RESEARCH ARTICLE Targeting the EIF2AK1 Signaling Pathway Rescues Red Blood Cell Production in SF3B1- Mutant Myelodysplastic Syndromes With Ringed Sideroblasts Vera Adema1, Feiyang Ma2, Rashmi Kanagal-Shamanna3, Natthakan Thongon1, Guillermo Montalban-Bravo1, Hui Yang1, Scott A. Peslak4,5, Feng Wang6, Pamela Acha7, Francesc Sole7, Pamela Lockyer1, Margherita Cassari8, Jaroslaw P. Maciejewski9, Valeria Visconte9, Irene Gañán-Gómez1, Yuanbin Song10, Carlos Bueso-Ramos3, Matteo Pellegrini11, Tuyet M. Tan12, Rafael Bejar12, Jennifer S. Carew13, Stephanie Halene14, Valeria Santini8, Gheath Al-Atrash15, Karen Clise-Dwyer15, Guillermo Garcia-Manero1, Gerd A. Blobel5, and Simona Colla1 ABSTRACT SF3B1 mutations, which occur in 20% of patients with myelodysplastic syndromes (MDS), are the hallmarks of a specific MDS subtype, MDS with ringed sideroblasts (MDS-RS), which is characterized by the accumulation of erythroid precursors in the bone marrow and primarily affects the elderly population. Here, using single-cell technologies and functional validation studies of primary SF3B1-mutant MDS-RS samples, we show that SF3B1 mutations lead to the activa- tion of the EIF2AK1 pathway in response to heme deficiency and that targeting this pathway rescues aberrant erythroid differentiation and enables the red blood cell maturation of MDS-RS erythroblasts. These data support the development of EIF2AK1 inhibitors to overcome transfusion dependency in patients with SF3B1-mutant MDS-RS with impaired red blood cell production. SIGNIFICANCE: MDS-RS are characterized by significant anemia. Patients with MDS-RS die from a shortage of red blood cells and the side effects of iron overload due to their constant need for transfu- sions. Our study has implications for the development of therapies to achieve long-lasting hematologic responses. INTRODUCTION in dysfunctional hemoglobin synthesis and RS formation (8–12). However, given that few studies have elucidated how Cancer-causing mutations can dysregulate transcriptional SF3B1MT affect distinct transcriptional states in erythroid programs and cause mutant cells to become dependent on progenitors and precursors, the dysregulated transcriptional certain molecular regulators or signaling pathways. The regu- programs induced by iron accumulation and the ways by lators or pathways that cancer cells become reliant upon vary which these programs arrest erythroid differentiation are based on the cell type and the dynamics of the cell states largely unknown. affected by the mutations (1). Here, using single-cell technologies, we sought to elu- Some of the best examples of hematologic alterations cidate how SF3B1MT disrupt the transcriptional states of that can induce specific cell identity–based morphologic and MDS-RS hematopoietic stem and progenitor cells (HSPC) clinical phenotypes are hotspot mutations in the splicing and downstream differentiating erythroblasts to find disease factor SF3B1 (SF3B1MT). SF3B1MT are highly recurrent initiat- dependencies that are not predicted by SF3B1MT-induced ing events in myelodysplastic syndromes (MDS), arising in aberrant splicing. the most primitive hematopoietic stem cells (HSC; ref. 2). Although found in different hematopoietic cell types (3), SF3B1MT preferentially deregulate erythroid progenitors and RESULTS are hallmarks of a subtype of MDS characterized by the accumulation of ringed sideroblasts (RS) in the bone marrow SF3B1MT Do Not Impair Erythropoiesis at the Level (BM), MDS-RS (4–7). of HSPCs Several studies have dissected the molecular mechanisms To dissect at the single-cell level the cellular and transcrip- by which SF3B1MT modulate RNA splicing in MDS-RS. The tomic changes induced by SF3B1MT in cells undergoing eryth- SF3B1MT-induced aberrant splicing of genes involved in heme roid differentiation, we performed single-cell RNA-sequencing biosynthesis and mitochondrial iron transport leads to the (scRNA-seq) analysis of lineage-negative (Lin−) CD34+ HSPCs abnormal deposition of iron in erythroid cells, resulting isolated from the BM of five untreated SF3B1-mutant MDS-RS 1 Department of Leukemia, The University of Texas MD Anderson Cancer of California, Los Angeles, California. 12Moores Cancer Center, Univerity Center, Houston, Texas. 2Division of Rheumatology, Department of Internal of California San Diego, San Diego, California. 13University of Arizona Medicine, Michigan Medicine, University of Michigan, Ann Arbor, Michigan. Cancer Center, Tucson, Arizona. 14Section of Hematology, Department of 3 Department of Hematopathology, The University of Texas MD Anderson Internal Medicine and Yale Comprehensive Cancer Center, Yale University Cancer Center, Houston, Texas. 4Division of Hematology/Oncology, School of Medicine, New Haven, Connecticut. 15Department of Stem Department of Medicine, Hospital of the University of Pennsylvania, Cell Transplantation and Hematopoietic Biology and Malignancy, The Philadelphia, Pennsylvania. 5Division of Hematology, Children’s Hospital University of Texas MD Anderson Cancer Center, Houston, Texas. of Philadelphia, Philadelphia, Pennsylvania. 6Department of Genomic V. Adema and F. Ma contributed equally to this article. Medicine, The University of Texas MD Anderson Cancer Center, Houston, Texas. 7MDS Research Group, Josep Carreras Leukaemia Research Corresponding Author: Simona Colla, MD Anderson Cancer Center, Depart­ Institute, Universitat Autonoma de Barcelona, Badalona, Spain. 8MDS ment of Leukemia, 1500 Holcombe Street, Houston, TX 77030. Phone: Unit, Azienda Ospedaliero Universitaria Careggi, University of Florence, 713-563-5312; E-mail: scolla@mdanderson.org Florence, Italy. 9Department of Translational Hematology and Oncology Blood Cancer Discov 2022;3:554–67 Research, Taussig Cancer Institute, Cleveland Clinic Foundation, Cleveland, doi: 10.1158/2643-3230.BCD-21-0220 Ohio. 10Department of Hematologic Oncology, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer This open access article is distributed under the Creative Commons Attribution- Medicine, Sun Yat-sen University Cancer Center, Guangzhou, P.R. China. NonCommercial-NoDerivatives 4.0 International (CC BY-NC-ND 4.0) license. 11 Department of Molecular, Cell and Developmental Biology, University ©2022 The Authors; Published by the American Association for Cancer Research NOVEMBER 2022 BLOOD CANCER DISCOVERY | 555 RESEARCH ARTICLE Adema et al. patients and two age-matched healthy donors (HD; Sup- in defining how cells are poised for further differentiation plementary Table S1). Our analyses identified 16 cellular (18), these data demonstrate that SF3B1MT do not arrest clusters driven by cellular differentiation profiles (Fig. 1A) erythropoiesis at the HSPC level but instead induce HSC that we defined based on the differential expression of vali- metabolic activation and promote differentiation toward the dated lineage-specific transcriptional factors (TF) and cellular erythroid lineage at the expense of megakaryopoiesis. markers (refs. 13–16; Supplementary Fig. S1A; Supplemen- tary Table S2). To dissect the changes in lineage specifica- SF3B1MT Arrest Erythroblasts’ Terminal tion induced by SF3B1MT, we compared the distribution of Differentiation and Activate the EIF2AK1-Induced HD and SF3B1-mutant HSPCs in the distinct hematopoietic Response Pathway to Heme Deficiency clusters and found that HSPCs from MDS-RS patients had To evaluate whether SF3B1MT affects terminal erythroid a predominant erythroid/megakaryocytic (Ery/Mk) differen- differentiation, we performed scRNA-seq analysis of mono- tiation (Fig. 1B). Importantly, MDS-RS Ery/Mk cells had nuclear cells (MNC) isolated from the BM of the same five higher expression of KLF1 (P = 0) and lower expression of SF3B1-mutant MDS-RS patients and two HDs whose HSPC FLI1 (P = 4.68−36), the two major regulators of Ery and Mk compartments we analyzed previously. Our analysis identified differentiation, respectively (17), which suggests that these 20 cellular clusters inclusive of all major BM cell types, which cells were more erythroid-biased than those from HDs were we defined according to known cellular surface markers and (Supplementary Fig. S1B). Differential expression analyses lineage potential (ref. 19; Fig. 2A; Supplementary Fig. S2A; revealed that SF3B1-mutant HSPCs undergoing Ery/Mk dif- Supplementary Table S5). As expected, differential analysis ferentiation, compared with their healthy counterparts, exhib- of BM cell type composition revealed that SF3B1-mutant ited upregulation of genes in the MYC signaling and oxidative MNCs, compared with HD MNCs, were characterized by an phosphorylation pathways (Fig. 1C; Supplementary Fig. S1C erythroid predominance (Supplementary Fig. S2B), which is and S1D), which underscores the state of metabolic activation consistent with the accumulation of RS in the BM of these in SF3B1-mutant cells. Similar results were observed when we patients (Supplementary Table S1). To determine the mecha- analyzed the expression profile of most immature hematopoi- nistic basis by which SF3B1MT affect erythroid maturation and etic cells (cluster 1; Supplementary Fig. S1E and S1F). impair the release of red blood cells into the peripheral blood To further evaluate whether the fate of SF3B1-mutant Ery/ (PB), we analyzed the frequency and expression profile of each Mk cells was biased toward erythroid differentiation at the erythroid cluster within the erythroblastic population. We expense of megakaryopoiesis, we performed single-cell assays identified 8 cellular clusters representing 7 different stages of for transposase-accessible chromatin with high-throughput erythroblast differentiation (Fig. 2B), which we defined based sequencing (scATAC-seq) to profile the chromatin acces- on the expression of previously validated surface markers, sibility landscape in Lin− CD34+ HSPCs isolated from three TFs, and effector molecules involved in erythropoiesis (20, SF3B1-mutant MDS-RS patients and two age- and gender- 21). Cells in the early stages of erythropoiesis, including phe- matched HDs (Supplementary Table S1). We identified 8 notypic burst-forming unit-erythroid cells (BFU-E; cluster cellular clusters with distinct TF binding motif enrichment in 18) and colony-forming unit-erythroid cells (CFU-E; cluster open chromatin regions (Fig. 1D; Supplementary Table S3). 