# paper.pdf ## Metadata - PDFFormatVersion=1.3 - IsLinearized=false - IsAcroFormPresent=false - IsXFAPresent=false - IsCollectionPresent=false - IsSignaturesPresent=false - Producer=macOS 版本15.7.3（版号24G419） Quartz PDFContext, AppendMode 1.1 - ModDate=D:20260701071832Z00'00' - CreationDate=D:20260701071805Z00'00' ## Contents ### Page 1 Supplementary Material1Supplementary Methods2Antibodies3Anti-RBM5 (Abclonal, #A10788; Cell Signaling Technology, #86425S, 1:1000 for4immunoblotting); anti-HOXA9 (Abclonal, #A1908, 1:1000 for immunoblotting); anti-5FLT3 (Sangon, #D221706, 1:2000 for immunoblotting); anti-ACTIN (Sigma, #A2228,61:5000 for immunoblotting); anti-GAPDH (Abclonal, #A19056; Thermo Fisher,7#AM4300, 1:5000 for immunoblotting); anti-DHX15 (Proteintech, #12265-1-AP,81:2000 for immunoblotting); anti-c-Myc (Abclonal, #A19032, 1:2000 for9immunoblotting, 1:100 for immunoprecipitation); anti-IgG (Abclonal, #AC005, 1:10010for immunoprecipitation); anti-IgG (Abclonal, #AC011, 1:100 for11immunoprecipitation). Pierce Anti-c-Myc Magnetic Beads (Thermo Scientific, #88842,121:100 for immunoprecipitation); anti-Myc-tag (Cell Signaling Technology, #2276S,131:2000 for immunoblotting, 1:100 for immunoprecipitation); anti-V5-tag (Abclonal,14#AE089,1:2000 for immunoblotting, 1:100 for immunoprecipitation); anti-mCherry15(Abcam, #ab167453, 1:1000 for immunoblotting); anti-HDAC2 (Santa Cruz, #sc-1681599, 1:500 for immunoblotting); anti-MAX (Santa Cruz, #sc-756, 1:1000 for17immunoblotting); anti-MYC (Cell Signaling Technology, #9402S, 1:1000 for18immunoblotting). Note: Myc-HaloTag is a fusion protein, so we use the MYC antibody19(Cell Signaling Technology, #9402S) for immunoblotting.2021Culture and transduction of AML patients’ primary cells22Primary samples were cultured in primary AML culture media consisting of human23SFEM II (Stem Cell Technologies, #09650) supplemented with 20 ng/ml IL-324(PeproTech, #78042), 100 ng/ml FLT3 ligand (Peprotech, #300-19), 50 ng/ml TPO25(Peprotech, #300-18), 100 ng/ml SCF (Peprotech, #300-07), Glutamine (Gibco,26#25030-081, 2 mM) and 1% penicillin/streptomycin (NCM Biotech, #C100C5).27Primary samples were then transduced with concentrated virus expressing shNT or28shRBM5 shRNA together with puromycin (Gibco,#A11138-03) and 8 μg/ml29polybrene (Beyotime, #C0351). Cells were incubated for 6 hrs and new media was30added. Following puromycin selection for 48 hrs, the transduced cells were plated in31methylcellulose for colony-forming assays or directly analyzed by flow cytometry.3233Cell culture34HEK293T and Plat-E cells were cultured in DMEM (HyClone, #SH300022.01)35supplemented with 10% FBS and 4.5 mM L-glutamine. Murine c-Kit+ cells were36selected by anti-mouse CD117 MicroBeads (Stemcell, #18757) from 10-weeks-old37primary CD45.2+ mice, and were cultured overnight in IMDM 10% FBS medium38supplemented with 100 ng/mL recombinant murine SCF (R&D Systems, #455-MC),3920 ng/mL recombinant murine IL-3 (Peprotech, #213-13), 20 ng/mL recombinant40murine IL-6 (Peprotech, #216-16), 50 ng/mL recombinant murine FLT3 (Peprotech,41#250-31L) and 20 ng/mL recombinant murine TPO (Peprotech, #315-14). For the42murine MLL-AF9 cell line, cells were cultured in RPMI1640 10% FBS supplemented43with 5 ng/mL recombinant murine IL-3 (Peprotech, #213-13). Mouse leukemic bone44marrow cells from Rbm5fl/fl and Rbm5Δ/Δ mice were grown in RPMI (Cellgro) medium45containing 10% FBS, 6 ng/ml mIL-3, 10 ng/ml mSCF, 10 ng/ml mIL-6 and 10 ng/ml46mGM-CSF (Peprotech, #315-03). The CD34+ hematopoietic stem and progenitor cells47(HSPCs) were mobilized from normal subjects by granulocyte colony-stimulating48 ### Page 2 factor, collected by apheresis, and enriched by immunomagnetic bead selection using49the EasySep™ Human CD34 Positive Selection Kit (Stem Cell Technologies, #18,056),50in accordance with the manufacturer’s protocol,and cells were cultured in Stem Span51SFEM II with 20 ng/mL human TPO, 100 ng/mL human SCF, 100 ng/mL human FLT352Ligand, 20 ng/mL human IL6 (PeproTech, #200-06) and penicillin/streptonmycin. All53cells were incubated at 37 °C in a humidified atmosphere containing 5% CO2. The54HEK293T cells (no. SCSP-502) were obtained from the Cell Bank of the Chinese55Academy (www.cellbank.org.cn). The Plat-E cells (retroviral packaging cell line, #RV-56101) was originally from Cell Biolab, Inc. Before use in experiments, cell line identity57was verified by STR analysis, and cells were confirmed as mycoplasma negative.5859Retroviral transduction60BM progenitor (c-Kit+ HSPCs) cells were enriched by the CD117 microbeads (Stem61Cell Technologies, #118757) from 8-week-old Rbm5fl/fl and Rbm5Δ/Δ mice. c-Kit+ cells62were transduced with viral supernatants containing murine stem cell virus (MSCV)-63driven MLL-AF9 fusion oncogene in a construct tagged with GFP reported from an64internal ribosomal entry site (MSCV-MLL-AF9-GFP) for 3 days in RPMI1640 medium65with 10% fetal calf serum (FCS) supplemented with 10ng/mL murine stem cell factor66(mSCF) 6 ng/mL murine IL-3 and 10 ng/mL murine GM-CSF, and 10 ng/mL murine67IL-6, followed by injection of about 500,000 (~10% GFP+) cells and about 500,00068bone marrow support cell from C57BL/6J mice per recipient mouse via tail vein69injection into lethally irradiated (900 cGy) C57BL/6J mice.7071CRISPR-Cas9-mediated gene targeting and vector construction72The Cas9-expression vector, Lenti-Cas9-Blast, was purchased from Addgene (#52962).73Cas9 protein was introduced to human AML cell lines and MLL-AF9 cells by lentiviral74transduction and selected with 10 μg/mL blasticidin (Gibco, #A11139-03) to generate75Cas9-stable cell lines. The sgRNA sequences were chosen from the CRISPR library or76designed by a CRISPR design tool (http://www.crisprscan.org/) and generated as77oligonucleotide pairs. After annealing, constructs were cloned into the Lenti-guide-78puro-IRES-GFP vector (a sgRNA-expression vector), which encodes the sgRNA. Cells79were transduced with sgRNA lentivirus and puromycin (Gibco, #A11138-03) selection80for 72 hours after transduction. The oligonucleotides encoding sgRNAs are listed in81Supplementary Table 1. For the rescue experiments, the cDNAs of Hoxa9, Flt3, Rbm582(WT), and Rbm5 (PAM-MUT) were cloned into the Lenti-TRE3G-HAtag-MND-Zeo83backbone by the Gibson Assembly cloning strategy. PCR amplifications for cloning84were performed using PrimeSTAR Max Premix (TaKaRa, #R045). A Gibson Assembly85Cloning Kit (Abclonal, #RK21020) was used following the manufacturer’s instructions.8687Flow cytometry88For the hematopoietic and leukemia stem cell and mature cell analysis, suspended89single cells were prepared from bone marrow and peripheral blood. Red cells were lysed90by ACK LYSING Buffer (VWR) before staining for all FACS analysis except for red91cell analysis. Cells were incubated with antibodies in FACS buffer (2% FBS in PBS)92on ice for 20 minutes in the dark. Anti-Gr-1-AlexaFluor 488 (Biolegend, #108419) and93Anti-CD11b-APC (Biolegend, #101211), Ter119-APC (Invitrogen, #2518964), CD19-94FITC (Biolegend, #152403), and CD3-APC (Invitrogen, #4323725) are used as lineage95markers. Lin cocktail-FITC (eBioscience, #2812114) or Lin cocktail-APC (BD96Pharmingen, #558074), anti-Sca1-PerCP/Cyanine5.5 (Biolegend, #122524), c-Kit-97APC/eFluor780 (eBioscience, #2481340), CD34-BV421(BD Horizon, #562608),98 ### Page 3 CD127-Alexa Fluor700 (eBioscience, #56-1271-82) and CD16,32-PE (Biolegend,99#101308) antibodies are for leukemic stem cell L-GMP population or hematopoietic100progenitor/stem cell (HSPC) population analysis; Lin cocktail-FITC (eBioscience,101#2812114), anti-Sca1-PerCP/Cyanine5.5 (Biolegend, #122524), c-Kit-APC-eFluor780102(eBioscience, #2481340), CD48-APC (Biolegend, #103412) and CD150-BV421 (BD103Biosciences, #562811) antibodies were used for hematopoietic stem/progenitor cell104population (HSPC) analysis. For the flow cytometric analysis of acute myeloid105leukemia (AML) samples, mononuclear cells were isolated from primary AML patient106specimens. To identify and characterize the leukemic stem cell (LSC) population, cells107were stained with a panel of antibodies. CD34-APC (Biolegend, #128611) and CD38-108FITC (Biolegend, #356610) antibodies were used for the analysis of leukemic stem109cells in AML primary samples.110111Competition-based cell proliferation assay112Cas9-expressing MLL-AF9 cells were transduced with lentiviral pXPR-guide-puro-113IRES-GFP linked with a GFP reporter. The percentage of GFP-positive cells was114measured at day 3, the initial time point. GFP% was measured every 3 days over the115time course. The relative change in the GFP% at each time point was then normalized116to the GFP% at the initial time point. This relative change was used to assess the impact117of individual sgRNAs on cellular proliferation, which reflects the fact that cells with a118genetic knockout are outcompeted by non-transduced cells in the cell culture. Flow119cytometry was performed on BD LSRFortessa (BD Biosciences) and analyzed with120FlowJo software (Treestar).121122Colony-formation and replating assay123BM progenitor (HSPC, that is, c-Kit+) cells were enriched by the CD117 microbeads124(Stem Cell Technologies, #18000) from 8-week-old Rbm5fl/fl and Rbm5Δ/Δ mice. A125colony assay for murine c-Kit+ cells was performed by plating 2,000 sorted c-Kit+ cells126on