UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

The role of macrophages in human erythropoiesis

Heideveld, E.

Publication date2018Document VersionFinal published versionLicenseOther

Link to publication

Citation for published version (APA):Heideveld, E. (2018). The role of macrophages in human erythropoiesis.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:17 Apr 2022

Esther Heideveld

The role of macrophages in human erythropoiesis Esther HeideveldAll roads lead to red

cells

The role of macrophages inhuman erythropoiesis

Esther Heideveld

The role of macrophages in human erythropoiesis

PhD thesis, University of Amsterdam, the Netherlands

Copyright © 2018 by Esther Heideveld

All rights reserved. No part of this thesis may be reproduced, stored or transmitted in any form or

by any means without prior permission from the author.

A digital version of this thesis can be found at www.dare.uva.nl

Cover design and layout: evelienjagtman.com

Printed by: Gildeprint

ISBN: 978-94-6233-944-6

Printing of this thesis was financially supported by Sanquin Research.

The role of macrophages inhuman erythropoiesis

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maex

ten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op vrijdag 25 mei 2018, te 12.00 uur

door

Esther Heideveld

geboren te Purmerend

Promotiecommissie

Promotor: prof. dr. C.E. van der Schoot AMC-Universiteit van Amsterdam

Copromotores: dr. E. van den Akker Sanquin Research

dr. M.M. von Lindern Sanquin Research

Overige leden: prof. dr. S.M. van Ham AMC-Universiteit van Amsterdam

prof. dr. M.P.J. de Winther AMC-Universiteit van Amsterdam

prof. dr. M. van Egmond Vrije Universiteit

prof. dr. J.N.J. Philipsen Erasmus MC

dr. A.M. Toye University of Bristol

dr. R. van Bruggen Sanquin Research

Faculteit der Geneeskunde

"We adore chaos because we love to produce order."

M.C. Escher

Contents

Chapter 1 General introduction 9

Chapter 2 CD14+ cells from peripheral blood positively regulate hematopoietic stem

and progenitor cell survival resulting in increased erythroid yield

39

Chapter 3 Large-scale in vitro production of red blood cells from human peripheral

blood mononuclear cells

67

Chapter 4 Glucocorticoids induce differentiation of monocytes towards macrophages

that share functional and phenotypical aspects with erythroblastic island

macrophages

91

Chapter 5 The proteomic landscape of erythroblastic island macrophages from

human fetal liver, bone marrow and in vitro cultures

123

Chapter 6 General discussion 145

Appendix Summary

Samenvatting

List of publications

PhD portfolio

Dankwoord

Curriculum Vitae

167

171

177

179

185

191

1

General introduction

CHAPTER 1

Introduction

11

Erythropoiesis during development

Red blood cells have a unique function as oxygen transporter and are required during the entire

lifespan of an individual. In human adults, red blood cells are formed in the bone marrow. However,

during development there are multiple waves of erythropoiesis at several different anatomic sites

(Figure 1). After gastrulation, primitive red blood cells are derived from the hemangioblast, a meso-

dermal precursor cell in the primitive streak that migrates to the yolk sac1-3. Here, the mesodermal

cells adhere to form blood islands, which are the origin of primitive hematopoiesis1, 4. The human

yolk sac produces mainly erythroid cells that are relatively large, short-lived, nucleated erythrocytes

and can be found in the embryo at about day 16-20 of development5-10. Macrophages and primitive

megakaryocytes have been described in the yolk sac as early as day 25 of development. However, at

8.5 weeks of development, hematopoiesis in the yolk sac entirely disappears11, 12. Although human yolk

sac hematopoietic development is difficult to study due to ethical issues and availability of tissue13, it

has been proposed that part of the tissue macrophages in adults are derived from the yolk sac and

are not replenished by monocyte-derived macrophages, which we will discuss in the next section14.

Primitivestreak

Hemangioblasts

Yolk sac

Fetal liver

AGM

Day 16/20 - Week 8.5(I)

Day 27/30 - Week 30(III)

Day 74 - After birth(IV)

HBB

HBGEMP HBG

HBE

HSC

Day 26 - Week ?(II)

HBG

Yolk sac

Yolk sac / Fetal liver

Fetal liver

Bone marrow

Figure 1. Hematopoietic development in the human embryo. (I) Hemangioblasts from the primitive streak migrate to the yolk sac and form blood islands, which produce HBE-expressing primitive erythrocytes from day 16-20 up to 8.5 weeks. (II) In parallel, EHT-derived EMP arise in the yolk sac around day 26 and develop into HBG-expressing definitive erythrocytes. In addition, yolk sac EHT-derived EMP migrate to the fetal liver and dif-ferentiate towards erythrocytes. (III) From day 27-30 until the 7th month of gestation, the first AGM EHT-derived HSC migrate to the fetal liver and develop into definitive erythrocytes, also expressing fetal hemoglobin HBG. (IV) At week 10.5 (day 74), HSC from the fetal liver colonize the bone marrow and differentiate towards definitive erythrocytes expressing adult hemoglobin HBB.

Shortly after the initial wave of primitive yolk sac-derived erythropoiesis, a second intermediate defin-

itive wave has been identified around day 26 of human development to putatively bridge the gap

between primitive and definitive hematopoietic stem cell (HSC)-derived cells. This transient wave

Chapter 1

12

in the yolk sac derives from transient multipotent erythromyeloid progenitors (EMP) that are gen-

erated from hemogenic endothelium in a process called endothelial-to-hematopoietic transition

(EHT). EMP are the first cells to colonize the fetal liver15, 16. Interestingly, this wave produces definitive

erythroblasts with a remarkable in vitro expansion potential in mice (>1 year)17. The first CD34+ HSC

are formed from endothelial cells in the aorta-gonad-mesophrenos (AGM) region through EHT and

start to colonize the liver at day 27 to 30 of development8, 18-24. During this high flux of definitive fetal

liver erythropoiesis from the 9th to the 24th week of gestation25, 26, HSC localize to the bone marrow

at 10.5 weeks of development, which is at time of birth the main organ for erythropoiesis. Erythroid

cells from the different developmental waves can be readily identified due to switching of hemoglobin

subunit expression27. The hemoglobin molecule is a tetramer and consists of two α-like globin peptide

chains and two subunits of the β-like globin peptides. During primitive erythropoiesis in the yolk sac

of the early embryo, abundant expression of the β-like globin epsilon (HBE) and α-like globin zeta

(HBZ) is found in erythroid cells. Thereafter, during production of definitive erythroid cells in the fetal

liver the β-like globin gamma (HBG) is expressed28. However, it must be noted that both EMP-derived

erythroid cells from yolk sac and fetal liver express HBG and thus this cannot be used to discriminate

between the second and third definitive wave. Markers that discriminate between these two definitive

waves or their progeny are in need. After birth, the gamma chain is exchanged for beta-globin (HBB

>97%) or delta-globin (HBD ~3%) and pairs with the alpha-globin (HBA) to form adult hemoglobin29,

30. During aging, the red marrow gets replaced by yellow marrow, and hematopoiesis shifts from the

long bones, to the pelvis, sternum, cranium, and vertebrae.

Within the human bone marrow, the differentiation from stem cell to erythrocyte follows several

stages based on cell morphology, colony forming capacity or marker expression by flow cytometry

(Figure 2). CD34+ HSC first differentiate to multipotent progenitors (MPP), which give rise to either

common lymphoid progenitors (CLP) or common myeloid progenitors (CMP). CLP can generate all

lymphoid cells including T-cell, B-cell and NK-cells31, while CMP can generate all myeloid cells but not

lymphoid cells32, 33. CMP develop into granulocyte-macrophage progenitors (GMP) or differentiate

to megakaryocyte-erythroid progenitors (MEP). All three populations express CD34, however, we

can discriminate between the populations using CD45RA and CD123 in the Lin-CD34+CD38+ cells:

CD123lowCD45RA- CMP, CD123lowCD45RA+ GMP and CD123-CD45RA- MEP32. Another report has been

published using BAH1.1 and CD45RA and described BAH1.1-CD45RA- CMP, BAH1.1-CD45RA+ GMP and

BAH1.1+CD45RA- MEP34. Colony assays also allow to follow commitment of cells to the hematopoietic

lineages. CMP can generate colony forming unit granulocyte, erythrocyte, monocyte (CFU-GEMM)

including all types of myeloid colonies, whereas GMP give granulocyte and macrophage colonies

(CFU-GM) and MEP give rise to the earliest erythroid colonies identified as erythroid burst forming

units or BFU-E. MEP will further develop into CD34-CD71highCD235alow pro-erythroblasts while grad-

ually losing CD34 expression and acquiring high expression of CD7132, 33, 35. Cells between the MEP

and pro-erythroblast stage give rise to colonies of the late-stage erythroid lineage (CFU-E). In addition,

variations to the general hematopoietic differentiation scheme have been emerging lately with the

Introduction

13

availability of single cell RNA-sequencing techniques. In these novel hematopoietic trees, the MEP

population plays a less prominent role, because HSC may directly differentiate into megakaryoid

cells without any precursors in vivo36. Pro-erythroblasts subsequently differentiate into basophilic,

polychromatophilic, and orthochromatic erythroblasts and gradually acquire expression of CD235a

(GPA)6, 37-39. During this process cells become smaller, the nucleus condensates and will be expelled

resulting in an enucleated reticulocyte that loses CD71 and enters the blood stream to mature into

an erythrocyte6. The enucleation process can be followed using the DNA-associated cell permeable

DRAQ5 as a marker for nucleated cells.

HSC MPP CMP MEP Pro Baso Poly Ortho Retic Ery

Bone marrow Circulation

CD235aCD34

DRAQ5

BFU-E

GMP Mono Macro

Neutro

Flow

Cytometry

ColonyAssay

Cellmorphology

CFU-E

CFU-GEMM

CD71

Figure 2. Differentiation of HSC to erythrocytes based on cell morphology, colony assay and flow cytometry. CD34+ HSC develop into CFU-GEMM in the adult bone marrow. Here, the decision is made if cells follow an erythroid differentiation program and differentiate to MEP, or a myeloid program and differentiate to GMP. GMP will further develop into monocytes (Mono) and macrophages (Macro) or neutrophils (Neutro), while MEP further develop into CD71+ pro-erythroblasts (Pro) and differentiate towards basophilic (Baso), polychromatophilic (Poly), and orthochromatic (Ortho) erythroblasts. During erythroid differentiation cells loose CD71 expression, while acquiring expression of CD235a. Orthochromatic erythroblasts enucleate and thereby loose the expression of DRAQ5 and reticulocytes (Retic) enter the blood stream to develop into an erythrocyte (Ery).

After the release of reticulocytes from the bone marrow, the spleen acts as a quality control during the

entire lifespan of the red cell. A small amount of the circulating human reticulocytes have large auto-

phagic vacuoles40. The amount of vacuoles was increased in splenectomized individuals and patients

Chapter 1

14

with sickle cell disease, however, these specific reticulocytes were cleared in a healthy recipient41-43.

This suggests that final remodeling of reticulocytes by removal of the autophagic vacuoles most likely

occurs in the spleen. Extrusion of the autophagic vesicles allows reticulocytes to reduce the surface

area and volume of the cell. It is believed that the autophagic vesicles are phagocytosed by red pulp

macrophages, as clodronate depletion of macrophages inhibits reticulocyte maturation in the circu-

lation44-46. However, Griffiths et al. observed the presence of mature erythrocytes in splenectomized

indivuals, which suggests that the final maturation of reticulocytes could occur independent of splenic

macrophages47. Besides this ill-defined role of splenic macrophages in the removal of autophagic

vesicles from early reticulocytes, splenic macrophages are also known for the clearance of intracellular

inclusions from the red cell membrane and removal of aged red blood cells from the circulation48-53.

In addition, clearance of old erythrocytes and iron recycling has also been observed in the liver54, 55.

During steady-state erythropoiesis, red blood cells are produced at a constant rate of two million

erythrocytes per second and this is restricted to the adult bone marrow in both human and mouse.

However, erythropoiesis in human and mouse is differently regulated during stress erythropoiesis.

Stress-induced erythropoiesis is distinct from basal steady-state erythropoiesis as various types of

stress like hypoxia, hemorrhage or (chronic) anemia induce a rapid production of new erythrocytes,

resulting in restored oxygenation. In humans, both steady-state and stress-induced erythropoiesis

are restricted to the bone marrow. Although some papers suggest that human stress erythropoiesis

also occurs in the spleen56, this has never been confirmed. Splenomegaly that has been observed

during hypoxia is not the result of de novo erythropoiesis, but the release of stored erythrocytes in

the spleen, which has been observed during normobaric hypoxia in humans that showed increased

spleen volume, spleen contraction and erythrocyte release into the bloodstream57. In contrast,

stress-induced erythropoiesis in mice is localized to the extramedullary sites and occurs outside the

medulla of the bone, mainly in the spleen but also the liver is involved58, 59.

Signaling during erythropoiesis

The mechanisms defining the “decision” to differentiate into either one or the other hematopoietic

lineages are not completely known. Currently, it is generally accepted that lineage specific growth

factors allow the survival of specific lineages through concomitant receptor expression. This “permits”

lineage commitment and hence is termed the permissive model. Erythropoietin (EPO) is produced by

the fetal liver and shifts to the kidney after birth and although the kidney produces most EPO (80%)

in adults, the liver retains its ability to produce 10-15% of total EPO production in human, which can

be even further increased to 30% upon stress due to low oxygen levels60, 61. EPO binds and activates

the EPO-receptor (EPOR) on immature erythroid cells and functions as a survival factor during eryth-

roid differentiation62, 63. The importance of EPO has also been shown by Epo-/- or EpoR-/- mice, which

are both embryonically lethal due to lack of mature erythrocytes64, 65. Another key regulator during

erythropoiesis is stem cell factor (SCF), which signals via the mast/stem cell growth factor receptor

KIT (CD117) to enhance growth and survival of erythroid progenitors. Interestingly, SCF in combina-

Introduction

15

tion with EPO has a synergistic effect on erythroid cell proliferation, which relies on synergistic signal

transduction activation64, 66-68. SCF is provided by the stromal niche cells in the bone marrow as a

membrane bound protein. Mutations that allow only soluble SCF display a severe erythroid deficiency

and Kit knockout mice are non-viable at embryonic day 14 due to reduced numbers of erythroid cells

in the fetal liver69. Furthermore, it has been shown that KIT plays a role in EPOR maintenance, which

results in erythroid progenitor survival upon EPO stimulation63. During stress-induced erythropoiesis,

the glucocorticoid receptor (GR) cooperates with KIT and EPOR and induces long-term proliferation

of immature erythroid cells without differentiation68, 70-78. Glucocorticoids, which are the ligands of

the GR, are produced in the adrenal gland. In absence of ligands, the GR is bound to HSP90 and its

co-chaperones and localizes outside the nucleus in the cytoplasm79. HSP90-independent modes of

GR extranuclear localization have been reported and involve direct interaction of the GR with EPOR80.

Binding of glucocorticoids release the GR from HSP90 and enable nuclear translocation. Binding of full

agonists such as hydrocortisone or dexamethasone also enable binding of transcriptional activators

to the C-terminal transactivation domain, while partial (ant)agonists do not81. These latter still enable

nuclear localization and transcriptional activation via the hormone independent N-terminal transac-

tivation domain and enable the repressive function of the GR. The GR requires homodimerization

to be active as a transcription factor that binds the glucocorticoid transactivation element. However,

the GR can also heterodimerise with transcription factors such as STAT5 to bind a compound site

consisting of a half-mer GR site and half-mer STAT5 site81. Although STAT5 plays a prominent role in

erythropoiesis, there is currently no evidence that the GR acts as a GR-STAT5 dimer as it does in the

mammary gland68. Transcriptional repressive and active roles of STAT5 are important in macrophage

functionality, but also here it is unknown if GR-STAT5 oligomers are formed or involved. In addition,

DNA-binding domain independent monomeric GR association with chromatin was found in com-

plex with AP1 or NFκB81. This interaction represses the activity of AP1 and NFκB and is important in

repression of the immune function of macrophages. Stress erythropoiesis requires the activity of the

dimerized GR, because Gr knockout mice or mice expressing a mutant incapable of dimerization (Grdim/

dim) lack any form of stress erythropoiesis and display severely decreased fetal liver erythropoiesis74, 75.

Direction to specific lineages is generally believed to occur via stochastic variations in the concen-

tration of specific transcription factors. A well-known example is the “decision” of CMP to become

GMP or MEP, which depends on the relative levels of PU.1 (GMP direction) and GATA1 (MEP direc-

tion). The transcription factor GATA1 plays a critical role in erythroid development, both in primitive

and definitive erythropoiesis, as disruption of Gata1 in mice results in embryonic lethality due to a

defect in primitive erythropoiesis, while GATA1-/-embryonic stem cell-derived pro-erythroblasts in vitro

develop a block in differentiation82-86. Furthermore, during maturation, it has been shown that both

EPOR and GATA1 levels are increased, especially when cells differentiate from pro-erythroblast into

basophilic erythroblasts. During terminal differentiation from basophilic erythroblasts to erythrocytes,

both GATA1 and EPOR are reduced87-90. Failure to reduce GATA1 impairs erythroid maturation91. In

contrast, PU.1 is a positive regulator of lymphoid and myeloid differentiation and prevents differen-

Chapter 1

16

tiation into the erythroid lineage. PU.1 is still expressed in early erythroid cells, but downregulated

when cells mature92. Ectopic overexpression of PU.1 in erythroblasts inhibits the differentiation and

can result in immortalized erythroblasts93, 94. Another example is the transition of a MEP towards the

erythroid or megakaryoid lineage, which depends on the relative abundance of KLF1 (erythroid) and

FLI1 (megakaryoid). KLF1, the erythroid Krüppel-like factor (or EKLF), is crucial for erythropoiesis as

Klf1-/- mice show defects in hemoglobin metabolism and die around embryonic day 1495, 96. In addi-

tion, FLI1 suppresses erythroid differentiation and direct cells towards the megakaryoid lineage97-100.

Interestingly, recently this transcription factor driven directive model was challenged, as single cell

HSC tracking through differentiation revealed that PU.1 and GATA1 are never expressed at the same

time points during HSC differentiation101.

Erythropoiesis is not only regulated by intrinsic and soluble extrinsic factors, but also by cell-cell inter-

actions. During the development from HSC to erythrocyte, the microenvironment plays a crucial role

as macrophages in the stroma, which are located near the HSC in the bone marrow, affect HSC and

erythroid cells through ill-defined mechanisms. It is important to know the underlying mechanisms

of erythroid-macrophage interactions, as this could help optimizing red blood cell cultures. In the

next sections, we focus on these supporting macrophages and give an overview of their origin, char-

acterization in mice and human and their role in the HSC and erythroid niche in the bone marrow.

Tissue resident macrophages and their origin

The bone marrow is a complex organ devoted to the generation of all blood cells. It is divided in var-

ious compartments or niches, presumably dependent on the presence of different cell populations,

secreted cytokines and chemokines. Within the bone marrow, hematopoietic stem and progenitor

cell (HSPC) homeostasis and erythropoiesis are co-regulated by macrophages. Macrophages are key

regulators of both innate and adaptive immunity, however, they are also known for their role in tissue

homeostasis, development and malignancy. Dependent on cues in the microenvironment, monocytic

cells differentiate into macrophages with various phenotypes and functions, and migrate to different

tissues. In 2000, Mills et al. described a model for macrophage activation in which two major opposing

macrophage activities were classified into subtypes: classical M1 or alternative M2 macrophages102.

M1 macrophages inhibit cell proliferation and induce a pro-inflammatory response, while M2 mac-

rophages are anti-inflammatory, promote cell proliferation and are known to be involved in tissue

repair and wound healing103-107. The M1/M2 model has been used predominantly as it is a simple way

to distinguish between the two functional properties of macrophages. However, it depicts M1 and

M2 activation as clearly distinct processes, while macrophage polarization is more complex108. For

instance, the M2 population has been further divided into M2a-d macrophages based on inducing

agents, marker expression and functionality irrespective of tissue residence109-111. This sub-classifi-

cation becomes even more complex upon describing resident macrophages in different tissues as

marker expression and functionality can be influenced by the specific niche in which these macro-

phages reside. This results in an array of different notifications and classifications for tissue resident

Introduction

17

macrophages making generalizations like the M2 sub-classification rather limited and oversimplified.

In the following sections, we will describe the different macrophages based on marker expression

and functionality and will refrain from classical definitions like M1 and M2.

Based on their origin, tissue resident macrophages can be divided into two subsets. One derives from

the yolk sac and is maintained by self-renewal and proliferation. Another population originates from

bone marrow myelopoiesis and from resulting circulating monocytes112. These monocytes are a het-

erogeneous group consisting of three subsets, the classical CD14++CD16-, intermediate CD14++CD16+

and non-classical CD14+CD16+ monocytes113, 114. The non-classical monocytes are known for their

patrolling function in the tissues and are therefore suggested to differentiate into resident macro-

phage populations that function in wound healing and tissue repair115. Until recently, it was believed

that all macrophages including tissue resident macrophages derived from monocytes. However,

the complexity and heterogeneity of macrophages have been underestimated. In the last decade,

interest has been grown to understand the development, relationship, function and origin of the

different macrophage subsets within the different tissues. The dogma, which came into existence

during the 1960s and 1970s, is increasingly challenged. This dogma dictates that tissue resident

macrophages derive from de novo monocytes produced during myelopoiesis from definitive HSC in

the bone marrow116, although the same researchers already in 1984 showed that macrophages in

the spleen had dual origins, some were derived from monocytes while others were dependent on

self-renewal117. The general 20th century simplistic view of HSC differentiating to a CMP which further

matures to a GMP and subsequently monocytes that exit the bone marrow has gained in resolution

with the discovery of a clonogenic progenitor118. This progenitor gives rise to monocytes, macro-

phages, and dendritic cells119, and more recently to a monocyte-restricted bone marrow precursor

termed common monocyte progenitor (cMOP)120. In any case, de novo monocytes may then home to

their respective tissues and further differentiate into tissue resident macrophages or pro-inflamma-

tory macrophages depending on the systemic need116, 121, 122. Nevertheless, several lacunae concerning

the presumed bone marrow origin of tissue resident macrophages within this dogma remained.

Hashimoto et al. reported that recovery of specific tissue resident macrophages (e.g., microglia or

Langerhans cells) after tissue damage did not involve donor cells but appeared to be of host origin112.

Interestingly, monocytopenic mice present with normal macrophage distributions in the tissues123.

Several other authors showed persistent and maintained macrophage populations independent of

monocyte production112, 124-134. These results suggested that certain macrophage populations did not

arise from de novo generated bone marrow monocytes. Further investigation using i) parabiotic mice,

ii) selective ablation using specific macrophage markers, iii) gene expression analysis, and iv) single

population tracing led to the notion that a selection of tissue macrophages are derived independently

from bone marrow myelopoiesis112. Surprisingly, these cells are able to undergo renewal divisions

in order to repopulate the tissue after injury or insult to the tissue. Phenotypically these cells are

completely different from the inflammatory macrophage108. Interestingly, host-derived recovery of

macrophages in specific tissues after whole body irradiation prior to bone marrow transplantations

Chapter 1

18

of mice can arise independently which suggests the presence of bone marrow myeolopoiesis-inde-

pendent macrophages. The origin of these specific macrophages, their renewal capacity, signaling

cues that maintain these cells and whether the tissue resident effector cell and renewal population

are different entities are currently vividly pursuited within the field. Recently, it has been shown that

minor populations of macrophages in specific tissues are originating from yolk sac myelopoiesis and

derive prior to the establishment of definitive HSC-dependent hematopoiesis14, 112. It suggests that a

subset of macrophages finds their origin in early embryogenesis and maintains their tissue presence,

functionality and renewal capacity throughout adulthood. Indeed, fate-mapping studies now show

that macrophages are present in skin, brain and other tissues before the onset of monocytes in the

bone marrow112, 129, 131, 135. Arterial macrophages are mainly generated from early and late EMP derived

from the yolk sac126, while intestinal macrophages are maintained by constant replenishment of

monocytes136. In addition, the first Langerhans cells are thought to derive from yolk sac EMP-derived

macrophages that migrate to the skin and are exchanged by fetal liver monocytes after the second

wave of hematopoiesis124, 137. This suggests that there could be even three sources of macrophages:

yolk sac-derived macrophages, and fetal liver or adult bone marrow-derived monocytes that differ-

entiate into macrophages in the tissue138, 139. The contribution of these embryo-derived macrophages

to tissue functionality and regeneration in particularly must be resolved in order to understand their

role in tissue homeostasis as well as in pathological conditions. Indeed, the importance of host-tissue

resident macrophages persistence can be clinically important140. One of these studies indicated that

host Langerhans cells in the skin may remain after bone marrow transplantation in patients, the

persistence of these cells is highly correlated with severity of graft-versus-host disease140, 141. In addi-

tion, the bone marrow-independent origin of these macrophages places serious constraints on the

ability of in vitro differentiation model systems to generate and study tissue resident macrophages.

To establish if other tissues are equally affected, the different macrophage populations in human

tissues must be characterized phenotypically, functionally and upon tissue insult. In addition, it is

unknown if these specific macrophages are able to transmigrate to different tissues. This is difficult

to assess as macrophage identity depends on the tissue niche and markers may differ from tissue

to tissue but also during transit from one tissue to the other. However, Hashimoto et al. showed that

exchange of macrophages between tissue is limited in parabiotic mice112. In conclusion, the monocyte/

macrophage system needs to be overhauled and hierarchically reclassified to include the novel class

of bone marrow hematopoietic-independent tissue resident (embryonic) macrophage populations.

Hematopoiesis and specifically erythropoiesis is inseparably connected with nursing macrophages

and the total erythroid flux may thus also dependent on the availability and regulation of these specific

macrophages. Next, we will review two important niches in which macrophages have been observed.

Macrophages within the stem cell niche

HSC reside in specific distinct microenvironments in the bone marrow, however these niches are

poorly understood as characterization of the specific locations within the bone marrow is difficult

to visualize and study in vivo as well as ex vivo. Although the markers that define HSC have been

Introduction

19

relatively well characterized in mice142, the markers that define human long-term repopulating HSC

have not been properly identified and these cells are currently mostly defined as lineage-negative

CD34+CD38- cells with inclusion of several other markers like CD49d, CD117, CD133 and the ability to

expel Hoechst through multi-drug resistance receptors143, 144. Some HSC have been found as single

cells in the trabecular cavities of the long bones. Others have been found in a highly vascularized

area towards the center of the bone marrow145-147. The bone marrow contains a set of bone marrow

resident macrophages of which some reside within the main HSC niches: the endosteal and peri-

vascular niche. Within these niches stroma provide the optimal environment for the bone marrow

and support the maintenance of HSC. These stromal cells contain osteoblasts, CXCL12-abundant

reticular cells (CAR cells), endothelial cells and macrophages. 60% of the long-term repopulating HSC

are localized near the vasculature, close to the endothelium lining the sinusoids148. The sinusoids

can be found close to the endosteal niche, but are more likely to be located at greater distances

and contain CAR cells that secrete factors to promote self-renewal of HSC. This perivascular niche

includes the sinusoids in the bone marrow and contains a specific subset of macrophages that sup-

port maintenance and proliferation of mesenchymal stem cells (MSC). In addition, they instruct HSC

to reside in the bone marrow149.

The endosteal niche contains quiescent HSC and is located in the proximity of the bone surface,

where osteoblasts line the membrane. This niche contains a specific subset of tissue resident mac-

rophages referred to as osteal macrophages or osteomacs which play diverse roles in bone biology.

Osteomacs can be found immediately adjacent to osteoblasts and regulate bone formation and

homeostasis. Furthermore, they regulate maintenance and proliferation of Nestin+ MSC. These

MSC express a variety of HSC retention factors and it is thought that macrophages ‘talk’ to MSC via

unknown secreted factors, excluding IL1, IL10, TNFα, and insulin-like growth factor 1 (IGF1), result-

ing in HSC retention in the bone marrow150-152. Indeed, macrophage Fas-induced apoptosis (MAFIA)

mice lack osteomacs and as a consequence have increased mobilized HSC151, 153. Ablation of CD169+

macrophages leads to HSC mobilization and defects in erythropoiesis154, 155. In mice, osteomacs have

been characterized by the expression of a variety of markers and it has been suggested to have a

counterpart in humans. Osteomacs are positive for F4/80, CD169, vascular cell adhesion molecule 1

(VCAM1), CD11b, CD68, CD115, macrophage-3 antigen (MAC3) and are negative for tartarate-resistant

acid phosphatase (TRAP)151, 154, 156. The commonly used marker F4/80 for murine macrophages has

a human ortholog, an EGF-like module containing mucin-like hormone receptor 1 (EMR1). However,

expression of EMR1 is absent on mononuclear phagocytic cells and is only restricted to eosinophilic

granulocytes157. EMR1 can therefore not be used to define human macrophages. In human bone

marrow, Chow et al. describes a macrophage population that shares characteristics with murine

bone marrow macrophages. The human ortholog for CD169+VCAM1+ murine macrophages was

identified as CD15-CD163+VCAM1+CD169+ human bone marrow macrophages155. It was believed

that the murine macrophages described by Winkler et al.151 were different from the murine CD169+

macrophages by Chow et al.158, however, as both macrophages have a similar function, share a sim-

Chapter 1

20

ilar marker expression profile, and are both located close to the bone and osteoblasts, we suggest

CD169+VCAM1+ macrophages described by Chow et al. could in fact be osteomacs. Haldar et al.

described a bone marrow macrophage population in mice with a similar molecular expression profile

compared to osteomacs (expression of F4/80+VCAM1+CD68+CD169+CD11blow) of which the devel-

opment is dependent on SPI-C transcription factor159. SPI-C was highly expressed in these specific

macrophages as well as red pulp macrophages in the spleen and mice deficient for Spi-c lack both

bone marrow and spleen macrophages159, 160.

Table 1. The different bone marrow macrophage populations present in human and mice based on key cell surface markers and functionality.

Mouse bone marrow macrophages

Name Phenotype Function

Bone marrow macrophages

Gr1+ CD115int F4/80+ CD169+ MHCIIint CD11cint CD68int CD11blow CX3CR1-

Maintenance and retention of HSC in bone marrow154

Tissue resident macrophages

CD169+ VCAM1+ F4/80+ TRAP- Support erythropoiesis155

Osteomacs F4/80+ CD115+ MAC3+ CD68+

F4/80+ CD11b+ Ly6G+

G-CSFR+ CD68+

Lining endosteal/periosteal bone surfaces, regulate osteoblast function150, 152

Maintenance of HSC niches and HSC retention151, 153

Support growth/survival of osteoblasts and inhibit HSPC mobilization156

Tissue resident SPI-C expressing macrophages

F4/80+ VCAM1+ CD68+ CD169+ CD11blow MHCIIint TREML4+ Zbtb46-

Heme degradation and iron recycling159

Bone marrow resident macrophages and monocytes

CD11b+ CX3CR1+ COX2+ a-SMA+ Ly6Cint/low CD115+ CD150+

Adjacent to HSPC and support maintenance and protection of HSPC from exhaustion during stress161

Central macrophages F4/80+ VCAM1+ CD169+ CD163+ Promote pro-erythroblast proliferation in erythroblastic islands155, 162-164

Human bone marrow macrophages

Name Phenotype Function

Osteomacs CD68+ Support osteoblast-mediated bone formation150

Tissue resident macrophages

CD45+ CD169+ VCAM1+ CD163+ CD15-

Unknown, might play a similar role as tissue resident macrophages in mice155

Central macrophages FcRI-III+ CD4+ CD31+ CD11a+ CD11c+ CD18+ HLA-DR+

Support pro-erythroblast proliferation in erythroblastic islands163

In monocytes, Spi-c is repressed by BACH1, however, when heme levels increase BACH1 protein in

monocytes is degraded. Spi-c is no longer repressed and monocytes differentiate to either bone

marrow or red pulp macrophages. When these macrophage numbers increase, heme will be removed

and Spi-c will be repressed by BACH1 which will stop monocyte differentiation159, 160. Whether this

mechanism is similar in all SPI-C expressing monocytes or whether it is restricted to a particular

Introduction

21

monocyte subset is still unclear. Ludin et al. identified a rare activated bone marrow monocyte and

macrophage subset in mice expressing the α-smooth muscle actin and cyclooxyrgenase 2 (COX2)161.

These cells were adjacent to HSC and have been shown to promote the expression of CXCL12 in

perivascular Nestin-positive cells and possibly the maintenance and protection of HSC from exhaus-

tion during stress. The different bone marrow macrophages identified in human and mice have been

summarized in Table 1 and show the complexity and lack of uniform descriptions. There is a need

for an in depth characterization of (bone marrow) macrophages with the use of cell surface markers,

flow cytometry and/or proteomics to allow further research into the roles of these putative different

macrophage population within the tissues and specifically the hematopoietic tissues as studied in

this thesis.

Macrophages within erythroblastic islands

The erythroblast island is a specific structure in the bone marrow comprising of a central macro-

phage surrounded by erythroid precursors at different stages of terminal differentiation. Bessis et al.

first described erythroblastic islands as the specialized microenvironmental compartments in which

mammalian erythroblasts proliferate and differentiate during their second stage of maturation165.

These islands consist of a central macrophage that extends cytoplasmic protrusions to a ring of

surrounding erythroblasts166. Erythroblastic islands have been demonstrated in vivo at all sites of

human definitive erythropoiesis (e.g., in the fetal liver and bone marrow). In addition, erythroblastic

islands may be reconstituted ex vivo upon culturing bone marrow or peripheral blood mononuclear

cells (PBMC) resulting in simultaneous differentiation of HSC into erythroblasts and monocytes into

specific macrophages162, 167. This latter process is dependent on glucocorticoid directed differentiation

of CD34+ cells. Bone marrow central macrophages support erythropoiesis by regulating erythroid

proliferation, differentiation, and enucleation and are believed to clear the expelled erythroid nuclei

surrounded by plasma membrane (or pyrenocytes) resulting from the enucleation process168, 169.

These islands, although one would expect them to reside near the sinusoidal capillary lumen as

reticulocytes will be released into the blood stream, are actually evenly distributed across the bone

marrow170. Quantitative light and electron microscopy of rat bone marrow indeed showed a differ-

ence in the composition of islands adjacent and nonadjacent to the sinusoids. Nonadjacent islands

contain more pro-erythroblasts, adjacent islands are rich in orthochromatophilic erythroblasts, while

the numbers of basophilic and polychromatophilic erythroblasts are comparable in both171. This

remarkable finding suggests dynamic islands which are able to migrate to the sinusoids when cells

within the island maturate. Islands harvested from human bone marrow contain 5-30 erythroblasts

per island. Morphologically, central macrophages are relatively large with many extrusion to facilitate

15-25 erythroid cells172. Phenotypically, these macrophages are relatively well-described in humans

and seem to express a set of M2 macrophage markers like CD163, CD169 and CD206112, 173-175. Fur-

thermore, they have been described to express Fc-receptor I-III, CD4, CD11a,bdim,c, CD18, CD31, CD36,

and HLA-DR (Table 1), however, there is no respiratory burst activity reminiscent of pro-inflammatory

macrophages and they do not express C2b or CD35163, 172.

Chapter 1

22

It has been shown that central macrophages in the bone marrow can bind erythroblasts through

various interactions. Adhesion molecules are the logical proteins to mediate the association between

erythroid cells and central macrophages176-179. Indeed, several have been reported to play a role in

binding of erythroid cells to macrophages, while others have been described to be involved in the

uptake of the pyrenocytes (Figure 3). In mice, CD169+ erythroid-supporting macrophages express

both the erythroblast adhesion receptor CD163 and VCAM1. VCAM1 bind to integrin-α4β1 (VLA4) on

the erythroid cells and blocking these molecules disrupts erythroblastic islands155, 175, 180, 181. CD163

scavenge hemoglobin-haptoglobin complexes, however, future studies are required to unravel

how the interaction between macrophages and erythroblasts via CD163 occurs175. Another integrin

involved in erythroblastic-island formation is ICAM4. This adhesion molecule is present on erythroid

cells during terminal differentiation and mediates binding between erythroid cells but also binding

to integrin-α5β1 (VLA5) on macrophages176, 180, 182-185. Interestingly, VLA5 is not only expressed by

macrophages but is also present on erythroid cells and binds to fibronectin (FN)-expressing stromal

cells, including macrophages186.

Figure 3. Erythroid cells bind to central macrophages and pyrenocytes can be phagocytosed via several mechanisms. Central macrophages express several adhesion molecules and other receptors that enhance the binding of pro-erythroblasts, and more differentiated erythroid cells to macrophages and support the recognition and phagocytosis of phosphatidylserine-exposed pyrenocytes (PS) via Gas6, Protein S or lactadherin.

Introduction

23

Other interactions are facilitated by the macrophage erythroblast attacher (MAEA), also known as

erythroblast macrophage protein (EMP), which is present on both macrophages and erythroid cells.

EMP-EMP interactions are involved in the tight binding of erythroid cells to macrophages and Emp-de-

ficient mice die perinatally due to a reduction in erythroblastic islands176, 187, 188. Recently, Sui et al.

reported that tropomodulin 3 (TMOD3) is present on both macrophages and erythroid cells in the

fetal liver and plays an important role in erythroblastic island formation. Tmod3-/- mice have defec-

tive adhesive interactions in erythroblastic islands in the fetal liver, resulting in impaired terminal

erythroid differentiation, survival, cell-cycle progression and reduced enucleation in erythroblasts189.

It is also thought that ephrins play a role in erythroblastic islands. Ephrin type-B receptor 4 (EPHB4)

is expressed on the early CD34+CD117+ HSPC cells from cord blood and bind to EPHB2 ligand-ex-

pressing macrophages. This results in a fast differentiation towards mature erythroid cells, while the

cells immediately downregulate EPHB4 and detach from the EPHB2 expressing cells190. Interestingly,

EPHB4-positive erythroid cells can also bind EPHB2-negative stromal cells, however, the erythroid cells

remained positive for EPHB4. Another study showed that the EPHB4-EPHB2 interaction is required

for bone marrow HSPC mobilization to the blood191. In addition, EPHB4 has been described as an

alternative candidate for EPOR, which stimulated by EPO increase human ovarian and breast tumor

growth192.

As described in the first section, erythroid cells gain expression of CD235a during terminal matura-

tion. It has been shown that CD235a+ erythroid cells engage via SIGLEC9 on neutrophils. A similar

mechanism may be involved in the binding of GPA-expressing erythroid cells to CD169 (SIGLEC1)

on macrophages193. Furthermore, in this late stage of erythropoiesis VLA4 and VLA5 expression on

erythroid cells is downregulated, presumably to facilitate reticulocyte release from the islands and

stroma to enter the blood stream where reticulocytes further mature into biconcave erythrocytes194,

195. Late in human erythroid differentiation Lutheran (BCAM) is expressed, which can bind laminin

10/11 localized at the bone marrow sinusoids to facilitate entrance into the peripheral circulation196,

197. Besides regulating erythroid output, another function of the central macrophage is pyrenocyte

phagocytosis during enucleation. In mice, clearance of pyrenocytes occurs via the TAM-receptor family

of tyrosine kinases TYRO3, AXL, and MERTK on central macrophages. Pyrenocytes expose phosphati-

dylserine (PS) and will be recognized via MERTK or TYRO3 in a protein S or GAS6-dependent manner,

whereas AXL only recognizes GAS6168, 169. The TAM-receptors play an important role in the phagocytic

ability of macrophages, however, MerTK-deficient mice do not show a defect in pyrenocyte clearance

suggesting that TAM-receptor family members AXL and TYRO3 may cause redundancy or other

compensatory in vivo mechanisms are present that have not been elucidated yet198. Indeed, mice

with a triple knock-out of all TAM-receptors fail to clear apoptotic cells in multiple tissues. These mice

develop normally, but eventually develop autoimmunity like systemic lupus erythematosus (SLE)199.

This is in line with studies that showed that SLE has been associated with failure of macrophages to

phagocytose apoptotic cells and pyrenocytes in both human and mice200-203. In addition, anemia is

found in about 50% of SLE patients and Toda et al. showed that embryos suffer from severe anemia

Chapter 1

24

caused by failure of macrophages to phagocytose pyrenocytes204. Besides the TAM-receptors, other

receptors have been shown to play a role in clearance of apoptotic bodies, such as TIMD4 which

binds phosphatidylserine-exposed pyrenocytes via GAS6 and stabilin (STAB), which directly binds to

phosphatidylserine205-207. In addition, integrin-α5 (ITGA5) expressed on macrophages can also recog-

nize PS-exposed pyrenocytes via lactadherin208.

Although the origin of erythroblastic island macrophages is unknown, they already appear in the

fetal liver prior to HSC homing. This suggests that these fetal liver macrophages may originate from

the first wave of HSC-independent hematopoiesis in the yolk sac14 209. However, studies with adult

Mcsf-deficientop/op mice showed a reduction in the amount of bone marrow resident macrophages

suggesting an HSC-dependent origin of tissue resident macrophages in adults210. The cues that

drive central macrophage differentiation from the HSC are still unknown. Ramos et al. described the

necessity of retinoblastoma (RB) mediated downregulation of PU.1 target gene inhibition through

inhibitor-of-differentiation 2 (ID2) for terminal differentiation of central macrophages211. However,

the origin of central macrophages, cues that regulate differentiation towards central macrophages

and the regulation of central macrophage numbers in the bone marrow remain ill-defined, and is in

need of more directed research. In fact, a general characterization of human fetal and bone marrow

erythroid island macrophages has not been performed and would be an excellent starting point for

further studies.

Do central macrophages primarily regulate erythropoiesis in response to anemia?

Interestingly, specific diphtheria toxin (DTX) mediated ablation of CD169+ macrophages, thus also

ablating central macrophages, results in perturbed erythropoiesis albeit without anemia. Lack of

anemia may be partly attributed to decreased clearance of aged erythrocytes in the spleen as splenic

macrophages responsible for clearance and reticulocyte quality control will also be ablated112, 212, 213.

Although macrophages are implicated in the process of erythropoiesis, the absence of anemia after

CD169+ macrophage ablation clearly shows that erythropoiesis can exist without a central macro-

phage. Indeed, in vitro erythropoiesis can be performed from CD34+ cells from HS(P)C isolated from

PBMC resulting in enucleated reticulocytes that can be used for transfusion purposes75, 77, 214, 215. In

addition, not all erythroid cells in the bone marrow interact with a macrophage216. In agreement

with this, Choi et al. showed that physical interactions between erythroid cells mediated through

the association of ICAM4 with Rho GTPase activating protein DLC1 during high density cultures

increased the enucleation and survival rate and thus ensures autonomous erythroid differentiation217.

These data indicate that central macrophages are largely dispensable for steady-state erythroblast

differentiation towards reticulocytes which includes the process of enucleation. However, induction

of anemia through e.g., phenylhydrazide treatments shows impaired erythroid recovery upon mac-

rophage ablation, indicating that fast responses to alter bone marrow erythroid output is regulated

by and dependent on the presence of the central macrophages112, 211. Interestingly, Rb-deficient mice

die perinatally due to defective erythropoiesis218 caused by a perturbed differentiation of fetal liver

Introduction

25

central macrophages leading to absence of erythroid islands. Deficiency in Id2, directing the inhibi-

tory function of PU.1 on specific target genes, rescued the defect observed in Rb-deficient fetal liver

macrophages. In addition, in Emp-deficient mice the interaction between macrophages and erythroid

cells is also perturbed. These mice also die perinatally, however, this is also partly attributed to an

intrinsic defect in enucleation176. These studies show that erythroid macrophage interactions are cru-

cial during ontogeny. At first glance these results look contradictory to adult erythropoiesis in which

the contribution of central macrophages to erythropoiesis is limited during steady-state. However,

as indicated the necessity of the central macrophage is clearly shown during stress erythropoiesis,

a time of severely increased erythroid output. Parallels to the enormous erythroid flux in fetal liver

necessary to produce sufficient erythrocytes demanded by the developing fetus and stress eryth-

ropoiesis in adults can be made and suggests that central macrophages are crucially important for

erythroid regenerative capacity. Indeed, recently it was shown that the erythrocytosis phenotype in

Polycythemia Vera (PV) can be reversed upon clodronate-containing liposomes mediated ablation of

macrophages carrying the Jak2V617F mutation, clearly indicating a role for macrophages in the patho-

physiology of PV211. Furthermore, the data showed that the proliferation of cells from patients with

this JAK2V617F mutation was increased upon co-culturing with macrophages. This notion is strengthened

by the observation that mice with these aberrant central macrophages have a PV-like phenotype211.

Sufficient macrophage-independent erythropoiesis

Stress erythropoiesisSteady state erythropoiesis

Anemia

+

Macrophage ablation

Poor and severely delayed recovery from anemia

--

Macrophage ablation-

CA

B D

Figure 4. Normal steady-state erythropoiesis consists of macrophage-dependent and independent eryth-ropoiesis. (A) Upon ablation of macrophages (B), macrophage-independent erythroid flux produces sufficient erythrocytes. However, upon anemia or stress, macrophage-dependent erythropoiesis is needed to adequately respond to produce enough erythrocytes to alleviate the shortage (C). Myeloablation, but also bone marrow transplantations that eliminate central macrophages which leads to the inability to respond to anemia and results in (severe) delay in recovery from anemia (D). Of note, we hypothesize the increase in macrophages in panel C.

Chapter 1

26

In addition, macrophage depletion in β-thalassemia mice lead to reduced erythroid proliferation

and reduced reticulocytosis, while erythroid differentiation was increased resulting in increased red

blood cells because of increased red blood cell survival and lifespan. Combined, these studies sug-

gest that central macrophage numbers and functionality regulates the response to anemia and puts

the central macrophage in the center spot of stress-induced erythropoiesis (Figure 4). However, it

remains unclear if erythroid cells dictate the number of central macrophages or if the number of

central macrophages dictate the erythroid flux. In addition, what regulates the number of central

macrophages in the bone marrow and can their number be pharmacologically influenced? These are

important but unaddressed questions to which the answers may have implications in disease treat-

ments. Models to facilitate investigating erythroid-macrophage interactions are not readily available

as to date, central, particularly human, macrophages are ill-defined. This results in a lack of studies

on culturing central macrophage-like cells in order to unravel the underlying mechanisms involved

in the erythroid-supportive function of central macrophages.

Macrophages after bone marrow transplantation

Macrophages have a prominent role during hematopoietic homeostasis within the bone marrow, which

also suggests that they control hematopoietic regeneration after bone marrow transplantation. Mac-

rophages may play a decisive role in hematopoietic recovery as conditioning before bone marrow

transplantation almost always involves myeloablative treatments. The different origins of tissue macro-

phages may also define specific macrophage populations within the adult bone marrow. It is expected

that myeloablation leads to the destruction of bone marrow resident macrophages, however, limited

data is present on the survival of the specific subsets or the recovery of macrophages within the bone

marrow after transplantation. In addition, it is unknown if specific subsets are derived from host or

donor. Langerhans cells were shown to resist high doses of irradiation and repopulate from the host

after congenic bone marrow transplantation, whereas monocytes were all of donor origin219. Similar

results have been observed for microglia220, 221, suggesting that both populations maintain themselves

independently of the contribution of bone marrow-derived circulating precursors, even after exposure

to lethal doses of irradiation. However, microglia can repopulate from bone marrow-derived myelopoi-

esis as donor microglia were found in the brain of transplanted mice indicating that at least a portion

of the microglia in the brain can be repopulated from bone marrow-derived hematopoiesis222. Interest-

ingly, treatment with EPO in the first weeks after transplantation does not result in faster recovery from

anemia223, 224. Indeed, in the first weeks after transplantation, high endogenous EPO levels in plasma

do not induce fast recovery indicating that erythropoiesis is severely perturbed in the first month after

transplantation. On the other hand, EPO treatment after this first month does lead to increased eryth-

roid recovery compared to non-treated patients225 which suggests that the erythroid system is only

able to respond to EPO after an initial period of regeneration. One study suggested that the number

of CD68+ macrophages after bone marrow transplantation correlates with erythroid flux in the bone

marrow226. Taken into account that the crucial role of the central macrophage in the regulation of stress

erythropoiesis facilitates rapid responses to anemia, it would be interesting to evaluate the number

Introduction

27

of central macrophages within the bone marrow after transplantation as the regeneration of central

macrophages and the response to EPO may be connected. Besides this tentative connection, these data

also suggest that central macrophages are lost during bone marrow ablation therapies and may be de

novo generated226. This makes the prospect of differentiating these cells from CD34+ HSC a possibility,

if the cues that dictate differentiation to these macrophage subsets are known, which is presently not

the case. In addition, increasing evidence suggests that besides central macrophages also osteomacs

within the HSC niche are ablated after chemotherapy resulting in a decreased functionality of the niche.

Therefore, it would be reasonable to conclude that homing of HSC to the bone marrow is affected due

to the absence of specific macrophages and thus selective protection or co-transplantation of these

macrophages may increase the transplantation efficiency and hematopoietic recovery. Of note, granu-

locyte colony stimulating factor (G-CSF) also efficiently mobilizes the macrophages that are responsible

for HSC retention in mice. These macrophages have been characterized by the expression of CX3CR1

(the fractalkine receptor), CSF-1R, or CD11b156, 227. This enables purification and isolation of these specific

macrophages from peripheral blood112, 151, 156 which can be used to understand the interaction between

macrophages and HSC. In addition, inhibitory measures can be developed that disrupt the interaction,

thereby only mobilize HSC and keep part of the HSC niche intact to receive the transplanted HSC.

Scope of this thesis

Annually, nearly 400.000 volunteer blood donations are made in the Netherlands228, however, in

future shortage of blood products may develop due to an increased need or decline in donor num-

bers in an aging population229-232. In addition, safe products for transfusion-dependent alloimmunized

patients, for which compatible donor blood is missing, are needed. In particular for patients with sickle

cell disease, that have been alloimmunized against different alloantigens by previous blood transfu-

sions, it can be difficult to find compatible blood. Therefore, research to in vitro generated, specific

matched blood group units of erythrocytes is essential to obtain a degree of donor independency

and to minimize donor-patient blood type variation. However, it is technically challenging to reach

red cell quantities that make up a unit of red blood cells given for transfusion.

Erythroid development in vivo requires support from surrounding cells, such as macrophages. How-

ever, the stage during which support cells influence erythropoiesis is not clearly defined and molecular

events that underlie these support functions are ill-defined. We have previously shown that the

erythroid yield from total PBMC is 10-15 fold increased compared to CD34+ cells isolated from a

similar amount of PBMC77. PBMC consist of several effector cells of which mainly CD3+ T cells, CD14+

monocytes/macrophages and CD19+ B cells. It has been shown that macrophages play an important

role during erythropoiesis, as central macrophages in erythroblastic islands support the erythroid

proliferation, differentiation and enucleation. In order to further optimize erythroid cultures, we aim

to examine if CD14+ monocytes/macrophages from PBMC could increase the erythroid yield and if so,

how these cells inflict their effect. As the erythroid cultures start from PBMC, CD34+ HSC in PBMC will

differentiate towards CD71+CD235a+ erythroblasts after 8 days of culture. This indicates that effector

Chapter 1

28

cells could affect the early HSC and/or more differentiated cells. We therefore aim to provide a new

flow cytometry panel in order to track the differentiation from HSC to erythroblast and investigate in

which stage effector cells inflect their effect (chapter 2).

In order to obtain massive amounts of red blood cells for routine red blood cell production, we aim

to further optimize erythroid culture systems to increase erythroid expansion, differentiation and

enucleation. We therefore divided the culture system in three phases: i) from HSPC to erythroblasts, ii)

expansion of erythroblasts and iii) differentiation of erythroblasts to erythrocytes. Within these stages

we examined when and which supplements had to be added to the culture for optimal erythroid

outgrowth. Furthermore, we explored alternative culture systems (bioreactors) as culturing large

numbers of red cells from culture dishes is infeasible (chapter 3).

Although the importance of central macrophages during erythropoiesis has been reported since the

80s, to date, these macrophages are still ill-defined in humans. In addition, there is no convenient human

model to study the interaction between macrophages and erythroid cells in erythroblastic islands. As

ex vivo central macrophages are difficult to obtain and culture conditions could affect the macrophage

phenotype, we investigated if human CD14+ monocytes could be differentiated towards erythroid-sup-

porting macrophages that function as erythroblastic island macrophages. Erythroid intrinsic effects of

glucocorticoids have been well documented, but it is also known that glucocorticoids affect monocyte

differentiation and macrophage function. As a dual role for glucocorticoids on the process of eryth-

ropoiesis may be present, we investigated the role of glucocorticoids on monocyte differentiation to

macrophages in comparison to cells that have been cultured in the absence of glucocorticoids. In order

to relate these cells to their in vivo counterparts, these macrophages were phenotypically characterized

and compared to fetal liver and adult bone marrow macrophages (chapter 4).

Most studies on erythroblastic island macrophages in the bone marrow or fetal liver have only been

described in mice. As a result, the human counterparts are still ill-characterized. In addition, it is

important to make a comparison between erythroblastic island macrophages from the fetal liver

(week 17-22), adult bone marrow and in vitro cultures, as these macrophages all share an eryth-

roid-support function. This data will give new insights which are relevant for both the erythroid and

the macrophage field. Therefore, we aimed to start an unbiased analysis of the proteome of human

macrophage populations from fetal liver and bone marrow and compare this to in vitro cultured

erythroid-supporting macrophages. This will give clues about the possible underlying mechanisms

and processes by which macrophages interact with erythroid cells and facilitate erythropoiesis. In

addition, novel markers may be identified to discriminate between macrophages from different ori-

gins, as there is a lack of uniform descriptions in the macrophage field (chapter 5).

Finally, we summarize the work described in this thesis and discuss our findings in the light of current

knowledge (chapter 6).

Introduction

29

References

1. Sabin FR. Studies on the origin of blood vessels and of red blood corpuscles as seen in the living blastoderm of chicks during

the second day of incubation. Carnegie Inst Wash Pub n°272, Contrib Embryol. 1920;9:214-262.

2. Murray PDF. The development in vitro of the blood of the early chick embryo. Proc Roy Soc London Serie B. 1932; 111:497–521.

3. Choi K, Kennedy M, Kazarov A, Papadimitriou JC, Keller G. A common precursor for hematopoietic and endothelial cells. Devel-

opment. 1998;125(4):725-732.

4. Maximov AA. Unterschungen über blut und bindegewebe. I. Die frühesten entwicklungsstadien der bluyt and bindegewebzellen

beim säugetier-embryo, bis zum anfang der blutbildung in der leber. Arch Mikr Anat 1909;73:444-450.

5. Luckett WP. Origin and differentiation of the yolk sac and extraembryonic mesoderm in presomite human and rhesus monkey

embryos. Am J Anat. 1978;152(1):59-97.

6. Fraser ST, Isern J, Baron MH. Maturation and enucleation of primitive erythroblasts during mouse embryogenesis is accompanied

by changes in cell-surface antigen expression. Blood. 2007;109(1):343-352.

7. Bloom W, Bartelmez GW. Hematopoiesis in young human embryos. Am J Anat. 1940;67:21-53.

8. Tavian M, Hallais MF, Peault B. Emergence of intraembryonic hematopoietic precursors in the pre-liver human embryo. Devel-

opment. 1999;126(4):793-803.

