The epigenetic regulator RINF (CXXC5) maintains SMAD7 expression in human immature erythroid cells and sustains red blood cells expansion.

The gene CXXC5, encoding a Retinoid-Inducible Nuclear Factor (RINF), is located within a region at 5q31.2 commonly deleted in myelodysplastic syndrome (MDS) and adult acute myeloid leukemia (AML). RINF may act as an epigenetic regulator and has been proposed as a tumor suppressor in hematopoietic malignancies. However, functional studies in normal hematopoiesis are lacking, and its mechanism of action is unknow. Here, we evaluated the consequences of RINF silencing on cytokineinduced erythroid differentiation of human primary CD34+ progenitors. We found that RINF is expressed in immature erythroid cells and that RINF-knockdown accelerated erythropoietin-driven maturation, leading to a significant reduction (~45%) in the number of red blood cells (RBCs), without affecting cell viability. The phenotype induced by RINF-silencing was TGFß-dependent and mediated by SMAD7, a TGFß- signaling inhibitor. RINF upregulates SMAD7 expression by direct binding to its promoter and we found a close correlation between RINF and SMAD7 mRNA levels both in CD34+ cells isolated from bone marrow of healthy donors and MDS patients with del(5q). Importantly, RINF knockdown attenuated SMAD7 expression in primary cells and ectopic SMAD7 expression was sufficient to prevent the RINF knockdowndependent erythroid phenotype. Finally, RINF silencing affects 5'-hydroxymethylation of human erythroblasts, in agreement with its recently described role as a Tet2- anchoring platform in mouse. Altogether, our data bring insight into how the epigenetic factor RINF, as a transcriptional regulator of SMAD7, may fine-tune cell sensitivity to TGFß superfamily cytokines and thus play an important role in both normal and pathological erythropoiesis.

Finely-tuned regulation of AMP-activated protein kinase is crucial for human adult erythropoiesis.

AMP-activated protein kinase (AMPK) is a heterotrimeric complex containing α, ß, and γ subunits involved in maintaining integrity and survival of murine red blood cells. Indeed, Ampk α1-/- , Ampk ß1-/- and Ampk γ1-/- mice develop hemolytic anemia and the plasma membrane of their red blood cells shows elasticity defects. The membrane composition evolves continuously along erythropoiesis and during red blood cell maturation; defects due to the absence of Ampk could be initiated during erythropoiesis. We, therefore, studied the role of AMPK during human erythropoiesis. Our data show that AMPK activation had two distinct phases in primary erythroblasts. The phosphorylation of AMPK (Thr172) and its target acetyl CoA carboxylase (Ser79) was elevated in immature erythroblasts (glycophorin Alow), then decreased conjointly with erythroid differentiation. In erythroblasts, knockdown of the α1 catalytic subunit by short hairpin RNA led to a decrease in cell proliferation and alterations in the expression of membrane proteins (band 3 and glycophorin A) associated with an increase in phosphorylation of adducin (Ser726). AMPK activation in mature erythroblasts (glycophorin Ahigh), achieved through the use of direct activators (GSK621 and compound 991), induced cell cycle arrest in the S phase, the induction of autophagy and caspase-dependent apoptosis, whereas no such effects were observed in similarly treated immature erythroblasts. Thus, our work suggests that AMPK activation during the final stages of erythropoiesis is deleterious. As the use of direct AMPK activators is being considered as a treatment in several pathologies (diabetes, acute myeloid leukemia), this observation is pivotal. Our data highlighted the importance of the finely-tuned regulation of AMPK during human erythropoiesis.

Comprehensive Proteomic Analysis of Human Erythropoiesis.

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.

Control of Pim2 kinase stability and expression in transformed human haematopoietic cells.