methylcellulose (Stem Cell Technologies, #M3434). CFU colonies corresponding127to erythroid progenitor cells (BFU-E), granulocyte-macrophage progenitor cells (CFU-128GM, CFU-G, and CFU-M), and multipotent granulocyte, erythroid, macrophage, and129megakaryocyte progenitor cells (CFU-GEMM) were scored 7 days after seeding. For130transformed murine LSK and leukemic cells, about 20,000 cells were plated, and131colonies were counted 7 days later. The CFA cells were collected, counted, and 20,000132cells were used for a second round of colony formation assay. For MLL-AF9-Cas9 cells133treated with sgRNAs, about 300 cells were seeded in methylcellulose (Stem Cell134Technologies, #M3231), and colonies were counted 7 days later. For PDX-derived cells135treated with shRNAs, about 50,000 cells were seeded in methylcellulose (Stem Cell136Technologies, #H4034), and colonies were counted after 7-10 days of incubation.137138In vitro limiting dilution assays (LDAs)139For in vitro LDAs, frozen BM progenitor cells collected from primary BMT mice140(Rbm5fl/fl and Rbm5Δ/Δ, three mice from each group, whose survival is close to the141median survival of the corresponding group) transduced with different combinations of142retroviruses were seeded into Mouse Methylcellulose Media (STEMCELL143Technologies, #M3434) along with 1 μg/mL puromycin. Five days later, the colony144cells were collected and replated into 96-well plates at varying cell numbers, with 24145replicates per group. After 5-7 days, the number of wells with colonies was counted.146ELDA software was used to analyze frequencies of leukemia stem/initiating cells147(LSCs/LICs).148 ### Page 4 RNA-sequencing (RNA-seq) analysis149RNA samples from three biological replicates were extracted from cultured cells using150TRIzol (Ambion by Life Technologies, #343911) according to the manufacturer’s151instructions. The RNA was then sent to the commercial company Novogene (Beijing,152China) for library preparation and next-generation sequencing. Raw transcript counts153were derived from FastQ files using the alignment-independent quantification tool154Salmon (https://combine-lab.github.io/salmon/), with standard settings. The raw count155matrix was imported into RStudio and used as input for DESeq2 analysis, following the156package's Vignette for normalization, differential gene expression analysis, and157unbiased clustering, including principal component analysis. The DESeq2 output was158used as input for pre-ranked GSEA to enrich for functional pathways and gene159signatures (https://www.gsea-msigdb.org/gsea/index.jsp).160161Quantitative PCR (qPCR) analysis for gene expression162Total RNA was extracted from AML cells using the FastPure Cell/Tissue Total RNA163Isolation Kit (Vazyme, #RC112-01). cDNA was generated using PrimeScript™ RT164Master Mix (Takara, #RR036Q) from 1 μg RNA and diluted 1:200 for qPCR analysis.165qPCR was performed with 1 μL of diluted cDNA, using biological and technical166replicates, and SYBR Green Master Mix (Vazyme, #Q511-03) on a QuantStudio 6 real-167time PCR system. The results were normalized to GAPDH expression. The primer168sequences used for qPCR are in Supplementary Table 1.169170Immunoblotting and Immunoprecipitation171Cells were washed with PBS and lysed in Cell Lysis Buffer for Western and IP172(Beyotime, #P0013) with a protease inhibitor cocktail. Lysates were heated to 95 °C in173SDS sample buffer, separated by SDS-PAGE, and transferred to a nitrocellulose174membrane. Membranes were blocked in 5% non-fat milk in PBS with 0.1% Tween-20,175probed with the indicated primary antibodies, and then incubated with an HRP-coupled176secondary antibody for 1 h at room temperature. Blots were visualized using enhanced177chemiluminescence detection reagents and exposed to X-ray film. Representative plots178of three biological replicates are shown. Immunoprecipitation (IP) was conducted using179antibodies or antibody-tagged beads, as indicated in the figure legends. Briefly, 300-1801,000 μg of proteins were prewashed with Protein A and G beads (Santa Cruz181Biotechnology,#sc-2003). Then, the proteins were incubated with the indicated182antibodies or MYC beads at 4 °C for either 4 to 6 h or overnight. Protein A or G beads183were then added to protein samples containing only antibodies, and the mixture was184incubated at 4 °C for an additional 2 h. All of the beads attached samples were washed185at least three times with lysis, SNNTE [50 mM Tris/HCl (pH 7.4), 5 mM EDTA, 5%186sucrose, 1% NP-40, and 0.5 M NaCl], or RIPA [50 mM Tris/HCl (pH 7.4), 150 mM187NaCl, 1% Triton X-100, 0.1% SDS, and 1% Na Deoxycholate] buffers. After washing,188the samples were boiled with SDS loading buffer and resolved by 10% SDS-PAGE.189Bound proteins were detected by immunoblotting with specific antibodies, as indicated190in the figure and figure legends. All antibodies used in this study are listed in191Supplementary Table 1.192193HaloTag Chromatin Immunoprecipitation sequencing (ChIPseq)194SEM AID-RBM5 cells were treated with 1 μM 5-Ph-IAA (MedChemExpress, #HY-195134653) for 24 hours. 