** 0.001 < p < 0.0001, ** 0.01 < p <0.001. n=3–10 for NT, 10–13 for TDP-43 KD, and 9 or 10 for hnRNPA1 KD. Number of spheres vary because spheres were removed from analysis when they touched the edge of the well. P-values from mixed-effects analysis (rather than ANOVA due to the inconsistent number of spheres per day) with Dunnett’s multiple comparison test.