CHAPTER FOUR

New Insights into the Mechanisms of Mammalian Erythroid Chromatin Condensation and Enucleation Peng Ji Department of Pathology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA E-mail: [email protected]

Contents 1. Introduction 2. Chromatin Condensation and Terminal Erythropoiesis 2.1 Recent Advances in Chromatin Condensation in Terminal Erythropoiesis 2.2 Caspases in Chromatin Condensation 3. Membrane and Cytoskeleton Changes during Terminal Erythropoiesis and Enucleation 3.1 Overview of the Building Components of Erythroid Membrane and Cytoskeleton 3.2 Cell Surface Makers Used to Isolate Cells of Different Developmental Stages 3.3 Cytoskeleton Proteins Required for Terminal Erythropoiesis and Enucleation 3.4 Microtubules and the Establishment of Nuclear Polarity Prior Enucleation 3.5 Vesicle Trafficking in Enucleation 4. Extracellular Environment in Enucleation 4.1 Macrophages and Erythroblastic Island in Enucleation 4.2 Integrins and Other Cell-Adhesion Molecules in Terminal Erythropoiesis and Enucleation 5. Concluding Remarks Acknowledgments References

160 162 162 163 164 164 166 168 171 172 173 173 175 176 177 178

Abstract A unique feature in mammalian erythropoiesis is the dramatic chromatin condensation followed by enucleation. This step-by-step process starts at the beginning of terminal erythropoiesis after the hematopoietic stem cells are committed to erythroid lineage. Although this phenomenon is known for decades, the mechanisms of chromatin condensation and enucleation remain elusive. Recent advances in cell and molecular biology have started to reveal the molecular pathways in the regulation of chromatin condensation, the establishment of nuclear polarity prior enucleation, and the International Review of Cell and Molecular Biology, Volume 316 ISSN 1937-6448 http://dx.doi.org/10.1016/bs.ircmb.2015.01.006

© 2015 Elsevier Inc. All rights reserved.

159

j

160

Peng Ji

rearrangement of actin cytoskeleton in enucleation. However, many challenging questions, especially whether and how the apoptotic mechanisms are involved in chromatin condensation and how to dissect the functions of many actin cytoskeleton proteins in cytokinesis and enucleation, remain to be answered. Here I review our current understanding of mammalian erythroid chromatin condensation and enucleation during terminal differentiation with a focus on more recent studies. I conclude with my perspective of future works in this rising topic in developmental and cell biology.

1. INTRODUCTION Mammalian erythropoiesis involves differentiation of hematopoietic stem cells to committed burst forming unitsderythroid (BFU-Es) followed by colony forming unitsderythroid (CFU-Es) (Lodish et al., 2010). Differentiation from CFU-Es to mature red blood cells, generally termed terminal erythropoiesis, is driven by multiple erythropoietin (Epo)-induced signaling transduction pathways. These pathways act individually or collectively to activate or repress genes that regulate cell differentiation and proliferation, and inhibit apoptosis (Broxmeyer, 2013; Bunn, 2013). During terminal erythropoiesis the erythroid nucleus gradually condenses. Among many distinctive features of erythroid cells from other somatic cells, nuclear and chromatin condensation and the following enucleation process, are the most unique cellular processes in mammals (Ji et al., 2011). Nuclear and chromatin condensation starts in the early stages of terminal erythropoiesis. In fact, the chromatin condensation status of the nucleus is one of the important features to define morphologically distinctive erythroblasts at different developmental stages. These morphologically recognizable erythroblasts include cells after the CFU-E stage of development. From early to late stages of terminal erythropoiesis, they are categorized into proerythroblasts, basophilic erythroblasts, polychromatic erythroblasts, and orthochromatin erythroblasts. These morphologically distinct erythroblasts correlate with their nuclear condensation status and are critical for the diagnosis of many erythroid-related diseases such as megaloblastic anemia and myelodysplastic syndromes (Kjeldsberg and Perkins, 2010). One of the important manifestations in these diseases is the asynchronous maturation of the cytoplasm and nucleus in which nuclear condensation lags behind the hemoglobin enrichment in the cytoplasm (Kjeldsberg, 2010). Proerythroblast undergoes four to five cell divisions accompanied with gradual chromatin condensation followed by enucleation. Chromatin condensation is pertinent to erythroid progenitors in all vertebrates, but enucleation is unique in mammalian erythroblast. Enucleation is believed

Erythroid Chromatin Condensation and Enucleation

161

to provide evolutional advantages in mammals in that mature erythroid cells gain extra spaces for hemoglobin and more flexibility to migrate through terminal capillaries. It was thought for a long period of time that enucleation only occurs in definitive erythroblasts, whereas primitive erythroid cells retain their nuclei in fetal circulation. An important advance in the field came with the discovery of intravascular enucleation of circulating primitive erythroid cells in mouse (Kingsley et al., 2004; McGrath et al., 2008). Thus, mammalian erythroid cells universally extrude their nuclei at different hematopoietic tissues at various developmental stages. Besides erythroid cells, mammalian lens epithelia and keratinocytes also lose their nuclei (Hanna et al., 1961; McCall and Cohen, 1991; Ji et al., 2011). The mechanisms of enucleation in these three cell types are distinct in that maturing lens epithelium and keratinocyte resemble more closely to apoptotic disassembly of their nuclei (Ishizaki et al., 1998; Weil et al., 1999). On the other hand, the highly condensed nucleus in the orthochromatic stage of terminal erythroblasts is expelled out of the cytoplasm instead of being disassembled in the cytoplasm. The extruded nuclei from terminal erythroblasts, so-called “pyrenocytes,” are quickly engulfed by macrophages in fetal liver and bone marrow (Yoshida et al., 2005; Toda et al., 2014). Therefore, pyrenocytes are not found in peripheral blood, bone marrow, or any other hematopoietic tissues in adults at steady state, nor they are present in any hematologic diseases. This suggests that defects in the clearance of the extruded nuclei could be detrimental to the body. Occasionally, in patients with bone-marrow-occupying diseases or myeloproliferative neoplasm, leukoerythroblastic picture of peripheral blood with rare circulating nucleated red blood cells can be seen (Swerdlow et al., 2008). However, there is no patient with congenital defects in enucleation, which further demonstrates the physiological significance of enucleation in mammals. Although these processes have been known for decades, the mechanisms of chromatin condensation and enucleation are unclear until more recently (Ji et al., 2011; Keerthivasan et al., 2011). Over the past decade, we have gained significant amount of knowledge in this expanding field of hematology. The advances include molecular pathways involved in actin cytoskeleton rearrangement during enucleation, establishment of nuclear polarity, and the involvement of macrophages in enucleation. These pathways are interconnected and highly regulated. Disruption of any of these processes will affect terminal erythropoiesis as a whole. However, this also generates a common dilemma in the study of the final stage of terminal erythropoiesis

162

Peng Ji

in which dissecting the functions of each player in these complex regulatory networks becomes challenging. In addition, several outstanding questions, such as the chromatin-dynamic changes in terminal erythropoiesis and the detailed mechanism of chromatin condensation, remain elusive. In this chapter, I start with an overview of the important historic work on terminal erythropoiesis focusing on chromatin condensation and enucleation. This is followed by detailed discussions of more recent advances in the field. I conclude this chapter with perspectives of future studies and theories that need to be validated.

