A role for reactive oxygen species in JAK2 V617F myeloproliferative neoplasm progression.

Although other mutations may predate the acquisition of the JAK2(V617F) mutation, the latter is sufficient to drive the disease phenotype observed in BCR-ABL-negative myeloproliferative neoplasms (MPNs). One of the consequences of JAK2(V617F) is genetic instability that could explain JAK2(V617F)-mediated MPN progression and heterogeneity. Here, we show that JAK2(V617F) induces the accumulation of reactive oxygen species (ROS) in the hematopoietic stem cell compartment of a knock-in (KI) mouse model and in patients with JAK2(V617F) MPNs. JAK2(V617F)-dependent ROS elevation was partly mediated by an AKT-induced decrease in catalase expression and was accompanied by an increased number of 8-oxo-guanines and DNA double-strand breaks (DSBs). Moreover, there was evidence for a mitotic recombination event in mice resulting in loss of heterozygosity of Jak2(V617F). Mice engrafted with 30% of Jak2(V617F) KI bone marrow (BM) cells developed a polycythemia vera-like disorder. Treatment with the anti-oxidant N-acetylcysteine (NAC) substantially restored blood parameters and reduced damages to DNA. Furthermore, NAC induced a marked decrease in splenomegaly with reduction in the frequency of the Jak2(V617F)-positive hematopoietic progenitors in BM and spleen. Altogether, overproduction of ROS is a mediator of JAK2(V617F)-induced DNA damages that promote disease progression. Targeting ROS accumulation might prevent the development of JAK2(V617F) MPNs.

Constitutive JunB expression, associated with the JAK2 V617F mutation, stimulates proliferation of the erythroid lineage.

The JAK2 V617F mutation, present in the majority of polycythemia vera (PV) patients, causes constitutive activation of JAK2 and seems to be responsible for the PV phenotype. However, the transcriptional changes triggered by the mutation have not yet been totally characterized. In this study, we performed a large-scale gene expression study using serial analysis of gene expression in bone marrow cells of a newly diagnosed PV patient harboring the JAK2 V617F mutation and in normal bone marrow cells of healthy donors. JUNB was one of the genes upregulated in PV, and we confirmed, by quantitative real-time PCR, an overexpression of JUNB in hematopoietic cells of other JAK2 V617F PV patients. Using Ba/F3-EPOR cell lines and primary human erythroblast cultures, we found that JUNB was transcriptionally induced after erythropoietin addition and that JAK2 V617F constitutively induced JunB protein expression. Furthermore, JUNB knockdown reduced not only the growth of Ba/F3 cells by inducing apoptosis, but also the clonogenic and proliferative potential of human erythroid progenitors. These results establish a role for JunB in normal erythropoiesis and indicate that JunB may play a major role in the development of JAK2 V617F myeloproliferative disorders.

Membrane-bound delta-4 notch ligand reduces the proliferative activity of primitive human hematopoietic CD34+CD38low cells while maintaining their LTC-IC potential.

To examine the role of the Notch ligand Delta-4 on hematopoietic stem cells, human CD34+CD38low cord blood cells were cocultured on S17 cells transduced with transmembrane Delta-4 (mbD4/S17) or an empty vector (C/S17). By the end of a 3-week culture, mbD4/S17 induced a 25-fold reduction in nucleated cell production, as compared to C/S17, by maintaining a higher proportion of cells in G0/G1 phase. A specific retention of a high proportion of CD34+ cells throughout the culture was observed with mbD4/S17, contrary to C/S17. Although mbD4/S17 promoted expansion of cells with the phenotype of committed lymphoid precursors (CD34+CD7+CD45RA+), these cells still retained their myeloid differentiation potential. mbD4/S17 maintained a higher LTC-IC frequency in output CD34+ cells, compared to C/S17, as in the subsets of cells having completed the same number of divisions on mbD4/S17. A Delta4-Fc protein (extracellular part of human Delta4 fused to Fc human IgG1 portion), immobilized on plastic, also reduced cell production and retained the LTC-IC potential. Transplantation of cells grown on mbD4/S17 into NOD/SCID mice showed no significant enhancement of the long-term repopulating ability. Thus, Delta4 appears to inhibit hematopoietic stem cell proliferation, in association with the maintenance of short-term lymphoid and myeloid repopulation capacity.

Human platelet glycoprotein VI function is antagonized by monoclonal antibody-derived Fab fragments.

