Impact of WINi on the transcriptome of MLLr cancer cells.

(A) Chemical structures of C6 and C16. (B) Crystal structures of C6 or C16 bound to the WIN Site of WDR5 with electrostatic surfaces mapped [PDB IDs: 6E23 (Aho et al., 2019a); 6UCS (Tian et al., 2020)]. The image shows a close-up view of the WIN site. (C) Superimposed WIN site-binding conformations of C6 (green) and C16 (blue). (D) Transcript levels as determined by QuantiGene™ analysis of representative WDR5-bound (color) or non-bound (grayscale) ribosomal protein genes in MV4;11 cells treated with a serial dilution range of either C6 (left) or C16 (right) and relative to DMSO-treated cells (n = 2-3; Mean ±SEM). Vertical dashed line indicates either 2 µM C6 (left) or 100 nM C16 (right). (E) Number of genes with significantly (FDR < 0.05) altered transcript levels following treatment of MV4;11 cells with C6 (2 µM) or C16 (100 nM) for 48 hours, as determined by RNA-Seq (n = 3). See Figure 1—source data 1 for complete output of RNA-Seq analysis. (F) Comparison of gene expression changes elicited by C6 (x-axis) and C16 (y-axis), represented as Log2 fold change (FC) compared to DMSO. WDR5-bound genes are colored red. Locations of RPL22L1 and ZMAT3 are indicated. (G) Overlap of genes with decreased (left) or increased (right) transcript levels in MV4;11 cells treated with C6 or C16. (H) Gene Set Enrichment Analysis (GSEA) showing the distribution of genes suppressed in MV4;11 cells in response to C6 (left) or C16 (right) against the list of all genes bound by WDR5 in those cells (Aho et al., 2019a). NES = normalized enrichment score. (I) Enrichment analysis of genes suppressed (left) or induced (right) by C6 or C16 in MV4;11 cells. KEGG and Hallmark.MSigDB pathways are shown. Fold enrichment of indicated pathways is presented on the x-axis, the number of genes is shown in italics in each bar, and colors represent - Log10 FDR. See Figure 1—source data 2 for additional GSEA (Hallmark) and ORA (Hallmark) analyses of differentially expressed genes. (J) Transcript level changes in WDR5-bound (left) and non-bound (right) RPGs elicited by C6 (top) or C16 (bottom).

Impact of WINi on the translatome of MLLr cancer cells.

(A) Volcano plots depicting alterations in translation efficiency (TE) induced by 48 hour treatment of MV4;11 cells with either 2 µM C6 (left) or 100 nM C16 (right) compared to DMSO (n = 2; Red indicates FDR < 0.05 and Log2 FC > 0.25), as determined by Ribo-Seq. (B) Number of mRNAs with significantly (FDR < 0.05 and Log2 FC > 0.25) altered TE levels following treatment of MV4;11 cells with C6 (2 µM) or C16 (100 nM) for 48 hours. See Figure 2—source data 1 for complete output of Ribo-Seq analysis. (C) Overlap of mRNAs with significantly decreased TE in response to C6- or C16-treatment. (D) Translation efficiencies (TE) of mRNAs in DMSO-treated MV4;11 cells plotted against translation efficiencies of mRNAs in cells treated with either C6 (left) or C16 (right). Red indicates mRNAs with significantly altered translation efficiencies following inhibitor treatment (FDR < 0.05 and Log2 FC > 0.25). (E) Numbers of differentially-translated mRNAs (ΔTE) in each quartile of genes (stratified by TE in DMSO) in cells treated with C6 (left) or C16 (right). (F) Enrichment analysis of common mRNAs suppressed by C6/C16 at the mRNA (blue) and translational (red; TE) level in MV4;11 cells. Hallmark.MSigDB pathways are shown. The x-axis indicates the number of suppressed genes in each category; the italic numbers are the corresponding FDR. See Figure 2—source data 2 for the full Hallmark.MSigDB analysis, as well as for Reactome and KEGG pathways. (G) Enrichment analysis of mRNAs suppressed translationally by C6/C16 but with no significant changes in mRNA levels. Gene Ontology (GO) Biological Process (BP) and Molecular Function (MF) categories are shown, as well as KEGG pathways. The x-axis displays -Log10 FDR; the number of mRNAs is shown in italics in each bar. See Figure 2—source data 3 for extended enrichment analyses, broken down by TE and mRNA direction changes. (H) TE changes in WDR5-bound (left) and non-bound (right) RPGs elicited by C6 (top) or C16 (bottom).