2. CHROMATIN CONDENSATION AND TERMINAL ERYTHROPOIESIS 2.1 Recent Advances in Chromatin Condensation in Terminal Erythropoiesis During terminal erythropoiesis from CFU-Es to orthochromatic erythroblasts, the volume of erythroid nucleus steadily decreases more than 10-folds (Ji et al., 2008). This drop in volume is associated with significant expression changes of major histones (our unpublished data). Genetic study using chromatin immunoprecipitation coupled to next generation sequencing (ChIP-seq) shows multiple posttranslational modifications on histone tails, including methylation and acetylation (Wong et al., 2011). However, it is not clear how these changes relate to chromatin condensation, which could be important since treatment of mouse erythroblasts with histone deacetylase (HDAC) inhibitors blocks chromatin condensation and enucleation (Popova et al., 2009; Ji et al., 2010). HDAC2 is specifically involved in this process as its knockdown by shRNA recapitulates the HDAC inhibitor results (Ji et al., 2010). The role of histone deacetylation in chromatin condensation and enucleation is further strengthened by the evidence that ectopic expression of histone acetyltransferase Gcn5, which is normally up-regulated by c-Myc, partially blocks nuclear condensation and enucleation (Jayapal et al., 2010). The level of Gcn5 is also regulated indirectly by miR-191, a microRNA that is normally down-regulated during terminal erythropoiesis (Zhang et al., 2011, 2012). The targets of miR-191 in erythroblasts include Riok3 and Mxi1, two erythroid-enriched and developmentally up-regulated genes. Mxi1 is an antagonist of c-Myc (Casc on and Robledo, 2012). Therefore, it negatively regulates Gcn5 and enucleation. On the other hand, Riok3 was suggested to directly inhibit Gcn5. Consistently, overexpression of miR-191, or knockdown of Mxi1

Erythroid Chromatin Condensation and Enucleation

163

or Riok3 blocked nuclear condensation and enucleation (Zhang et al., 2011). Despite the evidence, whether histone tail acetylation is directly involved in the chromatin remodeling and condensation is unknown. Moreover, the role of histone methylation in chromatin condensation and enucleation has not been directly investigated. Interestingly, there is a global DNA demethylation during erythropoiesis (Shearstone et al., 2011). How this is linked to chromatin condensation is unclear.

2.2 Caspases in Chromatin Condensation In addition to physiological conditions such as erythropoiesis and the development of other hematopoietic lineages, chromatin condensation is also observed in apoptosis. Indeed, apoptotic mechanisms are known to play important functions in erythropoiesis (Testa, 2004; Li and Yuan, 2008). This is also true from mounting evidence in other somatic organ systems demonstrating the essential roles of caspases in development (Li and Yuan, 2008). Specifically in mammals, inhibition of the caspase activities blocks terminal erythropoiesis at the basophilic stage (Zermati et al., 2001). Further studies showed that shRNA knockdown of caspase 3 in human erythroid cells leads to significant reduction of enucleated cells with no change in the fraction of apoptotic cells (Carlile et al., 2004). However, the roles of caspases in erythropoiesis remain elusive since mice with caspase 1, 3, or 9 knockout show no significant defect in erythropoiesis (Li and Yuan, 2008). In addition, the observed chromatin condensation and enucleation phenotype could be the consequence of the blockage of erythroid differentiation from proerythroblasts to basophilic erythroblasts after down-regulation of caspase 3. Furthermore, controversial studies showed that treatment of mouse spleen erythroblasts with caspase inhibitors failed to block enucleation and no evidence of caspase-induced cleavage of target proteins was observed in the late stage erythropoiesis (Chasis et al., 1989). These results indicate that erythroid chromatin condensation and enucleation may not use apoptotic machineries in a way similar to what is being enacted in apoptosis. Caspases could indirectly be involved in the chromatin condensation process through unknown mechanisms. To understand this, it becomes necessary to analyze the subcellular localization of various histones and caspases to determine their temporospatial changes at different stages of terminal erythropoiesis, as well as their relationship to nuclear condensation. In this respect, a recent study using simultaneous microscopy and flow cytometry has demonstrated that the vast majority of histones migrate into the cytoplasm of normal erythroblasts at the final stage of

164

Peng Ji

terminal erythropoiesis, which is mediated by exportin 7 (Hattangadi et al., 2014). Notably, the histone migration is observed only after the cell exits the last cell cycle in the terminal stage of differentiation before enucleation. Whether the same histone migration into the cytoplasm occurs in the early stages, and how exportin 7 is regulated during different stages of terminal erythropoiesis would be of interest to investigate. Although remaining to be murky, the effects of caspases in terminal erythropoiesis could possibly be mediated through nuclear enzymes that directly regulate chromatin condensation. Caspase 3 is known to activate endonucleases during apoptosis (Liu et al., 1999; Li and Yuan, 2008). Indeed, endonucleases such as DNase IIa and acinus are activated in erythropoiesis (Kawane et al., 2001; Zermati et al., 2001). These endonucleases could therefore be directly involved in the gradual nuclear condensation process once caspase 3 is activated in the early stage of terminal erythropoiesis. Future studies focusing on how caspases regulate the activation of endonucleases, as well as how caspases are activated, would help clarify the definitive role of caspases in mammalian erythropoiesis.

3. MEMBRANE AND CYTOSKELETON CHANGES DURING TERMINAL ERYTHROPOIESIS AND ENUCLEATION 3.1 Overview of the Building Components of Erythroid Membrane and Cytoskeleton Mature mammalian erythroid cells contain three major membrane cytoskeleton elements: actin filaments, microtubules, and intermediate filaments. Each of these elements is involved in the establishment of the unique membrane cytoskeleton of erythroid membrane and cytoplasm. Directly beneath the plasma membrane, the heterodimeric a and b spectrins form a2b2 tetramers, which are cross-linked with a short oligmer of actin to form a pentagonal or hexagonal lattice. The spectrin network is connected to plasma membrane mainly through protein 4.1 and ankyrin in the spectrin network to glycophorin C and anion transporter (band-3) on plasma membrane, respectively (Gratzer, 1981; Byers and Branton, 1985). These erythroid-specific membrane skeleton proteins are critical for the stability and flexibility of mature red cells. Mutations of several of these proteins are frequently observed in patients with hereditary spherocytosis and elliptocytosis (Tse and Lux, 1999). Biosynthesis of these erythroid-specific

Erythroid Chromatin Condensation and Enucleation

165

proteins is tightly regulated temporospatially during mammalian erythropoiesis (Liu et al., 2011). The microtubule network in mammalian red blood cells is less clear. In general, microtubule levels are low in circulating red blood cells (Simpson and Kling, 1967). This is distinct from the nucleated red blood cells in other vertebrates in which a microtubule bundle named “marginal band” is present along the equatorial plane of the cell (Behnke, 1970). The size of erythrocytes in vertebrates (except mammals) positively correlates with the number of marginal band microtubules (Goniakowska-Witali nska and Witali nski, 1976). Although the same marginal band is observed in primitive erythroid cells in mammals (van Deurs and Behnke, 1973), the roles of microtubule in circulating red cells and mammalian terminal erythropoiesis remains elusive until more recently. Studies that I discuss in detail in the next section illustrate that microtubule plays a role in the establishment of nuclear polarity during the terminal stages of mammalian erythropoiesis before enucleation (Thom et al., 2014). In this respect, microtubule-dependent local activation of phosphoinositide 3-kinase could be important for the localization of the nucleus to one side of the plasma membrane (Wang et al., 2012), where the highly condensed nucleus is believed to be extruded out of the cytoplasm. In contrast to the spectrineactin network, the intermediate filaments between mammalian red blood cells and their nucleated counterparts in other vertebrates are grossly distinct. Vimentin comprises a majority of the intermediate filaments in vertebrate erythrocytes. In avian red blood cells, vimentin anchor the nucleus to the plasma membrane to help maintain the biconvex ellipsoidal shape of the cells (Granger et al., 1982). While vimentin’s level gradually increases during the differentiation of avian erythroblasts, the opposite is seen in the maturing mammalian erythroblasts. Although loss of vimentin may not be sufficient for enucleation as rearrangements of many other cytoskeleton proteins are clearly required, it could play an important role to prepare the cells for enucleation (Sangiorgi et al., 1990). The highly condensed nucleus of the late stage erythroblast needs to establish a nuclear polarity in which the pycnotic nucleus migrates to one side of the cytoplasm. Perhaps loss of vimentin liberates the nucleus in this setting. To demonstrate this, ectopic expression of vimentin in the late stage of mammalian terminal erythropoiesis can be attempted to determine if the nuclear polarization is affected. It would also be interesting to determine the phenotypes of avian nucleated red cells if their vimentin levels are down-regulated during terminal differentiation.