Platelet interactions with adhesive ligands exposed at sites of vascular injury initiate the normal hemostatic response but may also lead to arterial thrombosis. Platelet membrane glycoprotein (GP)VI is a key receptor for collagen. Impairment of GPVI function in mice results in a long-term antithrombotic protection and prevents neointimal hyperplasia following arterial injury. On the other hand, GPVI deficiency in humans or mice does not result in serious bleeding tendencies. Blocking GPVI function may thus represent a new and safe antithrombotic approach, but no specific, potent anti-GPVI directed at the human receptor is yet available. Our aim was to produce accessible antagonists of human GPVI to evaluate the consequences of GPVI blockade. Amongst several monoclonal antibodies to the extracellular domain of human GPVI, one, 9O12.2, was selected for its capacity to disrupt the interaction of GPVI with collagen in a purified system and to prevent the adhesion of cells expressing recombinant GPVI to collagen and collagen-related peptides (CRP). While 9O12.2 IgGs induced platelet activation by a mechanism involving GPVI and Fc gamma RIIA, 9O12.2 Fab fragments completely blocked collagen-induced platelet aggregation and secretion from 5 microg mL-1 and fully prevented CRP-induced activation from 1.5 microg mL-1. 9O12.2 Fabs also inhibited the procoagulant activity of collagen-stimulated platelets and platelet adhesion to collagen in static conditions. Furthermore, 9O12.2 Fabs impaired platelet adhesion, and prevented thrombi formation under arterial flow conditions. We thus describe here for the first time a functional monoclonal antibody to human GPVI and demonstrate its effect on collagen-induced platelet aggregation and procoagulant activity, and on thrombus growth.

Cloning, expression, and function of BLAME, a novel member of the CD2 family.

The CD2 family is a growing family of Ig domain-containing cell surface proteins involved in lymphocyte activation. Here we describe the cloning and expression analysis of a novel member of this family, B lymphocyte activator macrophage expressed (BLAME). BLAME shares the structural features of the CD2 family containing an IgV and IgC2 domain and clusters with the other family members on chromosome 1q21. Quantitative PCR and Northern blot analysis show BLAME to be expressed in lymphoid tissue and, more specifically, in some populations of professional APCs, activated monocytes, and DCS: Retroviral forced expression of BLAME in hematopoietic cells of transplanted mice showed an increase in B1 cells in the peripheral blood, spleen, lymph nodes, and, most strikingly, in the peritoneal cavity. These cells do not express CD5 and are CD23(low)Mac1(low), characteristics of the B1b subset. BLAME may therefore play a role in B lineage commitment and/or modulation of signal through the B cell receptor.

Expression and function of the collagen receptor GPVI during megakaryocyte maturation.

In this report, the expression and function of the platelet collagen receptor glycoprotein VI (GPVI) were studied in human megakaryocytes during differentiation and maturation of mobilized blood and cord blood derived CD34(+) cells. By flow cytometry, using an anti-GPVI monoclonal antibody or convulxin, a GPVI-specific ligand, GPVI was detected only on CD41(+) cells including some CD41(+)/CD34(+) cells, suggesting expression at a stage of differentiation similar to CD41. These results were confirmed at the mRNA level using reverse transcription-polymerase chain reaction. GPVI expression was low during megakaryocytic differentiation but increased in the more mature megakaryocytes (CD41(high)). As in platelets, megakaryocyte GPVI associates with the Fc receptor gamma chain (FcRgamma). The FcR gamma chain was detected at the RNA and protein level at all stages of megakaryocyte maturation preceding the expression of GPVI. The other collagen receptor, alpha(2)beta(1) integrin (CD49b/CD29), had a pattern of expression similar to GPVI. Megakaryocytic GPVI was recognized as a 55-kDa protein by immunoblotting and ligand blotting, and thus it presented a slightly lower apparent molecular mass than platelet GPVI (58 kDa). Megakaryocytes began to adhere to immobilized convulxin via GPVI after only 8-10 days of culture, at a time when megakaryocytes were maturing. At this stage of maturation, they also adhered to immobilized collagen by alpha(2)beta(1) integrin-dependent and -independent mechanisms. Convulxin induced a very similar pattern of protein tyrosine phosphorylation in megakaryocytes and platelets including Syk, FcRgamma, and PLC(gamma)2. Our results showed that GPVI is expressed early during megakaryocytic differentiation but functionally allows megakaryocyte adherence to collagen only at late stages of differentiation when its expression increases.

Cloning, characterization, and functional studies of human and mouse glycoprotein VI: a platelet-specific collagen receptor from the immunoglobulin superfamily.

