ALBERT

Description: Extended Lagrangian Born-Oppenheimer molecular dynamics is developed and analyzed for applications in canonical (NVT) simulations. Three different approaches are considered: the Nosé and Andersen thermostats and Langevin dynamics. We have tested the temperature distribution under different conditions of self-consistent field (SCF) convergence and time step and compared the results to analytical predictions. We find that the simulations based on the extended Lagrangian Born-Oppenheimer framework provide accurate canonical distributions even under approximate SCF convergence, often requiring only a single diagonalization per time step, whereas regular Born-Oppenheimer formulations exhibit unphysical fluctuations unless a sufficiently high degree of convergence is reached at each time step. The thermostated extended Lagrangian framework thus offers an accurate approach to sample processes in the canonical ensemble at a fraction of the computational cost of regular Born-Oppenheimer molecular dynamics simulations.

Description: The Journal of Organic Chemistry DOI: 10.1021/jo501732q

Description: β-thalassemia is characterized by ineffective erythropoiesis and iron overload. Ineffective erythropoiesis causes iron overload by suppressing hepcidin, the main negative regulator of iron absorption and recycling, and is mediated by secretion of erythroferrone from bone marrow cells. Targeted treatment for ineffective erythropoiesis is unavailable. Furthermore, molecular mechanisms involved in ineffective erythropoiesis and the details of how erythropoiesis regulates iron metabolism are incompletely understood. Lastly, while loss of erythroferrone in β-thalassemic mice leads to partial reversal of iron overload [Kautz Blood 2015], erythroferrone ablated mice are still able to suppress hepcidin after phlebotomy [Kautz Nat Med 2014]. These finding provide evidence of additional regulatory crosstalk between erythropoiesis and iron metabolism. We hypothesize that bone-marrow derived exosomes regulate iron metabolism by modulating hepcidin. Exosomes are small extracellular vesicles derived from multi-vesicular bodies forming intraluminal vesicles which fuse with the plasma membrane and are released by many different cell types [Thery Nat Rev Immun 2002]. In light of their capacity for cell-cell communication and modification of the microenvironment, exosomes have been widely studied in multiple diseases [Valadi Nat Cell Bio 2007] despite which, erythropoiesis-derived exosomes and their role in iron metabolism regulation remain unexplored. Our preliminary data demonstrate that phlebotomy in wild type mice results in increased exosome concentration in serum and that exosomes are increased in th3/+ mouse serum (Figure 1a). Furthermore, hepcidin induction by exosome depleted-FBS is decreased relative to FBS (Figure 1b), and exosomes isolated from FBS induce hepcidin in a dose response manner in vitro (Figure 1c). We thus propose to explore the mechanistic relationship between exosomes and hepcidin regulation in β-thalassemia. Serum samples from patients with β-thalassemia major and age / gender matched controls were collected; all patients were treated with iron chelation therapy and all samples were collected immediately prior to transfusion. Exosome fractions were purified and analyzed in patients relative to controls. Although there is no difference in the number of exosomes or mean particle size within the exosomal fraction, exosomal protein content per volume of serum is significantly decreased in patients relative to controls. In addition, the treatment of primary wild type mouse hepatocytes with sera from patients and controls reveals the expected relatively decreased hepcidin induction in β-thalassemic patient sera treated hepatocytes relative to control sera; a similar difference is seen in hepatocytes treated with exosome-depleted sera from patients and controls (Figure 2a). These findings suggest that hepcidin suppression is a consequence of the exosome-free portion of serum from control and β-thalassemic samples. Furthermore, only exosomes derived from β-thalassemic patient sera induces hepcidin expression in primary wild type mouse hepatocyte cultures (Figure 2b). Lastly, exosomes derived from β-thalassemic patient sera do not affect ERK1/2 and STAT3 signaling in primary hepatocytes but increase SMAD1/5/8 (Figure 2c) and decrease AKT signaling (Figure 2d). Taken together, these findings demonstrate that exosomes enhance hepcidin expression via increased SMAD1/5/8 signaling, that increased hepcidin may influence multiple signaling pathways by an autocrine mechanism in response to exosomes, and that exosomes counterbalance hepcidin suppressive substances in the exosome-depleted serum from β-thalassemic samples. Our studies provide novel insights into the important previously unexplored mechanism of hepcidin regulation by exosomes in both physiologic and pathologic states. Disclosures Coates: apo pharma: Consultancy, Honoraria, Speakers Bureau; vifor: Consultancy, Honoraria; celgene: Consultancy, Honoraria, Other: steering committee of clinical study; agios pharma: Consultancy, Honoraria. Ginzburg:La Jolla Pharma: Membership on an entity's Board of Directors or advisory committees.