166

Peng Ji

3.2 Cell Surface Makers Used to Isolate Cells of Different Developmental Stages Membrane and cytoskeleton undergo dynamic changes during terminal differentiation and enucleation (Liu et al., 2011). Major erythroid-specific plasma membrane proteins are markedly increased, whereas adhesion molecules, such as a4 and b1 integrins and CD44, gradually decrease during terminal differentiation (Eshghi et al., 2007; Chen et al., 2009). Loss of adhesion molecules is important for the enucleated red blood cells to detach from bone marrow and enter circulation. These cell surface changes also make it possible to use membrane molecules as markers to differentiate erythroblasts at different developmental stages. In mouse, two well-established systems are widely used to isolate erythroid cells for the studies of erythropoiesis. The first system uses two cell surface markers, namely, transferrin receptor CD71 and glycoprotein-associated antigen Ter119 (Socolovsky et al., 1999; Zhang et al., 2003). In this system, Ter119-negative cells (mostly CFU-Es) were purified from fetal liver cells. When the cells are cultured in vitro, most of them express low level of CD71 at the beginning of culture. During the 2-day culture period in the presence of erythropoietin, the cells gradually gain surface CD71 that reaches the maximum level at approximately 24 h in culture. This is followed by a moderate decrease of CD71, whereas Ter119 level continuously increases throughout the terminal differentiation. With these two surface makers, cells can be gated into five different populations by flow cytometry. Figure 1(A) shows a flow cytometric analysis of total mouse fetal liver cells purified from embryonic day 13.5. In this plot, R1 represents proerythroblasts where most of the cells are Ter119 and CD71 negative. Cells in R2 and R3 are more differentiated and represent basophilic to polychromatic stage of terminal erythropoiesis. In fact, cells from R2 to R3 experience the most dramatic gene expression changes as evidenced by genome-wide RNA-sequencing analysis (Wong et al., 2011). Cells in R4 and R5 stages are mostly orthochromatic erythroblast. Enucleated reticulocytes can be found in R5. These different stages can be closely recapitulated in vitro in which proerythroblasts on day 0 differentiate through the basophilic and polychromatic stages on day 1, to orthochromatic erythroblasts and enucleated reticulocytes on day 2. Based on this system, a modified version was developed to quantitatively determine the enucleation efficiency by analyzing the percentage of DNA-stainingnegative reticulocytes (Figure 1(B)).

Erythroid Chromatin Condensation and Enucleation

167

Figure 1 Two broadly used flow cytometric systems for the analysis of mammalian terminal erythropoiesis. (A) Total fetal liver cells were analyzed by CD71 and Ter119. Populations from R1 to R5 represent cells from the early to late stages of terminal erythropoiesis. (B) Analysis of day 2 cultured mouse fetal liver cells using Hoechst 33342 and Ter119. The enucleated reticulocytes were quantified. (C) Ter119-positive total bone marrow cells were analyzed for CD44 expression in correlation with cell size (FSC).

The second system uses Ter119 and CD44 to differentiate the cells (Chen et al., 2009). CD44 is a cell surface adhesion molecule with a higher expression level in proerythroblast stage. It gradually decreases during differentiation and reaches the lowest level in the incipient reticulocyte. The same expression pattern is also shared by other adhesion molecules such as intercellular adhesion molecule 4 (ICAM-4), b1 integrin, and Lutheran (Chen et al., 2009), which could be significant physiologically to facilitate the release of enucleated red blood cells out of bone marrow. Using CD44 and Ter119, as well as forward scatter (FSC) to separate cells based on their size, the bone marrow Ter119-positive erythroid cells can be further divided into different populations as in Figure 1(C). Cells of CD44highFSChigh are

168

Peng Ji

mostly basophilic stage of erythroblasts. The enucleated reticulocytes can be identified in cells with CD44lowFSClow. The rapid proliferative nature of mouse erythroblast makes it an ideal system for in vitro studies of genes of interest in a fast turnaround time. However, this also causes one of the prominent drawbacks of the system in which synchronization of developmental stages becomes demanding, even with the help of surface markers discussed above. One strategy that we have recently developed to overcome this problem is to culture the purified Ter119-negative proerythroblasts in medium without erythropoietin but containing stem cell factor (SCF), which will maintain the progenitor stages of the proerythroblasts but provide time for the viral-transduced genes or shRNAs to express. The cells are cultured in SCF medium for up to 24 h, which is followed by routine culture in erythropoietin-containing medium. In this way, the functions of the gene of interest in the early stage of terminal erythropoiesis can be differentiated from those in the late stage. In addition, the purified Ter119-negative cells have extra time to synchronize mostly to the basophilic stage of differentiation in SCF medium (Zhao et al., 2014b).

3.3 Cytoskeleton Proteins Required for Terminal Erythropoiesis and Enucleation Given the unique membrane and cytoskeleton network in red blood cells, the question is whether these erythroid-specific membrane and cytoskeleton proteins are involved in the enucleation process. To this end, recent study using knockout mouse models have demonstrated that ankyrin and band 3 are not required for enucleation (Ji and Lodish, 2012). Our unpublished results also indicate that other erythroid-specific membrane and cytoskeleton proteins, such as protein 4.1 and b-adducin, are not involved in enucleation as well. Although these data provide clear evidence that embryonic depletion of erythroid-specific membrane and cytoskeleton proteins could be dispensable for enucleation, it is not clear whether acquired loss or mutations of these proteins would compromise terminal erythropoiesis and enucleation. In this setting, patients with mutations on red-cell-specific membrane proteins, who often develop congenital anemia, do not show circulating nucleated red blood cells (Tse and Lux, 1999). This strongly supports the conclusion that red-cell-specific membrane proteins are not essential for enucleation. In contrast to these erythroid-specific membrane and cytoskeleton proteins, actin is well known to play critical roles in terminal erythropoiesis,

Erythroid Chromatin Condensation and Enucleation

169

especially enucleation. Old experiment using actin inhibitor cytochalasin D demonstrated that block of actin polymerization completely inhibited erythroblast enucleation (Koury et al., 1989). This inhibitory effect is reversible when cytochalasin D is removed, indicating actin is involved in a dynamic process during enucleation. The significance of actin in enucleation is further consolidated by many recent mechanistic studies using various in vitro and in vivo models. Several of the recent important findings have demonstrated that Rac GTPases play essential roles in both early and late stages of terminal erythropoiesis (Ji et al., 2008; Konstantinidis et al., 2010, 2012). Inactivation of Rac GTPases using a specific inhibitor, NSC23766, in the early stage blocked cell differentiation and induces apoptosis, which is recapitulated using a red-cell-specific Rac1/2-doubleknockout mouse model (Kalfa et al., 2010; Konstantinidis et al., 2010). Treatment of erythroid cells with NSC23766 in the late stage of terminal erythropoiesis dramatically inhibited enucleation in a dose-dependent manner, which was similarly confirmed genetically using mouse model of Rac1, Rac2, and Rac3 triple deletion (Konstantinidis et al., 2012). Interestingly, double knockout of Rac1 and Rac2 does not seem to affect enucleation, suggesting Rac3’s compensatory role in the absence of Rac1 and Rac2 (Konstantinidis et al., 2012). Rac GTPases regulate enucleation, at least in part, through their downstream target mDia2 (Ji et al., 2008). mDia2 belongs to the mDia formin family proteins that are involved in the polymerization of linear actin filaments (Higgs, 2005; Faix and Grosse, 2006). In the late stage of terminal erythropoiesis, Rac GTPases induce the activation of mDia2, which mediates the formation of contractile actin ring (CAR) between the pycnotic nucleus and incipient reticulocyte. The role of mDia2 in the late stage of terminal erythropoiesis has been recently confirmed by mDia2-knockout mice. These mice show embryonic lethality at approximately embryonic day 12.5 with severe fetal anemia (Watanabe et al., 2013). As expected, fetal erythropoiesis, especially the late stage of terminal erythropoiesis, is significantly affected by the loss of mDia2. Interestingly, mDia2 appears to be essential for cytokinesis of late stage erythroblasts but not enucleation per se given the presence of large enucleated red blood cells when fetal liver erythroblasts from mDia2-null mice were cultured in vitro. This indicates that the originally observed enucleation defect after shRNA knockdown of mDia2 could be secondary to the cytokinesis abnormality in the late stage erythroblasts (Ji et al., 2008). However, it is also possible that enucleation could be directly affected by loss of mDia2 in vivo in fetal or adult