Injuries to the vessel wall and subsequent exposure of collagen from the subendothelial matrix result in thrombus formation. In physiological conditions, the platelet plug limits blood loss. However, in pathologic conditions, such as rupture of atherosclerotic plaques, platelet-collagen interactions are associated with cardiovascular and cerebral vascular diseases. Platelet glycoprotein VI (GPVI) plays a crucial role in collagen-induced activation and aggregation of platelets, and people who are deficient in GPVI suffer from bleeding disorders. Based on the fact that GPVI is coupled to the Fc receptor (FcR)-gamma chain and thus should share homology with the FcR chains, the genes encoding human and mouse GPVI were identified. They belong to the immunoglobulin (Ig) superfamily and share 64% homology at the protein level. Functional evidence demonstrating the identity of the recombinant protein with GPVI was shown by binding to its natural ligand collagen; binding to convulxin (Cvx), a GPVI-specific ligand from snake venom; binding of anti-GPVI IgG isolated from a patient; and association to the FcR-gamma chain. The study also demonstrated that the soluble protein blocks Cvx and collagen-induced platelet aggregation and that GPVI expression is restricted to megakaryocytes and platelets. Finally, human GPVI was mapped to chromosome 19, long arm, region 1, band 3 (19q13), in the same region as multiple members of the Ig superfamily. This work offers the opportunity to explore the involvement of GPVI in thrombotic disease, to develop alternative antithrombotic compounds, and to characterize the mechanism involved in GPVI genetic deficiencies. (Blood. 2000;96:1798-1807)

Hematopoietic-specific beta 1 tubulin participates in a pathway of platelet biogenesis dependent on the transcription factor NF-E2.

The cellular and molecular bases of platelet release by terminally differentiated megakaryocytes represent important questions in cell biology and hematopoiesis. Mice lacking the transcription factor NF-E2 show profound thrombocytopenia, and their megakaryocytes fail to produce proplatelets, the microtubule-based precursors of blood platelets. Using mRNA subtraction between normal and NF-E2-deficient megakaryocytes, cDNA was isolated encoding beta1 tubulin, the most divergent beta tubulin isoform. In NF-E2-deficient megakaryocytes, beta1 tubulin mRNA and protein are virtually absent. The expression of beta1 tubulin is exquisitely restricted to platelets and megakaryocytes, where it appears late in differentiation and localizes to microtubule shafts and coils within proplatelets. Restoring NF-E2 activity in a megakaryoblastic cell line or in NF-E2-deficient primary megakaryocytes rescues the expression of beta1 tubulin. Re-expressing beta1 tubulin in isolation does not, however, restore proplatelet formation in the defective megakaryocytes, indicating that other critical factors are required; indeed, other genes identified by mRNA subtraction also encode structural and regulatory components of the cytoskeleton. These findings provide critical mechanistic links between NF-E2, platelet formation, and selected microtubule proteins, and they also provide novel molecular insights into thrombopoiesis. (Blood. 2000;96:1366-1373)

Pathophysiology of thrombocytopenia and anemia in mice lacking transcription factor NF-E2.

Expression of the p45 subunit of transcription factor NF-E2 is restricted to selected blood cell lineages, including megakaryocytes and developing erythrocytes. Mice lacking p45 NF-E2 show profound thrombocytopenia, resulting from a late arrest in megakaryocyte differentiation, and a number of red blood cell defects, including anisocytosis and hypochromia. Here we report results of studies aimed to explore the pathophysiology of these abnormalities. Mice lacking NF-E2 produce very few platelet-like particles that display highly disorganized ultrastructure and respond poorly to platelet agonists, features consistent with the usually lethal hemorrhage in these animals. Thrombocytopenia was evident during fetal life and was not corrected by splenectomy in adults. Surprisingly, fetal NF-E2-deficient megakaryocyte progenitors showed reduced proliferation potential in vitro. Thus, NF-E2 is required for regulated megakaryocyte growth as well as for differentiation into platelets. All the erythroid abnormalities were reproduced in lethally irradiated wild-type recipients of hematopoietic cells derived from NF-E2-null fetuses. Whole blood from mice lacking p45 NF-E2 showed numerous small red blood cell fragments; however, survival of intact erythrocytes in vivo was indistinguishable from control mice. Considered together, these observations indicate a requirement for NF-E2 in generating normal erythrocytes. Despite impressive splenomegaly at baseline, mice lacking p45 NF-E2 survived splenectomy, which resulted in increased reticulocyte numbers. This reveals considerable erythroid reserve within extra-splenic sites of hematopoiesis and suggests a role for the spleen in clearing abnormal erythrocytes. Our findings address distinct aspects of the requirements for NF-E2 in blood cell homeostasis and establish its roles in proper differentiation of megakaryocytes and erythrocytes.