Description: Myelodysplastic syndrome (MDS) is a heterogeneous group of bone marrow stem cell disorders characterized by ineffective hematopoiesis, cytopenias, and transformation to acute leukemia (AML). Low risk MDS patients exhibit a longer median survival, the lowest rate of progression to AML, and account for approximately two-thirds of all MDS patients, requiring recurrent RBC transfusions to alleviate symptomatic anemia. Transfusion-dependence results in iron overload which is associated with reduced overall survival. Methods to diagnose and treat (e.g. iron chelator deferiprone (DFP)) iron overload are available, but because retrospective studies to date only identify a correlative, not a causal, relationship between iron overload and reduced survival, there is little consensus on whether benefits of its diagnosis and treatment outweigh risks in this patient population. We aimed to evaluate and characterize the effect of iron loading and iron chelation on ineffective hematopoiesis in NUP98-HOXD13 transgenic (NHD13) mice, a highly penetrant model with peripheral blood cytopenias and ineffective hematopoiesis with dysplasia, consistent with MDS in humans [Lin Blood 2005]. We hypothesize that iron overload in MDS has deleterious biological effects on hematopoiesis with increased likelihood of worsening disease, reversible with iron chelation therapy. Our preliminary data demonstrates that 1) DFP enhances erythroid differentiation in CD34+ cells from MDS patients; and 2) exogenous iron suppresses erythropoiesis in wild type (WT) mice. The current experiments reveal the effect of exogenous iron and DFP in NHD13 mice. Age and gender matched 5 month old mice, 5-10 mice per group, were injected with 20mg iron dextran and/or treated with DFP (175 mg/mouse) in the drinking water over 4 weeks. Treated mice were analyzed and compared with PBS injected NHD13 mice and WT controls. Our experiments demonstrate that NHD13 mice exhibit anemia and leukopenia, increased serum erythropoietin (Epo) concentration, and higher MCV despite no difference in reticulocyte count relative to WT controls (Table I) and no splenomegaly (Fig 1a). In addition, erythroblast ROS concentration (Fig 1b) and iron load in the liver (Fig 1c) increase without differences in transferrin saturation or hepcidin expression in NHD13 relative to WT mice. These findings demonstrate characteristics consistent with human MDS patients. We then analyzed the effects of additional iron loading, iron chelation, or a combination of both on parameters of iron metabolism and erythropoiesis in NHD13 mice. Iron-treated NHD13 mice exhibit higher WBC count, RBC count, and hemoglobin, both DFP- and iron-treated NHD13 mice exhibit decreased serum Epo, and DFP+iron-treated NHD13 mice exhibit decreased serum Epo while increasing RBC count and hemoglobin (Table I). These findings suggest that exogenous iron increases Epo responsiveness and extramedullary erythroid mass in NHD13 mice in a manner similar to what we observed in iron-treated thalassemic mice [Ginzburg Exp Hem 2009]. Furthermore, spleen size increased (Fig 1a) and erythroblast ROS decreased (Fig 1b) only in iron-treated NHD13 mice. These findings suggest that erythroblast ROS is unrelated to excess iron in NHD13 mice. Liver iron increased in iron-treated and decreased in DFP-treated NHD13 mice (Fig 1c) as expected. In addition, CD71 (TfR1) surface expression on bone marrow erythroblasts is suppressed in all treated NHD13 mice but only significantly decreased in DFP-treated mice (Fig 1d), suggesting expected iron restriction. Lastly, DFP- and DFP+iron-treated (but not iron-treated) NHD13 mice decreased the proportion of bone marrow erythroblasts (Fig 1e) and increased erythroid differentiation in NHD13 mouse bone marrow (Fig 1f). Taken together, our findings demonstrate the robustness of NHD13 mice as a model of MDS to study erythropoiesis, the utility of iron injection in NHD13 mice to mimic robust iron overload in MDS, and the effectiveness of DFP in enhancing erythroid differentiation, reversal of erythroid expansion, and Epo responsiveness in NHD13 mice. Additional experiments (i.e. RBC survival, analysis of bone marrow signaling, erythroferrone expression, and parameters of the iron restriction response) to further explore the dysregulation of iron metabolism in NHD13 mice are ongoing. Disclosures No relevant conflicts of interest to declare.