9. Palis J, Robertson S, Kennedy M, Wall C, Keller G. Development of erythroid and myeloid progenitors in the yolk sac and embryo

proper of the mouse. Development. 1999;126(22):5073-5084.

10. Palis J, Yoder MC. Yolk-sac hematopoiesis: the first blood cells of mouse and man. Exp Hematol. 2001;29(8):927-936.

11. Dommergues M, Aubeny E, Dumez Y, Durandy A, Coulombel L. Hematopoiesis in the human yolk sac: quantitation of erythroid

and granulopoietic progenitors between 3.5 and 8 weeks of development. Bone Marrow Transplant. 1992;9 Suppl 1:23-27.

12. Huyhn A, Dommergues M, Izac B, et al. Characterization of hematopoietic progenitors from human yolk sacs and embryos.

Blood. 1995;86(12):4474-4485.

13. Fukuda T. Fetal hemopoiesis. I. Electron microscopic studies on human yolk sac hemopoiesis. Virchows Arch B Cell Pathol.

1973;14(3):197-213.

14. Perdiguero EG, Klapproth K, Schulz C, et al. The Origin of Tissue-Resident Macrophages: When an Erythro-myeloid Progenitor

Is an Erythro-myeloid Progenitor. Immunity. 2015;43(6):1023-1024.

15. McGrath KE, Frame JM, Fromm GJ, et al. A transient definitive erythroid lineage with unique regulation of the beta-globin locus

in the mammalian embryo. Blood. 2011;117(17):4600-4608.

16. Frame JM, Fegan KH, Conway SJ, McGrath KE, Palis J. Definitive Hematopoiesis in the Yolk Sac Emerges from Wnt-Responsive

Hemogenic Endothelium Independently of Circulation and Arterial Identity. Stem Cells. 2016;34(2):431-444.

17. England SJ, McGrath KE, Frame JM, Palis J. Immature erythroblasts with extensive ex vivo self-renewal capacity emerge from the

early mammalian fetus. Blood. 2011;117(9):2708-2717.

18. Boisset JC, van Cappellen W, Andrieu-Soler C, Galjart N, Dzierzak E, Robin C. In vivo imaging of haematopoietic cells emerging

from the mouse aortic endothelium. Nature. 2010;464(7285):116-120.

19. Ivanovs A, Rybtsov S, Welch L, Anderson RA, Turner ML, Medvinsky A. Highly potent human hematopoietic stem cells first emerge

in the intraembryonic aorta-gonad-mesonephros region. J Exp Med. 2011;208(12):2417-2427.

20. Tavian M, Coulombel L, Luton D, Clemente HS, Dieterlen-Lievre F, Peault B. Aorta-associated CD34+ hematopoietic cells in the

early human embryo. Blood. 1996;87(1):67-72.

21. Medvinsky A, Dzierzak E. Definitive hematopoiesis is autonomously initiated by the AGM region. Cell. 1996;86(6):897-906.

22. North TE, de Bruijn MF, Stacy T, et al. Runx1 expression marks long-term repopulating hematopoietic stem cells in the midges-

tation mouse embryo. Immunity. 2002;16(5):661-672.

23. de Bruijn MF, Ma X, Robin C, Ottersbach K, Sanchez MJ, Dzierzak E. Hematopoietic stem cells localize to the endothelial cell layer

in the midgestation mouse aorta. Immunity. 2002;16(5):673-683.

24. Chen MJ, Yokomizo T, Zeigler BM, Dzierzak E, Speck NA. Runx1 is required for the endothelial to haematopoietic cell transition

but not thereafter. Nature. 2009;457(7231):887-891.

25. Palis J. Primitive and definitive erythropoiesis in mammals. Front Physiol. 2014;5:3.

26. Palis J, Malik J, McGrath KE, Kingsley PD. Primitive erythropoiesis in the mammalian embryo. Int J Dev Biol. 2010;54(6-7):1011-

1018.

27. Cantu I, Philipsen S. Flicking the switch: adult hemoglobin expression in erythroid cells derived from cord blood and human

induced pluripotent stem cells. Haematologica. 2014;99(11):1647-1649.

28. Peschle C, Mavilio F, Care A, et al. Haemoglobin switching in human embryos: asynchrony of zeta----alpha and epsilon----gamma-

globin switches in primitive and definite erythropoietic lineage. Nature. 1985;313(5999):235-238.

Chapter 1

30

29. Dover GJ, Boyer SH. Quantitation of hemoglobins within individual red cells: asynchronous biosynthesis of fetal and adult

hemoglobin during erythroid maturation in normal subjects. Blood. 1980;56(6):1082-1091.

30. Kunkel HG, Wallenius G. New hemoglobin in normal adult blood. Science. 1955;122(3163):288.

31. Kondo M, Weissman IL, Akashi K. Identification of clonogenic common lymphoid progenitors in mouse bone marrow. Cell.

1997;91(5):661-672.

32. Manz MG, Miyamoto T, Akashi K, Weissman IL. Prospective isolation of human clonogenic common myeloid progenitors. Proc

Natl Acad Sci U S A. 2002;99(18):11872-11877.

33. Akashi K, Traver D, Miyamoto T, Weissman IL. A clonogenic common myeloid progenitor that gives rise to all myeloid lineages.

Nature. 2000;404(6774):193-197.

34. Sanada C, Xavier-Ferrucio J, Lu YC, et al. Adult human megakaryocyte-erythroid progenitors are in the CD34+CD38mid fraction.

Blood. 2016;128(7):923-933.

35. Reya T, Morrison SJ, Clarke MF, Weissman IL. Stem cells, cancer, and cancer stem cells. Nature. 2001;414(6859):105-111.

36. Psaila B, Barkas N, Iskander D, et al. Single-cell profiling of human megakaryocyte-erythroid progenitors identifies distinct

megakaryocyte and erythroid differentiation pathways. Genome Biol. 2016;17:83.

37. Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, Lodish HF. Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to

decreased survival of early erythroblasts. Blood. 2001;98(12):3261-3273.

38. Koulnis M, Pop R, Porpiglia E, Shearstone JR, Hidalgo D, Socolovsky M. Identification and analysis of mouse erythroid progenitors

using the CD71/TER119 flow-cytometric assay. J Vis Exp. 2011(54).

39. Bell AJ, Satchwell TJ, Heesom KJ, et al. Protein distribution during human erythroblast enucleation in vitro. PLoS One.

2013;8(4):e60300.

40. Kent G, Minick OT, Volini FI, Orfei E. Autophagic vacuoles in human red cells. Am J Pathol. 1966;48(5):831-857.

41. Holroyde CP, Gardner FH. Acquisition of autophagic vacuoles by human erythrocytes. Physiological role of the spleen. Blood.

1970;36(5):566-575.

42. Mankelow TJ, Griffiths RE, Trompeter S, et al. Autophagic vesicles on mature human reticulocytes explain phosphatidylserine-pos-

itive red cells in sickle cell disease. Blood. 2015;126(15):1831-1834.

43. Mankelow TJ, Griffiths RE, Trompeter S, et al. The ins and outs of reticulocyte maturation revisited: The role of autophagy in sickle

cell disease. Autophagy. 2016;12(3):590-591.

44. Song SH, Groom AC. Sequestration and possible maturation of reticulocytes in the normal spleen. Can J Physiol Pharmacol.

1972;50(5):400-406.

45. Song SH, Groom AC. Scanning electron microscope study of the splenic red pulp in relation to the sequestration of immature

and abnormal red cells. J Morphol. 1974;144(4):439-451.

46. Rhodes MM, Koury ST, Kopsombut P, Alford CE, Price JO, Koury MJ. Stress reticulocytes lose transferrin receptors by an extrinsic

process involving spleen and macrophages. Am J Hematol. 2016;91(9):875-882.

47. Griffiths RE, Kupzig S, Cogan N, et al. Maturing reticulocytes internalize plasma membrane in glycophorin A-containing vesicles

that fuse with autophagosomes before exocytosis. Blood. 2012;119(26):6296-6306.

48. Crosby WH. Siderocytes and the spleen. Blood. 1957;12(2):165-170.

49. Schnitzer B, Sodeman T, Mead ML, Contacos PG. Pitting function of the spleen in malaria: ultrastructural observations. Science.

1972;177(4044):175-177.

50. Gottlieb Y, Topaz O, Cohen LA, et al. Physiologically aged red blood cells undergo erythrophagocytosis in vivo but not in vitro.

Haematologica. 2012;97(7):994-1002.

51. Mebius RE, Kraal G. Structure and function of the spleen. Nat Rev Immunol. 2005;5(8):606-616.

52. Kostova EB, Beuger BM, Klei TR, et al. Identification of signalling cascades involved in red blood cell shrinkage and vesiculation.

Biosci Rep. 2015;35(2).

53. van Zwieten R, Bochem AE, Hilarius PM, et al. The cholesterol content of the erythrocyte membrane is an important determinant

of phosphatidylserine exposure. Biochim Biophys Acta. 2012;1821(12):1493-1500.

54. Stijlemans B, Cnops J, Naniima P, et al. Development of a pHrodo-based assay for the assessment of in vitro and in vivo eryth-

rophagocytosis during experimental trypanosomosis. PLoS Negl Trop Dis. 2015;9(3):e0003561.

55. Theurl I, Hilgendorf I, Nairz M, et al. On-demand erythrocyte disposal and iron recycling requires transient macrophages in the

liver. Nat Med. 2016;22(8):945-951.

56. Kim TS, Hanak M, Trampont PC, Braciale TJ. Stress-associated erythropoiesis initiation is regulated by type 1 conventional

dendritic cells. J Clin Invest. 2015;125(10):3965-3980.

57. Richardson MX, Lodin A, Reimers J, Schagatay E. Short-term effects of normobaric hypoxia on the human spleen. Eur J Appl

Physiol. 2008;104(2):395-399.

58. Bozzini CE, Barrio Rendo ME, Devoto FC, Epper CE. Studies on medullary and extramedullary erythropoiesis in the adult mouse.

Am J Physiol. 1970;219(3):724-728.

Introduction

31

59. Lenox LE, Shi L, Hegde S, Paulson RF. Extramedullary erythropoiesis in the adult liver requires BMP-4/Smad5-dependent sig-

naling. Exp Hematol. 2009;37(5):549-558.

60. Jacobson LO, Goldwasser E, Fried W, Plzak L. Role of the kidney in erythropoiesis. Nature. 1957;179(4560):633-634.

61. Jelkmann W. Physiology and pharmacology of erythropoietin. Transfus Med Hemother. 2013;40(5):302-309.

62. Haase VH. Regulation of erythropoiesis by hypoxia-inducible factors. Blood Rev. 2013;27(1):41-53.

63. Kapur R, Zhang L. A novel mechanism of cooperation between c-Kit and erythropoietin receptor. Stem cell factor induces the

expression of Stat5 and erythropoietin receptor, resulting in efficient proliferation and survival by erythropoietin. J Biol Chem.

2001;276(2):1099-1106.

64. Wu H, Liu X, Jaenisch R, Lodish HF. Generation of committed erythroid BFU-E and CFU-E progenitors does not require erythro-

poietin or the erythropoietin receptor. Cell. 1995;83(1):59-67.

65. Suzuki N, Ohneda O, Takahashi S, et al. Erythroid-specific expression of the erythropoietin receptor rescued its null mutant mice

from lethality. Blood. 2002;100(7):2279-2288.

66. Wang W, Horner DN, Chen WL, Zandstra PW, Audet J. Synergy between erythropoietin and stem cell factor during erythropoiesis

can be quantitatively described without co-signaling effects. Biotechnol Bioeng. 2008;99(5):1261-1272.

67. Tan BL, Hong L, Munugalavadla V, Kapur R. Functional and biochemical consequences of abrogating the activation of multiple

diverse early signaling pathways in Kit. Role for Src kinase pathway in Kit-induced cooperation with erythropoietin receptor. J

Biol Chem. 2003;278(13):11686-11695.

68. Kolbus A, Blazquez-Domingo M, Carotta S, et al. Cooperative signaling between cytokine receptors and the glucocorticoid

receptor in the expansion of erythroid progenitors: molecular analysis by expression profiling. Blood. 2003;102(9):3136-3146.

69. Kapur R, Everett ET, Uffman J, et al. Overexpression of human stem cell factor impairs melanocyte, mast cell, and thymocyte

development: a role for receptor tyrosine kinase-mediated mitogen activated protein kinase activation in cell differentiation.

Blood. 1997;90(8):3018-3026.

70. Wessely O, Deiner EM, Beug H, von Lindern M. The glucocorticoid receptor is a key regulator of the decision between self-renewal

and differentiation in erythroid progenitors. EMBO J. 1997;16(2):267-280.

71. Wessely O, Bauer A, Quang CT, et al. A novel way to induce erythroid progenitor self renewal: cooperation of c-Kit with the

erythropoietin receptor. Biol Chem. 1999;380(2):187-202.

72. Reichardt HM, Kaestner KH, Tuckermann J, et al. DNA binding of the glucocorticoid receptor is not essential for survival. Cell.

1998;93(4):531-541.

73. von Lindern M, Zauner W, Mellitzer G, et al. The glucocorticoid receptor cooperates with the erythropoietin receptor and c-Kit

to enhance and sustain proliferation of erythroid progenitors in vitro. Blood. 1999;94(2):550-559.

74. Bauer A, Tronche F, Wessely O, et al. The glucocorticoid receptor is required for stress erythropoiesis. Genes Dev.

1999;13(22):2996-3002.

75. Leberbauer C, Boulme F, Unfried G, Huber J, Beug H, Mullner EW. Different steroids co-regulate long-term expansion versus

terminal differentiation in primary human erythroid progenitors. Blood. 2005;105(1):85-94.

76. Migliaccio G, Sanchez M, Masiello F, et al. Humanized culture medium for clinical expansion of human erythroblasts. Cell Trans-

plant. 2010;19(4):453-469.

77. van den Akker E, Satchwell TJ, Pellegrin S, Daniels G, Toye AM. The majority of the in vitro erythroid expansion potential resides

in CD34(-) cells, outweighing the contribution of CD34(+) cells and significantly increasing the erythroblast yield from peripheral

blood samples. Haematologica. 2010;95(9):1594-1598.

78. Migliaccio G, Di Pietro R, di Giacomo V, et al. In vitro mass production of human erythroid cells from the blood of normal donors

and of thalassemic patients. Blood Cells Mol Dis. 2002;28(2):169-180.

79. Sanchez ER. Chaperoning steroidal physiology: lessons from mouse genetic models of Hsp90 and its cochaperones. Biochim

Biophys Acta. 2012;1823(3):722-729.

80. Stellacci E, Di Noia A, Di Baldassarre A, Migliaccio G, Battistini A, Migliaccio AR. Interaction between the glucocorticoid and

erythropoietin receptors in human erythroid cells. Exp Hematol. 2009;37(5):559-572.

81. Vandevyver S, Dejager L, Tuckermann J, Libert C. New insights into the anti-inflammatory mechanisms of glucocorticoids: an

emerging role for glucocorticoid-receptor-mediated transactivation. Endocrinology. 2013;154(3):993-1007.

82. Pevny L, Lin CS, D’Agati V, Simon MC, Orkin SH, Costantini F. Development of hematopoietic cells lacking transcription factor

GATA-1. Development. 1995;121(1):163-172.

83. Fujiwara Y, Browne CP, Cunniff K, Goff SC, Orkin SH. Arrested development of embryonic red cell precursors in mouse embryos

lacking transcription factor GATA-1. Proc Natl Acad Sci U S A. 1996;93(22):12355-12358.

84. Weiss MJ, Keller G, Orkin SH. Novel insights into erythroid development revealed through in vitro differentiation of GATA-1

embryonic stem cells. Genes Dev. 1994;8(10):1184-1197.

85. Dzierzak E, Philipsen S. Erythropoiesis: development and differentiation. Cold Spring Harb Perspect Med. 2013;3(4):a011601.

Chapter 1

32

86. Ferreira R, Ohneda K, Yamamoto M, Philipsen S. GATA1 function, a paradigm for transcription factors in hematopoiesis. Mol Cell

Biol. 2005;25(4):1215-1227.

87. Wickrema A, Krantz SB, Winkelmann JC, Bondurant MC. Differentiation and erythropoietin receptor gene expression in human

erythroid progenitor cells. Blood. 1992;80(8):1940-1949.

88. Broudy VC, Lin N, Brice M, Nakamoto B, Papayannopoulou T. Erythropoietin receptor characteristics on primary human erythroid

cells. Blood. 1991;77(12):2583-2590.

89. Dalyot N, Fibach E, Ronchi A, Rachmilewitz EA, Ottolenghi S, Oppenheim A. Erythropoietin triggers a burst of GATA-1 in normal

human erythroid cells differentiating in tissue culture. Nucleic Acids Res. 1993;21(17):4031-4037.

90. Ribeil JA, Zermati Y, Vandekerckhove J, et al. Hsp70 regulates erythropoiesis by preventing caspase-3-mediated cleavage of

GATA-1. Nature. 2007;445(7123):102-105.

91. Whyatt D, Lindeboom F, Karis A, et al. An intrinsic but cell-nonautonomous defect in GATA-1-overexpressing mouse erythroid

cells. Nature. 2000;406(6795):519-524.

92. Back J, Dierich A, Bronn C, Kastner P, Chan S. PU.1 determines the self-renewal capacity of erythroid progenitor cells. Blood.

2004;103(10):3615-3623.

93. Quang CT, Wessely O, Pironin M, Beug H, Ghysdael J. Cooperation of Spi-1/PU.1 with an activated erythropoietin receptor inhibits

apoptosis and Epo-dependent differentiation in primary erythroblasts and induces their Kit ligand-dependent proliferation.

EMBO J. 1997;16(18):5639-5653.

94. Schuetze S, Stenberg PE, Kabat D. The Ets-related transcription factor PU.1 immortalizes erythroblasts. Mol Cell Biol.

1993;13(9):5670-5678.

95. Drissen R, von Lindern M, Kolbus A, et al. The erythroid phenotype of EKLF-null mice: defects in hemoglobin metabolism and

membrane stability. Mol Cell Biol. 2005;25(12):5205-5214.

96. Borg J, Patrinos GP, Felice AE, Philipsen S. Erythroid phenotypes associated with KLF1 mutations. Haematologica. 2011;96(5):635-

638.

97. Athanasiou M, Mavrothalassitis G, Sun-Hoffman L, Blair DG. FLI-1 is a suppressor of erythroid differentiation in human hema-

topoietic cells. Leukemia. 2000;14(3):439-445.

98. Ano S, Pereira R, Pironin M, et al. Erythroblast transformation by FLI-1 depends upon its specific DNA binding and transcriptional

activation properties. J Biol Chem. 2004;279(4):2993-3002.

99. Pereira R, Quang CT, Lesault I, Dolznig H, Beug H, Ghysdael J. FLI-1 inhibits differentiation and induces proliferation of primary

erythroblasts. Oncogene. 1999;18(8):1597-1608.

100. Siripin D, Kheolamai P, Y UP, et al. Transdifferentiation of erythroblasts to megakaryocytes using FLI1 and ERG transcription

factors. Thromb Haemost. 2015;114(3):593-602.

101. Hoppe PS, Schwarzfischer M, Loeffler D, et al. Early myeloid lineage choice is not initiated by random PU.1 to GATA1 protein

ratios. Nature. 2016;535(7611):299-302.

102. Mills CD, Kincaid K, Alt JM, Heilman MJ, Hill AM. M-1/M-2 macrophages and the Th1/Th2 paradigm. J Immunol. 2000;164(12):6166-

6173.

103. Gordon S. Alternative activation of macrophages. Nat Rev Immunol. 2003;3(1):23-35.

104. Goerdt S, Politz O, Schledzewski K, et al. Alternative versus classical activation of macrophages. Pathobiology. 1999;67(5-6):222-

226.

105. Ferrante CJ, Leibovich SJ. Regulation of Macrophage Polarization and Wound Healing. Adv Wound Care (New Rochelle).

2012;1(1):10-16.

106. Martinez FO, Sica A, Mantovani A, Locati M. Macrophage activation and polarization. Front Biosci. 2008;13:453-461.

107. Jetten N, Verbruggen S, Gijbels MJ, Post MJ, De Winther MP, Donners MM. Anti-inflammatory M2, but not pro-inflammatory M1

macrophages promote angiogenesis in vivo. Angiogenesis. 2014;17(1):109-118.

108. Martinez FO, Gordon S. The M1 and M2 paradigm of macrophage activation: time for reassessment. F1000Prime Rep. 2014;6:13.

109. Martinez FO, Helming L, Gordon S. Alternative activation of macrophages: an immunologic functional perspective. Annu Rev

Immunol. 2009;27:451-483.

110. Sironi M, Martinez FO, D’Ambrosio D, et al. Differential regulation of chemokine production by Fcgamma receptor engage-

ment in human monocytes: association of CCL1 with a distinct form of M2 monocyte activation (M2b, Type 2). J Leukoc Biol.

2006;80(2):342-349.

111. Roszer T. Understanding the Mysterious M2 Macrophage through Activation Markers and Effector Mechanisms. Mediators

Inflamm. 2015;2015:816460.

112. Hashimoto D, Chow A, Noizat C, et al. Tissue-resident macrophages self-maintain locally throughout adult life with minimal

contribution from circulating monocytes. Immunity. 2013;38(4):792-804.

113. Zawada AM, Rogacev KS, Rotter B, et al. SuperSAGE evidence for CD14++CD16+ monocytes as a third monocyte subset. Blood.

2011;118(12):e50-61.

Introduction

33

114. Wong KL, Tai JJ, Wong WC, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and non-

classical human monocyte subsets. Blood. 2011;118(5):e16-31.

115. Auffray C, Fogg D, Garfa M, et al. Monitoring of blood vessels and tissues by a population of monocytes with patrolling behavior.

Science. 2007;317(5838):666-670.

116. van Furth R, Cohn ZA. The origin and kinetics of mononuclear phagocytes. J Exp Med. 1968;128(3):415-435.

117. van Furth R, Diesselhoff-den Dulk MM. Dual origin of mouse spleen macrophages. J Exp Med. 1984;160(5):1273-1283.

118. Lavin Y, Mortha A, Rahman A, Merad M. Regulation of macrophage development and function in peripheral tissues. Nat Rev

Immunol. 2015;15(12):731-744.

119. Fogg DK, Sibon C, Miled C, et al. A clonogenic bone marrow progenitor specific for macrophages and dendritic cells. Science.

2006;311(5757):83-87.

120. Hettinger J, Richards DM, Hansson J, et al. Origin of monocytes and macrophages in a committed progenitor. Nat Immunol.

2013;14(8):821-830.

121. Virolainen M. Hematopoietic origin of macrophages as studied by chromosome markers in mice. J Exp Med. 1968;127(5):943-

952.

122. Morrison SJ, Weissman IL. The long-term repopulating subset of hematopoietic stem cells is deterministic and isolatable by

phenotype. Immunity. 1994;1(8):661-673.

123. Kuziel WA, Morgan SJ, Dawson TC, et al. Severe reduction in leukocyte adhesion and monocyte extravasation in mice deficient

in CC chemokine receptor 2. Proc Natl Acad Sci U S A. 1997;94(22):12053-12058.

124. Hoeffel G, Chen J, Lavin Y, et al. C-Myb(+) erythro-myeloid progenitor-derived fetal monocytes give rise to adult tissue-resident

macrophages. Immunity. 2015;42(4):665-678.

125. Lavin Y, Winter D, Blecher-Gonen R, et al. Tissue-resident macrophage enhancer landscapes are shaped by the local microen-

vironment. Cell. 2014;159(6):1312-1326.

126. Ensan S, Li A, Besla R, et al. Self-renewing resident arterial macrophages arise from embryonic CX3CR1(+) precursors and

circulating monocytes immediately after birth. Nat Immunol. 2016;17(2):159-168.

127. Wang J, Kubes P. A Reservoir of Mature Cavity Macrophages that Can Rapidly Invade Visceral Organs to Affect Tissue Repair.

Cell. 2016;165(3):668-678.

128. Perdiguero EG, Geissmann F. The development and maintenance of resident macrophages. Nat Immunol. 2016;17(1):2-8.

129. Ginhoux F, Greter M, Leboeuf M, et al. Fate mapping analysis reveals that adult microglia derive from primitive macrophages.

Science. 2010;330(6005):841-845.

130. Schulz C, Gomez Perdiguero E, Chorro L, et al. A lineage of myeloid cells independent of Myb and hematopoietic stem cells.

Science. 2012;336(6077):86-90.

131. Yona S, Kim KW, Wolf Y, et al. Fate mapping reveals origins and dynamics of monocytes and tissue macrophages under homeo-

stasis. Immunity. 2013;38(1):79-91.

132. Epelman S, Lavine KJ, Beaudin AE, et al. Embryonic and adult-derived resident cardiac macrophages are maintained through

distinct mechanisms at steady state and during inflammation. Immunity. 2014;40(1):91-104.

133. Guilliams M, De Kleer I, Henri S, et al. Alveolar macrophages develop from fetal monocytes that differentiate into long-lived cells

in the first week of life via GM-CSF. J Exp Med. 2013;210(10):1977-1992.

134. Jakubzick C, Gautier EL, Gibbings SL, et al. Minimal differentiation of classical monocytes as they survey steady-state tissues and

transport antigen to lymph nodes. Immunity. 2013;39(3):599-610.

135. Chorro L, Sarde A, Li M, et al. Langerhans cell (LC) proliferation mediates neonatal development, homeostasis, and inflamma-

tion-associated expansion of the epidermal LC network. J Exp Med. 2009;206(13):3089-3100.

136. Bain CC, Bravo-Blas A, Scott CL, et al. Constant replenishment from circulating monocytes maintains the macrophage pool in

the intestine of adult mice. Nat Immunol. 2014;15(10):929-937.

137. Hoeffel G, Wang Y, Greter M, et al. Adult Langerhans cells derive predominantly from embryonic fetal liver monocytes with a

minor contribution of yolk sac-derived macrophages. J Exp Med. 2012;209(6):1167-1181.

138. Swirski FK, Robbins CS, Nahrendorf M. Development and Function of Arterial and Cardiac Macrophages. Trends Immunol.

2016;37(1):32-40.

139. van de Laar L, Saelens W, De Prijck S, et al. Yolk Sac Macrophages, Fetal Liver, and Adult Monocytes Can Colonize an Empty Niche

and Develop into Functional Tissue-Resident Macrophages. Immunity. 2016;44(4):755-768.

140. Merad M, Hoffmann P, Ranheim E, et al. Depletion of host Langerhans cells before transplantation of donor alloreactive T cells

prevents skin graft-versus-host disease. Nat Med. 2004;10(5):510-517.

141. Collin M, Jardine L. A question of persistence: Langerhans cells and graft-versus-host disease. Exp Dermatol. 2014;23(4):234-235.

142. Oguro H, Ding L, Morrison SJ. SLAM family markers resolve functionally distinct subpopulations of hematopoietic stem cells and

multipotent progenitors. Cell Stem Cell. 2013;13(1):102-116.

Chapter 1

34

143. Schwarzenberger P, Spence S, Lohrey N, et al. Gene transfer of multidrug resistance into a factor-dependent human hemato-

poietic progenitor cell line: in vivo model for genetically transferred chemoprotection. Blood. 1996;87(7):2723-2731.

144. Zhong Q, Oliver P, Huang W, et al. Efficient c-kit receptor-targeted gene transfer to primary human CD34-selected hematopoietic

stem cells. J Virol. 2001;75(21):10393-10400.

145. Nilsson SK, Johnston HM, Coverdale JA. Spatial localization of transplanted hemopoietic stem cells: inferences for the localization

of stem cell niches. Blood. 2001;97(8):2293-2299.

146. Lo Celso C, Fleming HE, Wu JW, et al. Live-animal tracking of individual haematopoietic stem/progenitor cells in their niche.

Nature. 2009;457(7225):92-96.

147. Xie Y, Yin T, Wiegraebe W, et al. Detection of functional haematopoietic stem cell niche using real-time imaging. Nature.

2009;457(7225):97-101.

148. Kiel MJ, Yilmaz OH, Iwashita T, Yilmaz OH, Terhorst C, Morrison SJ. SLAM family receptors distinguish hematopoietic stem and

progenitor cells and reveal endothelial niches for stem cells. Cell. 2005;121(7):1109-1121.

149. Ehninger A, Trumpp A. The bone marrow stem cell niche grows up: mesenchymal stem cells and macrophages move in. J Exp

Med. 2011;208(3):421-428.

150. Chang MK, Raggatt LJ, Alexander KA, et al. Osteal tissue macrophages are intercalated throughout human and mouse bone

lining tissues and regulate osteoblast function in vitro and in vivo. J Immunol. 2008;181(2):1232-1244.

151. Winkler IG, Sims NA, Pettit AR, et al. Bone marrow macrophages maintain hematopoietic stem cell (HSC) niches and their deple-

tion mobilizes HSCs. Blood. 2010;116(23):4815-4828.

152. Cho SW. Role of osteal macrophages in bone metabolism. J Pathol Transl Med. 2015;49(2):102-104.

153. Burnett SH, Kershen EJ, Zhang J, et al. Conditional macrophage ablation in transgenic mice expressing a Fas-based suicide gene.

J Leukoc Biol. 2004;75(4):612-623.

154. Chow A, Lucas D, Hidalgo A, et al. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and

progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011;208(2):261-271.

155. Chow A, Huggins M, Ahmed J, et al. CD169(+) macrophages provide a niche promoting erythropoiesis under homeostasis and

stress. Nat Med. 2013;19(4):429-436.

156. Christopher MJ, Rao M, Liu F, Woloszynek JR, Link DC. Expression of the G-CSF receptor in monocytic cells is sufficient to mediate

hematopoietic progenitor mobilization by G-CSF in mice. J Exp Med. 2011;208(2):251-260.

157. Hamann J, Koning N, Pouwels W, et al. EMR1, the human homolog of F4/80, is an eosinophil-specific receptor. Eur J Immunol.

2007;37(10):2797-2802.

158. Westerterp M, Gourion-Arsiquaud S, Murphy AJ, et al. Regulation of hematopoietic stem and progenitor cell mobilization by

cholesterol efflux pathways. Cell Stem Cell. 2012;11(2):195-206.

159. Haldar M, Kohyama M, So AY, et al. Heme-mediated SPI-C induction promotes monocyte differentiation into iron-recycling

macrophages. Cell. 2014;156(6):1223-1234.

160. Kohyama M, Ise W, Edelson BT, et al. Role for Spi-C in the development of red pulp macrophages and splenic iron homeostasis.

Nature. 2009;457(7227):318-321.

161. Ludin A, Itkin T, Gur-Cohen S, et al. Monocytes-macrophages that express alpha-smooth muscle actin preserve primitive hema-

topoietic cells in the bone marrow. Nat Immunol. 2012;13(11):1072-1082.

162. Falchi M, Varricchio L, Martelli F, et al. Dexamethasone targeted directly to macrophages induces macrophage niches that

promote erythroid expansion. Haematologica. 2015;100(2):178-187.

163. Manwani D, Bieker JJ. The erythroblastic island. Curr Top Dev Biol. 2008;82:23-53.

164. Seu KG, Papoin J, Fessler R, et al. Unraveling Macrophage Heterogeneity in Erythroblastic Islands. Front Immunol. 2017;8:1140.

165. Bessis M. [Erythroblastic island, functional unity of bone marrow]. Rev Hematol. 1958;13(1):8-11.

166. Gifford SC, Derganc J, Shevkoplyas SS, Yoshida T, Bitensky MW. A detailed study of time-dependent changes in human red blood

cells: from reticulocyte maturation to erythrocyte senescence. Br J Haematol. 2006;135(3):395-404.

167. Allen TD, Dexter TM. Ultrastructural aspects of erythropoietic differentiation in long-term bone marrow culture. Differentiation.

1982;21(2):86-94.

168. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by central macrophages in erythroblastic islands.

Blood. 2014;123(25):3963-3971.

169. Yoshida H, Kawane K, Koike M, Mori Y, Uchiyama Y, Nagata S. Phosphatidylserine-dependent engulfment by macrophages of

nuclei from erythroid precursor cells. Nature. 2005;437(7059):754-758.

170. Mohandas N, Prenant M. Three-dimensional model of bone marrow. Blood. 1978;51(4):633-643.

171. Yokoyama T, Etoh T, Kitagawa H, Tsukahara S, Kannan Y. Migration of erythroblastic islands toward the sinusoid as erythroid

maturation proceeds in rat bone marrow. J Vet Med Sci. 2003;65(4):449-452.

172. Lee SH, Crocker PR, Westaby S, et al. Isolation and immunocytochemical characterization of human bone marrow stromal

macrophages in hemopoietic clusters. J Exp Med. 1988;168(3):1193-1198.

Introduction

35

173. Crocker PR, Werb Z, Gordon S, Bainton DF. Ultrastructural localization of a macrophage-restricted sialic acid binding hemagglu-

tinin, SER, in macrophage-hematopoietic cell clusters. Blood. 1990;76(6):1131-1138.

174. Takahashi K, Donovan MJ, Rogers RA, Ezekowitz RA. Distribution of murine mannose receptor expression from early embryo-

genesis through to adulthood. Cell Tissue Res. 1998;292(2):311-323.

175. Fabriek BO, Polfliet MM, Vloet RP, et al. The macrophage CD163 surface glycoprotein is an erythroblast adhesion receptor.

Blood. 2007;109(12):5223-5229.

176. Soni S, Bala S, Gwynn B, Sahr KE, Peters LL, Hanspal M. Absence of erythroblast macrophage protein (Emp) leads to failure of

erythroblast nuclear extrusion. J Biol Chem. 2006;281(29):20181-20189.

177. Lee G, Spring FA, Parsons SF, et al. Novel secreted isoform of adhesion molecule ICAM-4: potential regulator of membrane-as-

sociated ICAM-4 interactions. Blood. 2003;101(5):1790-1797.

178. Wang Z, Vogel O, Kuhn G, Gassmann M, Vogel J. Decreased stability of erythroblastic islands in integrin beta3-deficient mice.

Physiol Rep. 2013;1(2):e00018.

179. Spring FA, Griffiths RE, Mankelow TJ, et al. Tetraspanins CD81 and CD82 facilitate alpha4beta1-mediated adhesion of human

erythroblasts to vascular cell adhesion molecule-1. PLoS One. 2013;8(5):e62654.

180. 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(1):411-415.

181. 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(1):86-94.

182. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythro-

blastic island formation. Blood. 2006;108(6):2064-2071.

183. Ulyanova T, Jiang Y, Padilla S, Nakamoto B, Papayannopoulou T. Combinatorial and distinct roles of alpha(5) and alpha(4) integrins

in stress erythropoiesis in mice. Blood. 2011;117(3):975-985.

184. 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(8):2940-2950.

185. McGrath KE, Kingsley PD, Koniski AD, Porter RL, Bushnell TP, Palis J. Enucleation of primitive erythroid cells generates a transient

population of “pyrenocytes” in the mammalian fetus. Blood. 2008;111(4):2409-2417.

186. Tanaka R, Owaki T, Kamiya S, et al. VLA-5-mediated adhesion to fibronectin accelerates hemin-stimulated erythroid differentiation

of K562 cells through induction of VLA-4 expression. J Biol Chem. 2009;284(30):19817-19825.

187. Soni S, Bala S, Hanspal M. Requirement for erythroblast-macrophage protein (Emp) in definitive erythropoiesis. Blood Cells Mol

Dis. 2008;41(2):141-147.

188. Soni S, Bala S, Kumar A, Hanspal M. Changing pattern of the subcellular distribution of erythroblast macrophage protein (Emp)

during macrophage differentiation. Blood Cells Mol Dis. 2007;38(1):25-31.

189. 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(5):758-767.

190. Suenobu S, Takakura N, Inada T, et al. A role of EphB4 receptor and its ligand, ephrin-B2, in erythropoiesis. Biochem Biophys

Res Commun. 2002;293(3):1124-1131.

191. Kwak H, Salvucci O, Weigert R, et al. Sinusoidal ephrin receptor EPHB4 controls hematopoietic progenitor cell mobilization from

bone marrow. J Clin Invest. 2016;126(12):4554-4568.

192. Pradeep S, Huang J, Mora EM, et al. Erythropoietin Stimulates Tumor Growth via EphB4. Cancer Cell. 2015;28(5):610-622.

193. Lizcano A, Secundino I, Dohrmann S, et al. Erythrocyte sialoglycoproteins engage Siglec-9 on neutrophils to suppress activation.

Blood. 2017;129(23):3100-3110.

194. Kossiva L, Paterakis G, Tassiopoulos S, et al. Decreased expression of membrane alpha4beta1, alpha5beta1 integrins and

transferrin receptor on erythroblasts in splenectomized patients with beta-thalassemia intermedia. Parallel assessment of

serum soluble transferrin receptors levels. Ann Hematol. 2003;82(9):579-584.

195. Eshghi S, Vogelezang MG, Hynes RO, Griffith LG, Lodish HF. Alpha4beta1 integrin and erythropoietin mediate temporally distinct

steps in erythropoiesis: integrins in red cell development. J Cell Biol. 2007;177(5):871-880.

196. Gu Y, Sorokin L, Durbeej M, Hjalt T, Jonsson JI, Ekblom M. Characterization of bone marrow laminins and identification of

alpha5-containing laminins as adhesive proteins for multipotent hematopoietic FDCP-Mix cells. Blood. 1999;93(8):2533-2542.

197. Parsons SF, Lee G, Spring FA, et al. Lutheran blood group glycoprotein and its newly characterized mouse homologue specifically

bind alpha5 chain-containing human laminin with high affinity. Blood. 2001;97(1):312-320.

198. Lu Q, Gore M, Zhang Q, et al. Tyro-3 family receptors are essential regulators of mammalian spermatogenesis. Nature.

1999;398(6729):723-728.

199. Lu Q, Lemke G. Homeostatic regulation of the immune system by receptor tyrosine kinases of the Tyro 3 family. Science.

2001;293(5528):306-311.

Chapter 1

36

200. Gaipl US, Voll RE, Sheriff A, Franz S, Kalden JR, Herrmann M. Impaired clearance of dying cells in systemic lupus erythematosus.

Autoimmun Rev. 2005;4(4):189-194.

201. Munoz LE, Janko C, Schulze C, et al. Autoimmunity and chronic inflammation - two clearance-related steps in the etiopathogenesis

of SLE. Autoimmun Rev. 2010;10(1):38-42.

202. Nagata S. Apoptosis and autoimmune diseases. Ann N Y Acad Sci. 2010;1209:10-16.

203. Rothlin CV, Lemke G. TAM receptor signaling and autoimmune disease. Curr Opin Immunol. 2010;22(6):740-746.

204. Toda S, Nishi C, Yanagihashi Y, Segawa K, Nagata S. Clearance of Apoptotic Cells and Pyrenocytes. Curr Top Dev Biol. 2015;114:267-

295.

205. Meyer AS, Zweemer AJ, Lauffenburger DA. The AXL Receptor is a Sensor of Ligand Spatial Heterogeneity. Cell Syst. 2015;1(1):25-

36.

206. Nishi C, Toda S, Segawa K, Nagata S. Tim4- and MerTK-mediated engulfment of apoptotic cells by mouse resident peritoneal

macrophages. Mol Cell Biol. 2014;34(8):1512-1520.

207. D’Souza S, Park SY, Kim IS. Stabilin-2 acts as an engulfment receptor for the phosphatidylserine-dependent clearance of primary

necrotic cells. Biochem Biophys Res Commun. 2013;432(3):412-417.

208. Dasgupta SK, Abdel-Monem H, Guchhait P, Nagata S, Thiagarajan P. Role of lactadherin in the clearance of phosphatidylser-

ine-expressing red blood cells. Transfusion. 2008;48(11):2370-2376.

209. Palis J. Hematopoietic stem cell-independent hematopoiesis: emergence of erythroid, megakaryocyte, and myeloid potential in

the mammalian embryo. FEBS Lett. 2016;590(22):3965-3974.

210. Felix R, Cecchini MG, Hofstetter W, Elford PR, Stutzer A, Fleisch H. Impairment of macrophage colony-stimulating factor produc-

tion and lack of resident bone marrow macrophages in the osteopetrotic op/op mouse. J Bone Miner Res. 1990;5(7):781-789.

211. Ramos P, Casu C, Gardenghi S, et al. Macrophages support pathological erythropoiesis in polycythemia vera and beta-thalas-

semia. Nat Med. 2013;19(4):437-445.

212. de Back DZ, Kostova EB, van Kraaij M, van den Berg TK, van Bruggen R. Of macrophages and red blood cells; a complex love

story. Front Physiol. 2014;5:9.

213. Bratosin D, Mazurier J, Tissier JP, et al. Cellular and molecular mechanisms of senescent erythrocyte phagocytosis by macro-

phages. A review. Biochimie. 1998;80(2):173-195.

214. Giarratana MC, Rouard H, Dumont A, et al. Proof of principle for transfusion of in vitro-generated red blood cells. Blood.

2011;118(19):5071-5079.

215. Anstee DJ, Gampel A, Toye AM. Ex-vivo generation of human red cells for transfusion. Curr Opin Hematol. 2012;19(3):163-169.

216. Rhodes MM, Kopsombut P, Bondurant MC, Price JO, Koury MJ. Adherence to macrophages in erythroblastic islands enhances

erythroblast proliferation and increases erythrocyte production by a different mechanism than erythropoietin. Blood.

2008;111(3):1700-1708.

217. Choi HS, Lee EM, Kim HO, Park MI, Baek EJ. Autonomous control of terminal erythropoiesis via physical interactions among

erythroid cells. Stem Cell Res. 2013;10(3):442-453.

218. Iavarone A, King ER, Dai XM, Leone G, Stanley ER, Lasorella A. Retinoblastoma promotes definitive erythropoiesis by repressing

Id2 in fetal liver macrophages. Nature. 2004;432(7020):1040-1045.

219. Merad M, Sugie T, Engleman EG, Fong L. In vivo manipulation of dendritic cells to induce therapeutic immunity. Blood.

2002;99(5):1676-1682.

220. Ajami B, Bennett JL, Krieger C, Tetzlaff W, Rossi FM. Local self-renewal can sustain CNS microglia maintenance and function

throughout adult life. Nat Neurosci. 2007;10(12):1538-1543.

221. Garcia MR, Ledgerwood L, Yang Y, et al. Monocytic suppressive cells mediate cardiovascular transplantation tolerance in mice.

J Clin Invest. 2010;120(7):2486-2896.

222. Sergijenko A, Langford-Smith A, Liao AY, et al. Myeloid/Microglial driven autologous hematopoietic stem cell gene therapy corrects

a neuronopathic lysosomal disease. Mol Ther. 2013;21(10):1938-1949.

223. Link H, Brune T, Hubner G, et al. Effect of recombinant human erythropoietin after allogenic bone marrow transplantation. Ann

Hematol. 1993;67(4):169-173.

224. Biggs JC, Atkinson KA, Booker V, et al. Prospective randomised double-blind trial of the in vivo use of recombinant human

erythropoietin in bone marrow transplantation from HLA-identical sibling donors. The Australian Bone Marrow Transplant Study

Group. Bone Marrow Transplant. 1995;15(1):129-134.

225. Jaspers A, Baron F, Willems E, et al. Erythropoietin therapy after allogeneic hematopoietic cell transplantation: a prospective,

randomized trial. Blood. 2014;124(1):33-41.

226. Thiele J, Kvasnicka HM, Beelen DW, et al. Erythropoietic reconstitution, macrophages and reticulin fibrosis in bone marrow

specimens of CML patients following allogeneic transplantation. Leukemia. 2000;14(8):1378-1385.

227. Capoccia BJ, Shepherd RM, Link DC. G-CSF and AMD3100 mobilize monocytes into the blood that stimulate angiogenesis in vivo

through a paracrine mechanism. Blood. 2006;108(7):2438-2445.

Introduction

37

228. Wiersum-Osselton JC, Marijt-van der Kreek T, de Kort WL. Donor vigilance: what are we doing about it? Biologicals. 2012;40(3):176-

179.

229. Katalinic A, Peters E, Beske F, Pritzkuleit R. Projection of Morbidity 2030 and 2050: Impact for the National Health System and

Blood Supply. Transfus Med Hemother. 2010;37(3):155-159.

230. An MW, Reich NG, Crawford SO, Brookmeyer R, Louis TA, Nelson KE. A stochastic simulator of a blood product donation envi-

ronment with demand spikes and supply shocks. PLoS One. 2011;6(7):e21752.

231. Goodnough LT, Shander A, Brecher ME. Transfusion medicine: looking to the future. Lancet. 2003;361(9352):161-169.

232. Greinacher A, Weitmann K, Lebsa A, et al. A population-based longitudinal study on the implications of demographics on future

blood supply. Transfusion. 2016;56(12):2986-2994.

ErythroblastsMonocytes HSCPBMC

Survival Differentiation

2

CD14+ cells from peripheral blood positively regulate hematopoietic

stem and progenitor cell survival resulting in

increased erythroid yield

CHAPTER 2

Esther Heideveld, Francesca Masiello, Manuela Marra, Fatemehsadat Esteghamat, Nurcan Yağcı, Marieke von Lindern, Anna Rita F. Migliaccio, and Emile van den Akker

Haematologica 2015 Nov;100(11):1396-1406

Chapter 2

40

Abstract

Expansion of erythroblasts from human peripheral blood mononuclear cells is 4- to 15-fold more effi-

cient than that of CD34+ cells purified from peripheral blood mononuclear cells. In addition, purified

CD34+ and CD34- populations from blood do not reconstitute this erythroid yield, suggesting a role for

feeder cells present in blood mononuclear cells that increase hematopoietic output. Immunodeplet-

ing peripheral blood mononuclear cells for CD14+ cells reduced hematopoietic stem and progenitor

cell expansion. Conversely, this yield was increased upon co-culture of CD34+ cells with CD14+ cells

(full contact or transwell assays) or CD34+ cells re-constituted in conditioned medium from CD14+ cells.

In particular, CD14++CD16+ intermediate monocytes/macrophages enhanced erythroblast outgrowth

from CD34+ cells. No effect of CD14+ cells on erythroblasts themselves was observed. However, 2 days

of co-culturing CD34+ and CD14+ cells increased CD34+ cell numbers and colony-forming units 5-fold.

Proliferation assays suggested that CD14+ cells sustain CD34+ cell survival but not proliferation. These

data identify previously unrecognized erythroid and non-erythroid CD34- and CD34+ populations in

blood that contribute to the erythroid yield. A flow cytometry panel containing CD34/CD36 can be

used to follow specific stages during CD34+ differentiation to erythroblasts. We have shown modu-

lation of hematopoietic stem and progenitor cell survival by CD14+ cells present in peripheral blood

mononuclear cells which can also be found near specific hematopoietic niches in the bone marrow.

Human CD14+ cells enhance ex vivo HSPC survival

41

Introduction

Hematopoiesis occurs in niches that ensure specific interactions and cross-talk of hematopoietic cells

with the surrounding stromal cells and among different hematopoietic cells themselves. These niches

dictate processes such as lineage specification, cell survival and mobilization. Hematopoietic stem and

progenitor cells (HSPC) reside in perivascular niches and within the non-endosteal parenchyma1-4.

This hematopoietic niche consists of mesenchymal stem cells, osteoblasts, and hematopoietic effec-

tor cells, such as T regulatory cells and tissue-resident macrophages. The niche is important for

hematopoietic stem cell (HSC) homeostasis as well as hematopoietic lineage development including

erythropoiesis5. In mice, tissue-resident macrophages are important regulators of HSC retention

within the bone marrow6,7, and ablation of CD163+CD169+ macrophages leads to mobilization of

HSPC, committed progenitors and erythrocyte precursors8. These myelodepleted mice experience

compensated anemia with increased splenic erythroblasts. Increased erythrocyte survival in these

mice is likely due to reduced phagocytosis of aging red cells by red pulp macrophages.

Central tissue-resident macrophages also contribute to the erythroid islands in the bone marrow (the

erythron) that regulate erythroblast differentiation, the final stages of enucleation, and reticulocyte

maturation9-12. However, macrophage colony-stimulating factor (M-CSF)-deficient op/op mice and

CSFR-/- mice, which display reduced F4/80 tissue-resident macrophages, are characterized by hypo-

cellularity of the bone marrow and reduced erythroid burst-forming units (BFU-E)13, which indicates

a major effect of macrophages on HSPC. In addition, erythroblast proliferation in bone marrow was

unchanged in myeloablated mice and the ratios between the different stages during erythroblast

maturation were comparable8. Together, these observations suggest a major role for (tissue-resi-

dent) macrophages in the control of proliferation and/or differentiation of HSPC, which needs to be

investigated.

We have previously shown that erythroblast expansion from unsorted peripheral blood mononuclear

cells (PBMC) is 4- to 15-fold increased compared to CD34+ cells purified from a similar number of

PBMC14. This increased erythroid yield may partly be explained by the presence and differentiation

of CD34- cells in peripheral blood14. However, erythroid differentiation from CD34- PBMC does not

fully account for the increased erythroid yield from unselected PBMC. This suggests that additional

processes underlie the increased erythroid yield from total PBMC compared to CD34+ cells isolated

from PBMC. We hypothesize that PBMC with feeder/stromal cell-like properties can positively influ-

ence HSPC proliferation, differentiation, and/or survival to yield more erythroblasts. A role for these

effector cells during ex vivo culture may also reveal clues to their function in the bone marrow niche. In

this study we showed that human PBMC-derived CD14+ cells, in particular CD14++CD16+ intermediate

monocytes/macrophages, increased the erythroid yield from CD34+ HSPC in co-culture experiments.

Macrophages sustained HSPC that precede the erythroblast stage, which resulted in increased eryth-

roid expansion from CD34+ cells in ex vivo cultures.

Chapter 2

42

Methods

Cell sorting

CD3, CD19, CD14 and CD34 MicroBeads (Miltenyi Biotec, Bergisch Gladbach, Germany) were used

for magnetic-activated cell sorting (MACS) from PBMC (manufacturer’s protocol). Prior to sorting,

monocytes/macrophages were purified from PBMC by counterflow centrifugal elutriation (JE-6B Beck-

man-coulter centrifuge, Beckman Instruments Inc., Palo Alto, CA). Monocyte/macrophage subsets and

hematopoietic precursors were sorted on a FACS-Aria II/III (BD Biosciences, Oxford, UK).

Cell culture

Human cells were cultured in StemSpan (Stem Cell Technologies, Grenoble, France) supplemented

with stem cell factor (SCF, supernatant equivalent to 100ng/ml), erythropoietin (2U/ml, ProSpec, East

Brunswick, NJ, USA), dexamethasone (1mM; Sigma, St. Louis, MO, USA) and cholesterol-rich lipids

(40mg/ml; Sigma) as described elsewhere14,15. Informed consent was given in accordance with the

Declaration of Helsinki and Dutch national and Sanquin internal ethic boards. Conditioned media

were collected from CD14+ cells cultured for 2 days at 5-10x106 cells/8ml, filtered (0.22µm) and stored

at 4°C. Isolated CD34+ cells were cultured with conditioned media diluted 1:2 with fresh culture

medium. Media was replenished every 2 days.

Co-culture experiments

CD34+ cells were co-cultured with purified hematopoietic effector cells using ratios as found in

PBMC (1:100 CD14+ cells; 1:430 CD3+ cells or 1:25 CD19+ cells). CD34+ cells were co-cultured with

CD14++CD16-, CD14++CD16+ or CD14+CD16+ cells (at a ratio of 1:100).

Transwell assays

CD14+ and CD34+ cells were seeded into transwells (0.4µm polyester membrane; Corning, NY, USA)

with CD34+ cells inside the transwell and CD14+ cells in the well (at a ratio of 1:100). Cells were analyzed

after 2-8 days on the flow cytometer.

Colony assays

Colony assays were started with freshly purified, sorted, or cultured cells mixed with methacult

(Medium ColonyGel™ Cell Systems, Troisdorf, Germany). After 14 days, total colony numbers and

colony forming units were manually scored twice using a wide-field microscope (Axiovert 200M; Carl

Zeiss Inc., Thornwood, NY, USA).

Proliferation assays

CD34+ cells were washed in phosphate-buffered saline (PBS) supplemented with 0.1% bovine serum

albumin (BSA, Sigma; PSA). Cells were labeled with 0.5µM carboxyfluoroscein succinimidyl ester (CFSE;

Molecular Probes, Leiden, The Netherlands) for 8 min with gentle shaking at 37°C. Cells were washed

Human CD14+ cells enhance ex vivo HSPC survival

43

in 1% PSA and cultured with or without CD14+ cells (at a ratio of 1:100) for 2 days. CFSE dilution was

measured by flow cytometry.

Flow cytometry

Cells were washed in PBS and re-suspended in 1% PSA. Cells were incubated with primary antibodies

for 30 min at 4°C, washed in PBS and measured on a FACS Canto II or LSR Fortessa (both BD Biosci-

ences) and analyzed using FlowJo software (FlowJo v7.6.4/v10, Ashland, OR, USA). The antibodies are

detailed in the Supplementary Methods. For apoptosis experiments, cells were stained in 100µl 1x

binding buffer with Annexin V-APC (BD Biosciences; 1:200) at 4°C. After 30 min, 100µl binding buffer

and propidium iodide (50mg/ml; Sigma) were added.

Cytospins

Cells (5×105) were cytospun onto glass slides, fixed in methanol, and stained with May-Grünwald-Gi-

emsa stain (manufacturer’s protocol). Images were processed using Adobe Photoshop 9.0 (Adobe

Systems Inc., CA, USA).

Statistical analysis

The Student t-test was used for the statistical analysis. P-values ≤0.05 are considered statistically

significant.

Results

Hematopoietic cell fractions in peripheral blood mononuclear cells

The erythroid yield is significantly higher from total PBMC than from CD34+ cells purified from PBMC14.

To investigate which cells contribute to this phenomenon we first examined the presence, frequency

and erythroid potential of CD34+ and CD34- HSPC and their downstream precursor cells in PBMC.

PBMC contained low numbers of CD34+ cells (0.16±0.08%; Figure 1A) which could be subdivided into

69% HSC/common myeloid progenitors (CMP), 29% granulocyte-monocyte progenitors (GMP) and

1% megakaryocyte-erythroid progenitors (MEP; Figure 1B and Figure S1). Around 22% of the CD34+

cells consisted of CD34+CD38-CD36- long-term HSC (Figure S2A). The erythroblasts expanded and

differentiated from PBMC may potentially also originate from erythroblasts present in PBMC. We,

therefore, evaluated the frequency of erythroblasts in PBMC. T-cells (CD3) and monocytes/macro-

phages (CD14), further denoted as Lineage (Lin), were excluded from the analysis. Lin-CD34+CD36- (A;

HSPC morphology), Lin-CD34+CD36+ (B; blast morphology), Lin-CD34-CD36+ and Lin-CD34-CD36dim (C,D;

blast, basophilic, orthochromatic, polychromatic, reticulocytes) cells were identified in PBMC (Figure

1C,D and Table 1). The Lin-CD34-CD36- population E consisted primarily of the remaining assorted

cells (e.g., B cells, natural killer cells, CD34- HSC and other CD3-CD14-CD34-CD36- cells).