Impact of WINi on the ribosome inventory of MLLr cancer cells.

(A) Lysates from MV4;11 cells treated 24 or 72 hours with either 0.1% DMSO or 250 nM C16 were subjected to liquid chromatography coupled with tandem mass spectrometry and analyzed by label-free quantification (LFQMS). The table shows the number of proteins detected in DMSO and C16 samples and those with significantly altered levels at each time point (n = 4; adj. p-value < 0.05). See Figure 3—source data 1 for complete output of LFQMS analysis. (B) Volcano plot, showing protein level alterations in cells treated with C16 for 24 hours (red indicates adj. p-value < 0.05). The location of RPL22L1 is indicated. (C) As in (B) but for 72 hour treatment with C16. (D) Overlap of proteins significantly increased (top) or decreased (bottom) following 24- or 72-hour C16-treatment. (E) Protein level alterations induced by C16 in consensus p53 target proteins (Fischer, 2017) at the 24 and 72 hour treatment timepoints. Those proteins only altered in abundance at 24 hours are represented as blue dots; proteins only altered at 72 hours are red; proteins altered at both timepoints are grey. (F) As in (E) but for ribosomal proteins. (G) Changes in expression of proteins encoded by WDR5-bound (left) and non-bound (right) RPGs elicited by 24 (top) or 72 (bottom) hour treatment with C16. Note that, due to the magnitude of change, Log2(FC) for RPL22L1 is presented on a separate scale.

A two tier loss of function screen for modulators of the response to WINi.

(A) Two-tier screen design. In the first tier, Cas9-expressing MV4;11 cells were transduced with a genome-wide sgRNA library and treated with 2 µM C6 until a resistant cell population emerged. sgRNA representation in the pre-treatment population was compared to the post-treatment population. In the second tier, cells were transduced with a custom library of distinct sgRNAs targeting non-pan-essential "hits" from the first tier, cultured in the presence of DMSO, C6, or C16, and sgRNA representation in C6/C16-treated cultures compared to that from DMSO-treated cultures. Created with BioRender.com. (B) Volcano plot, showing gene-level changes in sgRNA representation from the first tier (orange indicates FDR < 0.05). Datapoints corresponding to TP53, RPL22, and CDKN2A are indicated. See Figure 4—source data 1 for full output of the Tier 1 screen. (C) Comparison of gene-level changes in sgRNA representation in C6- and C16-treated populations in the second tier screen, each compared to DMSO-treated populations (red indicates FDR < 0.05; black indicates non-targeting control sgRNAs). See Figure 4—source data 2 for full output of the Tier 2 screen. (D) Top: Overlap of genes from the Tier 2 screen with enriched (left) or depleted (right) sgRNAs in C6- and C16-treated MV4;11 populations, compared to the DMSO control. Bottom: Overlap of genes with enriched (left) or depleted (right) sgRNAs in the first versus second tiers of the screen. "Tier 1" contains only those genes targeted in the Tier 2 screen. "Tier 2" contains the intersection of genes with altered sgRNAs in both the C6 and C16 treatments. (E) Ranked heatmap, representing the mean gene-level Log2 fold change (FC) of sgRNAs from the C6 and C16 treatments in the Tier 2 screen, as well as gene enrichment analysis outputs. Note that "Signal transduction by p53 class mediator" is a GO:BP term (orange); "p53" assignments (yellow) were added by manual curation.

© 2024, BioRender Inc. Any parts of this image created with BioRender are not made available under the same license as the Reviewed Preprint, and are © 2024, BioRender Inc.

Identification of agents that synergize with WINi in MLLr cells.

