Ramos et al. report a crucial role for macrophages in erythroblast development in mice. Under conditions that induce new red blood cell formation, macrophage depletion impaired red blood cell recovery. Conversely, macrophage depletion normalized red blood cell counts in mouse models of polycythemia vera and ®-thalassemia, pointing to a potential new therapeutic strategy for these diseases. Findings similar to these are reported in an accompanying paper by Chow et al. Regulation of erythropoiesis is achieved by the integration of distinct signals. Among them, macrophages are emerging as erythropoietin-complementary regulators of erythroid development, particularly under stress conditions. We investigated the contribution of macrophages to physiological and pathological conditions of enhanced erythropoiesis. We used mouse models of induced anemia, polycythemia vera and β-thalassemia in which macrophages were chemically depleted. Our data indicate that macrophages contribute decisively to recovery from induced anemia, as well as the pathological progression of polycythemia vera and β-thalassemia, by modulating erythroid proliferation and differentiation. We validated these observations in primary human cultures, showing a direct impact of macrophages on the proliferation and enucleation of erythroblasts from healthy individuals and patients with polycythemia vera or β-thalassemia. The contribution of macrophages to stress and pathological erythropoiesis, which we have termed stress erythropoiesis macrophage-supporting activity, may have therapeutic implications.

Abstract Abstract 81 We investigated the contribution of macrophages to physiological and pathological conditions in which erythropoietic activity is enhanced. We utilized mouse models of a) anemia by phlebotomy-induced stress erythropoiesis (SE); b) increased erythropoiesis by erythropoietin (Epo) administration; c) Polycythemia Vera (Jak2V617F/+ or PV) and d) beta-thalassemia intermedia (Hbbth3/+ or BTI) in which macrophages were chemically depleted by injection of liposome-clodronate (LC). While chronic injection (up to 3 months) of LC in normal mice had little effect on steady state erythropoiesis, depletion of macrophages severely impaired recovery from anemia following phlebotomy and significantly limited the increase in hematocrit (Htc) in animals treated with Epo. To exclude that this effect was mediated by decreased serum iron parameters, we used mice iron overloaded by dietary means or affected by hemochromatosis (Hfe-KO and Hamp-KO). In these mice, recovery from anemia was still impaired following macrophage depletion, even though serum iron and transferrin saturation levels were elevated and unaffected by LC administration. In vitro studies using both mouse and human primary erythroblasts (EBs) indicated that EBs in S-phase were twice as many compared to EBs cultured in absence of macrophages. The numbers of terminally mature erythroid cells were up to six fold higher in co-culture conditions. Experiments using transwells indicate that direct contact between EBs and macrophages was required to generate this effect. Since our data highlighted an important role of macrophages in enhancing erythropoiesis, we investigated two disorders in which the pool of erythroid progenitor cells is expanded, such as PV and BTI. Chronic administration of LC in PV mice completely reversed splenomegaly and the Htc (P<0.001). BTI mice exhibited normal spleen, amelioration of ineffective erythropoiesis (by accelerating the differentiation of EBs to erythrocytes), improvement of red blood cell (RBC) morphology, red cell distribution width (RDW, P<0.001) and increased hemoglobin levels (∼2g/dL, P<0.01). This effect was due to an increased RBC lifespan following LC administration (P<0.001), which was associated with a decrease in hemichrome formation, but not with a reduction in erythophagocytosis. Our observations indicate that macrophages directly modulate stress- and pathological erythropoiesis. Several adhesion molecules participate in the formation of interactions within the erythroblastic islands, including integrins. Interestingly, βeta1integrin and its associated protein, focal adhesion kinase-1 (Fak1), were reported to be necessary for the compensatory response to anemia, suggesting that this pathway might be involved in the macrophage-EB cross-talk. More EBs co-cultured with macrophages retained cell surface expression of βeta1integrin molecule during the last stage of cell differentiation compared to EBs cultured alone, even though other differentiation markers did not shown any variation. Fak1 phosphorylation in EBs was induced by co-culturing them with splenic macrophages, suggesting that Fak1 signaling is one of the pathways activated in EBs through contact with macrophages. Administration of a FAK1 inhibitor (FAK1i) decreased proliferation of EB co-cultured with macrophages, while delayed recovery from anemia and decreased the spleen size in phlebotomized animals (40% decrease compared to phlebotomized control animals at day 4; P=0.032). Finally, short-term administration of FAK1i to BTI animals rapidly reverted splenomegaly with a concurrent reduction of erythroid expansion in both BM and spleen and led to amelioration of anemia, supported by increased RBCs count. Our data indicate that, while macrophages allow proper erythroid response under conditions of induced anemia or increased erythropoiesis in wt mice, they contribute to the pathological progression of PV and BTI. Activation of Fak1 promotes erythroid proliferation and pathological development, while its inhibition limits ineffective erythropoiesis and splenomegaly in BTI. In conclusion, we identified a new mechanism contributing to the pathophysiology of these disorders, which we believe will have critical scientific and therapeutic implications in the near future. Disclosures: Levine: Agios Pharmaceuticals: Research Funding. Rivella:Novartis Pharmaceuticals: Consultancy; Biomarin: Consultancy; Merganser Biotech: Consultancy, Equity Ownership, Research Funding; Isis Pharma: Consultancy, Research Funding.