Description: Patients with β-thalassemias manifest anemia, ineffective erythropoiesis, extramedullary hematopoiesis, splenomegaly, and systemic iron overload. Even in non-transfusion dependent patients, iron overload in β-thalassemia develops because of increased intestinal iron absorption, leading to multiple organ dysfunction if untreated and accounts for most of the deaths in this disease. The main regulator of body iron content and distribution is hepcidin, inhibiting iron absorption in duodenal enterocytes and release of stored iron from macrophages and hepatocytes. Despite iron overload in patients and mice with β-thalassemia, hepcidin levels are insufficiently increased, as ineffective erythropoiesis dominates hepcidin regulation. Relatively low hepcidin causes iron overload in β-thalassemia. Recent evidence demonstrates that erythroferrone (ERFE), an erythroid regulator of hepcidin, is increased in bone marrow and serum from β-thalassemic patients and th3/+ mice [Kautz Nat Gen 2014] and its loss results in increased hepcidin, partially reversing iron overload in th3/+ mice [Kautz Blood 2015]. In addition, bone marrow ERFE expression normalizes in TfR1 haploinsufficient th3/+ mice [Li Blood 2017]. We hypothesize that the loss of ERFE and TfR1 influences erythropoiesis and iron metabolism in complementary ways in th3/+ mice, and therefore aim to explore iron- and erythropoiesis-related parameters in th3/+ TfR1+/- ERFE-/- (triple mutant (TM)) mice. All models are on a C57BL6 background and have been crossed to generate 4-6 mice for analysis at 6 weeks of age. We confirm our previous reports [Li Blood 2017] that th3/+TfR1+/- mice have increased RBC count and hemoglobin, decreased MCV and reticulocyte count (Table I), and reduce splenomegaly (Fig 1a and 1b) relative to th3/+ mice. We also confirm that th3/+ ERFE-/- mice do not reverse splenomegaly or improve peripheral blood circulating erythroid parameters compared to th3/+ mice [Kautz Blood 2015] (Table I) but exhibit further increase in TfR1 in late stage erythroid precursors (Fig 1c). Analysis of the bone marrow reveals that total erythroid mass is unaltered in triple mutants relative to th3/+, th3/+ ERFE-/-, and th3/+ TfR1+/- mice, but the number of late erythroblasts (poly-E and ortho-E stages) is normalized to WT levels (Fig 1d), strongly suggesting that, unlike in th3/+ erythropoiesis, where the block in differentiation occurs at the poly-E stage, th3/+ TfR1+/- and especially triple mutant mice restore differentiation at this stage to generate a higher hemoglobin. No differences in erythroblast apoptosis or ROS concentration are evident in triple mutant relative to th3/+ ERFE-/- or th3/+ TfR1+/- mice. We also analyzed markers of Epo responsiveness and demonstrate that serum Epo and EpoR expression are increased in th3/+ relative to WT mice (Fig 1e and 1f), but while serum Epo is decreased, EpoR is further increased (Fig 1f). These findings suggest that Epo responsiveness is more optimized in triple mutant erythroblasts, enabling a smaller proportion of late stage erythroblasts to produce circulating RBCs with relatively less serum Epo. Remarkably, while neither th3/+ ERFE-/- and th3/+ TfR1+/- mice reverse iron overload or impact hepcidin expression at 6 weeks of age, triple mutant mice demonstrate fully normalized ratio of hepcidin expression relative to liver iron concentration (LIC) (Fig 1g). Taken together, these experiments provide evidence of the differential and additive effects of TfR1 and ERFE loss in th3/+ mice, with a predominantly erythropoietic benefit of TfR1 loss, a predominantly iron-homeostatic benefit of ERFE loss, and synergy of both in optimizing Epo responsiveness. Disclosures Ganz: Intrinsic LifeScience: Consultancy, Equity Ownership, Membership on an entity's Board of Directors or advisory committees; Silarus Pharma: Consultancy, Equity Ownership; Keryx Pharma: Consultancy, Research Funding; Gilead: Consultancy; Ablynx: Consultancy; Vifor: Consultancy; Akebia: Consultancy, Research Funding; La Jolla Pharma: Consultancy, Patents & Royalties: Patent licensed to La Jolla Pharma by UCLA.