Chapter 2

44

A B C D E1 2

3 1

1

1

1

3

E Lin- Lin-CD117-

Lin-CD38- Lin-CD38-CD117-

i ii

iii iv

CD

34

CD36

vi

Lin- Lin-BAH1.1-

v

A

Cel

l fre

quen

cy (%

)

C D 3 4 +

0 .0

0 .1

0 .2

0 .3

0 .4

B

CD

34

BAH

1.1

BAH

1.1

Day 0 Day 3

SSC

FSC-A FSC-A CD45RA CD45RA

D

CCD36

A B

E D C

SSC

FSC

-H

Lin

CD

34

FSC-A FSC-A FSC-A CD36

i iii ivLiving cells 63.4%

Single cells 97.6%

ii

Lin-19.8%

Figure 1. Defining hematopoietic progenitor cells and frequencies in PBMC. (A) The percentage of CD34+ cells in PBMC measured using flow cytometry (n=8). (B) Representative dot plots of CD34+ cells from panel A stained for CD45RA and BAH1.1 to discriminate between MEP (CD34+BAH1.1+CD45RA-), GMP (CD34+BAH1.1-CD45RA+) and CMP (CD34+BAH1.1-CD45RA-) at day 0 and at day 3 of ex vivo culture. The low abundant MEP population was quantified using a specific MEP marker BAH1.1 (characterization in Figure S1). (C) Gating strategy to show hematopoietic pre-cursors according to CD34/CD36 expression. Living cells were gated (FSC/SSC; i), followed by gating of single cells (FSC-A/FSC-H; ii). A dump-channel containing CD14+ and CD3+ cells, henceforth denoted as Lin, was setup to exclude specific populations (iii). Lin-negative cells are depicted in a CD34/CD36 dot plot (iv) and five distinct populations are indicated (A-E; frequencies in Table 1). (D) May-Grünwald-Giemsa stained cytospins of FACS-sorted populations as indicated in C (iv). (E) Representative CD34/CD36 dot plot as indicated in C (i), supplementing the Lin-negative cells with anti-CD117 (ii), anti-CD38 (iii) or both anti-CD38/anti-CD117 (iv) or BAH1.1 (vi; control in v). FSC(-A/H): forward-scatter(-area/height); SSC: side-scatter; MEP: megakaryocyte-erythroid progenitor; CMP: common myeloid progenitor; GMP: granulocyte-monocyte progenitor; PBMC: peripheral blood mononuclear cells; Lin: lineage.

Human CD14+ cells enhance ex vivo HSPC survival

45

Supplementing lineage depletion with CD117 (c-kit) antibodies resulted in severe reductions of pop-

ulations A and B indicating that the majority of these cells are CD117+ (Figure 1E). Supplementing

lineage depletion with CD38 resulted in loss of population D and reductions in populations A, B, C

and E. Combining CD117 and CD38 in the lineage depletion resulted in complete loss of populations

A, B and D and reductions in C and E (Figure 1E). These results suggest that the CD34+CD36+C-

D38+CD117+ population B is enriched for MEP. This was further confirmed through supplementing

lineage depletion with BAH1.1 which resulted in complete loss of population B and only a modest

loss in population A (Figure 1E). Populations A-E were seeded in colony assays. Populations A and B

contained the majority of the colony formation capacity, whereas populations C, D and E displayed

low to no colony-forming capability (Table 1). This is in line with our previously published data show-

ing that the CD34- population has severely limited colony-stimulating capacity14. Population A can be

discriminated from B by its greater ability to generate granulocyte-monocyte colony-forming units

(CFU-GM), indicating that population A is composed of more immature hematopoietic progenitors

(e.g., HSC, CMP, GMP), whereas population B contains MEP committed to the megakaryocyte-eryth-

roid lineage. In conclusion, the cells in population A are early HSPC, population B is composed of

MEP and cells in transition to become erythroblasts. Populations C and D are mixed pools of cells

that also contain erythroid cells at different stages.

Table 1. Frequency of the five populations identified in Figure 1C on the basis of CD34/CD36 staining in PBMC and respective colony formation capacity. Colonies shown per 500 cells. Results are presented as mean ± SD of three to five separate experiments.

Population Phenotype Frequency CFU-E BFU-E CFU-GM CFU-GEMM Total CFU

PBMC N/A N/A 0.75±0.03 0.5±0.3 0.2±0.1 0.006±0.005 0.8±0.7

A CD34+CD36

-0.14±0.05 320±32 112±91 67±46 4±6 263±281

B CD34+CD36

+0.04±0.05 272±16 152±130 11±14 4±5 303±306

C CD34-CD36

+0.96±0.52 0.1±0.01 0.1±0.01 0.1±0.002 0.1±0.05 0.35±0.05

D CD34-CD36

dim2.04±0.76 0 0 0 0 0

E CD34-CD36

-83.5±20 0 0.1±0.1 0.02±0.03 0.001±0.002 0.1±0.1

PBMC: peripheral blood mononuclear cells; CFU: colony-forming units; BFU-E: erythroid burst-forming units; E: erythroid; GM: granulocyte-monocyte; GEMM: granulocyte-erythrocyte-monocyte-megakaryocyte.

Tracking hematopoietic stem and progenitor cell differentiation to erythroblasts

The Lin-CD34/CD36 flow cytometer readout provides a novel tool to follow the differentiation of HSPC

through the MEP stage to erythroblasts. CD34+ cells were isolated from peripheral blood and cultured

ex vivo for 8 days. Differentiation of CD34+ cells was monitored daily by assessing Lin-/CD34/CD36/

CD71/CD235a membrane expression and GATA1 and erythropoietin receptor (EPOR) mRNA expres-

sion during culture (Figure 2 and Figure S2B,C). CD71 (transferrin receptor) and CD235a (glycophorin

A; GPA) are more classical markers for erythroblasts and subsequent stages of differentiation towards

erythrocytes16, but fail to identify the upstream progenitor progression of HSC towards erythroblasts.

At day 0, cells are primarily CD34+CD36- HSPC (P1) but progress towards CD34+CD36+ MEP (P2,3)

Chapter 2

46

AC

D34

CD34+

PBMC

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7Day 0

CD36

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

P1 P2 P3

P4

B

CD235a

CD

71

Day 1 Day 2 Day 3 Day 4 Day 5 Day 6 Day 7Day 011.1% 1.78% 5.39% 20.2% 49.2% 79.5% 85.4% 83.9%

0.30% 0.07%0.07% 0.22% 0.93% 57% 65.4% 76.6%

CD34+

PBMC

0

20

40

60

80

100

120

0

20

40

60

80

100

120CD34+ culture

Time (days)

0 1 2 3 4 5 6 7

CD

71high/C

D235a

dim(line; %

)Popu

latio

ns 1

-4 (%

of t

otal

)

P1P2P3P4

C D

0

20

40

60

80

100

120

0

20

40

60

80

100

120PBMC culture C

D71

high/CD

235adim

(line; %)

Time (days)

Popu

latio

ns 1

-4 (%

of t

otal

)

P1P2P3P4

0 1 2 3 4 5 6 7

Figure 2. Tracking HSPC differentiation to erythroblasts. CD34+ cells purified from PBMC and total PBMC (n=3) were cultured for 7 days. (A) Dot plots depicting delineated CD34/CD36 populations from a representative donor over time (CD34+ culture: top; total PBMC culture: bottom). The specific stages, hematopoietic stem and progenitor cells (HSPC; P1), megakaryocyte-erythroid progenitors (MEP; P2-P3) and erythroblasts (P4) are indi-cated within the dot plots. Note the progression from stage 1 to 4 during culture. (B) Dot plots depicting CD71/CD235a expression from CD34+ cell (top panels) and total PBMC (bottom panels) cultures over 7 days. Note that CD71highCD235adim cells appear gradually during culture. Histograms showing the percentage of cells in stage P1-4 of panel A for (C) CD34+ cell cultures and (D) total PBMC cultures. The sum of the percentages within P1-4 is set to 100% to better visualize the relative changes during culture. The line depicts the percentage of CD71high-

CD235adim erythroid cells in the total culture. Note that erythroid cells appear faster in the CD34+ cell culture than in total PBMC cultures. This is because CD34+ cell cultures do not contain any other cells than CD34+ cells at the onset of culture in contrast to total PBMC, which contain large numbers of T cells, B cells and CD14+ monocytes/macrophages during the first days of culture. However, these cells are outcompeted by the proliferating erythroid cells after 6-7 days (see also Figure 2B). PBMC: peripheral blood mononuclear cells.

during day 1-5, mirrored by an increase in GATA1 mRNA levels. From day 5 to day 7, the cells finally

become CD34-CD36+ erythroid cells (P4), accompanied by a 100-fold increase in GATA1 and EPOR

mRNA levels compared to levels in CD34+ cells at day 0. CD71highCD235adim erythroid cells start to

appear after 4 days (Figure 2B,C) and are the predominant population by day 7. This coincides with

the up-regulation of GATA1 and EPOR mRNA and markedly increased protein levels of gata1 and the

Human CD14+ cells enhance ex vivo HSPC survival

47

erythroid specific markers α/β-spectrin and band3 (SLC4A1; Figure S2B-D). This pattern of marker

expression is similar in cultures initiated from total PBMC, although the erythroid compartment is

enhanced compared to that from CD34+ cultures (Figure 2C,D). In conclusion, the Lin-CD34/CD36

scheme allows sorting for specific populations of HSPC, MEP and erythroblasts but also for cells

between these specific stages.

Expansion of erythroblasts from sorted peripheral blood mononuclear cell populations

Next, we investigated the contribution of each specific population in Figure 1C to the outgrowth of

erythroid cells from total PBMC (Figure 3A,B). Population A and B have the majority of the proliferative

potential. Eight days after initiation of the cultures, the cells stained negative for CD34, CD38 and

CD41, and positive for CD117, CD71 and CD235a (Figure S2E), indicating a pure erythroblast popu-

lation. To assess the relative contribution of each population to the erythroid expansion capacity of

total PBMC, the absolute number of cells in each fraction was corrected for the frequency with which

the initiating cells were present in PBMC (Table 1). The number of CD71highCD235adim erythroblasts

produced by each population at day 17 of culture was compared to the number of erythroblasts

obtained from total PBMC (Figure 3B). The CD34+ populations A and B accounted maximally for 43.6%

of the total erythroid yield from PBMC, and the CD34- fraction accounted for 5.6% of the erythroid

yield. A relatively small proportion (<3%) of this yield could be attributed to the Lin-CD34-CD36dim

population D in PBMC. Importantly, the total number of erythroblasts from populations A-E did not

recapitulate the erythroid yield from total PBMC.

B

Eryt

hrob

last

s (f

old

chan

ge)

A

Time (days)

Cel

l num

ber (

*106 )

0 5 1 0 1 5 2 0

0 .0 0 1

0 .0 1

0 .1

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0 PBMC

C D 34 + C D 36-

C D 34+ C D 36+

C D 34- C D 36+

C D 34- C D 36d im

C D 3 4- C D 3 6-

0

0,2

0,4

0,6

0,8

1

1,2

Figure 3. Cumulative yield of erythroblasts from sorted precursor populations present in PBMC is inferior to total PBMC erythroid yield. (A) One million cells of each population as defined in Figure 1C were cultured for 17 days. The growth curve depicts the cell number in millions as a function of time in days (n=5). (B) To normalize the yield observed in A to the frequencies of the individual populations observed in PBMC, the cell number at day 17 was multiplied by the frequencies presented in Table 1. The data are shown as a fold change with respect to the erythroblast yield of PBMC (n=5). PBMC: peripheral blood mononuclear cells.

Chapter 2

48

CD14+ monocytes/macrophages enhance erythroblast expansion from CD34+ cells

As we observed an increased yield of erythroblasts from total PBMC compared to CD34+ cells, we

hypothesized that PBMC contain a cell fraction that functions as ´helper/feeder cells´. PBMC mainly

contained CD3+ T cells, CD19+ B cells and CD14+ monocytes (Figure 4A). Depletion of CD14+ cells, but

not CD3+ or CD19+ cells, from PBMC consistently reduced expansion of CD71highCD235adim erythroid

cells by 35-40% (Figure 4B and Figure S3 for isolation/depletion efficiencies). The CD14+ fraction

did not contain cells with expansion capacity, which indicates that hematopoietic progenitors are

not co-depleted with CD14+ cells. Conversely, the addition of CD14+ cells to CD34+ cells increased

the number of erythroid cells (Figure 4C). In addition, co-culture of PBMC-derived CD14+ cells with

CD34+ cells isolated from cord blood or bone marrow also increased the erythroblast yield (Figure

S4A,B). PBMC-derived CD14+ cells expressed CD16, CD4, CD38 and CD163, but not CD117, CD34,

CD3 or CD19 (Figure S4C). The addition of CD3+ and CD19+ cells only marginally effected outgrowth

of erythroblasts from CD34+ cells. Together, the data indicate that CD14+ cells isolated from PBMC

positively influence hematopoiesis/erythropoiesis.

CD14+ monocytes/macrophages do not exert their effect on erythroblasts but increase the

numbers of CD34+ hematopoietic stem and progenitor cells during co-culture

CD14+ cells persisted in culture until day 7 (Figure S4D), at which time erythroblasts are emerging.

CD14+ cells could, therefore, elicit their effect on all stages from HSC to erythroblasts. To test this,

CD14+ cells were added to day 8 cultures existing of erythroblasts only; the expansion of erythro-

blasts was similar in the presence or absence of CD14+ cells (Figure 5A). To examine whether CD14+

cells exert their effect on HSPC prior to the erythroblast stage, PBMC-derived CD14+ and CD34+ cells

were co-cultured for 2 days. This resulted in a 5-fold increase in CD34+ cell numbers compared to

when CD34+ cells were cultured alone (Figure 5B). In addition, CD34+ cell numbers from cord blood

or bone marrow were 1.5-fold increased in the presence of PBMC-derived CD14+ cells (Figure S4E,F).

The CD14+ cell fraction did not contribute to erythroid outgrowth nor did these cells become CD34+

(Figure 5B,E). Population A (Lin-CD34+CD36-) was sorted and cultured in presence or absence of CD14+

cells for 2 days. Cell numbers increased 3-fold in population A and 5-fold in population B (Figure 5C).

However, the ratio between A and B did not increase upon co-culture (Figure S5A,B). These data

suggest that proliferation and/or survival of the HSPC and MEP populations is affected by CD14+ cells.

They predict that depletion of CD14+ cells from PBMC should decrease the total number of CD34+

cells during culture. Indeed, PBMC depleted for CD14+ cells resulted in decreases in populations A

and B within 2 days (Figure 5D and Figure S5C).

Human CD14+ cells enhance ex vivo HSPC survival

49

BAC

ell f

requ

ency

(%)

C D 3 + C D 1 4 + C D 1 9 +

2

4

8

1 6

3 2

6 4

1 2 8

C

02468

1012141618

CD34+ CD34+CD3+

CD34+CD14+

CD34+CD19+

Eryt

hrob

last

s (f

old

chan

ge)

P<0.05

0

0,2

0,4

0,6

0,8

1

1,2

1,4

Eryt

hrob

last

s (f

old

chan

ge) P<0.05 NS

Figure 4. CD14+ cells positively affect erythroblast yield. (A) The percentage of lymphocytes (anti-CD3), mono-cytes/macrophages (anti-CD14) and B cells (anti-CD19) was assessed in PBMC using flow cytometry (n=10). (B) PBMC were depleted for CD14+, CD3+ or CD19+ cells using magnetic beads coupled to specific antibodies (for efficiencies see Figure S3). Erythroid yield was assessed after 8 days of culture and depicted as fold changes with respect to the erythroid yield from total PBMC (n=3). (C) CD34+ cells were isolated from PBMC and co-cultured with isolated CD3+, CD14+ or CD19+ cells. The erythroid yield was assessed after 8 days and depicted as fold changes normalized to the yield from CD34+ cells alone (n=3-6). PBMC: peripheral blood mononuclear cells.

To determine whether co-culture is required or whether the effect is exerted immediately, we seeded

CD34+ cells in colony assays in the presence or absence of CD14+ cells, and with or without 2 days

of co-culture. Without co-culture, the presence or absence of CD14+ cells did not increase colony

numbers. However, co-culture of CD34+ with CD14+ cells resulted in a 5-fold increase in hematopoietic

colony formation compared to CD34+ cell cultures or cells seeded in colony assays at the initiation of

culture (Figure 5E). Note that CD14+ cells do not form any colonies and that co-cultures of CD34+ with

CD14+ cells yielded the same number of colonies as 2-day cultures of total PBMC. CD14+ cells may

increase CD34+ cell numbers in the erythroid biased culture conditions through: (i) enhanced lineage

specification; (ii) induction of cell survival; or (iii) promotion of proliferation. To examine an effect on

lineage specification the number of CFU-GM (myeloid), granulocyte-erythrocyte-monocyte-mega-

karyocyte colony-forming units (CFU-GEMM; mixed) and BFU-E (erythroid) colony-forming cells was

determined. The overall distribution of CFU-GM, CFU-GEMM and BFU-E after 2 days of co-culture

with CD14+ cells was not biased towards erythropoiesis (Figure 5F). This suggests that CD14+ cells do

not push CD34+ cells towards a lineage, but affect the overall yield of HSPC.

Chapter 2

50

B

D

CD

34+

cells

(fol

d ch

ange

)

01234567

CD34+ CD34+CD14+

CD14+

P<0.01

C

AEr

ythr

obla

sts

(x10

6 )

Time (days)0 1 2 3 4 5

0

2

4

6

8

1 0 EBL

EBL C D14+

Cel

l cou

nt (a

bsol

ute

num

bers

)

P<0.01

P<0.01

0

5000

10000

15000

20000

25000

30000

35000

WithoutCD14+

WithCD14+

A

B

0

0,2

0,4

0,6

0,8

1

1,2

PBMC

CD14-PBMC

E

Col

ony

num

ber c

hang

e (r

atio

)

0

0,4

0,8

1,2

1,6

2Day 0 Day2

F

00,20,40,60,8

11,21,4 CFU-GEMM

CFU-GM

BFU-E

Col

ony

form

ing

units

(ra

tio) Day 0 Day2

Cel

l cou

nt (f

old

chan

ge)

Figure 5. CD14+ cells do not affect erythroblasts, however they positively influence HSPC. (A) Erythroblasts were expanded from PBMC and were cultured with or without CD14+ cells for 4 additional days (n=6). (B) CD34+ cells were cultured in the presence or absence of CD14+ cells for 2 days. CD34+ cell numbers were assessed by multiplying the percentage of CD34+ cells by total cell counts. The data are depicted as fold changes normalized to CD34+ cell numbers of CD34+ cell cultures (n=4). Note the absence of CD34+ cells in the CD14+-only condition. (C) Population A, as defined in Figure 1C, was FACS-sorted from CD34+ cells and cultured for 2 days with or without CD14+ cells. Population A and B were quantified and absolute counts are displayed (n=3). (D) PBMC depleted for CD14+ cells were cultured for 2 days. CD34/CD36 populations were delineated and normalized to total PBMC. Note the reduction in population A and B when CD14+ cells were absent (n=3). (E-F) Total PBMC, CD34+ cells and CD34+ with CD14+ cells were subjected to colony assays at initiation (day 0) and after 2 days of culture (day 2). Total colony numbers were normalized to colony numbers from total PBMC (n=4) (E). Quantification of colony-forming units. Total colony number is set to 1 to visualize any bias towards a specific hematopoietic lineage (n=4) (F). EBL: erythroblasts; PBMC: peripheral blood mononuclear cells; CFU: colony-forming units; GEMM: granulocyte-eryth-rocyte-monocyte-megakaryocyte; GM: granulocyte-monocyte; BFU-E: erythroid burst-forming units.

Human CD14+ cells enhance ex vivo HSPC survival

51

The effect of CD14+ cells is independent of cell contact, in particular CD14++CD16+ intermediate

monocytes/macrophages enhance erythroblast expansion

To investigate whether CD14+ cells increase proliferation of CD34+ cells, the CD34+ cells were loaded

with CFSE and CFSE dilution was measured during 2 days of culture. Both in the presence and absence

of CD14+ cells, the CFSE signal in CD34+ cells was reduced to similar levels, indicating comparable

numbers of cell divisions (Figure 6A). When lineage specification and proliferation are similar, CD14+

cells most likely enhance survival of CD34+ cells. Indeed, the presence of CD14+ cells decreased

apoptosis of CD34+ cells (Figure 6B). However, this was not regulated via B-cell CLL/lymphoma2 (BCL2)

or BCL2-like 1 (BCL2L1) anti-apoptotic signaling pathways as mRNA levels of these proteins were

unchanged in co-culture conditions compared to the levels in CD34+ cultures alone (Figure S5D,E).

Monocytes/macrophages express various integrins, membrane-bound ligands of integrins, and other

sequestering membrane proteins. These molecules facilitate macrophage functions in innate immune

responses, and regulate cell-cell contact-mediated signal transduction in CD14+ and associated cells.

We hypothesized that cell contact may be important for CD14+ cells to exert their effect on CD34+

cells. However, a dose-dependent effect of the number of CD14+ cells in a transwell assay supported

the involvement of secreted factors rather than cell-cell contact (Figure 6C). These experiments did

not discriminate between an exclusive effect of CD14+ cells or an interplay between secreted factors

by CD34+ cells that regulate the factor secretion profile of CD14+ cells. To investigate this, condi-

tioned media from cultured CD14+ cells were added to cultures of isolated CD34+ cells. Similar to

CD14+ cells, supernatant increased expansion of erythroid cells, suggesting that the spectrum of

factors secreted by CD14+ cells is not influenced by CD34+ cell-induced signaling. Of note, the reduced

erythroid yield of the conditioned media (Figure 6D) compared to the transwell assays (Figure 6C) is

probably explained by: (i) the half-life of the secreted factors in the conditioned media, rather than

continuously replenished factors in the transwell setting; and (ii) the dilution of the supernatant 1:2

with fresh culture medium. Together, these results show that CD14+ cells exert their effect on HSPC

primarily through secreted factors.

CD14+ monocytes/macrophages from PBMC can be subdivided into three populations based on

CD14 and CD16 expression: classical (CD14++CD16-), intermediate (CD14++CD16+) and non-classical

(CD14+CD16+) monocytes/macrophages (Figure 7A)17. Each subpopulation has its own functional and

molecular characteristics18-20. To investigate whether a specific subset is responsible for the increased

erythroid yield, CD14++CD16- (P1), CD14++CD16+ (P2) or CD14+CD16+ (P3) subpopulations were sorted

and co-cultured with CD34+ cells (Figure 7B). Of note, the percentages of the specific subsets in PBMC

reflect literature values (Figure 7A). Co-culture of CD14++CD16+ cells (P2) or CD14++CD16- cells (P1)

with CD34+ cells resulted in 5-fold and 3-fold increases in erythroblast yield, respectively, compared

to culture of CD34+ cells alone (Figure 7B). Co-culturing CD34+ cells with CD14+CD16+ cells (P3) led to

only a modest increase in erythroid yield, indicating that this population is not involved. Additional

analysis showed that CD14++CD16+ intermediate cells (P2) can be further discriminated from classical

Chapter 2

52

A

CFSE

Cou

nt

FSC

CFSE

Unlabeled control

D0 CD34+

D2 CD34+

D2 CD34+CD14+

MFI

(a.u

.)

D 0C D 3 4 +

D 2C D 3 4 +

D 2C D 3 4 + C D 1 4 +

1 0 0

1 0 0 0

1 0 0 0 0

NS

D

00,5

11,5

22,5

33,5

44,5

CD34+ CD34+ sup

Eryt

hrob

last

s (f

old

chan

ge)

C

Eryt

hrob

last

s (f

old

chan

ge)

0

2.5

5

7.5

10

12.5

15

CD34+ CD34+CD14+ 100x

CD34+CD14+ 50x

CD34+CD14+ 10x

CD34+CD14+ CC

CD14+

B

0

10

20

30

40

50

60

70

80

CD34+ CD34+CD14+

P<0.01

Apop

totic

cel

ls (p

erce

ntag

e)An

nexi

n V

Propidium Iodide

Figure 6. CD14+ cells affect CD34+ cell survival by secreted factors. (A) PBMC-derived CD34+ cells were loaded with CFSE and cultured for 2 days with or without CD14+ cells. The CFSE dilution is displayed as arbitrary units (a.u.) of mean fluorescence intensity (MFI) (n=6). The histogram and dot plot show a representative experiment. (B) CD34+ cells were isolated from PBMC and cultured in a transwell assay with or without CD14+ cells. After 2 days, apoptosis was measured by assaying the cumulative number of annexinV/propidium iodide-positive cells, which is displayed as the percentage of apoptotic cells in the bar graph (n=4). Note the high baseline rate of apop-tosis (60%) of CD34+ short-term cultures on erythropoietin, stem cell factor and dexamethasone as previously described14. The dot plots depict a representative experiment. (C) PBMC-derived CD34+ cells were cultured with or without CD14+ cells in a transwell assay preventing cell-cell contact (using various concentrations of CD14+ cells) or in full contact co-culture (CC). The bar diagram displays the erythroid yield after 8 days of culture nor-malized to the erythroid yield (CD71highCD235adim) in CD34+ conditions without any CD14+ cells (n=4). (D) CD34+ cells isolated from PBMC were cultured for 8 days in conditioned medium as described in Methods. The erythroid yield was normalized to the erythroid yield observed in CD34+ cells (n=3). MFI: mean fluorescence intensity; CFSE: carboxyfluoroscein succinimidyl ester; FSC: forward-scatter; CC: co-culture.

Human CD14+ cells enhance ex vivo HSPC survival

53

and non-classical cells by the combination of CXCR4, CD163, CD169 (Figure 7C) and HLA-DR (data

not shown). All peripheral CD14+ cells are negative for the M2 macrophage marker CD206 (mannose

receptor). CD14+ cells cultured for 2 days showed differentiation towards a CD14++CD16+CD206+

macrophage population that also upregulate tissue residency markers like CXCR4, CD163 and CD169

(Figure 7C and Table S1). Note that CD16 expression is similar to intermediate monocytes freshly

isolated from peripheral blood. Low to no expression of CD209 was observed thereby excluding the

presence of dendritic cells (data not shown).

Discussion

Survival and proliferation of HSPC is crucial to prevent bone marrow failure in vivo, and to enable

expansion of HSPC for cellular therapy in vitro. For future transfusion medicine, the in vitro produc-

tion of erythrocytes is the Holy Grail to provide anemic patients with fully matched erythrocytes,

which requires a maximum expansion capacity during all stages of erythropoiesis. We previously

found that total PBMC cultures display a better expansion of erythroblasts compared to CD34+ cells

isolated from a similar amount of PBMC14. Unraveling the mechanisms that drive this process is

important to optimize in vitro erythroid expansion protocols and potentially unlocks clues of CD34+

cell homeostasis. Here, we show that CD14+ monocytes/macrophages in PBMC secrete factors that

enhance the expansion of CD34+ HSPC. In particular, it is the CD14++CD16+ subset of intermediate

monocytes/macrophages17 that is involved. Interestingly, they affect early CD34+CD36- HSPC and

CD34+CD36+ MEP but not erythroblasts. Besides erythroid expansion from CD34+ cells, also myeloid

colony forming potential increased following co-culture of CD34+ cells with CD14+ monocytes/macro-

phages. Analysis of cell divisions using CFSE labelling showed similar proliferation capacity of CD34+

cells in culture with or without CD14+ cells, indicating that the increased erythroid yield is through

activation of survival pathways.

Tissue-resident macrophages have been found in hematopoietic niches near the HSC where they

provide retention signals to HSPC6. Depletion of macrophages in mice reduces erythropoiesis not

because of impaired erythroblast proliferation per se, but rather as a result of HSPC and erythroblast

mobilization8. This is in agreement with our data showing that CD14+ cells enhance the number of

CD34+ HSPC but not the proliferation rate of human erythroblasts. We showed that CD14+ cells do not

alter lineage commitment or proliferation. Instead, we show that CD14+ cells support CD34+ cell sur-

vival, although this remains to be proven in vivo. Reduced BFU-E formation and hypocellularity of the

bone marrow was observed in M-CSF deficient op/op mice and CSFR-/- mice which are characterized

by reduced F4/80 tissue-resident macrophages13. The combination of retention signals and survival

signals could explain the hematopoietic phenotypes observed in op/op mice and underscores that

macrophages play a role in hematopoietic/erythropoietic homeostasis. Thus, besides being involved

in the retention of HSPC in the bone marrow, we discovered a potential additional role for

Chapter 2

54

A B

0

1

2

3

4

5

6

7

CD34+ CD34+P1

CD34+P2

CD34+P3

Eryt

hrob

last

s (fo

ld c

hang

e)

P<0.05

SSC

CD

16

FSC-A

CD14

P3

P2

P1

Monocytes90.1%

C

C las s ic

a l

Inte

rme d ia

te

N o n c las s ic

a l

D a y2

0 .1

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

C D 1 6

C las s ic

a l

Inte

rme d ia

te

N o n c las s ic

a l

D a y2

1 0 0

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0

C X C R 4

C las s ic

a l

Inte

rme d ia

te

N o n c las s ic

a l

D a y2

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0

C D 1 6 3

C las s ic

a l

Inte

rme d ia

te

N o n c las s ic

a l

D a y2

0 .1

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

C D 1 6 9

C las s ic

a l

Inte

rme d ia

te

N o n c las s ic

a l

D a y2

- 1 0 0

0

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

C D 2 0 6

MFI

(CD1

6)

MFI

(CXC

R4)

MFI

(CD2

06)

MFI

(CD1

69)

MFI

(CD1

63)

Figure 7. In particular CD14++CD16+ monocytes/macrophages positively influence the erythroblast yield. (A) Dot plots representing enrichment of monocytes after elutriation (left). Sorting strategy of the monocyte subsets on the basis of the expression of CD14 and CD16 (right): CD14++CD16- (P1), CD14++CD16+ (P2), and CD14+CD16+ (P3). (B) Sorted subpopulations P1, P2, and P3 were co-cultured with CD34+ cells. After 8 days CD14-CD71highCD235adim cell numbers were counted on the flow cytometer and depicted as fold changes with respect to the erythroid yield from CD34+ cells alone (n=5). (C) Distribution graphs displaying the geometric mean fluorescent intensity of CD16, CXCR4, CD163, CD169 and CD206 on freshly isolated monocyte populations in PBMC and after 2 days of culture (n=3-8). Table S1 shows the statistical significance between each population and marker. FSC(-A): forward-scatter(-area); SSC: side-scatter; MFI: mean fluorescence intensity.

Human CD14+ cells enhance ex vivo HSPC survival

55

macrophages in the survival of CD34+ HSPC during differentiation. An important remaining question is

whether survival and retention of CD34+ cells by the CD14+ monocytes/macrophages are the same, or

originate from independent signal transduction pathways. We observed that the increased survival of

CD34+ cells and subsequent yield of erythroblasts is caused by a secreted factor, whilst the retention

of CD34+ HSPC in the bone marrow was dependent on cell contact through VCAM-1 (CD106)6. This

favors a model in which these two macrophage functions can be separated.

CD14+ cells in peripheral blood are a mixed population of cells that includes phenotypes resem-

bling bone marrow tissue-resident macrophages (CD14+CD163+CD169+)6,8. We find that these cells

are present in the intermediate monocyte population in PBMC. This population also displays the

best HSPC support during co-culture experiments. During (co-)culture these monocytes differen-

tiate towards CD206+ macrophages that further upregulate tissue resident markers like CXCR4,

CD163, CD169 resembling tissue-resident M2 macrophages21,22. It is important to compare this

specific subset and their similarity to bone marrow isolated resident macrophages. The observed

effects are caused by secreted factors and hence the identity of these growth factors is of para-

mount importance to optimize in vitro erythropoiesis and to better understand HSPC survival in

the hematopoietic niches. At present, hematopoietic stem cell transplantation is performed with

total mobilized PBMC rather than purified CD34+ HSPC from PBMC or bone marrow. However, the

effect of different hematopoietic effector cells on HSPC engraftment is unclear. As we found that

CD14+ cells promote survival of CD34+ HSPC, a putative beneficial effect of CD14+ cells in HSPC

survival after engraftment is possible.

Analysis of total peripheral blood and PBMC-derived CD34+ cell cultures on the basis of lineage

depletion (CD3/CD14) and CD34/CD36 during time, revealed a gradual disappearance of CD34 and

appearance of CD36 during HSPC differentiation as recently observed for cord blood CD34+ cells23.

Colony assays confirmed that CD34+CD36- cells were HSPC, CD34+CD36+ cells were highly enriched

for MEP and CD34-CD36+ cells were erythroblasts. This Lin-CD34/CD36 staining provides a novel tool

to follow HSPC differentiation towards the erythroblasts and enables sorting of specific differentia-

tion intermediates. Subsequently, the erythroblasts can be further analyzed during differentiation to

enucleated reticulocytes using previously described protocols16,24. Here, we found that CD14+ cells

provide support to HSPC cells leading to increased survival of CD34+ cells. The observations explain

the increased erythroid expansion from total PBMC compared to CD34+ cells isolated from a similar

amount of PBMC and unravel a novel role for CD14+ cells in survival of hematopoietic progenitors.

Furthermore, we designed a novel tool to follow specific stages during CD34+ differentiation to eryth-

roblasts in adult erythropoiesis using a Lin-CD34/CD36 flow cytometry panel.

Chapter 2

56

Author contributions

EH, EA, FM and MM performed the experiments. FE performed the erythroblast/CD14+ co-culture

experiments. NY performed conditioned medium experiments. AM and EA designed the experiments,

supervised the study and analyzed the data. EA and EH wrote the manuscript. ML and AM co-super-

vised and helped writing the manuscript. The manuscript was critically revised by all authors. The

authors report no potential conflicts of interest.

Acknowledgements

We would like to thank Erik Mul (Sanquin Central Facility) and Gijs van Schijndel (Immunopathol-

ogy; Sanquin research) for technical assistance regarding the isolation of monocytes/macrophages

from PBMC by elutriation and FACS sorting and Thamar van Dijk (Erasmus Medical Center, Rotter-

dam, the Netherlands) for providing GATA1 antibodies. This study was supported by funding from

the Landsteiner Foundation for Blood Transfusion Research Landsteiner (LSBR1141; EA, EH, FE),

Sanquin (PPOC: 12-025; NY, ML, EA), NHLBI (HL116329-01; AM) and the National Cancer Institute

(P01-CA108671; AM), Centro Nazionale Sangue and Associazione Italiana Ricerca sul Cancro (AIRC;

FM, MM).

Human CD14+ cells enhance ex vivo HSPC survival

57

References

1. Kiel MJ, Iwashita T, Yilmaz OH, Morrison SJ. Spatial differences in hematopoiesis but not in stem cells indicate a lack of regional

patterning in definitive hematopoietic stem cells. Dev Biol. 2005;283(1):29-39.

2. Sugiyama T, Kohara H, Noda M, Nagasawa T. Maintenance of the hematopoietic stem cell pool by CXCL12-CXCR4 chemokine

signaling in bone marrow stromal cell niches. Immunity. 2006;25(6):977-988.

3. Lo Celso C, Wu JW, Lin CP. In vivo imaging of hematopoietic stem cells and their microenvironment. J Biophotonics. 2009;2(11):619-

631.

4. Mendez-Ferrer S, Michurina TV, Ferraro F, et al. Mesenchymal and haematopoietic stem cells form a unique bone marrow niche.

Nature. 2010;466(7308):829-834.

5. Nwajei F, Konopleva M. The bone marrow microenvironment as niche retreats for hematopoietic and leukemic stem cells. Adv

Hematol. 2013;2013:953982.

6. Chow A, Lucas D, Hidalgo A, et al. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and

progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011;208(2):261-271.

7. Winkler IG, Sims NA, Pettit AR, et al. Bone marrow macrophages maintain hematopoietic stem cell (HSC) niches and their deple-

tion mobilizes HSCs. Blood. 2010;116(23):4815-4828.

8. Chow A, Huggins M, Ahmed J, et al. CD169(+) macrophages provide a niche promoting erythropoiesis under homeostasis and

stress. Nat Med. 2013;19(4):429-436.

9. Mohandas N, Prenant M. Three-dimensional model of bone marrow. Blood. 1978;51(4):633-643.

10. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythro-

blastic island formation. Blood. 2006;108(6):2064-2071.

11. Rhodes MM, Kopsombut P, Bondurant MC, Price JO, Koury MJ. Adherence to macrophages in erythroblastic islands enhances

erythroblast proliferation and increases erythrocyte production by a different mechanism than erythropoietin. Blood.

2008;111(3):1700-1708.

12. Hanspal M, Golan DE, Smockova Y, et al. Temporal synthesis of band 3 oligomers during terminal maturation of mouse eryth-

roblasts. Dimers and tetramers exist in the membrane as preformed stable species. Blood. 1998;92(1):329-338.

13. Dai XM, Ryan GR, Hapel AJ, et al. Targeted disruption of the mouse colony-stimulating factor 1 receptor gene results in osteo-

petrosis, mononuclear phagocyte deficiency, increased primitive progenitor cell frequencies, and reproductive defects. Blood.

2002;99(1):111-120.

14. van den Akker E, Satchwell TJ, Pellegrin S, Daniels G, Toye AM. The majority of the in vitro erythroid expansion potential resides

in CD34- cells, outweighing the contribution of CD34+ cells and significantly increasing the erythroblast yield from peripheral

blood samples. Haematologica. 2010;95(9):1594-1598.

15. Migliaccio G, Sanchez M, Masiello F, et al. Humanized culture medium for clinical expansion of human erythroblasts. Cell Trans-

plant. 2010;19(4):453-469.

16. Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, Lodish HF. Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to

decreased survival of early erythroblasts. Blood. 2001;98(12):3261-3273.

17. Ziegler-Heitbrock L, Ancuta P, Crowe S, et al. Nomenclature of monocytes and dendritic cells in blood. Blood. 2010;116(16):e74-

80.

18. Zawada AM, Rogacev KS, Rotter B, et al. SuperSAGE evidence for CD14++CD16+ monocytes as a third monocyte subset. Blood.

2011;118(12):e50-61.

19. Schmidl C, Renner K, Peter K, et al. Transcription and enhancer profiling in human monocyte subsets. Blood. 2014;123(17):e90-

99.

20. Wong KL, Tai JJ, Wong WC, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and non-

classical human monocyte subsets. Blood. 2011;118(5):e16-31.

21. Moestrup SK, Moller HJ. CD163: a regulated hemoglobin scavenger receptor with a role in the anti-inflammatory response. Ann

Med. 2004;36(5):347-354.

22. Martinez FO, Gordon S, Locati M, Mantovani A. Transcriptional profiling of the human monocyte-to-macrophage differentiation

and polarization: new molecules and patterns of gene expression. J Immunol. 2006;177(10):7303-7311.

23. Li J, Hale J, Bhagia P, et al. Isolation and transcriptome analyses of human erythroid progenitors: BFU-E and CFU-E. Blood.

2014;124(24):3636-3645.

24. Chen K, Liu J, Heck S, Chasis JA, An X, Mohandas N. Resolving the distinct stages in erythroid differentiation based on dynamic

changes in membrane protein expression during erythropoiesis. Proc Natl Acad Sci U S A. 2009;106(41):17413-17418.

Chapter 2

58

Supplementary methods

Mobilized peripheral blood, cord blood and bone marrow cultures

Mononuclear cells were isolated from bone marrow and cord blood. CD34+ cells from both cord

blood and mobilized peripheral blood were isolated by MACS sorting. CD34+ cells from bone marrow

were FACS sorted. CD34+ cells were co-cultured with PBMC-derived CD14+ cells at a ratio of 1:100.

Media was replenished every two days. Informed consent from donors was given in accordance with

the Declaration of Helsinki.

Quantitative PCR analysis

To assess the relative GATA1 and EPOR mRNA levels during differentiation of CD34+ cells, aliquots

were taken at day 0, 2, 4 and 8 of culture. To assess the relative mRNA levels of BCL2 and BCL2L1

after co-culturing CD34+ cells with CD14+ cells, CD34+ cells were cultured in a transwell with or

without CD14+ cells for 2 days. RNA was extracted with TRIzol reagent (Invitrogen Technologies,

Carlsbad, CA) and complementary DNA was synthesized with random hexamers as described in

the product protocol (Invitrogen). RT-PCR was performed in duplicate with Express SYBR GreenER

reagents (Invitrogen) on the StepOnePlus RT PCR system (Applied Biosystems, Foster City, CA). Values

were normalized using S18 as a reference gene and calibrated relative to expression in CD34+ cells

at day 0 (for GATA1 and EPOR) and CD34+ cells cultured in the absence of CD14+ cells (for BCL2

and BCL2L1). The following primer sets were used: S18 (forward: 5’-GGACAACAAGCTCCGTGAAGA-3’,

reverse: 5’-CAGAAGTGACGCAGCCCTCTA-3’), GATA1 (forward: 5’-ATCACACTGAGCTTGCCACA-3’,

reverse: 5’-ACCAGAGCAGGATCCACAAA-3’), EPOR (forward: 5’-CTCATCCTCGTGGTCATCCT-3’, reverse:

5’-GCTGGAAGTTACCCTTGTGG-3’), BCL2 (forward: 5’-ATGTGTGTGGAGAGCGTCAA-3’, reverse:

5’-CGTACAGTTCCACAAAGGCA-3’) and BCL2L1 (forward: 5’-GCGTGGAAAGCGTAGACAAG-3’, reverse:

5’-TGCTGCATTGTTCCCATAGA-3’).

Western blotting

During CD34+ cell differentiation, aliquots were taken at day 0, 2, 4 and 8 of culture. Cells were pel-

leted, washed once with PBS and lysed (20mM Tris-HCl pH8.0, 138mM NaCl, 10mM EDTA, 100mM

NaF, 1% Nonidet P-40, 10% glycerol; Protease inhibitor cocktail V (Calbiochem, Darmstadt, Germany),

100μM PMSF (Sigma Aldrich; St. Louis, MO; 1:1000) and 1mM sodium orthovanadate (New England

Biolabs, Leusden, the Netherlands)). Protein concentrations were determined by Bradford assay

according to the manufacturer’s protocol (Bio-Rad, Hercules, CA, USA). Lysates were mixed with

Laemmli loading dye (Invitrogen Life Technologies, Breda, the Netherlands) and incubated for 5min

at 95°C. Proteins were separated on a 7.5% SDS-PAGE gel loading equal protein concentrations or

equal cell numbers and subsequently transferred to nitrocellulose membranes (Invitrogen). A 250-

kda precision marker was used as size standards (Invitrogen). Membranes were blocked with 5%

bovine serum albumin (BSA; Sigma) and subsequently incubated for 1hr with primary antibodies;

anti-RhoGDI (Abnova, Taipei, Taiwan; 1:1000), anti-gata1 (a kind gift from Thamar van Dijk, Erasmus

Human CD14+ cells enhance ex vivo HSPC survival

59

Medical Center, the Netherlands; 1:1000), anti-band3 (BRIC170; IBGRL; Bristol, UK; 1:1000) and anti-

α/β-spectrin (Sigma; 1:1000)) in Tris-buffered saline and 0.1% Tween 20 (pH8.0) containing 5% BSA.

Membranes were washed, followed by incubation with IRDye 800-conjugated secondary antibodies

(LI-COR Biosciences, Lincoln, NE, USA; 1:2500-10000). Fluorescence was detected on an Odyssey

Infrared Imaging system using Odyssey V3.0 software (both LI-COR Biosciences).

Flow cytometry

Antibodies used: BD Biosciences: anti-CD3 (PE 1:50; Pacific Blue 1:80), anti-CD14 (PE 1:50; Pacific

Blue 1:150), anti-CD16 (Pacific Blue 1:80), anti-CD19 (PE 1:50; PE-Cy7 1:20), anti-CD34 (FITC 1:10),

anti-CD38 (PE 1:50), anti-CD41 (APC 1:20), anti-CD42 (PE 1:20), anti-45RA (PE 1:10), anti-CD56 (PE 1:20),

anti-CD117 (PE 1:50), anti-CD169 (APC 1:100), anti-CD206 (APC 1:100) and anti-BAH1.1 (PE 1:100; APC

1:10); Miltenyi Biotec: anti-CD14 (APC 1:50; PE 1:50), anti-CD71 (APC 1:150) and anti-CD163 (PE 1:100);

Pelicluster (Amsterdam, The Netherlands): anti-CD4 (PE 1:50), anti-CD16 (FITC 1:100) and anti-CD36

(FITC 1:100); Dako (Glostrup, Denmark): anti-CD41 (PE 1:20) and anti-CD235a (PE 1:100); eBioscience

(Vienna, Austria): CD184 (PE 1:100); IQ Products (Groningen, The Netherlands): anti-CD34 (APC 1:10).

Chapter 2

60

Supplementary data

Table S1. Statistical analysis of marker expression between the different CD14+ populations in PBMC before and after two days of culture (Figure 7C).

P-values t-test CD16 CXCR4 CD163 CD169 CD206

Classical vs. Intermediate 0.00005 0.00317 0.00416 0.17957 N/A

Classical vs. Non-classical 0.00109 0.16771 0.03652 0.23951 N/A

Intermediate vs. Non-classical 0.00225 0.00812 0.01267 0.07984 N/A

Classical vs. Day 2 0.00152 0.03255 0.00460 0.05191 0.01646

Intermediate vs. Day 2 0.03347 0.03387 0.00492 0.05456 0.01714

Non-classical vs. Day 2 0.00378 0.03247 0.00315 0.03645 0.01727

Human CD14+ cells enhance ex vivo HSPC survival

61

A

C

Col

onie

s (3

00 c

ells

inpu

t)

3242

5

29

0

10

20

30

40

50

60

70

80

90non BFU-E

BFU-E

MEP CMP

B

Isot

ype

Isot

ype

CD

41C

D71

Isotype CD235a

Isotype CD42

iv

ii

iii

i

MEP

CMP GMPBAH

1.1

SSC

CD

34

FSC-A FSC-A CD45RA

Figure S1. Characterization of BAH1.1 as a MEP marker. (A) Gating strategy to define CMP, GMP and MEP populations. CD34+ cells were isolated from peripheral blood mononuclear cells (PBMC) and cultured for 4 days in StemSpan supplemented with IL6 (10ng/ml), IL3 (1ng/ml), SCF (100ng/ml) and N-PLATE (50ng/ml; thrombopoi-etin agonist). Cells were analyzed by flow cytometry. Living cells were gated on FSC/SSC, followed by a positive gate for CD34. The CD34+ cells were further separated on the basis of BAH1.1 and CD45RA. CD34+BAH1.1+C-D45RA- cells were FACS-sorted and subjected to colony assays (B) or cultured in specific media (C). (B) Colony assays indicating that CD34+BAH1.1+CD45RA- cells primarily display BFU-E colony forming ability. Note that CD34+BAH1.1-CD45RA-CMP produce both BFU-E and non BFU-E colonies. (C) CD34+BAH1.1+CD45RA- cells were cultured in megakaryopoiesis specific media containing StemSpan supplemented with SCF (100ng/ml), IL1β (1ng/ml), IL6 (10ng/ml) and N-PLATE (50ng/ml) for 4 days followed by 4 days with IL1β (1ng/ml) and N-PLATE (50ng/ml) only (i-ii) or erythropoiesis specific media containing StemSpan supplemented with erythropoietin (2U/ml), stem cell factor (100ng/ml) and dexamethasone (1μM) for 8 days (iii-iv). Cells were analyzed by flow cytometry. Both CD41a+CD42b- megakaryoblasts and CD41a+CD42b+ megakaryocytes were found in megakaryopoiesis specific media (ii; isotype in i) and CD71highCD235adim erythroblasts were found in erythropoiesis specific media (iv; isotype in iii). Together the colony assays and liquid cultures show that BAH1.1 can be used to delineate MEP from other hematopoietic lineages. FSC(-H): forward scatter; SSC: side scatter; MEP: megakaryocyte-eryth-roid progenitor; CMP: common myeloid progenitor; GMP: granulocyte-monocyte progenitor; BFU-E: erythroid burst-forming units.

Chapter 2

62

CD34 CD38 CD117 CD41 CD71 CD235a

AC

D34

Isot

ype

CD

117

CD

38

FSC-A Isotype CD36 CD36

i ii iii iv

Cou

nt

FSC FSC Isotype CD235a

CD

71

II III IVSSC

CD

71

ii iii iv

FSC

-H

i

Isot

ype

FSC-A Isotype CD235aFSC-A

E

C

0,1

1

10

100

1000

0 2 4 80,1

1

10

100

1000

0 2 4 8

Rel

ativ

e ex

pres

sion

(EPO

R)

Rel

ativ

e ex

pres

sion

(GAT

A1)

Time (days)Time (days)

B

D0 2 4 8 M 0 2 4 8

Anti-spectrin

Anti-band3

Anti-gata1

Anti-RhoGDI

Figure S2. Population definitions and marker expression on erythroblasts. (A) CD34+ cells isolated from PBMC were separated using CD36/CD38 and CD36/CD117 antibodies. The dot plot (iii) shows a representative experiment indicating that CD34+ cells (i) can be further separated into 72% CD38+ (short term hematopoietic stem cells; HSC), 22% CD38- (long term HSC) and 4% CD38+CD36+ (megakaryocyte-erythroid progenitors; MEP). CD117 positivity mirrors CD38 expression patterns (iv). The isotype control is displayed in ii. (B-C) RT-PCR analysis of GATA1 (B) and EPOR (C) mRNA levels in CD34+ cells during culture at day 0, 2, 4 or 8 normalized to S18 and relative to CD34+ cells at day 0 (n=3). (D) CD34+ cells from mobilized peripheral blood were cultured for 0, 2, 4, or 8 days and western blot analysis demonstrated increasing spectrin (250kda), band3 (100kda) and gata1 (50kda) protein expression during CD34+ cell differentiation, lanes were loaded according to equal protein concentrations (left) or to equal cell numbers (right). Western blot images show a representative result (n=3). Note that the strictly erythroid marker band3 (SCLa41) is only expressed on day 8. (E) Gating strategy to show erythroblasts according to CD71/CD235a expression. PBMC were cultured for 8 days and erythroblasts were gated based on size (FSC/SSC), single cells (FSC-H/FSC-A) and CD71/CD235a positivity. Representative histograms showing PBMC-derived erythroblasts which do not express CD34, CD38 or CD41 but are CD117+, CD71high and CD235adim. FSC(-H): forward scatter(-height); SSC: side scatter; M: precision marker.

Human CD14+ cells enhance ex vivo HSPC survival

63

A B C

0

20

40

60

80

PBMC CD3-PBMC

CD

3+ c

ells

(%)

CD

3

FSC-A

0

5

10

15

20

25

PBMC CD14-PBMC

CD14

+ ce

lls (%

)CD

14

FSC-A

0

2

4

6

8

10

PBMC CD19-PBMC

CD

19+

cel

ls (%

)C

D19

FSC-A

Isot

ype

CD

34

FSC-A FSC-A

D E

F G

Isot

ype

CD

3

FSC-A FSC-A

FSC-A FSC-A

Isot

ype

CD

14

FSC-A FSC-A

Isot

ype

CD

19

Figure S3. Depletion and isolation efficiencies of CD3+, CD14+ and CD19+ cells. Graphs display the percent-age of (A) CD3+, (B) CD14+ and (C) CD19+ cells before and after depletion by magnetic-activated cell sorting (n=3). Dot plots showing the MACS isolation efficiencies of (D) CD34+, (E) CD3+, (F) CD14+ and (G) CD19+ cells. The dot plots depict representative depletion and isolation experiments. PBMC: peripheral blood mononuclear cells; FSC: forward scatter.

Chapter 2

64

A

0

1

2

3

4

0

1

2

3

BCD34+ from cord blood CD34+ from bone marrow

Eryt

hrob

last

s (fo

ld c

hang

e)

Eryt

hrob

last

s (fo

ld c

hang

e)

0

0,5

1

1,5

0

0,5

1

1,5

2

FE

CD

34+

cells

(fo

ld c

hang

e)

CD

34+

cells

(fol

d ch

ange

)

CD34+ from cord blood CD34+ from bone marrow

P<0.05P<0.0001

P<0.001 P<0.05

C

CD56 CD34 Control APC CD14

CD169 Control PB CD14 CD16

Control PE CD14 CD38 CD4

CD3 CD117 CD19 CD163C

D14

+ ce

lls in

PB

MC

(%)

0

5

10

15

20

0 1 2 3 4 7

Time (days)

D

Figure S4. PBMC-derived CD14+ cells positively affect CD34+ cells isolated from cord blood and bone marrow. (A-B) CD34+ cells isolated from (A) cord blood or (B) bone marrow were cultured in the absence or presence of PBMC-derived CD14+ cells. After 7 (bone marrow) or 8 (cord blood) days, the erythroblasts yield was quantified by analyzing the number of CD71/CD235a positive cells by flow cytometry and depicted in a bar graph (n=3-4). (C) Representative histograms of CD14+ isolated cells using MACS sorting from PBMC stained with markers as indicated in the figure. Note the presence of CD16, CD163, CD4 and CD38 but absence of CD34, CD3, CD19 and CD117 indicating that MACS sorted CD14+ cells are not contaminated with HSPC, lymphocytes or B cells. (D) The percentage of CD14+ cells in total PBMC during time was assessed by flow cytometry (n=5). (E-F) CD34+ cells isolated from cord blood (E) or bone marrow (F) were cultured in the absence or presence of PBMC-derived CD14+ cells. After 2 (cord blood) or 3 (bone marrow) days, the number of CD34+ cells was ana-lyzed by flow cytometry and depicted as bar graphs (n=3-4). The blunted response of the more heterogeneous bone marrow- and cord blood-derived CD34+ populations underscores that co-culture with CD14+ cells mainly improves the survival of early, uncommitted CD34+ cells. PB: pacific blue.

Human CD14+ cells enhance ex vivo HSPC survival

65

CD36

CD

34

CD34+ CD34+CD14+

A

Popu

latio

n B

/A (

ratio

)

0

0,05

0,1

0,15

0,2

0,25

WithoutCD14+

WithCD14+

NSB

C

CD36

PBMCLin-

CD14-PBMCLin-

CD14+Lin-

CD

34

i ii iii

0

0,5

1

1,5

CD34+ CD34+CD14+0

0,5

1

1,5

CD34+ CD34+CD14+

Rel

ativ

e ex

pres

sion

(BC

L2)

Rel

ativ

e ex

pres

sion

(BC

L2L1

)

D ENS NS

Figure S5. CD14+ cells control the survival of HSPC. (A) Representative dot plots belonging to Figure 5C showing CD34/CD36 expression (populations defined in Figure 1C of sorted population A (Lin-CD34+CD36) in the presence or absence of CD14+ cells after 2 days of culture. (B) Ratio depicting the absolute cell numbers in population B (megakaryocyte-erythroid progenitors; MEP) divided by population A (HSPC) from the experiment in Figure 1C (n=3). (C) Representative CD34/CD36 dot plots from total PBMC (i), PBMC depleted for CD14+ cells (ii) and CD14+ cells (iii) belonging to the bar diagraphs of Figure 5D. RT-PCR analysis of (D) BCL2 and (E) BCL2L1 mRNA levels in CD34+ cells during a 2-day culture in transwell assays in the presence or absence of CD14+ cells. The bar graph depicts the relative mRNA levels normalized to the S18 reference gene with the level of mRNA in CD34+ cells without co-culture set to 1 (n=4). PBMC: peripheral blood mononuclear cells; Lin-: lineage-negative cells.