Expression of a foreign protein in human megakaryocytes and platelets by retrovirally mediated gene transfer.

Recent progress in the culture of human megakaryocytes (MKs) has led to the capacity to produce platelets in vitro. This capability enables investigation into the possibility of modifying platelet structure and/or function by genetically altering the MK. To this end, a cDNA for the murine CD9 (mCD9) cell surface protein was introduced into MK progenitors by retrovirally mediated gene transfer and subsequently detected in cultured MKs with a monoclonal antibody (MoAb) that specifically recognizes the murine protein. CD34+ human peripheral blood or marrow progenitors, enriched by immunomagnetic bead selection, were cultured for 5 days in the presence of growth factors, including stem cell factor and thrombopoietin, to induce MK progenitors into the cell cycle. The stimulated cells were then cocultured with the mCD9 retroviral producer cell line for 3 days, followed by culture in serum-depleted medium for 3 to 7 additional days. Flow cytometry analysis using the anti-CD9 MoAb and TAB, a MoAb recognizing human GPIIb, revealed that a large proportion (40-100%) of the MKs expressed mCD9. To ascertain whether these cells were capable of producing mCD9+ platelets, flow cytometry analysis was performed at a time when proplatelets were observed in the culture. mCD9 was detected in up to 59% of the TAB+ platelet-sized particles. Because deteriorating MKs can produce platelet-sized particles in vitro, experiments were performed to determine whether mCD9+ TAB+ particles were functionally active. Addition of phorbol myristate acetate resulted in the redistribution of P-selectin (CD62) from the alpha granule to the platelet surface as detected by MoAbs S12 and G5 in three-color flow cytometry analyses. These studies showed that up to 76% of the mCD9+ TAB+ particles were functionally active. The data show that retrovirally mediated gene transfer is a viable approach for genetically altering MK progenitors, resulting in platelets that express heterologous proteins.

Mice lacking transcription factor NF-E2 provide in vivo validation of the proplatelet model of thrombocytopoiesis and show a platelet production defect that is intrinsic to megakaryocytes.

Mechanisms of platelet production and release by mammalian megakaryocytes are poorly understood. We used thrombocytopenic knockout mice to better understand these processes. Proplatelets are filamentous extensions of terminally differentiated megakaryocytes that are thought to represent one mechanism of platelet release; however, these structures have largely been recognized in cultured cells and there has been no correlation between thrombocytopoiesis in vivo and proplatelet formation. Mice lacking transcription factor NF-E2 have a late arrest in megakaryocyte maturation, resulting in profound thrombocytopenia. In contrast to normal megakaryocytes, which generate abundant proplatelets, cells from these mice never produce proplatelets, even after prolonged stimulation with c-Mpl ligand. Similarly, megakaryocytes from thrombocytopenic mice with lineage-selective loss of transcription factor GATA-1 produce proplatelets very rarely. These findings establish a significant correlation between thrombocytopoiesis and proplatelet formation and suggest that the latter represents a physiologic mechanism of platelet release. We further show that proplatelet formation by normal megakaryocytes and its absence in cells lacking NF-E2 are independent of interactions with adherent (stromal) cells. Similarly, thrombocytopenia in NF-E2(-/-) mice reflects intrinsic defects in the megakaryocyte lineage. These observations improve our understanding of platelet production and validate the study of proplatelets in probing the underlying mechanisms.

Major effects of TPO delivered by a single injection of a recombinant adenovirus on prevention of septicemia and anemia associated with myelosuppression in mice: risk of sustained expression inducing myelofibrosis due to immunosuppression.

