Search

The role of ribosome biogenesis in erythroid development is supported by the recognition of erythroid defects in ribosomopathies in both Diamond-Blackfan anemia and 5q− syndrome. Whether ribosome biogenesis exerts a regulatory function on normal erythroid development is still unknown. In the present study, a detailed characterization of ribosome biogenesis dynamics during human and murine erythropoiesis showed that ribosome biogenesis is abruptly interrupted by the decline in ribosomal DNA transcription and the collapse of ribosomal protein neosynthesis. Its premature arrest by the RNA Pol I inhibitor CX-5461 targeted the proliferation of immature erythroblasts. p53 was activated spontaneously or in response to CX-5461, concomitant to ribosome biogenesis arrest, and drove a transcriptional program in which genes involved in cell cycle–arrested, negative regulation of apoptosis, and DNA damage response were upregulated. RNA Pol I transcriptional stress resulted in nucleolar disruption and activation of the ATR-CHK1-p53 pathway. Our results imply that the timing of ribosome biogenesis extinction and p53 activation is crucial for erythroid development. In ribosomopathies in which ribosome availability is altered by unbalanced production of ribosomal proteins, the threshold downregulation of ribosome biogenesis could be prematurely reached and, together with pathological p53 activation, prevents a normal expansion of erythroid progenitors.

In ribosomopathies, the Diamond-Blackfan anemia (DBA) or 5q- syndrome, ribosomal protein (RP) genes are affected by mutation or deletion, resulting in bone marrow erythroid hypoplasia. Unbalanced production of ribosomal subunits leading to a limited ribosome cellular content regulates translation at the expense of the master erythroid transcription factor GATA1. In RPS14-deficient cells mimicking 5q- syndrome erythroid defects, we show that the transcript length, codon bias of the coding sequence (CDS) and 3’UTR (untranslated region) structure are the key determinants of translation. In these cells, short transcripts with a structured 3’UTR and high codon adaptation index (CAI) showed a decreased translation efficiency. Quantitative analysis of the whole proteome confirmed that the post-transcriptional changes depended on the transcript characteristics that governed the translation efficiency in conditions of low ribosome availability. In addition, proteins involved in normal erythroid differentiation share most determinants of translation selectivity. Our findings thus indicate that impaired erythroid maturation due to 5q- syndrome may proceed from a translational selectivity at the expense of the erythroid differentiation program, and suggest that an interplay between the CDS and UTR may regulate mRNA translation.

The rare PEL-negative phenotype is one of the last blood groups with an unknown genetic basis. By combining whole-exome sequencing and comparative global proteomic investigations, we found a large deletion in the ABCC4/MRP4 gene encoding an ATP-binding cassette (ABC) transporter in PEL-negative individuals. The loss of PEL expression on ABCC4-CRISPR-Cas9 K562 cells and its overexpression in ABCC4-transfected cells provided evidence that ABCC4 is the gene underlying the PEL blood group antigen. Although ABCC4 is an important cyclic nucleotide exporter, red blood cells from ABCC4null/PEL-negative individuals exhibited a normal guanosine 3′,5′-cyclic monophosphate level, suggesting a compensatory mechanism by other erythroid ABC transporters. Interestingly, PEL-negative individuals showed an impaired platelet aggregation, confirming a role for ABCC4 in platelet function. Finally, we showed that loss-of-function mutations in the ABCC4 gene, associated with leukemia outcome, altered the expression of the PEL antigen. In addition to ABCC4 genotyping, PEL phenotyping could open a new way toward drug dose adjustment for leukemia treatment.