Description: Sickle cell disease (SCD) is a recessively inherited hemoglobin disorder; the most common and severe form is a consequence of homozygous βS mutation. High concentration of hemoglobin S damages RBC membranes, leading to hemolysis, vaso-occlusion, and inflammation, which together result in anemia, recurrent painful crises, and multiple end-organ damage (i.e. brain, kidney, lung, and bone). Anemia in SCD is multi-factorial. Causes include hemolysis, ineffective erythropoiesis, impaired iron utilization, insufficient erythropoietin responsiveness, and low oxygen affinity [Sherwood Blood 1986]. RBC transfusions are used to ameliorate symptoms and complications but indications are not clearly defined for adults. We postulate that impaired iron utilization in SCD is a consequence of complex regulation of hepcidin, the main hormone regulating circulating and systemic iron. We aim to explore correlation between hepcidin, inflammation and erythropoiesis to enhance our understanding of iron metabolism, hepcidin regulation, and pathophysiology of anemia in SCD. Ultimately this may yield more specific indications for RBC transfusion, and pave the way for novel therapeutic approaches. High hepcidin results in the sequestration of iron within macrophages and prevents further iron absorption in the gastrointestinal tract. Hepcidin expression is enhanced by iron and inflammation and suppressed by hypoxia and erythropoiesis. Distinguishing the relative contribution of these competing effects on hepcidin in SCD is complicated. For example, adult SCD patients in steady state have significantly lower hepcidin relative to heterozygous controls and do not correlate with markers of inflammation and erythropoiesis [Kroot Haematologica 2009]. In a randomized, placebo-controlled trial, inhaled steroids (mometasone) were administered to non-asthmatic SCD patients [Glassberg Am J Hematol 2017]. The results demonstrate improved pain scores and reduced hemolysis in the treated group although the mechanism of this effect is not known. In a subsequent analysis, markers of inflammation are suppressed in the treated group, correlating with macrophage activation. We thus evaluate hepcidin expression in stored serum samples from this trial of SCD patients treated with inhaled steroids or placebo. Samples before and after treatment were analyzed. Serum hepcidin quantification was performed with Hepcidin-25 ELISA kit (Intrinsic Lifesciences LLC). In addition, we evaluate if and how hepcidin levels correlate with pro-inflammatory markers measured with olink inflammation assay. Serum hepcidin concentration is not different in steroid- vs. placebo-treated SCD patients before or after treatment and no significant difference is observed in the hepcidin ratio after:before treatment between the two groups (Fig 1a). However, in the steroid group, hepcidin concentrations are increased in a higher proportion of patients although the difference does not reach statistical significance (Fig 1b). We thus hypothesize that hepcidin levels may be a marker of increased erythropoiesis relative to inflammation in SCD. Both IL-6 (Fig 2a) and IL-10 (Fig 2b), cytokines known to induce hepcidin expression during inflammation, decrease after treatment with steroids relative to placebo, and neither correlates with hepcidin concentration after treatment (Table I). Hepcidin is however inversely correlated with adenosine deaminase (ADA) and hemoglobin S and directly correlative with fibroblast growth factor 23 (FGF23), monocyte chemoattractant protein 1 (MCP1), and hemoglobin F in both placebo- and steroid-treated groups (Table I), potentially suggesting that hepcidin inversely correlates with expanded erythropoiesis. We thus use the newly available human ELISA kit for erythroferrone (ERFE) (Intrinsic Lifesciences LLC), a recently identified erythroid regulator of hepcidin [Kautz Nat Gen 2014], to evaluate correlation with hepcidin. Our results demonstrate that serum ERFE concentration decreases in steroid- (Fig 3a) but not placebo-treated SCD patients (Fig 3b), and no correlation with hepcidin is observed. Taken together, these data for the first time demonstrate a decrease in erythropoiesis and erythroid-regulation-induced hepcidin suppression in inhaled-steroid-treated SCD patients. Disclosures No relevant conflicts of interest to declare.