20 million treated or DMSO control cells were quickly washed196once with PBS and resuspended in 1.5 mL room temperature Fixing Buffer A (Covaris197truChIP Shearing Kit, #520154). About 150 μL of 11.1% formaldehyde (1% final) was198 ### Page 5 added, incubated at room temperature with rocking for 5 minutes, followed by a 5-199minute incubation with 87 μL Quenching Buffer (Covaris). Cells were rewashed with200PBS and flash frozen for storage at -80 °C. The nuclei were thawed on ice and lysed201according to the Covaris TruChIP protocol. The Lysis Buffer and Wash buffers202supplied with the kit were prepared using Protease Inhibitor Cocktail (Promega,203#G6521) to a 1x concentration. Prepared nuclei were sheared in 1mL shearing buffer204[10mM Tris-HCl, 1mM EDTA, and 0.01% SDS + protease inhibitors (Promega)] using205Covaris M220 at a duty factor of 10 and 200 cycles/burst for 10 minutes at 4 °C. The206sheared chromatin was cleared by centrifugation, and the supernatant was transferred207to a 15mL conical tube containing 10mL of IP buffer (100 mM Tris-HCl, 150 mM NaCl,2080.001% IGEPAL, and 1x protease inhibitors). HaloLink Resin (Promega, #G1913) was209prepared by washing 200 μL of resin in reaction with 1mL of IP buffer twice. Chromatin210and resin were incubated together overnight at 4 °C with rotation. The next day, the211resin was pelleted by centrifugation and washed three times with IP buffer plus PI.212Following the first wash, the resin was moved to a 1.5mL tube for subsequent washes.213DNA was de-crosslinked and eluted from the resin by 1 x TE, 1% SDS, proteinase K,214and 400 mM NaCl at 65 °C for 4 hours. DNA was purified using215phenol:chloroform:isoamyl alcohol precipitation for library preparation and sequencing216by NovaSeq.217218Supplementary Figure Legends219Supplemental Fig. S1. Rbm5 promotes murine AML development and220progression221222A. Representative flow cytometric analysis of the percentage of GFP+ cells after223infection with MSCV-MLL-AF9-GFP retrovirus in c-Kit+ cells derived from Rbm5fl/fl224and Rbm5Δ/Δ mice.225226B. Bar graph showing the mRNA levels of MLL-AF9 after being infected with MSCV-227MLL- AF9-GFP retrovirus in c-Kit+ cells derived from Rbm5fl/fl and Rbm5Δ/Δ mice.228 229C. Real-time qPCR was used to detect Rbm5 mRNA level in c-Kit+ cells and MLL-AF9230leukemia cells; GAPDH was used as an internal control.231 232D. Western blot was used to detect Rbm5 protein level in c-Kit+ cells and MLL-AF9233leukemia cells, respectively. GAPDH was used as a loading control.234235E. Rbm5 deletion significantly inhibits the second-round colony-forming ability of236leukemia cells derived from the spleen of MLL-AF9 leukemic mice in Fig. 1C.237Results are from Rbm5fl/fl (n238= 3) and Rbm5Δ/Δ (n = 3) mice. CFU: Colony-Forming Unit.239240F. Proliferation curve of primary MLL-AF9 leukemia cells from the moribund Rbm5fl/fl241and Rbm5Δ/Δ mice.242243G. Representative flow cytometry analysis of myeloid differentiation marker (Cd11b and244Gr1) in primary MLL-AF9 leukemia cells derived from Rbm5fl/fl and Rbm5Δ/Δ mice.245246H. Schematic diagram of the construction and study of the murine MLL-AF9 leukemia247 ### Page 6 cell model by CRISPR/Cas9-mediated Rbm5 depletion. Real-time qPCR and Western248blot were used to detect the expression level of Rbm5 in MLL-AF9-Cas9 cells after249targeting with two respective sgRNAs (Rbm5-sg1/sg2). ACTIN was used as an250internal control. Created in BioRender. Xin, Y. (2026)251https://BioRender.com/ny9uzov.252 253I. Cell counting assay was conducted on the sgNT, Rbm5-sg1, and Rbm5-sg2 targeted254MLL- AF9-Cas9 to monitor the proliferation of MLL-AF9 cells. sgNT used as a255negative control.256 257J. Competitive proliferation assay (CPA) was conducted in Cas9 stably expressed MLL-258AF9 after transduction with GFP reporter-based lentiviral sgRNAs (sgNT, Rbm5-sg1,259Rbm5-sg2) at about ~50% efficiency. The GFP% was quantified at days 3, 6, 9, 12,26015, and 18 by flow cytometry to evaluate the growth disadvantage. The guide RNA261targeting the non-target (sgNT) gene was included as a negative control. Flow262cytometry analysis of the percentage of GFP+ leukemia cells in bone marrow and263spleen of recipient mice in the sgNT (n=6) and