ALBERT

Description: Thalassemias are a heterogeneous group of red blood cell disorders ranging from a clinically severe phenotype requiring life-saving transfusions (thalassemia major) to a relatively moderate symptomatic disorder, sometimes requiring transfusions (thalassemia intermedia). Thalassemia minor, the least severe form of the disorder, is characterized by minimal to mild symptoms. Though considered a major cause of morbidity and mortality worldwide, there is still no universally available cure for thalassemia major. The reason for this is, at least in part, due to the lack of full understanding of pathophysiology of thalassemia. The underlying basis of thalassemia pathology is the premature apoptotic destruction of erythroblasts causing ineffective erythropoeisis. Normally, the assembly of adult hemoglobin (consisting of a tetramer of two α- and two β-globin chains) features a very tight coordination of α- and β-globin chain synthesis. However, in β-thalassemia, β-globin synthesis is diminished causing α-globin accumulation; while in α-thalassemia the opposite scenario occurs. Unpaired globin chains that accumulate in thalassemic erythroblasts are bound to heme. In addition, in β-thalassemia an erythroid specific protease destroys excess α-globin chains, likely leading to the generation of a pool of "free" heme in erythroblasts. "Unshielded" heme is toxic, but this toxicity will likely be augmented, if heme oxygenase 1 (HO-1) can release iron from heme. So far, virtually no information about the expression of HO-1 in erythroblasts has been produced. However, we have recently provided unequivocal evidence that this enzyme is present in several model erythroid cells1. Based on this novel and important finding, we hypothesize that in β-thalassemic erythroblasts HO-1-mediated release of iron from heme is the major culprit responsible for cellular damage. To test this hypothesis, we exploited the mouse model of β-thalassemia known as th3/th3. Our data indicates that HO-1 expression is increased in the liver, spleen and kidney of β-thalassemic mice compared to wild-type mice. Importantly, we observed that erythropoietin-mediated erythroid differentiation of fetal liver (FL) cells isolated from β-thalassemic fetuses have increased levels of HO-1 at mRNA and protein levels as well as a decrease in phosphorylated eIF2-α levels. Ferritin levels were increased in β-thalassemic FL cells suggesting increased heme catabolism and iron release. To investigate the contribution of HO-1 to the pathology associated with β-thalassemia, wild-type and thalassemic (th3/+) mice were injected with 40 µmoles/kg/d of tin-protoporphyrin IX (SnPP, HO-1 inhibitor) during a 4-week period, 3 times a week. Our results show that β-thalassemic mice injected with SnPP display a decrease in the spleen index, hemoglobin levels, red blood cell counts, reticulocyte counts and liver iron content when compared to PBS injected β-thalassemic mice. Additionally, HO-1 inhibition reduced ineffective erytropoiesis in β-thalassemia mice. Our results indicate that β-thalassemic erythroblasts have inappropriately high levels of "free" heme that is continuously degraded by HO-1. Further research is needed to determine whether iron liberated from heme by HO-1 is directly responsible for the damage of β-thalassemic erythroblasts. 1Garcia-Santos D, et al. Heme oxygenase 1 is expressed in murine erythroid cells where it controls the level of regulatory heme. Blood 123 (14): 2269-77, 2014. Disclosures No relevant conflicts of interest to declare.