Stage 1

PBMC

Stage 2 Stage 3

Erythroblasts Cultured RBC

3

Large-scale in vitro production of red blood cells from

human peripheral blood mononuclear cells

CHAPTER 3

Esther Heideveld*, Patrick Burger*, Marijke Thiel-Valkhof, Erica Sellink, Steven Heshusius, Anna Visser, Marieke von Lindern, and Emile van den Akker

* These authors contributed equally to this work

Chapter 3

68

Abstract

Transfusion of donor-derived red blood cells is the most common form of cellular therapy. How-

ever, the source depends on donor availability and carries a potential risk of alloimmunization and

blood borne diseases, especially in chronically transfused patients. In vitro cultured, customizable red

blood cells would negate these concerns and introduce precision medicine. Large-scale, cost effective

production depends on reducing cost-drivers and optimizing culture conditions. We adapted our

protocols to GMP culture requirements, which reproducibly provided pure erythroid cultures within

25 days with a 3.4x107 times expansion from peripheral blood mononuclear cells without prior CD34+

isolation. Expanded erythroblasts could be differentiated to CD71dimCD235a+CD44+CD117-DRAQ5-

cultured red blood cells. More than 90% of the cells enucleated and expressed adult hemoglobin

as well as the correct blood group antigens. Deformability and oxygen binding capacity of cultured

red blood cells was comparable to in vivo reticulocytes. This process was compatible with upscaling

using a G-Rex bioreactor with a capacity of 5L per reactor, allowing transition towards clinical studies

and application.

Large-scale in vitro production of red blood cells

69

Introduction

Blood transfusion is the most applied cellular therapy, with more than 80 million transfusion units

administered worldwide each year1. However, it is increasingly more difficult to find suitable donors

because of the antigen variation in modern multi-cultural societies and the risk for novel blood-born

infections, because of global traffic of individuals, including endemic areas. In the past decades,

oxygen-carrier substitutes have shown to be applicable in case of immediate emergency, but they

are less reliable to replace blood transfusions2. Better availability of erythrocytes with rare blood

groups will decrease the risk of alloimmunization in patients with chronic anemia that are in need of

recurrent transfusions, e.g., patients with hemoglobinopathies or with cancer. Cost-effective, large-

scale production of blood group matched erythrocytes may become essential to obtain a degree of

donor independency and to minimize donor-patient blood type variation. In addition, in vitro cultured

erythrocytes can be used as therapeutic delivery systems by releasing carried cargo at specific parts

of the body3. Several groups already cultured enucleated cultured red blood cells (cRBC) in vitro from

cord blood4-7. However, these cells produce fetal hemoglobin with a higher tendency to denature

and to cause membrane damage compared to adult hemoglobin8. A fetal phenotype shortens the

life span of erythrocytes from 120 to 40-90 days9. Moreover, cord blood as a source is limited, and

it is therefore imperative to investigate alternative sources to culture red blood cells expressing

adult hemoglobin. We have previously shown that enucleated cRBC can be generated starting from

peripheral blood mononuclear cells (PBMC) isolated from buffy coats10, 11. Not only are PBMC a better

accessible source than CD34+ cells isolated from cord blood, we have also shown that the erythroid

yield from PBMC is 10-15 fold increased compared to CD34+ cells isolated from a similar amount of

PBMC, because CD14+ cells present in PBMC increase the survival of CD34+ hematopoietic stem and

progenitor cells (HSPC)10, 11. Furthermore, using PBMC and thus dispensing the need to purify CD34+

cells simplifies the protocol and avoids expensive isolation procedures.

One transfusion unit contains about 2x1012 erythrocytes, reflecting the high requirement for eryth-

roblast expansion to obtain sufficient numbers of cRBC. Previous attempts to culture the required

number of enucleated cRBC from CD34+ cells isolated from PBMC were hampered by low expansion

or poor enucleation12, 13. Expansion of CD71highCD235adim erythroblasts can be prolonged by exploiting

the cooperative action of erythropoietin (EPO), stem cell factor (SCF) and glucocorticoids in a serum/

plasma-free environment10, 11, 14, while differentiation is induced by increasing concentrations of EPO

and dispensing SCF and glucocorticoids. Here, we describe a three stage GMP-grade culture protocol

using culture dishes or G-Rex bioreactors, both with high expansion and enucleation to generate

PBMC-derived cRBC. This three-stage culture protocol can be used for large-scale GMP-grade pro-

duction, yielding >90% enucleated reticulocytes with adult hemoglobinization.

Chapter 3

70

Methods

Cell culture

Human peripheral blood mononuclear cells (PBMC) were purified by density separation using Ficoll-

Paque (manufacturer protocol). Informed consent was given in accordance with the Declaration of

Helsinki and Dutch National and Sanquin Internal Ethic Boards. PBMC were cultured at 5-10x106

cells/ml (CASY® Model TCC; Schärfe System GmbH, Reutlingen, Germany) in HEMA-Def medium10,

15 supplemented with erythropoietin (2U/ml; ProSpec, East Brunswick, NJ), human recombinant

stem cell factor (100ng/ml; ITK Diagnostics BV, Uithoorn, The Netherlands), dexamethasone (1μM;

Sigma, St. Louis, MO) and 0.1% ultra-clean human serum albumin (cHSA; kindly provided by San-

quin Plasma Products, Amsterdam, The Netherlands), referred to as expanding medium (EM). Only

during the first two days of culture IL3 was added (1ng/ml; Miltenyi Biotec, Bergisch Gladbach,

Germany). Media was replenished every two days. Around day 6, when the first erythroblasts arose,

the cells were expanded in EM by maintaining cells at 1-2x106 cells/ml. EM cultures were compared

to alternative cell systems by replacing cHSA for 2.5% Albuman® (Sanquin, Amsterdam, The Neth-

erlands), 0.1% detoxified HSA16, recombinant HSA (Akron Biotech, Boca Raton, FL, USA), AB+ pool

plasma (Sanquin) or erythroblast cultures in StemSpan as previously described11. Differentiation

of erythroblasts was induced in differentiation medium (DM) containing HEMA-Def supplemented

with erythropoietin (10U/ml), 5% Omniplasma (Omnipharma S.A.L., Beirut, Lebanon), holotrans-

ferrin (700µg/ml; Sanquin) and heparin (5U/ml; LEO Pharma BV, Breda, The Netherlands). At day

two half a medium change was performed. At day 5, storage components were added and media

was refreshed (half) every two days until fully differentiated at day 10-12 of differentiation. Of note,

after obtaining a pure erythroblast populations, cells were either maintained in culture dishes as

described above, or transferred to a G-Rex system (WilsonWolf, stPaul, MN, USA). For reticulocyte

filtration see Supplementary Methods.

Flow cytometry

Cells were washed and resuspended in HEPES buffer supplemented with 1% HSA. Cells were incu-

bated with primary antibodies for 30min at 4C, measured on FACS Canto II or LSRFortessa (both BD

Biosciences, Oxford, UK) and analyzed using FlowJo software (FlowJo v10, Ashland, OR, USA). Thiazole

orange (TO; Sigma) was used to stain RNA as described previously17 and DRAQ5 (1:2500; Abcam,

Cambridge, UK) was used to stain nuclei. Antibodies are listed in Supplementary Methods.

RBC deformability

RBC deformability was measured by the Automated Rheoscope and Cell Analyzer (ARCA), in which

cells were elongated by shear flow in a viscous medium between parallel glass plates as described

previously18. Advanced imaging algorithms were used to locate cells and to calculate cellular deforma-

bility. All experiments were performed at a shear stress of 10 Pa, and for each experiment more than

3000 cells were judged and grouped in 30 bins according to increasing elongation or cell projection

Large-scale in vitro production of red blood cells

71

area (as a measure of membrane surface area). Both the extent of elongation (major cell radius

divided by minor cell radius) and the area (in mm2) was plotted against the normalized frequency of

occurrence.

HPLC

Culture lysates were prepared and stored at -80°C prior to analysis as described previously19. In short,

hemoglobin separation was performed within 5 days after drawing the blood by high performance

cation exchange liquid chromatography (HPLC) on Waters Alliance 2690 equipment (Waters, Milford,

MA, USA). The protocol consisted of a 30min elution over a combined 20-200mM NaCl and pH7.0-

6.6 gradient in 20mM BisTris/HCl and 2mM KCN. A PolyCAT A 100/4.6mm, 3mm, 1500Å column was

purchased (PolyLC, Columbia, MD, USA).

Coomassie

Ghosts (erythroid membranes) were generated as described before10, subjected to SDS-PAGE gels

(Bio-Rad, Hercules, CA, USA) and total proteins stained with Coomassie. In short, proteins were fixed in

30% ethanol, 2% (v/v) phosphoric acid ON. Wash 2 times 10min in 2% phosphoric acid and equilibrate

for 30min in 2% phosphoric acid containing 18% ethanol and 15% (w/v) ammonium sulphate. Stain by

diluting Coomassie Blue G-250 slowly in water from 0.2% to a final concentration of 0.02% (0.2mg/ml).

Cytospins

Cells were cytospun using Shandon Cytospin II (Thermo Scientific), dried and fixed in methanol.

Cells were stained with benzidin in combination with the Differential Quik Stain Kit (PolySciences,

Warrington, PA) (manufacturers protocol). Slides were dried, subsequently embedded in Entellan

(Merck-Millipore) and covered with a coverslip. Images were taken (Leica DM-2500; Germany).

Results

Large-scale erythroblast expansion from PBMC using a G-Rex bioreactor and GMP-grade

medium

We previously showed that a pure erythroblast population can be expanded from the progenitor pool

present in total PBMC without a CD34+ isolation step10, 11. Differentiation of these progenitor pools

in total PBMC to erythroblasts (stage I) and expansion of the erythroblast pool itself (stage II) both

depend on EPO, SCF and glucocorticoid agonist dexamethasone. The combination of these factors are

driving differentiation towards and expansion of erythroid cells, here termed expansion medium (or

EM), in which dexamethasone functions to delay onset of terminal differentiation of erythroblasts to

erythrocytes10, 14, 20, 21. However, this expansion potential was not properly assessed nor adjusted to a

completely defined serum/plasma-free GMP-grade culture system. We first tested distinct sources of

albumin. EM contains ultra-clean human serum albumin obtained after plasma fractionation (cHSA),

Chapter 3

72

which resulted in a significantly increased erythroid expansion potential compared to cultures in

which cHSA was replaced by plasma or Albuman®(Figure 1A and Figure S1A). The addition of plasma

or Albuman® resulted in a low erythroblast yield and premature differentiation of erythroblasts, indi-

cated by a loss of CD71 expression in conjunction with increased CD235a expression (Figure S1A)22,

23. On the contrary, erythroblasts cultured in the presence of cHSA or detoxified HSA (dHSA) showed

a significantly increased expansion potential with limited spontaneous differentiation (Figure 1A and

FigureS1A). In addition, the erythroid expansion in EM supplemented with cHSA, dHSA or xeno-free

recombinant HSA is comparable (Figure S1B). Of note, PBMC contain primarily T-cells, myeloid effector

cells and B-cells and only ca. 0.16% CD34+ HSPC that are capable of differentiating into erythroid

cells10, 24. Therefore, expansion curves using PBMC as a starting material show a drop in expansion

between day 0 and day 5, caused by a loss of these non-proliferating immune cells11.

Because upscaling cRBC for routine cRBC production in culture dishes is practically impossible and

expensive, a G-Rex bioreactor was used allowing high density cultures25, 26. We previously showed that

adherent CD14+ cells present in PBMC increase the erythroid yield by increasing CD34+ cell survival11,

however, the G-Rex bioreactor does not support adherent cells and thus compromises stage I yield

(data not shown). Therefore, cultures started from PBMC in stage I were performed in culture dishes

until a pure erythroblast population was obtained around day 7-10 of culture in EM supplemented

with cHSA. Subsequently, cultured erythroblasts were either transferred to the G-Rex system or

retained in culture dishes and could be maintained for at least 26 days of culture (Figure 1B). Transfer

to a G-Rex bioreactor briefly delayed expansion of erythroblast cultures, which caught up to a similar

expansion rate between day 15 and 25 of culture. Although the G-Rex bioreactor yielded slightly

less cells (3.2x106 vs. 3.3x107), less donor variation was observed. Erythropoiesis can be staged from

CD71highCD235dim erythroblasts (P1) to enucleated CD71-CD235+ reticulocytes (P4), with intermediate

stages in which cells first increased CD235a expression, then reduced CD71 expression (Figure 1C)23.

During erythroblast expansion in stage II, erythroblasts gained expression of CD235a from day 7 to

21 resulting in a pure population of CD71highCD235+ pro-erythroblasts after 10-14 days of expansion,

which could be expanded for at least 26 days (Figure 1C-E). Note that at day 7 of culture, erythroblasts

highly expressed CD71 and were CD235adim validating that this GMP-grade medium yield similar

cells as previously reported using commercial StemSpan media (Figure S1C)10, 11. Interestingly, cells

expanded in the G-Rex system maintained a less differentiated state for a prolonged culture period,

as shown by delayed CD235a upregulation and decreasing CD71 expression during expansion in

Figure 1E. In conclusion, a defined GMP-grade medium enabled erythroblast cultures culture from

a mixed PBMC cell pool with significantly delayed onset of spontaneous differentiation, resulting in

a large expansion potential. In addition, this system allowed erythroblast expansion in both culture

dishes and a G-Rex bioreactor.

Large-scale in vitro production of red blood cells

73

A B

C

D

Day 9 Day12 Day14 Day16

E

7 9 1 2 1 4 1 6 2 0 2 3 2 60

2 0

4 0

6 0

8 0

1 0 0

C u ltu re d is h

D a y s in E M c u ltu r e

Ce

lls

ing

ate

(%)

P 1

P 2

P 3

P 4

7 9 1 2 1 4 1 6 2 0 2 3 2 60

2 0

4 0

6 0

8 0

1 0 0

G -R e x

D a y s in E M c u ltu r e

Ce

lls

ing

ate

(%)

P 1

P 2

P 3

P 4

Day20 Day23 Day26

Culture dish

G-Rex

Day 7

0 1 0 2 0 3 01 0 -11 0 01 0 11 0 21 0 31 0 41 0 51 0 61 0 71 0 81 0 9

D a y s in E M c u ltu r e

Ex

pa

ns

ion

(fc

)

C u ltu re d is hG -R e x

*

0 1 0 2 0 3 01 0 -2

1 0 -1

1 0 0

1 0 1

1 0 2

1 0 3

1 0 4

1 0 5

1 0 6

D a y s in E M c u ltu r e

Ex

pa

ns

ion

(fc

)

c H S A

A lb u m and H S A

P la s m a

*

Culture dish

G-Rex

Figure 1. Efficient expansion of erythroblasts in plasma/serum-free GMP-grade medium. (A) PBMC were cultured towards erythroblasts in GMP-grade medium (EM) for 26 days. Graph showing the erythroblast expan-sion curve upon culture in the presence of ultra-clean HSA (cHSA or EM) or upon replacing cHSA with detoxified HSA (dHSA), Albuman® or plasma. Cell counts at day 0 were normalized to 1 PBMC at the start of culture. Mean ± SD (unpaired t-test, *P<0.05; n=4). (B) PBMC were cultured in culture dishes until a pure erythroblast population was obtained (at day 7). Erythroblasts were further expanded in a G-Rex bioreactor and compared to erythroblast expansion in culture dishes. Mean ± SD (two-way ANOVA, * P<0.05; n=4). (C-E) Representative dot plots (C) and quantification of multiple cultures (D-E) showing CD71 and CD235 progression during erythroblast expansion in culture dishes or the G-Rex bioreactor (P1, CD71highCD235a-; P2, CD71highCD235a+; P3, CD71dimCD235a+; P4, CD71-CD235a+).

Chapter 3

74

cRBC cultured in a G-Rex culture system fully enucleate in the absence of SCF

Erythroblast differentiation is induced by removing SCF and dexamethasone and increasing the EPO

concentration 5-fold. In addition, the holotransferrin concentration was increased and the differ-

entiation medium (DM) was supplemented with 5% Omniplasma and heparin to prevent medium

from clotting. Erythroblast differentiation is characterized by i) a transient proliferation burst with

decreased cell-cycle time resulting in a reduced cell volume22, 23, ii) hemoglobinization, iii) erythroid

specific protein expression of e.g., blood group antigens, and iv) enucleation. During the first days of

differentiation, in particular in culture dishes, proliferation was observed followed by cell growth arrest

(Figure 2A). Note that the number of cells after the initial short proliferation burst did not decrease,

suggesting no negative effect on culture viability. In contrast, cells differentiated in a G-Rex bioreactor

did not show increased expansion. During erythroblast differentiation, both erythroid cells cultured

in dishes and the G-Rex system gained expression of CD235a+ and lost expression of CD71, although

erythroblasts differentiated faster in culture dishes (Figure 2B-D). Note that CD71 expression was

upregulated for two days coinciding with the hemoglobinization stages after which it was downreg-

ulated and cells became negative (Figure S2A,B). In addition, differentiation was accompanied with a

decrease in cell size (FSC-A), an initial increase of KIT expression at day 1 followed by a sharp down

regulation. Furthermore, CD44 was progressively reduced in expression as reported before10. The

data indicates that the first 24-72 hours in erythroblast differentiation mark an important transition

stage.

At blood banks, erythrocytes are stored in media that contain specific components that protect

the viability, which were not present in the defined media. Therefore, a specific storage component

solution was added at day 5 of differentiation, when the first reticulocytes arose in culture. This

increased the number and frequency of enucleated cRBC, particularly in the G-Rex bioreactor (data

not shown). Nonetheless, initial differentiation experiments in the G-Rex system yielded low numbers

of enucleated cells compared to culture dishes (Figure S3A). We noticed that residual expansion

medium during the first days of differentiation negatively affected enucleation. In the G-Rex system,

upon initiating differentiation, approximately 90% EM could be replaced with DM in the G-Rex system

compared to 100% in culture dishes. Replacing 90% of the culture medium resulted in 56% enucle-

ation after 12 days of differentiation, while 84,6% enucleation was observed upon complete medium

replacement (Figure S3B). To further investigate whether this was caused by residual SCF or dexa-

methasone, SCF was removed from the culture medium two days prior to differentiation in culture

dishes, which resulted in a 1.5-fold increase in enucleation (56% vs. 85%) after 12 days of differen-

tiation (Figure S3C). A further reduction of the concentration dexamethasone (from 1µM to 10nM)

two days prior differentiation and a complete medium replacement at the start of differentiation

did not affect enucleation (Figure S3D). This indicates that residual SCF at the start of differentiation

negatively affect enucleation at the end of differentiation. Enucleation was observed from day 5

onwards, resulting in >90% enucleation after 12 days of differentiation in culture dishes and almost

85% enucleation in the G-Rex system (Figure 2E and Figure S3E).

Large-scale in vitro production of red blood cells

75

A

0 5 1 0 1 51 0 -1

1 0 0

1 0 1

D a y s in D M c u ltu r e

Ex

pa

ns

ion

(fc

)C u ltu re d is hG -R e x

*****

***********

*

BDay 1 Day 2 Day 5 Day 6

Day 7 Day 8 Day 9 Day 12

Culture dish

G-Rex

Culture dish

G-Rex

Figure 2. Differentiation of erythroblasts in culture dishes or a G-Rex bioreactor. (A) A pure population of erythroblasts was obtained from culture dishes. Erythroblast expansion and differentiation was either continued in culture dishes or G-Rex. Erythroblasts were differentiated (day 0 of differentiation) in DM for 12 days. Graph showing the expansion curve of erythroblasts in culture dishes or G-Rex in DM upon differentiation initiation. Mean ± SD (Two-way ANOVA, *P<0.05, ***P<0.001, ****P<0.0001; n=4). (B-D) Representative dot plots (B) and quantification of multiple cultures (C-D) showing expression of CD71 and CD235a (P1, CD71highCD235a-; P2, CD71highCD235a+; P3, CD71dimCD235a+; P4, CD71-CD235a+) during differentiation in culture dishes or the G-Rex bioreactor. (E) Enucleation of erythroid cells during differentiation in culture dishes and a G-Rex system. Mean ± SD (unpaired t-test, n=4). (F) Ratio of nuclei versus cRBC during differentiation in culture dishes or G-Rex (<1 means more cRBC than nuclei). Mean ± SD (unpaired t-test, *P<0.05, **P<0.01; n=3-4). (G) Enucleation of differ-entiating erythroblasts at day 10 of differentiation shown by flow cytometry using DRAQ5 and cytospin images stained with benzidine before and after leucoreduction filter.

Chapter 3

76

E F

GPre-filter Post-filter

6 7 8 9 1 0 1 1 1 2 1 30 .0

0 .5

1 .0

1 .5

2 .0

2 .5

D a y s o f d if fe re n t ia t io n

Ra

tio

nu

cle

i/c

RB

C C u ltu re d is h

G -R e x*

**

* *

0 1 2 5 6 7 8 9 1 20

2 0

4 0

6 0

8 0

1 0 0

C u ltu re d is h

D a y s in D M c u ltu r e

Ce

lls

ing

ate

(%)

P 1

P 2

P 3

P 4

0 1 2 5 6 7 8 9 1 20

2 0

4 0

6 0

8 0

1 0 0

G -R e x

D a y s in D M c u ltu r e

Ce

lls

ing

ate

(%)

P 1

P 2

P 3

P 4

C D

0 5 1 0 1 50

2 0

4 0

6 0

8 0

1 0 0

D a y s o f D M c u ltu r e

En

uc

lea

ted

ce

lls

(%) C u ltu re d is h

G -R e xns

Figure 2. Continued

Large-scale in vitro production of red blood cells

77

Figure 2F showed slight differences in the ratio nuclei/cRBC between culture dishes and a G-Rex bio-

reactor due to the stability of the cultured reticulocytes. The flow cytometry data and cytospin images

revealed that cRBC still contained pyrenocytes and some enucleated cells at the end of differentiation

(Figure 2G and Figure S3E). To purify reticulocytes, remaining nuclei and nucleated cells were filtered

using neonatal leucoreduction filters. This resulted in 99% removal of nuclei and 99% removal of

remaining nucleated cells resulting in a virtually pure population of enucleated cRBC (Figure 2G).

Of note, only G-Rex derived cRBC were further purified using a filter. These data indicate that in the

absence of SCF and dexamethasone erythroblasts can be differentiated to fully enucleated cRBC in

both the G-Rex system and culture dishes.

cRBC resemble mature reticulocytes that maintain blood group antigen expression

Figure 2G revealed a spheroid morphology of filtered cRBC, indicative of a reticulocyte population.

This indicates that cRBC were not fully mature erythrocytes. Nevertheless, expression of the major

membrane proteins α/β-spectrin, Band 3, protein 4.1, protein 4.2 and GAPDH were comparable

between cRBC from culture dishes and peripheral blood RBC (Figure 3A). Reticulocytes released

from the bone marrow have a low deformability, which increases during maturation towards eryth-

rocytes27. Both cRBC cultured in culture dishes and the G-Rex system showed a deformability index

recapitulating late peripheral blood reticulocytes (Figure 3B). Furthermore, cell pellets from cRBC

turned dark red and HPLC data showed that cRBC both derived from G-Rex and culture dishes

mainly express adult hemoglobin (HbA1 73.4% G-Rex vs. 46.4% dish; HbA2 2.1% G-Rex vs. 0.5% dish),

low levels of HbF (8.8% G-Rex vs. 4.1% dish) and some HbH (2.4% G-Rex vs. 4.3% dish) (Figure 3C,D).

Note that only G-Rex cRBC were purified using a filter before performing deformability assays and

HPLC analysis. In addition, hemoglobin oxygen association and dissociation rates at various oxygen

tension in cRBC derived from G-Rex were similar to peripheral blood RBC (Figure 3E). Blood group

antigen expression should be similar between RBC and cRBC that originate from the same donor28,

29. Blood group expression of the cRBC was assessed by flow cytometry and compared to the original

donor RBC. Donors negative for specific blood groups on erythrocytes remained negative after cul-

ture and blood group expression was in complete agreement with the original donors (Table 1). The

functional similarities between cRBC and peripheral blood RBC indicate that the three-stage culture

model using culture dishes or a G-Rex bioreactor that we described here (Figure 4) could be used

for large-scale cRBC cultures.

Chapter 3

78

A B

C D

E

B

AA

B

Shearstress

1 2 30 .0

0 .5

1 .0

1 .5

2 .0

D e fo rm a b iliy (A /B )F

req

ue

nc

y

E a r ly re t ic u lo c y te

L a te re tic u lo c y te

c R B C c u ltu re d is h

c R B C G -R e x

0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0-2,0

5,0

10,0

15,0

20,0

25,0 kweek 28-april UV_VIS_1mAU

min

1 - H

bH -

2,57

42

- 3,7

003

- 4,5

19

4 - H

bF -

6,29

05

- 7,3

82

6 - 8

,683

7 - 9

,205

8 - H

bA1

- 10,

287

9 - H

bA2

- 12,

767

WVL:415 nm

cRBC G-Rex

0,0 5,0 10,0 15,0 20,0 25,0 30,0 35,0-5,0

10,0

20,0

30,0

45,0 39 Dish UV_VIS_1mAU

min

1 - 2

,030 2

- HbH

- 2,

451

3 - 2

,886

4 - 3

,653

5 - 3

,974

6 - 5

,060

7 - H

bF -

6,93

8

8 - 8

,362

9 - 9

,702

10 -

HbA

1 - 1

1,20

9

11 -

HbA

2 - 1

3,60

0

WVL:415 nm

cRBC culture dish

Figure 3. cRBC are highly similar to donor RBC. (A) Ghosts from cRBC and peripheral blood RBC were lysed and subjected to SDS-PAGE. Gel was stained with Coomassie and depicts the most abundant proteins in erythrocyte membranes. (B) Graph showing the deformability under shear stress of RNAhighCD71high early reticulocytes (blue), RNA-CD71dim late reticulocytes (black) as described by17 and cRBC in culture dishes before filtering (red) or G-Rex after filtering (green). (C-D) Expression of hemoglobin variants as determined by HPLC in cRBC in culture dishes before filtering (C) or G-Rex after filtering (D). Note that cRBC cell pellets turned red after differentiation for 12 days. (E) Oxygen association and dissociation curve for peripheral blood erythrocytes (teal and red) and cRBC cultured in a G-Rex bioreactor (blue and green).

Large-scale in vitro production of red blood cells

79

Table 1. Blood group analysis of peripheral blood RBC (RBC) and cultured RBC (cRBC) using flow cytometry.

Blood group RBC1

RBC2

RBC3

cRBC1

cRBC2

cRBC3

A - - + - - +

B - - - - - -

RhD + + - + + -

RhE - + - - + -

RhC + - - + - -

Rhe + + + + + +

Rhc + + + + + +

K - - - - - -

k + + + + + +

Kp-a - - - - - -

Kp-b + + + + + +

Lu-a - - - - - -

Lu-b + + + + + +

P1 + - + + - +

S + - - + - -

s + + + + + +

Fya + - + + - +

Fyb - + + - + +

Jka - + - - + -

Jkb + - + + - +

M + - + + - +

N + + + + + +

Discussion

The potential to culture red blood cells for transfusion purposes has been recognized for a long time,

however, current attempts were still hampered by a low yield and incomplete differentiation to fully

biconcave erythrocytes30-38. We previously described a culture method to obtain pure erythroblasts

from PBMC10. Here, we optimized our culture protocol for GMP conditions, and describe a three-

stage culture system that allowed 107-fold expansion of erythroblasts derived from PBMC after 26

days of culture. The majority of the existing large-scale red blood cell cultures starts from purified

CD34+ cells derived from PBMC, cord blood or bone marrow39. Here, we initiate cultures from total

PBMC and dispense with expensive CD34+ cell isolation procedures. Thereby we take advantage

of not only all CD34+ progenitors, but also CD34- hematopoietic progenitors present in blood that

Chapter 3

80

- IL3- EPO- SCF- Dex

- EPO- SCF- Dex

Phase I Phase II Phase IIIStart of culture Expansion Differentiation

- EPO- SCF- Dex

- EPO- 5% plasma- Holotransferrin

Day 0 Day 2 Day 7-10 Day 0 Day12

Culture dish

G-Rex

cRBC

PBMC

Figure 4. Progression of erythroid cells during culture. Overview of the three-phase erythroid culture system using culture dishes or a G-Rex bioreactor.

have the capacity to differentiate towards erythroid cells10, 11. In addition, we recently showed that

CD14+ cells in PBMC increase CD34+ cell survival and thereby increase the erythroid yield compared

to CD34+ cells isolated from a similar amount of PBMC, showing another advantage to start cRBC

cultures from total PBMC11. The expansion curves shown here have been normalized to the input of

1 PBMC. However, only 0.16% of the total PBMC pool is a CD34+ HSPC11. Normalization to the actual

amount of CD34+ cells allows to compare our results to previously published cRBC culture protocols

starting from CD34+ cells and revealed an erythroblast fold expansion of 2x1010 per CD34+ cell within

26 days. In addition, we showed that these erythroblasts can be further differentiated to cRBC with

>90% enucleation, express adult hemoglobin and display a blood group expression and deformability

similar to donor RBC. This process was compatible with upscaling for clinical studies and applications

using a G-Rex bioreactor. The G-Rex bioreactor allowed for 3.2x106 expansion per PBMC, or 2x109 per

CD34+ cell. Moreover, less donor variation was observed compared to culture dishes and remaining

nuclei and enucleated cells after differentiation could be filtered to obtain a pure cRBC population.

This GMP-grade culture protocol uses glucocorticoids to induce erythroblasts proliferation whilst

inhibiting differentiation, as has been reported before10, 15, 20, 21, 40, 41. The addition of glucocorticoids

also supports erythropoiesis by differentiating peripheral blood monocytes or CD34+ cells to eryth-

roid supporting macrophages during culture from PBMC24, 42. In contrast to our plasma/serum-free

expansion, most large-scale red cell culture protocols use glucocorticoid agonists in combination

with serum and/or plasma during expansion. Here, we showed that the addition of plasma causes

premature differentiation of erythroblasts and negates the differentiation inhibiting effect of gluco-

corticoids4, 6, 43-45. Increased spontaneous differentiation upon addition of plasma during erythroblast

Large-scale in vitro production of red blood cells

81

expansion may be due to additional growth factors or other plasma components. Thus to ensure a

good expansion of erythroblasts using the combination of EPO, SCF and dexamethasone, neither

plasma nor serum should be added. This may proof to be important to reach the amount of cRBC

required for transfusion, but also to expand a workable number of pure erythroid cells from small

aliquots of blood from specific anemic patients. We have recently shown that 3ml of blood is sufficient

to culture enough cells to reprogram erythroblasts to induced pluripotent stem cells46. Dispensing

the necessity to first isolate CD34+ cells, which is less efficient from small amounts of blood, helps to

facilitate easy implementation in other laboratories and to study erythropoiesis in patients of which

limited sample volumes are available.

Here, we found that the isolation and manufacturing process of human albumin used as a media

supplement critically influences the erythroid expansion potential. Using ultra-clean, detoxified or

recombinant additive-free HSA deprived of bound substances (e.g., lipids, protein and steroids) sig-

nificantly increased the erythroid expansion potential. Albumin binds a multitude of substances

including proteins, metabolites and fatty acids, including toxins, drugs and other therapeutics. These

components that bind albumin potentially quench the albumin-binding-capacity47-49. Interestingly, the

process to manufacture Albuman®, which is clinically used to regulate blood circulation, includes a

step to saturate and thus restrain the binding potential of albumin, rendering it mostly inert. Replacing

cHSA in EM by Albuman® reduced erythroblast expansion potential, which could be reverted by char-

coal and ion exchanger treatment. This suggests that the binding and/or transport function of albumin

is important to ensure continued erythroblast expansion. The identification of albumin-binding factors

during erythroid cultures may be of interest, as it might help understanding culture inhibiting factors

that are intrinsically produced by erythroid cells and that are sequestered by albumin. Interestingly,

adding 5% Omniplasma during terminal differentiation in stage III, increases enucleation from 20-25%

to more than 90% (data not shown). Therefore, future research should focus on identifying the

components in plasma that are key to this increased enucleation potential. Once identified, these

components could be added at optimal concentrations during terminal erythroblast differentiation.

In 2011, Timmins et al. demonstrated an ultra-high yield of cultured red blood cells with >90% enucle-

ation from cord blood-derived CD34+ cells in the absence of plasma7. This reported erythroid yield was

similar to our serum/plasma-free PBMC-derived erythroid expansion. However, >90% enucleation

during terminal differentiation using our differentiation protocol was only recapitulated in presence

of plasma. Whether this reflects a difference in cord blood versus adult erythroid cultures remains

unknown but important to investigate.

Previously, we reported that cultures starting from frozen PBMC are similar with respect to erythroid

yield compared to fresh PBMC50. In addition, erythroblasts that are expanded at stage II can also

be readily frozen and defrosted (data not shown). This introduces a degree of flexibility concerning

the actual production and storage of eventual products that arise from erythroid cultures. The use

of adult PBMC also facilitates the availability of starting material and introduces the possibility to

Chapter 3

82

culture cRBC from the patient’s own cells. This is important considering alloimmunization caused by

blood group mismatches and matching from cord blood derived erythrocytes may be complicated. In

addition, with the advances of using therapeutic amounts of erythrocytes loaded with specific cargo,

using the patient’s own cells is also less complicated compared to unmatched blood, cord blood or

immortalized cell lines like hIPSC or immortalized erythroblasts51. Using adult peripheral blood as

starting material for cRBC cultures is an important step towards precision medicine. The synchro-

nized cultures contain young reticulocytes that have a theoretical lifespan of about 120 days. Chronic

anemia patients that receive blood transfusions every two months may benefit from transfusions

with in vitro cultured long-lived erythrocytes, potentially increasing the time between transfusions

and thereby reducing the costs. Currently we are working towards a clinical trial that will allow to test

the in vivo lifespan of transfused PBMC-derived cRBC.

Author Contributions

PB, MT-V, ES, EH, SH and AV performed the experiments. EA, EH, and PB designed the experiments,

analyzed the data and wrote the manuscript. ML contributed to the experiment design and writing

of the manuscript. The manuscript was critically revised by all authors. The authors declare no com-

peting financial interests.

Acknowledgements

We are grateful to Wilson Wolf Manufacturing (Saint Paul, MN, USA) for providing the G-Rex biore-

actors. In addition, we would like to thank Rob van Zwieten, Martijn Veldthuis and Jeffrey Berghuis

(Sanquin, dept. Blood Cell Research) for the technical assistance and data acquisition regarding the

HPLC data and deformability assays and we would also like to thank the Central Facility of Sanquin

for their assistance regarding FACS analysis. This work was supported by grants from the Landsteiner

Foundation (LSBR:1141; EA, EH), ZONMW (40-41400-98-1327; PB, MT-V, ES); ZONMW-TOP (40-00812-

98-12128; SH) and Sanquin (PPOR:15-30; AV).

Large-scale in vitro production of red blood cells

83

References

1. Carson JL, Grossman BJ, Kleinman S, et al. Red blood cell transfusion: a clinical practice guideline from the AABB*. Ann Intern

Med. 2012;157(1):49-58.

2. Palmer AF, Intaglietta M. Blood substitutes. Annu Rev Biomed Eng. 2014;16:77-101.

3. Yoo JW, Irvine DJ, Discher DE, Mitragotri S. Bio-inspired, bioengineered and biomimetic drug delivery carriers. Nat Rev Drug

Discov. 2011;10(7):521-535.

4. Giarratana MC, Kobari L, Lapillonne H, et al. Ex vivo generation of fully mature human red blood cells from hematopoietic stem

cells. Nat Biotechnol. 2005;23(1):69-74.

5. Miharada K, Hiroyama T, Sudo K, Nagasawa T, Nakamura Y. Efficient enucleation of erythroblasts differentiated in vitro from

hematopoietic stem and progenitor cells. Nat Biotechnol. 2006;24(10):1255-1256.

6. Giarratana MC, Rouard H, Dumont A, et al. Proof of principle for transfusion of in vitro-generated red blood cells. Blood.

2011;118(19):5071-5079.

7. Timmins NE, Athanasas S, Gunther M, Buntine P, Nielsen LK. Ultra-high-yield manufacture of red blood cells from hematopoietic

stem cells. Tissue Eng Part C Methods. 2011;17(11):1131-1137.

8. Kleihauer E, Braun H, Betke K. [Demonstration of fetal hemoglobin in erythrocytes of a blood smear]. Klin Wochenschr.

1957;35(12):637-638.

9. Pearson HA. Life-span of the fetal red blood cell. J Pediatr. 1967;70(2):166-171.

10. van den Akker E, Satchwell TJ, Pellegrin S, Daniels G, Toye AM. The majority of the in vitro erythroid expansion potential resides

in CD34(-) cells, outweighing the contribution of CD34(+) cells and significantly increasing the erythroblast yield from peripheral

blood samples. Haematologica. 2010;95(9):1594-1598.

11. Heideveld E, Masiello F, Marra M, et al. CD14+ cells from peripheral blood positively regulate hematopoietic stem and progenitor

cell survival resulting in increased erythroid yield. Haematologica. 2015;100(11):1396-1406.

12. Boehm D, Murphy WG, Al-Rubeai M. The potential of human peripheral blood derived CD34+ cells for ex vivo red blood cell

production. J Biotechnol. 2009;144(2):127-134.

13. Griffiths RE, Kupzig S, Cogan N, et al. Maturing reticulocytes internalize plasma membrane in glycophorin A-containing vesicles

that fuse with autophagosomes before exocytosis. Blood. 2012;119(26):6296-6306.

14. von Lindern M, Zauner W, Mellitzer G, et al. The glucocorticoid receptor cooperates with the erythropoietin receptor and c-Kit

to enhance and sustain proliferation of erythroid progenitors in vitro. Blood. 1999;94(2):550-559.

15. Migliaccio G, Sanchez M, Masiello F, et al. Humanized culture medium for clinical expansion of human erythroblasts. Cell Trans-

plant. 2010;19(4):453-469.

16. Beug H, Steinlein P, Bartunek P, Hayman MJ. Avian hematopoietic cell culture: in vitro model systems to study oncogenic trans-

formation of hematopoietic cells. Methods Enzymol. 1995;254:41-76.

17. Ovchynnikova E, Aglialoro F, Bentlage AEH, et al. DARC extracellular domain remodeling in maturating reticulocytes explains

Plasmodium vivax tropism. Blood. 2017;130(12):1441-1444.

18. van Zwieten R, van Oirschot BA, Veldthuis M, et al. Partial pyruvate kinase deficiency aggravates the phenotypic expression of

band 3 deficiency in a family with hereditary spherocytosis. Am J Hematol. 2015;90(3):E35-39.

19. van Zwieten R, Veldthuis M, Delzenne B, et al. Hemoglobin analyses in the Netherlands reveal more than 80 different variants

including six novel ones. Hemoglobin. 2014;38(1):1-7.

20. Leberbauer C, Boulme F, Unfried G, Huber J, Beug H, Mullner EW. Different steroids co-regulate long-term expansion versus

terminal differentiation in primary human erythroid progenitors. Blood. 2005;105(1):85-94.

21. von Lindern M, Deiner EM, Dolznig H, et al. Leukemic transformation of normal murine erythroid progenitors: v- and c-ErbB act

through signaling pathways activated by the EpoR and c-Kit in stress erythropoiesis. Oncogene. 2001;20(28):3651-3664.

22. Socolovsky M, Nam H, Fleming MD, Haase VH, Brugnara C, Lodish HF. Ineffective erythropoiesis in Stat5a(-/-)5b(-/-) mice due to

decreased survival of early erythroblasts. Blood. 2001;98(12):3261-3273.

23. Koulnis M, Pop R, Porpiglia E, Shearstone JR, Hidalgo D, Socolovsky M. Identification and analysis of mouse erythroid progenitors

using the CD71/TER119 flow-cytometric assay. J Vis Exp. 2011(54).

24. Heideveld E, Hampton-O’Neil LA, Cross SJ, et al. Glucocorticoids induce differentiation of monocytes towards macrophages that

share functional and phenotypical aspects with erythroblastic island macrophages. Haematologica. 2018;103(3):395-405.

25. Vera JF, Brenner LJ, Gerdemann U, et al. Accelerated production of antigen-specific T cells for preclinical and clinical applications

using gas-permeable rapid expansion cultureware (G-Rex). J Immunother. 2010;33(3):305-315.

26. Bajgain P, Mucharla R, Wilson J, et al. Optimizing the production of suspension cells using the G-Rex “M” series. Mol Ther Methods

Clin Dev. 2014;1:14015.

Chapter 3

84

27. Waugh RE, Mantalaris A, Bauserman RG, Hwang WC, Wu JH. Membrane instability in late-stage erythropoiesis. Blood.

2001;97(6):1869-1875.

28. Bony V, Gane P, Bailly P, Cartron JP. Time-course expression of polypeptides carrying blood group antigens during human

erythroid differentiation. Br J Haematol. 1999;107(2):263-274.

29. Southcott MJ, Tanner MJ, Anstee DJ. The expression of human blood group antigens during erythropoiesis in a cell culture system.

Blood. 1999;93(12):4425-4435.

30. Migliaccio AR, Whitsett C, Migliaccio G. Erythroid cells in vitro: from developmental biology to blood transfusion products. Curr

Opin Hematol. 2009;16(4):259-268.

31. Migliaccio AR, Masselli E, Varricchio L, Whitsett C. Ex-vivo expansion of red blood cells: how real for transfusion in humans? Blood

Rev. 2012;26(2):81-95.

32. Migliaccio AR, Grazzini G, Hillyer CD. Ex vivo generated red cells as transfusion products. Stem Cells Int. 2012;2012:615412.

33. Migliaccio AR, Palis J. Blood in a dish: In vitro synthesis of red blood cells. Drug Discov Today Dis Mech. 2011;8(1-2):e3-e8.

34. Shah S, Huang X, Cheng L. Concise review: stem cell-based approaches to red blood cell production for transfusion. Stem Cells

Transl Med. 2014;3(3):346-355.

35. Kim HO. In-vitro stem cell derived red blood cells for transfusion: are we there yet? Yonsei Med J. 2014;55(2):304-309.

36. Larochelle A. Generation of red blood cells in vitro: monitoring the process for improved efficiency. Cytotherapy. 2013;15(9):1043-

1045.

37. Migliaccio AR, Whitsett C, Papayannopoulou T, Sadelain M. The potential of stem cells as an in vitro source of red blood cells for

transfusion. Cell Stem Cell. 2012;10(2):115-119.

38. Singh VK, Saini A, Tsuji K, Sharma PB, Chandra R. Manufacturing blood ex vivo: a futuristic approach to deal with the supply and

safety concerns. Front Cell Dev Biol. 2014;2:26.

39. Stella CC, Cazzola M, De Fabritiis P, et al. CD34-positive cells: biology and clinical relevance. Haematologica. 1995;80(4):367-387.

40. Carotta S, Pilat S, Mairhofer A, et al. Directed differentiation and mass cultivation of pure erythroid progenitors from mouse

embryonic stem cells. Blood. 2004;104(6):1873-1880.

41. Migliaccio G, Di Pietro R, di Giacomo V, et al. In vitro mass production of human erythroid cells from the blood of normal donors

and of thalassemic patients. Blood Cells Mol Dis. 2002;28(2):169-180.

42. Falchi M, Varricchio L, Martelli F, et al. Dexamethasone targeted directly to macrophages induces macrophage niches that

promote erythroid expansion. Haematologica. 2015;100(2):178-187.

43. Neildez-Nguyen TM, Wajcman H, Marden MC, et al. Human erythroid cells produced ex vivo at large scale differentiate into red

blood cells in vivo. Nat Biotechnol. 2002;20(5):467-472.

44. Kupzig S, Parsons SF, Curnow E, Anstee DJ, Blair A. Superior survival of ex vivo cultured human reticulocytes following transfusion

into mice. Haematologica. 2017;102(3):476-483.

45. Zhang Y, Wang C, Wang L, et al. Large-Scale Ex Vivo Generation of Human Red Blood Cells from Cord Blood CD34(+) Cells. Stem

Cells Transl Med. 2017;6(8):1698-1709.

46. Varga E, Hansen M, Wust T, von Lindern M, van den Akker E. Generation of human erythroblast-derived iPSC line using episomal

reprogramming system. Stem Cell Res. 2017;25:30-33.

47. Klammt S, Brinkmann B, Mitzner S, et al. Albumin binding capacity (ABiC) is reduced in commercially available human serum

albumin preparations with stabilizers. Z Gastroenterol. 2001;39 Suppl 2:24-27.

48. Klammt S, Mitzner SR, Stange J, et al. Improvement of impaired albumin binding capacity in acute-on-chronic liver failure by

albumin dialysis. Liver Transpl. 2008;14(9):1333-1339.

49. Kragh-Hansen U, Chuang VT, Otagiri M. Practical aspects of the ligand-binding and enzymatic properties of human serum

albumin. Biol Pharm Bull. 2002;25(6):695-704.

50. Masiello F, Tirelli V, Sanchez M, et al. Mononuclear cells from a rare blood donor, after freezing under good manufacturing

practice conditions, generate red blood cells that recapitulate the rare blood phenotype. Transfusion. 2014;54(4):1059-1070.

51. Shi J, Kundrat L, Pishesha N, et al. Engineered red blood cells as carriers for systemic delivery of a wide array of functional probes.

Proc Natl Acad Sci U S A. 2014;111(28):10131-10136.

Large-scale in vitro production of red blood cells

85

Supplementary methods

Reticulocyte filtration

Cultured red blood cells were purified using a 60ml neonatal leukoreduction filter (Haemonetic BV,

Breda, The Netherlands). The filter was primed with IMDM containing 4% Omniplasma. Cultured cells

were filtered followed by a 2 times wash with 25ml priming medium. Cultured red blood cells were

resuspended in IMDM containing 4% Omniplasma. Cells were centrifuged at 1600rpm for 5 min in

order to obtain packed red blood cells.

Flow cytometry

Antibodies used: Acris (Herford, Germany): anti-CD235a (FITC 1:400; PE 1:400); BD Biosciences:

anti-CD117 (PE 1:50), anti-CD235a (PE 1:150); eBioscience (Vienna, Austria): anti-CD44 (APC 1:150);

Miltenyi Biotec: anti-CD71 (APC 1:100; VioBlue 1:200), anti-CD235a (VioBlue 1:200); and Life technolo-

gies (Carlsbad, CA, USA): anti-CD117 (FITC 1:100). The following antibodies were used for blood group

analysis: IBGRL research products (Bristol, UK): anti-A (Bric145; 1:1000), anti-B (BGRL2; 1:500), anti-M

(clone 6A7; 1:5000), anti-N (clone BRIC157; 1:1000); Sanquin Products (Amsterdam, The Netherlands):

anti-RhC (clone MS24; RBC 1:1000, cRBC undiluted), anti-Rhc (clone MS35; 1:200), anti-RhE (clone

MS260; RBC 1:80, cRBC undiluted), anti-Rhe (clone MS21/63; RBC 1:100, cRBC undiluted), P1 (clone

P3NIL100; 1:500), Kp-a (undiluted), kp-b (undiluted), Jk-a (MS15; undiluted), Jk-b (MS8; undiluted), Fy-a

(P3TIM; undiluted), Fy-b (undiluted), Lu-a (undiluted), Lu-b (undiluted); Merck-Millipore (Darmstadt,

Germany): anti-K (MS59; 1:50), anti-k (P3A118OL67; 1:100), anti-s (P3Y326Bn5; 1:400). Ortho Clinical

Diagnostics (Buckinghamshire, UK): anti-S (MNS3; undiluted). Anti-A, B, N, M were detected with an

anti-mouse IgG FITC-labeled antibody (Life Technologies; 1:200) was used and anti-C, c, E, e, Jk-a,

Jk-b, and P1 with an anti-human IgM PE-labeled antibody (Southern Biotech, Birmingham, AL; 1:500).

The remaining blood groups were detected with an anti-human IgG PE-labeled antibody (Southern

Biotech; 1:200).

Chapter 3

86

Supplementary data

A

BDay 7 Day 19

C

Plasma Albuman dHSA cHSA

0 5 1 0 1 5 2 01 0 -1

1 0 0

1 0 1

1 0 2

1 0 3

1 0 4

1 0 5

D a y s in E M c u ltu r e

Ex

pa

ns

ion

(fc

)

c H S A

rH S Ad H S A

ns

Figure S1. The presence of plasma/serum affects erythroblast expansion. (A) Representative dot plots of erythroblasts at day 12 of culture in EM in the presence of ultra-clean HSA (cHSA) or EM in which cHSA was replaced by plasma, Albuman® or detoxified HSA (dHSA) corresponding to Figure 1A (n=3-4). Note that the percentage of immature erythroblasts (P1) is 3-4 times higher in cHSA and dHSA compared to plasma and Albu-man®. This suggests that cHSA and dHSA prolong the immature state of erythroid cultures. (B) Erythroblasts cultured in EM for 18 days in the presence of cHSA, dHSA or recombinant HSA (rHSA) showed a similar expansion potential. Mean ± SD (two-way ANOVA; n=2). (C) Erythroblasts at day 7 and 19 of expansion in StemSpan, as described previously11.

Large-scale in vitro production of red blood cells

87

A

B

0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 20

2 0 0 0 0

4 0 0 0 0

6 0 0 0 0

8 0 0 0 0

1 0 0 0 0 0

1 2 0 0 0 0

C D 4 4

D a y s in d iffe re n t ia t io n

MF

I(n

orm

ali

zed

toIg

G)

0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 20

1 0 0 0 0

2 0 0 0 0

3 0 0 0 0

4 0 0 0 0

C D 7 1

D a y s a fte r d if fe re n tia t io n

MF

I(n

orm

ali

zed

toIg

G)

0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 20

2 0 0 0 0

4 0 0 0 0

6 0 0 0 0

8 0 0 0 0

1 0 0 0 0 0

C D 2 3 5 a

D a y s a fte r d if fe re n tia t io n

MF

I(n

orm

ali

zed

toIg

G)

0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 20

2 0 0 0

4 0 0 0

6 0 0 0

8 0 0 0

C D 1 1 7

D a y s a fte r d if fe re n tia t io n

MF

I(n

orm

ali

zed

toIg

G)

0 1 2 3 4 5 6 7 8 9 1 0 1 1 1 20

5 0 0 0 0

1 0 0 0 0 0

1 5 0 0 0 0

F S C -A

D a y s a fte r d if fe re n tia t io n

MF

I(n

orm

ali

zed

toIg

G)

Differentiation

Figure S2. Marker expression during erythroblast differentiation in culture dishes. (A) Erythroblasts were cultured in EM for 8 days and subsequently differentiated for 12 days in DM. Representative overlaying histograms showing the FSC-A during differentiation and the expression levels of CD71, CD235a, CD117 and CD44 analyzed by flow cytometry. (B) Graphs corresponding to panel A showing the mean fluorescence intensity (MFI; n=4).

Chapter 3

88

A B

C D

U n wa s h e d

Wa s h e d

0

2 0

4 0

6 0

8 0

1 0 0

En

uc

lea

ted

ce

lls

(%)

**

+ S C F-S

C F0

2 0

4 0

6 0

8 0

1 0 0

En

uc

lea

ted

ce

lls

(%)

****

0 5 1 0 1 50

2 0

4 0

6 0

8 0

1 0 0

D a y s o f D M c u ltu r e

En

uc

lea

ted

ce

lls

(%) C u ltu re d is h

G -R e x*

****

*****

1 0 0 0 n MD e x

1 0 n MD e x

0

2 0

4 0

6 0

8 0

1 0 0

En

uc

lea

ted

ce

lls

(%)

ns

EDay 0 Day 5 Day 10 Day 12

Culture dish

G-Rex

Figure S3. Enucleation of cRBC derived from culture dishes or a G-Rex bioreactor. (A) Graph showing the enucleation of erythroid cells during differentiation after supplementation into DM to initiate differentiation in a G-Rex bioreactor, compared to culture dishes corresponding to Figure 2E. Mean ± SD (paired t-test, *P<0.05, **P<0.01, ***P<0.001; n=4). (B) Erythroblasts expanded and differentiated in culture dishes with 90% (unwashed) or 100% (washed) replacement of expansion medium before differentiation medium. Mean ± SEM (paired t-test, **P<0.01; n=4). (C) Two days prior to differentiation in DM, SCF was either removed from or maintained in EM and enucleation was analyzed after 12 days of differentiation. Mean ± SEM (paired t-test, ****P<0.0001; n=4). (D) Two days prior to differentiation 1000nM or 10nM dexamethasone was added to the expansion medium. Expansion medium was replaced with differentiation medium and cells were analyzed for enucleation after 12 days. Mean ± SEM (paired t-test; n=4). (E) Representative dot plots of erythroblasts during differentiation in DM in culture dishes or G-Rex based on the expression of DRAQ5.

4

Glucocorticoids induce differentiation of monocytes

towards macrophages that share functional and phenotypical aspects with erythroblastic

island macrophages

CHAPTER 4

Esther Heideveld, Lea A. Hampton-O’Neil, Stephen J. Cross, Floris P.J. van Alphen, Maartje van den Biggelaar, Ashley M. Toye, and Emile van den Akker

Haematologica 2018 Mar;103(3):395-405

Chapter 4

92

Abstract

The classical central macrophage found in erythroblastic islands plays an important role in erythroblast

differentiation, proliferation and enucleation in the bone marrow. Convenient human in vitro models

to facilitate the study of erythroid-macrophage interactions are desired. Recently, we demonstrated

that cultured monocytes/macrophages enhance in vitro erythropoiesis by supporting hematopoietic

stem and progenitor cell survival. Herein, we describe that these specific macrophages also support

erythropoiesis. Human monocytes cultured in serum-free media supplemented with stem cell factor,

erythropoietin, lipids and dexamethasone differentiate towards macrophages expressing CD16,

CD163, CD169, CD206, CXCR4 and the phagocytic TAM-receptor family. Phenotypically, they resemble

both human bone marrow and fetal liver resident macrophages. This differentiation is dependent on

glucocorticoid receptor activation. Proteomic studies confirm that glucocorticoid receptor activation

differentiates monocytes to anti-inflammatory tissue macrophages with a M2 phenotype, termed

GC-macrophages. Proteins involved in migration, tissue residence and signal transduction/receptor

activity are upregulated whilst lysosome and hydrolase activity GO-categories are downregulated.

Functionally, we demonstrate that GC-macrophages are highly mobile and can interact to form clus-

ters with erythroid cells of all differentiation stages and phagocytose the expelled nuclei, recapitulating

aspects of erythroblastic islands. In conclusion, glucocorticoid-directed monocyte differentiation to

macrophages represents a convenient model system to study erythroid-macrophage interactions.

Modeling human erythroid-macrophage interactions

93

Introduction

In human bone marrow (BM) and fetal liver (FL), the production of erythrocytes through erythropoi-

esis occurs on erythroblastic islands1,2. These erythroblastic islands consist of a central macrophage

surrounded by erythroid cells at different stages of terminal differentiation and support proliferation,

differentiation and phagocytose the extruded nuclei (or pyrenocytes) of erythroid cells2-6. Chow et

al. described that mouse CD169+ (SIGLEC1) BM resident macrophages display a dual role promoting

erythropoiesis and retention of hematopoietic stem and progenitor cells (HSPC)7,8. Their absence

leads to mobilization of HSPC, reduced BM erythropoiesis and the inability to properly respond to

anemia7-10. It is, however, unclear whether CD169 identifies different macrophage populations or

indicates an intrinsic dual role for the same tissue macrophage. FL macrophages that are unable to

interact with erythroblasts due to disruption of the retinoblastoma tumor suppressor gene in mice

lead to embryonic death as erythroblasts fail to enucleate11. These data show that in vivo, macro-

phages are important in regulating erythropoiesis in adults and during development.

