REVIEW URRENT C OPINION
Macrophages and regulation of erythropoiesis Rebecca N. Jacobsen a,b, Andrew C. Perkins a,c,d, and Jean-Pierre Levesque a,b,c
Purpose of review The nature and function of macrophages at the center of erythroblastic islands is not fully understood. This review discusses novel findings on the phenotypic and molecular characterization of erythroblastic island macrophages, and their role in regulating normal and pathological erythropoiesis. Recent findings The phenotype to prospectively isolate erythroblastic island macrophages from mouse bone marrow has been identified. In-vivo depletion of erythroblastic island macrophages causes blockade of erythroblast maturation and delays erythropoietic recovery following chemical insults. The cytokine granulocyte colonystimulating factor arrests medullary erythropoiesis by depleting erythroblastic island macrophages from the bone marrow. In-vivo ablation of macrophages improves anemia associated with b-thalassemia and reduces red blood cell counts in the mouse model of polycythemia vera. The role of cell adhesion molecules regulating interactions between erythroblastic island macrophages and erythroblasts has been clarified, and mechanisms of pyrenocyte engulfment by erythroblastic island macrophages have been demonstrated to involve Mer tyrosine kinase receptor. Summary Prospective isolation of mouse erythroblastic island macrophages together with new genetic mouse models to specifically target erythroblastic island macrophages will enable molecular studies to better define their role in controlling erythroblast maturation. These studies have revealed the key role of erythroblastic island macrophages in regulating normal erythropoiesis and could be interesting targets to treat b-thalassemia or polycythemia vera. Keywords anemia, erythroblastic island, erythropoiesis, macrophage, red blood cell
INTRODUCTION Macrophages have two major roles in regulating red blood cell homeostasis, controlling the final maturation of erythroblasts into reticulocytes and clearing old erythrocytes. In 1958 Marcel Bessis first described erythroblastic islands, consisting of clusters of erythroid cells at various stages of differentiation surrounding a central macrophage, in bone marrow [1]. Similar erythroblastic islands were observed in the adult mouse bone marrow, spleen and fetal liver [2]. From these observations, it was proposed that erythroblastic island macrophages contribute to the final stages of erythroblast differentiation and maturation. In particular, they have been proposed to secrete cytokines essential to erythroblast survival and maintenance [3,4], transport iron to erythroblasts [5,6] to support the synthesis of large quantities of hemoglobin, and phagocytose and degrade erythroblast nuclei during nuclear extrusion, a step necessary to generate enucleated reticulocytes [7,8]. Until the www.co-hematology.com
mid-2000s, most studies were based on microscopy observations or in-vitro studies in culture as the phenotypic identification of erythroblastic island macrophages has been elusive. However in 2006, the gene encoding erythroblast macrophage protein (EMP also called macrophage erythroblast attacher or Maea), a protein expressed by both erythroblasts and erythroblastic island macrophages and promoting their heterotypic adhesion, was identified and a
Blood and Bone Diseases Program, Mater Research Institute – University of Queensland, Woolloongabba, bSchool of Medicine, cSchool School of Biomedical Sciences, University of Queensland, Saint Lucia and dDepartment of Haematology, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia Correspondence to Jean-Pierre Le´vesque, Associate Professor, PhD, Blood and Bone Diseases Program, Mater Research Institute – University of Queensland, TRI Building, 37 Kent Street, Woolloongabba, QLD 4102, Australia. Tel: +61 7 3443 7571; e-mail: [email protected]. edu.au Curr Opin Hematol 2015, 22:212–219 DOI:10.1097/MOH.0000000000000131 Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
KEY POINTS The phenotype of mouse erythroblastic island macrophages has been identified enabling their prospective isolation for molecular studies. In-vivo depletion of erythroblastic island macrophages stops erythropoiesis with loss of erythroblasts. Erythroblastic island macrophages are necessary to erythropoietic rescue following myeloablation or phenylhydrazine red cell poisoning. Elimination of erythroblastic island macrophages can normalize erythrocyte counts in polycythemia vera or b-thalassemia. New mechanisms by which erythroblastic island macrophages bind to erythroblasts and eliminate nucleus during erythroblast maturation are better understood.
differentiation (CD)169 or sialic acid binding immunoglobulin-like lectin1 (Siglec-1) was recognized as being present on the plasma membrane of erythroblastic island macrophages at the interface with erythroblast plasma membrane [9]. Other cell adhesion molecules such as EMP [10,11], vascular cell adhesion molecule-1 (VCAM-1 or CD106) [12], and aV integrins [13] were found to be present on erythroblastic island macrophages and play an important functional role in promoting adhesion of erythroblasts. Likewise, the mouse antigen recognized by the rat monoclonal antibody ER-HR3 was also found to be expressed on mouse erythroblastic island macrophages [14,15]. However, taken in isolation, none of these cell surface proteins is exclusively expressed on erythroblastic island macrophages precluding their prospective isolation for cellular and molecular studies. In a recent study [16 ], we combined antibodies specific for CD169, VCAM-1, and ER-HR3 which are all expressed by erythroblastic island macrophages together with myeloid specific antigens such as CD11b (aM integrin), F4/80 (antigen specific of mouse macrophages), and Ly-6G, an antigen mostly expressed by granulocytes but also subsets of macrophages. We identified by flow cytometry in the adult mouse bone marrow and spleen macrophages that were positive for all these cell surface proteins [i.e. CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ lymphocyte antigen (Ly)-6Gþ]. This subset of macrophages represented 1.3 0.2% of bone marrow leukocytes and 27 6% of macrophages identified in cell aggregates that were positive for both the erythroid marker Ter119 and the myeloid marker CD11b. Furthermore, we showed that daily administration of granulocyte colony-stimulating factor (G-CSF) selectively depleted this macrophage subset by 90% in the bone marrow and concomitantly blocked medullary erythropoiesis with a 90–50% reduction in the number of all erythroblast subsets and reticulocytes depending on the antibody panel used [16 ]. To further prove that CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ Ly-6Gþ macrophages are bona fide erythroblastic island macrophages, we selectively depleted CD169þ macrophages in Siglec1DTR/þ mice in which a simian diphtheria toxin receptor is knocked in the CD169 gene to specifically sensitize CD169þ macrophages to diphtheria toxin-mediated killing. Mice do not express diphtheria toxin receptors and are consequently resistant to this potent protein toxin. We also depleted phagocytes in vivo by intravenous injection of clodronate-loaded liposomes. In both these models, CD169þ macrophage depletion or phagocyte depletion caused loss of erythroblasts and reticulocytes in bone marrow and spleen &
knocked-out in the mouse. Following deletion of the EMP (Maea) gene, pups died perinatally. Embryos from E12.5 days of gestation and beyond were anemic with accumulation of nucleated erythrocytes in the blood and absence of erythroblastic island in the fetal liver [8]. Furthermore, EMP expressed on macrophages was necessary to establish erythroblastic island in vitro whereas EMP on erythroblasts was dispensable [8]. These data proved that erythroblastic island macrophages are necessary for terminal erythroblast maturation and enucleation. However, the absence of the precise phenotypic profile of erythroblastic island macrophages has hampered progress in understanding the mechanism by which erythroblastic island macrophages control erythroblast maturation in normal settings and how they contribute to diseased states, such as anemia resulting from myeloid leukemia and chronic inflammation, myeloproliferative diseases with expansion of the erythroid compartment such as polycythemia vera, and genetic erythropathies such as b-thalassemia. In the past 2 years, significant progress has been made in the phenotypic identification of erythroblastic island macrophages in mice and humans with a better understanding of their function in steady-state erythropoiesis or in disease states.
PHENOTYPIC CHARACTERIZATION OF ERYTHROBLASTIC ISLAND MACROPHAGES In the 1990s, many antigens expressed by erythroblastic island macrophages were identified particularly in the mouse. For instance, cluster of
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
213
Erythroid system and its diseases
concomitant with loss of CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ Ly-6Gþ macrophages in both tissues [16 ]. Interestingly, in these three models (G-CSF injection, CD169þ macrophage depletion, and phagocyte depletion), loss of erythroblasts was accompanied by accumulation of proerythroblasts suggesting a failure of maturation [16 ]. Of note however, unlike CD169þ macrophage or phagocyte depletion which affects all tissues in the body, thus both medullar and splenic erythropoiesis, G-CSF increased splenic erythropoiesis, erythroblastic island macrophages and cell aggregates containing erythroid cells and erythroblastic island macrophages despite the fact that both medullar and splenic erythroblastic island macrophages express the G-CSF receptor mRNA [16 ]. This suggests that the response of erythroblastic island macrophages to cytokines such as G-CSF varies depending on the tissue of residence. &
&
&
ROLE OF ERYTHROBLASTIC ISLAND MACROPHAGES IN ERYTHROPATHIES AND POLYCYTHEMIA VERA
&&
The effect of macrophage depletion on erythropoiesis was also studied by Paul Frenette’s and Stefano Rivella’s laboratories, with an emphasis on the role macrophages play in stress erythropoiesis and erythropathies such as polycythemia vera and b-thalassemia. Frenette’s group utilized both the Siglec1DTR/þ mouse model and phagocyte depletion using clodronate-loaded liposomes to study the effect of macrophage depletion in steady-state, as well as recovery from hemolytic anemia induced by hemoglobin-oxidizing phenylhydrazine (PHZ), phlebotomy, myeloablation, and also in a mouse model of polycythemia vera, whereas Rivella’s group only used the liposome-mediated phagocyte depletion to study similar conditions as well as b-thalassemia. Both groups observed a loss of developing erythroblasts from mouse bone marrow when phagocytes or CD169þ macrophages were depleted [17 ,18 ]. Also, prolonged administration of clodronateloaded liposomes (over 12 weeks) induced an anemia with characteristics of iron deficiency anemia, which was however not corrected by iron supplementation [18 ]. Therefore, erythroblastic island macrophages play an important role in erythropoietic recovery independently of iron transport and recycling. Both groups observed that phagocyte depletion with clodronate liposomes or depletion of CD169þ macrophages delayed erythropoietic recovery following PHZ insult [17 ,18 ]. Likewise, erythropoietic recovery following myeloablation and bone marrow transplantation was delayed when CD169þ macrophages were ablated following &&
&&
&&
214
www.co-hematology.com
&&
&&
transplant [17 ]. Taken together these data show that erythroblastic island macrophages are essential to support erythroblast maturation in steady-state as well as during erythropoietic rescue. Polycythemia vera is a myeloproliferative neoplasm driven by a V617F point mutation that constitutively activates the tyrosine kinase Janus kinase 2 (JAK2) which is downstream of many cytokine receptors including the erythropoietin receptor. Mice with the JAK2V617F mutation knocked into the endogenous mouse Jak2 gene [19] provides a robust mouse model of polycythemia vera with characteristics comparable to the human condition. Clodronate-loaded liposome treatment of mice with either early or well established polycythemia vera led to the normalization of hematocrits and red blood cell counts, decreased erythropoietic activity (reduced reticulocytosis), extramedullary hematopoiesis, and decreased splenomegaly. Polycythemia vera symptoms were reduced for up to 4 weeks following cessation of the liposome treatment and only a single liposome injection was required at this time point to return blood parameters to normal [18 ]. Therefore, phagocytes which comprise erythroblastic island macrophages play an important role in maturing erythroblasts carrying the malignant JAK2V617F mutation and could be a potential therapeutic target for this neoplasm. b-Thalassemia is characterized by increased but ineffective erythropoiesis caused by a mutation in the HBB gene, which reduces b-hemoglobin chain synthesis and failure to assemble functional hemoglobin. Hbbth3/þ mice, in which a mutation has been introduced in one allele of the Hbb gene, recapitulate the human b-thalassemia intermedia phenotype. Phagocyte depletion in these mice improved anemia and hemoglobin synthesis. Importantly, long-term administration of clodronate liposomes improved the symptoms of anemia and splenomegaly, possibly by increasing the lifespan of the red blood cells [18 ]. This suggests that macrophages can exacerbate anemia observed in b-thalassemia by accelerating degradation of defective erythrocytes.