Description: The transferrin receptor (TfR) is a membrane glycoprotein whose only clearly defined function is to mediate cellular uptake of iron (Fe) from a plasma glycoprotein, transferrin. Iron uptake from diferric transferrin (Tf) involves the binding of transferrin to the TfR followed by internalization of Tf within an endocytic vesicle by receptor-mediated endocytosis. Iron is then released from transferrin within endosomes by a combination of Fe3+ reduction by Steap3 (likely when transferrin is still bound to TfR) and a decrease in pH (~pH 5.5). Following this, Fe2+ is transported across the endosomal membrane by DMT1. Transferrin receptors are highly expressed on immature erythroid cells, placental tissue, and rapidly dividing cells, both normal and malignant. In proliferating nonerythroid cells the expression of TfR is negatively regulated post-transcriptionally by intracellular iron through iron responsive elements (IREs) in the 3' untranslated region (UTR) of transferrin receptor mRNA. IREs are recognized by specific cytoplasmic proteins (iron regulatory proteins; IRPs) that, in the absence of iron in the labile pool, bind to the IREs of transferrin receptor mRNA, preventing its degradation. On the other hand, the expansion of the labile iron pool leads to a rapid degradation of transferrin receptor mRNA that is not protected, since IRPs are not bound to it. However, some cells and tissues with specific requirements for iron probably evolved mechanisms that can override the IRE/IRP-dependent control of transferrin receptor expression. We previously documented that the TfR gene promoter contains an erythroid active element that stimulates the receptor gene transcription upon induction of hemoglobin synthesis (1). In this study we have demonstrated that incubation of erythroid cells with 5-aminolevulinic acid (ALA) increased TfR expression as well as iron incorporation into heme. This effect of ALA can be completely prevented by the inhibitors of heme biosynthesis (succinylacetone [blocks ALA dehydratase] or N-methylprotoporphyrin [blocks ferrochelatase]), indicating that the effect of ALA requires its metabolism to heme. The induction of TfR mRNA expression by ALA is primarily a result of increased mRNA synthesis, since the effect of ALA can be abolished by actinomycin D. Moreover, we found that the TfR promoter was activated in vitro by the addition of ALA or hemin to murine erythroleukemia (MEL) cells induced to differentiate using DMSO. Furtehermore, site-directed mutation of erythroid active element (1) in the TfR promoter abolished the effects of ALA or hemin. These results indicate that heme may directly or indirectly interact with the TfR promoter, consequently enhancing the gene expression. Hence, our results show that in erythroid cells heme serves as a positive feedback regulator that maintains high TfR levels thus ensuring adequate iron availability for hemoglobin synthesis. In conclusion, erythroid cells, which are the most avid consumers of iron in the organism, use a transcriptional mechanism to maintain very high transferrin receptor levels. 1 Chun-Nam Lok Ponka P. (2000) Identification of an Erythroid Active Element in the Transferrin Receptor Gene. J. Biol. Chem. 275: 24185-24190. Disclosures No relevant conflicts of interest to declare.

Description: In erythroid cells, more than 90% of transferrin-derived iron enters mitochondria where ferrochelatase inserts Fe2+ into protoporphyrin IX. However, the path of iron from endosomes to mitochondrial ferrochelatase remains elusive. The prevailing opinion is that, after its export from endosomes, the redox-active metal spreads into the cytosol and mysteriously finds its way into mitochondria through passive diffusion. An opposing view is that the highly efficient transport of iron toward ferrochelatase in erythroid cells requires a direct interaction between transferrin-endosomes and mitochondria ("kiss-and-run" hypothesis; Ponka Blood 89:1, 1997). Despite the longevity of the prevailing opinion, experimental evidence (Zhang et al. Blood 105:368, 2005; Sheftel et al. Blood 110: 125, 2007) only supports the latter hypothesis. Using 3D live confocal imaging of reticulocytes following their incubation with MitoTracker Deep Red (MTDR) and Alexa Green Transferrin (AGTf), we have demonstrated transient endosome-mitochondria interactions. We have also documented these interactions by a novel method exploiting flow sub-cytometry to analyze reticulocyte lysates labeled with MTDR and AGTf. We have thusly identified a population of particles labeled with both fluorescent markers, representing endosomes interacting with mitochondria. FACS followed by 2D confocal microscopy confirmed the association of both organelles in the double-labeled population. In the current study, we examined whether reticulocyte mitochondria interact with transferrin (Tf) in a cell-free system. Lysates of reticulocytes previously labeled with MTDR were incubated with AGTf for various time intervals. Examination of lysates by 2D confocal microscopy revealed a time-dependent increase in the number of mitochondria in contact with fluorescent Tf. This can be prevented by the presence of excess, unlabeled Fe2-Tf, but not by albumin (Fig.1). Moreover, the addition of unlabeled Fe2-Tf to reticulocyte lysates removed AGTf from mitochondria, indicating that mitochondria from reticulocyte lysates are associated with TfR that can reversibly bind Tf. In addition, we demonstrate that endosomes containing mutated recombinant holotransferrin, which cannot release iron, remain associated with mitochondria, while endosomes containing mutated recombinant apotransferrin, which cannot bind iron, are not associated with mitochondria. Our findings indicate that endosomes containing holo-Tf promote their attachment to, and drive the detachment of apo-Tf-endosomes from, mitochondria, respectively. By co-immunoprecipitation assay (from murine eryhroleukemia [MEL] cells and reticulocytes lysates), we purified the voltage-dependent anion channel 2 (VDAC2), which is located at the outer membrane of the mitochondrion (Graham, et al. Curr Top Dev Biol. 59: 87, 2004) with DMT1. We confirmed the colocalization of VDAC2 and DMT1 in MEL cells and reticulocytes by both immunofluorescence and confocal microscopy. Moreover, we found a significant decrease in the number of mitochondria in contact with Tf-endosomes after depletion of VDAC2 in MEL cells or after treatment of reticulocyte lysates with the mitochondrial uncoupler CCCP, further supporting the concept of a physical interaction between endosomes and mitochondria. To examine a possible role of DMT1-VDAC2 interactions in iron trafficking, we depleted MEL cells of VDAC2 or inhibited VADC2 using erastin (a specific VDAC2 inhibitor that alters its gating) followed by the measurement of 59Fe incorporation from 59Fe-Tf into heme. Our finding of decreased 59Fe incorporation into heme of MEL cells with silenced or inhibited VDAC2 supports the idea that this outer-membrane mitochondrial protein is involved in the interaction of endosomes with mitochondria. We are currently continuing to delineate the molecular mechanisms involved in endosome-mitochondria interactions focusing on the "signal(s)" that direct iron-carrying endosomes towards mitochondria, the players involved in the docking of endosomes to mitochondria and the "signal(s)" that determine the detachment of iron-free endosomes from mitochondria. Figure 1. Figure 1. Disclosures No relevant conflicts of interest to declare.

