Targeting the EIF2AK1 Signaling Pathway Rescues Red Blood Cell Production in SF3B1-Mutant Myelodysplastic Syndromes With Ringed Sideroblasts
- PMID: 35926182
- PMCID: PMC9894566
- DOI: 10.1158/2643-3230.BCD-21-0220
Targeting the EIF2AK1 Signaling Pathway Rescues Red Blood Cell Production in SF3B1-Mutant Myelodysplastic Syndromes With Ringed Sideroblasts
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 activation 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 transfusions. Our study has implications for the development of therapies to achieve long-lasting hematologic responses. This article is highlighted in the In This Issue feature, p. 476.
©2022 The Authors; Published by the American Association for Cancer Research.
Figures
![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.](https://cdn.statically.io/img/www.ncbi.nlm.nih.gov/pmc/articles/instance/9894566/bin/554fig1.gif)
![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-erythroid 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.](https://cdn.statically.io/img/www.ncbi.nlm.nih.gov/pmc/articles/instance/9894566/bin/554fig2.gif)
![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.](https://cdn.statically.io/img/www.ncbi.nlm.nih.gov/pmc/articles/instance/9894566/bin/554fig3.gif)
![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
significance 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.](https://cdn.statically.io/img/www.ncbi.nlm.nih.gov/pmc/articles/instance/9894566/bin/554fig4.gif)
![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.](https://cdn.statically.io/img/www.ncbi.nlm.nih.gov/pmc/articles/instance/9894566/bin/554fig5.gif)
Similar articles
-
SF3B1-mutated myelodysplastic syndrome with ring sideroblasts harbors more severe iron overload and corresponding over-erythropoiesis.Leuk Res. 2016 May;44:8-16. doi: 10.1016/j.leukres.2016.02.011. Epub 2016 Feb 27. Leuk Res. 2016. PMID: 26970172
-
Clinical Outcomes With Ring Sideroblasts and SF3B1 Mutations in Myelodysplastic Syndromes: MDS Clinical Research Consortium Analysis.Clin Lymphoma Myeloma Leuk. 2018 Aug;18(8):528-532. doi: 10.1016/j.clml.2018.05.016. Epub 2018 May 26. Clin Lymphoma Myeloma Leuk. 2018. PMID: 29937400
-
Real-world diagnostic testing patterns for assessment of ring sideroblasts and SF3B1 mutations in patients with newly diagnosed lower-risk myelodysplastic syndromes.Int J Lab Hematol. 2021 Jun;43(3):426-432. doi: 10.1111/ijlh.13400. Epub 2020 Nov 21. Int J Lab Hematol. 2021. PMID: 33220019 Free PMC article. Clinical Trial.
-
SF3B1-mutant MDS as a distinct disease subtype: a proposal from the International Working Group for the Prognosis of MDS.Blood. 2020 Jul 9;136(2):157-170. doi: 10.1182/blood.2020004850. Blood. 2020. PMID: 32347921 Free PMC article. Review.
-
Myelodysplastic syndromes with ring sideroblasts.Hematol Oncol. 2023 Oct;41(4):612-620. doi: 10.1002/hon.3125. Epub 2023 Feb 26. Hematol Oncol. 2023. PMID: 36794650 Review.
Cited by
-
Erythroid Differentiation Enhances RNA Mis-Splicing in SF3B1-Mutant Myelodysplastic Syndromes with Ring Sideroblasts.Cancer Res. 2024 Jan 16;84(2):211-225. doi: 10.1158/0008-5472.CAN-23-3038. Cancer Res. 2024. PMID: 37921711 Free PMC article.
-
Normal and dysregulated crosstalk between iron metabolism and erythropoiesis.Elife. 2023 Aug 14;12:e90189. doi: 10.7554/eLife.90189. Elife. 2023. PMID: 37578340 Free PMC article.
-
SF3B1 mutations in myelodysplastic syndromes: A potential therapeutic target for modulating the entire disease process.Front Oncol. 2023 Mar 17;13:1116438. doi: 10.3389/fonc.2023.1116438. eCollection 2023. Front Oncol. 2023. PMID: 37007111 Free PMC article. Review.
References
-
- Yoshida K, Sanada M, Shiraishi Y, Nowak D, Nagata Y, Yamamoto R, et al. . Frequent pathway mutations of splicing machinery in myelodysplasia. Nature 2011;478:64–9. - PubMed
Publication types
MeSH terms
Substances
Grants and funding
LinkOut - more resources
Full Text Sources
Medical
Research Materials
Miscellaneous