&&
&&
CRITICAL MOLECULES WHICH PARTICIPATE IN THE ERYTHROBLASTIC ISLAND INTERACTION Erythroid progenitor cells are not simply passive players within the erythroblastic island. They express cell surface proteins which make numerous productive and essential interactions with the central macrophage. For example, erythroid progenitor cells express ICAM4 which interacts with aV integrins on the erythroblastic island macrophage, a4-integrin and b1-integin which together form Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
the very late activation antigen (VLA)-4 receptor which interacts with VCAM-1 on erythroblastic island macrophages, and EMP expressed by both erythroblasts and erythroblastic island macrophages, which forms homotypic adhesive interactions between these two cell types. There are other less well studied molecules which participate (see Table 1 and Fig. 1). Of the cell adhesion molecules that mediate adhesion between erythroblasts and erythroblastic island macrophages, EMP seems to be the most critical in steady-state. Indeed, it is the only gene encoding a cell adhesion molecule whose deletion causes embryonic anemia and defective erythroblastic island formation during development [8]. ICAM4 knockout (KO) mice have no anemia in steady-state, however the formation of erythroblastic island in vivo and in vitro is reduced [13]. ICAM4 is expressed by erythroblasts. ICAM4 counter receptors on erythroblastic island macrophages are thought to include av integrins (CD51) as peptides blocking ICAM4–aV interaction reduce the formation of erythroblastic islands in vitro [13]. Deletion of the av integrin gene is embryonic lethal between E10 and E12 so its role in establishing erythroblastic island in the fetal liver is not known [20]. Conditional deletion of the av integrin gene specifically in macrophages would answer this question. As integrins are expressed as ab heterodimers, the identity of the av partner integrin on erythroblastic island macrophages has been of interest. A candidate partner is b3-integrin, thus forming an aVb3 dimer. In support of this hypothesis is the presence of mild anemia (lower hematocrit with enhanced number of reticulocytes and CD71þ erythroblasts in the blood) and mild erythroblastic island defects (lower number of erythroblasts per erythroblastic island) in b3/ mice [21]. Table 1. List of molecular interactions between the erythroblastic island macrophages and erythroid cells Macrophage
Erythroblast
VCAM-1
a4b1 integrin complexed with CD81, CD82, and CD151
avb3 integrin
ICAM4
EMP
EMP
MerTK
Phosphatidyl serine complexed with protein S
Axl
Unknown
CD169
Various sialylated proteins
CD163
Unknown
Axl, Axl tyrosine kinase receptor; CD, cluster of differentiation; EMP, erythroblast macrophage protein; ICAM4, intercellular adhesion molecule-4; MerTK, Mer tyrosine kinase receptor; VCAM-1, vascular cell adhesion molecule-1.
Recent evidence suggests a4b1 integrin exists in a large transmembrane complex which includes tetraspanin proteins CD81, CD82, and CD151, and these influence the functional interactions with VCAM-1 on erythroblastic island macrophages [22]. Conditional deletion of the Vcam1 gene, a4 integrin (Itga4) gene, or b1 integrin (Itgb1) gene in hematopoietic cells (germinal knock-out for these genes are embryonic lethal) results in increased trafficking and spontaneous mobilization of hematopoietic stem and progenitor cells in the blood [23,24]. In these three conditional mutants, basal erythropoiesis is normal however erythropoietic recovery following chemical insult with phenyl hydrazine is severely compromised in b1D/D mice with reduced numbers of erythroblasts in spleen and bone marrow. a4D/D and VCAM1D/D had less severe impairment of erythropoiesis rescue following phenyl hydrazine challenge [25]. This effect is autonomous to erythroid cells as conditional deletion of the a4 integrin gene in erythroid cells using EporCre mice led to similar impairment of rescue erythropoiesis following challenge with PHZ [26 ]. These mice also had abnormally elevated number of circulating erythroblasts [26 ]. In humans, blocking antibodies against a4-integrin/VLA-4 result in transient anemia and increased circulating nucleated erythroblasts suggesting they interfere with adhesion between erythroid progenitor cells and the erythroblastic island macrophages in vivo [27 ]. These antibodies also release hematopoietic stem and progenitor cells into the circulation [28], so VLA-4 is also important for hematopoietic stem cell (HSC) interactions with other stromal macrophages in the stem cell niche. Indeed, the erythroblastic island macrophage is probably just one of many specialized types of niche macrophage. Interestingly, antibodies (such as natalizumab) and small molecule inhibitors of VLA-4 are increasingly being used in the clinic to treat various disorders such as multiple sclerosis. Thus, hematologists are seeing referrals from concerned neurologists to exclude more sinister causes of leukoerythroblastic blood films, which are simply because of a transient mobilization of immature blood cells including nucleated erythroid cells usually retained by niche macrophages (including erythroblastic island macrophages) via the a4b1–VCAM1 interaction. &
&
&
EVIDENCE FOR A ROLE OF ERYTHROBLASTIC ISLAND MACROPHAGES IN HUMANS Although erythroblastic islands have been visualized in human bone marrow, the characteristics of the erythroblastic island macrophage and its
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
215
Erythroid system and its diseases
Reticulocyte Erythroblast
KLF-1
α4β1
EMP ?CD169 ?CD163
αvβ3
Pyrenocyte
EMP
ICAM4 VCAM1
Phosphatidyl serine Axl ?
Protein S MerTK
F4/80 Ly6G
Erythroblastic island macrophage
ER-HR3 KLF-1????
DNAse IIa
FIGURE 1. Model of interactions between the erythroblastic island macrophages and maturing erythroblasts. Erythroblastic island macrophages develop adhesive interactions with maturing erythroblasts using erythroblast macrophage protein (EMP), a3b1 integrin, vascular cell adhesion molecule-1 (VCAM-1), and possibly cluster of differentiation (CD) 169 and CD163 and their cognate counter receptors on erythroblasts. Kruppel like factor 1 (KLF1) in erythroblasts induces expression of intercellular adhesion molecule-4 (ICAM4). During the enucleation of erythroblasts, phosphatidyl serine flips from the inner to the outer layer of the plasma membrane, and binds to Mer tyrosine kinase receptor via bridging with protein S. This triggers an engulfment signal of the pyrenocyte by the erythroblastic island macrophage and subsequent degradation of the pyrenocyte nucleus by DNase IIa. The enucleated reticulocyte is released in the circulation.
relationship to the mouse erythroblastic island macrophage have not been fully explored. Ex-vivo cultures of erythroblasts isolated from human bone marrow in steady-state or from polycythemia vera or b-thalassemia patients showed increased proliferation and differentiation when cultured in the presence of macrophages isolated from the same individuals. This was not observed in cocultures wherein erythroblasts were separated from erythroblastic island macrophages by a semipermeable membrane [18 ]. Thus, direct contact between erythroblasts and erythroblastic island macrophages in humans is necessary for erythroblast differentiation and proliferation [18 ]. Further to this, Frenette’s group examined bone marrow aspirates from healthy donors and assessed the expression of CD169 and VCAM-1 on monocytes and macrophages. These two antigens were not expressed by CD15þCD14 granulocytes, or CD15þCD14þ monocytes. However, cells expressing both CD169 and VCAM-1 were present exclusively within the CD15CD163þ macrophage population [17 ] indicating that similar to mice, CD169 and VCAM-1 are both expressed on erythroblastic island macrophages in human bone marrow. Interestingly, following HSC transplantation in humans EMP protein and mRNA expression in mononuclear cells
from blood and bone marrow were dramatically upregulated with increased density of erythroblastic islands in the bone marrow [29]. In severe anemia patients, erythroblastic islands and EMP were barely detectable however erythropoietin treatment boosted both erythroblastic island density and EMP mRNA expression in these patients [29]. This suggests that EMP may also play an important role in establishing human erythroblastic islands.