Rbm5-sg1 (n=7) groups sacrificed264after 28 days post- transplantation.265 266K. Flowchart of Competitive proliferation assay (CPA) rescue assay by the construction267of a Doxycycline (Dox)-inducible cell line that introduces the TRE3G-Rbm5 PAM268MUT cDNA into MLL-AF9-Cas9 cells and the functional rescue experiment using269these cells. Schematic diagram of the sgRNA-resistant cDNA mutagenesis. The 22270bp DNA sequence and corresponding amino acids close to the sgRNA PAM region271(blue) in RBM5 cDNA with Rbm5-sg1 resistant mutant (PAM-MUT) are designed272with the non-sense mutated nucleotides highlighted in different colors (red and273orange). A schematic diagram of the inducible Tet-on system for Rbm5 PAM-274resistant cDNA expression is shown below. Western blot analysis of RBM5 protein275expression in MLL-AF9- Cas9-TRE3G-Rbm5 cells transduced with sgNT or Rbm5-276sg1, with or without Dox treatment. GAPDH was used as an internal control.277278L. Competitive proliferation assay in MLL-AF9-Cas9-TRE3G-Rbm5 cells transduced279with non- targeting sgNT or Rbm5-sg1, maintaining the GFP+ cell ratio at280approximately 50%.281282Data are presented as the mean ± SEM of three independent experiments. Statistical283significance was determined using one-way ANOVA with Tukey's multiple284comparison test (B, C and H), Student’s two-tailed unpaired t-test (E and K) and two-285way ANOVA with Sidak’s multiple comparison test (F and I) (* P < 0.05; ** P < 0.01;286*** P < 0.001; **** P < 0.0001).287288Supplemental Fig. S2. Rbm5 is dispensable for normal hematopoiesis289290A. The body weight of Rbm5Δ/Δ mice showed no difference compared to Rbm5fl/fl mice.291Body weight statistics of Rbm5fl/fl Rbm5 fl/Δ and Rbm5Δ/Δ mice at 6, 10, 17, and 25292weeks of age.293294B. The major organ development of hematopoietic system-specific Rbm5 conditional295knockout mice is normal. Statistical graphs of the weight and size of various tissues296 ### Page 7 and organs in Rbm5fl/fl and Rbm5Δ/Δ mice. Each point represents one mouse, n = 6.297 298C. The bar graph shows the proportions of various terminally mature cells peripheral299blood cells in Rbm5fl/fl (n=3) and Rbm5Δ/Δ (n=3) mice.300301D. Representative flow cytometry analysis of bone marrow and peripheral blood cells in302Rbm5fl/fl and Rbm5Δ/Δ mice, including T cells, B cells, myeloid cells, and erythrocytes.303304E. The three-part differential blood routine test of peripheral blood in Rbm5 conditional305knockout mice is normal. Peripheral blood was collected from the tail vein of Rbm5fl/fl,306Rbm5fl/Δ and Rbm5Δ/Δ mice at 6, 10, and 17 weeks of age for three-part differential307blood routine testing. WBC: white blood cells, RBC: red blood cells, Platelets:308platelets, HGB: hemoglobin, HCT: hematocrit, LYM: lymphocytes. Each point309represents one mouse, n = 8.310311Each dot represents an independent sample. Data are presented as the mean ± SEM.312Statistical significance was determined using Student’s two-tailed unpaired t-test (B313and C) and two-way ANOVA with Sidak’s multiple comparison test (A and E) (* P <3140.05; **P < 0.01; “n.s.” indicates no statistical significance).315316Supplemental Fig. S3. Rbm5 specifically sustains the activity of LSCs but not the317normal HSCs318319A. Heatmap of marker genes for 15 subpopulations identified through unbiased320clustering of the 51,517 cells in the panel.321322B. Stacked bar plot showing the proportions of different cell subpopulations from various323samples.324325C. UMAP plots showing the expression of Kmt2a, Flt3, and Hoxa9 genes in the MLL-326AF9 leukemia model and primary c-Kit+ cells, where color intensity represents327expression levels, and each dot represents a single cell.328329D. Volcano plot illustrating differentially expressed genes in hematopoietic330stem/progenitor cells (HSPCs) populations between Rbm5fl/fl and Rbm5Δ/Δ in the331primary c-Kit+ model.332333E. The bubble plot displays the top five enriched gene targets (ranked by the P value334from the enrichR dataset) among downregulated genes (left panel) and upregulated335genes (right panel) in the CD34+ leukemia stem cells (LSCs) population following336Rbm5 deletion in the MLL-AF9 leukemia model.337338F. Summary information for the genotype from the four AML patients including339KMT2A-rearranged and KMT2A-PTD mutation.340341G. The colony-forming numbers after knocking down RBM5 using shRNAs in patient-342derived xenograft #1(PDX#1) samples, which were expanded in NSG-3M mice from343originally the bone marrow primary cells of an AML patient (MLL-AF17).344345H. Representative flow cytometry analysis and statistical analysis of CD34 and CD38346 ### Page 8 after