170

Peng Ji

erythropoiesis. Tissue-specific or inducible mDia2-knockout mouse model would be of great help to dissect the function of mDia2 in the late stage of terminal erythropoiesis, as well as in adult erythropoiesis at steady state or during stress. Besides Rac GTPases and mDia2, recent genetic and biochemical studies have demonstrated that other regulatory factors are also involved in CAR formation and enucleation. Similar to mDia2, some of these regulators, such as non-muscle myosin IIB (Ubukawa et al., 2012), are among those that are critical for cytokinesis. Non-muscle myosin IIB is well characterized to interact with actin and functions in cell migration, adhesion, and cytokinesis in other cell types (Vicente-Manzanares et al., 2009). Consistent with the view of enucleation as an asymmetrical cytokinesis, inhibition of myosin IIB blocks enucleation in mouse and human erythroblasts (Ubukawa et al., 2012; Wang et al., 2012). Similar to mDia2 too, one should consider whether the enucleation defect caused by inhibition of myosin IIB is direct or secondary to the compromised cytokinesis following mitosis in the late stage of terminal erythropoiesis. This is especially a concern in the 2-day mouse fetal liver culture system due to the fast dividing nature of the mouse erythroblasts as discussed before. In this respect, human erythroblasts require an extended culture time so that it is possible to temporally dissect the functions of genes of interests in different developmental stages. Indeed, treatment of blebbistatin, an inhibitor of myosin IIB, at day 11 of cultured human erythroblasts when the cells are ready to enucleate, completely blocked enucleation, which provides convincing evidence of myosin IIB’s direct role in enucleation (Ubukawa et al., 2012). Tropomodulin3 (Tmod3) is another actin remodeling protein that has been recently found to be part of the enucleation regulatory system (Sui et al., 2013). The tropomodulin family proteins, specifically Tmod1, are well known as part of the spectrineactin network (Moyer et al., 2010). Tmod1 capping of the pointed end regulates the length of the actin filaments. Red blood cells with loss of Tmod1 show features resembling hereditary elliptocytosis. These cells also show increased expression of Tmod3, indicating a compensatory role of Tmod3 to rescue the red cells with loss of Tmod1 (Sui et al., 2013). Tmod3-knockout mice are lethal due to embryonic anemia and defects in fetal erythropoiesis. Embryonic lethality occurs after embryonic day 14.5, which makes it possible to investigate fetal erythropoiesis in these mice. Specifically, late stage fetal liver erythroblasts from Tmod3-knockout mice show failure of cell cycle exit, multilobular nuclear morphology, and aberrant F-actin assembly during enucleation,

Erythroid Chromatin Condensation and Enucleation

171

suggesting Tmod3-regulated actin-dynamic changes are important for the late stage of terminal erythropoiesis. Similar to Rac GTPases, Tmod3 also functions in the early stage of erythropoiesis as evidenced from the reduced BFU-E and CFU-E colony formations from Tmod3-knockout mice. This suggests that the actin regulatory machinery plays a general role throughout mammalian erythropoiesis. As discussed above, this also generates a common dilemma to explicitly dissect the functions of the individual regulatory protein in different aspects of actin remodeling such as mitosis and contractile actin ring formation in enucleating cell. Like Rac GTPases, many cytoskeleton regulatory proteins that are indispensable for the late stage of terminal erythropoiesis and enucleation, also play important functions in the early stage of terminal erythropoiesis. Through a targeted shRNA screening, recent work from our group has discovered more than 30 genes that play novel functions both in the early and late stages of terminal erythropoiesis (Zhao et al., 2014a). Among these genes, pleckstrin-2 (plek2) is particularly interesting given its specific and high-level expression in erythroid cells. Plek2 is a ubiquitously expressed paralog of pleckstrin-1 (plek1) involved in actin-dynamics. It contains a central DEP (Disheveled, Egl-10, Pleckstrin) domain flanked by two PH (Pleckstrin Homology) domains (Hu et al., 1999). Plek2 is required for T-cell cytoskeleton reorganization (Bach et al., 2007), but its roles in other hematopoietic cells were unknown prior to our study. We found that plek2 is regulated by erythropoietin signaling during terminal erythropoiesis. Knockdown of plek2 in the early stage of terminal erythropoiesis induced a dramatic defect in cell differentiation and proliferation. On the other hand, plek2 is not required for cell differentiation in the late stage of terminal erythropoiesis but still critical for enucleation. Mechanistically, in the early stage of terminal erythropoiesis when the level of reactive oxygen species (ROS) is high, plek2 binds to cofilin and prevents its mitochondrial entry. This pro-survival function of plek2 is not required in the late stage since ROS is decreased. However, plek2 continues to be important for actin cytoskeleton and enucleation through interaction with cofilin (Zhao et al., 2014a).

3.4 Microtubules and the Establishment of Nuclear Polarity Prior Enucleation Immediately before enucleation, the nucleus migrates to one side of the cytoplasm to establish polarity. The polarized nucleus is believed to be extruded from the site where nucleus and cytoplasm membrane contact, although this has not been directly observed in real time. Several lines of

172

Peng Ji

evidence indicate that microtubules are required in the nuclear polarization process. The molecular mechanisms behind this begin to be elucidated. The first time that microtubule was observed to be functionally important for enucleation came from studies of rat erythroid cells where microtubule inhibitor colchicine blocked nuclear extrusion in vitro and in vivo (Chasis et al., 1989). Recent technique advances in fluorescence microscopy have made it possible to reveal the detailed microtubule structural and localization changes during enucleation. Before erythroblast undergoes enucleation, microtubules form a cage-like structure surrounding the nucleus. Once the orthochromatic erythroblast exits the last cell cycle, the nucleus migrates to one side of the cytoplasm to establish polarity. Microtubules radiated from gamma tubulin-rich centroid are enriched on the opposite side. Treatment of the orthochromatic erythroblasts with colchicine disrupts this structure, leading to a diffuse distribution pattern of microtubule (Wang et al., 2012). The nuclear polarization process starts before contractile actin ring formation since adding cytochalasin-D before polarization does not additively worsen the colchicine effect (Konstantinidis et al., 2012). The microtubules are also dynamic in the nuclear polarization process characterized by the comet structure that are not uncommon to find (Wang et al., 2012). A more recent study has indicated that loss of microtubule motor dynein, through Trim58-mediated ubiquitination and degradation, mediates microtubule reorganization and the nuclear polarization process. Consistently, knockdown of Trim58 significantly affected enucleation in mouse fetal liver erythroblasts. Microtubules are known to regulate the localized phosphatidylinositol3-kinase (PI3K) activity in migrating neutrophils (Xu et al., 2005) and phagocytic macrophages (Khandani et al., 2007). One plausible mechanism is that microtubule indirectly controls PI3K activity to regulate contractile actin ring formation and enucleation. This is confirmed by a diffuse location of PI3K products, phosphatidylinositol (3,4)-bisphosphate (PtdIns(3,4)P2) and phosphatidylinositol (3,4,5)-trisphosphate (PtdIns(3,4,5)P3), when late stage erythroblasts were treated with nocodazole, compared to their normal plasma membrane localization.