Murine-based cellular models have provided and continue to provide many useful insights into the fundamental mechanisms of erythropoiesis, as well as insights into the pathophysiology of inherited and acquired red cell disorders. Although detailed information on many aspects of these cell models is available, comprehensive proteomic data are lacking. This is a critical knowledge gap, as proteins are effectors of most biologic processes. To address this critical unmet need, proteomes of the murine cell lines Friend erythroleukemia (MEL), GATA1 erythroid (G1ER), and embryonic stem cell–derived erythroid progenitor (MEDEP) and proteomes of cultured murine marrow–derived erythroblasts at different stages of terminal erythroid differentiation were analyzed. The proteomes of MEDEP cells and primary murine erythroid cells were most similar, whereas those of MEL and G1ER cells were more distantly related. We demonstrated that the overall cellular content of histones does not decrease during terminal differentiation, despite strong chromatin condensation. Comparison of murine and human proteomes throughout terminal erythroid differentiation revealed that many noted transcriptomic changes were significantly dampened at the proteome level, especially at the end of the terminal differentiation process. Analysis of the early events associated with induction of terminal differentiation in MEDEP cells revealed divergent alterations in associated transcriptomes and proteomes. These proteomic data are powerful and valuable tools for the study of fundamental mechanisms of normal and disordered erythropoiesis and will be of broad interest to a wide range of investigators for making the appropriate choice of various cell lines to study inherited and acquired diseases of the erythrocyte.

SUMMARY Mass spectrometry-based proteomics now enables the absolute quantification of thousands of proteins in individual cell types. We used this technology to analyze the dynamic proteome changes occurring during human erythropoiesis. We quantified the absolute expression of 6,130 proteins during erythroid differentiation from late burst-forming units-erythroid (BFU-Es) to orthochromatic erythroblasts. A modest correlation between mRNA and protein expression was observed. We identified several proteins with unexpected expression patterns in erythroid cells, highlighting a breakpoint in the erythroid differentiation process at the basophilic stage. We also quantified the distribution of proteins between reticulocytes and pyrenocytes after enucleation. These analyses identified proteins that are actively sorted either with the reticulocyte or the pyrenocyte. Our study provides the absolute quantification of protein expression during a complex cellular differentiation process in humans, and it establishes a framework for future studies of disordered erythropoiesis., In Brief Gautier et al. use quantitative mass spectrometry to determine the absolute proteome composition of human erythroid progenitors throughout the differentiation process and the quantitative distribution of proteins between reticulocytes and pyrenocytes after enucleation.

Although sickle cell disease (SCD) is a red cell disorder, many cell types, including endothelial cells and polymorphonuclear neutrophils (PMNs), contribute to its pathophysiology. In particular, activated PMNs have been implicated to play an important role in the initiation and propagation of vaso-occlusive events in SCD. Activated PMNs engage in a complex process of abnormal interactions with activated endothelial cells, platelets and circulating erythrocytes contributing to endothelial injury and decreased blood flow. In the present study, global proteomic analysis was performed using label-free mass spectrometry of PMNs from 4 SCD patients (SS) in steady state and from 4 control subjects (AA). We identified a total of 4,534 proteins both in AA and SS PMNs with 3,080 of these proteins identified in at least three samples for each condition. 50 proteins were significantly over-expressed in SS PMNs compared to AA PMNs (ratio > 1.4). STRING employed to monitor potential interaction between the overexpressed proteins showed that the main interactive clusters consist of STAT1 and STAT2, OAS 1, 2 and 3, and many Interferon Signaling Proteins i.e. IFIT1, IFIT2, IFIT 3, ISG15, ISG20, GBP2, IFI35, MX1 and MX2, TLR8 proteins (Fig. 1). This finding implies a strong activation of the type I interferon (IFN) signaling pathway in the SS PMNs (between 10 and 100-fold increase in SS vs AA). In addition, 33 proteins showed significantly lower expression in SS PMNs compared to AA PMNs. Among these were L-selectin (CD62L) and IL-17 receptor A (IL17RA) (p = 0.01). These findings are consistent with previously described phenotypes of aged neutrophils and acute inflammatory responses in SCD. Similar proteomic analysis performed on PMNs from SS patients treated with hydroxycarbamide (HC, n=4) showed that 14 proteins had significantly lower expression compared to untreated-SS patients (ratio Type I IFNs represent the major effector cytokines of the host immune response against viruses and other intracellular pathogens. These cytokines are produced via activation of STAT1 and of pattern recognition receptors, including the Toll-like receptor signaling network. To determine if type I IFN-α could be detected at the protein level in the plasma of SS patients, we used the novel digital-ELISA technology (SIMOA, Quanterix) developed by Wilson et al (J Lab Autom, 2016). Interestingly, we found an increased level of INFα in plasma from SS patients compared to AA (n=32) (p In summary, our novel proteomic analysis documents a high level expression of interferon signaling proteins, STAT1 and TLR8 in the proteome of neutrophils from SS patients and strongly suggests autoimmune or auto-inflammatory phenomena at basal state in SCD. Our results provide strong support for an important role for the innate immune system in the pathophysiology of SCD. Future studies will help determine the relationship between the plasmatic level of IFN-α and clinical complications and will establish if interferon signaling proteins and IFN-α could represent new therapeutic targets in SCD. Disclosures Hermand-Tournamille: Imara: Research Funding. Le Va Kim:Imara: Research Funding. Koehl:Imara: Research Funding.