&&
&&
&&
216
www.co-hematology.com
A ROLE FOR THE MACROPHAGE IN ENGULFMENT AND DIGESTION OF THE PYKNOTIC ERYTHROID NUCLEUS The central macrophage plays a critical role in the final steps of erythroid cell maturation. The final division of maturing erythroid cells is an asynchronous one; the pyknotic nucleus and other organelles are deposited into one daughter cell, called a pyrenocyte. The other daughter cell, a reticulocyte, is released from the erythroblastic island macrophage into the circulation. Asynchronous deposition of cell surface receptors (such as a4b1 integrin), membrane lipids such phosphatidyl serine, and cytoskeletal components is critically important for the retention of one daughter cell by the erythroblastic Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
island macrophage and the release of the other. The erythroblastic island macrophage expresses proteins specifically designed to undertake the tasks of adhesion, engulfment, and degradation of the pyrenocyte. There are likely to be many erythroblastic island macrophage specific as well as general macrophage proteins which coordinate these processes. This is a remarkably efficient process as a sibling pyrenocyte is produced for every one of the million reticulocytes produced every second. EMP is the key to this process as EMP/ mice have defective erythroblastic islands and erythroblast enucleation [8]. Tropomodulin3 (Tmod3) is a recently described actin-interacting protein that caps the pointed ends of actin fibers regulating their length and stability. Tmod3/ embryos die between E14.5 and E18.5 of anemia with defective erythroblastic island formation, reduced numbers of BFU-Es, CFU-Es, and erythroblasts in the fetal liver [30 ]. Interestingly, Tmod3 is necessary in both erythroblasts and erythroblastic island macrophages for efficient erythroblastic island formation [30 ]. Mer tyrosine kinase receptor is an example of a critical erythroblastic island macrophage receptor that is required for the engulfment of pyrenocytes via recognition of protein S bound to phosphatidyl serine on the cell membrane [31 ,32]. However, other related transmembrane tyrosine kinases such as Axl tyrosine kinase receptor may play a similar role as mice double KO for Mer tyrosine kinase receptor and Axl have more pronounced medullar and splenic anemia in steady-state than single KO for each gene [33] (Fig. 1). &&
&&
&
THE ROLE OF KRUPPEL LIKE FACTOR 1 IN COORDINATING ERYTHROBLASTIC ISLAND FUNCTION The transcription factor Kruppel like factor 1 (KLF1), previously called EKLF, is essential for definitive erythropoiesis in mice [34] and man (G. Magor et al., in preparation) and is a key regulator of many erythroid genes including those involved in hemoglobin production, nuclear compaction, enucleation, and iron metabolism [35,36]. A key feature of KLF1-deficient erythropoiesis in both species is very high numbers of circulating nucleated erythrocytes which is suggestive of a specific problem with enucleation. Klf1/ mouse and human cells have dramatically reduced levels of ICAM4 [37], and KLF1 protein directly binds the Icam4 gene promoter [35,36]. Thus, KLF1 directly regulates genes, which facilitates erythroid cell attachment to the erythroblastic island macrophage. Recent evidence suggests KLF1 might also play a surprising role intrinsically within erythroblastic
island macrophages. This is controversial as KLF1 was considered to be an erythroid cell specific transcription factor for decades. Recently, KLF1 protein or its DNA-binding activity has been reported within erythroblastic island macrophages [38]. However, it was not clear whether the KLF1 immunofluorescence signal was coming from the macrophage or ingested pyrenocytes. Common myeloid progenitors give rise to bipotent granulocyte monocyte progenitors and megakaryocyte erythroid progenitors. Macrophages derive from granulocyte monocyte progenitors and erythroid cells from megakaryocyte erythroid progenitors. The traditional dogma is that these two lineages are distinct and do not share an immediate common progenitor cell. There are few reports of hematopoietic colonies composed solely of macrophages and erythroid cells. A recent article by Bieker’s group has challenged these views about the ontogeny of macrophages versus erythrocytes, and also the restricted expression of KLF1 to erythroid cells. The authors introduced a single copy transgene construct, in which green fluorescent protein (GFP) reporter is driven by 950 bp of the Klf1 promoter, into embryonic stem cells [39 ]. This same minimal promoter drives erythroid specific expression in vivo [40]. The authors discovered a number of very interesting things. First, they found strong evidence for a clonal embryoid body-derived progenitor cell capable of differentiating into erythroid cells and F4/80þ macrophages, but not other cells. These colonies looked like erythroblastic islands [39 ]. The timing of development of these progenitors within embryoid bodies suggests they may be from the primitive wave of erythropoiesis. More needs to be done to examine the development of these erythroblastic island macrophages colonies in vivo. Are they restricted to the primitive wave, and do they seed the fetal liver directly from the yolk sac rather than from HSCs generated in the dorsal aorta? Do clonal erythroblastic island macrophage progenitor cells also exist in the bone marrow or spleen as a part of the normal ontogeny of hematopoiesis? Secondly and somewhat surprisingly, these authors found a significant percentage of F4/80þ macrophages in the fetal liver were GFPþ [39 ]. This is surprising given the fact that KLF1 is not expressed in myeloid cell lines [41] and not expressed in sorted primary macrophages or myeloid progenitor cells. Also, we have not been able to detect any Klf1 mRNA expression in CD11bþ F4/80þ CD169þ macrophages sorted from the adult mouse bone marrow by reversed transcription-PCR (AC Perkins, unpublished data). Nevertheless, the result by Xue et al. [39 ] is very convincing. It is important to note that &&
&&
&&
&&
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
217
Erythroid system and its diseases &&
Xue et al. [39 ] did not claim to detect KLF1 protein in central island macrophages, just GFP driven by the Klf1 promoter. KLF1 may be there but it is well known in the field that detection by immunofluorescence or western blotting is problematic with current antibodies. It remains an intriguing possibility that KLF1 protein is produced and functions as a transcriptional activator of key specialized genes in the erythroblastic island macrophage but not in other macrophage subtypes. In this way, KLF1 might coordinate functions such as iron metabolism and enucleation between the central macrophage and developing erythroid cells. One particularly interesting KLF1 target gene in erythroblastic island macrophages is Dnase2a. This gene encodes a nuclease which is essential for digestion of the pyrenocyte nucleus (Fig. 1). DNAse2a/ embryos die in mid-gestation from anemia with a marked excess of nucleated red blood cells [42]. Transplantation studies confirm the anemia is not cell autonomous to erythroid cells but because of a defect in the erythroblastic island macrophages with accumulation of undigested erythroid cell DNA, which induces an inflammatory response [42]. Evolutionary studies suggest the Klf1 gene was recently duplicated from the Klf2 gene situated 12.6 Mb distant on mouse chromosome 8. Interestingly, Klf1 has lodged directly adjacent to the DNAse2a gene. Porcu et al. [38] have shown Dnase2a is downregulated in Klf1/ mice and KLF1 can directly bind to the Dnase2a promoter in a CACC-box site-specific fashion to drive its expression in reporter assays in macrophage cell lines. This article strongly argues for a bona fide role for KLF1 in erythroblastic island macrophages. The results of Xue et al. [39 ] raise an alternative possibility. It is possible the Klf1 promoter can be interpreted by the transcriptional milieu in F4/80þ macrophages to drive GFP expression. That is, the Klf1 promoter might undertake two functions in vivo. In addition to regulation of Klf1 coding sequence in erythroid cells, it might directly activate the Dnase2a gene over a short distance in erythroblastic island macrophages. This could be via either enhancer type activity or some alternative splicing from the first exon of Klf1 (presumably noncoding region) to an exon of Dnase2a. We think this later scenario is possible and intriguing as it could coordinate the complex process of nuclear condensation and enucleation with engulfment and degradation by the erythroblastic island macrophage. This might occur via one regulatory element interpreted by two different cell types to generate two alternative functional outcomes, that is KLF1 production by erythroid cells and DNAse 2a production by erythroblastic island macrophages. &&
218
www.co-hematology.com
CONCLUSION The possibility to prospectively isolate erythroblastic island macrophages from the mouse will enable molecular studies to elucidate the mechanisms and the molecular machinery by which these macrophages control erythroblast maturation. Efforts will need to be devoted to a more precise phenotypic characterization of their human counterparts to validate studies in the mouse. The fact that macrophages are immune cells puts them in a central position to adapt erythropoiesis to infections and be prime actors of anemia of inflammation. Finally, the fact that in-vivo depletion of macrophages in mice can correct the erythropoietic symptoms of b-thalassemia and polycythemia vera further illustrates the key role of erythroblastic island macrophages in erythropathies and suggests that therapies targeting macrophages may have some benefit to treat these diseases. Acknowledgements None. Financial support and sponsorship J.-P.L. is supported by a Senior Research Fellowship (no. 1044091), R.N.J. by a Project Grant (no. 1046590) and A.C.P. by a Project Grant (1030143) from the National Health and Medical Research Council of Australia. Conflicts of interest The authors have no conflict of interest.
REFERENCES AND RECOMMENDED READING Papers of particular interest, published within the annual period of review, have been highlighted as: & of special interest && of outstanding interest 1. Bessis M. L’ilot erythroblastique, unite fonctionelle de la moelle osseuse. Rev Hematol 1958; 13:8–11. 2. Chasis JA, Mohandas N. Erythroblastic islands: niches for erythropoiesis. Blood 2008; 112:470–478. 3. Sawada K, Krantz SB, Dessypris EN, et al. Human colony-forming unitserythroid do not require accessory cells, but do require direct interaction with insulin-like growth factor I and/or insulin for erythroid development. J Clin Invest 1989; 83:1701–1709. 4. Manwani D, Bieker JJ. The erythroblastic island. Curr Top Dev Biol 2008; 82:23–53. 5. Bessis MC, Breton-Gorius J. Iron metabolism in the bone marrow as seen by electron microscopy: a critical review. Blood 1962; 19:635–663. 6. Leimberg MJ, Prus E, Konijn AM, et al. Macrophages function as a ferritin iron source for cultured human erythroid precursors. J Cell Biochem 2008; 103:1211–1218. 7. Skutelsky E, Danon D. On the expulsion of the erythroid nucleus and its phagocytosis. Anat Rec 1972; 173:123–126. 8. Soni S, Bala S, Gwynn B, et al. Absence of erythroblast macrophage protein (Emp) leads to failure of erythroblast nuclear extrusion. J Biol Chem 2006; 281:20181–20189. 9. Crocker P, Werb Z, Gordon S, et al. Ultrastructural localization of a macrophage-restricted sialic acid binding hemagglutinin, SER, in macrophagehematopoietic cell clusters. Blood 1990; 76:1131–1138. 10. Hanspal M, Hanspal JS. The association of erythroblasts with macrophages promotes erythroid proliferation and maturation: a 30-kD heparin-binding protein is involved in this contact. Blood 1994; 84:3494–3504.
Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al. 11. Hanspal M, Smockova Y, Uong Q. Molecular identification and functional characterization of a novel protein that mediates the attachment of erythroblasts to macrophages. Blood 1998; 92:2940–2950. 12. Sadahira Y, Yoshino T, Monobe Y. Very late activation antigen 4-vascular cell adhesion molecule 1 interaction is involved in the formation of erythroblastic islands. J Exp Med 1995; 181:411–415. 13. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythroblastic island formation. Blood 2006; 108:2064–2071. 14. de Jong JP, Voerman JS, van der Sluijs-Gelling AJ, et al. A monoclonal antibody (ER-HR3) against murine macrophages. I. Ontogeny, distribution and enzyme histochemical characterization of ER-HR3-positive cells. Cell Tissue Res 1994; 275:567–576. 15. Sonoda Y, Sasaki K. Hepatic extramedullary hematopoiesis and macrophages in the adult mouse: histometrical and immunohistochemical studies. Cells Tissues Organs 2012; 196:555–564. 16. Jacobsen RN, Forristal CE, Raggatt LJ, et al. Mobilization with granulocyte & colony-stimulating factor blocks medullar erythropoiesis by depleting F4/ 80 þ VCAM1 þ CD169 þ ER-HR3 þ Ly6Gþ erythroid island macrophages in the mouse. Exp Hematol 2014; 42:547–561; e544. Identification of the phenotype of mouse erythroblastic island macrophages enabling their prospective sorting from mouse bone marrow and spleen by flow cytometry. In-vivo depletion of CD169þ macrophages or phagocytes causes loss of erythroblastic island macrophages in bone marrow and spleen and arrest of medullary and splenic erythropoiesis with loss of erythroblasts. The cytokine GCSF used to mobilize hematopoietic stem and progenitor cells into the blood for transplantation blocks medullary erythropoiesis by depleting erythroblastic island macrophages in the bone marrow but not in the spleen. 17. Chow A, Huggins M, Ahmed J, et al. CD169þ macrophages provide a niche && promoting erythropoiesis under homeostasis and stress. Nat Med 2013; 19:429–436. First demonstration that in-vivo depletion of CD169þ macrophages or phagocytes stops medullary erythropoiesis by depleting erythroblastic island macrophages that express CD169. Furthermore, CD169þ macrophage depletion delays erythropoietic rescue after myeloablation or challenge with PHZ. Phagocyte depletion reduces hematocrit and red blood cell counts in a mouse model of polycythemia vera driven by JAK2V617F mutation. 18. Ramos P, Casu C, Gardenghi S, et al. Macrophages support pathological && erythropoiesis in polycythemia vera and [b]-thalassemia. Nat Med 2013; 19:437–445. In-vivo phagocyte depletion delays erythropoietic rescue after phlebotomy in wildtype mice and reduces hematocrit and red blood cell counts in a mouse model of polycythemia vera driven by JAK2V617F mutation. Chronic phagocyte depletion corrects anemia in a mouse model of b-thalassemia. 19. Mullally A, Lane SW, Ball B, et al. Physiological Jak2V617F expression causes a lethal myeloproliferative neoplasm with differential effects on hematopoietic stem and progenitor cells. Cancer Cell 2010; 17:584–596. 20. Bader BL, Rayburn H, Crowley D, et al. Extensive vasculogenesis, angiogenesis, and organogenesis precede lethality in mice lacking all av integrins. Cell 1998; 95:507–519. 21. Wang Z, Vogel O, Kuhn G, et al. Decreased stability of erythroblastic islands in integrin b3-deficient mice. Physiol Rep 2013; 1:e00018. 22. Spring FA, Griffiths RE, Mankelow TJ, et al. Tetraspanins CD81 and CD82 facilitate a4b1-mediated adhesion of human erythroblasts to vascular cell adhesion molecule-1. PLoS One 2013; 8:e62654. 23. Scott LM, Priestley GV, Papayannopoulou T. Deletion of {a}4 integrins from adult hematopoietic cells reveals roles in homeostasis, regeneration, and homing. Mol Cell Biol 2003; 23:9349–9360. 24. Ulyanova T, Scott LM, Priestley GV, et al. VCAM-1 expression in adult hematopoietic and nonhematopoietic cells is controlled by tissue-inductive signals and reflects their developmental origin. Blood 2005; 106:86–94. 25. Ulyanova T, Jiang Y, Padilla S, et al. Combinatorial and distinct roles of a5 and a4 integrins in stress erythropoiesis in mice. Blood 2011; 117:975–985.
26. Ulyanova T, Padilla SM, Papayannopoulou T. Stage specific functional roles of integrins in erythropoiesis. Exp Hematol 2014; 42:404–409. Conditional deletion of the a4 integrin gene in erythroblasts impairs erythropoietic recovery following challenge with PHZ. Conditional deletion of the a5 integrin gene in erythroblasts has no effect on erythropoietic recovery. 27. Robier C, Amouzadeh-Ghadikolai O, Bregant C, et al. The anti-VLA-4 anti& body natalizumab induces erythroblastaemia in the majority of the treated patients with multiple sclerosis. Mult Scler 2014; 20:1269–1272. Ninety-three percent of multiple sclerosis patients treated with the antia4 integrin natalizumab have erythroblastemia. None of the multiple sclerosis patients treated with chronic b-interferon experienced erythroblastemia. 28. Papayannopoulou T, Nakamoto B. Peripheralization of hemopoietic progenitors in primates treated with anti-VLA4 integrin. Proc Natl Acad Sci USA 1993; 90:9374–9378. 29. Mao X, Shi X, Liu F, et al. Evaluation of erythroblast macrophage protein related to erythroblastic islands in patients with hematopoietic stem cell transplantation. Eur J Med Res 2013; 18:9. 30. Sui Z, Nowak RB, Bacconi A, et al. Tropomodulin3-null mice are embryonic && lethal with anemia due to impaired erythroid terminal differentiation in the fetal liver. Blood 2014; 123:758–767. Deletion of the tropomodullin gene is embryonic lethal from day 14.5 of gestation in mice with severe fetal liver anemia, defective erythroblastic island assembly, defective erythroblast maturation, and enucleation. 31. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by & central macrophages in erythroblastic islands. Blood 2014; 123:3963–3971. Engulfment of pyrenocytes by erythroblastic island macrophages depends on the tyrosine kinase receptor Mer expressed by erythroblastic island macrophages. Phosphatidyl serine flips from the inner to the outer lipid bilayer of pyrenocyte plasma membrane and binds to Mer with protein S acting as a bridge. This signals engulfment by the macrophage. 32. McGrath KE. Red cell island dances: switching hands. Blood 2014; 123:3847–3848. 33. Tang H, Chen S, Wang H, et al. TAM receptors and the regulation of erythropoiesis in mice. Haematologica 2009; 94:326–334. 34. Perkins AC, Sharpe AH, Orkin SH. Lethal [b]-thalassaemia in mice lacking the erythroid CACCC-transcription factor EKLF. Nature 1995; 375:318–322. 35. Tallack MR, Perkins AC. KLF1 directly coordinates almost all aspects of terminal erythroid differentiation. IUBMB Life 2010; 62:886–890. 36. Siatecka M, Bieker JJ. The multifunctional role of EKLF/KLF1 during erythropoiesis. Blood 2011; 118:2044–2054. 37. Tallack MR, Magor GW, Dartigues B, et al. Novel roles for KLF1 in erythropoiesis revealed by mRNA-seq. Genome Res 2012; 22:2385–2398. 38. Porcu S, Manchinu MF, Marongiu MF, et al. Klf1 affects DNase II-a expression in the central macrophage of a fetal liver erythroblastic island: a non-cell-autonomous role in definitive erythropoiesis. Mol Cell Biol 2011; 31:4144–4154. 39. Xue L, Galdass M, Gnanapragasam MN, et al. Extrinsic and intrinsic control by && EKLF (KLF1) within a specialized erythroid niche. Development 2014; 141:2245–2254. Cultures from a single progenitor cell derived from mouse embryoid bodies give rise to complete erythroblastic islands with erythroblasts rosetting around a central macrophage suggesting that during the primitive wave of erythropoiesis, erythroblasts, and erythroblastic island macrophages may emerge from a common progenitor. Furthermore, in the fetal liver of embryos containing a fluorescent reporter under the control of the Klf1 gene promoter, both erythroblasts and macrophages are labeled suggesting that the Klf1 promoter is active in both erythroblasts and EI macrophages. 40. Xue L, Chen X, Chang Y, et al. Regulatory elements of the EKLF gene that direct erythroid cell-specific expression during mammalian development. Blood 2004; 103:4078–4083. 41. Perkins A. Erythroid Kruppel like factor: from fishing expedition to gourmet meal. Int J Biochem Cell Biol 1999; 31:1175–1192. 42. Kawane K, Fukuyama H, Kondoh G, et al. Requirement of DNase II for definitive erythropoiesis in the mouse fetal liver. Science 2001; 292:1546–1549. &
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
219
Macrophages and regulation of erythropoiesis Rebecca N. Jacobsen a,b, Andrew C. Perkins a,c,d, and Jean-Pierre Levesque a,b,c
Purpose of review The nature and function of macrophages at the center of erythroblastic islands is not fully understood. This review discusses novel findings on the phenotypic and molecular characterization of erythroblastic island macrophages, and their role in regulating normal and pathological erythropoiesis. Recent findings The phenotype to prospectively isolate erythroblastic island macrophages from mouse bone marrow has been identified. In-vivo depletion of erythroblastic island macrophages causes blockade of erythroblast maturation and delays erythropoietic recovery following chemical insults. The cytokine granulocyte colonystimulating factor arrests medullary erythropoiesis by depleting erythroblastic island macrophages from the bone marrow. In-vivo ablation of macrophages improves anemia associated with b-thalassemia and reduces red blood cell counts in the mouse model of polycythemia vera. The role of cell adhesion molecules regulating interactions between erythroblastic island macrophages and erythroblasts has been clarified, and mechanisms of pyrenocyte engulfment by erythroblastic island macrophages have been demonstrated to involve Mer tyrosine kinase receptor. Summary Prospective isolation of mouse erythroblastic island macrophages together with new genetic mouse models to specifically target erythroblastic island macrophages will enable molecular studies to better define their role in controlling erythroblast maturation. These studies have revealed the key role of erythroblastic island macrophages in regulating normal erythropoiesis and could be interesting targets to treat b-thalassemia or polycythemia vera. Keywords anemia, erythroblastic island, erythropoiesis, macrophage, red blood cell
INTRODUCTION Macrophages have two major roles in regulating red blood cell homeostasis, controlling the final maturation of erythroblasts into reticulocytes and clearing old erythrocytes. In 1958 Marcel Bessis first described erythroblastic islands, consisting of clusters of erythroid cells at various stages of differentiation surrounding a central macrophage, in bone marrow [1]. Similar erythroblastic islands were observed in the adult mouse bone marrow, spleen and fetal liver [2]. From these observations, it was proposed that erythroblastic island macrophages contribute to the final stages of erythroblast differentiation and maturation. In particular, they have been proposed to secrete cytokines essential to erythroblast survival and maintenance [3,4], transport iron to erythroblasts [5,6] to support the synthesis of large quantities of hemoglobin, and phagocytose and degrade erythroblast nuclei during nuclear extrusion, a step necessary to generate enucleated reticulocytes [7,8]. Until the www.co-hematology.com