(A) Peak synergy (> 0) and antagonism (< 0) ZIP Delta (δ) scores from synergy assays in which MV4;11 cells were treated for three days with 49 unique dose combinations of C16 and the indicated compound of interest (n = 4). See Figure 5—source data 1 for numerical ZIP Delta analysis output. (B) Heatmaps of MV4;11 cell growth inhibition at each dose of C16 and the indicated six compounds. The remaining five combinations tested are shown in Figure 5—figure supplement 1. (C) As in (A) but for MOLM13 cells. See Figure 5—source data 1 for numerical ZIP Delta analysis output. (D) As in (B) but for MOLM13 cells. The remaining five combinations tested are shown in Figure 5—figure supplement 2. (E) Number of genes with significantly (FDR < 0.05) altered transcript levels following treatment of MV4;11 cells with C16 (100 nM), mivebresib (Mibv; 2.5 nM), or the combination for 48 hours, as determined by RNA-Seq (n = 3). See Figure 5—source data 2 for complete output of RNA-Seq analysis. (F) UpSet plot, showing the overlap of genes suppressed (left) or induced (right) in response to C16, mivebresib, or the combination. (G) UpSet plot, showing the breakdown of Reactome "Translation" pathway genes suppressed in response to C16, mivebresib, or the combination. (H) Enrichment of Reactome Pathways in genes with increased transcripts following treatment of MV4;11 cells with C16, mivebresib, or the combination. See Figure 5—source data 3 for complete output of enrichment analyses.

WINi inactivate MDM4 in an RPL22-dependent manner.

(A) Differential alternative splicing events affected by C6/C16 treatment of MV4;11 cells were quantified by rMATS. The types of alternative splicing events are cartooned at left, and the number of significantly different events (> 5% Δψ; FDR < 0.05) common to C6/C16 depicted in the graph. See Figure 6—source data 2 for output of rMATS analysis. (B) Sashimi plot quantifying read junctions that span exons 5–7 of MDM4 in MV4;11 cells treated with DMSO (green) or C16 (blue). Numbers in the arcs display junction depth. The location of exons 5, 6, and 7 is depicted at the bottom; skipped exon 6 is highlighted in orange. (C) Viabilities of control (non-targeting: NT) and RPL22 knock out (KO) MV4;11, MOLM13, and K562 cells treated with a serial dilution range of C16 for 72 hours, relative to viability of DMSO-treated cells (n = 3; Mean ±SEM). (D) Western blot analysis of p53 levels in control (NT) and RPL22 knockout (KO) MV4;11 and MOLM13 cells treated with either 0.1% DMSO or C16 (MV4;11, 200 nM; MOLM13, 400 nM) for 72 hours. α-Actinin is loading control. Representative images from three biological replicates shown. Raw unprocessed gel images are presented in Figure 6—source data 5. (E) Heatmap, showing significant changes in the expression of consensus p53 target genes (Fischer, 2017) between the indicated pairwise comparisons of RNA-Seq datasets. Note that only consensus p53 target genes altered in expression by C16 in control (NT) cells are represented. (F) Sashimi plot quantifying read junctions that span exons 5–7 of MDM4 in RPL22KO MV4;11 cells treated with DMSO or C16. Numbers in the arcs display junction depth. The location of exons 5, 6, and 7 is depicted at the bottom; skipped exon 6 is highlighted in orange. Corresponding NT images are presented alongside RPL22KO images in Figure 6—figure supplement 3B). (G) Western blots, comparing the effects of 72 hours of DMSO (DM) or C16 treatment (MV4;11, 200 nM; MOLM13, 400 nM) of control (NT) or RPL22 knockout (KO) MV4;11 (left) or MOLM13 (right) cells on levels of MDM4, p21, RPL22L1, RPL22, and GAPDH (loading control). Representative images from three biological replicates are shown. Raw unprocessed gel images are presented in Figure 6—source data 9.

Transcript changes elicited by WINi in MLLr cancer cells.