Introduction Most blood antigen-carrying proteins have important functions not only in red blood cells (RBCs) but also in other tissues. Hence, because of their polymorphisms and the existence of natural null phenotypes, blood groups also represent powerful tools to investigate human diseases. The PEL-negative rare blood type is one of the last with an unknown genetic basis. Family studies showed that the PEL-negative phenotype was inherited as an autosomal recessive trait, but the PEL locus remained undetermined since the first description of the corresponding antibody in French-Canadian individuals in 1996. Results Combining whole-exome sequencing and comparative global proteomic approaches using genomic DNA and red cell membrane proteins from four unrelated French-Canadian PEL-negative individuals and from controls, we identified i) the Multidrug Resistance-Associated Protein 4 (MRP4), also known as ABCC4, as the only missing protein in all PEL-negative RBCs and ii) a large deletion removing the last 10 exons of the ABCC4 gene. Western blot analyses confirmed that ABCC4 was present in the membrane of PEL-positive control RBCs but totally absent in the PEL-negative RBCs (Fig 1A). Western blot and flow cytometry analyses of transfected HEK293 cells revealed that over-expression of ABCC4 results in a significant increase in PEL antigen cell-surface expression (Fig 1B). Conversely, similar analyses of ABCC4-knockout K562 cells generated by the CRISPR-Cas9 strategy showed that the complete loss of ABCC4 abolished the expression of the PEL antigen (Fig 1C). PEL-negative individuals are therefore the first ever reported population lacking the complete expression of the ABCC4 protein. Surprisingly, despite the expected biological importance of ABCC4 as a cyclic nucleotide transporter, no apparent mutual symptoms or diseases were identified after investigation of the clinical history of the 12 PEL-negative members within the 4 families analyzed in this study. However, platelet investigation from one PEL-negative individual supported the role of this protein in the modulation of platelet aggregation, as previously observed in ABCC4-knockout mice studies To better understand the function of ABCC4 in mature RBCs, we measured the intracellular cGMP level in RBCs from 5 PEL-negative individuals and 5 controls. We showed that the cGMP level was not significantly impaired in the absence of ABCC4 suggesting that another cGMP transporter could compensate for the absence of ABCC4 on the RBC membrane. Interestingly, proteomic and flow cytometry analyses indicated that among the five other ABC transporters (ABCC1, ABCA7, ABCB6, ABCC5 and ABCG2) expressed on RBCs, only ABCG2 was increased in PEL-negative/ABCC4null RBCs compared to controls (1.3-fold; p=0.0007). This suggested that the absence of ABCC4 in PEL-negative individuals could be compensated by the over-expression of ABCG2. Our data, together with previous studies indicating that both ABCC4 and ABCG2 export cGMP, strongly suggest that the over-expression of ABCG2 in PEL-negative RBCs represents a compensatory mechanism for the lack of ABCC4, which could explain the absence of hematological abnormalities in PEL-negative individuals. Discussion Using genetic, molecular and cellular approaches, we demonstrated that ABCC4 is the gene underlying the novel PEL blood group system and that homozygosity for a large deletion in this gene is responsible for the rare PEL-negative phenotype. ABCC4 represents the third ABC transporter carrying a blood group system after ABCG2 (JR) and ABCB6 (LAN). The expression level of ABCC4 is strongly involved in drug resistance and toxicity, and in the clinical course of leukemia. As ABCC4 carries the PEL antigen, PEL phenotyping could be useful in the adjustment of an optimal drug dose and prevention of chemotherapy side effects in leukemia. Our findings are of immediate and important relevance in transfusion medicine and in a personalized medicine approach for chemotherapy treatment follow-up in leukemia. Furthermore, the exceptional natural "knockouts" of ABC transporters are highly valuable in understanding their function, not only in erythroid cells, but also in other cells or tissues. Disclosures Hermine: Novartis: Research Funding; Celgene: Research Funding; AB Science: Consultancy, Equity Ownership, Honoraria, Research Funding.