Abstract Abstract 1035 β-Thalassemia is a disorder associated with abnormal β-globin production, leading to anemia, extramedullary hematopoiesis (EMH), a decreased lifespan of the red cells and iron overload. In this disorder erythropoiesis is ineffective due to increased erythroid apoptosis and erythroblast proliferation, as well as deficient differentiation. Recent evidence suggests that erythroid development, especially under conditions of anemia (stress erythropoiesis), is highly dependent on microenvironmental factors within the erythroid niche, potentially mediated by the interaction of erythroblasts with macrophages. However, little is known about the function of these cells in pathological anemias associated with abnormal erythropoiesis. Our goal was to study the role of macrophages in normal, stress and ineffective erythropoiesis (IE). Macrophages were eliminated by intravenous administration of clodronate-containing liposomes. Treatment was carried out for up to 12 weeks, serial measurements being made of erythropoietic and pathological parameters. As a model of stress erythropoiesis, phlebotomized wt mice were used. To study IE we utilized th3/+ mice, a model of β-thalassemia intermedia (TI). Clodronate treatment effectively depleted splenic and bone marrow (BM) macrophages as shown by FACS and immunohistochemical analyses. Depletion of macrophages in wt mice had little effect on steady state erythropoiesis. In contrast, clodronate treatment drastically impaired the response to stress erythropoiesis in these mice, as shown by the slow recovery from phlebotomy-induced anemia. This was associated with a very slow rate of RBC and reticulocyte production, suggesting that erythroid activity was markedly impaired. Accordingly, mice depleted of macrophages were unable to expand their pool of erythroid progenitors in the BM and spleen in response to anemia, suggesting that macrophages play a critical role in this process. A similar defect was observed in response to Epo stimulation, suggesting that an intact erythroid niche is essential for normal activity of Epo in promoting erythroid expansion. Interestingly, TI mice treated with clodronate exhibited an improvement of the thalassemic phenotype. Within 40 hours of clodronate treatment, mice showed an increase in hemoglobin (Hb), RBC and reticulocyte counts in the peripheral blood, and a reduction of extra-medullary hematopoiesis (increased ratio of mature to immature erythroid cells) and splenomegaly (P<0.05 for all parameters analyzed). This indicated a more effective erythropoiesis in the absence of macrophages, suggesting that these cells negatively influence erythroid development in this disorder. Improvement of anemia was maintained for up to 12 weeks of continuous treatment, and was associated with increased RBC counts. Under these conditions, serum iron was markedly decreased, potentially reducing iron delivery to maturing RBCs. Recent studies have suggested that lowering iron delivery to erythroblasts leads to improvement of the RBC phenotype in this disorder. Consistently, MCHs were decreased after macrophage depletion, which correlated with lower accumulation of alpha-globin/heme precipitates in the RBC membranes. Moreover, the RBC lifespan in clodronate-treated mice was increased compared to that in PBS controls. This difference was maintained even when these cells were transfused into wt mice, suggesting that it was not associated exclusively with deficient RBC clearance in macrophage-depleted mice. In conclusion, our data suggests that macrophages have two major roles in β-thalassemia: 1) to modulate iron availability for erythroid cells; 2) to impair erythroid development, as suggested by the amelioration of splenomegaly and EMH observed after clodronate treatment. We hypothesize that the macrophages within erythroblastic islands control erythropoiesis, acting as modulators of this process. Under conditions of stress erythropoiesis they positively influence erythroid development, promoting proliferation to increase the pool of erythroid cells. However, under conditions of chronic stress such as in TI, macrophages limit differentiation and promote excessive expansion of the erythron, contributing to IE. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 3187 ß-Thalassemia and sickle cell disease (SCD) are the most common genetic red blood cell (RBC) disorders characterized respectively by limited production of aberrant ß-globin chains. In both cases, chronic transfusions and iron chelation are required to treat the anemia and/or formation of abnormal RBC. In ß-thalassemia, anemia stimulates erythropoietin (Epo) synthesis, which in turn leads to increased erythropoiesis and development of hepatosplenomegaly, often resulting in the need for splenectomy. Recently, we demonstrated that erythroid cells from ß-thalassemic mice have a hyper-activation of Jak2, a kinase that mediates the signaling triggered by the binding of Epo to the Epo receptor. This led us to hypothesize that Jak2 inhibitors could be utilized to minimize erythroid expansion in this disorder, limiting splenomegaly. A Jak2 inhibitor (Tg101209 or Tg) was first tested in mice affected by ß-thalassemia intermedia (th3/+). Two doses of Tg (150 and 100 mg/Kg/day) were given orally for 10 days. Tg administration induced a mild decrease of hemoglobin levels (8.8±0.2, 8±0.2 and 7.8±0.2g/dL for placebo, Tg 100 mg/Kg and Tg 150 mg/Kg treated mice, respectively. p<0.05) and in the number of reticulocytes (approximately 75% of the levels seen in controls, p<0.05). Splenomegaly was also reduced in Tg-treated mice (up to 60%; p<0.05), the extent of this effect correlating with the dosage used. Reduction of splenomegaly was associated with a decrease in the number of erythroid progenitors in this organ (p<0.05) and trend toward normalization of the splenic architecture. These data support our hypothesis that, in ß-thalassemia, splenomegaly is associated with increased erythroid proliferation and it can be alleviated by administration of Jak2 inhibitors, with only a mild increase in anemia. We further tested the effect of Tg in other anemias associated with extramedulary hematopoiesis (EMH) and splenomegaly, including SCD. Administration of the drug to mice affected by SCD led to a significant worsening of anemia (more pronounced than that seen in th3/+ mice) and a proportional reduction of splenomegaly and EMH. We then evaluated the outcome of combining Tg with blood transfusion, a common therapy in b-thalassemia and SCD. In b-thalassemia, massively enlarged spleens are believed to sequester a significant proportion of circulating RBC, thereby limiting their lifespan and the efficacy of transfusion regimens. We hypothesize that decreasing splenomegaly by administration of Jak2 inhibitors could increase the efficacy of transfusion. This was first tested in th3/+ animals. In this case, transfusion alone was sufficient to increase the hemoglobin (Hb) levels approximately 3 g/dL and reduce the spleen size to 65% of that seen in non-transfused controls. In this model, the combined effect of transfusion and administration of Tg was more effective, the spleen size been 50% of non-transfused controls (p<0.05). We further tested this approach in mice affected by ß-thalassemia-major (th3/th3), for which transfusion is required for survival and massive splenomegaly develops rapidly. Administration of Tg together with transfusion led to a greater increase in Hb levels compared to transfusion alone (9.3±0.4 vs 7.3±0.5g/dL, p<0.05). This was likely a consequence of reduced splenomegaly and decreased sequestration of RBCs in Tg/transfused mice. Lastly, we tested combination therapy in a mouse model of SCD. Mice treated with Tg and transfusion exhibited slightly lower levels of Hb than transfused controls (Hb=9.7±0.2g/dL versus Hb=10.9±0.2g/dL). However, compared to the control, mice receiving combination therapy exhibited a larger percentage of donor RBCs, while endogenous erythropoiesis was markedly suppressed along with the production of sickle RBCs (1.3±0.3×106 RBC/ul compared to transfused-controls exhibiting 2.7±0.3×106 RBC/ul). In summary, administration of Jak2 inhibitors might reduce the production of pathological cells that, together with preservation of the splenic architecture, could minimize the propensity of patients to thrombotic events. Furthermore, suppression of endogenous erythropoiesis and reduction of the transfusion regimen would be expected to also reduce iron accumulation, making it easier to prevent its toxic effects through chelation therapy. Disclosures: No relevant conflicts of interest to declare.