3.5 Vesicle Trafficking in Enucleation In the very end stage of terminal erythropoiesis when nuclear polarity is already established in postmitotic erythroblasts, endocytic vesicle trafficking and vacuole formation between the incipient reticulocyte and pyrenocyte are reported to be involved in enucleation. Inhibition of clathrin-dependent

Erythroid Chromatin Condensation and Enucleation

173

vesicle trafficking blocked enucleation of primary fetal liver erythroblasts without affecting normal differentiation or proliferation. On the other hand, induction of vesicle formation through vacuolin-1 increased the percentage of enucleated cells (Keerthivasan et al., 2010). Mechanistically, clathrin binds to survivin and EPS15 to facilitate the vesicle formation. Knockdown of survivin in human erythroblasts significantly diminished cytoplasmic vesicles and enucleation efficiency, which can be rescued by vacuolin-1 (Keerthivasan et al., 2012). It appears that clathrin, EPS15, and survivin are among proteins of a supernumerary complex. The other components of this complex are currently unknown but could be involved in vesicle formation or movement as well in the terminal stage of erythropoiesis. In addition, vesicle trafficking and fusion seem to facilitate the excision of nucleus from the incipient reticulocyte. Further studies are required to determine the signaling pathways that mediate the cross talk between the actin cytoskeleton regulatory machineries and clathrin/survivin-mediated vesicle trafficking. It is noted that lipid rafts are also present in this region between reticulocyte and pyrenocyte (Konstantinidis et al., 2012). Whether and how vesicle trafficking and lipid rafts are connected will also be valuable to explore.

4. EXTRACELLULAR ENVIRONMENT IN ENUCLEATION 4.1 Macrophages and Erythroblastic Island in Enucleation It is well established that the extracellular environment is crucial for mammalian erythroblast enucleation. The first hint came from the close association of maturing erythroblasts with macrophages, which forms so-called “erythroblastic island” in hematopoietic tissues such as bone marrow, and possibly fetal liver and spleen (Chasis and Mohandas, 2008). The island is characterized by a central macrophage surrounded by a ring of maturing erythroblasts. Although the widespread presence of the erythroblastic island in the hematologic tissues still requires comprehensive histological investigation, convincing evidence from in vitro studies demonstrated that macrophages are actively involved in erythropoiesis including enucleation. The bone marrow macrophages specifically express CD169, which has been shown to be required for the retention of hematopoietic stem and progenitor cells in the mesenchymal stem cell niche (Chow et al., 2011). More recently, two studies have demonstrated that bone

174

Peng Ji

marrow macrophages are less important for steady-state erythropoiesis, but are particularly important during stress erythropoiesis (Chow et al., 2013; Ramos et al., 2013). Mice with macrophage depletion by clodronate show decreased recovery of phlebotomy or phenylhydrazine-induced anemia. Furthermore, macrophage depletion also alleviates the disease phenotypes in polycythemia vera and beta-thalassemia by modulating erythroid proliferation and differentiation (Chow et al., 2013; Ramos et al., 2013). These results revealed the overall functions of macrophages in erythropoiesis in vivo, which are long awaited since molecular studies of proteins that are involved in macrophageeerythroblast interaction have started more than a decade ago. Research in the 1990s revealed that macrophages interact with erythroblasts through erythroblast macrophage protein (Emp), which has a relatively high expression level on the surface of macrophages and differentiating erythroblasts (Hanspal and Hanspal, 1994). In vitro evidence showed that differentiation of both macrophages and erythroblasts requires Emp (Hanspal and Hanspal, 1994; Soni et al., 2007). This was confirmed by a more recent in vivo study using Emp-knockout mice. These mice die perinatally with increased nucleated red blood cells in fetal circulation and absence of erythroblastic island in fetal liver due to the substantially reduced macrophages (Soni et al., 2006). The defect in enucleation could also be intrinsic since in vitro culture of Emp-deficient erythroblasts with wild-type macrophages failed to enucleate (Soni et al., 2006). In this respect, Emp is indicated to regulate F-actin distribution to influence erythroblasts enucleation and macrophage differentiation. Nevertheless, detailed mechanistic studies are required to fully reveal the functions of Emp in erythroblastic island in vitro and in vivo, and how Emp cross talk with other signaling pathways such as Rac GTPases and mDia2 in regulating enucleation. Similar to Emp, the tumor suppressor retinoblastoma (Rb) protein also plays a cell intrinsic role in the differentiation of macrophages and erythroblast enucleation (Iavarone et al., 2004). Loss of Rb in mouse leads to embryonic death due to failure of erythroblast enucleation, primarily due to the disrupted macrophage and erythroblast interaction in erythroblastic island. This can be rescued by compound knockout of Id2, an inhibitor of transcription factor PU.1 that promotes macrophage differentiation (Iavarone et al., 2004). Rb also plays a role in the early stage of terminal erythropoiesis and mitochondria biogenesis (Sankaran et al., 2008). Given the general function of Rb in cell cycle control in different cell types

Erythroid Chromatin Condensation and Enucleation

175

(Korenjak and Brehm, 2005), future studies are needed to further pinpoint the specific molecular mechanism of Rb in mammalian erythropoiesis, especially enucleation.

4.2 Integrins and Other Cell-Adhesion Molecules in Terminal Erythropoiesis and Enucleation Integrins are actively involved in cellecell and cellemicroenvironment interactions. In erythroid cells, a4, a5, and b1 are the principal integrins expressed on the progenitors. Same as CD44, these integrins are gradually down-regulated during terminal erythropoiesis. This down-regulation parallels the loss of cell adhesion to fibronectin, which is one of the key components of erythroid microenvironment in bone marrow and fetal liver (Tada et al., 2006). a4b1 integrins are believed to provide survival cues to erythroblasts in the late stage of terminal erythropoiesis when erythropoietin signal diminishes (Eshghi et al., 2007). Besides fibronectin, a4b1 integrins also bind to vascular cell adhesion molecule 1 (VCAM-1) on macrophages. Monoclonal antibodies against a4 integrin and VCAM-1 blocked erythroblastic island formation (Sadahira et al., 1995). In addition to VCAM-1, av integrin subunit CD51 is also detected on the cell surface of macrophage. Interaction of av integrin with ICAM-4 expressed on the erythroblasts is critical for the erythroblastic island integrity (Lee et al., 2006). ICAM-4knockout mice show decreased erythroblastic island formation in vitro and in vivo, confirming the important role of macrophageeerythroblasts interaction. In addition, ICAM-4 was reported to interact with DLC-1 (deleted in liver cancer1) that is also expressed on erythroblasts. In vitro this interaction could promote differentiation and enucleation, but its significance in vivo is currently unclear (Choi et al., 2013). Loss of a4b1 integrins in the final stage of terminal erythropoiesis facilitates the release of incipient reticulocytes from bone marrow to circulation. The maturing erythroblasts lose surface integrins through two different approaches. The first approach is the gradual transcriptional downregulation. In fact, a majority of genes share this approach through terminal erythropoiesis to be down-regulated (Wong et al., 2011). In the second approach, the enucleating erythroblasts sort plasma membrane proteins between reticulocytes and pycnotic nuclei. Integrins were found to be sorted on both reticulocytes and the extruded nuclei, thus providing a fast way for the reticulocyte to shed unnecessary integrins. As introduced above, macrophages also engulf the extruded nuclei (pyrenocytes), as well as aged or senescence red blood cells to recycle iron. To engulf the nuclei,

176

Peng Ji

macrophage recognizes phosphatidylserine, an “eat me” signal that is also present in apoptotic cells, on the surface of pyrenocytes. Block of phosphatidylserine on the pyrenocytes prohibits their macrophage engulfment (Yoshida et al., 2005). On macrophage, MerTK, a receptor kinase expressed on the surface of macrophage, binds to phosphatidylserine through serum factor protein S. Macrophage from MerTK-knockout mice show decreased engulfment of pyrenocytes in an in vitro model of erythroblastic island (Toda et al., 2014). Since Tmod3 is also expressed in macrophage, its role in erythroid island formation was investigated. As expected, macrophage with loss of Tmod3 fails to form the island with wild-type erythroblasts, which indicates that Tmod3 is also involved in macrophageeerythroblast communication (Sui et al., 2013).