Previously, we found that blood-derived monocytes induced to differentiate using stem cell factor

(SCF), erythropoietin (EPO) and glucocorticoids enhance in vitro erythropoiesis by supporting HSPC

survival12. These macrophages display a tissue-resident profile expressing CD14 (lipopolysaccharide

[LPS]-receptor), CD16 (FcgRIII), scavenger receptor CD163, CD169, CD206 (mannose receptor), CXCR4

and minimal expression of dendritic cell-specific intercellular adhesion molecule 3-grabbing non-in-

tegrin (DC-SIGN)12. We hypothesized that these cultured monocyte-derived macrophages may have

a similar role as mouse CD169+ macrophages in both hematopoiesis and erythropoiesis. This would

provide an easy-to-use in vitro human model system to mimic erythroblastic islands allowing for the

study of functional interactions between macrophages and erythroid cells, which is currently limited

to harvesting BM or involves genetic modification13. A better understanding of the mechanism(s)

through which human macrophages interact and regulate erythroblast maturation and enucleation is

important in order to understand the pathology of erythropoietic disorders, such as erythrocytosis in

polycythemia vera or erythrophagocytosis in several types of hemolytic anemia, as well as to improve

in vitro erythroid differentiation protocols for erythrocyte production14,15.

In mice BM, erythroblasts are bound to macrophages via interactions between integrin-α4β1 on

erythroblasts and VCAM1 on macrophages, and blocking these molecules disrupts erythroblastic

islands16. Chow et al. described human BM macrophages as also express VCAM1. However, Ulyanova

et al. has shown that Vcam-/- mice do not display an erythroid phenotype during homeostasis or

phenylhydrazide-induced stress17. During terminal differentiation, erythroblasts enucleate resulting

in reticulocytes and pyrenocytes. The latter are also still encapsulated by plasma membrane. In mice,

clearance of pyrenocytes occurs via TAM-receptors on the central macrophages that recognize and

bind phosphatidylserine (PS) exposed on pyrenocytes resulting in phagocytosis in a protein S-de-

pendent manner18,19. The TAM-receptor family of tyrosine kinases (TYRO3, AXL, and MERTK) play an

Chapter 4

94

important role in the phagocytic ability of macrophages as triple knock-out mice fail to clear apoptotic

cells in multiple tissues. These mice develop normally, but eventually develop autoimmunity, such as

systemic lupus erythematosus (SLE)20. This is in line with studies showing that SLE has been associ-

ated with failure of macrophages to phagocytose apoptotic cells and pyrenocytes in both human and

mice21-24. In addition, anemia is found in about 50% of SLE patients; Toda et al. showed that embryos

suffer from severe anemia caused by failure of macrophages to phagocytose pyrenocytes25. These

data indicate that macrophages are essential during all stages of erythropoiesis, including enucle-

ation, and display inherent features that are indispensable to the functionality of these macrophages.

Herein, we show that peripheral blood monocytes can be differentiated to erythropoiesis supporting

macrophages that interact with erythroid cells, phagocytose pyrenocytes and phenotypically resemble

human CD169+ BM and FL macrophages.

Methods

Human materials

Human blood, BM and FL mononuclear cells were purified by density separation, following manufac-

turer’s protocol. Regarding blood, informed consent was given in accordance with the Declaration

of Helsinki, the Dutch National and Sanquin Internal Ethic Boards, and by the Bristol Research Ethics

Committee (REC; 12/SW/0199). Following informed consult, adult BM aspirates were obtained from

the sternum of patients undergoing cardiac surgery, and approved by the Medical Ethical Review

Board of the AMC (MEC:04/042#04.17.370). Fetal tissues (week 15-22) were obtained from elective

abortions contingent on informed consent and approval by the Medical Ethical Commission of the

Erasmus University Medical Center Rotterdam (MEC-2006-202).

Cell culture

CD14 and CD34 MicroBeads (Miltenyi Biotec, Gladbach, Germany) were used for cell isolation from

peripheral blood. CD14+ monocytes were cultured at 1.5-3x106 cells/well (CASY® Model TCC, Schärfe

System GmbH, Reutlingen, Germany) in a 12-well plate as described12. Cells were treated with 1-20μM

mifepristone (Sigma-Aldrich, Munich, Germany) directly after isolation or 4-24 hours after three days

of culture. CD34+ cells were differentiated towards erythroblasts12, with the addition of 1ng/ml IL-3

(R&D systems, Abingdon, UK) at the start of culture. Media was replenished every two days. After 8-10

days, cells were differentiated towards reticulocytes by removing dexamethasone, increased EPO

(10U/ml, ProSpec, East Brunswick, NJ, USA) and addition of heparin (5U/ml, LEO Pharma B.V., Breda,

The Netherlands), 5% pooled AB+ plasma and holotransferrin (700µg/ml, Sanquin, Amsterdam, The

Netherlands). Every other day, half the media was replenished. For co-culture experiments, CD14+

cells were differentiated with (GC-macrophages) or without dexamethasone for three days and

co-cultured with erythroblasts (day 8-10 of culture; ratio 1:1.5) or more differentiated erythroid cells

(day 6 of differentiation; ratio 1:4) for 24 hours.

Modeling human erythroid-macrophage interactions

95

Flow cytometry

Cells were washed in phosphate-buffered saline (PBS) and resuspended in 1% bovine serum albu-

min (BSA)/PBS. Cells were incubated with primary antibodies for 30min at 4C, measured on LSRII or

LSRFortessa (both BD Biosciences, Oxford, UK) and analyzed using FlowJo software (FlowJo v10; Tree

Star, Inc., Ashland, OR, USA) (antibodies listed in Supplementary methods).

Mass spectrometry

See Supplementary methods.

ImageStreamX and IncuCyte

GC-macrophages or unstimulated cells were incubated with 100µg/ml fluorescein isothiocyanate

(FITC)-labeled zymosan (S. cerevisiae; MP Biomedicals, Solon, OH, USA) for 40min at 37C. Zymosan was

removed and cells were fixed in 4% paraformaldehyde (PFA) for 20min at 4C. Cells were transferred

to 1% BSA/PBS and stained with human leukocyte antigen-antigen D-related R-phycoerythrin (HLA-DR

PE; BD Biosciences). Furthermore, erythroid cells at day 7 of differentiation were stained with Deep

Red Anthraquinone 5 (DRAQ5; Abcam, Cambridge, UK). Imaging was performed on the ImageStreamX

(Amnis Corporation, Seattle, WA, USA) and images were analyzed using IDEAS Application v6.1 soft-

ware (Amnis Corporation). For IncuCyte experiments see Supplementary methods.

Cytospins

Cells were cytospun using Shandon Cytospin II (Thermo Scientific), dried and fixed in methanol.

Cells were stained with benzidine and Differential Quik Stain Kit (PolySciences, Warrington, PA, USA)

following manufacturer’s instructions. Slides were dried, embedded in Entellan (Merck, Darmstadt,

Germany) and images were taken (Leica DM-2500, Germany).

Reverse transcription polymerase chain reaction analysis

Reverse transcription polymerase chain reaction (RT-PCR) was performed as described previously12.

Values were normalized using S18 and HPRT as a reference gene and calibrated relative to expression

of CD14+ monocytes at day 0 (primers listed in Supplementary methods).

Results

Glucocorticoid stimulation directs monocyte differentiation to CD16+CD163+CD169+CX-

CR4+CD206+ macrophages

We have previously found that purified peripheral blood CD14+ monocytes cultured in EPO, SCF,

lipids and dexamethasone differentiate within three days into CD163, CD169, CXCR4 and CD16-pos-

itive macrophages that, upon co-culture with CD34+ cells, significantly increase the erythroid yield12.

However, it remained unclear as to which growth factors were crucial to differentiate monocytes to

Chapter 4

96

macrophages supporting erythropoiesis. Therefore, we examined which growth factors or supple-

ments determined this differentiation cue. Flow cytometry analysis showed that dexamethasone,

exclusively, induces high expression of CD16 and CD163 in macrophages. The addition of EPO, SCF

or lipids does not contribute to this high expression (Figure 1A,B). CXCR4 expression was already

upregulated in the absence of dexamethasone but was further increased upon stimulation with dexa-

methasone and lipids, whilst the expression of tissue residency marker CD169 was also upregulated

but occurred in a dexamethasone-independent manner (Figure 1C,D). Figure S1A depicts distinct

morphological changes upon dexamethasone-induced differentiation between freshly isolated CD14+

monocytes and cultured CD14+ cells. Monocytes were incubated with mifepristone, which blocks

glucocorticoid receptor activation. Membrane and messenger ribonucleic acid (mRNA) expression of

CD16, CD163, and CD206 was significantly reduced by mifepristone treatment, and thus dependent

on glucocorticoid receptor transcriptional control (Figure 1E and Figure S1B,C). Although neither

Figure 1C nor Figure 1E show an effect of dexamethasone on the fluorescence intensity of CD169,

mRNA levels of CD169 were clearly increased upon stimulation of the glucocorticoid receptor and

reduced when cells were treated with mifepristone. In contrast, CXCR4 mRNA levels did not change

upon mifepristone treatment, but membrane expression was increased (Figure S1B). Monocyte dif-

ferentiation increases expression of DC-SIGN independent of dexamethasone, albeit to expression

levels that are significantly lower compared to dendritic cells (Figure 1E and Figure S1C)26. Note that

cultured monocytes in all conditions are a homogeneous population, as single peaks observed in

histograms and multi-color flow cytometry data revealed that monocytes stimulated with glucocorti-

coids are CD16+CD163+CD169+CXCR4+CD206+ cells (Figure S1C,D). Interestingly, flow cytometry data

revealed that monocytes that have been differentiated for three days in the presence of dexameth-

asone were unable to change their phenotype after 4 or 24 hours of mifepristone treatment. Only

CD163 expression was slightly reduced after 24 hours mifepristone treatment (Figure S1E). The data

indicates that glucocorticoid stimulation initiates an irreversible differentiation program of monocytes

towards CD16+CD163+CD169+CXCR4+CD206+ macrophages which is maintained for at least 17 days

of culture (Figure S2A,B).

Proteomics data revealed GC-macrophages display a distinct anti-inflammatory profile

To gain further insights into the dexamethasone-induced monocyte differentiation process, we

performed mass spectrometry-based quantitative proteomics on these cells after three days of

differentiation and compared this to non-glucocorticoid stimulated monocytes. A total of 3210 pro-

teins were quantified, and principal component analysis clearly separated glucocorticoid-stimulated

from non-stimulated cells (Figure 2A and Online Supplementary Table S1). Glucocorticoid stimula-

tion induced a distinct expression pattern compared to non-glucocorticoid stimulated monocytes,

as visualized in the volcano plot and corresponding heatmap of the 169 differentially expressed

proteins for individual donors (Figure 2B,C). Note that the expression of CD163 and CD206 (MRC1)

was highly induced after glucocorticoid receptor activation, corroborating the flow cytometry exper-

iments. The most differentially expressed proteins (n=169) were mapped to evaluate specific up- or

Modeling human erythroid-macrophage interactions

97

0

1

2

3

C X C R 4

MF

In

orm

ali

zed

(fc

)

*

E P O

L ip id s

S C F

D e x

0

1 0

2 0

3 0

C D 1 6 3

MF

In

orm

ali

zed

(fc

) *** **** **** *** **** **

E P O

L ip id s

S C F

D e x

0 .0

0 .5

1 .0

1 .5

2 .0

C D 1 6 9

MF

In

orm

ali

zed

(fc

)

ns

E P O

L ip id s

S C F

D e x

0

5

1 0

1 5

C D 1 6

MF

In

orm

ali

zed

(fc

)

** ** *** *** ** ***

E P O

L ip id s

S C F

D e x

D

B

C

A

E**

**

D 0 D 3

M if (2 0 M )M if (1 M )

0

5

1 0

1 5

2 0

2 5

C D 1 6

MF

In

orm

ali

zed

(fc

)

D e x (1 M )

* 0.0675**

0

5

1 0

1 5

2 0

2 5

C D 1 6 3

MF

In

orm

ali

zed

(fc

)

D 0 D 3

M if (2 0 M )M if (1 M )

D e x (1 M )

***

0

5

1 0

1 5

2 0

2 5

C X C R 4

MF

In

orm

ali

zed

(fc

)

D 0 D 3

M if (2 0 M )M if (1 M )

D e x (1 M )

*

0

1 0 0

2 0 0

3 0 0

4 0 0

5 0 0

C D 2 0 6

MF

In

orm

ali

zed

(fc

)

D 0 D 3

M if (2 0 M )M if (1 M )

D e x (1 M )

0.0737

N D

ns

0

2

4

6

D C -S IG N

MF

In

orm

ali

zed

(fc

)

D 0 D 3

M if (2 0 M )M if (1 M )

D e x (1 M )

0

1 0 0

2 0 0

3 0 0

4 0 0

C D 1 6 9

MF

In

orm

ali

zed

(fc

) ns

D 0 D 3

M if (2 0 M )M if (1 M )

D e x (1 M )

Figure 1. Glucocorticoid receptor activation directs CD14+ monocytes towards a tissue resident mac-rophage phenotype. (A-D) Distribution graphs displaying the relative geometric mean fluorescence intensity (MFI) of CD16, CD163, CD169 and CXCR4 on human monocytes (n=3-6) cultured for three days under various conditions (EPO, SCF, lipids or dexamethasone). MFI was normalized to isotype control and presented as fold change (fc). Mean ± SEM (two-way ANOVA, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001). (E) Relative expression of CD16, CD163, CD169, CXCR4, CD206 and DC-SIGN on CD14+ monocytes (n=3) directly after isolation from mononuclear cells (D0) and after culture in the presence or absence of dexamethasone (Dex) and/or mifepri-stone (Mif). MFI was normalized to isotype control and displayed as a fold change to day 0. Mean ± SEM (ratio paired t-test, *P<0.05, **P<0.01). EPO: erythropoietin; SCF: stem cell factor; ND: not detected; ns: not significant.

Chapter 4

98

downregulation of functionality-linked protein networks, based on Search Tool for the Retrieval of

Interacting Genes/Proteins (STRING) analysis (Figure 2D). CD163 and CD206 are part of an interac-

tome protein node that is specifically upregulated in dexamethasone-induced macrophages, and

includes M2 macrophage markers CSF1R, stabilin-1 (STAB1) and complement proteins C3AR1, C1QC,

and FcgRIIa (CD32) which has been associated with high phagocytic capacity of the cells. Moreover,

VSIG4 was upregulated in dexamethasone-induced macrophages, which is restricted to resting tissue

macrophages27, while ABCA1 was also upregulated, which has been highly associated with hemo-

globin-associated macrophages28. In addition, proteins with a positive regulation of cell migration

and motility, including DAB2, ADAM9, Serpine1 (PAI1) and CD81, are upregulated in dexametha-

sone-induced macrophages. Furthermore, a whole range of signaling receptors were upregulated,

amongst which are TGF and IFNg-receptors (TGFBR1 and IFNGR1) and IL13RA1. These proteins belong

to processes that are enriched as GO-term, e.g., membrane part, signal transducer activity, trans-

membrane receptor activity and molecular transduced activity. In addition, many immune regulatory

processes are also enriched (Figure 2E and Online Supplementary Table S2). Interestingly, members

of the cathepsin family involved in antigen presentation (e.g., CTSC, CTSL1, CTSD and CTSS) were

downregulated. A range of pro-inflammatory proteins, clustered within an interactome node, were

downregulated; these include lysosomal enzymes HEXA and HEXB, MANBA, saponin PSAP and GLB1,

in addition to other lysosome/hydrolase activity-related GO-categories (Figure 2D,E). In addition,

GO-categories associated with lipid metabolic processes were also downregulated in GC-macro-

phages. Furthermore, CHI3L1 and CD44 are highly upregulated in non-glucocorticoid stimulated cells

(Figure 2B). CHI3L1 is described as a pro-inflammatory factor29,30, while CD44 has been expressed on

pro-inflammatory tissue macrophages31. In conclusion, CD14+ monocytes that have been differenti-

ated in the presence of dexamethasone display a distinct anti-inflammatory proteomic profile and

are further denoted as GC-macrophages, while unstimulated cells have a more inflammatory profile.

GC-macrophages are motile and bind erythroblasts

GC-macrophages may, besides supporting the erythroid yield, also regulate terminal differentiation

of erythroblasts, recapitulating aspects of erythroblastic islands. In mice, it has been shown that BM

central macrophages can bind erythroblasts through various interactions: VCAM1-integrin-α4β116,32,

integrin-α5β1-ICAM433,34, erythroblast macrophage protein (EMP)-EMP4,35, or EphrinB2-EphrinB436.

Flow cytometry data revealed that GC-macrophages express common cell adhesion molecules (CAM),

such as integrins (α4 [ITGA4], β1,2 [ITGB1, ITGB2/CD18] and αL,M,X [ITGAL/CD11a, ITGAM/CD11b,

ITGAX/CD11c]), the immunoglobin (Ig) superfamily (ICAM1, PECAM, VCAM1) and E- and L-selectin

(Figure 3A and Figure S3A). Most of these CAM could be identified in the proteomics data, including

ICAM3, integrin-β5 and α5, however, VCAM1, selectins and EMP were not detected (Online Supple-

mentary Table S1). With the exception of integrin-β5, these CAM were not differentially expressed

between GC-macrophages and non-glucocorticoid stimulated cells. Erythroblasts expressed similar

ITGA4 levels compared to GC-macrophages, but exhibit a 10-fold reduction in ITGB1 expression and

low expression of ICAM1 and PECAM, whereas VCAM1 was not detected (Figure 3A). When differ

Modeling human erythroid-macrophage interactions

99

Figure 2. Proteome analysis of CD14+ monocytes cultured in the presence or absence of dexamethasone revealed two distinct macrophage populations. (A) Principal component analysis of GC-macrophages (red) versus non-glucocorticoid stimulated cells (blue) of four donors (indicated A-D). (B) Volcano plot (false discov-ery rate 0.05 S0 0.4) showing P-values (-log) versus difference of cells cultured for three days in the presence or absence of dexamethasone. (C) Heatmap of differentially expressed proteins based on Z-scored label-free quantification values. (D) Interaction analysis based on STRING (all interactions) of upregulated (red) and down-regulated (blue) proteins. (E) Enrichment analysis using BiNGO and enrichment mapper in GC-macrophages with upregulated (red) and downregulated (blue) processes.

Chapter 4

100

entiating erythroblasts towards reticulocytes (Figure S3B,C), the expression of CAM was reduced, as

expected, which potentially indicates a lower binding affinity of erythroid cells to macrophages during

erythroid differentiation. Next, we investigated whether GC-macrophages interact in vitro with eryth-

roid cells compared to non-glucocorticoid stimulated monocytes. Indeed, live imaging cells for 2.5

days showed that GC-macrophages are highly motile and non-stimulated macrophages are non-mo-

tile (Figure 3B), a finding which corroborates the increased expression of cell migration and motility

proteins (Figure 2D) whilst engaging twice as many erythroblasts (0.5 vs. 0.3, P<0.0001) at every time

point measured (Figure 3C,D). In addition, cytospins of macrophages co-cultured for 24 hours with

erythroblasts showed that the number of macrophages binding erythroblasts as well as the number

of erythroblasts bound was increased in GC-macrophages compared to non-GC macrophages (Figure

3E and Figure S3D). Nonetheless, no difference in interaction duration between erythroblasts and

macrophages from both conditions was observed (Figure S3E), suggesting that the unstimulated cells

possess some machinery to interact with erythroblasts. In conclusion, GC-macrophages are motile,

express a variety of CAM and form erythroblast interactions with increased frequency and numbers

per macrophage compared to cells cultured in the absence of dexamethasone.

A

GC -M

E B Ld 1

E B Ld 7

0

1 0

2 0

3 0

4 0

V C A M 1

MF

I(n

orm

ali

zed

)

N D N D

GC -M

E B Ld 1

E B Ld 7

01 02 03 04 05 0

5 0 0 06 0 0 07 0 0 0

**** ****

P E C A M

MF

I(n

orm

ali

zed

)

GC -M

E B Ld 1

E B Ld 7

01 02 03 04 05 0

5 0 0 01 0 0 0 01 5 0 0 0 **

***IC A M 1

MF

I(n

orm

ali

zed

)

GC -M

E B Ld 1

E B Ld 7

0

1 0 0

2 0 0

3 0 0

4 0 0

IT G A 4

MF

I(n

orm

ali

zed

) *

GC -M

E B Ld 1

E B Ld 7

0

1 0 0 0

2 0 0 0

3 0 0 06 0 0 09 0 0 0

1 2 0 0 0 *********

IT G B 1

MF

I(n

orm

ali

zed

)

B

E -Dex

+Dex

i ii iii iv

i ii iii iv

v

v

C

- D e x + D e x0

1

2

3

4

5

Av

era

ge

nu

mb

er

of

lin

ks

/M ****

- D e x + D e x0

2

4

6

8

Ma

xim

um

nu

mb

er

of

lin

ks

/M

**

D

+ D e x

0 .4 7 m m

0.4

7m

m

-D e x

0 .4 7 m m

0.4

7m

m

-D e x + D e x0 .0

0 .1

0 .2

0 .3

0 .4

Le

ng

thtr

av

ell

ed

by

M

(in

mm

)

**

Modeling human erythroid-macrophage interactions

101

A

GC -M

E B Ld 1

E B Ld 7

0

1 0

2 0

3 0

4 0

V C A M 1

MF

I(n

orm

ali

zed

)

N D N D

GC -M

E B Ld 1

E B Ld 7

01 02 03 04 05 0

5 0 0 06 0 0 07 0 0 0

**** ****

P E C A M

MF

I(n

orm

ali

zed

)

GC -M

E B Ld 1

E B Ld 7

01 02 03 04 05 0

5 0 0 01 0 0 0 01 5 0 0 0 **

***IC A M 1

MF

I(n

orm

ali

zed

)

GC -M

E B Ld 1

E B Ld 7

0

1 0 0

2 0 0

3 0 0

4 0 0

IT G A 4

MF

I(n

orm

ali

zed

) *

GC -M

E B Ld 1

E B Ld 7

0

1 0 0 0

2 0 0 0

3 0 0 06 0 0 09 0 0 0

1 2 0 0 0 *********

IT G B 1

MF

I(n

orm

ali

zed

)

B

E -Dex

+Dex

i ii iii iv

i ii iii iv

v

v

C

- D e x + D e x0

1

2

3

4

5

Av

era

ge

nu

mb

er

of

lin

ks

/M ****

- D e x + D e x0

2

4

6

8

Ma

xim

um

nu

mb

er

of

lin

ks

/M

**

D

+ D e x

0 .4 7 m m

0.4

7m

m

-D e x

0 .4 7 m m

0.4

7m

m

-D e x + D e x0 .0

0 .1

0 .2

0 .3

0 .4

Le

ng

thtr

av

ell

ed

by

M

(in

mm

)

**

Figure 3. GC-macrophages form erythroblast clusters with increased frequency and erythroblast com-position. (A) Expression of integrins ITGA4 and ITGB1 and adhesion molecules ICAM1, PECAM, and VCAM1 on GC-macrophages (Mφ) (n=6) and erythroblasts (EBL) at day 1 and 7 of differentiation (n=3-4). Mean fluorescence intensity (MFI) has been normalized to the isotype control. Mean ± SEM (unpaired t-test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001). (B) Scaled cell-displacement vector diagram (left; 20 representative macrophages in both conditions) and box-and-whisker plot (right; 68 representative macrophages in -Dex and 21 in +Dex) after three days of culture in the absence or presence of dexamethasone (Welch’s unpaired t-test, **P<0.01, n=5). (C-D) Co-culture of GC-macrophages or unstimulated cells with erythroblasts (unpaired t-test of 1153 macrophages (-Dex) and 749 (+Dex), **P<0.01, ****P<0.0001, n=5). Images were taken every hour during 64 hours of analysis. (C) Plot showing the average erythroblast-macrophage links for each macrophage. Mean ± SD. (D) 5-95% box plot showing the maximum number of links per macrophage. Mean is indicated by crosses. (E) Representative images of cytospins of GC-macrophages (+Dex) or unstimulated cells (-Dex) co-cultured with erythroblasts for 24 hours (in 50x magnification, panels i-ii or 100x magnification, panels iii-v; n=4). Dex: dexamethasone; ND: not detected.

GC-macrophages express TAM-receptor family members and phagocytose pyrenocytes

As CD169+CD163+ macrophages promote erythropoiesis8, we decided to examine whether GC-macro-

phages can provide a similar functional role in vitro. In mice, pyrenocytes are phagocytosed by central

macrophages in a Mer tyrosine kinase (MERTK)-dependent manner18. RT-PCR showed that GC-macro-

phages upregulate both MERTK and AXL mRNA compared to freshly isolated and non-glucocorticoid

stimulated monocytes (Figure 4A). MERTK expression was inhibited by mifepristone treatment during

the first three days of culture, whereas AXL was not, suggesting that AXL expression is induced via a

trans-regulated process while MERTK needs the transcriptional activity of the glucocorticoid receptor.

Note that TYRO3 levels are dexamethasone-independently increased. Besides TAM-receptors, other

PS-receptors on macrophages have been reported to be involved in clearing apoptotic bodies, such

as TIM337 (T-cell immunoglobulin and mucin-domain containing-3), STAB38 and CD300A39 (CMRF35-

like molecule 8). TIM3 mRNA levels were increased, albeit independently of dexamethasone (Figure

S4A). This was confirmed by mass spectrometry, as peptides corresponding to TIM3 were identified

in GC-macrophages (HAVCR2 in Online Supplementary Table S1). CD300A and STAB1 were also iden-

Chapter 4

102

tified, of which STAB1 was significantly increased in GC-macrophages compared to unstimulated cells.

Interestingly, proteomics data showed that lactadherin, a PS-binding glycoprotein which stimulates

phagocytosis of red blood cells by macrophages40, was significantly induced in GC-macrophages

compared to unstimulated cells. RT-PCR confirmed increased lactadherin mRNA levels but this was

dexamethasone-independent (Figure S4B). Moreover, both GC-macrophages and unstimulated cells

express DNASE2, a crucial protein required to degrade DNA within phagocytosed apoptotic bodies

or pyrenocytes in macrophages41.

Expression of TAM-receptors and other PS-receptors on GC-macrophages may be a pre-requisite

to phagocytose particles, cells or pyrenocytes in case of erythropoiesis. Figure 4B showed that the

number of GC-macrophages that phagocytose particles, in addition to the amount of zymosan par-

ticles per macrophage is higher (73% vs. 45%, 2.3 vs. 1.7, respectively) compared to unstimulated

cells. Subsequently, both unstimulated cells and GC-macrophages were co-cultured with a mixture

of differentiating erythroblasts, reticulocytes and pyrenocytes (Figure S3B,C) for 24 hours. Cytospin

analysis showed that both GC-macrophages and unstimulated cells bind erythroid cells (Figure 4C),

however, increased numbers of nucleated cells, reticulocytes and pyrenocytes bind to GC-macro-

phages compared to unstimulated cells (Figure 4D-F and Figure S4C). Note that all nucleated erythroid

cells are specifically aligned with their nucleus towards the macrophage as observed in vivo (Figure 4C).

Pyrenocytes, however, were almost solely phagocytosed by GC-macrophages (Figure 4F and Figure

S4D). Importantly, GC-macrophages and unstimulated cells did not overtly phagocytose nucleated

cells or reticulocytes (Figure S4E,F). Flow cytometry data showed that indeed both GC-macrophages

and unstimulated cells can bind erythroid cells, however, increased cluster formation was found

for GC-macrophages compared to unstimulated cells (Figure 4G). These results demonstrate that

GC-macrophages functionally resemble specific aspects of macrophages within the erythroblastic

island by binding erythroblasts and reticulocytes and phagocytosing pyrenocytes.

GC-macrophages share characteristics with CD163+ macrophages found in human BM and FL

To investigate whether GC-macrophages share phenotypical characteristics with macrophages found

in the two major erythropoietic organs during human development and adulthood (FL and BM,

respectively), mononuclear cells of both organs were analyzed. Between week 15 and 22 of human

development, the FL is primarily undertaking erythropoiesis, representing a median of 85% of the total

number of mononuclear cells compared to 29% in BM, with increased frequencies of CD71+CD235-

pro-erythroblasts in FL (Figure 5A,B). To prevent the presence of free immunogenic pyrenocytes and

to support erythroid cell requirements in the developing embryo, it is anticipated that the FL contains

significant amounts of erythroblastic islands and, thus, supporting macrophages. Indeed, Figure

5C shows a 6.5-fold increase in CD163+ FL macrophages compared to BM (3.3% vs. 0.5%). Further

characterization shows only subtle differences in expression of macrophage markers (Figure 5D and

Figure S5A,B), as both macrophage populations express high levels of CD163 and CD14 and have

intermediate levels of CD169, CD206 and VCAM1. CD163+ BM macrophages tend to express more

Modeling human erythroid-macrophage interactions

103

A

0 .0 0 1

0 .0 1

0 .1

1

1 0

1 0 0

1 0 0 0

A X L

Re

lati

ve

ex

pre

ss

ion

(fc

)

4 h rs

D e xM if

D 0 D 3

**** *

0 .0 1

0 .1

1

1 0

1 0 0

T Y R O 3

Re

lati

ve

ex

pre

ss

ion

(fc

)

4 h rs

D e xM if

D 0 D 3

ns

B

387 out of 530 Mϕ (73%)

588 out of 1285 Mϕ (45%)

-D e x + D e x

2

4

6

8

1 0

1 2P

ha

go

cy

tos

ed

pa

rtic

les

****

BF HLA-DR Zymosan Merge

+ De

x-D

ex

BF HLA-DR Zymosan Merge

4 h rs0 .0 0 0 1

0 .0 0 1

0 .0 1

0 .1

1

1 0

1 0 0

1 0 0 0

M E R T K

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* * 0.0747

D

G

-D e x + D e x0

1 0

2 0

3 0

4 0

5 0

CD

14

+C

D2

35

a+

ce

lls

(%)

**MFI

+Dex 17888

- Dex 9873

Isotype 10284

CD14+ cells

-D e x + D e x0

1

2

3

4

Py

ren

oc

yte

sli

nk

ed

toM ****

-D e x + D e x0

2

4

6

Re

tic

ulo

cy

tes

bo

un

dto

M

****

-D e x + D e x0

2

4

6

8

1 0

1 2

1 4

Nu

cle

ate

dc

ell

sb

ou

nd

toM

****

C

+Dex

-Dex

i ii iii iv v

vii iii ivi

*

*

E FFigure 4. GC-macrophages can bind erythroid cells and phagocytose pyrenocytes. (A) Relative mRNA expres-sion of TAM-receptor family members MERTK, AXL and TYRO3 on CD14+ cells (D0) cultured for three days (D3) in the presence or absence of dexamethasone (Dex). 20µM mifepristone (Mif) was added for three days or after three days for 4 hours (n=4). Mean ± SEM (ratio paired t-test, *P<0.05, ****P<0.0001). (B) Representative ImageStreamX images of zymosan (green) phagocytosed by HLA-DR (red) positive unstimulated macrophages (-Dex) and GC-macrophages (+Dex) (left), and corresponding 10-90% box plot showing the number of zymosan particles phagocytosed (right) (unpaired t-test of 1285 macrophages (-Dex) and 530 (+Dex), ****P<0.0001, n=3). (C-F) GC-macrophages and unstim-ulated cells were co-cultured for 24 hours with day 6 differentiated erythroid cells (unpaired t-test of 370 -Dex and 313 +Dex macrophages, ****P<0.0001, n=3). (C) Representative images of cytospins (in 50x magnification, panels i-ii or 100x magnification, panels iii-v). Macrophages bind nucleated erythroid cells (large arrow), reticulocytes (arrowhead) and phagocytose pyrenocytes (small arrow) and some erythroid cells during differentiation (asterisk). 10-90% box plots showing the number of nucleated cells (D) or reticulocytes (E) bound to macrophages (Mφ). (F) Scatter plot showing the number of pyrenocytes bound to or phagocytosed by macrophages. Mean ± SD. (G) Graph showing the binding of CD235a+ differentiated erythroid cells to GC-macrophages versus unstimulated cells. Corresponding histogram showing geometric mean of CD235a in FITC (n=4). Mean ± SEM (paired t-test, **P<0.01). HLA-DR: human leukocyte antigen-antigen D-related; MFI: mean fluorescence intensity; ns: not significant; BF: bright-field.

Chapter 4

104

D

G

-D e x + D e x0

1 0

2 0

3 0

4 0

5 0

CD

14

+C

D2

35

a+

ce

lls

(%)

**MFI

+Dex 17888

- Dex 9873

Isotype 10284

CD14+ cells

-D e x + D e x0

1

2

3

4

Py

ren

oc

yte

sli

nk

ed

toM ****

-D e x + D e x0

2

4

6

Re

tic

ulo

cy

tes

bo

un

dto

M

****

-D e x + D e x0

2

4

6

8

1 0

1 2

1 4

Nu

cle

ate

dc

ell

sb

ou

nd

toM

****

C

+Dex

-Dex

i ii iii iv v

vii iii ivi

*

*

E F

Figure 4. Continued

CXCR4, whereas CD163+ FL macrophages have higher expression of CD16. Table S3 displays the

comparison between the mean fluorescence intensity (MFI) of BM, FL, non-stimulated and GC-mac-

rophages and reveals that GC-macrophages phenotypically recapitulate macrophages found in the

FL and BM. GC-macrophages are more similar to BM macrophages (CD16 and CXCR4 expression),

however, they also share features of FL macrophages (CD206 expression). Unstimulated cells do not

express VCAM1, and have low expression of CD206, CD163, CD14 and CD16. Figure 5E,F shows that

both BM and FL CD163+ macrophages bind erythroid cells (46% in BM vs. 83% in FL), indicating that

CD163 purifies erythroid-supporting macrophages. Interestingly, FL macrophages have increased

interactions with CD71+CD235a+ cells compared to BM. The similarity of marker expression levels of

BM, FL and GC-macrophages and the fact that all three populations form erythroid clusters suggests

that GC-macrophages share phenotypic and functional characteristics with in vivo erythroid-support-

ing macrophages. GC-macrophages could thus be used as a substitute in vitro model to study the

supportive effects of macrophages on erythropoiesis.

Discussion 

We have previously shown that monocyte-derived macrophages can support erythropoiesis by

increased survival of HSPC12. Herein, we show that these macrophages derived from CD14+ mono-

cytes, are differentiated in a glucocorticoid-dependent manner (termed GC-macrophages), interact

with erythroid cells of all stages and phagocytose the extruded pyrenocytes. Besides these functional

aspects, GC-macrophages also share phenotypic characteristics with resident macrophages from both

Modeling human erythroid-macrophage interactions

105

human BM and FL, among which there is high expression of CD163 and CD206. Interestingly, CD163+

BM cells appear to be more heterogeneous compared to FL cells. GC-macrophages also phenotypically

resemble macrophages described recently by Belay et al., who employed a lentivirally introduced small

molecule responsive Mpl-based cell growth switch that enabled cord blood or BM CD34+ cells to be

differentiated to erythroid-supporting macrophages13. Similar to GC-macrophages, these cells express

CD14, CD163, CD169, CD206, VCAM1, ITGAM and ITGAX. Herein, we show that these macrophages

can also be differentiated from peripheral blood monocytes using dexamethasone, without the need

for genetic manipulation. Falchi et al. showed that in erythroid culture conditions, CD34+ cells can also

differentiate to macrophages that interact with erythroid cells, however, we can exclude this differenti-

ation pathway as the purified CD14+ monocytes we used to differentiate macrophages from peripheral

blood did not show hematopoietic colony potential or CD34+ contamination12.

The erythroid system is renowned for its rapid response to systemic decreases in oxygen pressure.

Together with elevated EPO levels, glucocorticoid levels also increase upon exposure to high altitude42.

EPO, SCF and glucocorticoids induce erythroblasts to proliferate whilst inhibiting differentiation43-46.

Elevated systemic EPO and glucocorticoids as a response to low-oxygen stress leads to increased

erythroid output due to augmented survival and proliferation of BM erythroblasts. To accommodate

this increased erythropoiesis, we hypothesize that the number of central macrophages must also be

increased or alternatively these cells would have to engage with more erythroblasts. Our flow cytometry

and cytospin data confirmed that GC-macrophages interact with erythroid cells of all stages, be that as it

may, this does not provide information on the longevity of the interactions, as these could be transient,

as previously implied47. Via live cell imaging we analyzed the interaction between GC-macrophages and

erythroblasts, which revealed GC-macrophages are more mobile compared to cells that were cultured in

the absence of dexamethasone, and this mobility, or “macrophage ranging”, results in more interactions

with erythroblasts. Higher mobility was accompanied by an increased expression of proteins involved in

migration and motility. High motility has previously been observed in CD34+differentiated macrophages

stimulated with dexamethasone47. Motility is an important functional aspect, as erythroblastic islands in

vivo form away from sinusoids and migrate to the sinusoidal endothelium to release reticulocytes into

the circulation48,49. Interestingly, this work also demonstrates that non-glucocorticoid-stimulated mono-

cytes can interact with erythroblasts, as they form interactions for the same length of time (1.8 hours

on average) when they encounter erythroblasts. This suggests that both populations express receptors

that allow engagement and interaction with erythroblasts, however, GC-macrophages have significantly

more interactions with erythroblasts per macrophage and bind a higher number of erythroblasts. Sur-

prisingly, GC-macrophages display low expression of VCAM1, suggesting that erythroblast interactions

may also occur in a VCAM1-independent manner. Indeed, Ulyanova et al. reported that Vcam1-/- mice do

not display a compromised erythroid stress response in spleen and BM17. Whether another interaction

substitutes for VCAM1 would need to be determined. The presented monocyte differentiation method-

ology has potential to be exploited as an imaging platform to delineate the hierarchy of contributions of

various receptors within the macrophage-erythroblasts in BM and GC-macrophages in future studies.

Chapter 4

106

B M F L0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

C D 1 4

MF

I(n

orm

ali

zed

)

B M F L0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

C D 1 6

MF

I(n

orm

ali

zed

) **

B M F L0

2 0 0 0 0

4 0 0 0 0

6 0 0 0 0

8 0 0 0 0

1 0 0 0 0 0

C D 1 6 3

MF

I(n

orm

ali

zed

)

B M F L0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

2 0 0 0 0

C X C R 4

MF

I(n

orm

ali

zed

) 0.0723

B M F L0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

2 0 0 0 0

C D 2 0 6

MF

I(n

orm

ali

zed

)

B M F L0

5 0 0 0

1 0 0 0 0

1 5 0 0 0

V C A M 1

MF

I(n

orm

ali

zed

)

B M F L0

2 0 0 0

4 0 0 0

6 0 0 0

8 0 0 0

C D 1 6 9

MF

I(n

orm

ali

zed

)

D

C

B

B M F L0

5 0

1 0 0

To

tal

ce

llfr

ac

tio

n(%

)

C D 7 1 + C D 2 3 5 a -

C D 7 1 + C D 2 3 5 a +

C D 7 1 -C D 2 3 5 a +

*

***

F

B M F L0

5 0

1 0 0

CD

16

3+

ce

llfr

ac

tio

n(%

)

C D 7 1 + C D 2 3 5 a -

C D 7 1 + C D 2 3 5 a +

C D 7 1 -C D 2 3 5 a +*

ATotal BM Total FL

ECD163+ BM CD163+ FL

B M F L0

1

2

3

4

5

6

CD

16

3+

ce

lls

(%)

**

Figure 5. CD163+ macrophage populations in human BM and FL. (A) Representative dot plots of erythroid cells characterized by the expression of CD71high and CD235a in total human BM (n=4) and FL mononuclear cells (n=4). (B) Graph belonging to panel A. Mean ± SD (unpaired t-test, ***P<0.001). (C) Graph showing the percentage of CD163+ cells present in the total cell population in BM (n=7) and FL (n=5). Mean ± SEM (unpaired t-test, **P<0.01). (D) Characterization of CD163+ macrophages in human BM (n=3-7) and FL (n=3-5) based on the expression of CD14, CD16, CD163, CD169, CXCR4, CD206 and VCAM1. Mean fluorescence intensity (MFI) has been normalized to the isotype control. Mean ± SEM (unpaired t-test, **P<0.01). (E) Representative dot plots of erythroid-macro-phage cluster formation of erythroid cells (CD71highCD235a+) with CD163+ BM (n=4) and FL (n=4) macrophages. (F) Graph belonging to panel E. Mean ± SD (unpaired t-test, *P<0.05). BM: bone marrow; FL: fetal liver.

Modeling human erythroid-macrophage interactions

107

We have also demonstrated, using proteomics and imaging, that GC-macrophages actively phagocytose

pyrenocytes and express the correct putative machinery to recognize pyrenocytes. The mechanism(s)

through which macrophages recognize reticulocytes but phagocytose pyrenocytes are ill-defined in

human erythropoiesis. Our proteomic study and RT-PCR data demonstrate that GC-macrophages

express all TAM-receptors, including MERTK and other PS-receptors, which may be used by GC-mac-

rophages to take up pyrenocytes. This work, alongside our ability to manipulate erythroblast protein

expression, now provides an excellent accessible model system to mechanistically understand how

macrophages promote erythropoiesis and eventually target pyrenocytes for phagocytosis and destruc-

tion. Furthermore, it is interesting to note that GC-macrophages interact preferably to the polarized

nuclear side of erythroid cells as observed in BM erythroblastic islands. In general, proteomic analysis

revealed an array of processes and proteins that are differentially regulated between GC-macrophages

and unstimulated cells. The data will allow further studies to delineate essential pathways that are key to

glucocorticoid stimulated differentiation of monocytes towards erythroid-supporting GC-macrophages.

This is probably the concerted action of multiple pathways.

Finally, our observations have important implications for our understanding of the dynamics of the mac-

rophage populations in human BM. We characterized both human BM and FL macrophages and found

that CD163+ FL macrophages define a homogeneous population. In contrast, CD163+ BM macrophages

show a more heterogeneous population, reflecting that CD163+ cells represent a mixed population

of myeloid cells. Both human BM and FL CD163+ macrophages are capable of binding erythroid cells,

however, this percentage is lower in BM (46%) compared to FL (83%). The FL is primarily performing

erythropoiesis at week 15-22 of embryonic development, which suggests that CD163 purifies mainly

central macrophages. The reduced erythroid-macrophage clusters in BM may reflect a more hetero-

geneous CD163+ population with possibly different functions. Changes observed in marker expression

of macrophages in both organs could thus be due to this heterogeneity in the BM population. CD163

isolation in combination with single cell RNA-sequencing may discriminate these different populations

and identify specific discriminatory cell surface markers to allow for functional experiments.

Albeit for decades it was believed that all macrophages originate from monocytes50, recent parabiosis

and fate-mapping studies showed that most resident macrophages are maintained independently of

monocytes51. However, Theurl et al. showed that resident Kupffer cells in the liver contain a mixture of

de novo hematopoiesis-derived and embryonic-derived macrophages. They identified an on-demand

mechanism to facilitate quick and transient increases in cells that can function as Kupffer cells but

originate from classical monocytes52. Taken together with our work, we hypothesize a new scenario

in which specific macrophages originate from different sources depending on the need of a specific

tissue or process. These processes may also occurs in other tissues in response to stress, like the BM.

The origin and homeostasis of human BM resident macrophages is presently ill-defined, if described

at all. Elevated glucocorticoid levels may lead to direct differentiation of monocytes and elevated

numbers of nursing central macrophages to facilitate the increased erythroid output in analogy to

Chapter 4

108

Kupffer cells. Active research aimed at unraveling the origin of tissue resident macrophages, which is

important to understand not only homeostatic but also pathogenic erythropoiesis in which a driving

role of macrophages has been implicated, such as polycythemia vera and β-thalassemia. Herein,

we provide evidence that monocytes can indeed differentiate in vitro to macrophages that support

erythropoiesis, providing a model to study such erythroid-macrophage interactions.

Author contributions

EH performed the experiments. LAH-O and EH performed the IncuCyte experiments and SJC and

LAH-O did IncuCyte data analysis. FPJA and MB performed the mass spectrometry data acquisition

and analysis. EH and EA designed the experiments and analyzed the data. EH and EA wrote the

manuscript. AMT contributed to experiment design, analysis and writing of the manuscript. The

manuscript was critically revised by all authors. The authors declare no competing financial interests.

Acknowledgements

The authors would like to thank the staff of the CASA clinic in Leiden for collecting human fetal tis-

sues and Dr. Tom Cupedo, Natalie Papazian and Martijn Bogaerts from the Erasmus Medical Center,

Rotterdam, for providing human fetal liver material. We also thank the Central Facility of Sanquin for

technical assistance regarding ImageStreamX. This work was supported by grants from the Land-

steiner Foundation (LSBR1141; EA, EH and LSBR1517; MB), a NHS Blood and Transplant (NHSBT) R&D

grant (WP15-05; AMT), National Institute for Health Research (NIHR) for a Blood and Transplant

Research Unit in Red Blood Cell Products at the University of Bristol in partnership with NHSBT

(AMT, LAH-O) and the and Wellcome Trust (105385/Z/14/Z; LAH-O and ISSF; SJC). This article presents

independent research partly funded by the NIHR. The views expressed are those of the authors

and not necessarily the NHS, the NIHR or the Department of Health.

Modeling human erythroid-macrophage interactions

109

References

1. Lee SH, Crocker PR, Westaby S, et al. Isolation and immunocytochemical characterization of human bone marrow stromal

macrophages in hemopoietic clusters. J Exp Med. 1988 Sep 01;168(3):1193-1198.

2. Mohandas N, Prenant M. Three-dimensional model of bone marrow. Blood. 1978 Apr;51(4):633-643.

3. 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 Oct 15;92(8):2940-2950.

4. Soni S, Bala S, Gwynn B, Sahr KE, Peters LL, Hanspal M. Absence of erythroblast macrophage protein (Emp) leads to failure of

erythroblast nuclear extrusion. J Biol Chem. 2006 Jul 21;281(29):20181-20189.

5. Hanspal M, Golan DE, Smockova Y, et al. Temporal synthesis of band 3 oligomers during terminal maturation of mouse eryth-

roblasts. Dimers and tetramers exist in the membrane as preformed stable species. Blood. 1998 Jul 1;92(1):329-338.

6. Sawada K, Krantz SB, Dessypris EN, Koury ST, Sawyer ST. Human colony-forming units-erythroid 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

May;83(5):1701-1709.

7. Chow A, Lucas D, Hidalgo A, et al. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and

progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011 Feb 14;208(2):261-271.

8. Chow A, Huggins M, Ahmed J, et al. CD169(+) macrophages provide a niche promoting erythropoiesis under homeostasis and

stress. Nat Med. Apr;19(4):429-436.

9. Winkler IG, Sims NA, Pettit AR, et al. Bone marrow macrophages maintain hematopoietic stem cell (HSC) niches and their deple-

tion mobilizes HSCs. Blood. 2010 Dec 2;116(23):4815-4828.

10. Heideveld E, van den Akker E. Digesting the role of bone marrow macrophages on hematopoiesis. Immunobiology. 2017

Jun;222(6):814-822.

11. Iavarone A, King ER, Dai XM, Leone G, Stanley ER, Lasorella A. Retinoblastoma promotes definitive erythropoiesis by repressing

Id2 in fetal liver macrophages. Nature. 2004 Dec 23;432(7020):1040-1045.

12. Heideveld E, Masiello F, Marra M, et al. CD14+ cells from peripheral blood positively regulate hematopoietic stem and progenitor

cell survival resulting in increased erythroid yield. Haematologica. 2015 Nov;100(11):1396-1406.

13. Belay E, Hayes BJ, Blau CA, Torok-Storb B. Human Cord Blood and Bone Marrow CD34+ Cells Generate Macrophages That

Support Erythroid Islands. PLoS One. 2017;12(1):e0171096.

14. Ramos P, Casu C, Gardenghi S, et al. Macrophages support pathological erythropoiesis in polycythemia vera and beta-thalas-

semia. Nat Med. 2013 Apr;19(4):437-445.

15. Jacobsen RN, Perkins AC, Levesque JP. Macrophages and regulation of erythropoiesis. Curr Opin Hematol. 2015 May;22(3):212-

219.

16. 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 Jan 01;181(1):411-415.

17. Ulyanova T, Phelps SR, Papayannopoulou T. The macrophage contribution to stress erythropoiesis: when less is enough. Blood.

2016 Sep 29;128(13):1756-1765.

18. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by central macrophages in erythroblastic islands.

Blood. 2014 Jun 19;123(25):3963-3971.

19. Yoshida H, Kawane K, Koike M, Mori Y, Uchiyama Y, Nagata S. Phosphatidylserine-dependent engulfment by macrophages of

nuclei from erythroid precursor cells. Nature. 2005 Sep 29;437(7059):754-758.

20. Lu Q, Lemke G. Homeostatic regulation of the immune system by receptor tyrosine kinases of the Tyro 3 family. Science. 2001

Jul 13;293(5528):306-311.

21. Gaipl US, Voll RE, Sheriff A, Franz S, Kalden JR, Herrmann M. Impaired clearance of dying cells in systemic lupus erythematosus.

Autoimmun Rev. 2005 Apr;4(4):189-194.

22. Munoz LE, Janko C, Schulze C, et al. Autoimmunity and chronic inflammation - two clearance-related steps in the etiopathogenesis

of SLE. Autoimmun Rev. 2010 Nov;10(1):38-42.

23. Nagata S. Apoptosis and autoimmune diseases. Ann N Y Acad Sci. 2010 Oct;1209:10-16.

24. Rothlin CV, Lemke G. TAM receptor signaling and autoimmune disease. Curr Opin Immunol. 2010 Dec;22(6):740-746.

25. Toda S, Nishi C, Yanagihashi Y, Segawa K, Nagata S. Clearance of Apoptotic Cells and Pyrenocytes. Curr Top Dev Biol. 2015;114:267-

295.

26. Ciudad MT, Sorvillo N, van Alphen FP, et al. Analysis of the HLA-DR peptidome from human dendritic cells reveals high affinity

repertoires and nonconventional pathways of peptide generation. J Leukoc Biol. 2017 Jan;101(1):15-27.

27. Vogt L, Schmitz N, Kurrer MO, et al. VSIG4, a B7 family-related protein, is a negative regulator of T cell activation. J Clin Invest.

2006 Oct;116(10):2817-2826.

Chapter 4

110

28. Finn AV, Nakano M, Polavarapu R, et al. Hemoglobin directs macrophage differentiation and prevents foam cell formation in

human atherosclerotic plaques. J Am Coll Cardiol. 2012 Jan 10;59(2):166-177.

29. Krause SW, Rehli M, Kreutz M, Schwarzfischer L, Paulauskis JD, Andreesen R. Differential screening identifies genetic markers of

monocyte to macrophage maturation. J Leukoc Biol. 1996 Oct;60(4):540-545.

30. Libreros S, Garcia-Areas R, Keating P, Carrio R, Iragavarapu-Charyulu VL. Exploring the role of CHI3L1 in “pre-metastatic” lungs

of mammary tumor-bearing mice. Front Physiol. 2013;4:392.

31. Hoban MD, Cost GJ, Mendel MC, et al. Correction of the sickle cell disease mutation in human hematopoietic stem/progenitor

cells. Blood. 2015 Apr 23;125(17):2597-2604.

32. 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 Jul 01;106(1):86-94.

33. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythro-

blastic island formation. Blood. 2006 Sep 15;108(6):2064-2071.

34. Ulyanova T, Jiang Y, Padilla S, Nakamoto B, Papayannopoulou T. Combinatorial and distinct roles of alpha(5) and alpha(4) integrins

in stress erythropoiesis in mice. Blood. 2011 Jan 20;117(3):975-985.

35. Soni S, Bala S, Hanspal M. Requirement for erythroblast-macrophage protein (Emp) in definitive erythropoiesis. Blood Cells Mol

Dis. 2008 Sep-Oct;41(2):141-147.

36. Suenobu S, Takakura N, Inada T, et al. A role of EphB4 receptor and its ligand, ephrin-B2, in erythropoiesis. Biochem Biophys

Res Commun. 2002 May 10;293(3):1124-1131.

37. Ocana-Guzman R, Torre-Bouscoulet L, Sada-Ovalle I. TIM-3 Regulates Distinct Functions in Macrophages. Front Immunol.

2016;7:229.

38. D’Souza S, Park SY, Kim IS. Stabilin-2 acts as an engulfment receptor for the phosphatidylserine-dependent clearance of primary

necrotic cells. Biochem Biophys Res Commun. 2013 Mar 15;432(3):412-417.

39. Simhadri VR, Andersen JF, Calvo E, Choi SC, Coligan JE, Borrego F. Human CD300a binds to phosphatidylethanolamine and

phosphatidylserine, and modulates the phagocytosis of dead cells. Blood. 2012 Mar 22;119(12):2799-2809.

40. Dasgupta SK, Abdel-Monem H, Guchhait P, Nagata S, Thiagarajan P. Role of lactadherin in the clearance of phosphatidylser-

ine-expressing red blood cells. Transfusion. 2008 Nov;48(11):2370-2376.

41. Kawane K, Fukuyama H, Kondoh G, et al. Requirement of DNase II for definitive erythropoiesis in the mouse fetal liver. Science.

2001 May 25;292(5521):1546-1549.

42. Humpeler E, Skrabal F, Bartsch G. Influence of exposure to moderate altitude on the plasma concentraton of cortisol, aldoste-

rone, renin, testosterone, and gonadotropins. Eur J Appl Physiol Occup Physiol. 1980;45(2-3):167-176.

43. van den Akker E, Satchwell TJ, Pellegrin S, Daniels G, Toye AM. The majority of the in vitro erythroid expansion potential resides

in CD34- cells, outweighing the contribution of CD34+ cells and significantly increasing the erythroblast yield from peripheral

blood samples. Haematologica. 2010 Sep;95(9):1594-1598.

44. Leberbauer C, Boulme F, Unfried G, Huber J, Beug H, Mullner EW. Different steroids co-regulate long-term expansion versus

terminal differentiation in primary human erythroid progenitors. Blood. 2005 Jan 01;105(1):85-94.

45. von Lindern M, Deiner EM, Dolznig H, et al. Leukemic transformation of normal murine erythroid progenitors: v- and c-ErbB act

through signaling pathways activated by the EpoR and c-Kit in stress erythropoiesis. Oncogene. 2001 Jun 21;20(28):3651-3664.

46. Bauer A, Tronche F, Wessely O, et al. The glucocorticoid receptor is required for stress erythropoiesis. Genes Dev. 1999 Nov

15;13(22):2996-3002.

47. Falchi M, Varricchio L, Martelli F, et al. Dexamethasone targeted directly to macrophages induces macrophage niches that

promote erythroid expansion. Haematologica. 2015 Feb;100(2):178-187.

48. Yokoyama T, Kitagawa H, Takeuchi T, Tsukahara S, Kannan Y. No apoptotic cell death of erythroid cells of erythroblastic islands

in bone marrow of healthy rats. J Vet Med Sci. 2002 Oct;64(10):913-919.