In hereditary hemochromatosis, mutations in HFE lead to iron overload through abnormally low levels of hepcidin. In addition, HFE potentially modulates cellular iron uptake by interacting with transferrin receptor, a crucial protein during erythropoiesis. However, the role of HFE in this process was never explored. We hypothesize that HFE modulates erythropoiesis by affecting dietary iron absorption and erythroid iron intake. To investigate this, we used Hfe-KO mice in conditions of altered dietary iron and erythropoiesis. We show that Hfe-KO mice can overcome phlebotomy-induced anemia more rapidly than wild-type mice (even when iron loaded). Second, we evaluated mice combining the hemochromatosis and β-thalassemia phenotypes. Our results suggest that lack of Hfe is advantageous in conditions of increased erythropoietic activity because of augmented iron mobilization driven by deficient hepcidin response. Lastly, we demonstrate that Hfe is expressed in erythroid cells and impairs iron uptake, whereas its absence exclusively from the hematopoietic compartment is sufficient to accelerate recovery from phlebotomy. In summary, we demonstrate that Hfe influences erythropoiesis by 2 distinct mechanisms: limiting hepcidin expression under conditions of simultaneous iron overload and stress erythropoiesis, and impairing transferrin-bound iron uptake by erythroid cells. Moreover, our results provide novel suggestions to improve the treatment of hemochromatosis.