The oncogenic Pim2 kinase is overexpressed in several haematological malignancies, such as multiple myeloma and acute myeloid leukaemia (AML), and constitutes a strong therapeutic target candidate. Like other Pim kinases, Pim2 is constitutively active and is believed to be essentially regulated through its accumulation. We show that in leukaemic cells, the three Pim2 isoforms have dramatically short half-lives although the longer isoform is significantly more stable than the shorter isoforms. All isoforms present a cytoplasmic localization and their degradation was neither modified by broad-spectrum kinase or phosphatase inhibitors such as staurosporine or okadaic acid nor by specific inhibition of several intracellular signalling pathways including Erk, Akt and mTORC1. Pim2 degradation was inhibited by proteasome inhibitors but Pim2 ubiquitination was not detected even by blocking both proteasome activity and protein de-ubiquitinases (DUBs). Moreover, Pyr41, an ubiquitin-activating enzyme (E1) inhibitor, did not stabilize Pim2, strongly suggesting that Pim2 was degraded by the proteasome without ubiquitination. In agreement, we observed that purified 20S proteasome particles could degrade Pim2 molecule in vitro. Pim2 mRNA accumulation in UT7 cells was controlled by erythropoietin (Epo) through STAT5 transcription factors. In contrast, the translation of Pim2 mRNA was not regulated by mTORC1. Overall, our results suggest that Pim2 is only controlled by its mRNA accumulation level. Catalytically active Pim2 accumulated in proteasome inhibitor-treated myeloma cells. We show that Pim2 inhibitors and proteasome inhibitors, such as bortezomib, have additive effects to inhibit the growth of myeloma cells, suggesting that Pim2 could be an interesting target for the treatment of multiple myeloma.

Defective nuclear localization of Hsp70 is associated with dyserythropoiesis and GATA-1 cleavage in myelodysplastic syndromes.

Normal human erythroid cell maturation requests the transcription factor GATA-1 and a transient activation of caspase-3, with GATA-1 being protected from caspase-3-mediated cleavage by interaction with the chaperone heat shock protein 70 (Hsp70) in the nucleus. Erythroid cell dysplasia observed in early myelodysplastic syndromes (MDS) involves impairment of differentiation and excess of apoptosis with a burst of caspase activation. Analysis of gene expression in MDS erythroblasts obtained by ex vivo cultures demonstrates the down-regulation of a set of GATA-1 transcriptional target genes, including GYPA that encodes glycophorin A (GPA), and the up-regulation of members of the HSP70 family. GATA-1 protein expression is decreased in MDS erythroblasts, but restores in the presence of a pan-caspase inhibitor. Expression of a mutated GATA-1 that cannot be cleaved by caspase-3 rescues the transcription of GATA-1 targets, and the erythroid differentiation, but does not improve survival. Hsp70 fails to protect GATA-1 from caspases because the protein does not accumulate in the nucleus with active caspase-3. Expression of a nucleus-targeted mutant of Hsp70 protects GATA-1 and rescues MDS erythroid cell differentiation. Alteration of Hsp70 cytosolic-nuclear shuttling is a major feature of MDS that favors GATA-1 cleavage and differentiation impairment, but not apoptosis, in dysplastic erythroblasts.

Polymeric IgA1 controls erythroblast proliferation and accelerates erythropoiesis recovery in anemia.

Anemia because of insufficient production of and/or response to erythropoietin (Epo) is a major complication of chronic kidney disease and cancer. The mechanisms modulating the sensitivity of erythroblasts to Epo remain poorly understood. We show that, when cultured with Epo at suboptimal concentrations, the growth and clonogenic potential of erythroblasts was rescued by transferrin receptor 1 (TfR1)-bound polymeric IgA1 (pIgA1). Under homeostatic conditions, erythroblast numbers were increased in mice expressing human IgA1 compared to control mice. Hypoxic stress of these mice led to increased amounts of pIgA1 and erythroblast expansion. Expression of human IgA1 or treatment of wild-type mice with the TfR1 ligands pIgA1 or iron-loaded transferrin (Fe-Tf) accelerated recovery from acute anemia. TfR1 engagement by either pIgA1 or Fe-Tf increased cell sensitivity to Epo by inducing activation of mitogen-activated protein kinase (MAPK) and phosphatidylinositol 3-kinase (PI3K) signaling pathways. These cellular responses were mediated through the TfR1-internalization motif, YXXΦ. Our results show that pIgA1 and TfR1 are positive regulators of erythropoiesis in both physiological and pathological situations. Targeting this pathway may provide alternate approaches to the treatment of ineffective erythropoiesis and anemia.

Transferrin receptor 2 is a component of the erythropoietin receptor complex and is required for efficient erythropoiesis.