mid-2000s, most studies were based on microscopy observations or in-vitro studies in culture as the phenotypic identification of erythroblastic island macrophages has been elusive. However in 2006, the gene encoding erythroblast macrophage protein (EMP also called macrophage erythroblast attacher or Maea), a protein expressed by both erythroblasts and erythroblastic island macrophages and promoting their heterotypic adhesion, was identified and a
Blood and Bone Diseases Program, Mater Research Institute – University of Queensland, Woolloongabba, bSchool of Medicine, cSchool School of Biomedical Sciences, University of Queensland, Saint Lucia and dDepartment of Haematology, Princess Alexandra Hospital, Woolloongabba, Queensland, Australia Correspondence to Jean-Pierre Le´vesque, Associate Professor, PhD, Blood and Bone Diseases Program, Mater Research Institute – University of Queensland, TRI Building, 37 Kent Street, Woolloongabba, QLD 4102, Australia. Tel: +61 7 3443 7571; e-mail: [email protected]. edu.au Curr Opin Hematol 2015, 22:212–219 DOI:10.1097/MOH.0000000000000131 Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
KEY POINTS The phenotype of mouse erythroblastic island macrophages has been identified enabling their prospective isolation for molecular studies. In-vivo depletion of erythroblastic island macrophages stops erythropoiesis with loss of erythroblasts. Erythroblastic island macrophages are necessary to erythropoietic rescue following myeloablation or phenylhydrazine red cell poisoning. Elimination of erythroblastic island macrophages can normalize erythrocyte counts in polycythemia vera or b-thalassemia. New mechanisms by which erythroblastic island macrophages bind to erythroblasts and eliminate nucleus during erythroblast maturation are better understood.
differentiation (CD)169 or sialic acid binding immunoglobulin-like lectin1 (Siglec-1) was recognized as being present on the plasma membrane of erythroblastic island macrophages at the interface with erythroblast plasma membrane [9]. Other cell adhesion molecules such as EMP [10,11], vascular cell adhesion molecule-1 (VCAM-1 or CD106) [12], and aV integrins [13] were found to be present on erythroblastic island macrophages and play an important functional role in promoting adhesion of erythroblasts. Likewise, the mouse antigen recognized by the rat monoclonal antibody ER-HR3 was also found to be expressed on mouse erythroblastic island macrophages [14,15]. However, taken in isolation, none of these cell surface proteins is exclusively expressed on erythroblastic island macrophages precluding their prospective isolation for cellular and molecular studies. In a recent study [16 ], we combined antibodies specific for CD169, VCAM-1, and ER-HR3 which are all expressed by erythroblastic island macrophages together with myeloid specific antigens such as CD11b (aM integrin), F4/80 (antigen specific of mouse macrophages), and Ly-6G, an antigen mostly expressed by granulocytes but also subsets of macrophages. We identified by flow cytometry in the adult mouse bone marrow and spleen macrophages that were positive for all these cell surface proteins [i.e. CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ lymphocyte antigen (Ly)-6Gþ]. This subset of macrophages represented 1.3 0.2% of bone marrow leukocytes and 27 6% of macrophages identified in cell aggregates that were positive for both the erythroid marker Ter119 and the myeloid marker CD11b. Furthermore, we showed that daily administration of granulocyte colony-stimulating factor (G-CSF) selectively depleted this macrophage subset by 90% in the bone marrow and concomitantly blocked medullary erythropoiesis with a 90–50% reduction in the number of all erythroblast subsets and reticulocytes depending on the antibody panel used [16 ]. To further prove that CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ Ly-6Gþ macrophages are bona fide erythroblastic island macrophages, we selectively depleted CD169þ macrophages in Siglec1DTR/þ mice in which a simian diphtheria toxin receptor is knocked in the CD169 gene to specifically sensitize CD169þ macrophages to diphtheria toxin-mediated killing. Mice do not express diphtheria toxin receptors and are consequently resistant to this potent protein toxin. We also depleted phagocytes in vivo by intravenous injection of clodronate-loaded liposomes. In both these models, CD169þ macrophage depletion or phagocyte depletion caused loss of erythroblasts and reticulocytes in bone marrow and spleen &
knocked-out in the mouse. Following deletion of the EMP (Maea) gene, pups died perinatally. Embryos from E12.5 days of gestation and beyond were anemic with accumulation of nucleated erythrocytes in the blood and absence of erythroblastic island in the fetal liver [8]. Furthermore, EMP expressed on macrophages was necessary to establish erythroblastic island in vitro whereas EMP on erythroblasts was dispensable [8]. These data proved that erythroblastic island macrophages are necessary for terminal erythroblast maturation and enucleation. However, the absence of the precise phenotypic profile of erythroblastic island macrophages has hampered progress in understanding the mechanism by which erythroblastic island macrophages control erythroblast maturation in normal settings and how they contribute to diseased states, such as anemia resulting from myeloid leukemia and chronic inflammation, myeloproliferative diseases with expansion of the erythroid compartment such as polycythemia vera, and genetic erythropathies such as b-thalassemia. In the past 2 years, significant progress has been made in the phenotypic identification of erythroblastic island macrophages in mice and humans with a better understanding of their function in steady-state erythropoiesis or in disease states.
PHENOTYPIC CHARACTERIZATION OF ERYTHROBLASTIC ISLAND MACROPHAGES In the 1990s, many antigens expressed by erythroblastic island macrophages were identified particularly in the mouse. For instance, cluster of
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
&
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
213
Erythroid system and its diseases
concomitant with loss of CD11bþ F4/80þ VCAM-1þ CD169þ ER-HR3þ Ly-6Gþ macrophages in both tissues [16 ]. Interestingly, in these three models (G-CSF injection, CD169þ macrophage depletion, and phagocyte depletion), loss of erythroblasts was accompanied by accumulation of proerythroblasts suggesting a failure of maturation [16 ]. Of note however, unlike CD169þ macrophage or phagocyte depletion which affects all tissues in the body, thus both medullar and splenic erythropoiesis, G-CSF increased splenic erythropoiesis, erythroblastic island macrophages and cell aggregates containing erythroid cells and erythroblastic island macrophages despite the fact that both medullar and splenic erythroblastic island macrophages express the G-CSF receptor mRNA [16 ]. This suggests that the response of erythroblastic island macrophages to cytokines such as G-CSF varies depending on the tissue of residence. &
&
&
ROLE OF ERYTHROBLASTIC ISLAND MACROPHAGES IN ERYTHROPATHIES AND POLYCYTHEMIA VERA
&&
The effect of macrophage depletion on erythropoiesis was also studied by Paul Frenette’s and Stefano Rivella’s laboratories, with an emphasis on the role macrophages play in stress erythropoiesis and erythropathies such as polycythemia vera and b-thalassemia. Frenette’s group utilized both the Siglec1DTR/þ mouse model and phagocyte depletion using clodronate-loaded liposomes to study the effect of macrophage depletion in steady-state, as well as recovery from hemolytic anemia induced by hemoglobin-oxidizing phenylhydrazine (PHZ), phlebotomy, myeloablation, and also in a mouse model of polycythemia vera, whereas Rivella’s group only used the liposome-mediated phagocyte depletion to study similar conditions as well as b-thalassemia. Both groups observed a loss of developing erythroblasts from mouse bone marrow when phagocytes or CD169þ macrophages were depleted [17 ,18 ]. Also, prolonged administration of clodronateloaded liposomes (over 12 weeks) induced an anemia with characteristics of iron deficiency anemia, which was however not corrected by iron supplementation [18 ]. Therefore, erythroblastic island macrophages play an important role in erythropoietic recovery independently of iron transport and recycling. Both groups observed that phagocyte depletion with clodronate liposomes or depletion of CD169þ macrophages delayed erythropoietic recovery following PHZ insult [17 ,18 ]. Likewise, erythropoietic recovery following myeloablation and bone marrow transplantation was delayed when CD169þ macrophages were ablated following &&
&&
&&
214
www.co-hematology.com
&&
&&
transplant [17 ]. Taken together these data show that erythroblastic island macrophages are essential to support erythroblast maturation in steady-state as well as during erythropoietic rescue. Polycythemia vera is a myeloproliferative neoplasm driven by a V617F point mutation that constitutively activates the tyrosine kinase Janus kinase 2 (JAK2) which is downstream of many cytokine receptors including the erythropoietin receptor. Mice with the JAK2V617F mutation knocked into the endogenous mouse Jak2 gene [19] provides a robust mouse model of polycythemia vera with characteristics comparable to the human condition. Clodronate-loaded liposome treatment of mice with either early or well established polycythemia vera led to the normalization of hematocrits and red blood cell counts, decreased erythropoietic activity (reduced reticulocytosis), extramedullary hematopoiesis, and decreased splenomegaly. Polycythemia vera symptoms were reduced for up to 4 weeks following cessation of the liposome treatment and only a single liposome injection was required at this time point to return blood parameters to normal [18 ]. Therefore, phagocytes which comprise erythroblastic island macrophages play an important role in maturing erythroblasts carrying the malignant JAK2V617F mutation and could be a potential therapeutic target for this neoplasm. b-Thalassemia is characterized by increased but ineffective erythropoiesis caused by a mutation in the HBB gene, which reduces b-hemoglobin chain synthesis and failure to assemble functional hemoglobin. Hbbth3/þ mice, in which a mutation has been introduced in one allele of the Hbb gene, recapitulate the human b-thalassemia intermedia phenotype. Phagocyte depletion in these mice improved anemia and hemoglobin synthesis. Importantly, long-term administration of clodronate liposomes improved the symptoms of anemia and splenomegaly, possibly by increasing the lifespan of the red blood cells [18 ]. This suggests that macrophages can exacerbate anemia observed in b-thalassemia by accelerating degradation of defective erythrocytes.