Abstract Abstract 4251 A deficient hepcidin response to iron is the principal mechanism responsible for increased iron uptake from the diet leading to iron overload. In hereditary hemochromatosis (HH), mutations in the HFE gene lead to iron overload through abnormally low levels of hepcidin. Interestingly, hepcidin has been shown to respond to a variety of stimuli, including iron, hypoxia, erythropoiesis and inflammation, requiring integration of the respective signals for its regulation. Further studies showed that HFE/Hfe could also modulate cellular iron uptake by associating with the transferrin receptor-1 (Tfrc), a crucial protein for iron uptake by erythroid cells. In addition, some studies have reported altered erythropoietic values in HH patients. Despite these findings, the role of Hfe in erythropoiesis was never explored. We hypothesized that Hfe influences erythropoiesis by two distinct mechanisms: 1) limiting hepcidin expression, thereby increasing iron availability, under conditions of simultaneous iron overload and stress erythropoiesis; 2) participating directly in the control of transferrin-bound iron uptake by erythroid cells. To test this hypothesis we investigated the role of Hfe in erythropoiesis, aiming to uncover the relative contribution of each of the aforementioned mechanisms. When erythropoiesis was challenged by phlebotomy, Hfe-KO animals were able to recover faster from anemia (p≤0.05) than either normal or iron overloaded wt mice. In Hfe-KO mice, despite their increased iron load, downregulation of hepcidin in response to phlebotomy or erythropoietin administration was comparable to that seen in wt mice. In contrast, iron overloaded wt mice showed increased hepcidin expression both at steady state and after erythropoietic stimulation compared to wt or Hfe-KO mice. In phlebotomized mice fed a standard diet, analysis of serum iron and transferrin saturation indicated that wt mice on the standard diet were able to increase their serum iron very rapidly. After 24 hours, both wt and Hfe-KO mice had similar serum iron and transferrin saturation levels. On the other hand, wt mice kept on an iron deficient diet over the course of phlebotomy, were unable to overcome the phlebotomy-induced anemia. In contrast, Hfe-KO mice fed the low iron diet were able to recover from anemia, although at a slower pace than either Hfe-KO or wt mice on a standard diet. These data indicate that gastrointestinal iron absorption in both wt and Hfe-KO mice is a major factor leading to recovery from anemia, although the excess iron in the liver of Hfe-KO mice contributes to restoration of the red blood cell reservoir. Phlebotomy is the main tool utilized to treat iron overload in HH patients. However, our data suggests that this treatment leads to both mobilization of iron from stores and increased gastrointestinal iron absorption. These observations suggest that patients might benefit from a controlled iron diet or from supplementation with hepcidin or an hepcidin agonist to limit iron absorption. Next, we determined that Hfe is expressed in erythroid cells and that it interacts with Tfrc in murine erythroleukemia cells. Moreover, we discovered that the level of Tfrc expression in Hfe-KO cells is 80% of that seen in wt cells, as measured by flow cytometry. This observation, together with measurement of iron uptake using 59Fe-saturated transferrin, indicated that Hfe-KO erythroid cells take up significantly more iron than wt cells. To confirm that Hfe plays a role in erythropoiesis independent from that in the liver, we transplanted Hfe-KO or wt bone marrow cells into lethally irradiated wt recipients and analyzed their recovery from phlebotomy. We observed that recovery from anemia was faster in Hfe→wt than in wt→wt and was associated with increased mean corpuscular hemoglobin levels, suggesting that lack of Hfe in the hematopoietic compartment can lead to increased hemoglobin production. In summary, our results indicate that lack of Hfe enhances iron availability for erythropoiesis by two distinct mechanisms. On the one hand, Hfe plays an important role in maintaining erythroid iron homeostasis by limiting the response of hepcidin to iron, particularly under conditions of erythropoietic stimulation. On the other hand, lack of Hfe contributes directly to increased iron intake by erythroid progenitors, even in the absence of iron overload. Disclosures: No relevant conflicts of interest to declare.

Abstract Abstract 2023 Poster Board I-1045 Macrophages represent an important link between erythropoiesis and iron metabolism. They support maturation and differentiation of erythroblasts and clear senescent red cells from circulation, thus recycling iron within the hematopoietic system. However, little is known about the role of macrophages in conditions of ineffective erythropoiesis like beta-thalassemia intermedia (β-TI) or sickle cell anemia (SCD). Our goal was to study the role of macrophages under such conditions, where there is a large expansion of the erythron, the lifespan of the red cells is decreased, and both iron absorption and organ iron content are increased. Clodronate liposomes have been shown to efficiently eliminate macrophages from spleen, liver and bone marrow when administered by intravenous injection. Therefore, we utilized this technique to eliminate macrophages from the spleen, liver and bone marrow of β-TI and SCD mice. Three different treatment schedules were used; a) acute high dose; b) medium term split dose; c) chronic low dose. Animals were analyzed for erythropoietic and pathological parameters. All treatments effectively depleted splenic and bone marrow macrophages as shown by flow citometric and immunohistochemical analyses. At doses that led to only a small decrease of hemoglobin (Hb) levels in normal mice, clodronate treatment induced high level of mortality in β-TI and SCD mice (70% and 75%, respectively). In contrast, administration of clodronate to mice affected by beta-thalassemia major, mice that do not make any adult or fetal Hb/mature red cells, caused no mortality suggesting that circulating abnormal red cells might be responsible for the mortality observed in β-TI and SCD mice. Pathological analysis of β-TI mice revealed the formation of thrombi in the lungs and heart to be the likely cause of death. Analysis of SCD mice is in progress. β-TI and SCD have been described as hypercoagulable states with a high risk of thrombosis. Damaged red cell membranes and ROS have been implicated as factors that increase the risk of thrombosis in β-TI and SCD patients. We hypothesize that clodronate treatment enhances this process. To test this hypothesis, we administered an anti-coagulant (heparin) and an anti-oxidant (NAC) to β-TI mice. Both these treatments resulted in increased survival from 31% to 65% compared to clodronate alone. Analysis of SCD animals is in progress. β-TI mice that survived clodronate treatment exhibited a milder thalassemic phenotype, characterized by increased Hb (∼1.5g/dL), HCT (∼7%) and RBC count (∼2e6 cells/ul), partially associated with a decreased RBC clearance. Ineffective erythropoiesis and splenomegaly were also ameliorated (∼40% decrease in spleen size). Animals depleted of macrophages also showed a marked decrease in serum iron parameters, but no difference in their organ iron concentrations. Analysis of hepcidin expression is in progress. To further corroborate these observations, β-TI mice are being crossed with mice expressing the human diphtheria toxin (CD11b-DTR) which allows for conditional depletion of macrophages. Our data suggests that macrophages play a major role in conditions of abnormal red cell production and ineffective erythropoiesis. In the former case, they have an important protective function related to the high risk of thrombosis associated with β-TI and SCD. In the latter, macrophages may act as negative modulators of erythropoietic development. The amelioration of splenomegaly and ineffective erythropoiesis observed suggests that erythroid cells might proliferate less and/or differentiate more in the absence of macrophages. One hypothesis is that macrophages sense erythroid maturation and/or differentiation and act as modulators of these processes to ensure the release of high quality red cells into the circulation. Second, macrophages, under conditions of iron overload, might produce inflammatory cytokines that further impair erythropoiesis in thalassemia. Analysis of RNA extracted from macrophages and of inflammatory cytokines in β-TI mice before and after clodronate treatment is in progress. Disclosures: No relevant conflicts of interest to declare.