9), were characterized by variable expression of the TF GATA2 Cells in cluster 5 were in the earliest differentiation state and and the surface markers CD34, CD36, CD117 (cKit), and were characterized by the highest activities of TFs involved CD105 (ENG), whereas cells undergoing terminal differentia- in retinoid and nuclear factor family signaling (Supplemen- tion, including pro-erythroblasts (Pro-E; cluster 4), basophilic tary Fig. S1G). Conversely, cells in clusters 2 and 3 showed erythroblasts (Baso-E; cluster 12), polychromatophilic eryth- the highest activities of TFs involved in Ery/Mk specifica- roblasts (Poly-E; cluster 7), and orthochromatic erythroblasts tion, including many members of the GATA family of TFs (Ortho-E; clusters 1 and 11), were characterized by progres- (Fig. 1E). Cells in clusters 1 and 4 were myeloid-biased and sively increased expression of the TFs GATA1 and KLF1, the had the highest activities of the TFs SPI1 (PU.1) and CEBPA erythropoietin and transferrin receptors (EPOR and TFRC, (Supplementary Fig. S1H). Differential analysis of chromatin i.e., CD71), the membrane protein glycophorin A (GYPA, i.e., accessibility among the samples showed that MDS-RS HSPCs CD235a), and hemoglobin subunits. Pre-erythrocytes (Pre-E; at the latest stage of Ery/Mk differentiation before line- cluster 13)—erythroblasts in the reticulocyte maturation age specification (cluster 2) had significantly decreased open phase, which occurs before the release of erythrocytes into chromatin peaks at the distal elements of genes involved the PB—expressed genes involved in hemoglobin synthe- in platelet production (hemostasis; Fig. 1F; Supplementary sis most highly and surface markers that characterize the Table S4). Because distal elements are better than promotors upstream stages of erythroblast differentiation most lowly Figure 1. SF3B1MT do not impair erythropoiesis at the level of HSPCs. A, UMAP plots of scRNA-seq data displaying pooled single Lin-CD34+ cells isolated from two HD (n = 2,324) and five SF3B1-mutant MDS-RS (n = 5,544) samples. Each dot represents one cell. Different colors represent the sample origin (top) and cluster identity (bottom). B, Distribution of HD (left) and MDS-RS (right) Lin-CD34+ cells among the clusters shown in A defined by distinct lineage differentiation profiles. Prolif, proliferative. C, Pathway enrichment analysis of the genes that were significantly upregulated in the MDS-RS Ery/Mk clusters as compared with the HD Ery/Mk clusters shown in A (adjusted P ≤ 0.05). The top 10 Hallmark gene sets are shown. D, UMAP plots of scATAC-seq data for pooled Lin-CD34+ cells isolated from 2 HD (n = 1,844) and 3 SF3B1-mutant MDS-RS (n = 5,203) samples. Each dot represents one cell. Different colors represent the sample origin (left) and cluster identity (right). E, Violin plots showing the activities of the TFs GATA1, GATA2, and GATA3 across the 8 clusters shown in D. F, Pathway enrichment analysis of the genes whose distal elements were enriched in open chromatin regions in HD cells from cluster 2 shown in D as compared with those of MDS-RS cells (adjusted P ≤ 0.05). The top 10 Reactome gene sets are shown. Baso, basophilic; DC, dendritic cell; Ery/Mk, erythroid/megakaryocytic; Granulo, granulocytic; HSC, hematopoietic stem cells; Mono, monocytic; MPP, multipotent progenitors; Myelo, myeloid; Prog, progenitors; UMAP, uniform manifold approximation and projection. 556 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE A B Er gh p Pro Ery Ery Prog HDs (hi /Mk olif) Er y P g y/M ro Er w p Prog (lo MDS-RS [1] 14% k P lif) y [1] 11% rog [2] 14% ro E [2] 10% r [3] 7% (h ry/M [3] 9% Er y ig k g HSC [10] 10% h HSC [10] 2% MK pr Pro Pro [0] 10% E ol g [0] 12% Pr ry/M if) [6] 6% (low k g og Pro MPP [6] 6% DC [13] 6% p r olif g Prog MPP ) [13] 3% Baso [12] 3% [12] 4% Prog [11] 6% g MPP Prog MK Pro [11] 4% Granulo lls [14] 1% og [14] 3% Ly e s o lif c cell MPP [9] 8% P r Lympho Prog m r [9] 5% hp DC lls Mono g if ph Hig ol P g Pro [5] 4% Immature proli rog ulo ro lls r [5] 7% High prolif ce o p P UMAP 2 Pr Myelo f ce gh [7] 2% no so i [7] 7% og P H Ba Prog Mo [8] 4% rog Immature [8] 7% an [4] 2% [4] 9% Gr Myelo High Prog [15] 3% [15] 1% UMAP 1 HDs MDS-RS HSC 1 MPP C Baso Prog MPP MYC targets V1 2 Ery/Mk Prog 3 5 7 Oxidative phosphorylation 8 MTORC1 signaling 10 14 Adipogenesis Immature Myelo Prog 0 4 Fatty acid metabolism 13 MYC targets V2 12 E2F targets Mono Prog 6 15 DNA repair 9 DC Prog Unfolded protein response UMAP 2 11 Interferon gamma response Granulo Prog 0 15 30 45 60 75 90 UMAP 1 –Log10 (P) HSC D HDs 5 Lympho/Myelo Prog MDS-RS Myelo Prog 6 MPP Myelo Prog 7 3 0 Lympho/Myelo Prog 1 2 4 UMAP 2 UMAP 2 Ery/Mk Prog UMAP 1 UMAP 1 Myelo Prog E F GATA1 GATA2 Hemostasis 6.6 6.6 MAPK family signaling cascades Adaptive immune system TF activity TF activity Metabolism of lipids 6.3 6.3 Cell–cell communication Signaling by receptor tyrosin kinases 6.0 6.0 Signaling by RHO GTPases NEF and signal transduction 5.7 Signaling by the B cell receptor (BCR) 0 1 2 3 4 5 6 7 0 1 2 3 4 5 6 7 FCGR3A-mediated IL10 synthesis GATA3 0 2 4 6 8 [0] Myelo Prog –Log10 (P) 8.2 [1] Lympho/Myelo Prog [2] Ery/Mk Prog TF activity 8.0 [3] MPP [4] Myelo Prog 7.8 [5] HSC [6] Myelo Prog [7] Lympho/Myelo Prog 0 1 2 3 4 5 6 7 NOVEMBER 2022 BLOOD CANCER DISCOVERY | 557 RESEARCH ARTICLE Adema et al. A 5 B HDs MDS-RS B-Lympho Poly-E 16 7 7 1 11 Ortho-E HSPC 1 11 Ortho-E DC 19 15 Pro-E 13 13 18 BFU-E 18 Pre-E 4 12 4 12 17 2 Ery Baso-E 9 9 8 CFU-E 14 10 0 6 Myelo UMAP 2 UMAP 2 UMAP 2 3 T-Lympho/NK UMAP 1 UMAP 1 UMAP 1 C D Poly-E Ortho-E Cellular responses to stress HDs Response of EIF2AK1 (HRI) to heme deficiency Cell cycle BFU-E BFU-E (14%) Cytosolic tRNA aminoacylation Pro-E Ba Neddylation so CFU-E (53%) Transcriptional regulation by TP53 -E Pro-E (17%) DNA repair Pro-E Cellular senescence CFU-E Baso-E (11%) Vesicle-mediated transport G2–M transition Poly-E (3%) Ortho-E (2%) 0 10 20 30 40 Cellular responses to stress Metabolism of RNA Cell cycle Respiratory electron transport Baso-E Transcriptional regulation by TP53 Response of EIF2AK1 (HRI) to heme deficiency BFU-E MDS-RS Vesicle-mediated transport G2–M transition BFU-E (3%) Mitotic G1 phase and G1–S transition Pre- E CFU-E (9%) Chromosome maintenance U- E 0 10 20 30 40 50 60 CF Pro-E (16%) Pro-E Baso-E (10%) Eukaryotic translation initiation Ortho-E Ba Poly-E (14%) Response of EIF2AK1 (HRI) to heme deficiency so Antigen processing: ubiquitination & proteasome degradation Poly -E Ortho-E (38%) Autophagy Poly-E Pre-E (10%) Membrane trafficking E - Respiratory electron transport Transcriptional regulation by TP53 Cytosolic tRNA aminoacylation Cooperation of PDCL and TRiC/CCT Protein localization 0 20 40 60 80 −Log10 (P) E F 20 *** Autophagic vesicles Autophagy 15 Cellular responses to stress Vesicle-mediated transport per cell Mitochondrial iron–sulfur cluster biogenesis Ortho-E Chaperone mediated autophagy 10 Foxo-mediated transcription Neutrophil degranulation 5 SF3B1 WT Transcriptional regulation by TP53 Metabolism of amino acids and derivatives 500 nm Negative regulators of DDX58/IFIH1 signaling 0 0 2 4 6 8 10 T T M W −Log10 (P ) 1 1 3B 3B SF SF MT SF3B1 500 nm 558 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE (Supplementary Fig. S2C). Differential analysis of the cluster and LAMPTOR5 (Supplementary Fig. S2H). These data are distribution between HD and SF3B1-mutant samples showed consistent with an attempt by Ortho-E to remove dysfunc- that SF3B1-mutant erythroblasts had an increased frequency tional mitochondria, alleviate the iron overload caused by of cells at the latest stages of erythroid differentiation, mainly SF3B1MT (8, 30), and maintain their survival. at the Ortho-E stage (Fig. 2C). To exclude the possibility that To functionally validate these findings, we analyzed the mor- the arrest of the SF3B1-mutant erythroblasts in the late