Informed consent was obtained from all patients. The protocol for this study was approved by the Institutional Review Board of Tohoku University Graduate School of Medicine and was based on the Ethical Principles for Medical Research Involving Human Subjects of the Declaration of Helsinki.
The protocol for culturing human erythroleukemia cell line K562 cells (ATCC CCL-243™, Manassas, VA), human embryonic kidney 293 T cells (HEK293T), and packaging cell line Plat-GP cells derived from HEK293T has been described previously.26.47.
HUDEP-2 cells26 were maintained in StemSpan serum-free expansion medium (STEMCELL Technologies, Vancouver, BC, Canada), supplemented with 50 ng/mL stem cell factor (SCF; PEPROTECH, Rocky Hill, NJ), 3 IU/mL d erythropoietin (EPO; Kyowa Hakko Kirin, Tokyo, Japan), 1 μg/mL doxycycline (DOX; Sigma-Aldrich, St. Louis, MO), and 1 μM dexamethasone (DEX; Sigma-Aldrich). Mouse mesenchymal stromal cell line (ATCC) OP9 cells were maintained in α-minimum essential medium (Thermo Fisher Scientific, Waltham, MA) supplemented with 20% (v/v) fetal bovine serum (FBS; Biological Industries USA, Cromwell, CT) and 1% (v/v) penicillin/streptomycin (Sigma-Aldrich). To induce erythroid differentiation, HUDEP-2 cells were seeded on OP9 cells for 6-7 days as previously described.25.
Codon optimized SF3B1WT or SF3B1K700E overexpression
As a human SF3B1 sequences are toxic to Escherichia coli48we used optimized codons SF3B1WT and SF3B1K700E expression vectors (Addgene Plasmid #82576 and #82577, respectively)49, encoded in pcDNA (Invitrogen, Carlsbad, CA). Each coding sequence was cloned into the retroviral vector pBABE-puro (Addgene Plasmid #1764)50.
For transient overexpression of SF3B1WT Where SF3B1K700Eeach pcDNA-based expression vector (10 μg) was transfected into K562 cells using Amaxa Cell Line Nucleofector II (Lonza, Cologne, Germany) with T-016 program27.47. For retroviral overexpression of SF3B1WT Where SF3B1K700E, pBABE-puro-based expression vector and VSV-G (Addgene plasmid #12259) were co-transfected into Plat-GP packaging cell lines (Cell Biolabs, San Diego, CA) using FuGene HD (Promega, Madison, WI). Seventy-two hours after transfection, the viral supernatant was used for infection. After spin infection of HUDEP-2 cells at 1300 × g for 2 h, 1 μg/mL puromycin (Sigma-Aldrich) was added to the medium to select transduced cells.
Primers used for amplification or detection of optimized codons SF3B1WT Where SF3B1K700E expression are shown online in Supplementary Table S9.
Detection of the SF3B1 mutation
Genomic DNA was extracted from whole BM lysates from MDS patients or cell lysates from SF3B1 mutant cell lines using a DNeasy Blood & Tissue kit (Qiagen NV, Hulsterweg, The Netherlands). Changes inside SF3B1 exons 14 to 16, where most SF3B1 mutations exist, were screened using high-resolution fusion analysis, as previously reported51, followed by confirmation of mutations with Sanger sequences when screening tests were positive. Primer sequences used for Sanger sequencing are listed online in Supplementary Table S9.
shRNA-mediated ABCB7 knockdown in HUDEP-2 cells
Reversal of human-based lentivirus ABCB7 The gene was made with the lentiviral shRNAmir pGIPZ (Clone ID: V3LHS_406787) (Open Biosystems, Huntsville, USA), as previously described47. The lentiviral vectors VSV-G and psPAX2 (Addgene plasmid #12260) were co-transfected into HEK293T cells; 72 h after transfection, the viral supernatant was used for infection, as in the retroviral overexpression protocol.
Quantitative RT-PCRs were performed as previously described27. Primers used for quantitative RT-PCR are listed online in Supplementary Table S9.