OBJECTIVE:CD36 has been shown to play a role in atherosclerosis in the apolipoprotein E-knockout (apoE(o)) mouse. We observed no difference in aortic lesion area between Western diet (WD)-fed LDLR(o) and LDLR(o)/CD36(o) mice. The objective was to understand the mechanism of CD36-dependent atherogenesis. METHODS AND RESULTS:ApoE(o) mice transplanted with bone marrow from LDLR(o)/CD36(o) mice had significantly less aortic lesion compared with those transplanted with LDLR(o) marrow. Reciprocal macrophage transfer into hyperlipidemic apoE(o) and LDLR(o) animals showed that foam cell formation induced by in vivo modified lipoproteins was dependent on the lipoprotein, not macrophage type. LDLR(o) and LDLR(o)/CD36(o) mice were fed a cholesterol-enriched diet (HC), and we observed significant lesion inhibition in LDLR(o)/CD36(o) mice. LDL/plasma isolated from HC-fed LDLR(o) mice induced significantly greater jnk phosphorylation, cytokine release, and reactive oxygen species secretion than LDL/plasma from WD-fed LDLR(o) mice, and this was CD36-dependent. HC-fed LDLR(o) mice had higher circulating levels of cytokines than WD-fed mice. CONCLUSIONS:These data support the hypothesis that CD36-dependent atherogenesis is contingent on a proinflammatory milieu that promotes the creation of specific CD36 ligands, not solely hypercholesterolemia, and may explain the greater degree/accelerated rate of atherosclerosis observed in syndromes associated with inflammatory risk.

A single mutation in the HFE gene (C282Y), a non-classical member of the MHC-I family, may lead to hereditary hemochromatosis (HH). Although the hallmark of HH is iron overload, several lines of evidence point to a distinct hematopoietic function for HFE:

Abstract Abstract 2020 Poster Board I-1042 β-thalassemia intermedia (TI) and major (TM) are characterized by Ineffective Erythropoiesis (IE). We hypothesized that the kinase Jak2 plays a major role in IE and splenomegaly. To test this hypothesis we administered a Jak2 inhibitor (TG101209) to mice affected by TI, showing that this treatment was associated with a marked decrease in IE, and a moderate decrease in hemoglobin (Hb) levels (∼1 g/dL). This last observation indicates that the use of a Jak2 inhibitor might exacerbate anemia in thalassemia. However, we hypothesized that using standard transfusion to treat TM mice would also be adequate to prevent any further anemia caused by Jak2 inhibition while still allowing for decreased splenomegaly. Therefore, we analyzed the erythropoiesis and iron metabolism in TM animals treated with a Jak2 inhibitor and transfused. Use of TG101209 in TM mice not only reduced the spleen size dramatically (0.42±0.15 g and 0.19±0.10 g respectively in transfused+placebo (N=4) vs transfused+TG101209 (N=8), P= 0.007), but also allowed the mice to maintain higher Hb levels (respectively 7.3±1.1 g/dl vs 9.3±1.2 g/dl, P=0.019). This was likely due to reduced spleen size and limited red cell sequestration. Contrary to TM mice treated with transfusion+placebo, no foci of extra-medullary hematopoiesis were detectable in the parenchema of mice treated with TG101209. Hamp1 expression inversely correlated with the spleen weight, possibly indicating that suppression of IE (due both to blood transfusion and TG101209 administration) had a positive effect on Hamp1 expression. Hb levels also directly correlated with Hamp1 expression in the same animals. In this case, however, only transfusion played a role in increasing Hamp1 expression, although TG101209 undoubtedly had a positive effect by reducing the spleen size and thereby indirectly increasing the Hb levels. The suppression of erythropoiesis by blood transfusion limits the extent of our interpretations as it may mask the effect of the Jak2 inhibitor. Therefore we hypothesized that the administration of a tailored and reduced dose of the drug could be effective in reducing the splenomegaly in non-transfused TI mice, without affecting the Hb levels. We also hypothesized that the suppression of erythropoiesis would also lead to increased Hamp1 expression in the presence of iron overload. Compared to mice treated with placebo (N=5), analysis of TI mice treated with a tailored dose of 100mg/kg/day of the drug (N=11) showed a significant decrease in spleen size (0.18±0.05 g and 0.27±0.05 g, P=0.006 for drug treated mice and placebo treated mice respectively). Of note no significant difference of Hb levels was detectable between the 2 groups. In the drug treated mice we observed a significant decrease of the immature erythroid cell population (P=0.012) and amelioration of the architecture of the spleen, with the reappearance of white pulp foci and a significant restoration of the splenic lymphocitic populations. Drug treated mice showed increased levels of Hamp1 mRNA that inversely correlated with the spleen weight, suggesting a direct feedback between erythropoietic rate and expression of Hamp1. To determine if the use of Jak2 inhibitors could be beneficial in a mouse model mimicking a human form of hereditary ellyptocytosis, we treated mice KO for the 4.1R protein isoforms with TG101209, in presence or absence of blood transfusions. These mice exhibit moderate splenomegaly and anemia and the drug treatment was effective in reducing the spleen weight and the associated IE. We also plan to analyze Sickle Cell mice that we are treating with a Jak2 inhibitor. In conclusion our data show that the administration of Jak2 inhibitors is efficient in decreasing the spleen size and ameliorating the pathologic iron metabolism in thalassemia, both in the presence or absence of blood transfusions. Moreover we show that Jak2 inhibitors could transform the therapeutic approach for other forms of anemias. Disclosures: No relevant conflicts of interest to declare.