&&
&&
CRITICAL MOLECULES WHICH PARTICIPATE IN THE ERYTHROBLASTIC ISLAND INTERACTION Erythroid progenitor cells are not simply passive players within the erythroblastic island. They express cell surface proteins which make numerous productive and essential interactions with the central macrophage. For example, erythroid progenitor cells express ICAM4 which interacts with aV integrins on the erythroblastic island macrophage, a4-integrin and b1-integin which together form Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
the very late activation antigen (VLA)-4 receptor which interacts with VCAM-1 on erythroblastic island macrophages, and EMP expressed by both erythroblasts and erythroblastic island macrophages, which forms homotypic adhesive interactions between these two cell types. There are other less well studied molecules which participate (see Table 1 and Fig. 1). Of the cell adhesion molecules that mediate adhesion between erythroblasts and erythroblastic island macrophages, EMP seems to be the most critical in steady-state. Indeed, it is the only gene encoding a cell adhesion molecule whose deletion causes embryonic anemia and defective erythroblastic island formation during development [8]. ICAM4 knockout (KO) mice have no anemia in steady-state, however the formation of erythroblastic island in vivo and in vitro is reduced [13]. ICAM4 is expressed by erythroblasts. ICAM4 counter receptors on erythroblastic island macrophages are thought to include av integrins (CD51) as peptides blocking ICAM4–aV interaction reduce the formation of erythroblastic islands in vitro [13]. Deletion of the av integrin gene is embryonic lethal between E10 and E12 so its role in establishing erythroblastic island in the fetal liver is not known [20]. Conditional deletion of the av integrin gene specifically in macrophages would answer this question. As integrins are expressed as ab heterodimers, the identity of the av partner integrin on erythroblastic island macrophages has been of interest. A candidate partner is b3-integrin, thus forming an aVb3 dimer. In support of this hypothesis is the presence of mild anemia (lower hematocrit with enhanced number of reticulocytes and CD71þ erythroblasts in the blood) and mild erythroblastic island defects (lower number of erythroblasts per erythroblastic island) in b3/ mice [21]. Table 1. List of molecular interactions between the erythroblastic island macrophages and erythroid cells Macrophage
Erythroblast
VCAM-1
a4b1 integrin complexed with CD81, CD82, and CD151
avb3 integrin
ICAM4
EMP
EMP
MerTK
Phosphatidyl serine complexed with protein S
Axl
Unknown
CD169
Various sialylated proteins
CD163
Unknown
Axl, Axl tyrosine kinase receptor; CD, cluster of differentiation; EMP, erythroblast macrophage protein; ICAM4, intercellular adhesion molecule-4; MerTK, Mer tyrosine kinase receptor; VCAM-1, vascular cell adhesion molecule-1.
Recent evidence suggests a4b1 integrin exists in a large transmembrane complex which includes tetraspanin proteins CD81, CD82, and CD151, and these influence the functional interactions with VCAM-1 on erythroblastic island macrophages [22]. Conditional deletion of the Vcam1 gene, a4 integrin (Itga4) gene, or b1 integrin (Itgb1) gene in hematopoietic cells (germinal knock-out for these genes are embryonic lethal) results in increased trafficking and spontaneous mobilization of hematopoietic stem and progenitor cells in the blood [23,24]. In these three conditional mutants, basal erythropoiesis is normal however erythropoietic recovery following chemical insult with phenyl hydrazine is severely compromised in b1D/D mice with reduced numbers of erythroblasts in spleen and bone marrow. a4D/D and VCAM1D/D had less severe impairment of erythropoiesis rescue following phenyl hydrazine challenge [25]. This effect is autonomous to erythroid cells as conditional deletion of the a4 integrin gene in erythroid cells using EporCre mice led to similar impairment of rescue erythropoiesis following challenge with PHZ [26 ]. These mice also had abnormally elevated number of circulating erythroblasts [26 ]. In humans, blocking antibodies against a4-integrin/VLA-4 result in transient anemia and increased circulating nucleated erythroblasts suggesting they interfere with adhesion between erythroid progenitor cells and the erythroblastic island macrophages in vivo [27 ]. These antibodies also release hematopoietic stem and progenitor cells into the circulation [28], so VLA-4 is also important for hematopoietic stem cell (HSC) interactions with other stromal macrophages in the stem cell niche. Indeed, the erythroblastic island macrophage is probably just one of many specialized types of niche macrophage. Interestingly, antibodies (such as natalizumab) and small molecule inhibitors of VLA-4 are increasingly being used in the clinic to treat various disorders such as multiple sclerosis. Thus, hematologists are seeing referrals from concerned neurologists to exclude more sinister causes of leukoerythroblastic blood films, which are simply because of a transient mobilization of immature blood cells including nucleated erythroid cells usually retained by niche macrophages (including erythroblastic island macrophages) via the a4b1–VCAM1 interaction. &
&
&
EVIDENCE FOR A ROLE OF ERYTHROBLASTIC ISLAND MACROPHAGES IN HUMANS Although erythroblastic islands have been visualized in human bone marrow, the characteristics of the erythroblastic island macrophage and its
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
215
Erythroid system and its diseases
Reticulocyte Erythroblast
KLF-1
α4β1
EMP ?CD169 ?CD163
αvβ3
Pyrenocyte
EMP
ICAM4 VCAM1
Phosphatidyl serine Axl ?
Protein S MerTK
F4/80 Ly6G
Erythroblastic island macrophage
ER-HR3 KLF-1????
DNAse IIa
FIGURE 1. Model of interactions between the erythroblastic island macrophages and maturing erythroblasts. Erythroblastic island macrophages develop adhesive interactions with maturing erythroblasts using erythroblast macrophage protein (EMP), a3b1 integrin, vascular cell adhesion molecule-1 (VCAM-1), and possibly cluster of differentiation (CD) 169 and CD163 and their cognate counter receptors on erythroblasts. Kruppel like factor 1 (KLF1) in erythroblasts induces expression of intercellular adhesion molecule-4 (ICAM4). During the enucleation of erythroblasts, phosphatidyl serine flips from the inner to the outer layer of the plasma membrane, and binds to Mer tyrosine kinase receptor via bridging with protein S. This triggers an engulfment signal of the pyrenocyte by the erythroblastic island macrophage and subsequent degradation of the pyrenocyte nucleus by DNase IIa. The enucleated reticulocyte is released in the circulation.
relationship to the mouse erythroblastic island macrophage have not been fully explored. Ex-vivo cultures of erythroblasts isolated from human bone marrow in steady-state or from polycythemia vera or b-thalassemia patients showed increased proliferation and differentiation when cultured in the presence of macrophages isolated from the same individuals. This was not observed in cocultures wherein erythroblasts were separated from erythroblastic island macrophages by a semipermeable membrane [18 ]. Thus, direct contact between erythroblasts and erythroblastic island macrophages in humans is necessary for erythroblast differentiation and proliferation [18 ]. Further to this, Frenette’s group examined bone marrow aspirates from healthy donors and assessed the expression of CD169 and VCAM-1 on monocytes and macrophages. These two antigens were not expressed by CD15þCD14 granulocytes, or CD15þCD14þ monocytes. However, cells expressing both CD169 and VCAM-1 were present exclusively within the CD15CD163þ macrophage population [17 ] indicating that similar to mice, CD169 and VCAM-1 are both expressed on erythroblastic island macrophages in human bone marrow. Interestingly, following HSC transplantation in humans EMP protein and mRNA expression in mononuclear cells
from blood and bone marrow were dramatically upregulated with increased density of erythroblastic islands in the bone marrow [29]. In severe anemia patients, erythroblastic islands and EMP were barely detectable however erythropoietin treatment boosted both erythroblastic island density and EMP mRNA expression in these patients [29]. This suggests that EMP may also play an important role in establishing human erythroblastic islands.