The JAK2V617F mutation is present in the majority of patients with polycythemia vera and one-half of those with essential thrombocythemia and primary myelofibrosis. JAK2V617F is a gain-of-function mutation resulting in constitutive JAK2 signaling involved in the pathogenesis of these diseases. JAK2V617F has been shown to promote S-phase entry. Here, we demonstrate that the CDC25A phosphatase, a key regulator of the G1/S cell-cycle transition, is constitutively overexpressed in JAK2V617F-positive cell lines, JAK2-mutated patient CD36+ progenitors, and in vitro–differentiated proerythroblasts. Accordingly, CDC25A is overexpressed in BM and spleen of Jak2V617F knock-in mice compared with wild-type littermates. By using murine FDC-P1–EPOR and human HEL and SET-2 cell lines, we found that JAK2V617F-induced CDC25A up-regulation was caused neither by increased CDC25A transcription or stability nor by the involvement of its upstream regulators Akt and MAPK. Instead, our results suggest that CDC25A is regulated at the translational level through STAT5 and the translational initiation factor eIF2α. CDC25A inhibition reduces the clonogenic and proliferative potential of JAK2V617F-expressing cell lines and erythroid progenitors while moderately affecting normal erythroid differentiation. These results suggest that CDC25A deregulation may be involved in hematopoietic cells expansion in JAK2V617F patients, making this protein an attracting potential therapeutic target.

Here, we demonstrate that the expression of the dual specificity phosphatase CDC25A, a key regulator of cell cycle progression, is deregulated in Ba/F3 cells expressing the oncogenic protein NPM/ALK and in human cell lines derived from NPM/ALK-positive anaplastic large cell lymphomas (ALCL). Both transcriptional and post-translational mechanisms account for the constitutive expression of the protein, and the PI3K/Akt pathway is essential for this process. Importantly, pharmacological inhibition of CDC25 dramatically inhibits the proliferation of NPM/ALK-expressing cells, while moderately affecting the proliferation of control Ba/F3 cells. RNA interference-mediated downregulation of CDC25A confirmed that NPM/ALK-expressing cells are highly dependent on this protein for their proliferation. Moreover, similar PI3K/AKt-mediated constitutive expression of CDC25A takes place down-stream of other hematological oncogenes, including BCR/ABL in Chronic Myeloid Leukemia and FLT3-ITD in Acute Myeloid Leukemia. Altogether, our data point to the functional link between hematopoietic oncogenic tyrosine kinases and the G(1) cell cycle regulator CDC25A, and we propose that this protein may be a potential therapeutic target in ALCL and other hematological malignancies.