(A) Crystal structures of C6 or C16 bound to the WIN Site of WDR5 with electrostatic surfaces mapped [PDB IDs: 6E23 (Aho et al., 2019a); 6UCS (Tian et al., 2020)]. (B) Viabilities of MV4;11 cells treated with a serial dilution range of either C6 (left) or C16 (right) for 72 hours, relative to viability of DMSO-treated cells (n = 3; Mean ±SEM). (C) As in (B) but for MOLM13 cells. (D) Transcript levels as determined by QuantiGene™ analysis of representative WDR5-bound (color) or non-bound (grayscale) ribosomal protein genes in MOLM13 cells treated with a serial dilution range of either C6 (left) or C16 (right) and relative to DMSO-treated cells (n = 3; Mean ±SEM). Vertical dashed line indicates either 2 µM C6 (left) or 100 nM C16 (right). (E) Transformed z-scores of genes with significantly altered transcript levels (RNA-Seq) in MV4;11 cells treated with either C6 (2 µM) or C16 (100 nM) for 48 hours, compared to DMSO-treatment. (F) Volcano plots, showing transcript level alterations in MV4;11 cells treated 48 hours with 2 µM C6 (left) or 100 nM C16 (right) compared to DMSO (n = 3; red indicates FDR < 0.05). (G) Dispersion plot describing the variance in gene expression for the RNA-seq data in a previous study (left) and this study (right).

Impact of WINi on RPL22L1 and p53 target gene expression.

(A) Venn diagram, showing the overlap of consensus p53 target genes (Fischer, 2017) with genes significantly induced by C6 or C16 in MV4;11 cells. (B) Graph showing the change in expression of the 91 common genes in (A) elicited by WIN site inhibitor (WINi) C6 (red) or C16 (blue) in MV4;11 cells, compared to DMSO. (C) Changes in expression (and FDR) of RPL22L1 elicited in response to C6 (red) or C16 (blue) treatment of K562 leukemia cells (Aho et al., 2019a) or five rhabdoid tumor cell lines [TTC642, KYM-1, G401, TM87-16, and TTC549; (Florian et al., 2022)].

WINi suppress bulk protein synthesis.

(A) Representative histograms from protein synthesis assays in MV4;11 cells treated 24, 48, or 96 hours with either 0.1% DMSO (blue), 2 µM C6 (red), or 100 nM C16 (orange). Cells were pulsed with O-propargyl-puromycin (OPP) to label nascent proteins, Alexa Fluor 647 linked to incorporated OPP in Click chemistry reactions, and fluorescence measured by flow cytometry analysis. MV4;11 cells treated 30 minutes with 100 µg/mL cycloheximide (“CHX”; green) serve as a positive control for inhibited protein synthesis. MV4;11 cells pulsed with DMSO (“No OPP”; black) serve as a control for background fluorescence. (B) Quantification of protein synthesis assays. Fluorescence from CHX-treated cells was set as the baseline, and fluorescence presented relative to DMSO-treated (DM) cells at each time point (n = 3; Normalized Geometric Mean ±SEM). P-values are represented by asterisks: ‘*’ < 0.05, ‘**’ < 0.01, ‘***’ < 0.001.

WINi suppress translation.

(A) Distribution of ribosome protected fragment (RPF) lengths in each Ribo-Seq sample/replicate. The length distribution of RPFs in mammalian Ribo-Seq experiments typically peaks at 30–31 nucleotides. (B) Proportion of RPFs mapping to the coding sequence (CDS) or 5’ or 3’ untranslated regions (UTR) of transcripts. Color of dots is the same as in (A). (C) Proportion of RPFs mapping to each reading frame in the 5’ UTR (left), the CDS (middle), and the 3’ UTR (right). Color of dots is the same as in (A). (D) Magnitudes of significant TE alterations of mRNAs in each quartile (stratified by TE in DMSO) in cells treated with C6 (left) or C16 (right). Color dots represent individual genes. Bottom, middle, and top horizontal lines of each box represent first, second, and third quartiles, respectively. Vertical lines extend to data points within 1.5-fold of the interquartile range. Black dots represent values beyond 1.5-fold of the inter-quartile range. (E) Changes in TE induced by C6 (left) or C16 (right) in MV4;11 cells, binned according to mRNA TOPscores (Philippe et al., 2020). Dashed lines represent the median; dotted lines indicate quartiles. Significance by t-test is indicated compared to group with TOPscore 0 to 1 (’*’ ≤ 0.05, ‘**’ ≤ 0.0001). (F) UpSet plot, showing the breakdown of genes encoding PRMT5 substrates (Radzisheuskaya et al., 2019) whose transcript levels and/or translation efficiencies decrease following WIN site inhibition (p-value calculated by hypergeometric test for over-representation of genes encoding PRMT5 substrates in genes with decreased translation efficiencies). (G) Overlap of C6/C16 common mRNAs with decreased abundance (RNA; blue) and those with decreased translation efficiency (TE; salmon), grouped according to the indicated Hallmark.MSigDB categories. (H) Overlap of all C6/C16 common mRNAs with altered abundance and decreased TE. (I) The top row of the heatmap displays the codon stability coefficient (CSC) for each codon (Wu et al., 2019) ranked from lowest (’Non-optimal’) to highest (’Optimal’). The middle row displays enrichment of each codon in mRNAs that are decreased at both the TE and mRNA levels (RNA+TE) versus those that show a decrease in TE without an accompanying decreased in mRNA abundance (TE only). Bottom row is -Log10 FDR.