knocking down RBM5 using two respective shRNAs (shRBM5#1 and347shRBM5#2) in PDX#1 samples.348349Data are presented as the mean ± SEM of three independent experiments. Statistical350significance was determined using Student’s two-tailed unpaired t-test (G and H) (* P351< 0.05; *** P < 0.001).352353Supplemental Fig. S4. Rbm5 loss leads to the suppression of LSCs self-renewal354genes, activation of myeloid differentiation genes.355356A. Sashimi plots demonstrating the deletion of exon 3 and exon 4 of Rbm5 in Rbm5Δ/Δ357MLL-AF9 cells. While red arrows indicate exons 3 and 4.358359B. Principal Component Analysis (PCA) of RNA-Seq data, where each dot represents360an independent sample.361362C. Bubble plot showing the top 5 GO-enriched pathways for differentially expressed363genes (downregulated and upregulated) in primary MLL-AF9 leukemia samples364following Rbm5 deletion.365366D. Real-time qPCR analysis of the representative downregulated mRNA levels of Rbm5367in primary leukemia cells derived from the spleen in the MLL-AF9 model.368369E. Real-time qPCR analysis of the representative downregulated mRNA levels of Hoxa9,370Flt3, Ikzf2, and Cd34 in primary leukemia cells derived from the spleen in the MLL-371AF9 model.372373F. Western blot analysis of RBM5, HOXA9 and FLT3 protein levels in leukemia cells374derived from the bone marrow and spleen of the MLL-AF9 mouse model upon RBM5375deletion. GAPDH was used as a loading control.376377Data are presented as the mean ± SEM of three independent experiments. Statistical378significance was determined using Student’s two-tailed unpaired t-test (D and E) (* P379< 0.05; ** P < 0.01).380381Supplemental Fig. S5. RBM5 controls the MYC transcriptional activity via382physical interaction and maintenance of its protein, and proper genomic383occupancy384385A. Real-time qPCR and Western blot analysis of MYC mRNA and protein levels in386MOLM13 cells upon RBM5 knockout. Data are presented as the mean ± SEM of387three independent experiments. **P < 0.01, Student’s two-tailed unpaired t-test.388389B. Real-time qPCR and Western blot analysis of MYC mRNA and protein levels in390Rbm5fl/fl and Rbm5Δ/Δ primary c-Kit+ hematopoietic stem/progenitor cells (HSPCs).391Data are presented as the mean ± SEM of three independent experiments. n.s.392indicates no statistical significance, Student’s two-tailed unpaired t-test.393394C. Proteins interacting with RBM5 identified in two independent IP-MS experiments,395with RBM5 and DHX15 serving as positive controls (left panel). The PPI (protein-396 ### Page 9 protein interaction) network displays proteins interacting with RBM5, as determined397by IP-MS(Immunoprecipitation-Mass Spectrometry) analysis (right panel).398399D. Bubble plot displaying the top 10 biological processes enriched by GO analysis for400proteins interacting with RBM5.401402E. Co-transfection of RBM5-MYCtag and MYC-V5tag expression vectors in 293T cells,403followed by co-immunoprecipitation using MYCtag and V5tag, respectively. Western404blot was performed to detect proteins interacting with MYCtag or V5tag.405406F. Co-Immunoprecipitation using HaloTag antibody beads in RBM5-AID SEM cells407that overexpressed the HaloTag version of MYC. Immunoprecipitation followed by408Western blot detection of miniAID-RBM5, Myc-HA-HaloTag and the endogenous409MYC protein levels were performed before and after RBM5 protein degradation.410411Supplemental Fig. S6. RBM5 controls the proper genomic occupancy of MYC412413A. Principal component analysis (PCA) of MYC-HaloTag ChIP-seq data in RBM5-AID414SEM cells untreated (-5-Ph-IAA) or treated with 1μM 5-Ph-IAA (+5-Ph-IAA) for 24415hours. Each condition had two independent replicates.416417B. Genome browser tracks of two canonical MYC target genes with decreased MYC-418HaloTag binding signal after RBM5 depletion (Left panel: the NPM1 locus; Right419panel: the NOP2 locus). Displayed are the ChIP-seq profiles of MYC-HaloTag in420RBM5-AID SEM cells untreated (-5-Ph-IAA) or treated with 1μM 5-Ph-IAA (+5-Ph-421IAA) for 24 hours. Each group has two biological replicates.422423C. Genome browser tracks of two example genes with no change of MYC-HaloTag424binding signal after RBM5 depletion (Left panel: the MARCHF9 locus; Right panel:425the FLT3 locus). Displayed are the ChIP-seq profiles of MYC-HaloTag in RBM5-426AID SEM cells untreated (- 5-Ph-IAA) or treated with 1μM 5-Ph-IAA (+5-Ph-IAA)427for 24 hours. Each group has two biological replicates.428429Supplementary Table Legends430Supplemental Table 1. Oligonucleotides, primers and antibodies used in this study.431432Supplemental Table 2. Differentially expressed genes (DEGs) in the Rbm5 