stages phology of erythroid cells isolated from the BM of SF3B1- of differentiation was the result of an increase in EPO in the mutant and SF3B1-wild-type (SF3B1WT) MDS-RS patients to BM of MDS-RS patients, we analyzed the changes induced assess differences in the ultrastructural features of autophagy. by EPO administration in the BM MNCs of one MDS-RS Transmission electron microscopy showed that SF3B1-mutant patient carrying the SF3B1K700E mutation before and after erythroblasts had significantly more autophagic vesicles than treatment with EPO as supportive therapy (Supplementary SF3B1WT erythroblasts do (Fig. 2F). To determine whether Fig. S2D; Supplementary Table S6). EPO administration did SF3B1MT instigated autophagy activation, we used CRISPR/ not induce any changes in either the frequencies of the eryth- Cas9 genome editing technology to introduce the SF3B1 mis- roid populations (Supplementary Fig. S2E) or these cells’ sense mutation c.2098A>G (K700E) into the erythroblastic cell transcriptional profile (Supplementary Table S7). line K562. This mutation results in the K700E amino acid sub- Together, these data suggest that SF3B1MT arrest erythro- stitution, which is recurrently detected in patients with SF3B1- poiesis at the final stage of erythroblast maturation by inhib- mutant MDS-RS (30). Consistent with our findings in primary iting the transition from Ortho-E to Pre-E, which results in samples, SF3B1-mutant cells in physiologic conditions or after the accumulation of terminally differentiated cells in the hemin-induced erythroid differentiation had EIF2AK1 pathway BM. Consistent with these observations, immunophenotypic upregulation (Supplementary Fig. S2I) and had significantly analysis showed that, compared with those from HDs, termi- more autophagic vesicles (Supplementary Fig. S2J and S2K) and nally differentiated erythroblasts from SF3B1-mutant MDS- higher expression of LC3B protein (Supplementary Fig. S2L) RS patients had a significantly higher frequency of cells in the than their parental counterparts did. Together, these data dem- orthochromatic stage V of differentiation (Supplementary onstrate that SF3B1MT arrest erythroid terminal differentiation Fig. S2F), as reported previously (22, 23). and activate the EIF2AK1-induced response pathway to main- To dissect the molecular mechanisms underlying the tain erythroblasts’ survival in response to heme deficiency. SF3B1MT-induced arrest of terminal erythroid differentiation, we analyzed the differential expression profiles between HD Hypomethylating Agent Therapy Inhibits and SF3B1-mutant cells at each stage of erythroblast differ- the EIF2AK1-Induced Response Pathway to entiation. These analyses revealed that SF3B1-mutant Pro- Heme Deficiency E, Baso-E, and Poly-E had significantly upregulated genes We hypothesized that if EIF2AK1 pathway activation plays involved in the response of eukaryotic translation initiation a role in arresting the maturation of terminally differen- factor 2 alpha kinase 1 (EIF2AK1) to heme deficiency, includ- tiated SF3B1-mutant erythroblasts, it should progressively ing EIF2AK1, EIF2S1, ATF4, and DDIT3 (Fig. 2D; Supplemen- decrease once patients achieve hematologic remission and tary Fig. S2G). EIF2AK1 is a metabolic stress-sensing kinase become transfusion-independent following hypomethylating that in homeostasis binds to its natural ligand, heme (24). agent (HMA) therapy. To test this hypothesis, we dissected at In conditions that limit heme production, EIF2AK1 phos- the single-cell level the cellular and transcriptomic changes phorylates its main target, EIF2S1, and leads to decreased induced by HMA therapy in Lin−CD34+ HSPCs and BM translation of globin mRNAs and increased expression of the MNCs, which were isolated from two SF3B1-mutant MDS-RS stress response effector ATF4 (25). ATF4 and its downstream patients at the time of diagnosis and at the time of hema- effector, DDIT3, are major regulators of autophagy (26) and tologic response following HMA therapy (Supplementary activate a set of targets involved in autophagy to maintain Table S1). Although HMA therapy did not overcome the cell survival in response to different types of stress stimuli, aberrant differentiation of HSCs toward the erythroid lineage including those affecting mitochondrial functions (27–29). (Fig. 3A; Supplementary Fig. S3A and S3B) or significantly Accordingly, we observed that SF3B1-mutant Ortho-E exhib- decrease the SF3B1MT burden in BM MNCs (Supplementary ited a significant upregulation of genes involved in mitochon- Fig. S3C; Supplementary Table S8), it did reduce the accu- drial autophagy (Fig. 2E), including PINK1, SQSTM1 (p62), mulation of erythroblasts in the BM (Fig. 3B; Supplementary Figure 2. SF3B1MT arrest erythroid terminal differentiation and activate the EIF2AK1-induced response pathway to heme deficiency. A, UMAP plots of scRNA-seq data for pooled single MNCs isolated from two HD (n = 5,049) and five SF3B1-mutant MDS-RS (n = 16,212) samples. Each dot represents one cell. Different colors represent the sample origin (left) and cluster identity (right). B, UMAP plot of scRNA-seq data from A showing the 7 different stages of erythroid differentiation in the total erythroblast population. C, Distribution of the 7 stages of erythroid differentiation in the HD (top) and MDS-RS (bottom) erythroblast populations shown in A. Arrows indicate the terminal steps of erythroid differentiation. D, Pathway enrichment analyses of the genes that were significantly upregulated in Pro-E (top), Baso-E (middle), Poly-E (bottom) from MDS-RS samples as compared with those from HD samples (P ≤ 0.01) The top 10 Reactome gene sets are shown. E, Pathway enrichment analysis of the genes that were significantly upregulated in Ortho-E from MDS-RS samples as compared with those from HD samples (P ≤ 0.01) The top 10 Reactome gene sets are shown. F, Left, the number of autophagic vesicles per cell in SF3B1WT and SF3B1-mutant erythroblasts from one representative SF3B1WT and one SF3B1-mutant sample. Statistically significant differences were detected using a two-tailed Student t test. Right, representative transmission electron microscopy images of erythroblasts from the BM sections of one SF3B1WT and one SF3B1-mutant MDS sample. Scale bars, 500 nm. Baso-E, basophilic erythroblasts; BFU-E, burst-forming unit-eryth- roid cells; B-Lympho, B-lymphocytes; CFU-E, colony formation unit-erythroid cells; DC, dendritic cells; Ery, erythroblasts; HSPC, hematopoietic stem and progenitor cells; Myelo, myeloid cells; Ortho-E, orthochromatic erythroblasts; Poly-E, polychromatophilic erythroblasts; Pre-E, pre-erythrocytes; Pro-E, pro-erythroblasts; T-Lympho/NK, T-lymphocytes, and natural killer cells; UMAP, uniform manifold approximation and projection. NOVEMBER 2022 BLOOD CANCER DISCOVERY | 559 RESEARCH ARTICLE Adema et al. A Diagnosis Response to HMA 7 Myelo Prog 3 Ery/Mk Prog Myelo Prog 1 Lympho Prog 8 0 MPP 5 6 2 Mk Prog UMAP 2 UMAP 2 4 HSC UMAP 1 UMAP 1 B Diagnosis Response to HMA Pre-E 13 12 2 Ortho-E Baso-E 5 6 Poly-E 0 Pro-E 3 T-Lympho/NK 4 BFU-E/CFU-E 11 1 HSPC 14 9 10 7 Mono 15 UMAP 2 UMAP 2 PC 8 B-Lympho UMAP 1 UMAP 1 C FU-E FU-E -E/C -E/C E E Pre- Pre- BFU BFU 7% 11% Pro-E 21% 27% Ortho-E 12% Ortho-E Pro-E 15% 16% 10% Baso-E 37% 29% 7% 8% Poly-E Poly-E Baso-E Diagnosis Response to HMA D Cellular responses to stimuli TCA cycle and respiratory electron transport Cytokine signaling in immune system Autophagy Transcriptional regulation by TP53 Response to EIF2AK1 (HRI) to heme deficiency Translation Synthesis of active ubiquitin: roles of E1 and E2 enzymes Protein localization Neutrophil degranulation 0 5 10 15 20 25 –Log10 (P) E Cellular responses to stimuli Cellular responses to chemical stress Translation Membrane traffiking Metabolism of RNA Transcriptional regulation by TP53 Respiratory electron transport Autophagy Neutrophil degranulation Cellular responses to starvation 0 10 20 30 40 50 –Log10 (P) 560 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE Fig. S3D and S3E; Supplementary Table S9) by inducing Supplementary Fig. S4D). These cells were characterized by these cells’ differentiation. Indeed, primary cells derived from the highest expression of both CD71 and CD235a (cluster SF3B1-mutant CD34+ MDS-RS cells cultured in erythroid 1) or CD235a (cluster 7; Supplementary Fig. S4E) and genes differentiation media for 8 days and then treated with 5-azac- involved in heme metabolism (Supplementary Fig. S4F; Sup- itidine for 3 days had a higher percentage of terminally dif- plementary Table S10). Differential expression analysis of these ferentiated CD235a (GYPA)+ erythroblasts than those treated clusters revealed that, as expected, EIF2AK1 sgRNA–treated with vehicle (21.9% vs. 7.62%; Supplementary Fig. S3F). MDS-RS erythroblasts were characterized by the downregula- Differential analysis of the cluster distribution in the eryth- tion of genes involved in the response of EIF2AK1 to heme roid compartment showed that, compared with those from deficiency (Fig. 4E), including the major effectors ATF4 and untreated patients, erythroblasts