49. Yokoyama T, Etoh T, Kitagawa H, Tsukahara S, Kannan Y. Migration of erythroblastic islands toward the sinusoid as erythroid

maturation proceeds in rat bone marrow. J Vet Med Sci. 2003 Apr;65(4):449-452.

50. van Furth R, Cohn ZA. The origin and kinetics of mononuclear phagocytes. J Exp Med. 1968 Sep 01;128(3):415-435.

51. Hashimoto D, Chow A, Noizat C, et al. Tissue-resident macrophages self-maintain locally throughout adult life with minimal

contribution from circulating monocytes. Immunity. 2013 Apr 18;38(4):792-804.

52. Theurl I, Hilgendorf I, Nairz M, et al. On-demand erythrocyte disposal and iron recycling requires transient macrophages in the

liver. Nat Med. 2016 Aug;22(8):945-951.

Modeling human erythroid-macrophage interactions

111

Supplementary methods 

Flow cytometry

Antibodies used: AbD Serotec (Bio-Rad, Veenendaal, The Netherlands): anti-CD54, ICAM1 (AF647

1:400), anti-CD209, DC-SIGN (PE 1:100); Acris (Herford, Germany): anti-CD235a (FITC 1:450; PE 1:450);

BD Biosciences: anti-CD11c, ITGAX (PE 1:10), anti-CD14 (Pacific Blue 1:150), anti-CD18, ITGB2 (FITC

1:20), anti-CD49d, ITGA4 (APC-H7 1:80), anti-CD62L, L-selectin (V450 1:300), anti-CD106, VCAM1 (APC

1:10), anti-CD169 (APC 1:100), anti-CD206 (APC 1:100); Beckman-Coulter (Fullerton, CA): anti-CD29,

ITGB1 (FITC 1:70); Bender MedSystems (Vienna, Austria): anti-CD62E, E-selectin (FITC 1:10); Biolegend

(ITK Diagnostics, The Netherlands): anti-CD11a, ITGAL (APC 1:50), anti-CD11b, ITGAM (BV421 1:50);

eBioscience (Vienna, Austria): anti-CD184, CXCR4 (PE 1:100); Miltenyi Biotec: anti-CD71 (VioBlue 1:200),

anti-CD163 (PE 1:100), anti-CD235a (VioBlue 1:200), propidium iodide (PI; 1:100); Pelicluster (Amster-

dam, The Netherlands): anti-CD16 (FITC 1:100), anti-CD31, PECAM (FITC 1:10). 

 

Mass spectrometry data acquisition and analysis 

Cells were lysed in 100μl 4% SDS, 100mM DTT, 100mM Tris.HCl pH7.5 and processed into tryptic peptides

using Filter Aided Sample Preparation1. Peptides were desalted and concentrated using Empore-C18

StageTips2 and eluted with 0.5%(v/v) acetic acid, 80%(v/v) acetonitrile. Sample volume was reduced by

SpeedVac and supplemented with 2% acetonitrile, 0.1% TFA to a final volume of 12μl. 3μl was injected

for mass spectrometry analysis. Tryptic peptides were separated by nanoscale C18 reverse phase chro-

matography coupled on line to an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific) via a

nanoelectrospray ion source (Nanospray Flex Ion Source, Thermo Scientific). Peptides were loaded on a

20cm 75–360µm inner-outer diameter fused silica emitter (New Objective) packed in-house with Repro-

Sil-Pur C18-AQ, 1.9μm resin (Dr Maisch GmbH). The column was installed on a Dionex Ultimate3000

RSLC nanoSystem (Thermo Scientific) using a MicroTee union formatted for 360μm outer diameter

columns (IDEX) and a liquid junction. The spray voltage was set to 2.15kV. Buffer A was composed of

0.5% acetic acid and buffer B of 0.5% acetic acid, 80% acetonitrile. Peptides were loaded for 17min at

300nl/min at 5% buffer B, equilibrated for 5min at 5% buffer B (17-22min) and eluted by increasing

buffer B from 5-15% (22-87min) and 15-38% (87-147min), followed by a 10min wash to 90% and a 5min

regeneration to 5%. Survey scans of peptide precursors from 400-1500m/z were performed at 120K

resolution (at 200m/z) with a 1.5x105 ion count target. Tandem mass spectrometry was performed by

isolation with the quadrupole with isolation window 1.6, HCD fragmentation with normalized collision

energy of 30, and rapid scan mass spectrometry analysis in the ion trap. The MS2 ion count target was

set to 1.5x104 and the max injection time was 35ms. Only those precursors with charge state 2-7 were

sampled for MS2. The dynamic exclusion duration was set to 60s with a 10ppm tolerance around the

selected precursor and its isotopes. Monoisotopic precursor selection was turned on. The instrument

was run in top speed mode with 3s cycles. All data were acquired with Xcalibur software. The RAW

mass spectrometry files were processed with the MaxQuant3 computational platform, version 1.5.2.8.

Proteins and peptides were identified using the Andromeda search engine by querying the human

Chapter 4

112

Uniprot database (release 2015-02, 89796 entries). Standard settings with the additional options match

between runs, Label Free Quantification (LFQ), and only unique and razor peptides for quantification

were selected. The generated ‘proteingroups.txt’ table was filtered for potential contaminants, reverse

hits and ‘only identified by site’ using Perseus4, version 1.5.1.6. The LFQ values were transformed in log2

scale and proteins were filtered for four valid values in at least one of the experimental groups. Missing

values were imputed by normal distribution (width=0.3, shift=1.8), assuming these proteins were close

to the detection limit. Quantitative significance (Principal Component Analysis and Volcano plot using

an FDR of 5% and S0 of 0.4) was performed by Perseus software. Interaction network analysis of the

most differentially expressed proteins was performed using STRING (version 10) using all parameters

and Score: 0.4005. The identified network was uploaded into Cytoscape6, version 3.5.1. Enrichment

analysis of the most differentially expressed proteins was performed using the Cytoscape (version

3.5.1) plug-in BiNGO7 (version 3.0.3) with an FDR threshold of 0.05 and enrichment mapper8 (version

2.1.0) with a P-value cut-off of 0.005, an FDR Q-value cut-off of 0.005, an overlap coefficient of 0.5 and

a combined constant of 0.5. 

 

IncuCyte data acquisition and analysis 

1.5x106 monocytes/well were seeded in a 12-well plate and cultured in the presence or absence of

dexamethasone. After three days, cells were gently washed with PBS and incubated for 45min with

CellTracker (ThermoFisher Scientific, Waltham, USA) Green CMFDA (5-chloromethylfluorescein diace-

tate) to monitor cell movement and location. Cells were co-cultured with 3x106 erythroblasts for 6hrs.

The wells were gently washed to remove excess erythroblasts. For imaging, medium was changed into

Iscove’s modified Dulbecco’s medium (IMDM without phenol red) supplemented with erythropoietin

(4U/ml), holotransferrin (700µg/ml) and 30% human serum. Plates were mounted on the IncuCyte

Zoom (Essen Biosciences) and once per hour real-time images at 25 spots per well were taken for a

68-hour time period. The spatial relationship between erythroblasts and macrophages was charac-

terised using Fiji9,10. Initially, lateral drift in the phase-contrast and fluorescence images over time was

corrected using the StackReg plugin11. A difference of Gaussian filter (approximating the equivalent

Laplacian of Gaussian12) was then applied to the phase-contrast channel to enhance features with

diameters matching those expected for erythroblasts. Erythroblasts were subsequently identified

with the TrackMate plugin using the Laplacian of Gaussian feature detector13. Fluorescence channel

images were processed with rolling-ball and Gaussian filters to remove inhomogeneity of illumination

and high frequency noise, respectively. The images were then thresholded using the Otsu method14

with a user-defined fixed multiplier offset and passed through the ImageJ particle analyser to identify

macrophages. Macrophages were tracked between frames using the Apache HBase (v1.3.1; Apache

Software Foundation, https://hbase.apache.org) implementation of the Munkres algorithm with costs

assigned based on object centroid separation15. Instances where objects in the phase-contrast channel

coincided with macrophages identified in the fluorescence channel were removed, as these likely cor-

responded to accidental detection of macrophages.  Finally, spatial relationships between erythroblasts

and macrophages were determined based on the maximum separation of object perimeters.  Multiple

erythroblasts could be assigned to a single macrophage. 

Modeling human erythroid-macrophage interactions

113

RT-PCR analysis 

The following primer sets were used: S18 (forward: 5’-GGACAACAAGCTCCGTGAAGA-3’, reverse:

5’-CAGAAGTGACGCAGCCCTCTA-3’), HPRT (forward: 5’-ATGGGAGGCCATCACATTGT-3’, reverse:

5’-ATGTAATCCAGCCAGGTCAGCAA-3’), MERTK (forward: 5’-ACCTCTGTCGAATCAAAGCCC-3’, reverse:

5’-GCACACTGGTTATGCTGA-AGGA-3’), AXL (forward: 5’-GTGGGCAACCCAGGGAATATC-3’, reverse:

5’-GTACTGTCCCGTGTCGGAAAG-3’), TYRO3 (forward: 5’-CAGCCGGTGAAGCTCAACT-3’, reverse:

5’-TGGCACACCTTCTACCGTGA-3’), TIM3 (forward: 5’-GACTTCACTGCAGCCTTTCC-3’, reverse:

5’-GATCCCTGCTCCGATGTAGA3’), Lactadherin (forward: 5’-GACAAGCAGGGCAACTTCAAC-3’, reverse:

5’-CAGGATGGGCGTCTCAAACAA3’), CD16 (forward: 5’-ACAGGTGCCAGACAAACCTC-3’, reverse:

5’-TTCCAGCTGTGACACCTCAG-3’), CD163 (forward: 5’-AATGGAAAAGGAGGCCATTC-3’, reverse:

5’-TGCTCCATTCAATAGTCCAGG-3’), CD169 (forward: 5’-GGGAGTACAAGTGCTCAGCC-3’, reverse:

5’-GCTTCTGCAGCTCAGTGTCA-3’), CXCR4 (forward: 5’-AGCAGGTAGCAAAGTGACGC-3’, reverse:

5’-ATAGTCCCCTGAGCCCATTT-3’), and CD206 (forward: 5’-TCCTGGTTTTTGCCTCTGTC-3’, reverse:

5’-GCACTGGGACTCACTGCAT-3’). 

Supplementary references

1. Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nat Methods.

2009;6(5):359-362.

2. Rappsilber J, Ishihama Y, Mann M. Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray,

and LC/MS sample pretreatment in proteomics. Anal Chem. 2003;75(3):663-670.

3. Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and pro-

teome-wide protein quantification. Nat Biotechnol. 2008;26(12):1367-1372.

4 .Tyanova S, Temu T, Sinitcyn P, et al. The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat

Methods. 2016;13(9):731-740.

5. Szklarczyk D, Franceschini A, Wyder S, et al. STRING v10: protein-protein interaction networks, integrated over the tree of life.

Nucleic Acids Res. 2015;43(Database issue):D447-452.

6. Lopes CT, Franz M, Kazi F, Donaldson SL, Morris Q, Bader GD. Cytoscape Web: an interactive web-based network browser.

Bioinformatics. 2010;26(18):2347-2348.

7. Maere S, Heymans K, Kuiper M. BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological

networks. Bioinformatics. 2005;21(16):3448-3449.

8. Merico D, Isserlin R, Stueker O, Emili A, Bader GD. Enrichment map: a network-based method for gene-set enrichment visual-

ization and interpretation. PLoS One. 2010;5(11):e13984.

9. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ: 25 years of image analysis. Nat Methods. 2012;9(7):671-675.

10. Schindelin J, Arganda-Carreras I, Frise E, et al. Fiji: an open-source platform for biological-image analysis. Nat Methods.

2012;9(7):676-682.

11. Thevenaz P, Ruttimann UE, Unser M. A pyramid approach to subpixel registration based on intensity. IEEE Trans Image Process.

1998;7(1):27-41.

12. Marr D, Hildreth E. Theory of edge detection. Proc R Soc Lond B Biol Sci. 1980;207(1167):187-217.

13. Tinevez JY, Perry N, Schindelin J, et al. TrackMate: An open and extensible platform for single-particle tracking. Methods.

2017;115:80-90.

14. Otsu N. A Threshold Selection Method from Gray-Level Histograms. IEEE Transactions on Systems, Man, and Cybernetics.

1979;9(1):62-66.

15. Munkres J. Algorithms for the Assignment and Transportation Problems. Journal of the Society for Industrial and Applied Math-

ematics. 1957;5(1):32-38.

Chapter 4

114

Supplementary data

Online Supplementary Table S1. List of proteins expressed in GC-macrophages compared to non-glu-cocorticoid stimulated cells. In blue downregulated proteins, in red upregulated proteins, n=4 (see www.haematologica.org/content/103/3/395)

Online Supplementary Table S2. BiNGO analysis of GC-macrophages combined up and down regulated GOBP GOCC GOMF. FDR<0.05, n=4 (see www.haematologica.org/content/103/3/395).

Table S3. Flow cytometry analysis of MFI macrophage marker expression on FL, BM, unstimulated cells (non-GC) and GC-macrophages. Unpaired t-test, *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001 (n=3-7).

MFI mean±SEM Experiments P-value SignificanceCD14BM vs. FL 5178±907.7 / 6642±2900 N=5 / N=4 0.6127 NSGC vs. non-GC 25545±1825 / 21606±4664 N=4 / N=4 0.4616 NSGC vs. BM 25545±1825 / 5178±907.7 N=4 / N=5 <0.0001 ****GC vs. FL 25545±1825 / 6642 ± 2900 N=4 / N=4 0.0015 **Non-GC vs. BM 21606±4664 / 5178±907.7 N=4 / N=5 0.0060 **Non-GC vs. FL 21606±4664 / 6642±2900 N=4 / N=4 0.0344 *CD16BM vs. FL 1803±342.4 / 8748±1695 N=4 / N=4 0.007 **GC vs. non-GC 3380±658 / 206.4±70.64 N=6 / N=5 0.0019 **GC vs. BM 3380±658 / 1803±342.4 N=6 / N=4 0.1062 NSGC vs. FL 3380±658 / 8748±1695 N=6 / N=4 0.0092 **Non-GC vs. BM 206.4±70.64 / 1803±342.4 N=5 / N=4 0.0014 **Non-GC vs. FL 206.4±70.64 / 8748±1695 N=5 / N=4 0.0007 ***CD163BM vs. FL 49689±7048 / 38955±6209 N=7 / N=5 0.3036 NSGC vs. non-GC 947777±24328 / 3334±1151 N=7 / N=5 0.0107 *GC vs. BM 947777±24328 / 49689±7048 N=7 / N=7 0.1004 NSGC vs. FL 947777±24328 / 38955±6209 N=7 / N=5 0.0891 NSNon-GC vs. BM 3334±1151 / 49689±7048 N=5 / N=7 0.0003 ***Non-GC vs. FL 3334±1151 / 38955±6209 N=5 / N=5 0.0005 ***CD169BM vs. FL 2053±972 / 739.6±270.1 N=7 / N=5 0.2948 NSGC vs. non-GC 3705±1272 / 186.7±107.4 N=6 / N=5 0.0341 *GC vs. BM 3705±1272 / 2053±972 N=6 / N=7 0.3168 NSGC vs. FL 3705±1272 / 739.6±270.1 N=6 / N=5 0.0675 NSNon-GC vs. BM 186.7±107.4 / 2053±972 N=5 / N=7 0.1418 NSNon-GC vs. FL 186.7±107.4 / 739.6±270.1 N=5 / N=5 0.0937 NSCXCR4BM vs. FL 11415±3563 / 2155±1368 N=3 / N=3 0.0723 NSGC vs. non-GC 19596±4359 / 9769±4426 N=7 / N=4 0.1774 NSGC vs. BM 19596±4359 / 11415±3563 N=7 / N=3 0.2896 NSGC vs. FL 19596±4359 / 2155±1368 N=7 / N=3 0.0362 *Non-GC vs. BM 9769±4426 / 11415±3563 N=4 / N=3 0.7956 NSNon-GC vs. FL 9769±4426 / 2155±1368 N=4 / N=3 0.2147 NS

Modeling human erythroid-macrophage interactions

115

Table S3. ContinuedCD206BM vs. FL 2618±1672 / 7569±3026 N=4 / N=5 0.2263 NSGC vs. non-GC 15144±1103 / 55±30.74 N=4 / N=4 <0.0001 ****GC vs. BM 15144±1103 / 2618±1672 N=4 / N=4 0.0008 ***GC vs. FL 15144±1103 / 7569±3026 N=4 / N=5 0.0713 NSNon-GC vs. BM 55±30.74 / 2618±1672 N=4 / N=4 0.1764 NSNon-GC vs. FL 55±30.74 / 7569±3026 N=4 / N=5 0.0647 NSVCAM1BM vs. FL 4524±2338 / 898±439.1 N=4 / N=4 0.1783 NSGC vs. non-GC 30.33±11.22 / 0±0 N=3 / N=3 0.0538 NSGC vs. BM 30.33±11.22 / 4524±2338 N=3 / N=4 0.1652 NSGC vs. FL 30.33±11.22 / 898±439.1 N=3 / N=4 0.1558 NSNon-GC vs. BM 0±0 / 4524±2338 N=3 / N=4 0.1629 NSNon-GC vs. FL 0±0 / 898±439.1 N=3 / N=4 0.1445 NS

Chapter 4

116

+Dex+Mif20

C

+Dex+Mif1+Dex-DexDay 0Isotype

Cells

(rel

ativ

e)Ce

lls(r

elat

ive)

B

0 .0 0 1

0 .0 1

0 .1

1

1 0

C D 1 6

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* *

0 .0 1

0 .1

1

1 0

1 0 0

C D 1 6 3

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

** *

0 .1

1

1 0

1 0 0

C D 1 6 9

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* ***

0 .1

1

1 0

C X C R 4

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

ns

0 .1

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

C D 2 0 6

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* ***

+Dex+Mif20+Dex+Mif1+Dex-DexDay 0Isotype

CD14+ -Dex +Dex

A

+Dex+Mif20

C

+Dex+Mif1+Dex-DexDay 0Isotype

Cells

(rel

ativ

e)Ce

lls(r

elat

ive)

B

0 .0 0 1

0 .0 1

0 .1

1

1 0

C D 1 6

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* *

0 .0 1

0 .1

1

1 0

1 0 0

C D 1 6 3

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

** *

0 .1

1

1 0

1 0 0

C D 1 6 9

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* ***

0 .1

1

1 0

C X C R 4

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

ns

0 .1

1

1 0

1 0 0

1 0 0 0

1 0 0 0 0

C D 2 0 6

Re

lati

ve

ex

pre

ss

ion

(fc

)

D e xM if

D 0 D 3

* ***

+Dex+Mif20+Dex+Mif1+Dex-DexDay 0Isotype

CD14+ -Dex +Dex

A

Modeling human erythroid-macrophage interactions

117

D

E

0

1 0

2 0

3 0

4 0

C D 1 6

MF

In

orm

ali

ze

d(f

c)

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

ns

0

5

1 0

1 5

2 0

C D 1 6 3M

FI

no

rma

liz

ed

(fc

)

**

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

0.0909

0

1 0

2 0

3 0 ns

C D 1 6 9

MF

In

orm

ali

ze

d(f

c)

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

0

2 0

4 0

6 0

8 0

C X C R 4

MF

In

orm

ali

ze

d(f

c)

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

ns

0

1 0 0 0

2 0 0 0

3 0 0 0

C D 2 0 6

MF

In

orm

ali

ze

d(f

c)

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

ns

0.0579

0

5

1 0

1 5

D C -S IG N

MF

In

orm

ali

ze

d(f

c)

D e x (1 M )

M if (2 0 M )

M if (1 M )

4 h rs 2 4 h rs

ns

Figure S1. Macrophage population characteristics. (A) Representative images of cytospins (in 100x magnifica-tion) of CD14+ monocytes directly after isolation, cultured CD14+ cells in the absence (-Dex) or presence (+Dex) of dexamethasone. Note that monocytes are smaller compared to macrophages. (B) Relative mRNA expression of CD16, CD163, CD169, CXCR4, and CD206 on CD14+ cells (D0) cultured for three days (D3) in the presence or absence of dexamethasone (Dex) and/or 20µM mifepristone (Mif) as indicated. Mean ± SEM (ratio paired t-test, *P<0.05, **P<0.01, ***P<0.001). (C) Representative histograms belonging to Figure 1E showing CD16, CD163, CD169, CXCR4, CD206 and DC-SIGN expression on monocytes (n=3) directly after isolation (Day 0) and after culture in the presence or absence of dexamethasone (Dex) and/or mifepristone (Mif1, 1µM; Mif20, 20µM) compared to isotype control. (D) Gating strategy of multi-color flow cytometry experiments showing cells, single cells, and expression of CD16, CD163, CD169, CXCR4 and CD206. (E) CD14+ monocytes were cultured for three days in the presence of dexamethasone and subsequently treated with 1-20µM mifepristone for 4 or 24 hours. Graphs displaying the geometric mean fluorescence intensity (MFI) of CD16, CD163, CD169, CXCR4, CD206 and DC-SIGN (n=2-4). Mean ± SEM (ratio paired t-test, **P<0.01), values normalized to day 0 in Figure 1E.

Chapter 4

118

B

ACe

lls(r

elat

ive)

Cells

(rel

ativ

e)

Day 8

Day 7

Day 3

Isotype

Day 8

Day 7

Day 3

Isotype

Figure S2. Long-term macrophage cultures with dexamethasone. (A) Representative histograms of CD16, CD163, CD169, CXCR4, CD206 expression on differentiated CD14+ monocytes at day 3, 7 and 8 of culture in the presence of dexamethasone (n=4). (B) Representative dot plots show that CD14+ monocytes cultured for 17 days in the presence of dexamethasone maintain CD163, CD169, and CD206 expression (n=3).

Modeling human erythroid-macrophage interactions

119

B

Reticulocytes

Nuclei + Nucleatedcells

Dif D7Dif D1Dif D0

D

-D e x + D e x0

5 0

1 0 0

E r y th ro b la s ts

Bin

din

gc

ap

ac

ity

of

M

(%)

U n b o u n d

1 E B L

2 E B L

3 E B L

4 E B L

5 E B L

6 E B L

**

******

CBF DRAQ5 DRAQ5

Differentiation

E

- D e x + D e x0

2 0

4 0

6 0ns

Av

era

ge

co

nta

ct

m

-EB

L(h

rs)

A

ITG

B 2

ITG

A L

ITG

A M

ITG

A X

E -se le

c t in

L -se le

c t in1 0 0

1 0 0 0

1 0 0 0 0

1 0 0 0 0 0

1 0 0 0 0 0 0M

FI

(no

rma

lize

d)

Figure S3. Integrin expression on GC-macrophages and erythroid differentiation in co-culture with GC-macrophages. (A) Expression of ITGB2, ITGAL, ITGAM, ITGAX, E-selectin and L-selectin on GC-macrophages belonging to Figure 3A-B (n=6). Mean fluorescence intensity (MFI) has been normalized to the isotype control. Mean ± SEM. (B) Representative dot plots of erythroblasts at day 0, 1, and 7 of differentiation (Dif). Upon differ-entiation CD71 expression is reduced and CD235a expression is increased (left panel) while cells also start to enucleate as DRAQ5 stains DNA (right panel) (n=4). (C) Representative ImageStreamX images of the maturation process of erythroblasts at day 7 of differentiation where the nuclei (red, DRAQ5 staining) is expelled from the erythroid cell (n=3). (D) Co-culture of GC-macrophages (+Dex) or unstimulated cells (-Dex) with erythroblasts for 24 hours (n=4). Bars present the percentage of unbound macrophages or macrophages (Mφ) bound to 1 to 6 erythroblasts. Mean ± SD (paired t-test, *P<0.05, **P<0.01, ***P<0.001). (E) Co-culture of GC-macrophages or unstimulated cells with erythroblasts (unpaired t-test of 1153 macrophages (-Dex) and 749 (+Dex), n=5). Images were taken every hour during 64 hours of analysis. Plot showing the mean duration of contact between macro-phages and erythroblasts.

Chapter 4

120

A

0 .0 1

0 .1

1

1 0

1 0 0

T IM 3R

ela

tiv

ee

xp

res

sio

n(f

c)

4 h rs

D e xM if

D 0 D 3

ns

0 .1

1

1 0

1 0 0

L a c ta d h e r in

Re

lati

ve

ex

pre

ss

ion

(fc

)

4 h rs

D e xM if

D 0 D 3

ns

B

C

-Dex

+D e x0

5 0

1 0 0

M

pe

rto

tal

M

(%)

U n b o u n d

N u c le a te d c e lls

R e tic u lo c y te s

P y re n o c y te s

C o m b in e d c e lls

*

D

-D e x + D e x0

5

1 0

1 5

2 0

P y re n o c y te s

M

lin

ke

dto

py

ren

oc

yte

s(%

)

B o u n d

P h a g o c y to s e d

0.0629

0.0785

E

-D e x + D e x0

2 0

4 0

6 0

8 0

1 0 0

N u c le a te d c e lls

Ph

ag

oc

yto

sin

gM

(%)

ns

F

-D e x + D e x0

2 0

4 0

6 0

8 0

1 0 0

R e tic u lo c y te s

Ph

ag

oc

yto

sin

gM

(%)

ns

Figure S4. Binding and phagocytosing capacity of GC-macrophages and unstimulated cells in co-culture with differentiated erythroid cells. Relative mRNA expression of TIM3 (A) and lactadherin (B) in cells from Figure 4A (n=4). Mean ± SEM (ratio paired t-test). (C-F) GC-macrophages or unstimulated cells were co-cultured for 24 hours with erythroid cells at day 6 of differentiation. Cytospins were analysed of 370 macrophages (-Dex) and 313 (+Dex) macrophages. (C) Percentage of macrophages (Mφ) that are unbound, or bound to nucleated cells, reticulocytes or pyrenocytes or a combination of erythroid cells. Mean ± SD (unpaired t-test, *P<0.05, n=3). (D) Percentage of macrophages that bind or phagocytose pyrenocytes. Mean ± SD (unpaired t-test, n=3). (E) Percentage of macrophages phagocytosing nucleated cells. Mean ± SD (paired t-test, n=3). (F) Percentage of macrophages phagocytosing reticulocytes. Mean ± SD (paired t-test, n=3).

Modeling human erythroid-macrophage interactions

121

Bone marrowA

CD14 CD16 CD163 CD169

CXCR4 CD206 VCAM1

Cel

ls(r

elat

ive)

B

Fetal liver

CD14 CD16 CD163 CD169

CXCR4 CD206 VCAM1

Cel

ls(r

elat

ive)

Figure S5. Characterization of human BM and FL macrophages. Representative histograms belonging to Figure 5D showing the expression of macrophage markers CD14, CD16, CD163, CD169, CXCR4, CD206 and VCAM1 (filled) on human BM (A) and FL (B) compared to isotype control (unfilled).

5

The proteomic landscape of erythroblastic island

macrophages from human fetal liver, bone marrow and

in vitro cultures

CHAPTER 5

Esther Heideveld, Floris P.J. van Alphen, Maartje van den Biggelaar, and Emile van den Akker

Submitted

Characterizing human EBI macrophages

125

Erythropoiesis is organized in erythroblastic islands (EBI) and these structures have been observed

in fetal liver (FL) during human development and within adult bone marrow (BM)1, 2. Within EBI, the

central macrophage supports erythroid cell proliferation, differentiation, maturation and enucleation

via adhesion molecules (e.g., ICAM4, VCAM1), soluble factors and pathways regulating pyrenocyte

phagocytosis2-9. Albeit macroscopically identified, efforts to functionally and phenotypically define

EBI macrophages from human FL and BM is incomplete and mainly described in mice10-13. CD163

and CD169 have been found to identify macrophages involved in erythroid support and human

macrophages from FL and adult BM additionally express CD14, CD16, CXCR4, CD206, and VCAM111-

14. Unbiased characterization of protein expression in human counterparts is needed to validate

expression of mechanistically important proteins identified in mice and to uncover novel erythropoi-

esis-supporting pathways. We have reported that human monocytes stimulated with glucocorticoids

differentiate towards erythroid-supporting macrophages (GC-macrophages), closely resembling FL

and BM macrophages14, 15. Including model systems into a global analysis and assuming that macro-

phages from FL and BM provide similar support to erythropoiesis, convergion to shared pathways

that focus on common erythroid support functions may be possible. Here, we report a comparative

proteomic expression profile of CD163+ macrophages from human FL, adult BM and in vitro cultured

GC-macrophages. The findings provide new insights into and confirm previous molecular pathways

that are key to the identity and function of these erythropoiesis-supporting macrophages.

Human FL and adult BM contain a small fraction of CD163+ macrophages12, 14, which were isolated by

magnetic cell sorting (Figure S1A,B). CD163+ FL macrophages showed a homogeneous macrophage

morphology with predominantly ill-defined ragged margins and a spongy nucleus (Figure 1A). The dark

cytoplasmic granules likely correspond to ingested pyrenocytes, indicating contact with erythroid cells

and suggests isolation of functional central macrophages7. CD163+ BM macrophages are smaller, have

distinct margins and had a vacuolated cytoplasm (Figure 1B)16. Although reported to be phenotypi-

cally heterogeneous, this was not reflected in their morphology14. Interestingly, erythroid cells were

still attached to BM and FL macrophages despite magnetic sorting, which was reported before17. In

contrast to CD163+ BM macrophages, nucleated erythroid cells and reticulocytes were engulfed by

FL macrophages, hinting to an additional clearance role. GC-macrophages have very distinct margins,

a compact nucleus, a heavily vacuolated cytoplasm and their cell size was comparable to FL macro-

phages (Figure 1C). In conclusion, FL, BM and GC-macrophages share morphological features, but also

display distinct morphological aspects possibly reflecting niche-instructed differences. Besides low

but consistent co-isolation of interacting erythroid cells, no other contaminating cells were observed.

Chapter 5

126

A

B

C

CD163+ FL macrophages

CD163+ BM macrophages

CD163+ GC-macrophages

i ii iii iv

i ii iii iv

i ii iii iv

* **

*

Figure 1. Morphological analysis of purified human CD163+ FL and BM macrophages and in vitro culture GC macrophages. Representative giemsa/benzidine stained cytospin images of CD163+ FL (A), BM (B) and GC (C) macrophages with 50x (upper) and 100x (lower) magnification. CD163+ in vivo macrophages bind and/or phagocytose erythroblasts (arrowhead), terminally differentiated erythroid cells (big arrow), reticulocytes (asterix) and pyrenocytes (small arrow).

To gain further insight in the molecular pathways involved in EBI structures, we compared the pro-

teome of CD163+ FL (n=3), BM (n=5) and GC-macrophages (n=4). 3536 proteins were identified in

all replicates of at least one of the three populations (Online Supplementary Table S1). Hierarchical

clustering of the Pearson regression correlation matrix in Figure 2A showed high correlation between

replicates within groups ( >0.97 for GC; >0,87 for FL and BM), confirming the uniform cellular mor-

phology. Correlation between samples was on average 0.84, indicating high similarity between CD163+

macrophages and in vitro cultured GC-macrophages. Out of 3536 proteins, 234 were differentially

expressed that cluster in 8 blocks (Figure 2B). Proteins specifically associated with in vitro or in vivo

CD163+ macrophages in clusters 3 and 5 did not enrich into specific processes (data not shown).

Characterizing human EBI macrophages

127

FL-enriched processes (clusters 1,2,6) associated with the VE-cadherin/catenin pathway and hemoglo-

bin function (Figure S2A-C). In agreement with the different waves of erythropoiesis during ontogeny,

HBG1, HBG2 and HBZ were restricted to FL macrophages, probably as a consequence of erythroid

phagocytosis (Figure S2D). In addition, FL macrophages express HBM, originally described as a pseu-

do-alpha2 globin gene but recently found to be expressed in erythroid cells18, 19. Upregulated proteins

in BM macrophages (clusters 4,7,8) reveal significant enrichment of multiple processes (Figure S3A). In

addition, protein-protein network analysis showed two predominant clusters associated with coagu-

lation and adult hemoglobin HBB and HBD (Figure S3B-D). Although cultured GC-macrophages never

encountered erythroid cells before analysis, HBA1 and HBB were found in all three populations albeit

significantly lower expressed in GC-macrophages (Figure S3D). LPS and interferon-induced hemoglo-

bin expression has been reported in mouse macrophages20, 21. However, the function of hemoglobin

in macrophages is unknown. In conclusion, CD163+ FL, BM and GC macrophages are highly similar

in protein expression and expression of erythroid specific proteins suggest active erythroid support.

Next, we continued with a more stringent list restricted to proteins either identified in all replicates or

not identified at all within the other populations (Figure 2C, Online Supplementary Table S1). Within

these 2050 proteins, 83% (n=1702) were identified in all macrophage populations, 0.8% were uniquely

expressed in BM, 7.8% in GC-macrophages and 2.9% in FL. Principal component analysis separated

based on tissue origin (PC2) and in vitro culture (PC1; Figure 2D). Gene ontology analysis revealed

that these 1702 common proteins function with a select number of enriched processes revealing

their myeloid identity (Figure 2E). The data validated CD11a (ITGAL), CD11b (ITGAM), CD11c (ITGAX),

CD13 (ANPEP), CD14, CD36, CD169 (SIGLEC1), CD206 (MRC1), and coagulation factor XIII (F13A1) as

EBI macrophage markers previously found expressed on BM or FL macrophages, including mature

tissue macrophage proteins VSIG4, MPEG1, and SGPL1 (Online Supplementary Table S1)1, 13, 14, 22-26.

In addition, Figure 3 depicts novel surface antigens not assigned to EBI macrophages before and

expressed in all analyzed macrophages: CD29 (ITGB1), CD41 (ITGA2B), CD47, CD54 (ICAM1), CD59,

CD61 (ITGB3), CD63, CD71 (TFRC), CD82, CD84, CD85c (LILRB5), CD91 (LRP1), CD97, CD98 (SLC3A2),

CD107a (LAMP1), CD107b (LAMP2), CD147 (BSG), CD156c (ADAM10), CD180, CD205 (LY75), IGF2R

(CD222), CD281 (TLR2). Within the commonly expressed proteins, 32 were differentially expressed

(or 1.9% of total), including 4 novel EBI-allocated membrane proteins: CD63, LILRB5, F13A1 and

ANPEP (Figure S4).

EBI macrophages support erythropoiesis by interacting with erythroid cells of all maturation stages13,

17, 27. CD169 (SIGLEC1) and SIGLEC9 were both identified in all populations and could support bind-

ing to erythroid cells via GPA28. Tmod3-/- mice display defective adhesive interactions in FL EBI and

impaired erythroid differentiation29. Figure S5A showed that TMOD3 was identified in all three mac-

rophage populations. In addition, integrins CD11a and CD29 that bind to ICAM4 on erythroblasts

were expressed by all macrophages (Figure 3)30-32. Interestingly, VCAM1 (CD106) and CD163L1 were

commonly expressed by BM and FL macrophages. Recently it was shown however that Vcam-/- mice

Chapter 5

128

A

-25-20-15-10

-505

1015202530

-30 -20 -10 0 10 20 30

Com

pene

nt 2

(27.

6%)

Component 1 (41.4%)

FL

BM

GC

C

E

D

B

12

3

4

5

6

7

8

GC FL BM

Figure 2. Proteomic comparison of human CD163+ FL, BM and GC macrophages. (A) Correlation plots (lower left half) and heatmap of Pearson correlation coefficients (top right half) of FL (n=3), BM (n=5) and GC-macrophages (n=4) of the 3536 proteins that were identified in at least all replicates of at least one of the three populations in all donors in Online Supplementary Online Supplementary Table S1. Note that the scale runs from 0.75 to 1. P-val-ues are depicted in Table S2. (B) Heatmap and hierarchical clustering (based on average Euclidean distance and pre-processed with K-means) using Z-scored LFQ-values of the regulated proteins (n=234) (ANOVA FDR 0.05 S0 0.4). (C) Venn diagram depicting the qualitative distribution of the 2050 proteins that were either quantified in all replicates in human FL, BM and GC-macrophages or not identified at all within one or two populations. (D) Principal component analysis of FL, BM and GC-macrophages on the 1702 proteins that were quantified in all populations. (E) Similarity matrix of the significantly enriched gene ontology terms (row and columns) found within the 1702 proteins identified in all populations (R Package Gogadget) using 3536 proteins identified as reference. The overlap index indi-cates the distances calculated using Gogadget. Dark blue clusters show GO-terms that share similarity (Figure S2).

Characterizing human EBI macrophages

129

display normal erythropoiesis27. Another key regulator of EBI is EMP (MAEA), which was only detected

in GC-macrophages (Figure S5A). Emp-deficient mice die perinatally due to reduced EBI4, 33, 34. The

heme-catalyzing proteins HMOX1 and HMOX2 were identified in all macrophages, although HMOX1

is significantly lower expressed in GC-macrophages (Figure S5A). HMOX1-null patients have anemia,

microcytosis an abnormal iron metabolism, and Hmox1-deficient mice have reduced EBI BM macro-

phages, severely depleted splenic macrophages and altered adhesion molecule expression on both

erythroblasts and EBI macrophages resulting in the inability to form BM EBI35.

Another characteristic of EBI macrophages is phagocytosis of pyrenocytes. Mouse BM EBI mac-

rophages phagocytose phosphatidylserine-exposed pyrenocytes via MERTK in a GAS6 or Protein

S dependent manner7. Figure S5B showed that TAM-receptor family members are expressed by

all groups. FL and BM macrophages expressed AXL and TIMD4 and GC-macrophages expressed

MERTK, which all have similar function6, 36. Stabilin (STAB) binds phosphatidylserine to aid phagocy-

tosis37. STAB1 was identified in all FL and GC-macrophages and some BM macrophages (3 out of 5),

while downstream proteins essential for STAB function RAC1, ELMO1 and ROCK1 were detected in

all macrophages (Figure S5B, Online Supplementary Table S1). Interestingly, STAB2 was restricted

to FL macrophages. Macrophage-capping protein (CAPG) which is important for the phagocytic

capacity of macrophages in mice38, 39 was identified in all macrophages (Figure S5B). Furthermore,

several caspases, cathepsin-associated proteins, and cytokines, such as TGF-β1 (TGFB1), IL16 and

IL18 were identified in all populations, suggesting a paracrine function (Online Supplementary Table

S1). DNASE2, which plays an important role during DNA degradation after phagocytosis, was iden-

tified in all FL and GC-macrophages, and some BM macrophages (3 out of 5) (Figure S5B). Besides

commonly expressed proteins, the proteomics data revealed some uniquely expressed proteins in

EBI macrophages. CD163+ macrophages from BM can be discriminated from FL macrophages using

BM-restrictive CD72 and CD66c (CEACAM6) and FL-restrictive CD26 (DPP4), CD144 (CDH5), CD49d

(ITGA1), and ‘ITGA9’. GC-macrophages exclusively express CD49d (ITGA4), CD85h (LILRA2), CD166

(ALCAM), CD232 (PLXNC1), CD184 (CXCR4), CD305 (LAIR1), and CD328 (SIGLEC7) (Online Supple-

mentary Table S1).

In conclusion, a unique resource is provided with the unbiased proteomic analyses of human CD163+

BM and FL EBI macrophages. Several markers display similar expression and from this a common

novel erythroid supporting macrophage phenotype could be deduced. Restricting the analysis to in

vivo purified CD163+ cells, VCAM1, CD317 (BST2), CD163L1 and ITGB5 may be additionally used as

markers. The proteome comparison of GC-macrophages with FL and BM showed that 83% of the pro-

teome was shared between these three macrophage populations. High similarity to in vitro cultured

macrophages suggests that specific monocyte differentiation cues may lead to erythroid-supporting

macrophages. The identification of co-expressed proteins within the studied populations now allows

for functional studies in vitro and in vivo.

Chapter 5

130

F LB M G

C

2 5

3 0

3 5

C D 2 9 ( IT G B 1 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 4

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 6 1 ( IT G B 3 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 8 4

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 5 9

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 5 4 (IC A M 1 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 4 7

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 3 (A N P E P )

LF

Q-v

alu

e

*

F LB M G

C

2 5

3 0

3 5

C D 8 2

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 7 1 (T F R C )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 4 1 ( IT G A 2 B )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 1 c ( IT G A X )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 1 b ( IT G A M )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 1 a ( IT G A L )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 3 6

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 6 3

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 9 1 (L R P 1 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 8 5 c (L IL R B 5 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 9 8 (S L C 3 A 2 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 9 7

LF

Q-v

alu

e

Characterizing human EBI macrophages

131

F LB M G

C

2 5

3 0

3 5

C D 2 8 1 (T L R 2 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 2 2 2 (IG F 2 R )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 2 0 6 (M R C 1 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 2 0 5 (L Y 7 5 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 8 0

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 5 6 c (A D A M 1 0 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 4 7 (B S G )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 0 7 b (L A M P 2 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C D 1 0 7 a (L A M P 1 )L

FQ

-va

lue

F LB M G

C

2 5

3 0

3 5

C D 1 6 9 (S IG L E C 1 )

LF

Q-v

alu

e

Figure 3. Common protein expression of macrophages from different origin. Graphs presenting the LFQ values of previously described and novel EBI-associated proteins for CD163+ macrophages from FL, BM and in vitro culture.

Author contributions

EH performed the experiments. FPJA and MB performed the mass spectrometry data acquisition

and analysis. EH and EA designed the experiments, analyzed the data and wrote the manuscript. The

manuscript was critically revised by all authors. The authors declare no competing financial interests.

Acknowledgements

We are grateful to the staff of the CASA clinic in Leiden for collecting human fetal tissues and Dr.

Tom Cupedo, Natalie Papazian and Martijn Bogaerts from the Erasmus Medical Center Rotterdam for

providing human fetal liver material. We also would like to thank Marten Hansen (Sanquin Research,

Department of Hematopoiesis, Amsterdam) for his technical support concerning R-package for

GO-gadgets. This work was supported by grants from the Landsteiner Foundation (LSBR1141; EA,

EH and LSBR1517; MB).

Chapter 5

132

References

1. Lee SH, Crocker PR, Westaby S, et al. Isolation and immunocytochemical characterization of human bone marrow stromal

macrophages in hemopoietic clusters. J Exp Med. 1988;168(3):1193-1198.

2. Mohandas N, Prenant M. Three-dimensional model of bone marrow. Blood. 1978;51(4):633-643.

3. 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(8):2940-2950.

4. Soni S, Bala S, Gwynn B, Sahr KE, Peters LL, Hanspal M. Absence of erythroblast macrophage protein (Emp) leads to failure of

erythroblast nuclear extrusion. J Biol Chem. 2006;281(29):20181-20189.

5. Sawada K, Krantz SB, Dessypris EN, Koury ST, Sawyer ST. Human colony-forming units-erythroid 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(5):1701-1709.

6. Nishi C, Toda S, Segawa K, Nagata S. Tim4- and MerTK-mediated engulfment of apoptotic cells by mouse resident peritoneal

macrophages. Mol Cell Biol. 2014;34(8):1512-1520.

7. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by central macrophages in erythroblastic islands.

Blood. 2014;123(25):3963-3971.

8. Toda S, Nishi C, Yanagihashi Y, Segawa K, Nagata S. Clearance of Apoptotic Cells and Pyrenocytes. Curr Top Dev Biol. 2015;114:267-

295.

9. Yoshida H, Kawane K, Koike M, Mori Y, Uchiyama Y, Nagata S. Phosphatidylserine-dependent engulfment by macrophages of

nuclei from erythroid precursor cells. Nature. 2005;437(7059):754-758.

10. Heideveld E, van den Akker E. Digesting the role of bone marrow macrophages on hematopoiesis. Immunobiology.

2017;222(6):814-22.

11. Fabriek BO, Polfliet MM, Vloet RP, van der Schors RC, Ligtenberg AJ, Weaver LK, et al. The macrophage CD163 surface glycoprotein

is an erythroblast adhesion receptor. Blood. 2007;109(12):5223-5229.

12. Chow A, Lucas D, Hidalgo A, et al. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and

progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011;208(2):261-271.

13. Chow A, Huggins M, Ahmed J, et al. CD169(+) macrophages provide a niche promoting erythropoiesis under homeostasis and

stress. Nat Med. 2013;19(4):429-436.

14. Heideveld E, Hampton-O’Neil LA, Cross SJ, et al. Glucocorticoids induce differentiation of monocytes towards macrophages that

share functional and phenotypical aspects with erythroblastic island macrophages. Haematologica. 2018;103(3):395-405.

15. Heideveld E, Masiello F, Marra M, et al. CD14+ cells from peripheral blood positively regulate hematopoietic stem and progenitor

cell survival resulting in increased erythroid yield. Haematologica. 2015;100(11):1396-1406.

16. Fauser AA, Messner HA. Identification of megakaryocytes, macrophages, and eosinophils in colonies of human bone marrow

containing neurtophilic granulocytes and erythroblasts. Blood. 1979;53(5):1023-1027.

17. 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(1):411-415.

18. Kim MS, Pinto SM, Getnet D, et al. A draft map of the human proteome. Nature. 2014;509(7502):575-581.

19. Goh SH, Lee YT, Bhanu NV, et al. A newly discovered human alpha-globin gene. Blood. 2005;106(4):1466-1472.

20. Saha D, Patgaonkar M, Shroff A, Ayyar K, Bashir T, Reddy KV. Hemoglobin expression in nonerythroid cells: novel or ubiquitous?

Int J Inflam. 2014;2014:803237.

21. Liu L, Zeng M, Stamler JS. Hemoglobin induction in mouse macrophages. Proc Natl Acad Sci U S A. 1999;96(12):6643-6647.

22. Slayton WB, Juul SE, Calhoun DA, Li Y, Braylan RC, Christensen RD. Hematopoiesis in the liver and marrow of human fetuses at 5

to 16 weeks postconception: quantitative assessment of macrophage and neutrophil populations. Pediatr Res. 1998;43(6):774-

782.

23. Bhoopat L, Taylor CR, Hofman FM. The differentiation antigens of macrophages in human fetal liver. Clin Immunol Immunopathol.

1986;41(2):184-192.

24. Kappelmayer J, Bacsko G, Birinyi L, Zakany R, Kelemen E, Adany R. Consecutive appearance of coagulation factor XIII subunit A

in macrophages, megakaryocytes, and liver cells during early human development. Blood. 1995;86(6):2191-2197.

25. Edwards JA, Jones DB, Evans PR, Smith JL. Differential expression of HLA class II antigens on human fetal and adult lymphocytes

and macrophages. Immunology. 1985;55(3):489-500.

26. Thiele J, Braeckel C, Wagner S, et al. Macrophages in normal human bone marrow and in chronic myeloproliferative disorders:

an immunohistochemical and morphometric study by a new monoclonal antibody (PG-M1) on trephine biopsies. Virchows Arch

A Pathol Anat Histopathol. 1992;421(1):33-39.

Characterizing human EBI macrophages

133

27. 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(1):86-94.

28. Lizcano A, Secundino I, Dohrmann S, et al. Erythrocyte sialoglycoproteins engage Siglec-9 on neutrophils to suppress activation.

Blood. 2017;129(23):3100-3110.

29. 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(5):758-767.

30. Lee G, Lo A, Short SA, et al. Targeted gene deletion demonstrates that the cell adhesion molecule ICAM-4 is critical for erythro-

blastic island formation. Blood. 2006;108(6):2064-2071.

31. Ulyanova T, Jiang Y, Padilla S, Nakamoto B, Papayannopoulou T. Combinatorial and distinct roles of alpha(5) and alpha(4) integrins

in stress erythropoiesis in mice. Blood. 2011;117(3):975-985.

32. Spring FA, Parsons SF, Ortlepp S, et al. Intercellular adhesion molecule-4 binds alpha(4)beta(1) and alpha(V)-family integrins

through novel integrin-binding mechanisms. Blood. 2001;98(2):458-466.

33. Soni S, Bala S, Kumar A, Hanspal M. Changing pattern of the subcellular distribution of erythroblast macrophage protein (Emp)

during macrophage differentiation. Blood Cells Mol Dis. 2007;38(1):25-31.

34. Soni S, Bala S, Hanspal M. Requirement for erythroblast-macrophage protein (Emp) in definitive erythropoiesis. Blood Cells Mol

Dis. 2008;41(2):141-147.

35. Fraser ST, Midwinter RG, Coupland LA, et al. Heme oxygenase-1 deficiency alters erythroblastic island formation, steady-state

erythropoiesis and red blood cell lifespan in mice. Haematologica. 2015;100(5):601-610.

36. Meyer AS, Zweemer AJ, Lauffenburger DA. The AXL Receptor is a Sensor of Ligand Spatial Heterogeneity. Cell Syst. 2015;1(1):25-

36.

37. D’Souza S, Park SY, Kim IS. Stabilin-2 acts as an engulfment receptor for the phosphatidylserine-dependent clearance of primary

necrotic cells. Biochem Biophys Res Commun. 2013;432(3):412-417.

38. Parikh SS, Litherland SA, Clare-Salzler MJ, Li W, Gulig PA, Southwick FS. CapG(-/-) mice have specific host defense defects that

render them more susceptible than CapG(+/+) mice to Listeria monocytogenes infection but not to Salmonella enterica serovar

Typhimurium infection. Infect Immun. 2003;71(11):6582-6590.

39. Witke W, Li W, Kwiatkowski DJ, Southwick FS. Comparisons of CapG and gelsolin-null macrophages: demonstration of a unique

role for CapG in receptor-mediated ruffling, phagocytosis, and vesicle rocketing. J Cell Biol. 2001;154(4):775-784.

Chapter 5

134

Supplementary methods

Human materials

Human blood (n=4), BM (n=5) and FL (n=3) mononuclear cells were purified by density separa-

tion using Ficoll-Paque (manufacturer protocol). For human blood, informed consent was given

in accordance with the Declaration of Helsinki and the Dutch National and Sanquin Internal Ethic

Boards. Human adult BM aspirates were obtained after informed consent from the sternum of

patients undergoing cardiac surgery, and approved by the Medical Ethical Review Board of the AMC

(MEC:04/042#04.17.370). Human fetal tissues were obtained from elective abortions contingent on

informed consent and approval by the Medical Ethical Commission of the Erasmus University Medical

Center Rotterdam (MEC-2006-202). Gestational age was determined by ultrasonic measurement of

the skull or femur diameter and ranged from 17-21 weeks.

Cell culture and isolation

GC-macrophages were generated by culturing CD14+ cells isolated from PBMC for three days as

described previously1. CD163+ macrophages were isolated from both FL and BM by a two-step

purification protocol. First, cells were stained with 20μl/70x106 cells anti-CD163 PE (Miltenyi Biotec,

Bergisch Gladbach, Germany) for 30min at 4C. Thereafter, cells were incubated for 30min at 4C with

50μl/70x106 cells anti-mouse IgG MicroBeads (Miltenyi Biotec). Magnetic-activated cell sorting was

started with an LS column and succeeded by a MS column. After purification, cells were measured

on the LSRFortessa (BD Biosciences, Oxford, UK) and analyzed using FlowJo software (FlowJo v10;

Tree Star, Inc., Ashland, OR, USA).

Cytospins

Cells were cytospun using Shandon Cytospin II (Thermo Scientific), dried and fixed in methanol. Cells

were stained with benzidine in combination with the Differential Quik Stain Kit (PolySciences, War-

rington, PA) (manufacturer’s protocol). Slides were dried, subsequently embedded in Entellan (Merck,

Darmstadt, Germany), covered with a coverslip and images were acquired on a Leica microscope

(Leica DM-2500, Germany).

Mass spectrometry data acquisition and analysis

Purified CD163+ FL macrophages, BM macrophages and cultured GC-macrophages samples were pre-

pared for mass spectrometry and data acquisition analysis was performed as previously described1.

Peptides were prepared from whole-cell lysates using the filter-aided sample preparation (FASP)

method2, and peptides were desalted and concentrated using C18 StageTips3. MS analyses were per-

formed on a Dionex U3000 RSLC coupled to an Orbitrap Fusion Tribrid mass spectrometer (Thermo

Fisher Scientific). Peptides were separated on a ReproSil-Pur C18-AQ, 1.9 μm resin (Dr Maisch; 75–360

μm inner-outer diameter fused silica emitter [New Objective] and 20-cm length) with a 162-min gra-

dient. The instrument was run in top speed mode with 3-s cycles. The RAW MS files were processed

Characterizing human EBI macrophages

135

with the MaxQuant computational platform4 (v1.5.2.8) using human sequences from the Uniprot-

Swiss-prot database (Uniprot; release 2015-02; 89,796 entries) with a false discovery rate below 1%

for both peptides and proteins. Standard settings with the additional options match between runs,

LFQ, and unique peptides for quantification were selected. The generated “proteingroups.txt” table

was filtered for potential contaminants, reverse hits, and “only identified by site”. The LFQ values were

transformed in log2 scale, the biological replicates per experimental condition grouped, and proteins

were filtered to contain at least all valid values in one of the experimental groups (n=3536). Missing

values were imputed by normal distribution (width = 0.3; shift = 1.8), assuming these proteins were

close to the detection limit. Quantitative significance (Pearson correlation coefficient, principal-com-

ponent analysis, ANOVA and hierarchical clustering) was performed by Perseus software5 (v1.5.1.6).

To identify the proteins with the most prominent differences in expression profiles within the different

macrophage subsets, we used the built-in ANOVA functions in PERSEUS using an FDR of 5% and

S0 of 0.4. This identified a total of 234 differentially expressed proteins which were clustered using

hierarchical clustering based on k-means. For the individual or combined clusters, interaction network

analysis was performed using STRING (v10.5, settings: molecular action, all parameters except text-

mining and gene fusion, medium confidence of 0.400). Further enrichment analysis of the individual

or combined clusters was performed using R Package Gogadget, using the 3536 proteins quantified

by the mass spectrometer as a background, using a cut-off value of 0.056.

Supplementary references

1. Heideveld E, Hampton-O’Neil LA, Cross SJ, et al. Glucocorticoids induce differentiation of monocytes towards macrophages that

share functional and phenotypical aspects with erythroblastic island macrophages. Haematologica. 2018;103(3):395-405.

2. Wisniewski JR, Zougman A, Nagaraj N, Mann M. Universal sample preparation method for proteome analysis. Nat Methods.

2009;6(5):359-362.

3. Rappsilber J, Ishihama Y, Mann M. Stop and go extraction tips for matrix-assisted laser desorption/ionization, nanoelectrospray,

and LC/MS sample pretreatment in proteomics. Anal Chem. 2003;75(3):663-670.

4. Cox J, Mann M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and pro-

teome-wide protein quantification. Nat Biotechnol. 2008;26(12):1367-1372.

5. Tyanova S, Temu T, Sinitcyn P, et al. The Perseus computational platform for comprehensive analysis of (prote)omics data. Nat

Methods. 2016;13(9):731-740.

6. Nota B. Gogadget: An R Package for Interpretation and Visualization of GO Enrichment Results. Mol Inform. 2017;36(5-6).

Chapter 5

136

Supplementary data

Tabl

e S1

. Lis

t of p

rote

ins

expr

esse

d in

FL

mac

roph

ages

, BM

mac

roph

ages

and

GC-

mac

roph

ages

. Se

e on

line

at h

ttp:

//hd

l.han

dle.

net/

1124

5.1/

3cd5

12b4

-a92

2-4e

17-8

0d5-

7749

a7fa

ad8e

.

