From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
Extruded Erythroblast Nuclei Are Bound and Phagocytosed by a Novel
Macrophage Receptor
ByLi-Bo Qiu, Harriet Dickson, M.A. Nasser Hajibagheri, and Paul R. Crocker
Resident bone marrow macrophages (RBMM) play an important role in clearance of nuclei extrudedfrom late-stage
erythroblasts (Eb).To investigatethe nature of macrophage
receptorsinvolvedin this process, extruded erythroblast nuclei (EEN)were purified by cultivation of erythroblasts with
erythropoietin, followed by density gradient centrifugation.
By electron microscopy, the majority of free nuclei had an
intact plasma membrane. EEN bound avidly to RBMM in a
divalent cation-independent manner at both 4°C and 37°C.
The use of specific monoclonal antibodies (MoAbs) and inhibitorsshowed that this adhesive interactionwas not mediated by previously characterized macrophage receptors involved in recognition of eitherdevelopinghematopoietic
cells or apoptotic cells.The EEN receptorwas expressed on
resident macrophages isolated from murine bone marrow,
spleen, lymph node, and peritoneal cavity, but was completely absent from alveolar macrophages. Despite high levels of EEN binding to RBMM, few were phagocytosed even
after prolongedculture.Phorbol myristate acetate (PMA)
was found to stimulate phagocytosis, suggesting that this
is a regulated process. These findings indicatethat €EN are
recognized by a novel macrophage receptor and
that recognition may be triggered during the membrane remodeling
that accompanies enucleation.
0 1995 by The American Society of Hematology.
IN
vent inappropriate phagocytosis of both the Eb and the newly
developed reticulocytes. Nothingisknown of themechanisms that permit macrophages to discriminate between an
extruded nucleus destined for phagocytosis and a reticulocyte that detaches and enters the circulation.
One possibility is that differences in membrane composition form a molecular basis for the macrophage recognition
events described above. Previous studies have shownthat
the plasma membrane surrounding the nucleus is selectively
enriched in glycoconjugates recognized by concanavalin A
(Con
The nuclear plasma membrane also bindsweakly
to positively charged colloidal iron oxide, in contrast to the
reticulocyte membrane.','"," This reduced negative charge
of extruded erythroblast nuclei (EEN) is a possible explanation for why they are avidlyboundand phagocytosed by
macrophages in vivo."
In previous studies, macrophages isolated from hematopoietic tissues were found to express two distinct receptors
for developing hematopoietic cells, the erythroblast receptor
(EbR) and sialoadhesin.'2,'3EbR mediates divalent cationdependent binding of in
Eb
whereas sialoadhesin
is a divalent cation-independent receptor that is believed to
interact selectively with myeloid precursors.",''~'7
So far, no studies have investigated the nature of receptors
for EEN on resident bone marrow macrophages (RBMM).
A major technical obstacle is the isolation of sufficient numbers of purified EEN to permit direct binding assays. In the
present study, we first developed a method in which a relatively large and homogeneous population of fluorescently
labeled EEN could be isolated from Eb cultured in the presence of erythropoietin. Binding assays were then performed
with RBMM. Our studies showed that EEN wereavidly
bound by RBMM in a divalent cation-independent manner.
The failure of a range of specific antibodies and inhibitors
to block this interaction, as well as the characteristics of
binding andthe pattern of expression on various macrophage
populations, suggest thattheEEN
receptor (EENR) is a
novel macrophage adhesion and phagocytosis receptor.
ADULT MAMMALS, the last step in development of
the erythroid lineage is the production of erythrocytes
by extrusion of nuclei from late-stage erythroblasts (Eb). In
this process, the nucleus moves to one pole of the Eb and
isthen actively expelled, together with an intact plasma
membrane and a small amount of cytoplasm.'.2 The precise
mechanisms that control this expulsion are unknown. However, itisan intrinsic property of the Eb, as enucleation
occurs in isolated Eb cultured in the presence of erythropoietin3 and actin polymerization is required as shown by the
inhibition of nuclear extrusion in the presence of cytochalasin D.4 In addition, immunofluorescence studies have
shown that enucleation is accompanied by cytoskeletal reorganization involving actin, spectrin, and microtubules, all
of which become selectively localized withinthenascent
reticuI~cyte.~.~
In the bone marrow, developing Eb, from the proerythroblast stage onwards, are in close contact with central resident
macrophage^.^,' These associations can be observed in sections of bone marrow as distinct anatomical structures that
have been described as erythroblastic
After enucleation, the nucleus is rapidly phagocytosed by macrophages,'
whereas the reticulocyte detaches from the macrophage surface to enter the blood stream. As proliferating Eb remain
in close contact with the macrophage plasma membrane up
until enucleation, controlling mechanisms must exist to pre-
From the Imperial Cancer Research Fund, Institute of Molecztlur
Medicine, Oxford University, Oxford UK.