from HMA-treated patients DDIT3 (Supplementary Fig. S4G). EIF2AK1-mediated pathway who achieved hematologic remission had a decreased fre- downregulation was associated with the significantly increased quency of Ortho-E (Fig. 3C). Differential expression profiles of expression of genes that regulate heme metabolism and are untreated and HMA-treated cells at different stages of termi- involved in erythroid differentiation (Fig. 4F; Supplementary nal erythroid differentiation revealed that HMA-treated Pro-E, Fig. S4H). Similar results were obtained from bulk RNA-seq Baso-E, and Poly-E (clusters 3, 6, and 5, respectively) had sig- analysis of three other SF3B1-mutant MDS-RS samples treated nificant downregulation of the major effectors of the EIF2AK1 with NT or EIF2AK1 sgRNAs at 13 days of erythroid differen- response (Fig. 3D; Supplementary Fig. S3G). Accordingly, the tiation (Supplementary Fig. S4I). Notably, downregulation of expression of autophagy genes was significantly decreased in EIF2AK1 signaling was associated with an increased expression HMA-treated Ortho-E (clusters 2 and 12; Fig. 3E; Supplemen- of genes involved in heme biosynthesis and mitochondrial tary Fig. S3H). Together, these data suggest that HMA therapy iron transport, including all regulators of protoporphyrin IX can overcome EIF2AK1 pathway activation in terminally dif- biosynthesis, such as ALAD, ALAS2, HMBS, UROD, ABCB6, ferentiated erythroblasts in patients who have had hemato- CPOX, TMEM14C, PPOX, and FECH (Supplementary Fig. S4J). logic remission and have become transfusion-independent. Given that the missplicing of TMEM14C diminishes porphyrin abundance within the mitochondria, resulting in iron accu- Inhibition of EIF2AK1 Overcomes the Accumulation mulation and heme deficiency, and that the increased expres- of RS and Enables Red Blood Cell Production sion of TMEM14C partially overcomes RS formation (12), We hypothesized that the inhibition of EIF2AK1 pathway these results may explain why EIF2AK1-depleted erythroblasts activation induces the maturation of SF3B1-mutant erythro- undergo terminal differentiation. As the SF3B1MT burden did blasts. To test this hypothesis, we used CRISPR/Cas9 genome not differ between NT sgRNA– and EIF2AK1 sgRNA–treated editing technology to delete EIF2AK1 in primary erythroblasts MDS-RS cells (Supplementary Fig. S4K), these data suggest derived from a three-phase erythroid differentiation culture that EIF2AK1 inhibition overcomes the arrest of erythroid dif- system of SF3B1-mutant CD34+ cells (31). EIF2AK1 deple- ferentiation induced by SF3B1MT. tion in MDS-RS cells using previously validated single-guide RNAs (sgRNA; ref. 32; Fig. 4A) significantly increased ter- minally differentiated CD71 (TFRC)+CD235a (GYPA)+ and DISCUSSION CD71−CD235a+ erythroblasts at the expense of less-differen- Despite mechanistic insights into SF3B1MT-induced tum- tiated CD71+CD235a− cells after 13 days of culture (Fig. 4B; origenesis, effective therapies targeting SF3B1MT-induced Supplementary Fig. S4A). In contrast, EIF2AK1 depletion in aberrations in MDS have not yet been developed, and thera- HD CD34+ cells had no significant effect on erythroid dif- peutic approaches based on global splicing inhibition have ferentiation (Supplementary Fig. S4B and S4C). To validate fallen short of expectations. Results from a phase I clinical these results, we performed an scRNA-seq analysis of three trial of the oral SF3B1 modulator H3B-8800 in patients MDS-RS samples treated with nontargeting (NT) or EIF2AK1 with myeloid neoplasms did not recapitulate the results sgRNAs at day 13 of erythroid differentiation (Fig. 4C). Dif- of the preclinical studies performed using mouse models ferential analysis of the cluster distribution showed that, com- (33). These studies demonstrate that the widespread inhi- pared with those treated with the NT sgRNAs, cells treated bition of splicing activity does not yield significant clini- with EIF2AK1 sgRNAs had a significantly higher frequency cal improvements (34). Taken together, these data suggest of terminally differentiated cells (clusters 1 and 7; Fig. 4D; that alternative therapeutic approaches are needed and that Figure 3. Hypomethylating agent therapy inhibits the EIF2AK1-induced response pathway to heme deficiency in terminally differentiated cells in patients who became transfusion independent. A, UMAP plots of scRNA-seq data for pooled single Lin-CD34+ cells isolated from two SF3B1-mutant MDS-RS patients at the time of diagnosis (n = 2,372) and at the time of response to HMA therapy (n = 1,551). Each dot represents one cell. Different colors represent the sample origin (left) and cluster identity (right). B, UMAP plots of scRNA-seq data for pooled single MNCs isolated from three SF3B1-mutant MDS-RS patients at the time of diagnosis (n = 6,089) and response to HMA therapy (n = 6,156). Each dot represents one cell. Different colors represent the sample origin (left) and cluster identity (right). C, Distribution of the stages of erythroid differentiation in the total erythroblast population shown in B at the time of diagnosis (left) and at the time of response to HMA therapy (right). Arrows indicate Ortho-E. D, Pathway enrichment analysis of the genes in the MDS-RS Pro-E, Baso-E, and Poly-E clusters shown B that were significantly downregulated at the time of response to HMA therapy as compared with the time of diagnosis (adjusted P ≤ 0.05). The top 10 Reactome gene sets are shown. E, Pathway enrichment analysis of the genes in the MDS-RS Ortho-E shown in B that were significantly downregulated at the time of response to HMA therapy as compared with the time of diagnosis (P ≤ 0.01). The top 10 Reactome gene sets are shown. Baso-E, basophilic erythroblasts; BFU-E, burst-forming unit-erythroid cells; B-Lympho, B-lymphocytes; CFU-E, colony formation unit-erythroid cells; Ery/Mk, erythroid/megakaryocytic; HSC, hematopoietic stem cells; HSPC, hematopoietic stem and progenitor cells; Lympho, lymphoid; Mk, megakaryocytic; Mono, monocytes; MPP, multipotent progenitors; Myelo, myeloid; NK, natural killer cells; Ortho-E, orthochromatic erythroblasts; PC, plasma cells; Poly-E, polychromatophilic erythroblasts; Pre-E, pre-erythrocytes; Pro-E, pro-erythroblasts; Prog, progenitors; T-Lympho, T-lymphocytes; UMAP, uniform manifold approximation and projection. NOVEMBER 2022 BLOOD CANCER DISCOVERY | 561 RESEARCH ARTICLE Adema et al. A sgRNA B NT EIF2AK1 80 * 100 ** 4 * % of CD71 CD235a % of CD71 CD235a % of CD71 CD235a Vinculin 80 3 EIF2AK1 60 60 cells cells cells 40 2 Phospho-EIF2S1 40 20 20 1 LC3B 0 0 0 T 1 T 1 T 1 AK AK AK N N N F2 F2 F2 C EI EI EI NT sgRNA 9 EIF2AK1 sgRNA 8 5 7 4 3 10 0 6 2 1 UMAP 2 UMAP 2 UMAP 1 UMAP 1 D 0 (20%) 0 (13%) 1 (6%) 1 (21%) 10 2 (16%) 2 (11%) CD71 9 9 10 0 CD235a 8 0 3 (14%) 8 3 (7%) 7 CD71 CD71 4 (10%) CD71 4 (8%) 6 1 CD235a 7 1 CD235a CD235a 5 (11%) 5 (7%) 5 6 (6%) 6 6 (7%) 2 5 4 3 7 (3%) 4 2 7 (10%) 3 8 (6%) 8 (6%) NT sgRNA 9 (6%) EIF2AK1 sgRNA 9 (6%) 10 (3%) 10 (4%) E Peptide chain elongation Peptide chain elongation Cell cycle, mitotic Cell cycle Cytosolic tRNA aminoacylation mRNA splicing–major pathway Response of EIF2AK1 (HRI) to heme deficiency TCA cycle and respiratory electron transport Cellular senescence Cytosolic tRNA aminoacylation APC/C-mediated degradation of cell cycle proteins Chromosome maintenance Gluconeogenesis DNA strand elongation Metalloprotease DUBs Response of EIF2AK1 (HRI) to heme deficiency Factors involved in megakaryocyte development DNA repair G2–M transition mRNA splicing–minor pathway 0 20 40 60 80 100 0 20 40 60 80 100 −Log10 (P ) −Log10 (P ) F MYC targets V1 Heme metabolism UV response Up TNFA signaling via NFKB MTORC1 signaling P53 pathway Apoptosis Adipogenesis Heme metabolism Interferon alpha response P53 pathway Interferon gamma response Oxidative phosphorylation Androgen response TNFA signaling via NFKB IL2 STAT5 signaling Reactive oxygen species pathway Hypoxia DNA repair Xenobiotic metabolism 0 5 10 15 20 25 0 5 10 15 −Log10 (P ) −Log10 (P ) Figure 4. Inhibition of EIF2AK1 overcomes the accumulation of RS and enables red blood cell production. A, Representative western blot analysis of EIF2AK1, phospho-EIF2S1, and LC3B in nontargeting (NT) sgRNA– and EIF2AK1 sgRNA–treated SF3B1-mutant MDS-RS cells at day 13 of culture. Vinculin was used as a loading control. B, Frequencies of CD71+CD235a− (left), CD71+CD235a+ (middle), and CD71−CD235a+ (right) erythroblasts in NT sgRNA– or EIF2AK1 sgRNA–treated MDS-RS samples (n = 6) at day 13 of culture. Each symbol represents one sample; lines connect paired samples. Statistical signifi- cance was calculated using paired t tests. C, UMAP plots of scRNA-seq data for single cells from NT sgRNA–treated (n = 30,307) or EIF2AK1 sgRNA–treated (n = 36,825) cells from 3 pooled SF3B1-mutant MDS-RS samples at day 13 of erythroid culture. Each dot represents one cell. Different colors represent the sample origin (left) and cluster identity (right). Dotted lines indicate terminally differentiated erythroblasts. D, Distribution of cells from NT (left) and EIF2AK1 (right) sgRNA–treated SF3B1-mutant MDS-RS samples among the clusters shown in C. Arrows indicate clusters 1 and 7. E, Pathway enrichment analyses of the significantly downregulated genes in EIF2AK1 sgRNA–treated SF3B1-mutant MDS-RS cells from clusters 1 (left) and 7 (right; adjusted P ≤ 0.05). The top 10 Reactome or Hallmark gene sets are shown. F, Pathway enrichment analyses of the significantly upregulated genes in EIF2AK1 sgRNA– treated SF3B1-mutant MDS-RS cells from clusters 1 (left) and 7 (right; adjusted P ≤ 0.05). The top 10 Hallmark gene sets are shown. 