In beta-thalassemia, the mechanism driving ineffective erythropoiesis (IE) is insufficiently understood. We analyzed mice affected by beta-thalassemia and observed, unexpectedly, a relatively small increase in apoptosis of their erythroid cells compared with healthy mice. Therefore, we sought to determine whether IE could also be characterized by limited erythroid cell differentiation. In thalassemic mice, we observed that a greater than normal percentage of erythroid cells was in S-phase, exhibiting an erythroblast-like morphology. Thalassemic cells were associated with expression of cell cycle-promoting genes such as EpoR, Jak2, Cyclin-A, Cdk2, and Ki-67 and the antiapoptotic protein Bcl-X(L). The cells also differentiated less than normal erythroid ones in vitro. To investigate whether Jak2 could be responsible for the limited cell differentiation, we administered a Jak2 inhibitor, TG101209, to healthy and thalassemic mice. Exposure to TG101209 dramatically decreased the spleen size but also affected anemia. Although our data do not exclude a role for apoptosis in IE, we propose that expansion of the erythroid pool followed by limited cell differentiation exacerbates IE in thalassemia. In addition, these results suggest that use of Jak2 inhibitors has the potential to profoundly change the management of this disorder.

β-Thalassemia intermedia and major are characterized by ineffective erythropoiesis (IE), requiring sporadic or chronic blood transfusions, respectively. Some of the major consequences of IE are extra-medullary hematopoiesis (EMH), splenomegaly and systemic iron overload mediated by transfusion therapy and down-regulation of hepcidin. Using mouse models of β-thalassemia intermedia (th3/+) and major (th3/th3), and human specimens we investigated IE in this disorder. Th3/+ and th3/th3 erythroid cells were analyzed with respect to rates of apoptosis and degrees of cell proliferation and differentiation. We found that there was both a relative and absolute expansion of the immature erythroid progenitor cell fraction in thalassemic mice compared to cells in the final stages of differentiation. Further investigation of the thalassemic erythroid cells in vivo and in vitro indicated that a larger number of the thalassemic cells are associated with the phosphorylated form of the Jak2 protein kinase than in normal mice. In fact, their proliferation was prevented by TG101209, a Jak2 inhibitor. Similar compounds are currently utilized or being considered for use in the treatment of myeloproliferative diseases such as polycythemia vera. In order to assess the potential of Jak2 inhibitors in limiting IE in β-thalassemia, we administered TG101209 to th3/+ mice. We found that both 10 and 18 days of treatment were sufficient to dramatically reduce the spleen size and the percentage of immature erythroid progenitors therein compared with administration of a placebo. However, these changes were associated with decreasing hemoglobin levels (Blood , 2008, 112:875–85). We speculated that this problem could be overcome by administration of blood transfusions during treatment, an extension of current management in thalassemia as noted above. To test this hypothesis, we administered TG101209 to th3/th3 mice in conjunction with regular blood transfusions. Administration of TG101209 to transfused th3/+ mice is in progress. Our preliminary data on th3/th3 mice indicates that simultaneous administration of TG101209 and transfused blood not only reverses splenomegaly, but also results in higher Hb levels (N ≥ 3). The increased hemoglobin levels observed, compared to those in mice treated with transfusion therapy alone, suggest that the use of Jak2 inhibitors may reduce the amount of blood per transfusion and/or the rate of transfusion required in thalassemia. Since the increased iron absorption in thalassemic mice is a direct consequence of IE, treatment with TG101209 could also be beneficial in ameliorating this process. According to several observations, suppression of erythropoiesis should lead to increased Hamp1 expression in the presence of iron overload. Therefore, under conditions of Jak2 inhibition, Hamp1 transcription is expected to increase. To test this hypothesis we are currently analyzing organ iron levels and the expression of iron-related genes in drug-treated mice. In conclusion, although our study does not exclude a role for apoptosis in the IE of β-thalassemia, we have demonstrated that increased cell proliferation and limited cell differentiation play a significant role in this process. Moreover, we show for the first time that a Jak2 inhibitor is effective in decreasing the spleen size of thalassemic mice. This could represent a completely new approach to the treatment of splenomegaly in β-thalassemia patients, perhaps coupled with blood transfusion. That administration of a Jak2 inhibitor reverses splenomegaly and also ameliorates the degree of iron overload could provide an opportunity to gain new insight into the dynamic processes of iron absorption and erythropoiesis in this pathological condition.