Description: Thalassemias are a heterogeneous group of red blood cell (RBC) disorders ranging from a clinically severe phenotype requiring lifesaving transfusions (thalassemia major) to a relatively moderate symptomatic disorder, sometimes requiring transfusions (thalassemia intermedia). Though considered a major cause of morbidity and mortality worldwide, there is still no universally available cure for thalassemia major. The reason for this is, at least in part, due to the lack of full understanding of pathophysiology of thalassemia. The underlying basis of thalassemia pathology is the premature apoptotic destruction of erythroblasts causing ineffective erythropoeisis. In β-thalassemia, β-globin synthesis is diminished causing α-globin accumulation. Unpaired globin chains that accumulate in thalassemic erythroblasts are bound to heme. Moreover, in β-thalassemia an erythroid-specific protease destroys excess α-globin chains, likely leading to the generation of a pool of "free" heme in erythroblasts. Physiologically, heme can be degraded only via heme oxygenases (HO). Circulating erythrocytes contain the majority of heme destined for catabolism; this process takes place primarily in splenic and hepatic macrophages following erythrophagocytosis of senescent RBC. Heme oxygenase, in particular its heme-inducible isoform HO1, has been extensively studied in hepatocytes and many other non-erythroid cells. Recently, we have provided unequivocal evidence that this enzyme is present in erythroid progenitors as well as their differentiated progenies.1 "Unshielded" heme is toxic, but this toxicity will likely be augmented, if HO1 releases iron from heme. We hypothesize that in β-thalassemic erythroblasts HO1-mediated release of iron from heme is the major culprit responsible for cellular damage. Additionally, it has been shown that prevention of heme-derived iron release from splenic and hepatic macrophages improves β-thalassemia phenotype2. Therefore, suppression of HO1-mediated heme catabolism from senescent RBC could be beneficial in reversing thalassemic phenotype. To test this hypothesis, we exploited the mouse model of β-thalassemia known as th3/th3; we obtained these mice from Dr. Stefano Rivella. Our data indicates that HO1 expression is increased in the liver of β-thalassemic mice as compared to wild type mice. Importantly, we observed that erythropoietin-mediated erythroid differentiation of fetal liver (FL) cells from β-thalassemic fetuses increased HO1 mRNA and protein levels to a higher degree than in their wild type counterparts. Ferritin levels were increased in β-thalassemic FL cells suggesting increased heme catabolism and iron release from the tetrapyrrole macrocycle. To investigate the contribution of HO1 to the pathology associated with β-thalassemia, wild type and thalassemic (th3/+) mice were injected intraperitoneally with 40 µmoles/kg/d of tin-protoporphyrin IX (SnPP, HO inhibitor) during a 4-weeks, 3-times a week. Our results show that β-thalassemic mice injected with SnPP have increased hemoglobin levels and red blood cell counts, and display a decrease in the spleen index, reticulocyte counts and liver iron content when compared to PBS-injected β-thalassemic mice. Furthermore, while hepcidin levels remain unchanged, liver ferroportin expression decreases in SnPP-injected β-thalassemic mice. Our results indicate that β-thalassemic erythroblasts have high levels of HO1, which would be expected to degrade any "free" heme. Further research is needed to determine whether iron liberated from heme by HO1 is directly responsible for the damage of β-thalassemic erythroblasts. 1GarciaSantos D, et al. Heme oxygenase 1 is expressed in murine erythroid cells where it controls the level of regulatory heme. Blood 123 (14): 226977, 2014. 2Nai A, et al. Deletion of TMPRSS6 attenuates the phenotype in a mouse model of β-thalassemia. Blood 119 (21): 5021, 2012. Disclosures No relevant conflicts of interest to declare.