KO433leukemia stem cells (LSCs).434435Supplemental Table 3. Enrichment results of EnrichR after Rbm5 knockout in the436CD34+ leukemia stem cells (LSCs) population within the MLL-AF9 model single-cell437data.438439Supplemental Table 4. Differentially expressed genes (DEGs) analysis between Rbm5440knockout (KO) and control MLL-AF9 leukemia cells.441442 ### Page 10 Supplemental Table 5. Summary of KEGG and GSEA enrichment results for443Differentially expressed genes (DEGs) in Rbm5-knockout versus control MLL-AF9444leukemia cells.445446Supplemental Table 6. Identified by mass spectrometry analysis, the list of proteins447interacting with RBM5 in OCIAML2 cells with exogenous RBM5 overexpression.448 ### Page 11 AGFP Neg0.53% 9.67% 8.72%CB EMLL-AF9 SS-A Rbm5 fl/fl Rbm5∆/∆ **** c-Kit MLL-AF9 Relative MLL-AF9 mRNA Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆0 20 40 60 80 * CFU per 10,000 cell Rbm5 fl/fl Rbm5 ∆/∆0.00 0.02 0.04 0.06 MLL-AF9 leukemia cells from spleenFCD11bGr1Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆D GRelative Rbm5 mRNA MLL-AF9c-Kit+ 0.000 0.005 0.010 0.015 0.020 0.025 Rbm5 fl/fl Rbm5 ∆/∆Rbm5 fl/fl Rbm5 ∆/∆* ** * * * * *** CFU per 20,000 cell Rbm5 fl/fl Rbm5 ∆/∆0 50 100 150 200 1st 2nd Myeloid differentiation markerGAPDH RBM5 c-Kit+ MLL-AF9 Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆1 0.05 2.46 0.03 130 kDa 37 kDa Days Cell number (x104) Liquid culture expansion ** 0 1 2 3 4 5 0 50 100 150 200 250 Rbm5 fl/flRbm5∆/∆* ** n.s. n.s. ** Liquid culture in MLL-AF9Cas9 Cell number (x104) 0 1 2 3 4 5 0 100 200 300 400 500 sgNT Rbm5-sg2 Rbm5-sg1 RBM5 cDNA (PAM-MUT)-/+DoxMLL-AF9Cas9- RBM5-OECPA Rescue assay sgNT orRbm5-sg1MND ZeocinTet3G +Dox (on) -Dox (off) x HA-Rbm5PAM-MUT Tet-O Rbm5-sg1sgNT Dox +- +- RBM5 GAPDHMLL-AF9Cas9-RBM5-OE 130 kDa 37 kDa0 30 60 90 120-Dox Rbm5-sg1 +Dox +Dox-Dox sgNT GFP% (nomalized to D3) D3 D6D9D12D15D18H I J K L MCPA validation MLL-AF9Lenti-Cas9Lenti-sgRNA-GFP (sgNT, sgRbm5) ACTIN RBM5Rbm5-sg1 Rbm5-sg2 MLL-AF9Cas9 sgNT 130 kDa 42 kDa D3 D6D9D12D15D18CPA in MLL-AF9Cas90 30 60 90 120GFP% (nomalized to D3) sgNT Rbm5-sg1 Rbm5-sg2 Spleen (leukemia burden) Bone marrow (leukemia burden) * sgNT Rbm5-sg10 10 20 30 GFP+ (%)sgNT RBM5-sg1 0 10 20 30 40 50 GFP+ (%) * mRNA level % ** *** sgNT Rbm5-sg1 Rbm5-sg20.0 0.3 0.6 0.9 1.2Figure S1 ### Page 12 BA DHeart Liver Spleen Lung Kidney 0.00 0.02 0.04 0.06 0.08 Coefficients Rbm5 fl/fl Rbm5∆/∆n.s. n.s. n.s. n.s.n.s. 0 5 10 15 20 0 5 10 15 20 25 Cell number (10*9/L) Rbm5 fl/fl Rbm5∆/∆Rbm5fl/∆0 5 10 15 20 0 2 4 6 8 10 weeksweeks RBC(K/ul) 0 5 10 15 20 0 50 80 100 120 140 160 weeks HGB(g/L) 0 5 10 15 20 0 800 800 1000 1200 1400 1600 Platelets(K/ul) 0 5 10 15 20 0.0 0.3 0.6 0.7 0.8 0.9 1.0 LYM% weeksweeks 0 5 10 15 20 0.0 0.2 0.4 0.6 weeks HCT% n s. n s. n s. * ** n s. n s.n s. n s. n s. n s. n s. n s.n s. n s. n s. n s. n s. n s.n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s. n s.ECd3 Cd19Cd11b Gr1 Ter119 SS-ARbm5-/- Rbm5+/+ ErythrocyteMyleiodT cell/B cell ErythrocyteMyleiodT cell/B cellMature cells in bone marrow Mature cells in peripheral bloodRbm5-/- Rbm5+/+ Cd3 Cd19 Cd11b Gr1 Ter119 SS-AT cell B cell Myleiod Erythrocyte 0 50 100 150 % of PB Cell Peripheral blood n.s. n.s. n.s.n.s. Rbm5 fl/fl Rbm5∆/∆C0 5 10 20 30 0 15 15 20 25 30 35 Body weight /g Weeks Rbm5 fl/fl Rbm5∆/∆Rbm5fl/∆n s. n s. n s. n s. n s. n s. n s. n s.Figure S2 ### Page 13 Flt3Kmt2aHoxa9MLL-AF9 leukemia Primary c-Kit+ 0.00 0.25 0.50 0.75 1.00 0.00 0.25 0.50 0.75 1.00 Percent %MLL-AF9 leukemia Primary c-Kit+Log2 Exp6.00.0HSPCs Rbm5∆/∆ /Rbm5fl/flB A C E G HshNT CD34+CD38- shRBM5#1 shRBM5#2 CD34+CD38- CD34+CD38- CD38 CD34 Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆Rbm5 fl/fl Rbm5∆/∆D3.99% 1.49% 1.80% Log2FoldChange−1 0 1 2 3 0 20 40 60−Log10(P value) Down = 4 Up = 3 CFU in PDX#1 CD34+CD38- relative frequency CFU /50, 000 cell0.0 0.5 1.0 1.5shNT shRBM5 0 20 40 60 80 *** * shNT shRBM5 LSC in PDX#1Primary BM cells from AML patient (MLL-AF17) Three round expansion in NSG-3MFunction assay: CFU/ LSCLenti-shRNAs (shNT, shRbm5) PDX#1 cellsB230303A05RikScrg1Fpr1Mmp8Dhrs9Thbs1Adam8Cxcr2RetnlgMmp9Ceacam10Slfn4AA467197Arg2Clec4dAnkrd33bOrm1CampFmnl2Acod1Il1bStfa3Stfa2Cstdc5Stfa2l1Pf4Irf8Ctla2aNrxn1Angpt1Prkg1Mir99ahgMt2Cst7Nkg7MpoCtsgPrtn3ElaneMs4a3Gstm1H2afxTop2aNusap1CenpeCenpfUbe2cHist1h3cMns1Mrpl52SpicCdh2Col5a2Col11a1Dlc1DcnPappaCol1a2SparcCcl2Fn1Apba1CfhPid1Ifi207LplSlfn5Clec4a3FcmrMs4a1IghdCpmBlkCcr7Gm8369Scd1Vpreb1Igll1Gm30948Gfra11700027J07RikCplx2Gfra2Lef1Prss34Cd200r3Fcer1aIl6Cyp11a1Mcpt8Cdh1Csrp3Sox6RhagCldn13Spta1GypaAnk1Trim10Slc25a21Car1Car2Smim1Klf1Klra13-psCd3gCd3eKlra7Samd3Klrc2Xcl1Gm2682IghaJchainIgkcIghg2bIglv1Ighg3Ighg2cIgkv12-46Activated