Tabl

e S2

. Sta

tist

ical

ana

lysi

s of

CD

163+ F

L m

acro

phag

es, B

M m

acro

phag

es, a

nd G

C-m

acro

phag

es. P

-val

ues

are

depi

cted

of F

L (n

=3),

BM (n

=5) a

nd G

C-m

acro

phag

es (n

=4) c

orre

spon

ding

to F

igur

e 2A

.

BM1

BM 2

BM 3

BM 4

BM 5

FL 1

FL 2

FL 3

GC

1G

C 2

GC

3G

C4

NaN

0.87

756

0.89

7801

0.86

463

0.88

6552

0.79

3836

0.80

7698

0.78

6622

0.74

8491

0.77

0522

0.75

3938

0.76

983

BM 1

0.87

756

NaN

0.86

9605

0.89

0744

0.90

6365

0.78

3522

0.81

0256

0.77

7001

0.76

6854

0.79

505

0.76

6965

0.79

4907

BM 2

0.89

7801

0.86

9605

NaN

0.90

8425

0.89

7386

0.80

5299

0.81

271

0.79

2109

0.77

188

0.79

3343

0.77

680.

7931

9BM

3

0.86

463

0.89

0744

0.90

8425

NaN

0.90

8193

0.77

8361

0.79

471

0.76

3745

0.81

6464

0.83

7323

0.81

8176

0.84

1178

BM 4

0.88

6552

0.90

6365

0.89

7386

0.90

8193

NaN

0.85

9686

0.88

2065

0.85

547

0.86

8545

0.88

628

0.87

0965

0.88

0057

BM 5

0.79

3836

0.78

3522

0.80

5299

0.77

8361

0.85

9686

NaN

0.94

3241

0.95

3925

0.81

0672

0.81

3054

0.81

1201

0.81

2006

FL 1

0.80

7698

0.81

0256

0.81

271

0.79

471

0.88

2065

0.94

3241

NaN

0.94

3683

0.83

5057

0.83

9825

0.83

519

0.83

8904

FL 2

0.78

6622

0.77

7001

0.79

2109

0.76

3745

0.85

547

0.95

3925

0.94

3683

NaN

0.79

9149

0.80

202

0.79

8787

0.79

5623

FL 3

0.74

8491

0.76

6854

0.77

188

0.81

6464

0.86

8545

0.81

0672

0.83

5057

0.79

9149

NaN

0.97

4263

0.97

2021

0.97

3856

GC

1

0.77

0522

0.79

505

0.79

3343

0.83

7323

0.88

628

0.81

3054

0.83

9825

0.80

202

0.97

4263

NaN

0.97

4603

0.97

8207

GC

2

0.75

3938

0.76

6965

0.77

680.

8181

760.

8709

650.

8112

010.

8351

90.

7987

870.

9720

210.

9746

03N

aN0.

9780

2G

C 3

0.76

983

0.79

4907

0.79

319

0.84

1178

0.88

0057

0.81

2006

0.83

8904

0.79

5623

0.97

3856

0.97

8207

0.97

802

NaN

GC

4

Characterizing human EBI macrophages

137

A

B

Total FL

FSC-A

SSC

-A

Total FL CD163+ FL

Total BM Total BM CD163+ BM

FSC-A

SSC

-A

CD

163

CD

163

CD

163

i ii iii

i iiiii CD

163

Figure S1. Purification of CD163+ macrophages from fetal liver and adult bone marrow. Representative dot plots showing the total cell population (i), the percentage of CD163+ macrophages before MACS isolation (ii), and the CD163+ macrophage population after MACS isolation (iii) in both FL (A) and BM (B).

Chapter 5

138

A

C

D

B

F LB M G

C

2 5

3 0

3 5

H B Z

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

H B G 1

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

H B G 2

LF

Q-v

alu

e

N D N D

GO-terms upregulated in FL

Hemoglobin complex

Molecular carrier activity

Oxygen carrier activity

Oxygen binding

Oxygen transport

Gas transport

F LB M G

C

2 5

3 0

3 5

R H O C

LF

Q-v

alu

eN D N D

F LB M G

C

2 5

3 0

3 5

C T N N D 1L

FQ

-va

lue

F LB M G

C

2 5

3 0

3 5

C T N N A 1

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C T N N B 1

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

C D H 5

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

H B M

LF

Q-v

alu

e

N D

Figure S2. Specific processes upregulated in CD163+ FL macrophages. (A) Table showing the GO-terms that were upregulated in FL compared to BM (clusters 1,2,6 of the heatmap in Figure 2B) using R Package Gogadget. (B) Protein-protein network analysis using STRING-analysis on proteins upregulated in FL compared to BM (clusters 1,2,6 of the heatmap in Figure 2B) depicts few molecular pathways associated. (C,D) Graphs presenting the LFQ values of hemoglobins (C) and the VE-cadherin (CDH5) and catenin pathway (D) upregulated in panel A for all CD163+ macrophages and in vitro cultured GC-macrophages.

Characterizing human EBI macrophages

139

A B

A B

Figure S3. Specific processes upregulated in CD163+ BM macrophages. (A) GO-terms associated with proteins that were upregulated in BM compared to FL (clusters 4,7,8 of the heatmap in Figure 2B), a similarity matrix of the significantly enriched gene ontology terms (row and columns) is shown using R Package Gogadget. The overlap index indicates the distances calculated using Gogadget. Dark blue clusters show GO-terms that share similarity. (B) STRING-analysis on proteins from panel C depicts an elaborate network of processes specifically involved in CD163+ BM macrophages (cluster 4,7,8). (C-E) Graphs presenting the LFQ values of the Serpin protease pathway (C) and hemoglobins (D) upregulated in panel B for all CD163+ macrophages and in vitro cultured GC macrophages. (E) Graphs presenting the LFQ values of the remaining hemoglobins identified in Online Supplementary Table S1.

A B

A B

Chapter 5

140

C

F LB M G

C

2 5

3 0

3 5

K N G 1

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

A P O H

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

A 2 M

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

S E R P IN C 1

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

P L G

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

S E R P IN C 1

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

S E R P IN A 3

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C F H

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

A 1 B G

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

O R M 1

LF

Q-v

alu

eN D N D

F LB M G

C

2 5

3 0

3 5

H B A 1

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

H B B

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

H P

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

H B D

LF

Q-v

alu

e

N D N D

D

F LB M G

C

2 5

3 0

3 5

H R GL

FQ

-va

lue

N D N D

F LB M G

C

2 5

3 0

3 5

F 1 3 A 1

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C 3

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

A H S G

LF

Q-v

alu

e

N D

Figure S3. Continued

Characterizing human EBI macrophages

141

Cluster 1 Cluster 2 Cluster 3

BM

FL

GC

LYZ

S100

A9MNDA

S100

A8MPO

AZU1

CTSG

ELAN

EBP

IHP LTF

SUN2

ANPEP

CPM

CAPN

2CA

PGTSPO

LGALS3

FSCN

1HM

OX1

EPB4

1L2

LILRB5

HBA1

FABP

3ME1

BLVR

BGA

TMFCGR

TC3 TTR

A2M

HBB

Figure S4. Significantly differentially expressed proteins within the 1702 commonly expressed proteins by all macrophages analyzed. Unsupervised hierarchical clustering of the differential proteins found within the 1702 commonly expressed proteins (n=32) and the different macrophage populations analyzed. Three clusters are indicated and colors represent Z-scored LFQ values.

Chapter 5

142

F LB M G

C

2 5

3 0

3 5

H M O X 2

LF

Q-v

alu

eF L

B M GC

2 5

3 0

3 5

H M O X 1L

FQ

-va

lue

F LB M G

C

2 5

3 0

3 5

V C A M 1

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

T M O D 3

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

S IG L E C 9

LF

Q-v

alu

e

A

F LB M G

C

2 5

3 0

3 5

C D 1 6 9 (S IG L E C 1 )

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

E M P (M A E A )

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

D N A S E 2

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

S T A B 2

LF

Q-v

alu

e

N D N D

F LB M G

C

2 5

3 0

3 5

A X L

LF

Q-v

alu

e

N D

B

F LB M G

C

2 5

3 0

3 5

M E R T K

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

T IM D 4

LF

Q-v

alu

e

N D

F LB M G

C

2 5

3 0

3 5

S T A B 1

LF

Q-v

alu

e

F LB M G

C

2 5

3 0

3 5

C A P G

LF

Q-v

alu

e

Figure S5. Protein expression associated with EBI functionality. Graphs presenting the LFQ values of EBI-as-sociated proteins for CD163+ macrophages from FL, BM and in vitro cultured GC macrophages, corresponding to (A) erythroid-macrophage binding and (B) phagocytosis and DNA degradation.

6

General discussion

CHAPTER 6

General discussion

147

The research presented in this thesis concerns the identification, phenotypic analysis and role of

human tissue resident macrophages as support cells during erythropoiesis. Unraveling the mech-

anisms through which macrophages drive erythroid expansion and differentiation is important not

only to understand the normal processes involved in erythropoiesis, but also to start understand-

ing specific erythroid pathologies in which support cells play an important but ill-defined role (e.g.,

myelodysplastic syndrome, polycythemia vera and beta-thalassemia). In addition, this data could

be used to design and optimize new model systems to study interactions between erythroid cells

and macrophages and to optimize current in vitro erythropoiesis protocols with the aim to produce

transfusion-ready red cells, the Holy Grail of transfusion medicine.

Support of hematopoietic stem and progenitor cells

In this thesis, we demonstrated in chapter 2 that during the expansion phase of erythroblast cultures

from peripheral blood mononuclear cells (PBMC) containing StemSpan supplemented with erythro-

poietin (EPO), SCF, and dexamethasone, CD14+ cells present in PBMC increase the erythroblast yield.

This research was initiated by our previous finding that the erythroid yield from total PBMC is 10-15

fold higher compared to CD34+ cells isolated from the same amount of PBMC1. The data of chapter 2

and chapter 4 showed that CD14+ cells have a supportive function during in vitro erythropoiesis. Our

novel flow cytometry tool based on lineage depletion (CD3/CD14/CD19; Lin-) and the expression of

CD34 and CD36 described in chapter 2 allows for easy tracking of erythroid commitment from CD34+

hematopoietic stem and progenitor cells (HSPC) and sorting/isolating of specific stages during differ-

entiation from HSPC to erythroblasts. The data suggests that CD14+ cells do not promote erythroid

expansion, but increase both CD34+CD36- HSPC and CD34+CD36+ megakaryocyte-erythroid progen-

itor (MEP) cell numbers through cell survival and thereby increase the erythroid yield. CD14+ cells are

a heterogeneous population consisting of classical, intermediate and non-classical monocytes based

on the expression of CD14 and CD162. In chapter 2 we showed that in particular the intermediate

monocytes (CD14++CD16+) increase the erythroid yield. Not only the erythroid yield from PBMC was

increased, but CD14+ cells also positively affect CD34+ cells from human bone marrow and cord blood,

although to a lower extent, which might be caused by inter-donor variabilities. CD169+ macrophages

are in close proximity to HSPC and retain HSPC within that specific niche3. Besides this function and

given the data in chapter 2, CD14+ cells in the bone marrow could also affect HSPC survival in vivo.

In addition, both HSPC and MEP were increased upon co-culturing with CD14+ cells, which suggests

that other lineages, in particular megakaryocyte cultures, but also stem cell expansion protocols could

benefit from co-cultures with CD14+ cells during the first days of culture. Because CD34+ cell survival

does not rely on cell-cell contact but on factors that are secreted by CD14+ cells, we have tested the

effect of IL10 as a usual tissue resident macrophage secreted interleukin suspect on HSPC expansion

and survival by adding IL10 to CD34+ cells. However, no positive or negative effect was observed.

Because it was unknown what other survival factors are secreted by CD14+ cells, proteomic analyses

were performed on CD14+ monocytes supernatant of two donors cultured for 24 hours. From this

data we identified two interesting candidates: a ligand for IL33R, IL33, which modulates migration

Chapter 6

148

of hematopoietic stem cells (HSC)4, 5 and IL31, which has been shown to enhance HSPC survival in

mouse bone marrow6. In future, these mass spectrometry experiments should be repeated, using

supernatant of CD14+ cells cultured for two to three days.

Alternatively, CD34+ cells could be cultured in fractionated conditioned medium and tested on CD34+

cell survival, which limits the potential candidates from mass spectrometry. The effect of potential can-

didates identified in this screen should then be tested by adding these factors to CD34+ cell cultures

and/or using neutralizing antibodies while co-culturing CD34+ cells with CD14+ cells. Identification of

these factors allows for proper use in culture protocols for various lineages and could give hints to

underlying mechanisms of CD14+ cells on HSPC to better understand HSPC survival in the hemato-

poietic niche. Several potential candidates for HSPC survival have been reported, such as IL3, IL6, IL7,

IL11, FLT3L, TPO7-9. However, it is unknown if these candidates are secreted by our cultured cells, as

we did not identify these proteins in our preliminary proteomic analysis of CD14+ cell supernatant.

Interestingly, recently published data on monocyte subsets report that the intermediate subtype

may be transitional monocytes that display the phagocytic features of the classical monocytes and

the inflammatory characteristics of the non-classical monocytes10. These cells give the support to

hematopoietic cells as observed in this thesis and secrete TNF-a, IL-1B, IL6, IL8, and IL10 according

to others11-13. Although identification of unique specific growth factors, interleukins and chemokines

is important, the mode of action on hematopoietic cells will probably be a combination of several

agents and may be dependent on the differentiation stage of HSPC, making assignments of specific

growth factors to the observed biological effects a complex matter.

Interestingly, we also showed in chapter 2 that there is residual erythroid potential in CD34- cells from

PBMC (population E, which is also Lin-), probably because PBMC contain CD34- HSC, as has previously

been shown in cord blood14. Within the CD34- population, particular cells have the capacity to develop

into CD34+ cells within two days of culture and will further develop into erythroblasts1. Bhatia et al.

already showed in 1998 that Lin-CD34- cells are capable of long-term repopulation in severe combined

immunodeficiency (SCID) mice to multiple lineages and are therefore thought to be the most imma-

ture HSC in humans, also called SCID-repopulating cells (SRC)15. It is believed that these CD34- HSC

that are negative for HLA-DR, FLT3 and KIT, but positive for CD133 and GPI-80 can be mobilized by

G-CSF treatment in a similar fashion to G-CSF mobilized CD34+ cells15-19, which is in agreement with our

unpublished data that showed the presence of CD34-CD133+ cells in PBMC. In order to further opti-

mize the erythroid cultures, it would be interesting to examine the role of CD14+ cells on CD34- HSC

and investigate if CD14+ cells could also affect CD34- cell survival or support differentiation towards

CD34+ HSC. Although CD34+ cell cultures are stroma-independent, human Lin-CD34- cord blood or

bone marrow-derived cells could only been maintained in bone marrow stroma-dependent cultures

(i.e. HESS-5 cell lines) in which the cells gained the ability to form colonies while expression of CD34

was acquired. This indicates that the microenvironment including macrophages may give cues that

affect CD34- HSC20-23. Preliminary data was obtained on erythroid cultures from PBMC, PBMC depleted

General discussion

149

for CD34 (CD34-) or PBMC depleted for both CD34 and CD14 (CD34-CD14-). Erythroblast outgrowth

after 13 days of culture from CD34- cells was extremely low compared to total PBMC (4.5 to 16-fold

reduction in erythroblasts from CD34- cells). In addition, after 13 days of culture, CD34-CD14- cells do

hardly form erythroblasts compared to CD34- cells of the same donor (4-fold reduction). This suggests

that CD14+ cells also positively influence the erythroid yield from CD34- cells or indirectly affect CD34+

cells that develop from the CD34- cell population, which should be further investigated by short-term

cultures of two days. In conclusion, the data in chapter 2 showed that CD14+ cells increase CD34+ cell

survival and thereby increase the erythroid yield, however, it is unclear if the erythroid yield could be

further increased by a supportive role of CD14+ cells on CD34- cells that are also present in PBMC.

Upon culturing, we noticed that purified monocytes cultured in erythroblast expansion medium differ-

entiate towards tissue resident cells expressing CD14, CD16, CD163, CD169, CD206, and CXCR4 within

two to three days. Although we recently exchanged StemSpan for a serum/plasma-free GMP-grade

medium for erythroid cultures in chapter 3, this did not affect the phenotype and function of these

macrophages. Co-expression of CD169 on cultured CD14+ cells suggests that these cells might have

a dual role within the hematopoietic stem cell niche, as the CD169+ macrophages described by Chow

et al. are important for HSPC retention in the bone marrow3. However, this paper was focused on HSC

migration, and therefore no data on HSC survival rates in the bone marrow after CD169+ macrophage

ablation was obtained and remains to be investigated. Together, these data suggest that after trans-

plantation, CD14+ cells may be beneficial for the survival and retention of CD34+ HSPC in the bone

marrow. Interfering with macrophage functionality in the bone marrow to increase HSC mobilization

may produce a negative side effect on the survival of progenitors in the hematopoietic niche. If HSPC

survival is regulated by macrophages in vivo, this adds additional functionality to the already known

retention signals from CD169+ macrophages in the bone marrow, which are dependent on cell-cell

contact through VCAM13. Recently, it has been shown that HSC retention in the spleen of mice with

stress-induced erythropoiesis also occurs via VCAM1 on macrophages24. In chapter 4 and chapter

5 we showed that bone marrow macrophages consists of a heterogeneous pool, while macrophages

in the fetal liver are homogeneous. Isolating these two macrophage populations to assay their surviv-

al-inducing potential on CD34+ cells will be instrumental to examine if these functions are performed

by the same or different macrophages. In addition, more research is needed to investigate if these

two effects (survival and maintenance) can be functionally separated, maybe through understanding

the transcriptional regulation of the secreted factors and the adhesion molecules that have been

shown to be important in erythroblastic island formation.

Ex vivo and in vitro erythroblastic island macrophages

In our culture system the synthetic glucocorticoid dexamethasone is present to inhibit erythroblast

differentiation. During systemic decreases in oxygen pressure due to stress or high altitudes, the

erythroid system has a fast response in order to increase erythroid output by increasing EPO and

glucocorticoid levels25-28. In chapter 4 we have shown that glucocorticoids have a clear effect on

Chapter 6

150

monocyte differentiation. Whether the addition of dexamethasone is also involved in the survival

of HSPC mediated by the soluble factor(s) secreted from the monocytes during the first days of

the culture still needs to be investigated. Phenotypically, these cells remain CD14+CD16+CD163+C-

D169+CD206+CXCR4+ macrophages suggesting that dexamethasone is primarily involved in the first

steps of monocyte differentiation towards these specific supportive macrophages (GC-macrophages).

Proteomic analysis showed that GC-macrophages have an anti-inflammatory phenotype and are

distinct from the cells that were cultured in the absence of dexamethasone, which showed a more

pro-inflammatory profile. Furthermore, GC-macrophages shared the majority of their proteome

with human bone marrow and fetal liver macrophages as shown in chapter 5. We hypothesized

that elevated glucocorticoid levels may increase the number of erythroblastic island macrophages

during stress-induced erythropoiesis as a result of directed differentiation of monocyte populations

to regulate the increased erythroid output. Indeed, we show in chapter 4 that GC-macrophages bind

cultured erythroblasts and reticulocytes, and phagocytose the expelled erythroid nuclei and thereby

share phenotypical and functional characteristics with central macrophages in erythroblastic islands.

Falchi et al. already reported in 2015 that human CD34-derived erythroid cultures containing gluco-

corticoids contained a very small population (1-3%) of macrophages29. These CD163+ macrophages

were highly motile and showed interactions with erythroblasts as we also observed in CD14-derived

macrophages in chapter 4. In addition, these CD34-derived macrophages shared morphological

features with GC-macrophages. Interestingly, although we culture a homogeneous population of

CD169+ macrophages, Falchi et al. identified a CD169- macrophage population that established tight

interactions with erythroid cells, while the CD169+ macrophages have multiple rapid loose interactions

and promote erythroblast proliferation. Because this study used CD34+ cells as starting material

instead of CD14+ cells and they cultured the cells with an additional supplement (IL-3), it is unclear if

these CD169+ macrophage populations are similar to GC-macrophages30, 31.

In mice, it has been shown that in steady-state hematopoiesis erythroid-macrophage support is

occurring via VCAM1, whereas Ulyanova et al. observed that Vcam1-/- mice do not display a compro-

mised erythroid stress response in the spleen and bone marrow, despite significantly lower absolute

levels of macrophages in the spleen32. This fits the notion that macrophages themselves may not

be necessary for the proliferation and differentiation during erythropoiesis upon optimal growth

factors (increased EPO and glucocorticoid levels during anemic situations), which we also observe in

our cultures. In addition, another group reported that Spi-c-/- mice had decreased VCAM1 expressing

macrophages, however, erythropoiesis was normal at steady-state conditions32. Interestingly, our data

in chapter 5 showed that VCAM1 was only identified in fetal liver and bone marrow macrophages

and not in GC-macrophages, although we showed in chapter 4 that very low levels of VCAM1 were

detected by flow cytometry on GC-macrophages. This suggests that erythroblast interactions may

also occur VCAM1-independent. Whether another interaction like EMP, which was only identified in

GC-macrophages, substitutes for VCAM1 would need to be determined. Although GC-macrophages

have increased erythroid binding capacity, bind more erythroblasts per macrophage, have more

General discussion

151

interaction with different stages of erythroblasts in particular reticulocytes, and bind and phagocytose

pyrenocytes compared to cells that were not cultured in the presence of dexamethasone, future

studies should include cultures in suboptimal conditions with reduced growth factors levels. Because

of non-limiting growth factors in vitro, erythroid expansion occurs efficiently and independently of

macrophages, as showed in this thesis and previous work1, 33, 34. In vivo erythropoiesis is suboptimal

and hence niche cells, together with systemic increases in growth factor concentrations (e.g., EPO and

glucocorticoids), can provide a regulatory axis to quickly increase the erythroid output in response

to stress. This may explain why co-cultures of erythroblasts with CD14+ cells in EPO, SCF and dexa-

methasone had no additional effect on erythroid expansion compared to erythroblast cultures in

the absence of CD14+ cells (chapter 2). We hypothesize that reducing EPO levels in the cultures may

positively affect erythroblast proliferation upon co-culture with GC-macrophages. However, GC-mac-

rophages may also produce EPO. Fetal liver macrophages in preterm infants are the primary source

of EPO35 and as we show in chapter 5, GC-macrophages share many characteristics with fetal liver

macrophages. Indeed, preliminary data showed that EPO was identified in CD14+ monocyte super-

natant after 24 hours of culture, however, no additional proteomics was performed on supernatant

from GC-macrophages. In addition, removing all supplements from the medium could change the

phenotype and function of GC-macrophages and results would thus be difficult to interpret. Mac-

rophages may still provide ill-defined nuances to erythropoiesis that are currently not implemented

during in vitro cultures.

In chapter 4 we showed that both macrophages and erythroblasts express several adhesion mole-

cules that are described to be involved in erythroblast-macrophage binding. Further research into the

last stage of erythroid maturation could identify which interactions are essential for erythroid support

and which molecular mechanisms are involved, by using neutralizing antibodies against the adhesion

molecules on GC-macrophages or their ligands on erythroid cells. Knowledge about erythroid-mac-

rophage interactions could then be used to further optimize erythrocyte cultures, e.g., by coating

culture dishes with specific integrins that support erythroid maturation or enucleation. Besides a role

for macrophages in erythroid binding, erythroblastic island macrophages are also known to bind and

phagocytose the expelled erythroid nuclei, the pyrenocyte. Enucleation normally takes place in the

adult bone marrow where central macrophages immediately phagocytose pyrenocytes via MERTK in

mice36. In the same chapter, we showed that GC-macrophages not only express the TAM-family recep-

tors, but also other phagocytosing-associated receptors, which suggests that GC-macrophages have

the correct machinery to recognize and actively phagocytose pyrenocytes. Interestingly, the data in

chapter 5 showed that MERTK was present in all GC-macrophages, but was not identified in erythro-

blastic island macrophages in the bone marrow and only found in some fetal liver macrophages (1 out

of 3 donors). In addition, TIMD4 and another TAM-family member AXL were restricted to erythroblastic

island macrophages in the fetal liver and bone marrow. Taken together, the exact mechanism(s)

through which macrophages recognize erythroid cells but phagocytose pyrenocytes are ill-defined in

human erythropoiesis and should be further studied in detail. We showed in chapter 4 that especially

Chapter 6

152

MERTK expression is reduced by glucocorticoid receptor inhibition through mifepristone treatment.

In order to investigate the role of TAM-receptors (in particular MERTK) in phagocytosis of pyrenocytes

by macrophages, co-cultures of erythroid cells and GC-macrophages could be supplemented with

mifepristone. Preliminary data already revealed that erythroid-macrophages clusters are reduced

upon mifepristone treatment. Mifepristone has also been used in the clinic for various purposes,

e.g., abortion pill or treatment for uterine myoma37, 38. It would be interesting to see if this treatment

affects the amount or phenotype of bone marrow macrophages and specifically within erythroblastic

islands. In addition, patients with MERTK mutations have been associated with phagocytic-dependent

diseases, such as retinitis pigmentosa and autoimmune disorders like systemic lupus erythematosus

(SLE)39-41. Patients with SLE had reduced MERTK expressing CD14+CD16+ cells and because CD163

expression is positively correlated with MERTK expression, these cells have reduced expression of

CD163. Furthermore, several groups reported increased serum levels of soluble MERTK and AXL in

lupus patients, which suggests that increased soluble TAM-receptors reflect decreased expression

of these receptors on the cell membrane41-43. Interestingly, pyrenocytes have never been found in

peripheral blood or bone marrow in human individuals. This strengthens the hypothesis that macro-

phages have multiple mechanisms to phagocytose these highly immunogenic pyrenocytes, as defects

in clearance of pyrenocytes in both mouse (by triple TAM-receptor KO) and human is supposedly

detrimental to the body.

In chapter 3 we discussed the possibility of using human induced pluripotent stem cells (hIPSC) or

immortalized cell lines like HUDEP, HIDEP and BEL-A, as starting material to produce in vitro eryth-

rocytes for transfusion. However, erythroid cells derived from this material display low enucleation

rates44-46. In contrast, erythroid cells derived from CD34+ cells or PBMC have >90% enucleation,

independently of macrophages. Nevertheless, the addition of bone marrow macrophages or GC-mac-

rophages to these cultures may still be beneficial for reticulocyte stability. Merryweather-Clarke et

al. reported that erythroid cells from hIPSC follow an embryonic or primitive program (e.g., high

expression of HBA, HBZ, ARID3A and low expression of SOX6, MYB, BCL11A), and thus macrophages

could indeed play a supporting role in enucleation47. Studies by Douays group showed that nucle-

ated hIPSC-derived erythroid cells undergo enucleation upon injection into mice48-50. However, these

experiments have never been reproduced and it is unknown where enucleation takes place after

transfusion. It is therefore crucial to know what drives erythroid cell enucleation and via which mech-

anisms pyrenocytes are cleared from the system. Interestingly, we noticed in chapter 4 that the

nuclei within erythroid cells all polarize towards the GC-macrophages. Understanding this polarity

may also be important to increase the enucleation rate of erythroid cultures from immortalized cell

lines. Erythroblast enucleation requires PI3K-dependent cell polarization, resulting in highly localized

phospholipids PtdIns(3,4)P2 and PtdIns(3,4,5)P3 at the cytoplasmic side of the plasma membrane.

However, this process occurs independent of macrophages51. Whether a polarized lipid composi-

tion contributes to the specific alignment of erythroid cells and whether this is further regulated by

macrophages remains to be determined. Moreover, nuclei polarization has been reported to be

General discussion

153

TRIM58 dependent52. TRIM58 is a E3 ubiquitin ligase superfamily member transcriptionally activated

by GATA1 prior to enucleation in mice. It has been linked to human erythrocyte traits53-55 and regulates

polarization and extrusion of mouse erythroblast nuclei in vitro via dynein and kinesins. However,

Thom et al. showed that TRIM58 does not affect erythroblast proliferation, survival, hemoglobiniza-

tion and expression of erythroid markers52. In addition, this group proposed that purification of CC1

protein could disengage aberrantly retained dynein from the nucleus of TRIM58 KO erythroblasts

and thereby may rescue enucleation (unpublished data by Thom et al.). More research is required

to investigate if TRIM58 or KIF5B also plays a role in human terminal differentiation and if the use

of GC-macrophages could support terminal differentiation. The work in this chapter provides an

excellent accessible human model system to understand how macrophages promote erythroid cells

during all stages of erythropoiesis.

In vivo, macrophages are important in regulating erythropoiesis both in adults and during develop-

ment, however, molecular markers that discriminate the different macrophage populations from

fetal liver and bone marrow are ill-defined, making isolation and further study difficult. In chapter 4

we performed a phenotypical comparison between GC-macrophages and CD163+ macrophages iso-

lated from bone marrow and fetal liver and in chapter 5 we compared the proteome of these three

erythroblastic island macrophage populations. For the first time, we show that erythroblastic island

macrophages from human fetal liver and bone marrow share a highly similar proteomic profile with

cultured monocyte-derived GC-macrophages. This raises the question if glucocorticoid treatment

not only increases erythroblast expansion in the bone marrow, but also increases the number of

erythroblastic island macrophages and thereby affect the erythroid yield. Indeed, Ulyanova et al.

showed that upon stress erythropoiesis in mice the number of splenic macrophages were increased.

However, there was no effect on macrophage numbers in the bone marrow32. In addition, it should

be investigated if glucocorticoids are required for central macrophage differentiation using gluco-

corticoid receptor (GR) knockout or Grdim/dim mice. For decades it was believed that all macrophages

originate from monocytes56. However, recent studies showed that most resident macrophages are

maintained independently of monocytes as lung, splenic red pulp, peritoneum, and bone marrow

tissue resident macrophages rely on self-renewal in steady-state conditions57. Contradictory results

have been shown by Landsman et al. who showed that in lungs adoptively transferred monocytes

contribute to and replace lung macrophages after depletion58, 59. In addition, Theurl et al. showed

an on-demand mechanism that allows for a fast but transient increase of Kupffer-like cells that are

derived from classical monocytes60. This opens up a whole new scenario through which specific

macrophages originate from different sources depending on the need of a specific tissue or process.

It also raises many questions about the role of fetal liver macrophages after birth, will they differen-

tiate towards Kupffer cells, do they undergo apoptosis, or do they migrate to the bone marrow and

maintain their function as central macrophages in erythroblastic islands? In addition, the transient

monocyte differentiation to supplement existing tissue resident pools of macrophages may also be

present in other tissues in response to stress, which could explain why GC-macrophages derived

Chapter 6

154

from monocytes could phenotypically and functionally resemble resident macrophages from the

bone marrow or fetal liver when mimicking stress erythropoiesis by the presence of glucocorticoids.

Glucocorticoid treatment of patients with autoimmune uveitis is associated with an enrichment of

CD14++CD16+ intermediate monocytes, however, endogenous glucocorticoid production may also

affect the treatment itself61. Specific red cell diseases have been described to be associated with

glucocorticoid signaling. Cushing’s syndrome is characterized by increased glucocorticoid levels and

erythrocytosis62, while Diamond-Blackfan anemia patients are treated with glucocorticoids to alleviate

their anemia63, 64. In addition, it has been shown that erythroblasts from polycythemia vera patients

that are characterized by erythrocytosis have a specific isoform of the glucocorticoid receptor (GR-β),

which is less responsive to glucocorticoids65. This suggests that glucocorticoids may also increase

erythrocyte numbers by increased erythroblastic islands and not only by erythroblast expansion.

In chapter 4 and chapter 5 we showed that CD163+ bone marrow macrophages are a heteroge-

neous population containing CD163dim and CD163high cells. Although purification by MACS isolation

enriched for CD163high macrophages, proteomic analysis data in chapter 5 showed that the correla-

tion between all bone marrow donors was lower compared to macrophages from the fetal liver or

cultured GC-macrophages. On the other hand, a single CD163+ macrophage population was observed

in fetal liver, and this population shared 83% of the identified proteins with CD163high macrophages

from the bone marrow. Previous studies reported heterogeneity in bone marrow macrophages66, 67

and our preliminary data indicate that CD163dim macrophages are more immature and do not express

tissue resident markers like VCAM1. Differences in marker expression of macrophages derived from

both organs observed in chapter 4 could thus be due to the heterogeneity in the bone marrow

population. Although the fetal liver is primarily performing erythropoiesis at week 9-24 of embryonic

development, it would be interesting to analyze fetal liver material <9 weeks in order to see when

CD163+ macrophages first arise during development and how this relates to erythroid development.

A study by Slayton et al. showed that already after 5 to 6 weeks after conception, human CD68+

macrophages are found in the fetal liver parenchyma and hepatic sinusoids68, while the first HSC

arise in the aorta-gonad-mesophrenos (AGM) region around week 4 post conception. This indicates

that fetal liver macrophages could be derived from HSC and/or from the yolk sac, even though flow

cytometry data in chapter 4 clearly showed a homogeneous macrophage population in the fetal liver

between week 15 and 22 of development. An important question that remains is how we should

characterize and discriminate between macrophages. The discrimination of M1 and M2 macrophages

is outdated as many studies showed that there are many macrophage subsets. Indeed, a proteomic

study by Brown et al. showed that the majority of the proteome between LPS- or IFNγ-induced M1

macrophages is similar, although the specific macrophage function is clearly different69. The majority

of proteins that were restricted to LPS- or IFNγ-induced M1 macrophages in this study were even

identified in our proteomics data of erythroblastic island macrophages in chapter 5, while CD163+

bone marrow, fetal liver and GC-macrophages all have a M2 profile. This makes the identification and

discrimination of macrophage subsets even more complex. It is therefore difficult to say what defines

General discussion

155

a macrophage subset. Macrophage polarization may be a gradual process, because macrophages

easily adapt to their microenvironment and change their functionality based on the need. Single cell

RNA-sequencing might discriminate between different origin of macrophages and may give new cues

on how to discriminate between the different subsets.

From in vitro red cell generation to a transfusion product

Protocols to culture red blood cells for transfusion purposes have been existing for a long time,

however, current attempts were still hampered by a low yield and incomplete differentiation to fully

biconcave erythrocytes70-78. The majority of the existing large-scale red blood cell cultures uses CD34+

cells derived from PBMC, cord blood or bone marrow79, while a minority uses PBMC as starting mate-

rial. It has been shown that the erythroid yield from PBMC is 10-15 increased compared to CD34+

cells1. Furthermore, some culture systems use glucocorticoids to induce erythroblasts to proliferate

whilst inhibiting differentiation, which has been shown over the last years in the group of Beug, von

Lindern, van den Akker and Migliaccio and others1, 33, 34, 80-82. In 2010, Migliaccio et al. showed massive

expansion of PBMC-derived erythroblasts using SCF, EPO, IL3 and dexamethasone or estradiol in

serum-free culture conditions81. However, only few large-scale cell studies use glucocorticoids with-

out the addition of serum and/or plasma, which directly differentiates erythroblasts and negates the

effect of glucocorticoids as shown in chapter 3. In 2002, the group of Douay reported the use of

cord blood-derived CD34+ for erythroid cultures using serum-free medium with BSA, ferric nitrate,

transferrin, insulin, lipids, hydrocortisone and additional supplements during culture (e.g., day 0-7:

FLT3, TPO, SCF; day 8-14: SCF, EPO, IGF1; day 15-22: EPO, IGF1). However, only 4% of these cells

enucleated after 17 days and they mainly expressed fetal hemoglobin48. The same group optimized

their protocol for CD34+ cells in 2005, by adding EPO and plasma during the entire culture, and IL3

during the first days of culture and removing FLT3 and TPO from the culture system. In addition, only

during the expansion phase hydrocortisone was used. Subsequently, cells were co-cultured with

stromal cells (e.g., mouse MS5 or human MSC) towards 90-100% enucleated reticulocytes without

the addition of additional cytokines, whereas only 2% enucleation was observed in the absence of

stromal cells83. Other studies also used stromal cell lines to culture red blood cells ex vivo84-86, however,

these methods are not feasible for routine red blood cell production. In 2011, the group of Douay

adapted their previously published protocol and reported that they could generate 68% enucleated

reticulocytes without stromal support. Interestingly, they were the first to report in-human blood

transfusion of around 14mL of cultured red blood cells (theoretically their cultures yield 3.6-27mL

of packed red cells per 10.000 CD34+ cells)87. However, the culture system described by this group is

also not feasible for large-scale red cell production, as it contains plasma and uses CD34+ cells after

G-CSF mobilization as starting material. A study by Griffiths et al. showed that 5mL of packed enucle-

ated red blood cells could be generated from PBMC-derived CD34+ cells albeit serum and/or plasma

was needed during these cultures88. Recently, the same group showed that under GMP conditions

but still including serum, CD34+ cells derived from both PBMC and cord blood could be cultured and

differentiated (>105 amplification) towards 60-63% enucleated reticulocytes, yielding around 10mL

Chapter 6

156

of packed red cells after filtering89. Another recent study by Zhang et al. optimized culture conditions

and developed a bottle turning device culture system for cord blood-derived CD34+ cell cultures to

generate red blood cells with a 50% enucleation rate90. In addition, these cultures also contain fetal

bovine serum. In contrast, Timmins et al. demonstrated in 2011 an ultra-high-yield of red blood cells

with >90% enucleation rate from cord blood-derived CD34+ cells in the absence of plasma and glu-

cocorticoids91. This group also used a 1L bioreactor to increase the erythroid yield.

In chapter 3 we described a three-stage culture system to generate around 25mL of packed red

cells from PBMC. We optimized existing erythroid cultures, by exchanging the commercial, propri-

ety-non-defined and expensive StemSpan for inexpensive completely characterized GMP-grade

medium. PBMC were used as starting material, because we showed in chapter 2 that CD14+ cells in

PBMC increase the erythroid yield by supporting CD34+ cell survival. In addition, starting from PBMC

dispenses the extremely expensive CD34 isolation scheme. Replacing plasma and serum for stripped

human serum albumin during the expansion phase maintained erythroblasts for a prolonged period

(>26 days), whilst inhibiting terminal differentiation. Removal of SCF and dexamethasone, increasing

the concentration of EPO and holotransferrin, and the addition of omniplasma and heparin differ-

entiates erythroblasts towards >90% fully enucleated cultured red blood cells that express adult

hemoglobin and maintain blood group expression. We are currently investigating if we can exchange

the addition of omniplasma during the last stage of erythroid differentiation for plasma-free supple-

ments by aiming to identify the components in plasma that allows efficient differentiation (e.g., using

PEG precipitation and plasma fractionation). As plasma not only contains proteins, but also lipids

and steroids, heat inactivation will determine if proteins in plasma are important for enucleation.

In addition, lipidomics on cultured red blood cells and freshly isolated red blood cells will give clues

to the differences in plasma membrane composition, which may need to be amended to further

resemble in vivo erythrocytes. Besides, routine red blood cell production in culture dishes will be

infeasible. Therefore, erythroblasts were expanded and differentiated in G-Rex bioreactors in order

to meet the required amount of red cells for one unit of packed red blood cells. A draw-back of the

currently used bioreactor remains in scalability, as these bioreactors are not provided with a capacity

over 5L due to physical constraints and are very costly if multiple G-Rex culture systems were to be

employed. Therefore, we currently focus on the design of a novel bioreactor.

Another disadvantage of our culture method described in chapter 3 is that people still have to donate

blood in order to start red cell cultures, which is in particularly invasive when anemic patient-specific

blood has to be used. Therefore, it would be more convenient to start with an immortal source, such

as hIPSC or immortalized cell lines. Recently, a human immortalized adult erythroid cell line (Bristol

Erythroid Line Adult, or BEL-A) was generated by transducing CD34+ cells from adult bone marrow

with a Tet-inducible HPV16-E6/E7 construct46. These cells have similar progression to the erythroid

pathway and erythroblast expansion compared to CD34+ cells from peripheral blood. However, the

enucleation rates are still around 30%. Using BEL-A cell lines for large-scale erythroid cultures requires

General discussion

157

further optimization of the enucleation and reticulocyte stability. Data from others and unpublished

data from our group showed that hIPSC can be generated from human cultured erythroblasts using

the Yamanaka lentiviral factor set Oct4, Sox2, c-Myc, and Klf4 or episomal reprogramming92, 93. By

generating embryoid bodies or using the more efficient single cell cultures, pluripotent cells can be

further directed towards erythroid cells94, 95 (and unpublished data by Varga et al.). However, these hIP-

SC-derived erythroid cells still express fetal hemoglobin. Transfusions with erythroid cells expressing

fetal hemoglobin would not immediately be detrimental, as patients with persistent fetal hemoglobin

do not show any phenotype96. Interestingly, patients with sickle cell disease and thalassemia could

even benefit from transfusions with high fetal hemoglobin expression in the erythroid cells, as this

could immediately reduce the symptoms caused by mutated adult hemoglobin. Recently, the group

of Maryeux performed proteomic analysis on cultured human erythroid cells97. Additional RNA-se-

quencing of erythroid cell cultures during differentiation will further add as a control to flow cytometry

and functional assays, but was not ready for analysis upon writing this thesis. However as indicated,

hIPSC-derived erythroid cells have poor enucleation rates, which will be the focus of future research.

Knowledge of erythroid-macrophage interactions may provide an approach to study this, e.g., by

co-cultures of hIPSC with GC-macrophages. In addition, reticulocytes that enter the bloodstream

will further differentiate towards erythrocytes in vivo, therefore, we possibly do not need biconcave

erythrocytes as a transfusion product. However, this can only be properly tested in a clinical trial and

therefore still allows research to focus on the differentiation of reticulocytes towards erythrocytes.

It is still very expensive to produce an in vitro transfusion product and the costs to produce 25mL

of cRBC, though difficult to assess at the moment, would be approximately 100.000 euro per prod-

uct. Major cost drivers are holotransferrin, SCF and EPO. However, these costs may be significantly

decreased if in-house developed recombinant versions are produced, which might reduce the costs

to approximately 5.500 euro per product. This is in the range of a unit rare blood from the frozen

blood bank of 3.676 euro. The synchronized cultures all contain young reticulocytes that have a the-

oretical lifespan of about 120 days. Chronic anemia patients that receive blood transfusions every

two months may benefit from transfusion with in vitro cultured long-lived erythrocytes, potentially

increasing the time between transfusions and thereby reducing the costs. A clinical trial will allow to

test the in vivo lifespan of transfused PBMC-derived reticulocytes and investigate the optimal period

of repeated transfusions, because transfused patients could experience a rapid drop in hemoglobin

when erythrocytes die around the same period.

The data in this thesis provide evidence that macrophages are crucial during human erythropoiesis

and positively influence erythroid cell cultures. We have provided new tools to study the role of mac-

rophages during erythropoiesis in depth, by tracking hematopoietic stem cell differentiation towards

erythropoiesis showing that macrophages support the survival of HSPC. Furthermore, these macro-

phages functionally and phenotypically resemble erythroblastic island macrophages from fetal liver

and bone marrow and can be used as a human model to study erythroid-macrophage interactions.

Chapter 6

158

Altogether, we have unraveled important mechanisms of erythroid-supporting macrophages, which

can be applied to current protocols to optimize in vitro production of red blood cells. We are currently

ready to start phase I clinical trials with 25mL of autologous blood from healthy individuals using

PBMC as starting material. However, in future, we expect to generate universal donor red blood cells

derived from hIPSC or other immortalized cell lines. CRISPR/Cas9 may be used to remove immuno-

genic features like blood groups that are not necessary for the functionality of the cells, such as Kel,

Kidd and Duffy. Both patients in need of regular blood transfusions as well as patients that need a

single blood transfusion after blood loss may then be transfused with in vitro cultured red blood cells.

General discussion

159

References

1. van den Akker E, Satchwell TJ, Pellegrin S, Daniels G, Toye AM. The majority of the in vitro erythroid expansion potential resides

in CD34(-) cells, outweighing the contribution of CD34(+) cells and significantly increasing the erythroblast yield from peripheral

blood samples. Haematologica. 2010;95(9):1594-1598.

2. Ziegler-Heitbrock L, Ancuta P, Crowe S, et al. Nomenclature of monocytes and dendritic cells in blood. Blood. 2010;116(16):e74-

80.

3. Chow A, Lucas D, Hidalgo A, et al. Bone marrow CD169+ macrophages promote the retention of hematopoietic stem and

progenitor cells in the mesenchymal stem cell niche. J Exp Med. 2011;208(2):261-271.

4. Yagami A, Orihara K, Morita H, et al. IL-33 mediates inflammatory responses in human lung tissue cells. J Immunol.

2010;185(10):5743-5750.

5. Kim J, Kim W, Le HT, et al. IL-33-induced hematopoietic stem and progenitor cell mobilization depends upon CCR2. J Immunol.

2014;193(7):3792-3802.

6. Broxmeyer HE, Li J, Hangoc G, et al. Regulation of myeloid progenitor cell proliferation/survival by IL-31 receptor and IL-31. Exp

Hematol. 2007;35(4 Suppl 1):78-86.

7. Buza-Vidas N, Antonchuk J, Qian H, et al. Cytokines regulate postnatal hematopoietic stem cell expansion: opposing roles of

thrombopoietin and LNK. Genes Dev. 2006;20(15):2018-2023.

8. Nandurkar HH, Robb L, Tarlinton D, Barnett L, Kontgen F, Begley CG. Adult mice with targeted mutation of the interleukin-11

receptor (IL11Ra) display normal hematopoiesis. Blood. 1997;90(6):2148-2159.

9. Testa U, Fossati C, Samoggia P, et al. Expression of growth factor receptors in unilineage differentiation culture of purified

hematopoietic progenitors. Blood. 1996;88(9):3391-3406.

10. Mukherjee R, Kanti Barman P, Kumar Thatoi P, Tripathy R, Kumar Das B, Ravindran B. Non-Classical monocytes display inflam-

matory features: Validation in Sepsis and Systemic Lupus Erythematous. Sci Rep. 2015;5:13886.

11. Cros J, Cagnard N, Woollard K, et al. Human CD14dim monocytes patrol and sense nucleic acids and viruses via TLR7 and TLR8

receptors. Immunity. 2010;33(3):375-386.

12. Ward JR, West PW, Ariaans MP, et al. Temporal interleukin-1beta secretion from primary human peripheral blood monocytes

by P2X7-independent and P2X7-dependent mechanisms. J Biol Chem. 2010;285(30):23147-23158.

13. Wong KL, Tai JJ, Wong WC, et al. Gene expression profiling reveals the defining features of the classical, intermediate, and non-

classical human monocyte subsets. Blood. 2011;118(5):e16-31.

14. Matsuoka Y, Sumide K, Kawamura H, et al. Human cord blood-derived primitive CD34-negative hematopoietic stem cells (HSCs)

are myeloid-biased long-term repopulating HSCs. Blood Cancer J. 2015;5:e290.

15. Bhatia M, Bonnet D, Murdoch B, Gan OI, Dick JE. A newly discovered class of human hematopoietic cells with SCID-repopulating

activity. Nat Med. 1998;4(9):1038-1045.

16. Sonoda Y. Immunophenotype and functional characteristics of human primitive CD34-negative hematopoietic stem cells: the

significance of the intra-bone marrow injection. J Autoimmun. 2008;30(3):136-144.

17. Takahashi M, Matsuoka Y, Sumide K, et al. CD133 is a positive marker for a distinct class of primitive human cord blood-derived

CD34-negative hematopoietic stem cells. Leukemia. 2014;28(6):1308-1315.

18. Matsuoka Y, Sumide K, Kawamura H, Nakatsuka R, Fujioka T, Sonoda Y. GPI-80 expression highly purifies human cord blood-de-

rived primitive CD34-negative hematopoietic stem cells. Blood. 2016;128(18):2258-2560.

19. Matsuoka Y, Takahashi M, Sumide K, et al. CD34 Antigen and the MPL Receptor Expression Defines a Novel Class of Human

Cord Blood-Derived Primitive Hematopoietic Stem Cells. Cell Transplant. 2017;26(6):1043-1058.

20. Goodell MA, Brose K, Paradis G, Conner AS, Mulligan RC. Isolation and functional properties of murine hematopoietic stem cells

that are replicating in vivo. J Exp Med. 1996;183(4):1797-1806.

21. Fujisaki T, Berger MG, Rose-John S, Eaves CJ. Rapid differentiation of a rare subset of adult human lin(-)CD34(-)CD38(-) cells

stimulated by multiple growth factors in vitro. Blood. 1999;94(6):1926-1932.

22. Nakamura Y, Ando K, Chargui J, et al. Ex vivo generation of CD34(+) cells from CD34(-) hematopoietic cells. Blood. 1999;94(12):4053-

4059.

23. Nakamura Y, Yahata T, Muguruma Y, et al. Angiopoietin-1 supports induction of hematopoietic activity in human CD34- bone

marrow cells. Exp Hematol. 2007;35(12):1872-1883.

24. Dutta P, Hoyer FF, Grigoryeva LS, et al. Macrophages retain hematopoietic stem cells in the spleen via VCAM-1. J Exp Med.

2015;212(4):497-512.

25. Humpeler E, Skrabal F, Bartsch G. Influence of exposure to moderate altitude on the plasma concentraton of cortisol, aldoste-

rone, renin, testosterone, and gonadotropins. Eur J Appl Physiol Occup Physiol. 1980;45(2-3):167-176.

Chapter 6

160

26. Bauer A, Tronche F, Wessely O, et al. The glucocorticoid receptor is required for stress erythropoiesis. Genes Dev.

1999;13(22):2996-3002.

27. Kodama T, Shimizu N, Yoshikawa N, et al. Role of the glucocorticoid receptor for regulation of hypoxia-dependent gene expres-

sion. J Biol Chem. 2003;278(35):33384-33391.

28. Jelkmann W. Physiology and pharmacology of erythropoietin. Transfus Med Hemother. 2013;40(5):302-309.

29. Falchi M, Varricchio L, Martelli F, et al. Dexamethasone targeted directly to macrophages induces macrophage niches that

promote erythroid expansion. Haematologica. 2015;100(2):178-187.

30. Ebner S, Hofer S, Nguyen VA, et al. A novel role for IL-3: human monocytes cultured in the presence of IL-3 and IL-4 differ-

entiate into dendritic cells that produce less IL-12 and shift Th cell responses toward a Th2 cytokine pattern. J Immunol.

2002;168(12):6199-6207.

31. Suzuki H, Katayama N, Ikuta Y, et al. Activities of granulocyte-macrophage colony-stimulating factor and interleukin-3 on mono-

cytes. Am J Hematol. 2004;75(4):179-189.

32. Ulyanova T, Phelps SR, Papayannopoulou T. The macrophage contribution to stress erythropoiesis: when less is enough. Blood.

2016;128(13):1756-1765.

33. Leberbauer C, Boulme F, Unfried G, Huber J, Beug H, Mullner EW. Different steroids co-regulate long-term expansion versus

terminal differentiation in primary human erythroid progenitors. Blood. 2005;105(1):85-94.

34. von Lindern M, Deiner EM, Dolznig H, et al. Leukemic transformation of normal murine erythroid progenitors: v- and c-ErbB act

through signaling pathways activated by the EpoR and c-Kit in stress erythropoiesis. Oncogene. 2001;20(28):3651-3664.

35. Ohls RK, Li Y, Trautman MS, Christensen RD. Erythropoietin production by macrophages from preterm infants: implications

regarding the cause of the anemia of prematurity. Pediatr Res. 1994;35(2):169-170.

36. Toda S, Segawa K, Nagata S. MerTK-mediated engulfment of pyrenocytes by central macrophages in erythroblastic islands.

Blood. 2014;123(25):3963-3971.

37. Creinin M, Blumenthal P, Shulman L. Mortality associated with mifepristone-misoprostol medical abortion. MedGenMed.

2006;8(2):26.

38. Seth S, Goel N, Singh E, Mathur AS, Gupta G. Effect of mifepristone (25 mg) in treatment of uterine myoma in perimenopausal

woman. J Midlife Health. 2013;4(1):22-26.

39. Al-Khersan H, Shah KP, Jung SC, Rodriguez A, Madduri RK, Grassi MA. A novel MERTK mutation causing retinitis pigmentosa.

Graefes Arch Clin Exp Ophthalmol. 2017;255(8):1613-1619.

40. Cheong HS, Lee SO, Choi CB, Sung YK, Shin HD, Bae SC. MERTK polymorphisms associated with risk of haematological disorders

among Korean SLE patients. Rheumatology (Oxford). 2007;46(2):209-214.

41. Zhu H, Sun X, Zhu L, et al. The expression and clinical significance of different forms of Mer receptor tyrosine kinase in systemic

lupus erythematosus. J Immunol Res. 2014;2014:431896.

42. Zizzo G, Guerrieri J, Dittman LM, Merrill JT, Cohen PL. Circulating levels of soluble MER in lupus reflect M2c activation of mono-

cytes/macrophages, autoantibody specificities and disease activity. Arthritis Res Ther. 2013;15(6):R212.

43. Ballantine L, Midgley A, Harris D, Richards E, Burgess S, Beresford MW. Increased soluble phagocytic receptors sMer, sTyro3

and sAxl and reduced phagocytosis in juvenile-onset systemic lupus erythematosus. Pediatr Rheumatol Online J. 2015;13:10.

44. Lapillonne H, Kobari L, Mazurier C, et al. Red blood cell generation from human induced pluripotent stem cells: perspectives

for transfusion medicine. Haematologica. 2010;95(10):1651-1659.

45. Kurita R, Suda N, Sudo K, et al. Establishment of immortalized human erythroid progenitor cell lines able to produce enucleated

red blood cells. PLoS One. 2013;8(3):e59890.

46. Trakarnsanga K, Griffiths RE, Wilson MC, et al. An immortalized adult human erythroid line facilitates sustainable and scalable

generation of functional red cells. Nat Commun. 2017;8:14750.

47. Merryweather-Clarke AT, Tipping AJ, Lamikanra AA, et al. Distinct gene expression program dynamics during erythropoiesis from

human induced pluripotent stem cells compared with adult and cord blood progenitors. BMC Genomics. 2016;17(1):817.

48. Neildez-Nguyen TM, Wajcman H, Marden MC, et al. Human erythroid cells produced ex vivo at large scale differentiate into red

blood cells in vivo. Nat Biotechnol. 2002;20(5):467-472.

49. Mazurier C, Douay L, Lapillonne H. Red blood cells from induced pluripotent stem cells: hurdles and developments. Curr Opin

Hematol. 2011;18(4):249-253.

50. Kobari L, Yates F, Oudrhiri N, et al. Human induced pluripotent stem cells can reach complete terminal maturation: in vivo and

in vitro evidence in the erythropoietic differentiation model. Haematologica. 2012;97(12):1795-1803.

51. Wang J, Ramirez T, Ji P, Jayapal SR, Lodish HF, Murata-Hori M. Mammalian erythroblast enucleation requires PI3K-dependent

cell polarization. J Cell Sci. 2012;125(Pt 2):340-349.

52. Thom CS, Traxler EA, Khandros E, et al. Trim58 degrades Dynein and regulates terminal erythropoiesis. Dev Cell. 2014;30(6):688-

700.