Robert W. Grady, Johannes Gerdes and Stefano Rivella de Sousa, Eliezer A. Rachmilewitz, John D. Hood, M. Domenica Cappellini, Patricia J. Giardina, Scholzen, Amy Chadburn, YiFang Liu, Margrit Kernbach, Bettina Baron-Lühr, Matteo Porotto, Maria Ilaria V. Libani, Ella C. Guy, Luca Melchiori, Raffaella Schiro, Pedro Ramos, Laura Breda, Thomas -thalassemia β erythropoiesis in Decreased differentiation of erythroid cells exacerbates ineffective

Patients with β-thalassemia hyper absorb dietary iron, most of which is stored in the liver. They also suffer from ineffective erythropoiesis (IE) which leads to hepatosplenomegaly, often requiring a splenectomy. We have been conducting a series of studies utilizing the th3/+ mouse model of thalassemia intermedia to investigate the absorption, distribution and erythroid utilization of iron. Here we focus on changes in the iron content of liver and spleen resulting from diets containing low (2.5 ppm), sufficient (35 ppm) and high (200 ppm) levels of iron, and assess the impact of splenectomy on its distribution. The high iron diet was standard rodent chow while the others were defined diets. Th3/+ mice were either bred or generated by transplantation of th3/+ hematopoietic stem cells from E14.5 fetal livers into lethally irradiated wild type (+/+) recipients. Wild type controls were similarly obtained. Splenectomy of bred and recipient mice was performed at 5 weeks of age and bone marrow transplantation (BMT) at 8 weeks. Non-transplanted mice were placed on the test diets at 8 weeks of age, and transplanted mice at 11 weeks. All animals were sacrificed after 4 weeks on the test diets, and livers and spleens harvested for determination of their iron content by atomic absorption. Group sizes ranged from 3 to 10 mice (median 7). In general, the mean organ iron content of mice fed the high iron diet was not significantly different from that of the animals fed the iron sufficient diet, while those fed the low iron diet had reduced levels of tissue iron. Over the course of the 4-week feeding study, the iron content of the livers and spleens of +/+ mice fed the 35-ppm diet increased 39% and 202%, respectively, while the corresponding values of those fed the 2.5-ppm diet were −21% and 30%. The changes in the liver and spleen of th3/+ mice were 79% and 32% (35-ppm diet) and 14% and 12% (2.5-ppm diet) compared to the values at baseline. The latter values, those at 8 weeks of age, were 1.8- and 30-fold higher in the th3/+ mice, the massive accumulation of iron in the spleen undoubtedly resulting from IE. Where iron intake (liver plus spleen) was low, it went preferentially to the spleen, undoubtedly to sustain erythropoiesis. Groups of splenectomized +/+ mice were also fed the three diets for 4 weeks. The mean iron content of their livers was similar to that of non-splenectomized animals. Similar studies of th3/+ mice are now in progress. A second set of studies is being conducted in transplanted +/+ and th3/+ mice, the goal being to determine whether or not the absorption and distribution of iron is the same as in bred animals. Again, the organ iron content of those mice fed the high iron diet was similar to that of the animals fed the iron sufficient diet. In the case of the transplanted +/+ animals fed iron sufficient diets, the mean iron contents of the livers and spleens were 64% and 186% increased after 4 weeks of feeding, values not markedly different from those of bred animals. The corresponding values on the 2.5-ppm diet were 27% and 72%, again the pattern being similar. The transplanted th3/+ animals accumulated significantly less iron in these organs than those that were bred. However, the rate at which they accumulated this iron was 10 to 20 times higher than that of the other groups studied, including the transplanted +/+ mice, perhaps reflecting a synergistic effect of BMT and IE on iron absorption. Mice fed the 35-ppm diet had only 75% and 46% as much iron in their livers and spleens, the animals fed the 2.5-ppm diet having even less (35% and 23%) while again showing preferential diversion of iron to the spleen. Splenectomizing the animals resulted in further increasing the liver iron, more that 2.5-fold in those fed the low iron diet. The hemoglobin levels of all the mice evaluated were unchanged as a result of the dietary studies, except for a 20% decrease seen in bred +/+ mice fed the low iron diet. We are currently studying splenectomized transplanted th3/+ mice as well as doing feeding studies of 5-months duration. In summary, a low iron diet has a marked effect on the iron levels of liver and spleen, which are accentuated under conditions of IE. Secondly, more iron is absorbed under conditions of IE than is needed for erythropoiesis, the excess being shuttled to the liver for storage.