&&
&&
&&
216
www.co-hematology.com
A ROLE FOR THE MACROPHAGE IN ENGULFMENT AND DIGESTION OF THE PYKNOTIC ERYTHROID NUCLEUS The central macrophage plays a critical role in the final steps of erythroid cell maturation. The final division of maturing erythroid cells is an asynchronous one; the pyknotic nucleus and other organelles are deposited into one daughter cell, called a pyrenocyte. The other daughter cell, a reticulocyte, is released from the erythroblastic island macrophage into the circulation. Asynchronous deposition of cell surface receptors (such as a4b1 integrin), membrane lipids such phosphatidyl serine, and cytoskeletal components is critically important for the retention of one daughter cell by the erythroblastic Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al.
island macrophage and the release of the other. The erythroblastic island macrophage expresses proteins specifically designed to undertake the tasks of adhesion, engulfment, and degradation of the pyrenocyte. There are likely to be many erythroblastic island macrophage specific as well as general macrophage proteins which coordinate these processes. This is a remarkably efficient process as a sibling pyrenocyte is produced for every one of the million reticulocytes produced every second. EMP is the key to this process as EMP/ mice have defective erythroblastic islands and erythroblast enucleation [8]. Tropomodulin3 (Tmod3) is a recently described actin-interacting protein that caps the pointed ends of actin fibers regulating their length and stability. Tmod3/ embryos die between E14.5 and E18.5 of anemia with defective erythroblastic island formation, reduced numbers of BFU-Es, CFU-Es, and erythroblasts in the fetal liver [30 ]. Interestingly, Tmod3 is necessary in both erythroblasts and erythroblastic island macrophages for efficient erythroblastic island formation [30 ]. Mer tyrosine kinase receptor is an example of a critical erythroblastic island macrophage receptor that is required for the engulfment of pyrenocytes via recognition of protein S bound to phosphatidyl serine on the cell membrane [31 ,32]. However, other related transmembrane tyrosine kinases such as Axl tyrosine kinase receptor may play a similar role as mice double KO for Mer tyrosine kinase receptor and Axl have more pronounced medullar and splenic anemia in steady-state than single KO for each gene [33] (Fig. 1). &&
&&
&
THE ROLE OF KRUPPEL LIKE FACTOR 1 IN COORDINATING ERYTHROBLASTIC ISLAND FUNCTION The transcription factor Kruppel like factor 1 (KLF1), previously called EKLF, is essential for definitive erythropoiesis in mice [34] and man (G. Magor et al., in preparation) and is a key regulator of many erythroid genes including those involved in hemoglobin production, nuclear compaction, enucleation, and iron metabolism [35,36]. A key feature of KLF1-deficient erythropoiesis in both species is very high numbers of circulating nucleated erythrocytes which is suggestive of a specific problem with enucleation. Klf1/ mouse and human cells have dramatically reduced levels of ICAM4 [37], and KLF1 protein directly binds the Icam4 gene promoter [35,36]. Thus, KLF1 directly regulates genes, which facilitates erythroid cell attachment to the erythroblastic island macrophage. Recent evidence suggests KLF1 might also play a surprising role intrinsically within erythroblastic
island macrophages. This is controversial as KLF1 was considered to be an erythroid cell specific transcription factor for decades. Recently, KLF1 protein or its DNA-binding activity has been reported within erythroblastic island macrophages [38]. However, it was not clear whether the KLF1 immunofluorescence signal was coming from the macrophage or ingested pyrenocytes. Common myeloid progenitors give rise to bipotent granulocyte monocyte progenitors and megakaryocyte erythroid progenitors. Macrophages derive from granulocyte monocyte progenitors and erythroid cells from megakaryocyte erythroid progenitors. The traditional dogma is that these two lineages are distinct and do not share an immediate common progenitor cell. There are few reports of hematopoietic colonies composed solely of macrophages and erythroid cells. A recent article by Bieker’s group has challenged these views about the ontogeny of macrophages versus erythrocytes, and also the restricted expression of KLF1 to erythroid cells. The authors introduced a single copy transgene construct, in which green fluorescent protein (GFP) reporter is driven by 950 bp of the Klf1 promoter, into embryonic stem cells [39 ]. This same minimal promoter drives erythroid specific expression in vivo [40]. The authors discovered a number of very interesting things. First, they found strong evidence for a clonal embryoid body-derived progenitor cell capable of differentiating into erythroid cells and F4/80þ macrophages, but not other cells. These colonies looked like erythroblastic islands [39 ]. The timing of development of these progenitors within embryoid bodies suggests they may be from the primitive wave of erythropoiesis. More needs to be done to examine the development of these erythroblastic island macrophages colonies in vivo. Are they restricted to the primitive wave, and do they seed the fetal liver directly from the yolk sac rather than from HSCs generated in the dorsal aorta? Do clonal erythroblastic island macrophage progenitor cells also exist in the bone marrow or spleen as a part of the normal ontogeny of hematopoiesis? Secondly and somewhat surprisingly, these authors found a significant percentage of F4/80þ macrophages in the fetal liver were GFPþ [39 ]. This is surprising given the fact that KLF1 is not expressed in myeloid cell lines [41] and not expressed in sorted primary macrophages or myeloid progenitor cells. Also, we have not been able to detect any Klf1 mRNA expression in CD11bþ F4/80þ CD169þ macrophages sorted from the adult mouse bone marrow by reversed transcription-PCR (AC Perkins, unpublished data). Nevertheless, the result by Xue et al. [39 ] is very convincing. It is important to note that &&
&&
&&
&&
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
217
Erythroid system and its diseases &&
Xue et al. [39 ] did not claim to detect KLF1 protein in central island macrophages, just GFP driven by the Klf1 promoter. KLF1 may be there but it is well known in the field that detection by immunofluorescence or western blotting is problematic with current antibodies. It remains an intriguing possibility that KLF1 protein is produced and functions as a transcriptional activator of key specialized genes in the erythroblastic island macrophage but not in other macrophage subtypes. In this way, KLF1 might coordinate functions such as iron metabolism and enucleation between the central macrophage and developing erythroid cells. One particularly interesting KLF1 target gene in erythroblastic island macrophages is Dnase2a. This gene encodes a nuclease which is essential for digestion of the pyrenocyte nucleus (Fig. 1). DNAse2a/ embryos die in mid-gestation from anemia with a marked excess of nucleated red blood cells [42]. Transplantation studies confirm the anemia is not cell autonomous to erythroid cells but because of a defect in the erythroblastic island macrophages with accumulation of undigested erythroid cell DNA, which induces an inflammatory response [42]. Evolutionary studies suggest the Klf1 gene was recently duplicated from the Klf2 gene situated 12.6 Mb distant on mouse chromosome 8. Interestingly, Klf1 has lodged directly adjacent to the DNAse2a gene. Porcu et al. [38] have shown Dnase2a is downregulated in Klf1/ mice and KLF1 can directly bind to the Dnase2a promoter in a CACC-box site-specific fashion to drive its expression in reporter assays in macrophage cell lines. This article strongly argues for a bona fide role for KLF1 in erythroblastic island macrophages. The results of Xue et al. [39 ] raise an alternative possibility. It is possible the Klf1 promoter can be interpreted by the transcriptional milieu in F4/80þ macrophages to drive GFP expression. That is, the Klf1 promoter might undertake two functions in vivo. In addition to regulation of Klf1 coding sequence in erythroid cells, it might directly activate the Dnase2a gene over a short distance in erythroblastic island macrophages. This could be via either enhancer type activity or some alternative splicing from the first exon of Klf1 (presumably noncoding region) to an exon of Dnase2a. We think this later scenario is possible and intriguing as it could coordinate the complex process of nuclear condensation and enucleation with engulfment and degradation by the erythroblastic island macrophage. This might occur via one regulatory element interpreted by two different cell types to generate two alternative functional outcomes, that is KLF1 production by erythroid cells and DNAse 2a production by erythroblastic island macrophages. &&
218
www.co-hematology.com
CONCLUSION The possibility to prospectively isolate erythroblastic island macrophages from the mouse will enable molecular studies to elucidate the mechanisms and the molecular machinery by which these macrophages control erythroblast maturation. Efforts will need to be devoted to a more precise phenotypic characterization of their human counterparts to validate studies in the mouse. The fact that macrophages are immune cells puts them in a central position to adapt erythropoiesis to infections and be prime actors of anemia of inflammation. Finally, the fact that in-vivo depletion of macrophages in mice can correct the erythropoietic symptoms of b-thalassemia and polycythemia vera further illustrates the key role of erythroblastic island macrophages in erythropathies and suggests that therapies targeting macrophages may have some benefit to treat these diseases. Acknowledgements None. Financial support and sponsorship J.-P.L. is supported by a Senior Research Fellowship (no. 1044091), R.N.J. by a Project Grant (no. 1046590) and A.C.P. by a Project Grant (1030143) from the National Health and Medical Research Council of Australia. Conflicts of interest The authors have no conflict of interest.
REFERENCES AND RECOMMENDED READING Papers of particular interest, published within the annual period of review, have been highlighted as: & of special interest && of outstanding interest 1. Bessis M. L’ilot erythroblastique, unite fonctionelle de la moelle osseuse. Rev Hematol 1958; 13:8–11. 2. Chasis JA, Mohandas N. Erythroblastic islands: niches for erythropoiesis. Blood 2008; 112:470–478. 3. Sawada K, Krantz SB, Dessypris EN, et al. Human colony-forming unitserythroid do not require accessory cells, but do require direct interaction with insulin-like growth factor I and/or insulin for erythroid development. J Clin Invest 1989; 83:1701–1709. 4. Manwani D, Bieker JJ. The erythroblastic island. Curr Top Dev Biol 2008; 82:23–53. 5. Bessis MC, Breton-Gorius J. Iron metabolism in the bone marrow as seen by electron microscopy: a critical review. Blood 1962; 19:635–663. 6. Leimberg MJ, Prus E, Konijn AM, et al. Macrophages function as a ferritin iron source for cultured human erythroid precursors. J Cell Biochem 2008; 103:1211–1218. 7. Skutelsky E, Danon D. On the expulsion of the erythroid nucleus and its phagocytosis. Anat Rec 1972; 173:123–126. 8. Soni S, Bala S, Gwynn B, et al. Absence of erythroblast macrophage protein (Emp) leads to failure of erythroblast nuclear extrusion. J Biol Chem 2006; 281:20181–20189. 9. Crocker P, Werb Z, Gordon S, et al. Ultrastructural localization of a macrophage-restricted sialic acid binding hemagglutinin, SER, in macrophagehematopoietic cell clusters. Blood 1990; 76:1131–1138. 10. Hanspal M, Hanspal JS. The association of erythroblasts with macrophages promotes erythroid proliferation and maturation: a 30-kD heparin-binding protein is involved in this contact. Blood 1994; 84:3494–3504.