562 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE EIF2AK1 Heme biosynthesis BFU-E SF3B1MT SF3B1WT CFU-E Pro-E RS survival Heme Heme deficiency Hemoglobin Baso-E Poly-E EIF2AK1 EIF2AK1 Ortho-E Globin chains Active Inactive P ATF4 Globin mRNA eIF2a eIF2a Autophagy Differentiation Heme biosynthesis genes and mitochondrial iron transporters Figure 5. Proposed model. Under normal physiologic conditions, heme binds to EIF2AK1 and represses its activation. In conditions that limit heme production, such as those induced by SF3B1MT in terminally differentiated erythroblasts, EIF2AK1 is activated. EIF2AK1 pathway activation promotes RS survival and inhibits erythroid maturation by increasing the expression of ATF4, which in turn upregulates the expression of genes involved in autophagy, transcriptionally inhibits genes involved in heme biosynthesis and mitochondrial iron transport, and translationally inhibits globin production. Targeting EIF2AK1 pathway activation by depleting EIF2AK1 rescues erythroid differentiation and red blood cell production. targeting specific downstream effectors of SF3B1MT in eryth- RS formation in our functional validation studies owing to roblasts may represent a novel strategy to overcome ineffec- the limited number of primary cells available for our assays. tive erythropoiesis before other secondary genetic lesions lead Growing evidence suggests that the EIF2AK1 pathway to MDS progression. affects erythropoiesis in response to several types of insults. In our study, we used single-cell technologies to elucidate The CRISPR/Cas9-mediated deletion of EIF2AK1 in sickle the biological mechanisms of the SF3B1MT-induced arrest of erythroblasts increases fetal hemoglobin production and erythropoiesis. We demonstrated that SF3B1MT significantly reduces sickling in cultured erythroid cells (31, 32). EIF2AK1- inhibit the differentiation of Ortho-E to Pre-E, thus impair- mediated fetal hemoglobin regulation proceeds via the modu- ing the release of red blood cells into the PB. We analyzed the lation of the fetal hemoglobin repressor BCL11A (36). A molecular pathways that were transcriptionally upregulated clinical study based on the transplantation of autologous in Ortho-E independently of splicing alterations and identi- HSCs transduced by lentiviral vectors delivering shRNAs tar- fied the EIF2AK1 response to heme deficiency as a potential geting BCL11A in patients with sickle cell disease demon- driver of the SF3B1MT-induced arrest of erythroid differen- strated a significant improvement of anemia and other disease tiation. EIF2AK1 pathway activation in terminally differ- manifestations (37), suggesting that pharmacologically inhib- entiated SF3B1-mutant erythroblasts resulted in increased iting of EIF2AK1 may lead to similar clinical outcomes. Thus, ATF4 expression, which in turn upregulated the expression EIF2AK1 inhibition represents one of the most promising of genes involved in autophagy and RS survival. We would, therapeutic approaches for sickle cell anemia to date. therefore, hypothesize that EIF2AK1 signaling exacerbates The EIF2AK1 effector DDIT3 is overexpressed in SF3B1WT RS formation and anemia in SF3B1-mutant MDS-RS patients MDS HSCs (38). DDIT3 overexpression delays the erythroid by reducing oxidative stress in response to heme deficiency differentiation of MDS CD34+ cells by affecting the activa- through the EIF2AK1-induced inhibition of globin protein tion of erythroid transcriptional programs. The inhibition of and biosynthetic heme enzyme expression and translation DDIT3 in CD34+ cells isolated from MDS patients with anemia (25, 35). Consistent with this hypothesis, EIF2AK1 inhibi- restores proper erythroid differentiation (38). Our study showed tion significantly increased the expression of both mito- that the activation of the EIF2AK1 pathway occurs in the later chondrial heme biosynthetic enzymes and iron transporters stages of erythroblast maturation, as SF3B1MT-induced aberrant and improved the terminal differentiation of SF3B1-mutant splicing primarily affected the function of terminally differenti- MDS-RS erythroblasts (Fig. 5). We could not directly assess ated erythroblasts with the highest mitochondrial content. NOVEMBER 2022 BLOOD CANCER DISCOVERY | 563 RESEARCH ARTICLE Adema et al. Although the EIF2AK1-mediated response to heme defi- number #194539, Sigma-Aldrich) treatment (1.5 μg/mL). Mutant ciency is not completely conserved in mice, mouse models in clones were screened by PCR using the following primers: 5′-GCATC which the gene Eif2ak1 has been knocked out have normal AAGCTTGGTACCGATATTCTTATGGGCTGTGCCATCTTGC-3′ hematologic indices (36). Moreover, EIF2AK1 depletion is and 5′-ACTTAATCGTGGAGGATGATAGGTAATTGGTGGATTTA CCTTTCCTCT3′. well tolerated in human erythroid cells in steady-state condi- SF3B1K700E mutation was confirmed by DNA sequencing. To tions (31, 32), as also confirmed in our study. These data sug- induce erythroid differentiation, parental and SF3B1K700E K562 cells gest that the EIF2AK1 pathway could be a therapeutic target were treated with 30 μmol/L hemin (catalog number # 51280, Sigma- for not only sickle cell disease but also MDS-RS or other types Aldrich) for 3 days, as described previously (23). of disorders with impaired red blood production. In conclusion, our study highlights the importance of dis- Fluorescence-Activated Cell Sorting secting the conserved molecular mechanisms of the erythroid FACS of Lin−CD34+ cells was performed using antibodies against stress response to develop novel approaches that achieve last- CD2 (RPA-2.10), CD3 (SK7), CD14 (MφP9), CD19 (SJ25C1), CD20 ing hematologic remission in patients with different types of (2H7), CD34 (581), CD56 (B159), and CD235a (HIR2; all from BD ineffective erythropoiesis or aberrant erythroblast functions. Biosciences); CD4 (S3.5), CD11b (ICRF44), and CD33 (P67.6; all from Thermo Fisher Scientific); CD7 (6B7; BioLegend); and CD10 (SJ5-1B4; Leinco Technologies), as we described previously. Samples METHODS were acquired with a BD Influx Cell Sorter (BD Biosciences). Human Primary Samples Cell Culture BM aspirates from 17 patients with SF3B1-mutant MDS-RS who MACS-purified CD34+ cells were subjected to a three-phase in vitro were referred to the Department of Leukemia at MD Anderson Cancer culture system as described previously (36). Briefly, for phase I medium, Center, the Cleveland Clinic Foundation, the University of California StemSpan SFEM II (catalog number #09605, Stem Cell Technologies) San Diego, or the Josep Carreras Leukaemia Research Institute were was supplemented with 100 ng/mL of human stem cell factor (catalog obtained after the approval of the corresponding Institutional Review number #300-07, PeproTech), 1 ng/mL of IL3 (catalog number #200- Boards and in accordance with the Declaration of Helsinki. Diagnosis 03, PeproTech), 3 units per mL of erythropoietin (catalog number of MDS-RS was assigned according to the World Health Organization #100-64, PeproTech), 200 μg/mL of holotransferrin (catalog number criteria (39). SF3B1MT were determined by a targeted amplicon-based #T4132, Sigma-Aldrich), 10 μg/mL of insulin (catalog number # I9278, next-generation sequencing (40). All available samples from patients Sigma-Aldrich), 5% human male A/B plasma (catalog number #H4522, with MDS-RS and SF3B1MT were included in the study. Baseline BM Sigma-Aldrich), and 10 μg/mL of heparin (catalog number #7980, aspirates were collected from patients before any treatment. For Stem Cell Technologies). For phase II medium, IL3 was withdrawn patients for whom baseline samples were available, sequential BM after 9 days of culture. For phase III medium, the cells were cultured samples were collected during HMA treatment. The clinical charac- with StemSpan SFEM II supplemented with 3 units per mL of eryth- teristics of the individuals with SF3B1-mutant MDS-RS are shown ropoietin, 1 mg/mL of holotransferrin, 10 μg/mL of insulin, 5% human in Supplementary Table S1. BM samples from HDs were obtained A/B plasma, and 10 μg/mL of heparin. Cells were maintained in phase from AllCells, the Department of Stem Cell Transplantation at MD I from the day of collection until day 8, at which time the cells under- Anderson Cancer Center, and the Cleveland Clinic Foundation. Writ- went ribonucleoprotein-based gene editing and were transitioned to ten informed consent was obtained from all donors. phase II medium. On day 13 of culture, the cells were transitioned to MNCs were collected from each BM sample immediately after BM phase III medium. Samples for