Submitted June 14, 1994; accepted November 1. 1994.
Supported by the ImperialCancerResearch
Fund. L-B.Q. wus
supported by the Royal Society, UK; and H.D. was the recipient of
an MRC Clinical Training Fellowship.
M.A.N.H. is at Imperiul Cancer Research Fund, Electron Microscopy Unit, London UK.
Addressreprintrequeststo
Paul R. Crocker, PhD, ICRF Labs.
Institute of MolecularMedicine, John RadclifleHospital,Headington,Oxford,OX3 9DU, UK.
The publication costs ofthis article were defruyed
in part by page
chargepayment. This article must thereforebeherebymarked
"advertisement" in accordance with 18 U.S.C. section 1734 solely to
indicate this jitct.
0 1995 by The American SocieQ of Hematology.
0006-4971/9S/8~06-0020$3.00/0
1630
MATERIALS AND METHODS
Animals. Male or female Balbk mice were bred at Clare Hall.
Imperial Cancer Research Fund (ICRF; South Mimms, Hertfordshire, UK) and used at 8 to 12 weeks of age.
Blood, Vol 85, No 6 (March 15).
1995:
pp 1630-1639
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
MACROPHAGE RECOGNITION OF EXTRUDED NUCLEI
Media and reagents. Dulbecco’s modified Eagle’s medium
(DMEM), RPMI 1640, and phosphate-buffered saline (PBS) without
Ca++ and Mg’+ were obtained from Clare Hall. Fetal calf serum
(FCS) was purchased from Advanced Protein Products Ltd (Brockmoor, UK) and was routinely heat-inactivated for 30 minutes at
56°C. Recombinant murine erythropoietin, collagenase, and DAPI
(4’,6-Diamidine-2-phenylindoledihydrochloride) were obtained
from Boehringer Mannheim (Lewes, UK). Vibrio cholerae sialidase
(1 Behringwerke UlmL) was from Calbiochem (La Jolla, CA). Biotinylated anti-rat IgG and ABC Vectastain Kit were obtained from
Vector Laboratories (Peterborough, UK). Streptavidin microbeads
were obtained from Miltenyi Biotec (Bergisch-Gladbach, Germany).
DiI ( I . 1 ’-Dioctadecyl-3.3,3’,3’-tetramethylindocarbocyanineperchlorate) was purchased from Molecular Probes (Eugene, OR). All
other reagents were purchased from Sigma Chemical CO (Poole,
UK). Sheep redblood cells (RBC) were from Flow Laboratories
(Irvine, Scotland) and were stored at 4°C and used within 2 weeks.
Cell lines and antibodies. The macrophage-like cell lines RAW
264.7 and P388D1 were provided by the cell bank at the Sir William
Dunn School of Pathology, Oxford University (Oxford, UK). The
Mm1 macrophage-like cell line” was provided by Dr T. Masuda
(Kyoto University, Kyoto, Japan). NIH 3T3 cells were obtained
from the ICRF Cell Bank. The following rat monoclonal antibodies
(MoAbs) with specificities shown were used as supernatants or ascites at saturating concentrations: F4/80, specific for mature mouse
macrophages”; 5C6, which binds CDllb on mouse myeloid cells?’;
SER-4 against mouse sialoadhesin”; 2F8, which recognizes the
mouse macrophage scavenger receptor? S7, S 1 1, and S 18 directed
to the murine CD43 antigen on myeloid cells23;FA1 1, which recognizes macrosialin, the murine CD68 homologue24; and ER.HR3
against murine stromal macrophage^.^^
Separation of Eb and EEN. Eb were enriched from adult mouse
bone marrow cells by depletion of myeloid cells. Mice were killed
by inhalation of COz, and bone marrow cells were flushed from
both femora and tibiae. Cells were dissociated by pipetting in PBS
containing 5 mmol/L EDTA and 1% bovine serum albumin (BSA;
PBWBSA), suspended at 108/mL,and incubated with MoAb 5C6
for 15 minutes on ice. Cells were then incubated for 15 minutes
with biotinylated anti-rat IgG ( I O pg per 10’ cells) in PBSBSA,
followed by 10 minutes with streptavidin magnetic microbeads (100
pL per 10’ cells) in PBS plus 5 mmol/L EDTA. Throughout the
procedure, cells were kept on ice, and after each incubation they
were washed twice with the appropriate buffer. The labeled cell
suspension was then passed through a magnetic cell separation column (MACS; Miltenyi Biotec), and nonmagnetic cells were collected. These included the majority of Eb, contaminating RBC, reticulocytes, and B lymphocytes. This cell suspension was layered onto
a self-generated 70% Percoll continuous density gradient (5,OOOg,
4°C for 20 minutes) and centrifuged at 500g at 4°C for 30 minutes.