5. CONCLUDING REMARKS Our understanding of chromatin condensation and enucleation in mammalian terminal erythropoiesis has been significantly improved in the past decade. Among the progresses, it becomes increasingly clearer how actin and microtubule networks coordinate in regulating and preparing the orthochromatic erythroblast to exit cell cycle, establish nuclear polarity, and extrude its pycnotic nucleus (Figure 2). However, many aspects of this unique biology phenomenon remain to be explored. Among the unknown areas of terminal erythropoiesis, chromatin condensation deserves a special attention. The contradictory theory of caspases in terminal erythropoiesis and chromatin condensation needs to be settled. Current evidence leans toward the functional significance of caspases in different stages of terminal erythropoiesis, but this function of caspases is completely unknown. Given the fundamental difference of the nuclear condensation during apoptosis and terminal erythropoiesis, it is unlikely that caspases enact their destructive capabilities in the condensing nucleus of erythroblasts. However, terminal erythroblasts indeed evolve strategies to protect themselves from caspase cleavage. For example, GATA1 is protected by Hsp-70 from caspase-3-mediated cleavage (Ribeil et al., 2007). Defective Hsp-70 localization in erythroblasts can lead to GATA1 cleavage in myelodysplastic syndromes (Frisan et al., 2012) and beta-thalassemia (Arlet et al., 2014). These studies reveal the role of heat shock protein Hsp-70 in protecting major transcription factor GATA1 in terminal erythropoiesis. Hsp-70 could also protect other major regulatory

Erythroid Chromatin Condensation and Enucleation

177

Figure 2 Chromatin condensation and enucleation require multiple regulatory pathways. During mammalian terminal erythropoiesis, the nucleus undergoes gradual condensation. Caspases are known to be involved in this process but the mechanism is unclear. In addition, chromatin condensation also involves histone deacetylation regulated by HDACs and Gcn5, whose levels are further controlled by miR-191 and its downstream targets. After the cell exits the last mitosis, the highly condensed nucleus migrates to one side of the cytoplasm to establish polarity, which is mediated by microtubules regulated by Trim58 and dynein. Recent studies have also broadened our understanding of actin cytoskeleton in enucleation, which involves Rac GTPases, mDia2, myosin IIB, Plek2, and Tmod3. Vesicle trafficking mediated by survivin complex is believed to promote the final cleavage of incipient reticulocyte and pyrenocyte.

proteins that are yet to be identified. Nevertheless, the outstanding question is why caspase 3 needs to be activated during normal differentiation. This seems to be a common question in development since non-apoptotic activation of caspases is also observed in other organ systems during development (Li and Yuan, 2008). A common role of caspases in development could be shared among various organ systems, from which we can adopt in terminal erythropoiesis. Caspases could also play a unique role in erythropoiesis to cleave selected substrates yet to be discovered.

ACKNOWLEDGMENTS I thank Dr Baobing Zhao for the help with Figure 1. I thank members in my laboratory, Drs. Baobing Zhao and Yang Mei, for helpful discussions. I acknowledge the support by American Society for Hematology scholar award, NIH pathway to independence award (R00HL102154), and National Cancer Institute grant (U54CA143869).

178

Peng Ji

REFERENCES Arlet, J.-B., Ribeil, J.-A., Guillem, F., Negre, O., Hazoume, A., Marcion, G., Beuzard, Y., Dussiot, M., Moura, I.C., Demarest, S., de Beauchêne, I.C., Belaid-Choucair, Z., Sevin, M., Maciel, T.T., Auclair, C., Leboulch, P., Chretien, S., Tchertanov, L., Baudin-Creuza, V., Seigneuric, R., Fontenay, M., Garrido, C., Hermine, O., Courtois, G., 2014. HSP70 sequestration by free a-globin promotes ineffective erythropoiesis in b-thalassaemia. Nature 514, 242e246. Bach, T.L., Kerr, W.T., Wang, Y., Bauman, E.M., Kine, P., Whiteman, E.L., Morgan, R.S., Williamson, E.K., Ostap, E.M., Burkhardt, J.K., Koretzky, G.A., Birnbaum, M.J., Abrams, C.S., 2007. PI3K regulates pleckstrin-2 in T-cell cytoskeletal reorganization. Blood 109, 1147e1155. Behnke, O., 1970. A comparative study of microtubules of disk-shaped blood cells. J. Ultrastruct. Res. 31, 61e75. Broxmeyer, H.E., 2013. Erythropoietin: multiple targets, actions, and modifying influences for biological and clinical consideration. J. Exp. Med. 210, 205e208. Bunn, H.F., 2013. Erythropoietin. Cold Spring Harb. Perspect. Med. 3, a011619. Byers, T.J., Branton, D., 1985. Visualization of the protein associations in the erythrocyte membrane skeleton. Proc. Natl. Acad. Sci. USA 82, 6153e6157. Carlile, G.W., Smith, D.H., Wiedmann, M., 2004. Caspase-3 has a nonapoptotic function in erythroid maturation. Blood 103, 4310e4316. Casc on, A., Robledo, M., 2012. MAX and MYC: a heritable breakup. Cancer Res. 72, 3119e3124. Chasis, J.A., Mohandas, N., 2008. Erythroblastic islands: niches for erythropoiesis. Blood 112, 470e478. Chasis, J.A., Prenant, M., Leung, A., Mohandas, N., 1989. Membrane assembly and remodeling during reticulocyte maturation. Blood 74, 1112e1120. Chen, K., Liu, J., Heck, S., Chasis, J.A., An, X., Mohandas, N., 2009. Resolving the distinct stages in erythroid differentiation based on dynamic changes in membrane protein expression during erythropoiesis. Proc. Natl. Acad. Sci. USA 106, 17413e17418. Choi, H.S., Lee, E.M., Kim, H.O., Park, M.-I., Baek, E.J., 2013. Autonomous control of terminal erythropoiesis via physical interactions among erythroid cells. Stem Cell Res. 10, 442e453. Chow, A., Huggins, M., Ahmed, J., Hashimoto, D., Lucas, D., Kunisaki, Y., Pinho, S., Leboeuf, M., Noizat, C., Van Rooijen, N., Tanaka, M., Zhao, Z.J., Bergman, A., Merad, M., Frenette, P.S., 2013. CD169⁺ macrophages provide a niche promoting erythropoiesis under homeostasis and stress. Nat. Med. 19, 429e436. Chow, A., Lucas, D., Hidalgo, A., Méndez-Ferrer, S., Hashimoto, D., Scheiermann, C., Battista, M., Leboeuf, M., Prophete, C., Van Rooijen, N., Tanaka, M., Merad, M., Frenette, P.S., 2011. Bone marrow CD169þ macrophages promote the retention of hematopoietic stem and progenitor cells in the mesenchymal stem cell niche. J. Exp. Med. 208, 261e271. van Deurs, B., Behnke, O., 1973. The microtubule marginal band of mammalian red blood cells. Z. Anat. Entwicklungsgesch 143, 43e47. Eshghi, S., Vogelezang, M.G., Hynes, R.O., Griffith, L.G., Lodish, H.F., 2007. Alpha4beta1 integrin and erythropoietin mediate temporally distinct steps in erythropoiesis: integrins in red cell development. J. Cell Biol. 177, 871e880. Faix, J., Grosse, R., 2006. Staying in shape with formins. Dev. Cell 10, 693e706. Frisan, E., Vandekerckhove, J., de Thonel, A., Pierre-Eugene, C., Sternberg, A., Arlet, J.-B., Floquet, C., Gyan, E., Kosmider, O., Dreyfus, F., Gabet, A.-S., Courtois, G., Vyas, P., Ribeil, J.-A., Zermati, Y., Lacombe, C., Mayeux, P., Solary, E., Garrido, C., Hermine, O., Fontenay, M., 2012. Defective nuclear localization of Hsp70 is associated