Description: Erythropoiesis normally occurs in the bone marrow within the pelvis and femur, and both erythropoiesis and bone metabolism are susceptible to changes in iron homeostasis. Thus, hematopoietic and osteoid systems require coordination of iron metabolism during stress or ineffective erythropoiesis. Recently, a more extensive understanding of the crosstalk between iron metabolism and erythropoiesis revealed that a bone marrow secreted protein, erythroferrone (ERFE), is a negative regulator of hepcidin [Kautz Nat Gen 2014]. Hepcidin in turn is the main negative regulator of iron absorption and recycling [Nemeth Science 2004] and its suppression enables an increase in iron availability during stress erythropoiesis. Diseases of ineffective erythropoiesis, such as β-thalassemia, with chronic erythroid expansion, are associated with thinning of cortical bone, leading to decreased bone mineral density [Haidar Bone 2011; Vogiatzi Bone 2006]. Mechanisms underlying coordination of erythropoiesis and bone metabolism are incompletely understood. However, because ERFE functions to suppress hepcidin by sequestering BMPs [Arezes Blood 2018], and because BMPs are crucially important for bone metabolism [Hogan Genes Dev 1996], we hypothesize that ERFE may be involved in coordinating iron metabolism, erythropoiesis, and bone homeostasis. Lastly, osteoblast expression of TfR2 was found to inhibit bone formation by activating BMP-p38MAPK signaling and expression of the Wnt inhibitor Sclerostin, protein product of the SOST gene [Rauner Nat Med 2019]. We thus propose to explore the role of ERFE in disordered bone metabolism in β-thalassemia. In vitro data demonstrates that osteoblasts from wild type (WT) mice express ERFE and this expression is enhanced by BMP2/6/7 (Figure 1a and 1b). Furthermore, osteoblasts from ERFE-/- mice exhibit enhanced bone mineralization (6.8-fold increased von Kossa staining, measured by image J) (Figure 1c), increased expression of osteoblast-specific markers (e.g. osterix (OSX))(Figure 1d), and higher SOST expression (Figure 1e) relative to WT osteoblasts. We anticipate that if TfR2 is central to bone metabolism, ERFE-/- osteoblasts may exhibit a decrease in TfR2; our results demonstrate only a trend toward decreased TfR2 in ERFE-/- osteoblasts (Figure 1f). In addition, we propose that ERFE is a negative regulator of osteoblast activity, predicting that ERFE loss in th3/+ mice would enhance bone mineral density. To this end, we analyzed bone mineral density and histomorphometry in WT, ERFE-/-, th3/+, and th3/+ERFE-/- mice. Surprisingly, although no differences are evident between WT, ERFE-/-, and th3/+ femora, th3/+ERFE-/- mice exhibit a decrease in bone mineral density and bone volume / total volume (BV/TV) (Figure 2a-2b) with a trend toward enhanced femoral mineral apposition rate (Figure 2c) relative to th3/+ mice. These results indicate enhanced osteoblast activity without increased bone formation. Because bone mineralization is a composite of the relative osteoblast and osteoclast activity, we hypothesize that osteoclast activity is further enhanced in th3/+ ERFE-/- mice. TRAP staining demonstrates a significantly increased number of osteoclasts in ERFE-/- relative to WT as well as th3/+ ERFE-/- relative to th3/+ femora (Figure 2d). Our studies demonstrate that ERFE, like other members of the TNFα superfamily [Lu J Bone Miner Res 2011], negatively regulates OSX which