Eb (greater than 90% pure on May-Grunwald-Giemsa-stained cytopreps) were collected from the lowest density fraction of the gradient,
washed twice in PBS, and resuspended in DMEM plus 20% FCS.
To obtain and isolate EEN, the purified Eb were cultured at 37°C
in a 7% COz atmosphere in DMEM containing 20% FCS and 1.5
U/mL murine erythropoietin. After 24 hours of culture, the cells
were harvested, washed in PBS, and layered onto a self-generated
70% Percoll density gradient as described above. EEN were collected
from the highest density fractions of the gradient, washed twice in
PBS, and resuspended in assay solution at an appropriate concentration. Cells from different fractions were analyzed by staining cytocentrifuge smears with May-Grunwald-Giemsa.
Electron microscopy. Eb and EEN were isolated as described
above. Cell pellets were fixed in 2.5% (wt/vol) glutaraldehyde in
0. I molk Sorensens phosphate, pH 7.4, for 2 hours at room temperature. After washing (three times for 5 minutes) with Sorensens’s
1631
phosphate buffer, the pellets were postfixed in 1% osmium tetraoxide
in the same buffer for 60 minutes. The samples were processed by
dehydration in graded series of ethanol followed by propylene oxide
and embedded in araldite resin (Taab Laboratories Equipment Ltd,
Reading, UK). Sections were stained for 10 minutes in a 2% (wt/
vol) solution of saturated lead citrate and examined with a Jeol 1200
electron microscope (Jeol, Tokyo, Japan).
Isolation of macrophage populations. RBMM were isolated as
described previously with some modifications.26Briefly, adult femoral and tibial bone marrow plugs were digested with 0.05% collagenase in RPMI for 45 to 60 minutes with constant rotation. Hematopoietic clusters enriched for RBMM were separated from single cells
by centrifugation at 500 rpm for 10 minutes on a self-generated
density gradient of 30% Percoll prepared as described above. Clusters were resuspended in RPMI plus 10%FCS and allowed to adhere
to 5-mm diameter, 12-spot glass slides (pH-086; Hendley, Essex,
UK) for 3 to 4 hours at 37°C in 5% CO?. Hematopoietic cells were
detached from adherent RBMM by incubation in PBS for 20 minutes
followed by flushing with PBS using a wide-bore plastic pipette.
Resident peritoneal macrophages were obtained from untreated mice
after lavage of the peritoneal cavity with PBS.Bronchoalveolar macrophages were collected by tracheal cannulation followed by lavage
withPBS plus 0.5mmol/L EDTA. Spleen and mesenteric lymph
node macrophages were obtained by mincing tissues from mice and
digesting them with 0.05% collagenase in RPMI for 60 minutes.
Undigested material was discarded. The macrophage-like cell lines
Mml, J774, and P388D1 were cultured in DMEM plus 10% FCS.
Before binding assays, the cells were washed twice in RPMI, resuspended in RPMI plus 10% FCS, and plated onto 12-spot glass slides
at 500 to 1,OOO adherent cells per spot. Nonadherent cells were
removed by gentle, direct pipetting with RPMI after 3 to 4 hours of
incubation at 37°C in 5% COz.
Binding and phagocytosis assays. Binding assays were performed in media either with (DMEM plus 20 mmollL HEPES) or
without (PBS plus 0.1 mmol/L EDTA) divalent cations. Slides with
adherent macrophages were rinsed three times in the assay medium.