Erythropoietin (Epo) is required for erythroid progenitor differentiation. Although Epo crosslinking experiments have revealed the presence of Epo receptor (EpoR)-associated proteins that could never be identified, EpoR is considered to be a paradigm for homodimeric cytokine receptors. We purified EpoR-binding partners and identified the type 2 transferrin receptor (TfR2) as a component of the EpoR complex corresponding to proteins previously detected in cross-linking experiments. TfR2 is involved in iron metabolism by regulating hepcidin production in liver cells. We show that TfR2 and EpoR are synchronously coexpressed during the differentiation of erythroid progenitors. TfR2 associates with EpoR in the endoplasmic reticulum and is required for the efficient transport of this receptor to the cell surface. Erythroid progenitors from TfR2(-/-)mice show a decreased sensitivity to Epo and increased circulating Epo levels. In human erythroid progenitors, TfR2 knockdown delays the terminal differentiation. Erythroid cells produce growth differentiation factor-15, a cytokine that suppresses hepatic hepcidin production in certain erythroid diseases such as thalassemia. We show that the production of growth differentiation factor-15 by erythroid cells is dependent on both Epo and TfR2. Taken together, our results show that TfR2 exhibits a non hepatic function as a component of the EpoR complex and is required for efficient erythropoiesis.

HSP27 controls GATA-1 protein level during erythroid cell differentiation.

Heat shock protein 27 (HSP27) is a chaperone whose cellular expression increases in response to various stresses and protects the cell either by inhibiting apoptotic cell death or by promoting the ubiquitination and proteasomal degradation of specific proteins. Here, we show that globin transcription factor 1 (GATA-1) is a client protein of HSP27. In 2 models of erythroid differentiation; that is, in the human erythroleukemia cell line, K562 induced to differentiate into erythroid cells on hemin exposure and CD34(+) human cells ex vivo driven to erythroid differentiation in liquid culture, depletion of HSP27 provokes an accumulation of GATA-1 and impairs terminal maturation. More specifically, we demonstrate that, in the late stages of the erythroid differentiation program, HSP27 is phosphorylated in a p38-dependent manner, enters the nucleus, binds to GATA-1, and induces its ubiquitination and proteasomal degradation, provided that the transcription factor is acetylated. We conclude that HSP27 plays a role in the fine-tuning of terminal erythroid differentiation through regulation of GATA-1 content and activity.

Phenotypic and genotypic characteristics of mastocytosis according to the age of onset.

Adult's mastocytosis is usually associated with persistent systemic involvement and c-kit 816 mutation, while pediatrics disease is mostly limited to the skin and often resolves spontaneously. We prospectively included 142 adult patients with histologically proven mastocytosis. We compared phenotypic and genotypic features of adults patients whose disease started during childhood (Group 1, n = 28) with those of patients whose disease started at adult's age (Group 2, n = 114). Genotypic analysis was performed on skin biopsy by sequencing of c-kit exons 17 and 8 to 13. According to WHO classification, the percentage of systemic disease was similar (75 vs. 73%) in 2 groups. C-kit 816 mutation was found in 42% and 77% of patients in groups 1 and 2, respectively (p<0.001). 816 c-kit mutation was associated with systemic mastocytosis in group 2 (87% of patients with systemic mastocytosis vs. 45% with cutaneous mastocytosis, p = 0.0001). Other c-kit activating mutations were found in 23% of patients with mastocytosis' onset before the age of 5, 0% between 6 and 15 years and 2% at adults' age (p<0.001). In conclusion, pathogenesis of mastocytosis significantly differs according to the age of disease's onset. Our data may have major therapeutic relevance when considering c-kit-targeted therapy.

Apaf-1 Deficiency Causes Chromosomal Instability.

Apaf-1 is an essential component of the apoptosome, the molecular complex assembled in response to mitochondrial cytochrome c release that promotes caspase activation. Apaf-1 expression is suppressed in some malignant tumors, in particular melanoma as well as cervical and colorectal carcinoma, in which the loss of Apaf-1 expression marks tumor progression and poor prognosis. Recent results from our laboratory demonstrate that Apaf-1 has an apoptosis-unrelated function that may well account for its role as a tumor suppressor. The knockout of apaf-1 (in mice), the knockdown of Apaf-1 (in human cells) and loss of function mutations of ced-4 (the Caenorhabditis elegans ortholog of Apaf-1) compromise the arrest of DNA synthesis in response to DNA damage, in a context in which apoptosis does not occur. Here, we show that the depletion of Apaf-1 also sensitizes cells to chromosomal instability induced by different types of DNA damage such as cisplatin, UVC light and gamma-irradiation. These results unravel a hitherto unsuspected role for Apaf-1 in the maintenance of genomic stability, independently from its function in the cell death machinery.