is critical for osteoblast function (Figure 3a). Thus, suppression of ERFE results in more OSX (Figure 1d), enhanced mineralization (Figure 1c), and higher SOST expression (Figure 1e) which results in the secretion of Sclerostin (Figure 3b). Sclerostin both feeds back to suppress Wnt signaling to decrease osteoblast function and increases RANKL production to stimulate osteoclast differentiation (Figure 3b). Taken together, ERFE functions as a negative regulator of both osteoblast and especially osteoclast activity such that its loss leads to more osteoclast activity and results in decreased bone mineral density in β-thalassemia. These findings provide novel insights into the complex interplay between regulation of iron metabolism and bone homeostasis in diseases of dysregulated erythropoiesis, when ERFE expression is increased, and support the rationale to further explore the role of ERFE and TfR2 in this crosstalk in β-thalassemia. Disclosures Fleming: Protagonist: Membership on an entity's Board of Directors or advisory committees; Silence Therapeutics: Consultancy; Ultragenyx: Consultancy. Rivella:Disc medicine, Protagonist, LIPC, Meira GTx: Consultancy; Meira GTx, Ionis Pharmaceutical: Membership on an entity's Board of Directors or advisory committees. Ginzburg:La Jolla Pharma: Membership on an entity's Board of Directors or advisory committees.

Description: Erythroferrone (ERFE) is a hormone produced by erythroblasts in response to erythropoietin. ERFE acts as a regulator of hepcidin expression during stress erythropoiesis. By suppressing hepcidin expression, ERFE contributes to the mobilization of dietary and stored iron necessary for recovery from blood loss after hemorrhage. Furthermore, overproduction of ERFE plays a pathogenic role in β-thalassemia and other anemias with ineffective erythropoiesis, where it contributes to hepcidin suppression and consequent iron overload. Development of a method to quantify serum ERFE in mice would improve our ability to study the pathobiology of this erythroid regulator. A dual polyclonal sandwich ELISA was developed to quantify ERFE in mouse serum. Purified recombinant mouse ERFE was used to immunize rabbits and goats, and high titer antibodies were purified from serum via protein A. Western blotting of reduced ERFE protein demonstrated that both the capture and detection antibody specifically recognized mouse ERFE, weakly recognized recombinant human ERFE, but neither antibody recognized mouse, rabbit or human TNF-alpha. Antibody was biotinylated and screened to determine the optimal antibody pairs. ELISA optimization established the standard curve range from 0 to 4 ng/ml. With a 10% sample dilution the lower limit of quantitation (LLOQ) was 0.1 ng/ml. Average spike recovery of ERFE in 3 different mouse sera (0.75 - 24 ng/ml) ranged from 93-105% (mean 99%). Dilutional linearity of the same spiked samples ranged from 93-104% (mean 99%). Intra- and inter-assay precision was 4.9% and 3.9%, respectively, over a concentration range of 0.57 - 16.3ng/ml. The effect of phlebotomy on serum ERFE in 6- and 8-week-old male C57BL/6 mice (n=3 each) was examined 24 h after removal of 0.5 ml of blood (Figure 1). At time 0, all the mice had serum ERFE levels below the limit of detection, but serum ERFE had increased to a mean of 1.9 ng/ml at 24 hours (P=0.03). This increase in ERFE is associated with a decrease in hepcidin from 209 ng/ml to 104 ng/ml (p