Adenoviral vectors may be useful tools to deliver a cytokine in vivo. A single intravenous injection of an adenovirus vector containing the human thrombopoietin (TPO) cDNA (AdRSVhuTPO) was able to induce a thrombocytosis for more than 6 weeks in SCID mice, associated with a megakaryocyte (MK) hyperplasia in different organs. A marrow and spleen fibrosis was observed at 6 weeks. In immunocompetent mice, a single AdRSVhuTPO injection led to a moderate and transient thrombocytosis without myelofibrosis. To evaluate the usefulness of TPO for the prevention of secondary side-effects during an aplastic period, mice were subjected to a myeloablative regimen 7 days after the intravenous AdRSVhuTPO injection. In this setting, TPO prevented mortality by accelerating hematological recovery. Survival was essentially related to an improvement in the leukopenia since all control mice died from septicemia. However, the effects of TPO may be potentiated by the release of inflammatory cytokines following the adenovirus infection; AdRSV beta galactosidase injected-mice had higher numbers of BFU-E and CFU-GM in the marrow than PBS-injected mice. Myelosuppression induced transient immunosuppression responsible for a sustained expression and elevation of platelet numbers for at least 5 months. These results further suggest that TPO may be an effective therapy in diminishing hematological complications related to myeloablative regimens, but emphasize that immunosuppression secondary to myelosuppression may lead to sustained expression associated with a risk of thrombosis and myelofibrosis when delivered by adenovirus vectors.

Thrombopoietin (TPO) knockout phenotype induced by cross-reactive antibodies against TPO following injection of mice with recombinant adenovirus encoding human TPO.

Adenovirus vectors have emerged as potent agents for gene transfer. Immune response against the vector and the encoded protein is one of the major factors in the transient expression following in vivo gene transfer. A single injection of an adenovirus encoding human thrombopoietin (TPO) into mice induced transient thrombocytosis, followed by a chronic immune thrombocytopenia. Thrombocytopenic mice had anti-human TPO Abs of the IgG2a and IgG1 isotypes. Thrombocytopenic mice sera neutralized more efficiently human than murine TPO, and exhibited no detectable anti-murine TPO Abs. Despite their low affinity for murine TPO, anti-TPO Abs induced a TPO knockout-like phenotype, i.e., low number of marrow megakaryocytes and of all kinds of hemopoietic progenitors. Hybridomas derived from a thrombocytopenic mouse revealed cross-reactivity of all of the secreted anti-TPO Ab isotypes. Mice subjected to myelosuppression after virus injection showed that anti-human TPO of IgG1 and IgG2a isotypes disappeared. Thus, sustained human TPO production was responsible for platelet elevation for at least 5 mo. Compelling results showed that elevated IgG2a/IgG2b ratios are always associated with thrombocytopenia, whereas low ratios are associated with tolerance or normal platelet counts. Finally, we hypothesize that in humans some chronic thrombocytopenia associated with a low TPO plasma level are due to anti-TPO Abs.

Mpl ligand or thrombopoietin: biological activities.

Thrombopoietin (TPO) or Mpl ligand is the primary physiological regulator of platelet production. This cytokine is the most potent stimulator of the proliferation and differentiation of MK progenitor and precursor cells in vitro. It also acts additively or synergistically with several cytokines on progenitor cells from various hematopoietic lineages, including the primitive stem cells. The factor is an extremely potent thrombocytopoietic agent when administrated to normal animals, and it accelerates platelet and erythropoietic recovery in several models of myelosuppression. Phase I/II clinical trials are ongoing with no detectable adverse effects. Mpl ligand does not induce platelet aggregation, but it lowers the platelet sensitivity to physiological dose of agonists. In experimental mouse models, high and chronic dose of Mpl ligand results in myelofibrosis. TPO is constantly produced by the liver and the kidney; its plasmatic clearance occurs by binding to its receptor expressed on megakaryocytes and platelets. However, the full spectrum of the biological effects of this new cytokine is not fully understood, in particular its the role in the terminal stage of platelet production. In the near future, it is likely that new insights will be obtained in the physiopathological mechanisms underlying abnormal platelet production in human.

Myelofibrosis: experimental models and human studies.

Thrombopoietin (TPO) is the central regulator of megakaryocytopoiesis and thrombocytopoiesis. Preclinical data and human studies have so far shown that the recombinant molecule is safe to administer and associated with very little toxicity. Nevertheless, different experimental animal models have revealed that a chronic exposure to very high doses of TPO could result in myeloproliferative syndromes with a spectrum of pathological features in common with human idiopathic myelofibrosis (PMF). A number of investigators have researched whether TPO or its receptor Mpl were involved in the pathogenesis of human myeloproliferative syndromes which are also characterized by a predominant megakaryocytic involvement, in PMF and primitive essential thrombocythemia. In both diseases, megakaryocyte (MK) progenitors develop autonomously in serum-deprived cultures. This spontaneous MK development is also observed at limiting dilution demonstrating that MK escape the normal regulatory controls. Furthermore, this abnormal MK proliferation and maturation is neither due to an autocrine stimulation by TPO nor by point mutation or deletion in the coding region of the c-mpl gene. This paper will review the data that have been reported to date on the effects of an overexpression of Mpl ligand and related molecules on the induction of experimental myelofibrosis and highlight recent insights into the pathogenesis of PMF.