Erythroid Chromatin Condensation and Enucleation

179

with dyserythropoiesis and GATA-1 cleavage in myelodysplastic syndromes. Blood 119, 1532e1542. Goniakowska-Witali nska, L., Witali nski, W., 1976. Evidence for a correlation between the number of marginal band microtubules and the size of vertebrate erythrocytes. J. Cell Sci. 22, 397e401. Granger, B.L., Repasky, E.A., Lazarides, E., 1982. Synemin and vimentin are components of intermediate filaments in avian erythrocytes. J. Cell Biol. 92, 299e312. Gratzer, W.B., 1981. The red cell membrane and its cytoskeleton. Biochem. J 198, 1e8. Hanna, C., Bicknell, D.S., O’Brien, J.E., 1961. Cell turnover in the adult human eye. Arch. Ophthalmol. 65, 695e698. Hanspal, M., Hanspal, J.S., 1994. The association of erythroblasts with macrophages promotes erythroid proliferation and maturation: a 30-kD heparin-binding protein is involved in this contact. Blood 84, 3494e3504. Hattangadi, S.M., Martinez-Morilla, S., Patterson, H.C., Shi, J., Burke, K., AvilaFigueroa, A., Venkatesan, S., Wang, J., Paulsen, K., Gorlich, D., Murata-Hori, M., Lodish, H.F., 2014. Histones to the cytosol: exportin 7 is essential for normal terminal erythroid nuclear maturation. Blood 124, 1931e1940. Higgs, H.N., 2005. Formin proteins: a domain-based approach. Trends Biochem. Sci. 30, 342e353. Hu, M.H., Bauman, E.M., Roll, R.L., Yeilding, N., Abrams, C.S., 1999. Pleckstrin 2, a widely expressed paralog of pleckstrin involved in actin rearrangement. J. Biol. Chem. 274, 21515e21518. Iavarone, A., King, E.R., Dai, X.-M., Leone, G., Stanley, E.R., Lasorella, A., 2004. Retinoblastoma promotes definitive erythropoiesis by repressing Id2 in fetal liver macrophages. Nature 432, 1040e1045. Ishizaki, Y., Jacobson, M.D., Raff, M.C., 1998. A role for caspases in lens fiber differentiation. J. Cell Biol. 140, 153e158. Jayapal, S.R., Lee, K.L., Ji, P., Kaldis, P., Lim, B., Lodish, H.F., 2010. Down-regulation of Myc is essential for terminal erythroid maturation. J. Biol. Chem. 285, 40252e40265. Ji, P., Jayapal, S.R., Lodish, H.F., 2008. Enucleation of cultured mouse fetal erythroblasts requires Rac GTPases and mDia2. Nat. Cell Biol. 10, 314e321. Ji, P., Lodish, H.F., 2012. Ankyrin and band 3 differentially affect expression of membrane glycoproteins but are not required for erythroblast enucleation. Biochem. Biophys. Res. Commun. 417, 1188e1192. Ji, P., Murata-Hori, M., Lodish, H.F., 2011. Formation of mammalian erythrocytes: chromatin condensation and enucleation. Trends Cell Biol. 21, 409e415. Ji, P., Yeh, V., Ramirez, T., Murata-Hori, M., Lodish, H.F., 2010. Histone deacetylase 2 is required for chromatin condensation and subsequent enucleation of cultured mouse fetal erythroblasts. Haematologica 95, 2013e2021. Kalfa, T.A., Pushkaran, S., Zhang, X., Johnson, J.F., Pan, D., Daria, D., Geiger, H., Cancelas, J.A., Williams, D.A., Zheng, Y., 2010. Rac1 and Rac2 GTPases are necessary for early erythropoietic expansion in the bone marrow but not in the spleen. Haematologica 95, 27e35. Kawane, K., Fukuyama, H., Kondoh, G., Takeda, J., Ohsawa, Y., Uchiyama, Y., Nagata, S., 2001. Requirement of DNase II for definitive erythropoiesis in the mouse fetal liver. Science 292, 1546e1549. Keerthivasan, G., Liu, H., Gump, J.M., Dowdy, S.F., Wickrema, A., Crispino, J.D., 2012. A novel role for survivin in erythroblast enucleation. Haematologica 97, 1471e1479. Keerthivasan, G., Small, S., Liu, H., Wickrema, A., Crispino, J.D., 2010. Vesicle trafficking plays a novel role in erythroblast enucleation. Blood 116, 3331e3340. Keerthivasan, G., Wickrema, A., Crispino, J.D., 2011. Erythroblast enucleation. Stem Cells Int. 2011, 1e9.

180

Peng Ji

Khandani, A., Eng, E., Jongstra-Bilen, J., Schreiber, A.D., Douda, D., SamavarchiTehrani, P., Harrison, R.E., 2007. Microtubules regulate PI-3K activity and recruitment to the phagocytic cup during Fcgamma receptor-mediated phagocytosis in nonelicited macrophages. J. Leukocyte Biol. 82, 417e428. Kingsley, P.D., Malik, J., Fantauzzo, K.A., Palis, J., 2004. Yolk sac-derived primitive erythroblasts enucleate during mammalian embryogenesis. Blood 104, 19e25. Kjeldsberg, C.R., Perkins, S.L., 2010. Practical Diagnosis of Hematologic Disorders. Kjeldsberg, S.L.P.C.R., 2010. Practical Diagnosis of Hematologic Disorders, fifth ed. American Society for Clinical Pathology. Konstantinidis, D.G., George, A., Kalfa, T.A., 2010. Rac GTPases in erythroid biology. Transfus. Clin. Biol. 17, 126e130. Konstantinidis, D.G., Pushkaran, S., Johnson, J.F., Cancelas, J.A., Manganaris, S., Harris, C.E., Williams, D.A., Zheng, Y., Kalfa, T.A., 2012. Signaling and cytoskeletal requirements in erythroblast enucleation. Blood 119, 6118e6127. Korenjak, M., Brehm, A., 2005. E2F-Rb complexes regulating transcription of genes important for differentiation and development. Curr. Opin. Genet. Dev. 15, 520e527. Koury, S.T., Koury, M.J., Bondurant, M.C., 1989. Cytoskeletal distribution and function during the maturation and enucleation of mammalian erythroblasts. J. Cell Biol. 109, 3005e3013. Lee, G., Lo, A., Short, S.A., Mankelow, T.J., Spring, F., Parsons, S.F., Yazdanbakhsh, K., Mohandas, N., Anstee, D.J., Chasis, J.A., 2006. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythroblastic island formation. Blood 108, 2064e2071. Li, J., Yuan, J., 2008. Caspases in apoptosis and beyond. Oncogene 27, 6194e6206. Liu, J., Mohandas, N., An, X., 2011. Membrane assembly during erythropoiesis. Curr. Opin. Hematol. 18, 133e138. Liu, X., Zou, H., Widlak, P., Garrard, W., Wang, X., 1999. Activation of the apoptotic endonuclease DFF40 (caspase-activated DNase or nuclease). Oligomerization and direct interaction with histone H1. J. Biol. Chem. 274, 13836e13840. Lodish, H., Flygare, J., Chou, S., 2010. From stem cell to erythroblast: regulation of red cell production at multiple levels by multiple hormones. IUBMB Life 62, 492e496. McCall, C.A., Cohen, J.J., 1991. Programmed cell death in terminally differentiating keratinocytes: role of endogenous endonuclease. J. Invest. Dermatol. 97, 111e114. McGrath, K.E., Kingsley, P.D., Koniski, A.D., Porter, R.L., Bushnell, T.P., Palis, J., 2008. Enucleation of primitive erythroid cells generates a transient population of “pyrenocytes” in the mammalian fetus. Blood 111, 2409e2417. Moyer, J.D., Nowak, R.B., Kim, N.E., Larkin, S.K., Peters, L.L., Hartwig, J., Kuypers, F.A., Fowler, V.M., 2010. Tropomodulin 1-null mice have a mild spherocytic elliptocytosis with appearance of tropomodulin 3 in red blood cells and disruption of the membrane skeleton. Blood 116, 2590e2599. Popova, E.Y., Krauss, S.W., Short, S.A., Lee, G., Villalobos, J., Etzell, J., Koury, M.J., Ney, P.A., Chasis, J.A., Grigoryev, S.A., 2009. Chromatin condensation in terminally differentiating mouse erythroblasts does not involve special architectural proteins but depends on histone deacetylation. Chromosome Res. 17, 47e64. Ramos, P., Casu, C., Gardenghi, S., Breda, L., Crielaard, B.J., Guy, E., Marongiu, M.F., Gupta, R., Levine, R.L., Abdel-Wahab, O., Ebert, B.L., Van Rooijen, N., Ghaffari, S., Grady, R.W., Giardina, P.J., Rivella, S., 2013. Macrophages support pathological erythropoiesis in polycythemia vera and b-thalassemia. Nat. Med. 19, 437e445. Ribeil, J.-A., Zermati, Y., Vandekerckhove, J., Cathelin, S., Kersual, J., Dussiot, M., Coulon, S., Moura, I.C., Zeuner, A., Kirkegaard-Sørensen, T., Varet, B., Solary, E., Garrido, C., Hermine, O., 2007. Hsp70 regulates erythropoiesis by preventing caspase-3-mediated cleavage of GATA-1. Nature 445, 102e105.