Haploinsufficiency in genes encoding ribosomal proteins (RP) or ribosome-associated proteins either by mutation or deletion leads to a predominant erythroid phenotype. In acquired 5q- myelodysplasic syndrome (MDS), the macrocytic anemia has been linked to the monoallelic deletion of RPS14 gene which results in altered ribosome biogenesis. Because of the defective maturation of the small 40S ribosome subunit, the RPL5/RPL11/5SrARN complex, normally involved in the assembly of the large 60S subunit assembly, accumulates and inhibits E3-ligase-HDM2 leading to the stabilization and activation of p53 resultig in cell cycle arrest, increased apoptosis and defective differentiation of maturing erythroblasts. In the present work, we hypothesized that p53 could play a key role in the control of normal erythroid differentiation by ribosome biogenesis and we further investigated the involvement of the decreased pool of ribosome on erythroid defects in the 5q- syndrome. To investigate the first hypothesis, we studied the kinetics of ribosome biogenesis in human primary erythroblasts by mass spectrometry after pulse-SILAC. We noted that ribosome renewal collapses begining at the polychromatophilic erythroblast stage. We subsequently used the pharmacological agent CX-5461 to inhibit RNA polymerase I. When ribosome biogenesis in proerythroblasts is blocked by CX-5461, p53 is activated and proerythroblasts enter the terminal differentiation by expressing GATA1-erythroid target genes without apoptosis. By ChIP-seq in primary erythroblasts, we demonstrated that p53 binds to 1289 genes including 263 genes specifically activated by CX-5461, 6 of them being upregulated during CX-5461-induced erythroid differentiation and 3 of them being known GATA1 targets. We further used an shRNA strategy to demonstrate that one of these genes is required to permit entry into terminal erythroid differentiation when ribosome biogenesis is abrogated. We thus showed that normal erythroid differentiation is controlled by ribosome biogenesis through a p53-dependent checkpoint. In 5q- syndrome, ribosome biogenesis is continuously decreased along all stages of erythropoiesis and in contrast to the normal conditions, erythroid differentiation is defective with an excess of apoptosis affecting mature erythroblasts. We previously reported that GATA1 is targeted by a caspase-dependent cleavage since GATA1 is not protected by its chaperone HSP70 in the nucleus of MDS erythroblasts. We confirmed that GATA1 protein is decreased in 5q- primary erythroblasts and in RPS14 shRNA-expressing normal erythroblasts. To obtain further insights into the defective erythroid maturation of RPS14-deleted erythroblasts, we developed an inducible shRNA to RPS14 in the UT7/Epo cell line. Polysome profiling confirmed the decrease of 40S subunit and absolute quantification of RP by deep proteomics demonstrated a 50% decrease of RPS in conjunction with 50% reduction of ribosome content in these cells. GATA1 expression was decreased and was only partially rescued by treatment with either caspase inhibitor qVD or proteasome inhibitor, bortezomib. We then tested the hypothesis of a decrease in GATA1 translation, as previously shown in a shRNA RPS19 Diamond-Blackfan model, by analyzing and comparing the global transcriptome and the translatome corresponding to transcripts present in high molecular weight polysomes using Affymetrix HTA 2.0 microarrays. We observed a decoupling between transcriptome and translatome suggesting a selectivity of translational defects. Thermodynamic characteristics i.e. the fold energy, energy per base and length of the 3'UTR and the energy per base of the 5'UTR (Vienna RNA Package, UCSC genome browser) were the determinants of transcript selection on the polysome. The shortest transcripts with a highly structured 3'UTR including GATA1 were the transcripts which were less effectively translated. Consistently, the diminution of GATA1 protein was associated with a decrease of its target genes. Our results suggest that GATA1 is a potentially interesting therapeutic target. In summary, our results show that ribosome biogenesis controls erythroid differentiation via a p53-dependent transcriptional regulation and that a reduction of the ribosome pool leads to a selective translation at the expense of erythroid master gene GATA1. Disclosures Fontenay: Celgene: Research Funding.