Expression profiling analysis
For RNA-seq analysis, total RNA was extracted from whole BM cells of MDS patients (see Supplementary Table S8 online) or from cell pellets when analyzing cell lines using the TRIzol reagent. For library preparation, SMARTer Ultra Low RNA Kit (Illumina, San Diego, CA) and Illumina TruSeq Stranded mRNA Library Kit were used for clinical samples and NEBNext RNA Library Preparation Kit Ultra II for Illumina was used for cell lines. Libraries were sequenced on an Illumina NovaSeq6000 (Otogenetics, Norcross, GA, USA). Sequence data was mapped to the human reference genome, hg19/GRCh37, using HISAT2 (version 2.2.1) (http://daehwankimlab.github.io/hisat2/)52. The normalized expression level of each gene was calculated as transcripts per million using StringTie (version 2.1.7) (https://ccb.jhu.edu/software/stringtie/)53.
Microarray analysis was performed using a Human Oligo 25k chip (Toray, Tokyo, Japan), and subsequent GO enrichment analyzes were performed using Metascape38. For overall normalization, the background value was subtracted and then adjusted to an average signal value of 25, and genes with >100 were analyzed.
DEG analysis for the GSE114922 dataset was performed with the web tool, iDEP.951 (http://bioinformatics.sdstate.edu/idep/)37 based on standardized metering data obtained from GREIN. We selected the DESeq2 method for DEG identification by setting the false discovery rate to 0.05 and a minimum fold change to 1.5.
Alternative splicing analysis
AS events from each sample were analyzed using MISO software (version 0.5.4) (https://miso.readthedocs.io/en/fastmiso/) with exoncentric analysis separately for A3SS, alternative splice site 5′ (A5SS), mutually exclusive exons (MXE), retained introns (RI) and skipping exons (SE)34. Human Genome Alternative Events (hg19) v2.0 (https://miso.readthedocs.io/en/fastmiso/annotation.html) was used for MISO annotation. Among the AS events extracted by the “compare miso” command, we considered significant the AS events passing all the filtering criteria, which in our study is the following: (a) at least one inclusion reading and one reading of exclusion, such that (b) the sum of the inclusion and exclusion readings is at least 10, (c) the Δ PSI (percent splicing) is ≥ 0.20 and (d) the factor Bayes is at least 10. AS events were expressed with read coverage visualized using IGV32 or the Sashimi plot command in IGV or MISO.
As splice variant isoforms containing premature stop codons (PTCs) are targeted by NMD, which decreases the number of splice variant isoforms54, the less expressed splice variant isoforms are sometimes difficult to detect. Cycloheximide (Nacalai Tesque, Inc., Japan), an NMD inhibitor, was added to HUDEP-2 cells stably expressing SF3B1WT or SF3B1K700E at a final concentration of 100 ug/mL19.
Production of cytospin slides and staining
Preparation and staining of Cytospin with May–Grünwald–Giemsa stain (Merck KgaA, Darmstadt, Germany) or Prussian blue (ScyTek Laboratories, Inc., West Logan, UT) was performed as previously described.25. RS was defined as erythroblasts with ≥ five iron granules surrounding at least one-third of the nuclear lesion55.
An electron microscope (H-7600; Hitachi) was used. Sample preparation protocols were described by Saito et al.25.
Western blot analysis
Western blot was performed using whole cell extracts as previously described25. The expression level was quantified by densitometry with ImageJ56, with respect to the expression level of the field. The main antibodies used were: α-tubulin (CP06; EMD Millipore, Billerica, MA, USA), SF3B1 (27684-1-AP; Proteintech, Rosemont, USA), ALAS2 (ab184964; Abcam, Cambridge , UK), ABCB7 (LS-B13035, Lifespan BioSciences, Seattle, WA, USA), MAP3K7 (#5206, Cell Signaling Technology, Danvers, MA, USA), RNH1 (10345-1-AP, Proteintech, USA), FLAG (#14793, Cell Signaling Technology) and GATA-1 (#3535, Cell Signaling Technology).
All statistical analyzes were performed using GraphPad Prism 9 (GraphPad Software, San Diego, CA, USA, http://www.graphpad.com/). Nonparametric analysis was adopted due to the small sample size. The Mann-Whitney U test was used to compare two groups and the Kruskal-Wallis test to compare equal to or greater than three groups, followed by Dunn’s test. Statistical significance was set at p