General discussion

161

53. van der Harst P, Zhang W, Mateo Leach I, et al. Seventy-five genetic loci influencing the human red blood cell. Nature.

2012;492(7429):369-375.

54. Kamatani Y, Matsuda K, Okada Y, et al. Genome-wide association study of hematological and biochemical traits in a Japanese

population. Nat Genet. 2010;42(3):210-215.

55. Gieger C, Radhakrishnan A, Cvejic A, et al. New gene functions in megakaryopoiesis and platelet formation. Nature.

2011;480(7376):201-208.

56. van Furth R, Cohn ZA. The origin and kinetics of mononuclear phagocytes. J Exp Med. 1968;128(3):415-435.

57. Hashimoto D, Chow A, Noizat C, et al. Tissue-resident macrophages self-maintain locally throughout adult life with minimal

contribution from circulating monocytes. Immunity. 2013;38(4):792-804.

58. Landsman L, Varol C, Jung S. Distinct differentiation potential of blood monocyte subsets in the lung. J Immunol. 2007;178(4):2000-

2007.

59. Landsman L, Jung S. Lung macrophages serve as obligatory intermediate between blood monocytes and alveolar macrophages.

J Immunol. 2007;179(6):3488-3494.

60. Theurl I, Hilgendorf I, Nairz M, et al. On-demand erythrocyte disposal and iron recycling requires transient macrophages in the

liver. Nat Med. 2016;22(8):945-951.

61. Liu B, Dhanda A, Hirani S, et al. CD14++CD16+ Monocytes Are Enriched by Glucocorticoid Treatment and Are Functionally

Attenuated in Driving Effector T Cell Responses. J Immunol. 2015;194(11):5150-5160.

62. Miljic P, Miljic D, Cain JW, Korbonits M, Popovic V. Pathogenesis of vascular complications in Cushing’s syndrome. Hormones

(Athens). 2012;11(1):21-30.

63. Gasser C. [Aplastic anemia (chronic erythroblastophthisis) and cortisone]. Schweiz Med Wochenschr. 1951;81(50):1241-1242.

64. Vlachos A, Ball S, Dahl N, et al. Diagnosing and treating Diamond Blackfan anaemia: results of an international clinical consensus

conference. Br J Haematol. 2008;142(6):859-876.

65. Varricchio L, Masselli E, Alfani E, et al. The dominant negative beta isoform of the glucocorticoid receptor is uniquely expressed

in erythroid cells expanded from polycythemia vera patients. Blood. 2011;118(2):425-436.

66. Chow A, Huggins M, Ahmed J, et al. CD169(+) macrophages provide a niche promoting erythropoiesis under homeostasis and

stress. Nat Med. 2013;19(4):429-436.

67. Gordon S, Pluddemann A, Martinez Estrada F. Macrophage heterogeneity in tissues: phenotypic diversity and functions. Immunol

Rev. 2014;262(1):36-55.

68. Slayton WB, Juul SE, Calhoun DA, Li Y, Braylan RC, Christensen RD. Hematopoiesis in the liver and marrow of human fetuses at 5

to 16 weeks postconception: quantitative assessment of macrophage and neutrophil populations. Pediatr Res. 1998;43(6):774-

782.

69. Brown J, Wallet MA, Krastins B, Sarracino D, Goodenow MM. Proteome bioprofiles distinguish between M1 priming and activation

states in human macrophages. J Leukoc Biol. 2010;87(4):655-662.

70. Migliaccio AR, Whitsett C, Migliaccio G. Erythroid cells in vitro: from developmental biology to blood transfusion products. Curr

Opin Hematol. 2009;16(4):259-268.

71. Migliaccio AR, Masselli E, Varricchio L, Whitsett C. Ex-vivo expansion of red blood cells: how real for transfusion in humans? Blood

Rev. 2012;26(2):81-95.

72. Migliaccio AR, Grazzini G, Hillyer CD. Ex vivo generated red cells as transfusion products. Stem Cells Int. 2012;2012:615412.

73. Migliaccio AR, Palis J. Blood in a dish: In vitro synthesis of red blood cells. Drug Discov Today Dis Mech. 2011;8(1-2):e3-e8.

74. Shah S, Huang X, Cheng L. Concise review: stem cell-based approaches to red blood cell production for transfusion. Stem Cells

Transl Med. 2014;3(3):346-355.

75. Kim HO. In-vitro stem cell derived red blood cells for transfusion: are we there yet? Yonsei Med J. 2014;55(2):304-309.

76. Larochelle A. Generation of red blood cells in vitro: monitoring the process for improved efficiency. Cytotherapy. 2013;15(9):1043-

1045.

77. Migliaccio AR, Whitsett C, Papayannopoulou T, Sadelain M. The potential of stem cells as an in vitro source of red blood cells for

transfusion. Cell Stem Cell. 2012;10(2):115-119.

78. Singh VK, Saini A, Tsuji K, Sharma PB, Chandra R. Manufacturing blood ex vivo: a futuristic approach to deal with the supply and

safety concerns. Front Cell Dev Biol. 2014;2:26.

79. Stella CC, Cazzola M, De Fabritiis P, et al. CD34-positive cells: biology and clinical relevance. Haematologica. 1995;80(4):367-387.

80. Carotta S, Pilat S, Mairhofer A, et al. Directed differentiation and mass cultivation of pure erythroid progenitors from mouse

embryonic stem cells. Blood. 2004;104(6):1873-1880.

81. Migliaccio G, Sanchez M, Masiello F, et al. Humanized culture medium for clinical expansion of human erythroblasts. Cell Trans-

plant. 2010;19(4):453-469.

82. Migliaccio G, Di Pietro R, di Giacomo V, et al. In vitro mass production of human erythroid cells from the blood of normal donors

and of thalassemic patients. Blood Cells Mol Dis. 2002;28(2):169-180.

Chapter 6

162

83. Giarratana MC, Kobari L, Lapillonne H, et al. Ex vivo generation of fully mature human red blood cells from hematopoietic stem

cells. Nat Biotechnol. 2005;23(1):69-74.

84. Baek EJ, Kim HS, Kim S, Jin H, Choi TY, Kim HO. In vitro clinical-grade generation of red blood cells from human umbilical cord

blood CD34+ cells. Transfusion. 2008;48(10):2235-2245.

85. Fujimi A, Matsunaga T, Kobune M, et al. Ex vivo large-scale generation of human red blood cells from cord blood CD34+ cells by

co-culturing with macrophages. Int J Hematol. 2008;87(4):339-350.

86. Xi J, Li Y, Wang R, et al. In vitro large scale production of human mature red blood cells from hematopoietic stem cells by cocul-

turing with human fetal liver stromal cells. Biomed Res Int. 2013;2013:807863.

87. Giarratana MC, Rouard H, Dumont A, et al. Proof of principle for transfusion of in vitro-generated red blood cells. Blood.

2011;118(19):5071-5079.

88. Griffiths RE, Kupzig S, Cogan N, et al. Maturing reticulocytes internalize plasma membrane in glycophorin A-containing vesicles

that fuse with autophagosomes before exocytosis. Blood. 2012;119(26):6296-6306.

89. Kupzig S, Parsons SF, Curnow E, Anstee DJ, Blair A. Superior survival of ex vivo cultured human reticulocytes following transfusion

into mice. Haematologica. 2017;102(3):476-483.

90. Zhang Y, Wang C, Wang L, et al. Large-Scale Ex Vivo Generation of Human Red Blood Cells from Cord Blood CD34(+) Cells. Stem

Cells Transl Med. 2017;6(8):1698-1709.

91. Timmins NE, Athanasas S, Gunther M, Buntine P, Nielsen LK. Ultra-high-yield manufacture of red blood cells from hematopoietic

stem cells. Tissue Eng Part C Methods. 2011;17(11):1131-1137.

92. Takahashi K, Yamanaka S. Induction of pluripotent stem cells from mouse embryonic and adult fibroblast cultures by defined

factors. Cell. 2006;126(4):663-676.

93. Varga E, Hansen M, Wust T, von Lindern M, van den Akker E. Generation of human erythroblast-derived iPSC line using episomal

reprogramming system. Stem Cell Res. 2017;25:30-33.

94. Yang CT, French A, Goh PA, et al. Human induced pluripotent stem cell derived erythroblasts can undergo definitive erythropoi-

esis and co-express gamma and beta globins. Br J Haematol. 2014;166(3):435-448.

95. Mao B, Huang S, Lu X, et al. Early Development of Definitive Erythroblasts from Human Pluripotent Stem Cells Defined by

Expression of Glycophorin A/CD235a, CD34, and CD36. Stem Cell Reports. 2016;7(5):869-883.

96. Borg J, Papadopoulos P, Georgitsi M, et al. Haploinsufficiency for the erythroid transcription factor KLF1 causes hereditary

persistence of fetal hemoglobin. Nat Genet. 2010;42(9):801-805.

97. Gautier EF, Ducamp S, Leduc M, et al. Comprehensive Proteomic Analysis of Human Erythropoiesis. Cell Rep. 2016;16(5):1470-

1484.

A

SummarySamenvatting

List of publicationsPhD portfolio

DankwoordCurriculum Vitae

APPENDIX

Summary

167

Summary

Despite the many donors who donate blood, there are patients for which it is difficult, if not impossible, to find compatible (interchangeable) blood. This mostly applies to patients with sickle cell disease that have been immunized against different non-self antigens (so-called alloantigens) by previous blood transfusions. The option to culture red blood cells to provide matched blood for these patients has been recognized for decades. Due to technical advances, research concerning the optimization of red blood cell cultures to reach quantities that resemble a unit of red blood cells given for transfusion have gained momentum over the last few years. These studies primarily investigate the intrinsic maturation program of red blood cells. However, red blood cell development is occurring on specialized support cells, the macrophages. Within this thesis, we aimed to optimize cell cultures from stem cell to red blood cell by focusing on the supporting macrophages that are known to have an essential but ill-defined role in the production of red blood cells.

Red blood cells or erythrocytes originate from the bone marrow. Within the bone marrow, a small fraction of blood stem cells (hematopoietic stem cells) is present which, in the right environment with a proper balance of growth factors, can develop into all blood cells in the body. The environment plays an important role as the cells located near the hematopoietic stem cells in the bone marrow support the stem cells to further maturate into one of the many blood cell lineages, a process called hematopoiesis. The development of a hematopoietic stem cell to a red blood cell is termed erythropoiesis. One of the cells located near the stem cells in the bone marrow is the macrophage (Greek makros = large, phagein = to eat), a certain type of white blood cell classically functioning to engulf old, dead or damaged cells and to clear pathogens. Multiple variants of macrophages have been found in the body and within a tissue fulfilling specific functions. In the bone marrow, macrophages have a supportive function in addition to their traditional clearing function and support precursor red blood cells, the erythroblasts. Such a macrophage is surrounded by erythroid cells and forms an islet structure, an erythroblastic island. These macrophages support erythroblast expansion (proliferation) and develop (differentiate) towards a red blood cell through mostly unknown or ill-defined mechanisms. There are several differentiation stages between erythroblasts and red blood cells. In the final stage, erythroid cells expel their nucleus (enucleation) resulting in enucleated reticulocytes and the individual nuclei surrounded by a plasma membrane, the pyrenocytes. The reticulocyte develops further into a red blood cell the moment it enters the blood stream. Macrophages also play an important role during this last stage, as they engulf and clear the expelled erythroid nuclei.

As we need hematopoietic stem cells for red blood cell production, and bone marrow material is difficult to obtain, peripheral blood has been used as starting material. Blood contains a very small fraction (on average 0.16%) of hematopoietic stem cells that have been mobilized from the bone marrow. This fraction can be isolated by density separation, which separates blood in three layers: 1) red blood cells and granulocytes, 2) white blood cells, and 3) plasma and platelets. The second layer, termed peripheral blood mononuclear cells, or PBMC, also contains the hematopoietic stem cells. Previously, it has been shown that the erythroblast yield from PBMC is 10-15 fold increased

Appendix

168

compared to purified hematopoietic stem cells isolated from the same amount of PBMC. A possible explanation would be that particular cells present in PBMC support the stem cells and thereby indirectly affect the amount of erythroblasts. We investigated this in chapter 2 and showed that several white blood cells are present in PBMC, in particular T-cells (58%), B-cells (6%), and monocytes and macrophages (18%). All blood cells can be distinguished by antigens (so-called CD antigens) present at the surface of the cells. First, we developed a novel way to characterize the stages between hematopoietic stem cells (CD34+CD36-) and erythroblasts (CD34-CD36+) in order to easily track cells during differentiation. Thereafter, T-cells (CD3), B-cells (CD19) and monocytes (CD14) were removed from PBMC one by one and the remaining cells were further cultured towards erythroblasts. The amount of erythroblasts was significantly reduced in PBMC depleted for monocytes/macrophages. In addition, we showed that co-cultures of purified hematopoietic stem cells with monocytes/macrophages result in a similar erythroblast yield compared to total PBMC. The results show that monocytes/macrophages do not have an effect on the erythroblasts themselves, but increase hematopoietic stem cell survival by the secretion of specific factors early during culture. This eventually results in an increased erythroblast yield.

Chapter 3 shows a culture method to efficiently culture red blood cells from PBMC. This culture system consists of three stages: i) the culture of PBMC towards erythroblasts, ii) expanding erythroblasts, and iii) erythroblast to reticulocyte differentiation. To mimic the correct environment, a specific cocktail of growth factors was added to the medium at each stage of culture. With the addition of growth factors cells survive, proliferate, and differentiate. In the first stage, stem cell factor (SCF) stimulates the proliferation of hematopoietic stem cells. The addition of erythropoietin (better known as EPO), which is normally produced by the kidney, supports the formation of red blood cells. In order to prevent cells already from differentiating to red blood cells, dexamethasone was added. This is a glucocorticosteroid, an adrenal gland hormone with an anti-inflammatory function, which is increased in the body during stress erythropoiesis when many erythroid cells have to be produced in the bone marrow. In the second stage, this specific cocktail is also used to increase the number of erythroblasts without differentiation. When a sufficient amount of erythroblasts is obtained, both SCF and dexamethasone are removed from the culture medium to allow the cells to further differentiate and to expel their nuclei. Furthermore, the EPO concentration is increased, and extra holotransferrin, an iron-binding and -transporting protein, is added. Iron-ion bound hemoglobin is an essential part of red blood cells, because it supports oxygen transport to different body parts. Finally, plasma is added to stimulate cell differentiation together with an anti-coagulant heparin to prevent the medium from clotting. In order to meet the required numbers of red blood cells for one transfusion, an infeasible amount of culture dishes is needed. We therefore examined alternative culture systems. Stage one was performed in culture dishes, however, in the second stage the culture dishes were exchanged for a bioreactor with optimal gas exchange due to a gas permeable bottom (G-Rex system). With this system, we can increase the cell culture density, in addition, larger culture volumes can be used and replenishing the medium is more efficient. This method allows to produce a yield of erythroid cells that is sufficient to start a clinical trial to test the lifespan of transfused cultured red blood cells and investigate the optimal period of repeated transfusions with cultured blood.

Summary

169

Macrophages play an important role in erythropoiesis. However, the underlying mechanism is not yet known, because there is no convenient human model to study this. Therefore, in chapter 4, we investigated whether we can differentiate PBMC-derived monocytes towards macrophages that we can use as a model to study the interaction between macrophages and different stages of red cells. As we used dexamethasone in the first and second stage of our culture system, we first examined the effect of this synthetic hormone on monocyte differentiation. In chapter 2, we showed that monocytes cultured in the presence of EPO, SCF, and dexamethasone differentiate towards tissue resident macrophages. Here, we show that glucocorticosteroids specifically direct all monocytes towards macrophages that phenotypically resemble bone marrow macrophages present in erythroblastic islands, expressing CD14, CD16, CD163, CD169, CD206 and CXCR4. Since these macrophages have been cultured with glucocorticosteroids, we refer to these cells as GC-macrophages. We actually show that these macrophages are a different type of macrophage compared to cells cultured without dexamethasone. Furthermore, GC-macrophages share functional characteristics with macrophages from erythroblastic islands: i) they bind erythroblasts and other stages of erythroid cells, and ii) they take up the erythroid nuclei when they are expelled by the erythroblasts, and iii) they are highly motile. Based on these results, we propose that GC-macrophages can be used as a model to study erythroblastic islands. The chapter concludes with antigen characterization of bone marrow macrophages, as well as macrophages from the fetal liver, because red cells are produced in the liver until the 7th month of fetal development instead of the bone marrow.

As a follow-up on this chapter, in chapter 5 we extensively compared CD163 fetal liver macrophages, CD163 bone marrow macrophages, and cultured CD163 GC-macrophages. The cell morphology between the three populations has a lot in common, also because they all bind erythroblasts and reticulocytes and take up erythroid nuclei. Based on protein expression, the results show that there is few variation between macrophages from various donors. Although the macrophages have a different origin and were derived from various tissues (liver or bone marrow) or never been in a tissue (GC-macrophages), the most important proteins associated with erythroblastic islands were identified in all three populations. This further strengthens the hypothesis that GC-macrophages can be used to study the interaction between red cells and macrophages. The method developed in this thesis to study erythroblastic-islands is important for further research in red blood cell cultures for transfusion purposes.

Finally, chapter 6 summarizes all chapters enclosed in this thesis and the results are discussed in a general discussion.

Samenvatting

171

Samenvatting

Ondanks de vele donoren die jaarlijks bloed doneren, is het voor sommige patiënten moeilijk, zo niet onmogelijk, om compatibel (uitwisselbaar) bloed te vinden. Dit geldt voornamelijk voor patiënten met sikkelcelziekte die zijn geïmmuniseerd met verschillende lichaamsvreemde antigenen (zogenaamde allo-antigenen) door eerdere bloedtransfusies. De mogelijkheid van het kweken van rode bloedcellen om deze patiënten gematcht bloed te bieden wordt al decennialang onderkend. Door technische vooruitgang is onderzoek naar het efficiënt kweken van rode bloedcellen om hoeveelheden te kunnen bereiken die overeenkomen met een zak bloed voor transfusie de laatste paar jaar in een stroomversnelling geraakt. In deze studies wordt voornamelijk gekeken naar het intrinsieke rijpingsprogramma van rode bloedcellen. Echter, rode bloedcellen ontwikkelen zich in een omgeving met gespecialiseerde verzorgende cellen, de macrofagen. In deze thesis is onderzoek gedaan naar het optimaliseren van celkweken van stamcel tot rode bloedcel. Daarbij is gekeken naar de rol van verzorgende macrofagen die een belangrijke maar veelal onbekende rol vervullen in de aanmaak van rode bloedcellen.

Rode bloedcellen oftewel erytrocyten, worden in het beenmerg gemaakt. In het beenmerg is een kleine fractie bloed stamcellen (hematopoëtische stamcellen) aanwezig die in de juiste omgeving, met de juiste balans in groeifactoren kunnen uitrijpen naar alle bloedcellen in het lichaam. De omgeving speelt hierbij een belangrijke rol, omdat de cellen die naast de hematopoëtische stamcellen liggen in het beenmerg de stamcellen stimuleren om uit te rijpen naar een van de vele bloedcellen, een proces dat hematopoëse heet. Het uitrijpen van een hematopoëtische stamcel naar een rode bloedcel wordt erytropoëse genoemd. Een van de cellen die in het beenmerg naast de stamcellen ligt is de macrofaag (uit het Grieks makros = groot, phagein = eten), een bepaald type witte bloedcel dat bekend staat om het opruimen van oude, dode of beschadigde cellen en ziekteverwekkers. Er zijn verschillende soorten macrofagen in het lichaam welke in verschillende type weefsels specifieke functies uitoefenen. In het beenmerg hebben de macrofagen naast de klassieke opruimfunctie ook een verzorgende functie en zorgen ze voor de voorlopers van rode bloedcellen, de erytroblasten. Deze macrofaag wordt omringt door erytroïde cellen en vormt een soort eilandje, een erytroblast-eiland. Deze macrofaag zorgt ervoor dat de erytroblasten vermeerderen (prolifereren) en uitrijpen (differentiëren) tot een rode bloedcel door veelal onbekende of niet goed beschreven mechanismen. Tussen erytroblast en rode bloedcel zitten meerdere uitrijpingsstadia. In het laatste stadium stoten de erytroïde cellen hun celkern uit (enucleatie) en ontstaan de geënucleeerde reticulocyten en de afzonderlijke celkernen omgeven door een plasmamembraan, de pyrenocyten. De reticulocyt rijpt vervolgens verder tot een rode bloedcel op het moment dat het de bloedbaan binnen komt. Ook in dit laatste stadium spelen macrofagen een belangrijke rol, omdat ze de celkern opruimen die wordt uitgestoten.

Omdat hematopoëtische stamcellen nodig zijn voor het kweken van rode bloedcellen, en beenmerg moeilijk te verkrijgen is, wordt perifeer bloed als startmateriaal gebruikt. Bloed bevat namelijk een zeer kleine fractie (gemiddeld 0.16%) hematopoëtische stamcellen die uit het beenmerg zijn gemigreerd. Om deze fractie te isoleren wordt het bloed op dichtheid gescheiden in drie lagen: 1)

Appendix

172

rode bloedcellen en granulocyten, 2) witte bloedcellen, en 3) plasma en bloedplaatjes. De tweede laag noemen we ook wel perifere bloed mononucleaire cellen, oftewel PBMC, en in deze laag bevinden zich ook de hematopoëtische stamcellen. Eerder onderzoek heeft aangetoond dat er 10-15 keer meer erytroblasten kunnen worden gekweekt van PBMC dan van gezuiverde hematopoëtische stamcellen geïsoleerd van eenzelfde hoeveelheid PBMC. De reden hiervoor is mogelijk dat bepaalde cellen in PBMC een goede ondersteuning bieden aan de stamcellen en daarmee de hoeveelheid erytroblasten indirect beïnvloeden. Dit hebben we onderzocht in hoofdstuk 2 waar we laten we zien dat er in PBMC verschillende witte bloedcellen zitten, waaronder voornamelijk T-cellen (58%), B-cellen (6%) en monocyten en macrofagen (18%). Alle cellen in het bloed zijn te onderscheiden aan de hand van antigenen (zogenaamde CD-antigenen) die zich aan de oppervlakte van de cel bevinden. Allereerst hebben we een nieuwe manier ontwikkeld om de stadia tussen hematopoëtische stamcellen (CD34+CD36-) en erytroblasten (CD34-CD36+) te karakteriseren, zodat de differentiatie gemakkelijker kan worden gevolgd. Daarna zijn de T-cellen (CD3), B-cellen (CD19) en monocyten (CD14) één voor één uit de PBMC verwijderd en de overige cellen verder gekweekt tot erytroblasten. De hoeveelheid erytroblasten was aanzienlijk minder wanneer de monocyten/macrofagen uit de PBMC waren verwijderd. Bovendien hebben we laten zien dat de hoeveelheid erytroblasten vergelijkbaar was met de opbrengst uit PBMC als de gezuiverde hematopoëtische stamcellen samen werden gekweekt met monocyten/macrofagen. De resultaten laten zien dat monocyten/macrofagen geen effect hebben op de erytroblasten zelf, maar vroeg in kweek zorgen dat er meer stamcellen overleven door het uitscheiden van specifieke factoren. Dit leidt uiteindelijk tot een toename van het aantal erytroblasten.

In hoofdstuk 3 laten we een kweekmethode zien om efficiënt rode bloedcellen te kweken vanuit PBMC. Dit kweeksysteem is in drie fasen op te delen: i) het kweken van PBMC naar erytroblast, ii) het expanderen van de erytroblasten en iii) het differentiëren van erytroblast naar reticulocyt. Om de juiste omgeving na te bootsen, voegen we in elke fase een specifieke cocktail aan groeifactoren toe aan het kweekmedium. De toevoeging van groeifactoren zorgt dat de cellen overleven, prolifereren en differentiëren. In de eerste fase zorgt stamcelfactor (SCF) voor de proliferatie van de hematopoëtische stamcellen. De toevoeging van erytropoëtine (beter bekend als EPO) dat normaal gesproken in de nieren wordt aangemaakt, stimuleert de vorming van rode bloedcellen. Om te voorkomen dat de cellen alvast gaan differentiëren naar rode bloedcellen, wordt dexamethason toegevoegd. Dit is een glucocorticosteroïde, een bijnierschorshormoon met een ontstekingsremmende functie, welke is verhoogd in het lichaam tijdens stress-erytropoëse als er opeens veel erytroïde cellen moeten worden geproduceerd in het beenmerg. In de tweede fase wordt dezelfde cocktail ook gebruikt om de hoeveelheid erytroblasten te vermeerderen zonder differentiatie. Als er voldoende erytroblasten zijn verkregen, wordt zowel SCF als dexamethason weggelaten uit het kweekmedium om te zorgen dat de cellen verder kunnen differentiëren en erytroblasten hun celkern kunnen uitstoten. Ook wordt de concentratie aan EPO verhoogd en wordt extra holotransferrine, een ijzerbindend en -transporterend eiwit dat standaard in het medium zit, toegevoegd. IJzer ionen in het hemoglobine zijn namelijk een essentieel onderdeel van rode bloedcellen, omdat deze zuurstof naar de verschillende lichaamsdelen helpt te transporteren. Tot slot zorgt de toevoeging van plasma ervoor dat de cellen beter differentiëren, waarbij het antistollingsmedicijn heparine voorkomt dat het medium gaat samenklonteren. Om de hoeveelheid rode bloedcellen die nodig is voor één transfusie

Samenvatting

173

te produceren, zijn er zoveel kweekschalen nodig dat dit inefficiënt wordt. Daarom hebben we gekeken naar een andere manier van kweken. We hebben fase één uitgevoerd in kweekschalen en zijn vervolgens vanaf de tweede fase overgestapt op een bioreactor met optimale gasuitwisseling door een gasdoorlaatbare bodem (G-Rex systeem). Hierin kunnen per oppervlakte veel meer cellen gekweekt worden, daarnaast past er een groter volume kweekmedium in dat efficiënter kan worden vervangen. Met deze methode kan een hoeveelheid rode cellen worden geproduceerd die voldoende is om een klinische trial te starten om de levensduur van gekweekte rode bloedcellen na transfusie te testen en te onderzoeken wat de optimale periode is van herhaalde transfusie met gekweekt bloed.

Dat macrofagen een belangrijke rol spelen in de erytropoëse was al bekend. Echter, de wijze waarop ze hun functie uitoefenen is nog niet helemaal duidelijk, omdat er geen goed humaan model beschikbaar is om dit te bestuderen. In hoofdstuk 4 is daarom onderzocht of we monocyten uit PBMC kunnen uitgroeien tot macrofagen die we kunnen gebruiken als model om de interactie tussen macrofagen en verschillende rode cel stadia te bestuderen. Omdat we in ons kweeksysteem in zowel de eerste als tweede fase dexamethason gebruiken, hebben we eerst gekeken wat het effect van dit synthetische hormoon is op monocyt differentiatie. We hebben in hoofdstuk 2 laten zien dat monocyten gekweekt met EPO, SCF en dexamethason differentiëren naar weefsel residente (blijvende) macrofagen. Hier laten we zien dat glucocorticosteroïden heel specifiek alle monocyten aanstuurt om te differentiëren naar macrofagen die qua fenotype lijken op beenmerg macrofagen die aanwezig zijn in erytroblast-eilandjes, omdat ze CD14, CD16, CD163, CD169, CD206 en CXCR4 tot expressie brengen. Omdat deze macrofagen gekweekt zijn met een glucocorticosteroïde, noemen we ze GC-macrofagen. We laten namelijk zien dat deze macrofagen een ander type macrofaag zijn dan de cellen die zonder dexamethason zijn gekweekt. Daarnaast hebben GC-macrofagen ook functionele karaktereigenschappen gemeenschappelijk met macrofagen uit erytroblast-eilandjes: i) ze binden aan erytroblasten en alle andere stadia rode cellen en ii) ze nemen de celkern op van erytroblasten die net hun celkern hebben uitgestoten en iii) ze zijn heel beweeglijk. Met deze resultaten stellen we voor dat de GC-macrofagen kunnen worden gebruikt als model voor het bestuderen van erytroblast-eilandjes. Het hoofdstuk sluit af met een antigeen karakterisatie van beenmerg macrofagen, maar ook van macrofagen uit de foetale lever, omdat de productie van rode bloedcellen tot de 7e maand van de foetale ontwikkeling plaatsvindt in de lever in plaats van het beenmerg.

Als vervolg op dit hoofdstuk, wordt in hoofdstuk 5 een uitgebreide vergelijking gemaakt tussen CD163 foetale lever macrofagen, CD163 beenmerg macrofagen en gekweekte CD163 GC-macrofagen. De morfologie van de drie cel populaties vertoont veel overeenkomsten, ook omdat ze allemaal erytroblasten en reticulocyten binden en erytroïde celkernen opnemen. Op basis van de eiwitten die ze tot expressie brengen laten de resultaten zien dat er weinig verschil zit in de macrofagen tussen de verschillende donoren. Ondanks dat de macrofagen een andere oorsprong hebben en zich in een ander type weefsel bevonden (lever of beenmerg) of nooit in een weefsel zijn geweest (GC-macrofagen), zijn de belangrijkste eiwitten die geassocieerd zijn met erytroblast-eilandjes in alle drie de populaties geïdentificeerd. Dit versterkt de hypothese dat GC-macrofagen kunnen worden gebruikt om de interactie tussen rode bloedcellen en macrofagen te bestuderen. De

Appendix

174

in dit proefschrift ontwikkelde methode om erytroblast-eilandjes te kunnen bestuderen is belangrijk voor verder onderzoek naar het kweken van rode bloedcellen voor transfusie doeleinden.

Tot slot beslaat hoofdstuk 6 een korte samenvatting van alle hoofdstukken die in deze thesis zijn bijgesloten en worden de bevindingen besproken in een algemene discussie.

List of publications

177

List of publications

Heideveld E, Masiello F, Marra M, Esteghamat F, Yağcı N, von Lindern M, Migliaccio AR, van den Akker E.CD14+ cells from peripheral blood positively regulate hematopoietic stem and progenitor cell survival resulting in increased erythroid yield. Haematologica 2015 Nov;100(11):1396-1406.

Heideveld E, van den Akker E.Digesting the role of bone marrow macrophages on hematopoiesis. Immunobiology 2017 Jun;222(6):814-822.

Heideveld E, Hampton-O’Neil L, Cross SJ, van Alphen FPJ, van den Biggelaar M, Toye AM, van den Akker E. Glucocorticoids induce differentiation of monocytes towards macrophages that share functional and phenotypical aspects with erythroblastic island macrophages. Haematologica 2018 Mar;103(3):395-405.

Heideveld E, van Alphen FPJ, van den Biggelaar M, van den Akker E.The proteomic landscape of erythroblastic island macrophages from human fetal liver, bone marrow and in vitro cultures. Submitted.

PhD portfolio

179

PhD portfolio

PhD training

Courses and workshops Year ECTS*

Sanquin Science Course 2014 0.7

Laboratory Animals 2014 4.2

Advanced Immunology 2014 2.9

Oral Presentation 2014 0.3

Eductional Skills Training 2014 0.2

Synthesis of Evidence 2014 0.5

Browsing Genomes with Ensembl 2014 0.3

New Developments in Flow cytometry 2014 0.3

APROVE Pimp My Sollicitatie 2017 0.2

Grant Writing 2017 0.3

APROVE Career Event 2017 0.5

Pitch Like a CHAMP 2017 0.5

Work Your Talent 2017 0.5

Seminars and masterclasses Year ECTS

Department Meetings 2013-2017 8.0

Journal Club 2013-2017 3.0

Landsteiner Lectures and Guest Speakers 2013-2017 2.0

Landsteiner Masterclasses 2013-2017 1.5

Sanquin Science Day 2014 2014 0.5

Sanquin Science Day 2016 2016 0.5

Sanquin Science Day 2017 2017 0.5

National conferences Year ECTS

Dutch Society for Cell Biology (DSCB)

- From Stem Cells to Aging, Amsterdam 2013 0.3

- Vesicles on the Move, Amsterdam 2014 0.3

- Building & Repairing Tissues, Amsterdam 2015 0.3

Dutch Society of Stem Cell Research (DSSCR)

- 7th Annual Meeting, Groningen 2014 0.3

- 8th Annual Meeting, Utrecht 2015 0.3

- 9th, Annual Meeting, Utrecht 2016 0.3

- 10th Annual Meeting, Utrecht 2017 0.3

Appendix

180

Dutch Hematology Congress (DHC)

- 8th DHC, Arnhem 2014 0.6

- 9th DHC, Arnhem 2015 0.6

- 10th DHC, Arnhem 2016 0.6

- 11th DHC, Arnhem 2017 0.6

Molecular Aspects of Hematopoietic Disorders (MAHD)

- Hematology Lectures, Rotterdam 2014 0.7

- Hematology Lectures, Rotterdam 2017 0.7

Tumour Infiltrating Myeloid Cell Compartment (TIMCC)

- The Complexity of Tumour-Host Immune 2016 0.3

Interactions, Leiden

Dutch Society for Immunology (DSI)

- Networking in Immunity, Lunteren 2017 0.6

Sanquin Spring Seminars

- Iron Metabolism and Anemia, Amsterdam 2017 0.6

International conferences Year ECTS

American Society of Hematology (ASH)

- 55th, New Orleans, LA, USA 2013 1.4

- 58th, San Diego, CA, USA 2016 1.4

International Stem Cell Forum (ISCF)

- Regenerative Medicin & Stem Cells, Amsterdam 2014 0.3

European Macrophage Dendritic Society (EMDS)

- Krakow, Poland 2015 0.9

- Amsterdam, the Netherlands 2016 0.9

- Madrid, Spain 2017 0.9

Red Cells Gordon Research (GRS/GRC)

- 2017 seminar, Newport, RI, USA 2017 0.6

- 2017 conference, Newport, RI, USA 2017 1.4

PhD portfolio

181

Presentations Year ECTS

Department Meetings 2013-2017 2.0

Landsteiner Lectures 2013-2017 0.6

Journal Club 2013-2017 0.3

Invited seminar MAHD, Rotterdam 2014 0.5Human CD14+ monocytes/macrophages from peripheral

blood positively regulate hematopoietic stem and progenitor cells

Invited seminar MAHD, Rotterdam 2017 0.5The intimate relationship between red cells and macrophages

Invited seminar, MRC, Edinburgh, UK 2017 0.5The role of macrophages in erythropoiesis

Poster presentation ASH 2013 0.5Identification of specific hematopoietic populations present in peripheral

blood mononuclear cells that increase erythroid output directly or through

feeder/stromal cell-like properties

Oral presentation DHC 2014 0.5Identification of monocytic populations present in peripheral blood

mononuclear cells that increase the erythroid output

Poster presentation DSSCR 2014 0.5Human CD14+ monocytes/macrophages from peripheral blood positively

regulate hematopoietic stem and progenitor cells

Poster presentation Sanquin Science Day 2014 0.5Human CD14+ monocytes/macrophages from peripheral blood positively

regulate hematopoietic stem and progenitor cells      

Oral presentation DHC 2015 0.5Characterization of human CD14+ cells in PBMC that are responsible for

increased erythroid yield

Oral presentation EMDS 2015 0.5Characterization of human CD14+ cells in peripheral blood that are

esponsible for hematopoietic stem and progenitor cell survival

Oral presentation DHC 2016 0.5CD14+ cells from peripheral blood positively regulate hematopoietic

stem and progenitor cell survival resulting in increased erythroid yield

Poster presentation EMDS 2016 0.5Cultured human macrophages as a model for central macrophages in

erythroblastic islands

Poster Presentation Sanquin Science Day 2016 0.5Cultured human macrophages as a model for central macrophages in

erythroblastic islands

Poster presentation ASH 2016 0.5Cultured human macrophages as a model for central macrophages in

erythroblastic islands

Oral presentation DHC 2017 0.5Cultured human macrophages as a model for central macrophages in

erythroblastic islands

Appendix

182

Poster presentation Sanquin Spring Seminar 2017 0.5Modeling human erythroblastic islands

Poster presentation GRS 2017 1.0Modeling human erythroblastic islands

Poster presentation GRC 2017 1.0Modeling human erythroblastic islands

Poster presentation EMDS 2017 0.5Proteomic comparison of CD163+ macrophages in human fetal liver,

bone marrow and cultured central macrophage-like cells

Poster presentation Sanquin Science Day 2017 0.5Proteomic comparison of CD163+ macrophages in human fetal liver,

bone marrow and cultured central macrophage-like cells

Additional tasks Year ECTS

DJOB committee member 2014-2016 1.5

FBCS committee member 2016-2017 0.8

Teaching Year ECTS

Bachelor student 2014 0.3

Bachelor Student 2014 0.3

HLO Student 2016 1.5

Awards and Prizes Year

Dutch TOP Publications, DHC 2015

Best Poster Presentation Award, Sanquin Spring Seminar 2017

Poster Prize, Gordon Red Cell Seminar 2017

*1 ECTS credit equals 28 hours workload

185

Dankwoord

Dankwoord

Misschien zou het dankwoord beter tot zijn recht komen aan het begin van dit boekje. Velen van jullie

staan namelijk niet op mijn artikelen, maar zonder jullie was het nooit zo mooi uit de verf gekomen.

Ieder heeft mij op zijn eigen manier geholpen: directe betrokkenheid bij mijn project of indirect om even

te ontspannen buiten het lab. Sommigen kennen mij al jaren, voor anderen lijkt het alsof we elkaar al

jaren kennen, en soms zijn we uiteindelijk onze eigen weg gegaan. Ik wil jullie allemaal bedanken en

ik kijk dan ook heel gelukkig terug op de afgelopen jaren en de vriendschappen die zijn opgebouwd!

Emile, ik vind het een eer dat ik de eerste PhD student ben die bij jou promoveert. Die allereerste dag

dat je mij als masterstudent hielp met het isoleren van PBMCs weet ik nog als de dag van gisteren.

Pas om 3 uur ’s middags waren we klaar en was er tijd voor lunch. Ik zag er al tegenop om elke

dag zo laat te lunchen, waarna jij zei ‘als ik het zelf doe, ben ik om 11 uur al klaar!’. Inmiddels heb ik

zoveel PBMC en MACS isolaties gedaan, dat ik durf te wedden (voor een fles rode wijn) dat het mij

vóór 11en ook wel lukt. Door mijn leuke tijd tijdens mijn stage ben ik bij je teruggekomen als PhD

student, en dat was een hele goede zet! Emile, ik heb bewondering voor de hoeveelheid kennis die

je hebt, de orde die verstopt zit in de chaos waardoor (alleen) jij je data feilloos weet terug te vinden.

Ik heb meegemaakt hoe onze groep en daarmee ook jij bent gegroeid, van gezellig samen in het

lab cellen kweken, naar een echte groepsleider: een gepassioneerde en enthousiaste onderzoeker,

altijd druk, nooit op zijn plek aanwezig, en in de vrije uurtjes nog wat stiekeme experimentjes doen.

Die vrijheid om secret experiments te doen heb ik zeer gewaardeerd. ‘Een goed idee moet je altijd

uitproberen, als het werkt, laat het me weten, en anders..wat niet weet, wat niet deert.’. Je hebt mijn

ogen geopend om meer van de wereld te willen zien en ik heb gemerkt dat ik het ontzettend leuk

vind om te netwerken. Bedankt dat je mij de mogelijkheid bood om naar de ASH (New Orleans en

San Diego), de Gordon Red Cell (Newport), de EMDS (Krakau en Madrid) en alle congressen binnen

Nederland te gaan. Daarnaast mocht ik ook nog geheel onverwachts een maand naar Bristol om een

mooi figuur te maken voor ons artikel. En dan te bedenken dat ik voor 2013 nog nooit in een vliegtuig

had gezeten! Naarmate de 4 jaar verstreken (en helaas de gaatjes in je shirts toenamen) ben je ook

steeds persoonlijker geworden en was het leuk om meer van jouw leven te weten te komen. Ik denk

dat we een goed team zijn samen, jij kijkt naar het geheel en ik let op de details (al heb ik je misschien

wel tot gekheid gedreven met mijn geordendheid, planningen en deadlines). Zo, en nu ben ik wel

benieuwd wat er allemaal over mij in jouw excel file staat!

Marieke, met jouw altijd vrolijke en opgewekte humeur en passie voor de wetenschap vind jij zelfs

in alle drukte nog tijd om even wat te bespreken of gewoon gezellig te babbelen over muziek, PhD

studenten uit de tijd dat je zelf nog promoveerde of over andere Vrinden. Bedankt voor alle input

tijdens de werkbesprekingen, maar ook tijdens het schrijven van de artikelen en dit boekje. Samen

met Emile vormen jullie een sterk (maar soms ook wel wat chaotisch) team! Ik weet nu dat het niet

erg is als je op de eerste rij zit te knikkebollen tijdens mijn presentatie, dit doe je immers alleen als

Appendix

186

je het vertrouwen hebt dat het toch wel goed komt. En dat vind ik een mooie gedachte om aan vast

te houden. Mocht je nog eens zin hebben om naar een concert te komen luisteren, dan ben je altijd

van harte uitgenodigd.

Ash, you could have been my third copromotor. Thank you for the opportunity to (ad hoc) visit your

lab to finish the last figure of the paper and celebrate the publication some months later while writing

my thesis. I really felt like being part of your group. It is treasured memories like Christmas dinner,

your birthday party (best cocktails I ever had), having lunch at St. Nicholas Market and (Wii) dancing

in the living room, that make me realize I will definitely return to Bristol again (and again, and again).

Ellen, eigenlijk was jij in de laatste maanden pas echt betrokken bij mijn project. Ik heb bewondering

voor de hoeveelheid kennis die je hebt en hoe je altijd verrassend interessante vragen weet te

stellen tijdens werkbesprekingen en tijdens het nalezen van dit proefschrift, zelfs als je niet echt in

het onderwerp thuis bent. Mijn ketting uit New Orleans bewaar ik als aandenken aan de leuke tijd op

Sanquin, want hoe konden wij weten dat we die niet zomaar mochten oprapen!

Alle PI’s, zowel van U2 als Y3 en Y4, bedankt voor alle input tijdens de werkbesprekingen, de leuke

discussies bij de koffieautomaat of het doorsturen van interessante artikelen.

Ik wil graag iedereen uit Emile’s en Marieke’s groep bedanken voor de fijne werkbesprekingen op de

donderdagen en de leuke gesprekken tijdens het kweken op het ML1, stainings maken voor de FACS

of cytospins kleuren. In het bijzonder Nurcan, jij bent voor mij nog steeds de moeder van de groep,

bedankt voor al jouw hulp tijdens mijn PhD, maar ook zeker al daarvoor tijdens mijn stage. Het is

jammer dat je nooit meer teruggekomen bent op mijn project, want ik weet zeker dat we een goed

team zijn! Patrick, hèhè, eindelijk iemand in de groep die iets weet van macrofagen! Dankjewel voor

het delen van jouw kennis en ook voor de mooie samenwerking met Erica en Marijke die resulteerde

in hoofdstuk 3. Franca, fijn dat er meer mensen zijn die houden van orde en regels! Elina, thanks for

the fun times in the lab, the NvvH and during lunch breaks at Wonders.

A100, it was a real pleasure meeting you, everyone has been so friendly and kind! In particular Lea,

you are the perfect guide, both in the lab as well as outside. I enjoyed working with you, drinking tea

at Bluebird, discovering secret bars in Bristol and staying at your comfortable sleeping bed. Good

luck with your final months and hope to see you soon!

Erik, Mark en Simon, bedankt voor jullie hulp en geduld bij de FACS, ImageStream, elutra en sort.

Jullie vrolijke humeur zorgde er altijd voor dat mijn lange experimenten wat korter leken. Floris

en Maartje, ik wil jullie bedanken voor alle hulp bij de mass spectrometry experimenten van de

macrofagen. Zonder de kennis en hulp van jullie allemaal als Research Facility waren de hoofdstukken

nooit zo mooi geworden (of had ik ze überhaupt nooit kunnen schrijven).

Dankwoord

187

Mariette, niets is zo lekker als een koffie met melkschuim! Bedankt dat je altijd klaar stond om bloed

(van mij of voor mij) te prikken. Sinds maart heb je er een IJsland-fan bij!

Mo, Anita, Kaoutar en Fatima, maar ook John, wat fijn dat er mensen zijn die orde en overzicht

weten te scheppen bij al die chaotische onderzoekers. Bedankt voor jullie hulp, regelwerk en geduld!

Bedankt (oud)-IHEP’ers! In het bijzonder, Carlijn, jij bent altijd in voor een goede borrel en een hoop

vrolijkheid! Ik moet nog even oefenen denk ik, want ik weet anders niet of ik onze weddenschap wel

haal ;) Tamara, bedankt voor jouw gezelligheid, ook tijdens de spelletjesavonden, al hebben we meer

tijd besteed aan lekker eten en kletsen dan aan spelletjes! Ik wacht nog steeds op het moment dat

je een ritje op mijn fiets komt maken (geloof me, mijn zadel is echt niet zo erg). Giso, als ik jou zie

begin ik spontaan te MACSen, zelfs als je even langs komt terwijl je niet meer op de afdeling werkt.

Wanneer gaan we weer de Dam-tot-Dam fietsen? Felix and Fiamma, you were always there, late at

night, when everyone left and I really wanted to talk. Thank you for being there (and for the cookies

Felix!). Benoit, Mister Green, were would Sanquin be without you? Plastic containers, department

pictures, classical music in the lab..thank you! Marion, zonder jouw gekleurde lijstjes op het ML1

was het nooit zo schoon en opgeruimd geweest op het lab! Jalenka, Ilse en Corina, bedankt voor

jullie vrolijkheid, gezelligheid en leuke gesprekken! Pleun, het is alweer veel te lang geleden dat wij

tijd hadden voor bier en een goed gesprek! Christina, hoe meer hobo’s, hoe beter. Ammarina, fijn

om te weten dat sommigen het nog kouder hebben dan ik. Florentine, het kookboek is er, nu nog

wachten tot sinterklaas, want ik hoop dat onze bakavonden een vervolg krijgen. Felipe, Sabrina (rood!

ze zijn ROOD!), and Sulima, thank you for all the fun times at P1.

Fietsmaatjes: Sem, bedankt voor de gezellige fietstochtjes van en naar werk op de maandagen,

al kwam ik altijd wel iets te laat. Gelukkig kon ik compenseren door bij de maandagpraatjes de

afkortingen van Research uit te leggen. Julian, mijn Purmerendse vouwfietsmaatje, super leuk om

een jaar lang met de trein te reizen en samen te fietsen! Nahuel, met de grootst mogelijke moeite

probeerde ik jou te overtuigen van mijn groenere fietsroute…300 meter extra? Echt niet. Max and

Marea, luckily you were always there if I didn’t want to cycle home.

De G-unit, bedankt voor jullie vrolijkheid, gezelligheid en gekkigheid, vooral in de laatste maanden

dat ik aan het schrijven was. Waarschijnlijk is het nu een stuk rustiger in het hoekje bij de trap.

Arthur, ik heb respect voor je doorzettingsvermogen en snelheid met hardlopen (en bier drinken)

en blijf vooral zo doorgaan (vooral met dat eerste)! Zoltan, Max and Erik, never try to disturb anyone

counting cells ;) Robin, tot onze verassing hebben wij onze hele jeugd maar een paar meter van elkaar

doorgebracht, wie had dat gedacht. Bedankt voor je vrolijkheid en humor!

Gillian, pepernoten bakken, boulderen, koken uit Ottolenghi, dat zijn toch wel de eerste dingen die

in mij opkomen als ik aan jou denk. Bedankt voor de leuke gesprekken, het zweten op de klimmuur

Appendix

188

en de wandelingetjes in de pauze. We moeten denk ik wel even ons best doen om deze tradities in

ere te houden, maar ik weet zeker dat het ons lukt!

Francesca, jij bent denk ik de enige Italiaanse die het hier in Nederland al te warm vindt om

in de zon te zitten. Bedankt voor je gezelligheid, zowel op werk, als na werk, als tijdens onze

spelletjesavonden. Dobbiamo ancora fare la nostra prova, quindi non so se riesco a questo, perché,

nonostante il nostro buon inizio con un libro di lingua esercizio. Conosco solo una frase in italiano:

sono caduto dalla bici.

Roos, wat ben jij toch gezellig en vol energie, altijd in voor een goed gesprek of lekker wat kletsen

tijdens een koffietje. Ik ben blij dat ik je in de laatste maanden van mijn PhD heb leren kennen, al had

dat véél eerder mogen zijn wat mij betreft!

Anne, ik heb leuke herinneringen aan onze gezellige avondjes in Utrecht. Als het aan mij ligt, worden

dat er nog veel meer! Ik ben blij dat je nieuwe plek echt is wat je zocht, en ik wens je heel veel succes

met het afronden van je PhD (je bent er echt bijna!!). De uilenbeker uit New Orleans staat nog steeds

op je te wachten, dus laten we snel een keer afspreken!

Nahuel, de man van de labello en raak-nooit-zijn-haar-en-schoenen-aan. Al waren we zowat gelijk

begonnen, ik heb je pas in de laatste jaren beter leren kennen. Samen naar San Diego was een leuk

avontuur, al kon je mijn gekletst maar net aan. Ik kom graag al jouw dulce de leche toetjes proeven,

maar dan wacht ik denk ik eerst tot je klaar bent in de keuken. Ik bewonder je dat je ogenschijnlijk

relaxed je PhD hebt afgerond en wens je heel veel succes met je nieuwe baan!

Marieke, door jou is een stukje van mijn steen afgebrokkeld. Dankjewel voor alle leuke Star Wars &

Pizza-avonden en je gezelligheid tijdens de NvvH, NvvI en de ASH in San Diego, ook al heb ik je de oren

van je hoofd gepraat en kwamen onze verschillende karakters soms boven. Ik hoop dat we blijven

afspreken buiten werk (ook nu met je lieve gezinnetje) en wens je veel succes met het afronden van

je PhD, je bent er bijna!

Max, the guy that never forgets to moisturize his hands. You are a great addition to the group, both

at work and at borrels after work. Although I wanted to finish within 4 years, I am glad I needed some

extra months as I found a very enjoyable and talkative friend in you. But what to do with all the cookies

now I can’t share them with you?

Steven, de man met de leuke t-shirt prints, koekiemonster en mede-Italië-fietser. Zelfs na een

superkorte nacht in alle gezelligheid foetale levers opwerken met mij naast je in de flow hood, dat

kunnen maar weinig mensen opbrengen. En daarna durfde je ook nog samen op congres naar

Newport, inclusief stedentrip naar Boston en New York onder het continue gekletst van mij. Ik

Dankwoord

189

waardeer je luisterend oor, de ontzettend goede gesprekken en discussies en wil je bedanken voor

je vertrouwen en gezelligheid. Ik ben blij dat jij mijn paranimf bent!

Marea, de nuchtere, retro, mango-lover die nooit (?) haar principes overboord zal zetten. Aan jou heb

ik een ongelooflijk leuke, grappige en oprechte vriendin overgehouden! Samen antibodies ordenen,

inkopen doen voor de borrels, koffie drinken bij het NKI, gezellige spelletjesavonden (volgende keer

knip ik mijn nagels extra kort) en oude Sanquin pennen verzamelen. Er is denk ik maar één ding dat

tussen ons in staat, en dat is jouw personal space ;) Ik ben blij dat jij mijn paranimf bent!

Ik bedank ook graag alle muzikanten met wie ik de afgelopen jaren heb samen gespeeld of heb

leren kennen na afloop van een concert. In het bijzonder Nieuwe Maten, bedankt voor de gezellige

repetities, mooie concerten en alle energie die ik van jullie heb gekregen door samen muziek te

maken!

Anh, jij bent zo mogelijk nog geordender dan ik! Na 16 jaar elkaar uit het oog te zijn verloren, waren

we alweer snel up-to-date. Vietnamees eten, wijn drinken op je balkon, lijstjes doornemen, (te)veel

kussens kopen of opeens de woonkamer verven, met jou weet je nooit wat de dag zal brengen!

Bedankt voor je positiviteit en gezelligheid.

Dewi, jij kent me echt als geen ander. Ik denk met plezier terug aan al onze Harry Potter en Austen

avonden, Avatar, Sex and the City, koekjes bakken en samen ontbijten met Kerst of Pasen. Ik voel

me altijd enorm welkom en thuis bij jou, en daar hoort Arjan natuurlijk ook bij. Bedankt voor je

vriendschap al die jaren!

Lieve Michel, kurs op de vlaai, enigszins onverwacht sta je hier dan in mijn dankwoord. Ik ben heel

erg blij dat ik jou heb leren kennen en ik wil je bedanken voor je vertrouwen, fijne gesprekken en alle

gezelligheid die jij met je mee brengt. Ik kijk al uit naar de vele avonturen die gaan komen!

Lieve papa, mama en zus. Ik wil jullie bedanken voor alle leuke en mooie momenten die we samen

hebben gedeeld. Saskia, een gezamenlijk potje starten was een heel goed idee! Ik heb genoten

van de zussenvakanties, de gezellige weekendjes weg of gewoon eventjes bijkletsen met een latte

macchiato caramel. Ik ben ontzettend blij met jou als zus! Papa en mama, wie had ooit gedacht dat

ik hier nu zou staan. Van stil en verlegen naar een aanwezige spraakwaterval (niet ongewoon in de

familie). Bedankt voor alle steun die ik van jullie heb gehad, de gezellige telefoontjes als ik op de fiets

zat naar werk, maar bovenal heb ik van jullie meegekregen dat als je iets wilt, je nooit op moet geven,

door moet zetten en er helemaal voor moet gaan.

Liefs,

Esther

Curriculum Vitae

191

Curriculum Vitae

Esther Heideveld was born on the 17th of August 1988 in Purmerend, The Netherlands. She attended high school at ‘Het Da Vinci College’ in Purmerend and graduated in 2005 (HAVO). She continued with pre-university education (VWO) and additionally started the preparatory program oboe at the Amsterdam Conservatory in 2005. Because of her interest in both music and medical biology, she started her study oboe at the Amsterdam Conservatory with Jan Kouwenhoven (former oboist at the Royal Concertgebouw Orchestra) as well as Biomedical Sciences at the Free University in Amsterdam in 2007, however, in 2009 she decided to focus on Biomedical Sciences. After performing an internship on the toxicity of non-dioxin-like PCBs in zebrafish embryos at the Laboratory of Health Protection Research and Ecological Risk Assessment at the RIVM, Bilthoven, Esther obtained her Bachelor of Science degree in 2011. She then decided to choose the Biomedical Sciences master program with a specialization in Immunology and Oncology. For her master, she performed a 6-month internship at the department of Hematopoiesis at Sanquin, Amsterdam. Here, she investigated the effect of different human hematopoietic lineages on hematopoietic stem cells and erythroid outgrowth and the reprogramming capacity of erythroblasts. Additionally, she performed a second internship in the group of Dr. J de Rooij at the Hubrecht Institute, Utrecht, on designing FRET probes for αE-catenin in order to study the cadherin complex in cell-cell adhesion. In 2013, she obtained her Master of Science degree in Biomedical Sciences and decided to continue her work at Sanquin, starting a PhD project in the department of Hematopoiesis under supervision of Dr. E van den Akker. This project focused on in vitro culture of red blood cells and the role of macrophages during erythropoiesis. During her PhD she has visited the laboratory of Dr. Ashley M. Toye at the University of Bristol (UK). The results of her research are described in this thesis.