High thrombopoietin production by hematopoietic cells induces a fatal myeloproliferative syndrome in mice.

To evaluate the effects of long-term, high-dose exposure to thrombopoietin (TPO), lethally irradiated mice were grafted with bone marrow cells infected with a retrovirus carrying the murine TPO cDNA. Mice were studied for 10 months after transplantation. In plasma, TPO levels were highly elevated (10(4) U/mL) throughout the course of the study. All mice developed a lethal myeloproliferative disorder evolving in two successive phases. During the first phase (7-9 weeks posttransplant), platelet and white blood cell (WBC) counts rose four- and ten-fold, respectively, whereas hematocrits decreased slightly to 29% +/- 3%. The WBC were mainly mature granulocytes, but myeloid precursor cells were invariably observed as well as giant platelets with an irregular granule distribution. The striking features were a massive hyperplasia of megakaryocytes and granulocytes in the spleen and bone marrow and a hypoplasia of erythroblasts in bone marrow. Total numbers of megakaryocyte colony-forming cell, burst-forming unit-erythroid, and granulocyte macrophage colony-forming cells were increased but colony-forming unit-erythroid numbers decreased. From 10 weeks posttransplant and thereafter, WBC, platelets, and red blood cell numbers declined dramatically. The absolute numbers of progenitor cells were very low in the spleen and bone marrow, but sharply increased in the blood and peritoneal cavity. Extramedullary hematopoiesis was observed in several organs. Histologic sections of the spleen and bones revealed severe fibrosis and osteosclerosis. The mean survival time was 7 months posttransplant and mice died with severe pancytopenia. Notably, two mice died between 3 and 4 months posttransplant with a leukemic transformation. This disorder was transplantable into secondary recipients who developed an attenuated form of the disease similar to the one previously described (Yan et al, Blood 86:4025, 1995). Taken together, our data show that high and persistent TPO production by transduced hematopoietic cells in mice results in a fatal myeloproliferative disorder that has a number of features in common with human idiopathic myelofibrosis.

Thrombopoietin does not induce lineage-restricted commitment of Mpl-R expressing pluripotent progenitors but permits their complete erythroid and megakaryocytic differentiation.

In this study, we examined the in vitro and in vivo effects of forced expression of Mpl-R (the thrombopoietin receptor) on the progeny of murine hematopoietic stem cells. Bone marrow cells from 5-FU-treated mice were transduced with retroviral vectors containing the human Mpl-R cDNA, or the neomycine gene as a control. After 7 days cocultivation on virus-producer cells, GpE86-Mpl-R or Gp86-Neo, the types of hematopoietic progenitor cells responding to thrombopoietin (TPO) were studied by clonogenic assays. Mpl-R-infected cells gave rise to CFU-GEMM, BFU-E, CFU-MK, but not CFU-GM while Neo-infected cells produced only megakaryocytic colonies. In addition, when nonadherent cells from GpE86-Mpl-R cocultures were grown with TPO as the only stimulus for 7 days, a marked expansion of CFU-GEMM, BFU-E, and CFU-MK was observed, while no change in CFU-GM number was seen. Erythroid and megakaryocytic maturation occurred in the presence of TPO while a block in granulocytic differentiation was observed at the myeloblast stage. The direct effects of TPO on Mpl-R-transduced progenitor cells were demonstrated by single cell cloning experiments. To analyze the effects of the constitutive expression of Mpl-R on the determination of multipotent progenitors (CFU-S) and long-term repopulating stem cells, Mpl-R- or Neo-infected cells were injected into lethally irradiated recipient mice. No difference was seen in (1) the number of committed progenitor cells contained in individual CFU-S12 whether colonies arose from noninfected or Mpl-R-infected CFU-S; (2) the mean numbers of progenitor cells per leg or spleen of mice reconstituted with Mpl-R- or Neo-infected cells, 1 or 7 months after the graft; and (3) the blood parameters of the two groups of animals, with the exception of a 50% reduction in circulating platelet counts after 7 months in mice repopulated with Mpl-R-infected bone marrow cells. These results indicate that retrovirus-mediated expression of Mpl-R in murine stem cells does not modify their ability to reconstitute all myeloid lineages of differentiation and does not result in a preferential commitment toward the megakaryocytic lineage.