western blot and flow cytometry analy- aspiration using the standard gradient separation approach with Ficoll- ses were harvested on day 13 or 15 of culture, respectively. Paque PLUS (catalog number #45-001-752, Thermo Fisher Scientific). To evaluate whether HMA therapy induces erythroblast differen- MNCs were cryopreserved and stored in liquid nitrogen until use. For tiation, erythroblasts derived from SF3B1-mutant CD34+ MDS-RS cell-sorting applications, MNCs were enriched in CD34+ cells using cells were cultured in phase I differentiation medium for 8 days and magnetic-activated cell sorting (MACS) with the CD34 Microbead Kit then treated with vehicle or 0.1 μmol/L 5-azacitidine (catalog number (catalog number #130-046-702, Miltenyi Biotec) and further purified #A2385, Sigma-Aldrich) for 3 days. Samples for flow cytometry analy- by fluorescence-activated cell sorting (FACS) as described below. sis were harvested on day 13 of culture. Generation of the SF3B1K700E Cell Line Ribonucleoprotein Electroporation in CD34+ Cells The K562 cell line was obtained by ATCC in 2017. Mycoplasma testing was routinely performed on the cell line, and cell identity Ribonucleoprotein-based gene editing was performed as described was validated by short tandem repeat DNA fingerprinting before previously (32). Briefly, CD34+ cells were electroporated with the pre- performing any experiment. Cells were cultured in 10% FBS (catalog viously validated EIF2AK1-targeting sgRNAs TTGTTGGCTATCACA number #S11150, Atlanta Biologicals) Iscove’s modified Dulbecco’s CCGCG and ATAGTCGAGAGAAACAAGCG or the nontargeting medium (catalog number #12440053, Thermo Fisher Scientific). Cells sgRNAs GCACTACCAGAGCTAACTCA and GTACGTCGGTATAACT for genome editing were used after 4 passages from reception. The CCTC, together with the Cas9 Nuclease V3 (catalog number #1081060, CRISPR/Cas9-based extreme genome editing system developed by Integrated DNA Technologies) using the P3 Primary Cell 4D-Nucleofec- Biocytogen (https://biocytogen.com/gene-editing) was used to gener- tor X Kit S (catalog number # V4XP-3032, Lonza). Chemically modified ate the heterozygous SF3B1K700E cell line. The nonconserved region sgRNAs were purchased from Synthego. of the SF3B1 intron 14 (near exon 15) was targeted by Cas9 endo- nuclease using the following guide RNA sequence: 5′-GGACAGCT Flow Cytometry GTCCTAAAATTT GGG-3′. The repaired template DNA included For erythroblast staging analysis, HD and SF3B1-mutant MDS-RS at least 1-kb 5′ and 3′ homologous arms, the puromycin gene, and BM MNCs were analyzed as described previously (22). Briefly, MNCs the K700E mutation in exon 15 (the sequence of the repaired DNA were stained with the following antibodies: PE-CD45 (clone HI30, template can be provided upon request). CRISPR plasmids and BD Biosciences), PE-CY7-CD235a (clone HI264, BioLegend), APC- template DNA (Biocytogen) were introduced into the K562 cells by CD49d (clone 9F10, BioLegend), and FITC–band 3 (a generous gift electroporation. Single clones were selected by puromycin (catalog from Dr. Xiuli An, New York Blood Cancer Center). 564 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE For erythroid differentiation analyses, cells on day 15 of culture cells and determine the marker genes for each cluster, we used the were stained with antibodies against PE-CD45, APC-CD71 (clone FindClusters and FindAllMarkers functions, respectively. Differential mA712, BD Biosciences), and FITC-CD235a (clone HIR2, BD Bio- expression analysis of the samples was performed using the Find- sciences). Dead cells were excluded with DAPI staining (catalog Markers function and the Wilcoxon rank-sum test. The Benjamini– number D1306, Thermo Fisher Scientific). Samples were acquired Hochberg procedure was applied to adjust the false discovery rate. with a BD LSRFortessa Cell Analyzer, and the cell populations were The normalized expression matrix for the scRNA-seq data was analyzed using FlowJo software (version 10.5.3, FlowJo, LLC). extracted and uploaded to the CytoTRACE web tool (https://cytotrace. stanford.edu/; ref. 41), and the output CytoTRACE score for each cell Western Blots was then plotted on the UMAP. Cells were washed with a 10% FBS/phosphate-buffered saline solu- tion (catalog numbers #16-140-07 and #14-190-250, respectively, Total RNA-seq Analysis Thermo Fisher Scientific), and pellets were resuspended in the Mam- Total RNA from NT sgRNA– or EIF2AK1 sgRNA–treated SF3B1- malian Cell and Tissue Extraction Kit buffer (catalog number #K269- mutant MDS-RS CD34+ cells at 13 days of erythroid differentiation 500, BioVision Incorporated) and incubated for 15 minutes with gentle was purified using the PicoPure RNA Isolation Kit (catalog number shaking. Lysates were then collected after centrifugation at 12,000 #KIT0204, Thermo Fisher Scientific). Quality checks and quantifica- rpm for 20 minutes at 4°C. The amount of protein was quantified tions were performed with the 2100 Expert Bioanalyzer (Agilent). using the Qubit Protein Assay Kit (catalog number #Q33212, Thermo Illumina low input total RNA libraries were prepared using the Fisher Scientific) and a Qubit Fluorometer (Thermo Fisher Scientific). SMARTer Stranded Total RNA-seq Kit v2–Pico Input Mammalian Sodium dodecyl sulfate–polyacrylamide gel electrophoresis and west- (catalog number #634412; Takara Bio). Briefly, 10 ng of total RNA ern blotting were performed following standard protocols. Blotted was converted to cDNA using Takara’s SMART (Switching Mecha- membranes were incubated with primary polyclonal antibodies against nism At 5′ End of RNA Template) technology. Illumina adapters EIF2AK1 (MBS2538144, Mybiosource), phospho-EIF2S1 (ab32157, with specific barcodes (Supplementary Table S11) were added to the Abcam), LC3B (2775, Cell Signaling), and vinculin (clone hVIN-1, double-stranded cDNA using 12 cycles of PCR. The PCR products Sigma-Aldrich), and with secondary anti-mouse and anti-rabbit digital were purified using Ampure Beads (Agencort); then, ribosomal cDNA antibodies (R1005 and R1006, respectively, Kindle Biosciences LLP). was specifically depleted using ZapR V2. The remaining cDNA frag- Membranes were developed using the SuperSignal West Pico PLUS ments were enriched by 12 cycles of PCR amplification using Illumina Chemiluminescent Substrate (catalog number #34580, Thermo Fisher universal library primers. The enriched library was purified to yield Scientific) in a KwikQuant Imager (Kindle Biosciences LLP). the final cDNA library. Libraries were combined and sequenced on the Illumina NextSeq500 sequencer using the High Output 150 cycle scRNA-seq Analysis flow cell with the 75nt PE format. The mapping of the RNA-seq reads to the human reference genome hg19 was performed with Tophat2. Lin−CD34+ and MNC samples were enriched by FACS. Sample prepa- The quality of the reads was verified using the FASTQC software ration and sequencing were performed at MD Anderson Cancer Center’s (v0.11.8). HTseq software (v0.11.2) was used to summarize the gene- Advanced Technology Genomics Core. Sample concentration and cell expression counts from the Tophat2 alignment data after sorting the suspension viability were evaluated using a Countess II FL Automated BAM files. Unsupervised analysis was performed using PCA and hier- Cell Counter (Thermo Fisher Scientific). Samples were normalized for archical clustering to verify any outliers and the overall similarities or input onto the Chromium Single-Cell A Chip Kit (catalog number differences among all samples. Differential expression analysis was #PN120236, 10X Genomics), and single cells were lysed and barcoded performed using the DEseq2 package, as we described previously (42). for reverse transcription. Equal amounts of each uniquely indexed sam- ple library were pooled together. Pooled libraries were sequenced using a NovaSeq6000 SP 100-cycle flow cell (Illumina). Sequencing analysis scATAC-seq Analysis was carried out using 10X Genomics’ Cell Ranger software, version 3.0.2. The scATAC Low Cell Input Nuclei Isolation protocol (10X The adapter trimming was done automatically using the Cell Genomics) was used to isolate nuclei from FACS-purified cells. The Ranger software. Alignment was performed using “Star” as a part of trypan blue exclusion assay was performed to check for intact nuclei. the Cell Ranger analysis. Cell Ranger default settings were used for The remaining extracted nuclei were used for the consecutive steps of all quality thresholds. The sequencing was run as a paired-end using the scATAC-seq library preparation protocol following 10X Genom- the following run format: 28 cycles Read 1, 8 cycles Index 1, 0 cycles ics guidelines. Equal molar concentrations of uniquely indexed sam- Index2, and 91 cycles Read2. Barcodes and unique molecular identi- ples were pooled together. Pooled libraries were sequenced using fiers were sequenced in Read1, and the transcriptional information a NextSeq500 150-cycle flow cell (Illumina). Reads were aligned was sequenced in Read2. The samples were single-indexed. to the human genome (hg38), and peaks were called using the After