Nonapoptotic role for Apaf-1 in the DNA damage checkpoint.

Apaf-1 is an essential factor for cytochrome c-driven caspase activation during mitochondrial apoptosis but has also an apoptosis-unrelated function. Knockdown of Apaf-1 in human cells, knockout of apaf-1 in mice, and loss-of-function mutations in the Caenorhabditis elegans apaf-1 homolog ced-4 reveal the implication of Apaf-1/CED-4 in DNA damage-induced cell-cycle arrest. Apaf-1 loss compromised the DNA damage checkpoints elicited by ionizing irradiation or chemotherapy. Apaf-1 depletion reduced the activation of the checkpoint kinase Chk1 provoked by DNA damage, and knockdown of Chk1 abrogated the Apaf-1-mediated cell-cycle arrest. Nuclear translocation of Apaf-1, induced in vitro by exogenous DNA-damaging agents, correlated in non-small cell lung cancer (NSCLC) with the endogenous activation of Chk-1, suggesting that this pathway is clinically relevant. Hence, Apaf-1 exerts two distinct, phylogenetically conserved roles in response to mitochondrial membrane permeabilization and DNA damage. These data point to a role for Apaf-1 as a bona fide tumor suppressor.

A novel epidermal growth factor receptor inhibitor promotes apoptosis in non-small cell lung cancer cells resistant to erlotinib.

Non-small cell lung cancer (NSCLC) with activating mutations in the epidermal growth factor receptor (EGFR) responds to EGFR tyrosine kinase inhibitors such as erlotinib. However, secondary somatic EGFR mutations (e.g., T790M) confer resistance to erlotinib. BMS-690514, a novel panHER/vascular endothelial growth factor receptor (VEGFR) inhibitor described here, exerted antiproliferative and proapoptotic effects on NSCLC cell lines, with prominent efficacy on H1975 cells expressing the T790M mutation. In this model, BMS-690514 induced a G(1) cell cycle arrest, as well as ultrastructural hallmarks of apoptosis, mitochondrial release of cytochrome c, and activation of caspases involved in the intrinsic (e.g., caspase-2, caspase-3, caspase-7, and caspase-9), but not in the extrinsic (e.g., caspase-8), pathway. Caspase inhibition conferred partial protection against BMS-690514 cytotoxicity, pointing to the involvement of both caspase-dependent and caspase-independent effector mechanisms. Transcriptome analyses revealed the up-regulation of proapoptotic (e.g., Bim, Puma) and cell cycle inhibitory (e.g., p27(Kip1), p57(Kip2)) factors, as well as the down-regulation of antiapoptotic (e.g., Mcl1), heat shock (e.g., HSP40, HSP70, HSP90), and cell cycle promoting [e.g., cyclins B1, D1, and D3; cyclin-dependent kinase 1 (CDK1); MCM family proteins; proliferating cell nuclear antigen (PCNA)] proteins. BMS-690514-induced death of H1975 cells was modified in a unique fashion by a panel of small interfering RNAs targeting apoptosis modulators. Down-regulation of components of the nuclear factor-kappaB survival pathway (e.g., p65, Nemo/IKK gamma, TAB2) sensitized cells to BMS-690514, whereas knockdown of proapoptotic factors (e.g., Puma, Bax, Bak, caspase-2, etc.) and DNA damage-related proteins (e.g., ERCC1, hTERT) exerted cytoprotective effects. BMS-690514 is a new pan-HER/VEGFR inhibitor that may become an alternative to erlotinib for the treatment of NSCLC.

Hsp70 regulates erythropoiesis by preventing caspase-3-mediated cleavage of GATA-1.