WINi impair translation of mitochondrial ribosomal proteins.

(A) Top: Transcript level changes in mitochondrial ribosomal protein genes elicited by C6 or C16, as indicated. Bottom: Translational efficiency (TE) changes in mitochondrial ribosomal protein genes elicited by C6 or C16. All of the mitochondrial RPGs are nuclear-encoded; none have detectable binding of WDR5.

Distribution of peptide/protein intensities in LFQMS analysis.

(A) Peptide intensities of all proteins detected in each mass spectrometry run before (left) and after normalization (right). (B) Magnitudes of significant protein level alterations within each decile (stratified by protein intensity in DMSO samples) in MV4;11 cells treated 24 hours with 250 nM C16. Red points represent ribosome proteins. (C) Box plot representation of data presented in (B). Bottom, middle, and top horizontal line of each box represents first, second, and third quartiles, respectively. Vertical lines extend to data points within 1.5-fold of the interquartile range. Black dots represent values beyond 1.5- fold of the interquartile range. (D) Number of proteins increased or decreased within each decile.

Enrichment analysis of proteins with altered expression in response to C16 treatment.

(A) Graphs showing enrichment of proteins in GO BP (top) and Hallmark.MSigDB (bottom) pathways that are induced by C16 treatment at 24 (blue) or 72 (green) hours. The x-axis displays -Log10 (FDR); the number of proteins in each category is given in italics. (B) Graphs showing enrichment of proteins in GO BP (top) and Hallmark.MSigDB (bottom) pathways that are suppressed by C16 treatment at 24 (red) or 72 (brown) hours. The x-axis displays -Log10 (FDR); the number of proteins in each category is given in italics. See Figure 3—source data 2 for output of the full enrichment analyses.

WIN site inhibitors suppress rRNA levels.

(A) In-gel fluorescence assay detecting metabolically labeled rRNA (top) isolated from MV4;11 cells treated 24, 48, or 96 hours with either DMSO (0.1%), C6 (2 µM), or C16 (100 nM), and pulsed with 2’-azido-2’-cytidine (AzCyd). As a positive control for inhibited rRNA synthesis, MV4;11 cells were treated one hour with 5 nM actinomycin D (“ActD”). As a control for background labeling, MV4;11 cells were pulsed with DMSO (“No AzCyd”). Fluorescent probes were covalently linked to incorporated AzCyd in Click chemistry reactions. Total RNA (bottom) was detected by SYBR stain. (B) Quantification of metabolic rRNA labeling (n = 3; Mean ± SEM). P-values are indicated. Raw unprocessed gel images are presented in Figure 3—source data 3.

C16 induces redistribution of nucleophosmin from the nucleolus to the nucleoplasm.

(A) Representative immunofluorescent images of MV4;11 cells treated with Vehicle (DMSO control), C16 (100 nM), or ActD (5 nM) for the times indicated and stained for Nucleophosmin (NPM1, green), gH2A.X pSer139 (Double-stranded break marker, magenta) and Hoechst (blue). DNA damage arises upon cell death following drug treatment. Scale bars are 5 μm. (B) Quantification of the ratio of nucleolar to total NPM1 in the cells described in A. C6, C16, and ActD treatment disrupted NPM1 localization. P-values are represented by asterisks: ‘**’ =0.0012, ‘****’ <0.0001.

Genome wide CRISPR screen identifies genes that influence response to C6/C16.