sequencing, raw reads were aligned to the human genome cellranger-atac count pipeline. Individual samples were merged using (hg38), and the digital expression matrix was generated using cell- the cellranger-atac pipeline to generate the peak-barcode matrix and ranger count. Individual samples were merged to generate the digital TF-barcode matrix. To identify specific TF activity for each cell cluster, expression matrix using cellranger aggr. The R package Seurat was we used the R package Seurat to analyze the TF-barcode matrix. used to analyze the digital expression matrix. Cells with fewer than The raw counts were normalized by the sequencing depth for each 100 genes and fewer than 500 unique molecular identifiers detected cell and scaled for each TF using the NormalizeData and ScaleData were removed from further analysis. The Seurat function Normalize- functions. PCA and UMAP were applied to reduce the dimensions of Data was used to normalize the raw counts. Variable genes were iden- the data, and the first two dimensions were plotted. The FindClusters tified using the FindVariableGenes function. The Seurat ScaleData function was used to cluster the cells. The FindAllMarkers function function was used to scale and center expression values in the data was used to determine the TF markers for each cluster. Differential set for dimensional reduction. Default parameters were used for the analysis of TF activity in the samples was performed using the Find- Seurat functions. When needed, samples were integrated using the Markers function and the Wilcoxon rank-sum test. Cluster identity Seurat functions FindIntegrationAnchors and IntegrateData. Princi- was determined based on the activity of master regulators of lineage pal component analysis (PCA) and uniform manifold approximation commitment, as we (43) and others (44, 45) described previously. and projection (UMAP) were used to reduce the dimensions of the Briefly, the HSC cluster was defined based on the lowest activity data, and the first two dimensions were used in plots. To cluster the of lineage-specific markers and highest activity of TFs that regulate NOVEMBER 2022 BLOOD CANCER DISCOVERY | 565 RESEARCH ARTICLE Adema et al. stemness, such as the retinoic acid receptor, NRF, and EGR families Authors’ Disclosures of TFs. Lymphoid and myeloid bipotent-biased clusters were defined J.P. Maciejewski reports grants from the Leukemia and Lymphoma based on the coactivities of major regulators of lymphoid differentia- Society, the U.S. Department of Defense, and NIH during the conduct tion (e.g., POU4F, TBX families of TFs) and myeloid differentiation of the study. R. Bejar reports other support from Aptose Biosciences, (RUNX1 and CEBPA); myeloid clusters were defined based on the personal fees from Gilead and Epizyme outside the submitted work. highest activities of the TFs SPI1 and CEBPA (master regulators of J.S. Carew reports other support from Majestic Therapeutics, LLC myelopoiesis); erythroid/megakaryocyte clusters were defined based outside the submitted work; in addition, J.S. Carew has a patent on the highest activity of the GATA family of TFs. 9926326 issued. S. Halene reports grants and other support from NIH Cluster-specific peaks were determined using the FindAllMarkers and other support from The Frederick A DeLuca Foundation during function, and differentially accessible peaks between the samples the conduct of the study; personal fees from FORMA Therapeutics were determined using the FindMarkers function. Each peak was outside the submitted work; in addition, S. Halene has a pending associated with a specific gene based on its distance to that gene’s patent. V. Santini reports personal fees from BMS, AbbVie, Geron, transcription start site (TSS). Peaks overlapping with a promoter Gilead, Takeda, Novartis, and Menarini outside the submitted work. region (−1,000 bp, +100 bp) of any TSS were annotated as peaks in K. Clise-Dwyer reports grants from NCI during the conduct of the promoters, whereas peaks not in promoter regions but within 200 study. No disclosures were reported by the other authors. kb of the closest TSS were annotated as peaks in the distal elements. Peaks not mapped in either the promoters or distal elements were Authors’ Contributions annotated as peaks in intergenic regions. V. Adema: Data curation, formal analysis, methodology. F. Ma: Data curation, software, methodology. R. Kanagal-Shamanna: Targeted Deep Sequencing Methodology. N. Thongon: Methodology. G. Montalban-Bravo: Selected samples were sequenced at MD Anderson’s Advanced Tech- Data curation. H. Yang: Methodology. S.A. Peslak: Supervision. nology Genomics Core facility. We used a custom SureSelect panel F. Wang: Software, formal analysis. P. Acha: Methodology. F. Sole: of 200 genes (Agilent Technologies) that are recurrently mutated in Supervision. P. Lockyer: Methodology. M. Cassari: Methodol- hematologic malignancies. DNA from MNCs was fragmented and ogy. J.P. Maciejewski: Supervision. V. Visconte: Methodology. bait-captured in solution according to the manufacturer’s proto- I. Ganan-Gomez: Methodology. Y. Song: Resources, formal analysis. cols. Captured DNA libraries were then sequenced using a NovaSeq C. Bueso-Ramos: Supervision. M. Pellegrini: Supervision. T.M. Tan: sequencer (Illumina) with 76-bp paired-end reads. Raw sequencing Methodology. R. Bejar: Supervision. J.S. Carew: Methodology. data from the Illumina platform were converted to fastq format and S. Halene: Supervision. V. Santini: Supervision. G. Al-Atrash: Super- aligned to the hg19 reference genome using the Burrows-Wheeler vision. K. Clise-Dwyer: Methodology. G. Garcia-Manero: Resources, Aligner. The aligned BAM files were subjected to mark duplication, supervision. G.A. Blobel: Supervision. S. Colla: Conceptualization, realignment, and recalibration using Picard and GATK (https:// resources, supervision, investigation, writing–original draft, writing– www.broadinstitute.org/gatk/guide/best-practices?bpm=DNAseq). review and editing. Preprocessed BAM files were then analyzed to detect single-nucleotide variants and small indels using the MuTect and Pindel algorithms, Acknowledgments respectively, against a virtual normal sequence developed in-house. This work was supported by philanthropic contributions to The University of Texas MD Anderson AML/MDS Moon Shot, by the NIH Transmission Electron Microscopy through MD Anderson’s Leukemia SPORE grant (P50 CA100632), by Fixed MNC samples were washed in 0.1 mol/L sodium cacodylate a Ladies Leukemia League grant to R. Kanagal-Shamanna, and by an buffer (catalog number #RT12310, Electron Microscopy Sciences) and NIH/NIDDK 5U54DK106857 grant to the Animal Modeling Core treated with 0.1% Millipore-filtered cacodylate buffered tannic acid of the Yale Cooperative Center of Excellence in Hematology. S. Colla (catalog number #RT21700, Electron Microscopy Sciences), postfixed is a Scholar of the Leukemia and Lymphoma Society. N. Thongon with 1% buffered osmium tetroxide (catalog number #RT19172, Elec- was supported by the 2020 ASH restart award. This work used MD tron Microscopy Sciences), and stained with 1% Millipore-filtered ura- Anderson’s South Campus Flow Cytometry and Cellular Imaging nyl acetate (catalog number #RT22400, Electron Microscopy Sciences). Core, the Advanced Technology Genomics Core, and the High- The samples were dehydrated in increasing concentrations of ethanol, Resolution Electron Microscopy Facility, which are supported in part and then infiltrated and embedded in LX-112 medium (catalog number by the NIH (NCI) through MD Anderson’s Cancer Center Support #LX112-21212, Ladd Research Industries). The samples were polymer- Grant (P30 CA16672). The authors thank Joseph Munch and Kelly ized in a 60°C oven for approximately 3 days. Ultrathin sections were cut Soltysiak for assistance with manuscript editing. in an Ultracut microtome (Leica), stained with uranyl acetate and lead citrate (catalog number #RT17800, Electron Microscopy Sciences), and Note examined with a JEM 1010 transmission electron microscope (JEOL) at Supplementary data for this article are available at Blood Cancer an accelerating voltage of 80 kV. Digital images were obtained using an Discovery Online (https://bloodcancerdiscov.aacrjournals.org/). AMT Imaging System (Advanced Microscopy Techniques Corp). Received November 29, 2021; revised April 26, 2022; accepted July Statistical Analysis 29, 2022; published first August 3, 2022. Flow cytometry data were analyzed with Prism 8 software (Graph- Pad). The figure legends indicate the statistical test(s) used in each REFERENCES experiment. Statistical significance was represented as *, P < 0.05; **, 1. Bradner JE, Hnisz D, Young RA. Transcriptional addiction in cancer. P < 0.01; ***, P < 0.001; ****, P < 0.0001. Figure 5 was made using Cell 2017;168:629–43. Biorender.com. 2. Mian SA, Rouault-Pierre K, Smith AE, Seidl T, Pizzitola I, Kizilors A, et al. SF3B1 mutant MDS-initiating cells may arise from the haemat- Data Availability opoietic stem cell compartment. Nat Commun 2015;6:10004. Data sets generated by scRNA-seq are accessible at GEO under 3. Awada H, Kerr CM, Durmaz A, Adema V, Gurnari C, Pagliuca S, et al. accession number GSE188367. data sets generated by bulk RNA-seq Clonal trajectories and cellular dynamics of myeloid neoplasms with are accessible at GEO under accession number GSE202722. SF3B1 mutations. Leukemia 2021;35:3324–28. 