Ineffective erythropoiesis (IE) in β-thalassemia has been attributed to erythroid cell death mediated by apoptosis or hemolysis during the maturation process. Historically, ferrokinetic studies in this disease suggested that 60%–80% of erythroid precursors die in the marrow or extramedullary sites. However, several observations have challenged this view. First, the number of apoptotic erythroid cells in patients is low compared to net expansion of the erythroid cell pool. Second, hemolytic markers in β-thalassemic patients are normal or only slightly increased, unless additional pathological conditions appear. Third, our most recent study (Blood, Gardenghi et al, 2007 Jun 1) demonstrated that GI iron absorption in β-thalassemia is increased by the dysregulation of genes such as hepcidin and ferroportin that control iron absorption, resulting in iron levels that exceed the amount required for erythropoiesis. We have undertaken a detailed investigation using cohorts of mice (n>30 per genotype) with β-thalassemia intermedia ( th3 /+) and major ( th3/th3 ). Using these models, we have previously shown that the severity of anemia (as low as 1 g/dL) inversely correlates with the total number of nucleated erythroid cells (»100 fold compared to wild-type (wt) mice). Cytological analysis has clearly shown that thalassemic spleen specimens were comprised of a homogeneous pre-erythroblastic population. In contrast, the percentage of apoptotic cells and the level of hemolytic markers, such as bilirubin and lactic acid dehydrogenase, slightly increased or were not different compared to wt mice. While not excluding a role for apoptosis, our observations suggest that control of the cell cycle and maturation of erythroid precursors play an important role in IE. We then explored whether the erythroid cell cycle was dysregulated in our model system. We found that erythropoietin (Epo) levels were raised in thalassemic animals by as much as three orders of magnitude. Binding of Epo to its receptor (EpoR), activates antiapoptotic and cell cycle promoting genes, through activation of Jak2 and Stat5. By Western blot we demonstrated up-regulation of EpoR, Stat5 and the antiapoptotic protein BclXL, as well as that of proliferation promoting genes, such as CycA and Cdk2, in purified thalassemic erythroid cells compared to those of wt animals. This data was confirmed by staining both wt and thalassemic liver and spleen sections using the proliferation markers Ki67 and Mcm3, by clonogenic assay and by analysis of the percentage of erythroid cells in S-phase after BrdU injection. In the latter case, we observed 22%, 30% and 44% BrdU + cells from wt, th3 /+ and th3/th3 mice, respectively. In addition, freshly purified thalassemic erythroid cells proliferate faster in vitro than normal cells, a phenomenon blocked by AG490, a Jak2 inhibitor. Significantly, we have been able to reproduce results from our animal studies in humans, comparing normal and thalassemic blood and spleen specimens. In conclusion, we propose that IE in β-thalassemia is likely to be the result of altered cell proliferation and impaired cell differentiation, which in turn limit apoptosis, thereby mimicking tumor-like behavior.

Progressive iron overload is the most salient and ultimately fatal complication of beta-thalassemia. However, little is known about the relationship among ineffective erythropoiesis (IE), the role of iron-regulatory genes, and tissue iron distribution in beta-thalassemia. We analyzed tissue iron content and iron-regulatory gene expression in the liver, duodenum, spleen, bone marrow, kidney, and heart of mice up to 1 year old that exhibit levels of iron overload and anemia consistent with both beta-thalassemia intermedia (th3/+) and major (th3/th3). Here we show, for the first time, that tissue and cellular iron distribution are abnormal and different in th3/+ and th3/th3 mice, and that transfusion therapy can rescue mice affected by beta-thalassemia major and modify both the absorption and distribution of iron. Our study reveals that the degree of IE dictates tissue iron distribution and that IE and iron content regulate hepcidin (Hamp1) and other iron-regulatory genes such as Hfe and Cebpa. In young th3/+ and th3/th3 mice, low Hamp1 levels are responsible for increased iron absorption. However, in 1-year-old th3/+ animals, Hamp1 levels rise and it is rather the increase of ferroportin (Fpn1) that sustains iron accumulation, thus revealing a fundamental role of this iron transporter in the iron overload of beta-thalassemia.

Previously, absence of the scavenger receptor CD36 or SRA was shown to be protective against atherosclerosis in the apoE KO after 12 weeks on western diet. A recent report (JCI 115. 2192 (2005)) showed no effect for SRA, and protection in females only for CD36 by en face analysis after 8 weeks on western diet. Here we determined if combined absence of CD36 and SRA had greater effect in the apoE KO. We analyzed lesions from CD36/apo E double KOs (dkos) and SRA/apoE dkos as controls. These mice are derived from those used in the JCI study. Because atherosclerosis is a multistep process involving many different factors over time, we also investigated whether absence of CD36 was protective at later time points in the apoE KO. Methods For the tko study, mice were on western diet for 12 weeks; for the long term study, mice were on diet for 20 and 35 weeks. Aortas were dissected and subjected to en face analysis. After 12 weeks on diet, tko males had 58% less lesion area compared with apoE KO males; tko females had 63% less lesion area. Male CD36/apoE dkos had 61% less lesion compared with apoE KO males; CD36/apoE dko females had 81% less lesion. These results indicate no added benefit to absence of SRA compared with absence of CD36 alone in this model. At 20 and 35 weeks, lesions were 25–28% smaller in the CD36/apoE dkos compared to apoE KOs. However, most impressive here was the difference in the gross appearance of the aortas: the apoE KO aortas were sclerotic and nearly occluded by lesion, whereas in the CD36/apoE dkos, the lesions were more punctate. Funding: NIH HL70083, HL072942, & HL46403