(A) Tier 1 screen: Daily cell counts of MV4;11 Cas9 and MV4;11 Cas9 + GeCKOv2 (Library) populations treated with either DMSO or 2 µM C6. The two replicates of this screen are shown separately. (B) Normalized counts of each sgRNA (x-axis) in the GeCKOv.2 library targeting TP53 in the initial transduced cells (red; not visible on this scale) and the C6-treated population (blue). Data represents means of replicates; ‘*’ indicates FDR < 0.05. (C) As in (B) but for sgRNAs targeting CDKN2A. (D) Schematic of the CDKN2A gene locus with indicated sites complementary to Tier 1 and Tier 2 screen sgRNAs. Red sgRNAs increase in representation in CRISPR screens. (E) miRNet 2.0 (Chang and Xia, 2023) analysis of the 27 miRNAs enriched in the Tier 1 screen produced a single significant hit corresponding to the KEGG p53 signaling pathway. The miRNAs are represented as blue boxes and target genes as red circles; the connections between them are indicated. (F) As in (B) but for sgRNAs targeting RPL22. (G) Volcano plots, showing gene-level changes from the Tier 2 screen in sgRNA representation in C6- (left) and C16- (right) treated populations compared to DMSO control cultures (orange indicates FDR < 0.05). (H) Graph depicting gene-level Log2 FC and FDR values for genes that were flagged as C6- (squares) or C16- (circles) specific in the Tier 2 screen. (I) GO enrichment analysis of the 57 C6/C16 common genes emerging from Tier 2 of the screen. Italics represent the number of genes in each category.

C16 is synergistic with multiple agents in MV4;11 cells.

(A) Heatmaps of MV4;11 cell growth inhibition at each dose of C16 and the indicated five compounds. (B) Heatmaps of δ scores from MV4;11 cells at each dose combination of C16 and the indicated agents.

C16 is synergistic with multiple agents in MOLM13 cells.

(A) Heatmaps of MOLM13 cell growth inhibition at each dose of C16 and the indicated five compounds. (B) Heatmaps of δ scores from MOLM13 cells at each dose combination of C16 and the indicated agents.

Impact of C16 and mivebresib on RPG and p53 target gene expression.

(A) Transcript level changes in WDR5-bound (left) and non-bound (right) RPGs elicited by C16 (top), mivebresib (Mivb; middle), or the combination (bottom). (B) Heatmap, showing significant changes in the expression of consensus p53 target genes (Fischer, 2017) induced by C16, mivebresib (Mivb; middle) or the combination (bottom) in MV4;11 cells.

WINi alter the abundance of alternatively-spliced mRNA isoforms.

(A) Western blots comparing the effects of 72 hour DMSO (DM) or C16 treatment of MV4;11 (top) or MOLM13 (bottom) cells on levels of p53 and GAPDH (loading control). Representative images from three biological replicates are shown. Raw unprocessed gel images are presented in Figure 6—source data 1. (B) Differential alternative splicing events impacted by C6 (red) or C16 (blue) treatment of MV4;11 cells were quantified by rMATS. The number of significantly different events (> 5% Δψ; FDR < 0.05) for each WIN site inhibitor are depicted in the graph. "RI" is retained intron; "MEX" is mutually exclusive exons; "A3SS" is alternative 3’ splice site; "A5SS" is alternative 5’ splice site; "SE" is skipped exon. See Figure 6—source data 2 for output of rMATS analysis. (C) Sashimi plot quantifying read junctions that span exons 12–17 of KRBA1 in MV4;11 cells treated with DMSO (green) or C16 (blue). Numbers in the arcs display junction depth. Skipped exon 15 is highlighted in orange. (D) As in (C) but for read junctions that span exons 20–22 of TTF2. Skipped exon 21 is highlighted in orange. (E) As in (C), but for read junctions that span exons 2 and 3 of RPL22L1. The location of exons 2 and 3 is depicted at the bottom. Splicing of exon 2 to the distal acceptor site in exon 3 results in an mRNA encoding RPL22L1a (orange); splicing to the proximal acceptor site in exon 3 results in an mRNA encoding RPL22L1b (yellow). (F) Left: Representation of amplicons used to discriminate between different MDM4 (top) and RPL22L1 (bottom) isoforms via semi-quantitative PCR. Right: Results of semi-quantitative PCR analysis for the various isoforms of MDM4 and RPL22L1, and a GAPDH control, in MV4;11 or MOLM13 cells treated for 48 hours with DMSO or C16 (MV4;11, 100 nM; MOLM13, 250 nM; n = 3). All three biological replicates for DMSO and C16 are shown. Raw unprocessed gel images for the data in (F) are presented in Figure 6—source data 3. (G) Left: Representation of amplicons used to discriminate between MDM4 (top) and RPL22L1 (bottom) isoforms via RT-qPCR. Right: Results of RT-qPCR analysis for the various isoforms of MDM4 (top) and RPL22L1 (bottom) in MV4;11 or MOLM13 cells treated for 48 hours with DMSO or C16 (MV4;11, 100 nM; MOLM13, 250 nM; n = 3; Mean ±SEM). For each amplicon, isoform levels are expressed relative to the DMSO control. P-values are represented by asterisks: ‘*’ ≤ 0.05, ‘**’ ≤ 0.01, ‘***’ ≤ 0.001.