566 | BLOOD CANCER DISCOVERY NOVEMBER 2022 AACRJournals.org EIF2AK1 Pathway Activation in SF3B1-Mutant MDS-RS RESEARCH ARTICLE 4. Yoshida K, Sanada M, Shiraishi Y, Nowak D, Nagata Y, Yamamoto R, and survival of erythroid precursors in iron deficiency. EMBO J et al. Frequent pathway mutations of splicing machinery in myelod- 2001;20:6909–18. ysplasia. Nature 2011;478:64–9. 25. Chen JJ. Translational control by heme-regulated eIF2alpha kinase 5. Papaemmanuil E, Cazzola M, Boultwood J, Malcovati L, Vyas P, during erythropoiesis. Curr Opin Hematol 2014;21:172–8. Bowen D, et al. Somatic SF3B1 mutation in myelodysplasia with ring 26. Kroemer G, Marino G, Levine B. Autophagy and the integrated stress sideroblasts. N Engl J Med 2011;365:1384–95. response. Mol Cell 2010;40:280–93. 6. Malcovati L, Papaemmanuil E, Bowen DT, Boultwood J, Della Porta MG, 27. Milani M, Rzymski T, Mellor HR, Pike L, Bottini A, Generali D, et al. Pascutto C, et al. Clinical significance of SF3B1 mutations in myelodys- The role of ATF4 stabilization and autophagy in resistance of breast plastic syndromes and myelodysplastic/myeloproliferative neoplasms. cancer cells treated with bortezomib. Cancer Res 2009;69:4415–23. Blood 2011;118:6239–46. 28. Rouschop KM, van den Beucken T, Dubois L, Niessen H, Bussink J, 7. Visconte V, Makishima H, Jankowska A, Szpurka H, Traina F, Jerez A, Savelkouls K, et al. The unfolded protein response protects human et al. SF3B1, a splicing factor is frequently mutated in refractory ane- tumor cells during hypoxia through regulation of the autophagy mia with ring sideroblasts. Leukemia 2012;26:542–5. genes MAP1LC3B and ATG5. J Clin Invest 2010;120:127–41. 8. Visconte V, Avishai N, Mahfouz R, Tabarroki A, Cowen J, Sharghi- 29. B’Chir W, Maurin AC, Carraro V, Averous J, Jousse C, Muranishi Y, Moshtaghin R, et al. Distinct iron architecture in SF3B1-mutant et al. The eIF2alpha/ATF4 pathway is essential for stress-induced myelodysplastic syndrome patients is linked to an SLC25A37 splice autophagy gene expression. Nucleic Acids Res 2013;41:7683–99. variant with a retained intron. Leukemia 2015;29:188–95. 30. Malcovati L, Karimi M, Papaemmanuil E, Ambaglio I, Jadersten M, 9. Dolatshad H, Pellagatti A, Fernandez-Mercado M, Yip BH, Malcovati L, Jansson M, et al. SF3B1 mutation identifies a distinct subset of myel- Attwood M, et al. Disruption of SF3B1 results in deregulated expression odysplastic syndrome with ring sideroblasts. Blood 2015;126:233–41. and splicing of key genes and pathways in myelodysplastic syndrome 31. Grevet JD, Lan X, Hamagami N, Edwards CR, Sankaranarayanan L, Ji X, hematopoietic stem and progenitor cells. Leukemia 2015;29:1092–103. et al. Domain-focused CRISPR screen identifies HRI as a fetal hemo- 10. Dolatshad H, Pellagatti A, Liberante FG, Llorian M, Repapi E, globin regulator in human erythroid cells. Science 2018;361:285–90. Steeples V, et al. Cryptic splicing events in the iron transporter ABCB7 32. Peslak SA, Khandros E, Huang P, Lan X, Geronimo CL, Grevet JD, and other key target genes in SF3B1-mutant myelodysplastic syn- et al. HRI depletion cooperates with pharmacologic inducers to ele- dromes. Leukemia 2016;30:2322–31. vate fetal hemoglobin and reduce sickle cell formation. Blood Adv 11. Shiozawa Y, Malcovati L, Galli A, Sato-Otsubo A, Kataoka K, Sato Y, et al. 2020;4:4560–72. Aberrant splicing and defective mRNA production induced by somatic 33. Lee SC, Dvinge H, Kim E, Cho H, Micol JB, Chung YR, et al. Modulation spliceosome mutations in myelodysplasia. Nat Commun 2018;9:3649. of splicing catalysis for therapeutic targeting of leukemia with muta- 12. Clough CA, Pangallo J, Sarchi M, Ilagan JO, North K, Bergantinos R, tions in genes encoding spliceosomal proteins. Nat Med 2016;22:672–8. et al. Coordinated mis-splicing of TMEM14C and ABCB7 causes ring 34. Steensma DP, Wermke M, Klimek VM, Greenberg PL, Font P, sideroblast formation in SF3B1-mutant myelodysplastic syndrome. Komrokji RS, et al. Phase I first-in-human dose escalation study of Blood 2021;139:2038–49. the oral SF3B1 modulator H3B-8800 in myeloid neoplasms. Leuke- 13. Laurenti E, Doulatov S, Zandi S, Plumb I, Chen J, April C, et al. The tran- mia 2021;35:3542–50. scriptional architecture of early human hematopoiesis identifies multi- 35. Chen JJ. Regulation of protein synthesis by the heme-regulated level control of lymphoid commitment. Nat Immunol 2013;14:756–63. eIF2alpha kinase: relevance to anemias. Blood 2007;109:2693–9. 14. Laurenti E, Frelin C, Xie S, Ferrari R, Dunant CF, Zandi S, et al. CDK6 36. Huang P, Peslak SA, Lan X, Khandros E, Yano JA, Sharma M, et al. The levels regulate quiescence exit in human hematopoietic stem cells. HRI-regulated transcription factor ATF4 activates BCL11A transcrip- Cell Stem Cell 2015;16:302–13. tion to silence fetal hemoglobin expression. Blood 2020;135:2121–32. 15. Velten L, Haas SF, Raffel S, Blaszkiewicz S, Islam S, Hennig BP, et al. 37. Esrick EB, Lehmann LE, Biffi A, Achebe M, Brendel C, Ciuculescu MF, Human haematopoietic stem cell lineage commitment is a continu- et al. Post-transcriptional genetic silencing of BCL11A to treat sickle ous process. Nat Cell Biol 2017;19:271–81. cell disease. N Engl J Med 2021;384:205–15. 16. Belluschi S, Calderbank EF, Ciaurro V, Pijuan-Sala B, Santoro A, 38. Berastegui N, Ainciburu M, Romero JP, Alfonso-Pierola A, Philippe C, Mende N, et al. Myelo-lymphoid lineage restriction occurs in the Vilas-Zornoza A, et al. Transcriptional regulation of HSCs in aging human haematopoietic stem cell compartment before lymphoid- and MDS reveals DDIT3 as a potential driver of transformation. primed multipotent progenitors. Nat Commun 2018;9:4100. BioRxiv 2021:2021.09.08.459384. 17. Paul F, Arkin Y, Giladi A, Jaitin DA, Kenigsberg E, Keren-Shaul H, 39. Vardiman JW, Thiele J, Arber DA, Brunning RD, Borowitz MJ, et al. Transcriptional heterogeneity and lineage commitment in mye- Porwit A, et al. The 2008 revision of the World Health Organization loid progenitors. Cell 2015;163:1663–77. (WHO) classification of myeloid neoplasms and acute leukemia: 18. Yu VWC, Yusuf RZ, Oki T, Wu J, Saez B, Wang X, et al. Epigenetic rationale and important changes. Blood 2009;114:937–51. memory underlies cell-autonomous heterogeneous behavior of hemat- 40. Kanagal-Shamanna R, Singh RR, Routbort MJ, Patel KP, Medeiros LJ, opoietic stem cells. Cell 2017;168:944–5. Luthra R. Principles of analytical validation of next-generation sequenc- 19. Hay SB, Ferchen K, Chetal K, Grimes HL, Salomonis N. The Human ing based mutational analysis for hematologic neoplasms in a CLIA- Cell Atlas bone marrow single-cell interactive web portal. Exp Hema- certified laboratory. Expert Rev Mol Diagn 2016;16:461–72. tol 2018;68:51–61. 41. Gulati GS, Sikandar SS, Wesche DJ, Manjunath A, Bharadwaj A, 20. Mello FV, Land MGP, Costa ES, Teodosio C, Sanchez ML, Barcena P, Berger MJ, et al. Single-cell transcriptional diversity is a hallmark of et al. Maturation-associated gene expression profiles during normal developmental potential. Science 2020;367:405–11. human bone marrow erythropoiesis. Cell Death Discov 2019;5:69. 42. Ganan-Gomez I, Yang H, Ma F, Montalban-Bravo G, Thongon N, 21. Huang P, Zhao Y, Zhong J, Zhang X, Liu Q, Qiu X, et al. Putative Marchica V, et al. Stem cell architecture drives myelodysplastic syn- regulators for the continuum of erythroid differentiation revealed drome progression and predicts response to venetoclax-based ther- by single-cell transcriptome of human BM and UCB cells. Proc Natl apy. Nat Med 2022;28:557–67. Acad Sci U S A 2020;117:12868–76. 43. Thongon N, Ma F, Santoni A, Marchesini M, Fiorini E, Rose A, et al. 22. Hu J, Liu J, Xue F, Halverson G, Reid M, Guo A, et al. Isolation and Hematopoiesis under telomere attrition at the single-cell resolution. functional characterization of human erythroblasts at distinct stages: Nat Commun 2021;12:6850. implications for understanding of normal and disordered erythro- 44. Corces MR, Buenrostro JD, Wu B, Greenside PG, Chan SM, Koenig JL, poiesis in vivo. Blood 2013;121:3246–53. et al. Lineage-specific and single-cell chromatin accessibility charts 23. Lieu YK, Liu Z, Ali AM, Wei X, Penson A, Zhang J, et al. SF3B1 human hematopoiesis and leukemia evolution. Nat Genet 2016;48: mutant-induced missplicing of MAP3K7 causes anemia in myelodys- 1193–203. plastic syndromes. Proc Natl Acad Sci U S A 2022;119:e2111703119. 45. Buenrostro JD, Corces MR, Lareau CA, Wu B, Schep AN, Aryee MJ, et al. 24. Han AP, Yu C, Lu L, Fujiwara Y, Browne C, Chin G, et al. Heme- Integrated single-cell analysis maps the continuous regulatory land- regulated eIF2alpha kinase (HRI) is required for translational regulation scape of human hematopoietic differentiation. Cell 2018;173:1535–48. NOVEMBER 2022 BLOOD CANCER DISCOVERY | 567