Volume 22 Number 3 May 2015
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Macrophages and regulation of erythropoiesis Jacobsen et al. 11. Hanspal M, Smockova Y, Uong Q. Molecular identification and functional characterization of a novel protein that mediates the attachment of erythroblasts to macrophages. Blood 1998; 92:2940–2950. 12. Sadahira Y, Yoshino T, Monobe Y. Very late activation antigen 4-vascular cell adhesion molecule 1 interaction is involved in the formation of erythroblastic islands. J Exp Med 1995; 181:411–415. 13. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythroblastic island formation. Blood 2006; 108:2064–2071. 14. de Jong JP, Voerman JS, van der Sluijs-Gelling AJ, et al. A monoclonal antibody (ER-HR3) against murine macrophages. I. Ontogeny, distribution and enzyme histochemical characterization of ER-HR3-positive cells. Cell Tissue Res 1994; 275:567–576. 15. Sonoda Y, Sasaki K. Hepatic extramedullary hematopoiesis and macrophages in the adult mouse: histometrical and immunohistochemical studies. Cells Tissues Organs 2012; 196:555–564. 16. Jacobsen RN, Forristal CE, Raggatt LJ, et al. Mobilization with granulocyte & colony-stimulating factor blocks medullar erythropoiesis by depleting F4/ 80 þ VCAM1 þ CD169 þ ER-HR3 þ Ly6Gþ erythroid island macrophages in the mouse. Exp Hematol 2014; 42:547–561; e544. Identification of the phenotype of mouse erythroblastic island macrophages enabling their prospective sorting from mouse bone marrow and spleen by flow cytometry. In-vivo depletion of CD169þ macrophages or phagocytes causes loss of erythroblastic island macrophages in bone marrow and spleen and arrest of medullary and splenic erythropoiesis with loss of erythroblasts. The cytokine GCSF used to mobilize hematopoietic stem and progenitor cells into the blood for transplantation blocks medullary erythropoiesis by depleting erythroblastic island macrophages in the bone marrow but not in the spleen. 17. Chow A, Huggins M, Ahmed J, et al. CD169þ macrophages provide a niche && promoting erythropoiesis under homeostasis and stress. Nat Med 2013; 19:429–436. First demonstration that in-vivo depletion of CD169þ macrophages or phagocytes stops medullary erythropoiesis by depleting erythroblastic island macrophages that express CD169. Furthermore, CD169þ macrophage depletion delays erythropoietic rescue after myeloablation or challenge with PHZ. Phagocyte depletion reduces hematocrit and red blood cell counts in a mouse model of polycythemia vera driven by JAK2V617F mutation. 18. Ramos P, Casu C, Gardenghi S, et al. Macrophages support pathological && erythropoiesis in polycythemia vera and [b]-thalassemia. Nat Med 2013; 19:437–445. In-vivo phagocyte depletion delays erythropoietic rescue after phlebotomy in wildtype mice and reduces hematocrit and red blood cell counts in a mouse model of polycythemia vera driven by JAK2V617F mutation. Chronic phagocyte depletion corrects anemia in a mouse model of b-thalassemia. 19. Mullally A, Lane SW, Ball B, et al. Physiological Jak2V617F expression causes a lethal myeloproliferative neoplasm with differential effects on hematopoietic stem and progenitor cells. Cancer Cell 2010; 17:584–596. 20. Bader BL, Rayburn H, Crowley D, et al. Extensive vasculogenesis, angiogenesis, and organogenesis precede lethality in mice lacking all av integrins. Cell 1998; 95:507–519. 21. Wang Z, Vogel O, Kuhn G, et al. Decreased stability of erythroblastic islands in integrin b3-deficient mice. Physiol Rep 2013; 1:e00018. 22. Spring FA, Griffiths RE, Mankelow TJ, et al. Tetraspanins CD81 and CD82 facilitate a4b1-mediated adhesion of human erythroblasts to vascular cell adhesion molecule-1. PLoS One 2013; 8:e62654. 23. Scott LM, Priestley GV, Papayannopoulou T. Deletion of {a}4 integrins from adult hematopoietic cells reveals roles in homeostasis, regeneration, and homing. Mol Cell Biol 2003; 23:9349–9360. 24. Ulyanova T, Scott LM, Priestley GV, et al. VCAM-1 expression in adult hematopoietic and nonhematopoietic cells is controlled by tissue-inductive signals and reflects their developmental origin. Blood 2005; 106:86–94. 25. Ulyanova T, Jiang Y, Padilla S, et al. Combinatorial and distinct roles of a5 and a4 integrins in stress erythropoiesis in mice. Blood 2011; 117:975–985.
26. Ulyanova T, Padilla SM, Papayannopoulou T. Stage specific functional roles of integrins in erythropoiesis. Exp Hematol 2014; 42:404–409. Conditional deletion of the a4 integrin gene in erythroblasts impairs erythropoietic recovery following challenge with PHZ. Conditional deletion of the a5 integrin gene in erythroblasts has no effect on erythropoietic recovery. 27. Robier C, Amouzadeh-Ghadikolai O, Bregant C, et al. The anti-VLA-4 anti& body natalizumab induces erythroblastaemia in the majority of the treated patients with multiple sclerosis. Mult Scler 2014; 20:1269–1272. Ninety-three percent of multiple sclerosis patients treated with the antia4 integrin natalizumab have erythroblastemia. None of the multiple sclerosis patients treated with chronic b-interferon experienced erythroblastemia. 28. Papayannopoulou T, Nakamoto B. Peripheralization of hemopoietic progenitors in primates treated with anti-VLA4 integrin. Proc Natl Acad Sci USA 1993; 90:9374–9378. 29. Mao X, Shi X, Liu F, et al. Evaluation of erythroblast macrophage protein related to erythroblastic islands in patients with hematopoietic stem cell transplantation. Eur J Med Res 2013; 18:9. 30. Sui Z, Nowak RB, Bacconi A, et al. Tropomodulin3-null mice are embryonic && lethal with anemia due to impaired erythroid terminal differentiation in the fetal liver. Blood 2014; 123:758–767. Deletion of the tropomodullin gene is embryonic lethal from day 14.5 of gestation in mice with severe fetal liver anemia, defective erythroblastic island assembly, defective erythroblast maturation, and enucleation. 31. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by & central macrophages in erythroblastic islands. Blood 2014; 123:3963–3971. Engulfment of pyrenocytes by erythroblastic island macrophages depends on the tyrosine kinase receptor Mer expressed by erythroblastic island macrophages. Phosphatidyl serine flips from the inner to the outer lipid bilayer of pyrenocyte plasma membrane and binds to Mer with protein S acting as a bridge. This signals engulfment by the macrophage. 32. McGrath KE. Red cell island dances: switching hands. Blood 2014; 123:3847–3848. 33. Tang H, Chen S, Wang H, et al. TAM receptors and the regulation of erythropoiesis in mice. Haematologica 2009; 94:326–334. 34. Perkins AC, Sharpe AH, Orkin SH. Lethal [b]-thalassaemia in mice lacking the erythroid CACCC-transcription factor EKLF. Nature 1995; 375:318–322. 35. Tallack MR, Perkins AC. KLF1 directly coordinates almost all aspects of terminal erythroid differentiation. IUBMB Life 2010; 62:886–890. 36. Siatecka M, Bieker JJ. The multifunctional role of EKLF/KLF1 during erythropoiesis. Blood 2011; 118:2044–2054. 37. Tallack MR, Magor GW, Dartigues B, et al. Novel roles for KLF1 in erythropoiesis revealed by mRNA-seq. Genome Res 2012; 22:2385–2398. 38. Porcu S, Manchinu MF, Marongiu MF, et al. Klf1 affects DNase II-a expression in the central macrophage of a fetal liver erythroblastic island: a non-cell-autonomous role in definitive erythropoiesis. Mol Cell Biol 2011; 31:4144–4154. 39. Xue L, Galdass M, Gnanapragasam MN, et al. Extrinsic and intrinsic control by && EKLF (KLF1) within a specialized erythroid niche. Development 2014; 141:2245–2254. Cultures from a single progenitor cell derived from mouse embryoid bodies give rise to complete erythroblastic islands with erythroblasts rosetting around a central macrophage suggesting that during the primitive wave of erythropoiesis, erythroblasts, and erythroblastic island macrophages may emerge from a common progenitor. Furthermore, in the fetal liver of embryos containing a fluorescent reporter under the control of the Klf1 gene promoter, both erythroblasts and macrophages are labeled suggesting that the Klf1 promoter is active in both erythroblasts and EI macrophages. 40. Xue L, Chen X, Chang Y, et al. Regulatory elements of the EKLF gene that direct erythroid cell-specific expression during mammalian development. Blood 2004; 103:4078–4083. 41. Perkins A. Erythroid Kruppel like factor: from fishing expedition to gourmet meal. Int J Biochem Cell Biol 1999; 31:1175–1192. 42. Kawane K, Fukuyama H, Kondoh G, et al. Requirement of DNase II for definitive erythropoiesis in the mouse fetal liver. Science 2001; 292:1546–1549. &
1065-6251 Copyright ß 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.co-hematology.com
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
219