Impact of RPL22 loss on the response of MLLr cells to WINi.

(A) Western blot analysis of RPL22 expression in MV4;11, MOLM13, and K562 cells electroporated with Cas9 and either scrambled non-targeting (NT) control or RPL22-targeting sgRNAs. GAPDH and α-Actinin are loading controls. Representative images from three biological replicates shown. Raw unprocessed gel images are presented in Figure 6—source data 4. (B) GI50 values of C16 in non-targeted (NT) and RPL22 knock out (KO) MV4;11, MOLM13, and K562 cells in a 72 hour assay (n = 3; Mean ±SEM). (C) Number of genes with significantly (FDR < 0.05) altered transcript levels following treatment of RPL22KO or control (NT) cells treated with DMSO or 100 nM C16 for 48 hours, as determined by RNA-Seq (n = 4). See Figure 6—source data 6 for complete output of RNA-Seq analysis. (D) Volcano plots, showing pairwise transcript level alterations in NT (control) and RPL22KO MV4;11 cells treated 48 hours with DMSO or 100 nM C16 (red indicates FDR < 0.05). The location of transcripts from ZMAT3 and RPL22L1 are indicated. (E) Transcript level changes in WDR5-bound (left) and non-bound (right) RPGs in each of the indicated pairwise comparisons of RNA-Seq datasets. (F) Enrichment analysis of genes differentially induced by C16 in RPL22KO cells compared to control (NT) cells. KEGG and Hallmark.MSigDB pathways are shown. Fold enrichment of indicated pathways is presented on the x-axis, the number of genes is shown in italics in each bar, and colors represent -Log10 FDR. See Figure 6—source data 7 for complete enrichment analyses. (G) As in (F) but for suppressed genes. (H) Transcript level changes in mitochondrial ribosomal protein genes elicited by C16 in NT or RPL22KO cells.

Impact of RPL22 loss on the abundance of alternatively-spliced mRNA isoforms in MV4;11 cells.

(A) Differential alternative splicing events affected by C16 treatment of control (NT) or RPL22 knockout (KO) MV4;11 cells were quantified by rMATS. The types of alternative splicing events are cartooned at left, and the number of significantly different events (> 5% Δψ; FDR < 0.05) depicted in the graph. See Figure 6—source data 8 for output of rMATS analysis. (B) Sashimi plot quantifying read junctions that span exons 5–7 of MDM4 in NT or RPL22KO MV4;11 cells treated with DMSO or C16. Numbers in the arcs display junction depth. The location of exons 5, 6, and 7 is depicted at the bottom; skipped exon 6 is highlighted in orange. Note that RPL22KO images are also represented in Figure 6F. (C) As in (B) but for read junctions that span exons 2 and 3 of RPL22L1. The location of exons 2 and 3 is depicted at the bottom. Splicing of exon 2 to the distal acceptor site in exon 3 results in an mRNA encoding RPL22L1a (orange); splicing to the proximal acceptor site in exon 3 results in an mRNA encoding RPL22L1b (yellow).