Caspase-3 is activated during both terminal differentiation and erythropoietin-starvation-induced apoptosis of human erythroid precursors. The transcription factor GATA-1, which performs an essential function in erythroid differentiation by positively regulating promoters of erythroid and anti-apoptotic genes, is cleaved by caspases in erythroid precursors undergoing cell death upon erythropoietin starvation or engagement of the death receptor Fas. In contrast, by an unknown mechanism, GATA-1 remains uncleaved when these cells undergo terminal differentiation upon stimulation with Epo. Here we show that during differentiation, but not during apoptosis, the chaperone protein Hsp70 protects GATA-1 from caspase-mediated proteolysis. At the onset of caspase activation, Hsp70 co-localizes and interacts with GATA-1 in the nucleus of erythroid precursors undergoing terminal differentiation. In contrast, erythropoietin starvation induces the nuclear export of Hsp70 and the cleavage of GATA-1. In an in vitro assay, Hsp70 protects GATA-1 from caspase-3-mediated proteolysis through its peptide-binding domain. The use of RNA-mediated interference to decrease the Hsp70 content of erythroid precursors cultured in the presence of erythropoietin leads to GATA-1 cleavage, a decrease in haemoglobin content, downregulation of the expression of the anti-apoptotic protein Bcl-X(L), and cell death by apoptosis. These effects are abrogated by the transduction of a caspase-resistant GATA-1 mutant. Thus, in erythroid precursors undergoing terminal differentiation, Hsp70 prevents active caspase-3 from cleaving GATA-1 and inducing apoptosis.

Heat shock proteins 27 and 70: anti-apoptotic proteins with tumorigenic properties.

Heat shock proteins (HSP) HSP27 and HSP70 are expressed in response to a wide variety of physiological and environmental insults including anticancer chemotherapy, thus allowing the cell to survive to lethal conditions. Several mechanisms account for the cytoprotective effect of HSP27 and HSP70. (1) Both proteins are powerful chaperones. (2) They both inhibit key effectors of the apoptotic machinery at the pre and post-mitochondrial level. (3) They participate in the proteasome-mediated degradation of proteins under stress conditions, thereby contributing to the so called "protein triage". In cancer cells, the expression of HSP27 and/or HSP70 is abnormally high, and both HSP27 and HSP70 may participate in oncogenesis and in resistance to chemotherapy. In rodent models, HSP27 or HSP70 over-expression increases tumor growth and metastatic potential. The depletion or inhibition of HSP27 and HS70 frequently reduces the size of the tumors and even can cause their complete involution (for HSP70). Therefore, the inhibition of HSP70 and HSP27 has become a novel strategy of cancer therapy.

Effect of tyrosine kinase inhibitor STI571 on the kinase activity of wild-type and various mutated c-kit receptors found in mast cell neoplasms.

Systemic mastocytosis (SM) is a rare disease caused by an abnormal mast cell accumulation in various tissues. Two classes of constitutive activating c-kit mutations are found in SM. The most frequent class occurs in the catalytic pocket coding region with substitutions at codon 816 and the other in the intracellular juxtamembrane coding region. Therefore, kinase inhibitors that block mutated c-kit activity might be used as therapeutic agents in SM. Here, we show that STI571 inhibits both wild-type and juxtamembrane mutant c-kit kinase activity, but has no effect on the activity of the D816 V mutant. Accordingly, STI571 selectively decreases the survival of normal mast cell and of mast cell lines either with juxtamembrane c-kit mutations, but not that of tumoral mast cell from patient with SM or of mast cell lines with the D816 V mutation. Therefore, STI571 is not a good candidate to treat SM and specific kinase inhibitors should be designed to inhibit constitutive activating mutations at codon 816.

Human T-cell lymphotropic virus oncoprotein Tax represses TGF-beta 1 signaling in human T cells via c-Jun activation: a potential mechanism of HTLV-I leukemogenesis.

Human T-cell leukemia virus I is the etiologic agent of adult T-cell leukemia (ATL), an aggressive T-cell malignancy. The viral oncoprotein Tax, through the activation of nuclear factorkappaB (NF-kappaB), CCAAT-enhancer binding protein (CREB), and activated protein-1 (AP-1) pathways, is a transcriptional regulator of critical genes for T-cell homeostasis. In ATL cells, activated AP-1 complexes induce the production of transforming growth factor beta1 (TGF-beta1). TGF-beta1 is an inhibitor of T-cell proliferation and cytotoxicity. Here we show that, in contrast to normal peripheral T cells, ATL cells are resistant to TGF-beta1-induced growth inhibition. The retroviral transduction of the Tax protein in peripheral T cells resulted in the loss of TGF-beta1 sensitivity. Transient transfection of Tax in HepG2 cells specifically inhibited Smad/TGF-beta1 signaling in a dose-dependent manner. In the presence of Tax transfection, increasing amounts of Smad3 restored TGF-beta1 signaling. Tax mutants unable to activate NF-kappaB or CREB pathways were also able to repress Smad3 transcriptional activity. Next we have demonstrated that Tax inhibits TGF-beta1 signaling by reducing the Smad3 DNA binding activity. However, Tax did not decrease the expression and the nuclear translocation of Smad3 nor did it interact physically with Smad3. Rather, Tax induced c-Jun N-terminal kinase (JNK) activity and c-Jun phosphorylation, leading to the formation of Smad3/c-Jun complexes. Whereas c-Jun alone abrogates Smad3 DNA binding, cotransfection of Tax and of a dominant-negative form of JNK or a c-Jun antisense-restored Smad3 DNA binding activity and TGF-beta1 responsiveness. In ATL and in normal T cells transduced by Tax, c-Jun was constitutively phosphorylated. Thus, we describe a new function of Tax, as a repressor of TGF-beta1 signaling through JNK/c-Jun constitutive activation, which may play a critical role in ATL leukemogenesis.