neutrophils CD34+ HSPCs Erythrocytes Erythroid precursors Inflammatory monocytes/macrophages Inflammatory neutrophils Mast cells Mature B cells Mature neutrophils neutrophil/granulocytic precursor Plasma cells Pre-B cells Proliferating (cycling)/stem cells T cells Tissue-resident macrophages MYB 21317192 TEAD4 22529382 AF4 28076791 AF4 26711339 ENL 26711339 Overlap 33 51 73 15.0 17.5 20.0-Log10(P value) TAL1 20887958 NELFA 20434984 TCF7 22412390 E2F1 18555785 CREM 20920259 Overlap 200 250 42 44 46 -Log10(P value)Up in enrichR_ChEADown in enrichR_ChEAFAML#1 KMT2A-AF9AML#2 KMT2A-PTDAML#3 KMT2A-ELL ID Genotyping KMT2A-AF17 AML#4Figure S3 ### Page 14 6588283758832442114283294312122082932845574341444543011457646613201081129320912712437284763330717461619845782484842839943694781396143113396405431447845441044227856260113105272106133011437355205264148421006196810705585284514579507139401659239087775106406646536324437913921471639179491525843245012202531384911461110831212416961651527105981554894231516819713206576305831362814691374482922602229428142464214319169171141112821248686674491111659425539135125928017285068563973680356321101614279284624546687247111349291168912411311639563488723341063210105723979864224863180475637731572204577284351431212499515101163392768107617412 107627730 107638048 107648367A B D-100 -50 0 50 100 -100 0 100 Dim1 (38.6%) Dim2 (17.8%) n=3 Rbm5fl/fl n=3 Rbm5 ∆/∆Rbm5 ∆/∆Rbm5fl/flRBM5HOXA9 FLT3GAPDHSpleen Rbm5 fl/fl Rbm5 ∆/∆Bone marrow RBM5 HOXA9 FLT3 GAPDH Rbm5 fl/fl Rbm5 ∆/∆E FAnterior/posterior pattern specification Postsynapse organization Regionalization Regulation of synapse organization Regulation of synapse structure or activity Count5 6 Myeloid leukocyte migration Skin development Leukocyte migration Chemotaxis Regulation of immune effector process Count 15.0 17.5 20.09 11 13 4 8 12 16 Down_GO Up_GOC-Log10(P Value) -Log10(P Value) 130 kDa 37 kDa 130 kDa 35 kDa 130 kDa 37 kDa 130 kDa 35 kDa exon3exon40.00 0.02 0.04 0.06 Rbm5 fl/fl Rbm5 ∆/∆** Relative mRNA level *0.000 0.002 0.004 0.006 0.008 0.010 **0.000 0.005 0.010 0.015 **Ikzf2Hoxa9 Flt3 Cd34Rbm5 fl/fl Rbm5 ∆/∆Rbm5 fl/fl Rbm5 ∆/∆0.00 0.01 0.02 0.03 Rbm5 fl/fl Rbm5∆/∆** Relative mRNA level Relative mRNA level Relative mRNA level Relative mRNA levelRbm5Spleen0.00 0.05 0.10 0.15 Rbm5 fl/fl Rbm5 ∆/∆Figure S4 ### Page 15 E BA CInput IP:IgG IP:MYCtag V5tag MYCtag Input IP:IgG IP:V5tag MYCtag RBM5-MYCtag + MYC-V5tag V5tagD0 50 100 0 50 100 Razor + unique peptides_rep1 Razor + unique peptides_rep2 RBM5 DHX15 RBM39 RBM17 MYC total = 658 IP-MASS OCIAML2FATP−dependent protein folding chaperonerRNA bindingStructural constituent of cytoskeleton Unfolded protein binding Helicase activity Ribonucleoprotein complex binding Catalytic activity, acting on RNA Cadherin binding ATP hydrolysis activity Structural constituent of ribosome Count 20 40 6020 30 40 50 60Enriched pathway for RBM5 interacting proteins-Log10(P ValueGAPDHminiAID-RBM5MYC (endogenous)GAPDHMyc-HA-HaloTagSEMminiAID-RBM5 clone 8Marker +Myc-HA-HaloTag 5-Ph-IAA (1μM) - - +cell lysate for immunoprecipitation MYC-antibodyRBM5Western blot forRBM5 and MYCHEK293T cellsRBM5-MYCtag + MYC-V5tagV5-antibodyY MYCor YMOLM13 RBM5 ACTIN MYC KO NT 1 0.492 1 0.422MYC mRNA levelDHX15MYBBP1AMCM3RBM17H1-4MYCMYBH2AC14SPI1H1-5ZNF346CTCFRBM5SIRT7PHF6WDR5GTF3C3EIF4A1H3-3B IDH2GTF3C5 CXXC1RBM39ZCCHC17PHF5ANONOZFP91RBBP4TRIM28NSD2RUVBL2RUVBL1ZNF598LMNB1Primary c-Kit+GAPDH MYCRbm5 fl/fl Rbm5 ∆/∆#1 #2 #3 #1 #2 #3RBM5 0.9 0.92 1.19 0.62 0.74 0.9 0.92 1.0 1.1 0.16 0.12 0.090.0 0.5 1.0 1.5MYC protein level n.s.Rbm5 fl/fl Rbm5 ∆/∆ 130 kDa 42 kDa 70 kDa 50 kDa 130 kDa 37 kDa 70 kDa 50 kDa 130 kDa 70 kDa 50 kDa 130 kDa 70 kDa 50 kDa 125 kDa 70 kDa 38 kDa 90 kDa 70 kDa 50 kDa 37 kDaNT KO 0.0 0.5 1.0 1.5 **0.00 0.05 0.10 0.15 n.s.Rbm5 fl/fl Rbm5 ∆/∆Myc mRNA levelFigure S5 ### Page 16 A B CPC1 (95.6% variance) PC2 (3.4% variance) -20 -10 0 10 20 4 2 0 -2 -4 +5-Ph-IAA +5-Ph-IAA -5-Ph-IAA -5-Ph-IAA hg38 chr5:171,385,116-171,412,900NPM10-31 0-31 0-31 0-31 0-31 Example genes with decreased MYC-HaloTag binding hg38 chr12:6,556,653-6,572,073NOP20-31 0-31 0-31 0-31 0-31 hg38 chr12:57,753,040-57,762,475 0-26 0-26 0-26 0-26 0-26MARCHF9FLT3hg38 chr13: 28,091,334-28,112,234FLT3MYC-HaloTag ChIP-seq 0-99 0-99 0-99 0-99 0-99 Input -5-Ph-IAA -5-Ph-IAA +5-Ph-IAA +5-Ph-IAA Input -5-Ph-IAA -5-Ph-IAA +5-Ph-IAA +5-Ph-IAANPM1 locus NOP2 locus Input -5-Ph-IAA -5-Ph-IAA +5-Ph-IAA +5-Ph-IAA Input -5-Ph-IAA -5-Ph-IAA +5-Ph-IAA +5-Ph-IAAMARCHF9 locus FLT3 locus Example genes with no change of MYC-HaloTag binding MYC-HaloTag ChIP-seq MYC-HaloTag ChIP-seqMYC-HaloTag ChIP-seq PCA for MYC-HaloTag ChIP-seqFigure S6