Description: Normal hemoglobinization of immature red blood cells (RBC) requires iron (Fe) uptake from transferrin (Tf), mediated by Tf receptors (TfR). Following the binding of Fe(III)2-Tf to TfR on the erythroid cell membrane, the Tf-TfR complexes are internalized via endocytosis, following which Fe is released from Tf by a process involving endosomal acidification and reduction by Steap3. Fe2+ is then transported across the endosomal membrane by the divalent metal transporter 1 (DMT1). Unfortunately, the post-endosomal path of Fe within cells remains elusive or is, at best, controversial. It has been commonly accepted that a low molecular weight intermediate chaperones Fe in transit from endosomes to mitochondria and other sites of utilization; however, this much sought Fe binding intermediate has never been identified. In erythroid cells, more than 90% of Fe has to enter mitochondria where ferrochelatase, the final enzyme in the heme biosynthetic pathway that inserts Fe2+ into protoporphyrin IX, resides. Indeed, strong evidence exists for specific targeting of Fe toward mitochondria in developing red blood cells in which Fe acquired from Tf continues to flow into mitochondria even when the synthesis of protoporphyrin IX is suppressed. Thus, it has been hypothesized (Ponka P. Blood 89:1, 1997) that, in hemoglobin-producing cells, there is a direct relaying of iron from the endosomal machinery to that of the mitochondria. Numerous reports from our laboratory support this hypothesis: 1) Iron acquired from Tf accumulates in mitochondria even when the synthesis of protoporphyrin IX is inhibited (Richardson et al. Blood 87:3477,1996); 2) Endosome mobility is essential for the efficient incorporation of 59Fe from 59Fe-Tf-labeled endosomes into heme (Zhang et al. Blood 105:368, 2005) and, 3) Confocal laser microscopy shows that in reticulocytes, endosomes continuously traverse the cytosol and touch mitochondria (Sheftel et al. Blood 110: 125, 2007). Based on this, we propose that erythroid precursors have special adaptations that facilitate the high rate of iron transport from endosomes to mitochondria to meet the exceptionally high demand for heme synthesis. Our lab has previously shown, using 3D live confocal imaging, that the iron delivery pathway in developing RBC involves a transient interaction of endosomes with mitochondria. To further demonstrate the interaction of these organelles, we used a novel method based on flow cytometry analyses (flow sub-cytometry) of lysates obtained from reticulocytes with fluorescently labeled endosomes (Alexa Green Transferrin) and mitochondria (MitoTracker Deep Red). Using this strategy, we have identified three distinct populations: endosomes, mitochondria, and a population double-labeled with both fluorescent markers representing endosomes interacting with mitochondria. This strategy has been used in studies on reticulocytes and erythroblasts subjected to various experimental conditions. In this study, we intended to identify molecular partners involved in the endosme-mitochondria interaction. Using co-immunoprecipitation and pull-down strategies, we attempted to recognize proteins interacting with the extra-endosomal (intracellular) loops of DMT1, which may be involved in interactions with mitochondria. The co-immunoprecipitated proteins were separated based on their molecular weights, stained using Coomassie and/or Silver gel and identified by mass spectrometry and western blotting. Using these strategies, we co-immunoprecipitated (from MEL cells and reticulocytes lysates) proteins that were pulled down with DMT1. Using this approach, we have identified the voltage-dependent anion channel (VDAC), which is located at the outer membrane of the mitochondria (Graham, et al. Curr Top Dev Biol. 59: 87, 2004) as one of DMT1 interacting partners using western blotting and specific antibodies against VDAC. These results indicate the physical contact between endosomes and mitochondria. In addition, to define the possible role of DMT1-VDAC interactions in mediating iron uptake, we used a siRNA approach to silence VDAC expression in MEL cells and then measured 59Fe incorporation into heme. These studies revealed decreased 59Fe incorporation into MEL cells with silenced VDAC. Our findings provide a strong support for the hypothesis that this outer-membrane mitochondrial protein is involved in the interaction with endosomes. Disclosures No relevant conflicts of interest to declare.