Specific involvement of caspases in the differentiation of monocytes into macrophages.

Caspases are cysteine proteases involved in apoptosis and cytokine maturation. In erythroblasts, keratinocytes, and lens epithelial cells undergoing differentiation, enucleation has been regarded as a caspase-mediated incomplete apoptotic process. Here, we show that several caspases are activated in human peripheral blood monocytes whose differentiation into macrophages is induced by macrophage colony-stimulating factor (M-CSF). This activation is not associated with cell death and cannot be detected in monocytes undergoing dendritic cell differentiation in the presence of interleukin-4 (IL-4) and granulocyte-macrophage colony-stimulating factor (GM-CSF). The mechanisms and consequences of caspase activation were further studied in U937 human monocytic cells undergoing phorbol ester-induced differentiation into macrophages. Differentiation-associated caspase activation involves the release of cytochrome c from the mitochondria and leads to the cleavage of the protein acinus while the poly(ADP-ribose)polymerase remains uncleaved. Inhibition of caspases by either exposure to the broad-spectrum inhibitor benzyloxycarbonyl-Val-Ala-(DL)-Asp-fluoromethylketone (z-VAD-fmk) or expression of the p35 baculovirus inhibitory protein or overexpression of Bcl-2 inhibits the differentiation process. In addition, z-VAD-fmk amplifies the differentiation-associated production of radical oxygen species in both phorbol ester-differentiated U937 cells and M-CSF-treated monocytes, shifting the differentiation process to nonapoptotic cell death. Altogether, these results indicate that caspase activation specifically contributes to the differentiation of monocytes into macrophages, in the absence of cell death.

Platelet formation is the consequence of caspase activation within megakaryocytes.

Platelets are formed from mature megakaryocytes (MKs) and arise from the development of long and thin cytoplasmic extensions called proplatelets. After platelet release, the senescent MKs (nucleus surrounded by some cytoplasm) undergo cell death by apoptosis. To explore the precise role of apoptosis in proplatelet formation, we grew human MKs from CD34(+) cells and assessed the possible role of caspases. Proteolytic maturation of procaspase-3 and procaspase-9 was detected by immunoblots in maturing MKs as well as in proplatelet-bearing MKs and senescent MKs. Cleavage of caspase substrates such as gelsolin or poly adenosine diphosphate (ADP)-ribose polymerase (PARP) was also detected. Interestingly, activated forms of caspase-3 were detected in maturing MKs, before proplatelet formation, with a punctuate cytoplasmic distribution, whereas a diffuse staining pattern was seen in senescent and apoptotic MKs. This localized activation of caspase-3 was associated with a mitochondrial membrane permeabilization as assessed by the release of cytochrome c, suggesting an activation of the intrinsic pathway. Moreover, these MKs with localized activated caspase-3 had no detectable DNA fragmentation. In contrast, when apoptosis was induced by staurosporine, diffuse caspase activation was seen; these MKs had signs of DNA fragmentation, and no proplatelet formation occurred. The pan-caspase inhibitor z-VAD.fmk as well as more specific inhibitors of caspase-3 and caspase-9 blocked proplatelet formation, whereas an inhibitor of calpeptin had no effect. Overexpression of Bcl-2 also inhibited proplatelet formation in maturing MKs. Thus, localized caspase activation is causal to proplatelet formation. We conclude that proplatelet formation is regulated by a caspase activation limited to only some cellular compartments.