Erythroid Chromatin Condensation and Enucleation

181

Sadahira, Y., Yoshino, T., Monobe, Y., 1995. Very late activation antigen 4-vascular cell adhesion molecule 1 interaction is involved in the formation of erythroblastic islands. J. Exp. Med. 181, 411e415. Sangiorgi, F., Woods, C.M., Lazarides, E., 1990. Vimentin downregulation is an inherent feature of murine erythropoiesis and occurs independently of lineage. Development 110, 85e96. Sankaran, V.G., Orkin, S.H., Walkley, C.R., 2008. Rb intrinsically promotes erythropoiesis by coupling cell cycle exit with mitochondrial biogenesis. Genes Dev. 22, 463e475. Shearstone, J.R., Pop, R., Bock, C., Boyle, P., Meissner, A., Socolovsky, M., 2011. Global DNA demethylation during mouse erythropoiesis in vivo. Science 334, 799e802. Simpson, C.F., Kling, J.M., 1967. The mechanism of denucleation in circulating erythroblasts. J. Cell Biol. 35, 237e245. Socolovsky, M., Fallon, A.E., Wang, S., Brugnara, C., Lodish, H.F., 1999. Fetal anemia and apoptosis of red cell progenitors in Stat5a-/-5b-/- mice: a direct role for Stat5 in Bcl-X(L) induction. Cell 98, 181e191. Soni, S., Bala, S., Gwynn, B., Sahr, K.E., Peters, L.L., Hanspal, M., 2006. Absence of erythroblast macrophage protein (Emp) leads to failure of erythroblast nuclear extrusion. J. Biol. Chem. 281, 20181e20189. Soni, S., Bala, S., Kumar, A., Hanspal, M., 2007. Changing pattern of the subcellular distribution of erythroblast macrophage protein (Emp) during macrophage differentiation. Blood Cells Mol. Dis. 38, 25e31. Sui, Z., Nowak, R.B., Bacconi, A., Kim, N.E., Liu, H., Li, J., Wickrema, A., An, X.-L., Fowler, V.M., 2013. Tropomodulin3-null mice are embryonic lethal with anemia due to impaired erythroid terminal differentiation in the fetal liver. Blood 123, 758e767. Swerdlow, S.H., Cancer, I.A.F.R.O., World Health Organization, 2008. WHO Classification of Tumours of Haematopoietic and Lymphoid Tissues. World Health Organization. Tada, T., Widayati, D.T., Fukuta, K., 2006. Morphological study of the transition of haematopoietic sites in the developing mouse during the peri-natal period. Anat. Histol. Embryol. 35, 235e240. Testa, U., 2004. Apoptotic mechanisms in the control of erythropoiesis. Leukemia 18, 1176e1199. Thom, C.S., Traxler, E.A., Khandros, E., Nickas, J.M., Zhou, O.Y., Lazarus, J.E., Silva, A.P.G., Prabhu, D., Yao, Y., Aribeana, C., Fuchs, S.Y., Mackay, J.P., Holzbaur, E.L.F., Weiss, M.J., 2014. Trim58 Degrades dynein and regulates terminal erythropoiesis. Dev. Cell 30, 688e700. Toda, S., Segawa, K., Nagata, S., 2014. MerTK-mediated engulfment of pyrenocytes by central macrophages in erythroblastic islands. Blood 123, 3963e3971. Tse, W.T., Lux, S.E., 1999. Red blood cell membrane disorders. Br. J. Haematol. 104, 2e13. Ubukawa, K., Guo, Y.-M., Takahashi, M., Hirokawa, M., Michishita, Y., Nara, M., Tagawa, H., Takahashi, N., Komatsuda, A., Nunomura, W., Takakuwa, Y., Sawada, K., 2012. Enucleation of human erythroblasts involves non-muscle myosin IIB. Blood 119, 1036e1044. Vicente-Manzanares, M., Ma, X., Adelstein, R.S., Horwitz, A.R., 2009. Non-muscle myosin II takes centre stage in cell adhesion and migration. Nat. Rev. Mol. Cell. Biol. 10, 778e790. Wang, J., Ramirez, T., Ji, P., Jayapal, S.R., Lodish, H.F., Murata-Hori, M., 2012. Mammalian erythroblast enucleation requires PI3K-dependent cell polarization. J. Cell Sci. 125, 340e349. Watanabe, S., De Zan, T., Ishizaki, T., Yasuda, S., Kamijo, H., Yamada, D., Aoki, T., Kiyonari, H., Kaneko, H., Shimizu, R., Yamamoto, M., Goshima, G., Narumiya, S., 2013. Loss of a Rho-regulated actin nucleator, mDia2, impairs cytokinesis during mouse fetal erythropoiesis. Cell Rep. 5, 926e932.

182

Peng Ji

Weil, M., Raff, M.C., Braga, V.M., 1999. Caspase activation in the terminal differentiation of human epidermal keratinocytes. Curr. Biol. 9, 361e364. Wong, P., Hattangadi, S.M., Cheng, A.W., Frampton, G.M., Young, R.A., Lodish, H.F., 2011. Gene induction and repression during terminal erythropoiesis are mediated by distinct epigenetic changes. Blood 118, e128e38. Xu, J., Wang, F., Van Keymeulen, A., Rentel, M., Bourne, H.R., 2005. Neutrophil microtubules suppress polarity and enhance directional migration. Proc. Natl. Acad. Sci. USA 102, 6884e6889. Yoshida, H., Kawane, K., Koike, M., Mori, Y., Uchiyama, Y., Nagata, S., 2005. Phosphatidylserine-dependent engulfment by macrophages of nuclei from erythroid precursor cells. Nature 437, 754e758. Zermati, Y., Garrido, C., Amsellem, S., Fishelson, S., Bouscary, D., Valensi, F., Varet, B., Solary, E., Hermine, O., 2001. Caspase activation is required for terminal erythroid differentiation. J. Exp. Med. 193, 247e254. Zhang, J., Socolovsky, M., Gross, A.W., Lodish, H.F., 2003. Role of Ras signaling in erythroid differentiation of mouse fetal liver cells: functional analysis by a flow cytometry-based novel culture system. Blood 102, 3938e3946. Zhang, L., Flygare, J., Wong, P., Lim, B., Lodish, H.F., 2011. miR-191 regulates mouse erythroblast enucleation by down-regulating Riok3 and Mxi1. Genes Dev. 25, 119e124. Zhang, L., Sankaran, V.G., Lodish, H.F., 2012. MicroRNAs in erythroid and megakaryocytic differentiation and megakaryocyte-erythroid progenitor lineage commitment. Leukemia 26, 2310e2316. Zhao, B., Keerthivasan, G., Mei, Y., Yang, J., McElherne, J., Wong, P., Doench, J.G., Feng, G., Root, D.E., Ji, P., 2014a. Targeted shRNA screening identified critical roles of pleckstrin-2 in erythropoiesis. Haematologica 99, 1157e1167. Zhao, B., Mei, Y., Yang, J., Ji, P., 2014b. Mouse fetal liver culture system to dissect target gene functions at the early and late stages of terminal erythropoiesis. J. Vis. Exp. e51894.