Description: Delivery of iron (Fe) to most cells occurs following the binding of diferric transferrin (Tf) to its cognate receptors on the cell membrane following which the Tf-receptor complexes are internalized via endocytosis. Iron is then released from Tf within endosomes by a combination of Fe3+ reduction by Steap3 and a decrease in pH (~ pH 5.5). Subsequently, Fe2+ is transported through the endosomal membrane by DMT1. In erythroid cells, more than 90% of Fe has to enter mitochondria where ferrochelatase, the final enzyme in the heme biosynthetic pathway that inserts Fe2+ into protoporphyrin IX, resides. The intracellular path of iron from endosomes to ferrochelatase is still obscure or, at best, controversial. The prevailing opinion is that Fe, after its export from endosomes, spreads into the cytosol, from where the metal mysteriously finds its way into mitochondria. An opposing view is that the highly efficient transport of Fe toward ferrochelatase in erythroid cells requires a direct interaction between transferrin-endosomes and mitochondria ("kiss-and-run" hypothesis;Ponka Blood 89:1, 1997). Despite the longevity of the prevailing opinion, experimental evidence (Richardson et al. Blood 87:3477, 1996; Zhang et al. Blood 105:368, 2005; Sheftel et al. Blood 110: 125, 2007) only supports the latter hypothesis, which sees favorable reception among Cell Biologists (McBride BMC Biology 13:8, 2015). Our laboratory has demonstrated, using both 2D and 3D live confocal imaging, that the intracellular Fe pathway in erythroid cells indeed involves a transient interaction of endosomes with mitochondria. To furtherdemonstrate the contact between these organelles, we have developed a novel method based on flow cytometry analysis ("flow sub-cytometry") of lysates obtained from reticulocytes with fluorescently labeled mitochondria (MitoTracker Deep Red; MTDR) and endosomes (Alexa Green Transferrin; AGTf). Using this strategy, we have identified three distinct populations: endosomes, mitochondria, and a population double-labeled with both fluorescent markers representing endosomes interacting with mitochondria. The size of the double-labeled population increases with the incubation time and plateaus in approximately 20 min. In this study, we examined whether reticulocyte mitochondria interact with Tf in a cell-free system. Lysates obtained by freeze-thawing of reticulocytes previously labeled with MTDR were incubated with AGTf for various time intervals. Examination of lysates by 2D confocal microscopy has revealed a time-dependent increase in the number of mitochondria being in contact with Tf-endosomes (fig 1: Images of mitochondria and endosomes; 20 min incubation with AGTf). This can be prevented by Fe2-Tf, but not by albumin, added to lysates. Moreover, the addition of unlabeled Fe2-Tf to reticulocyte lysates removed AGTf from mitochondria. We conclude that mitochondria from freeze-thawed reticulocyte lysates are associated with TfR that can reversibly bind Tf. We have also embarked on uncovering molecular partners involved in the endosome-mitochondria interactions. Using co-immunoprecipitation and pull-down strategies, we have attempted to detect proteins interacting with the intracellular loops of DMT1 that could be candidates for interactions with mitochondria. The co-immunoprecipitated proteins were separated based on their molecular weights, stained using Coomassie and/or Silver gel and identified by mass spectrometry followed by western blotting. We co-immunoprecipitated (from murine eryhroleukemia [MEL] cells and reticulocytes lysates) proteins that were pulled down with DMT1. One of the proteins that we have recognized is the voltage-dependent anion channel (VDAC), which is located at the outer membrane of the mitochondrion (Graham, et al. Curr Top Dev Biol. 59: 87, 2004). The identity of DMT1 was confirmed by western blotting using specific antibodies against VDAC. These results further support the concept of the physical interaction between endosomes and mitochondria. To examine a possible role of DMT1-VDAC interactions in iron trafficking, we silenced the expression of VDAC in MEL cells followed by the measurement of 59Fe incorporation from 59Fe-Tf into heme. Our finding of decreased 59Fe incorporation into heme of MEL cells with silenced VDAC supports the idea that this outer-membrane mitochondrial protein is involved in the interaction with endosomes. Figure 1. Figure 1. Disclosures No relevant conflicts of interest to declare.

Description: Key Points Iron released from heme by HO 1 contributes to the pathophysiology of thalassemia. Inhibition of HO 1 is of therapeutic value for the treatment of thalassemia.