Background We and others have shown that normal human erythroid cell maturation requires a transient activation of caspase-3 at late stages of maturation (Zermati et al, J Exp Med 2001). We further documented that, in human erythroblasts, the chaperone HSP70 is constitutively expressed and, at late stages of maturation, translocates into the nucleus and protects GATA-1, the master transcriptional factor critical for erythropoiesis, from caspase-3 cleavage (Ribeil et al, Nature 2007). During the maturation of human β-TM erythroblasts, HSP70 is sequestrated by excess of α-globin chains in the cytoplasm and as a consequence, GATA-1 is no longer protected from caspase-3 cleavage resulting in end-stage maturation arrest and apoptosis (Arlet et al, Nature 2013). Understanding the molecular mechanisms that regulate the localization of HSP70 during erythroid differentiation may help to find new therapeutic targets to reduce ineffective erythropoiesis in beta-thalassemia. Methods CD34 positive cells from normal and thalassemic peripheral blood were cultured in IMDM/BIT media in the presence of SCF, IL3, IL6 for seven days and subsequently cultured for additional 7 to 9 days in media containing SCF, IL3 and Epo. Erythroblasts differentiation, HSP70 localization were analysed by FACS, AMNIS stream, confocal microscopy and western blot analysis. RNAseq and proteomic analysis of highly purified erythroid cells at all distinct stages of differentiation were used to assess expression levels of various exportins. Duolink and Octet analyses were used to assess protein proximity and affinity of interactions, respectively. Results During erythroid differentiation, Hikeshi, the cognate nuclear importin of HSP70, is constitutively expressed and enables HSP70 nucleus entry as assessed by siRNA experiments. However, its expression was not regulated during erythroid differentiation. In contrast, exportin expression analysis showed marked differences in expression levels of XPO1 and XPO7 during erythroid differentiation. XPO1 expression being reduced at the time of c-kit down-regulation and caspase 3 activation while there was a marked increase in XPO7 expression at the late stages of terminal erythroid differentiation. XPO1 interacted in vivo (Duolink analysis) and in vitro with HSP70 (Octet analysis). Likewise, the previously described HSP70 S400A mutant (in the Leucine-rich Nuclear Export Sequence), which is constitutively located in the nucleus interacted with XPO1 with lower affinity compared to HSP70 WT. Stem Cell Factor (SCF) starvation and Pi3k inhibition led to decreased in vivo HSP70/XPO1 interactions. However, neither phosphorylation of HSP70 nor XPO1 were detected by Nanopro and proteomic analysis, and XPO1 expression was not regulated by Pi3K pathway. Expression of RanGTP Activating Protein (RanGAP), a protein critical for XPO1/cargo interaction, was down-regulated at the moment of caspase 3 activation during erythroid maturation, which may explain the decrease in HSP70/XPO-1 interactions. Inhibitors of XPO1 (leptomycin B and KPT 251) were able to induce HSP70 nuclear localization at early stages of differentiation (proE). In erythroid progenitors from β-TM patients, treatment with the Selective Inhibitor of Nuclear Export compound KPT-251 rescued nuclear HSP70 localization and GATA1 expression, and resulted in improved of erythroid terminal differentiation, without cytotoxicity, of thalassemic erythroid progenitors. Conclusion XPO1 is a major regulator of erythropoiesis through the regulation of HSP70 nuclear localization and is a potential new target to decrease ineffective erythropoiesis of thalassemia. Specific XPO-1 inhibitors currently in clinical development are being tested for potential therapy in thalassemic erythroid progenitors. Disclosures No relevant conflicts of interest to declare.

*the first two authors are co-first authors Introduction. Erythropoiesis is a complex process starting from pluripotent medullary progenitors and leading to the production of highly specialized and enucleated erythrocytes. Two successive phases are generally distinguished: an amplification phase with intense proliferation of morphologically similar progenitors and a terminal differentiation phase with few cell divisions and strong cellular modifications. Although erythropoiesis is a continuous process, these modifications allow the identification of specific maturation stages and the passage from one stage to the following one seems to correlate with a cell division. Several transcriptomic analyses of erythroid differentiation have been published but only few and very limited proteomic studies have been reported. Since post transcriptomic modifications are responsible for a large part of the proteome variations, a direct proteomic analysis of the erythroid differentiation is required to accurately assess the modifications that occur during this process. Results. For this study, we used CD34+ cord blood progenitors and an optimized three step cell culture method allowing the production of highly synchronized cell populations of erythroid cells at various differentiation stages. Several cellular populations from erythroid progenitors up to reticulocytes were analyzed by a label-free analysis and mass spectrometry that led to the absolute quantification of more than 6000 proteins with a false discovery rate of less than 1% (n=3). Moreover, the relative expression of well-known stage-specific erythroid markers such as TFRC, BAND3 or GLUT1, transcription factors, heme biosynthesis enzymes followed the expected pattern. To complete this study, we performed a quantitative analysis of the repartition of proteins between the generated reticulocytes and the expelled nucleus (pyrenocyte). To do that, pyrenocytes and reticulocytes were sorted by FACS according to size, Hoechst 33342 and glycophorin A labelling. Equal numbers of reticulocytes and pyrenocytes were used to prepare peptides that were analyzed by mass spectrometry after iTRAQ labelling. These experiments allowed the quantitative repartition of 1153 proteins including most erythrocyte-specific membrane proteins. Conclusion. All these results significantly increase our knowledge of the protein expression pattern during erythropoiesis and should constitute a valuable data base for subsequent studies regarding both physiological and disordered erythropoiesis. Disclosures No relevant conflicts of interest to declare.