Cells and nuclei, prepared as described above, were washed three
times in PBS and resuspended in assay medium. Forty microliters
of the cell suspension (5 X 106/mLof Eb or EEN; 1% vol/vol sheep
RBC) were added to adherent macrophages on each spot ofthe
slides. These slides were placed in a humidified chamber and incubated for 45 minutes at 37°C in 5% CO2. Unbound cells or nuclei
were removed by gently rinsing the slides in the assay medium.
Cells were then fixed for 15 minutes at room temperature in PBS
containing either 4% paraformaldehyde for fluorescence staining or
0.25% (vol/vol) glutaraldehyde for immunohistochemistry and then
rinsed twice in PBS. Macrophages were identified morphologically
by phase-contrast microscopy or by immunolabeling with F4/80.
EEN and Eb were identified by simultaneous DiI and DAPI fluorescence. Sheep RBC were identified by phase-contrast microscopy in
binding assays. Macrophages were considered positive if they bound
three or more cells. Assays were routinely performed in duplicate,
and 150 to 200 macrophages were scored on each spot.
Phagocytosis assays were performed in DMEM plus 20 mmoliL
HEPES. At the end of the 45-minute incubation period, nonphagocytosed EEN were removed from macrophages by incubation in 100
,ug/mLof trypsin at 37°C for 20 minutes, followed by gentle pipetting. Macrophages were then fixedin 4% paraformaldehyde,
rinsed, and mounted in 50% glycerol in PBS. Phagocytosis of EEN
was determined by both phase-contrast and fluorescence microscopy.
Macrophages were scored as positive if they contained one or more
EEN. At least 200 macrophages on duplicate spots were counted
for each data point.As a control for phagocytosis, zymosan was
dissolved inPBS at 1 pg/mL, boiled for 30 minutes, andwashed
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1632
twice in PBS before use; approximately lo6 particles were added to
each spot.
Fluorescentstaining.
Double fluorescence was used to identify
EEN with an intact plasma membrane. The Eb plasma membrane
was labeled before culture with the red fluorescent lipophilic dye,
DiI, as described previo~sly,~'
with some modifications. Eb, purified
as above, were resuspended in DMEM plus 10% FCS. Cells were
then incubated with DiI at 25 pg/mLfor 30 minutes at 3 7 T , washed
three times in PBS, and cultured with erythropoietin as described
above.
Staining of nuclei with DAPI was performed at the end of the
binding assays. After fixation, cells on slides were incubated with
1 pg/mL DAPI in PBS for 10 minutes at 3 7 T , rinsed twice with
PBS, and mounted in 50%glycerol in PBS. The efficiency of staining
was checked by fluorescence microscopy.
Immunohistochemistry. Cells on slides were fixed for 10 minutes
at room temperature in 0.25% glutaraldehyde in PBS and rinsed
three times in PBS. Cell surface antigens were detected by immunoperoxidase labeling using the avidin-biotin-peroxidase detection system." Cells were then counterstained with Mayer's hematoxylin and
observed by light microscopy.
Enzyme treatments. Enzymes were dissolved in RPM1 and added
to cells at varying concentrations. Digestion was routinely performed
for 60 minutes at 37°C and stopped by addition of FCS to a final
concentration of 20%, followed by a further 20-minute incubation
at 37°C. Cells were washed three times in the assay medium before
binding assays.
QIU ET AL
Table 1. Purification of EEN From Mouse Bone Marrow
Stage of Purification
marrow
bone
Total
Myeloid-depleted
cells
Percall-purified
Eb
Percall-purified EEN
Nucleated Cell Yield
2.1 x 108
9 X 107
6.5 X 10'
5.6 x 10'
Bone marrow cells were flushed from both femora and tibiae from
seven Balb/c male mice. Resultsare from a representative experiment
of more than 50 performed.
t- SD) condensed, pyknotic
nuclei that were similar morphologically to extruded nuclei obtained in other studies using
fresh hematopoietic tissue.'.3.5.'".'' However,thesenuclei
were of two types. In a typical experiment, 63% had a clear
rim of plasma membrane and representedauthentic EEN,
and the remaining 37% were bare nuclei, lacking a plasma
membrane(Fig 2). It islikelythatthebarenucleiwere
produced by mechanical disruption during cell separation.
As only the EEN are
involved in recognition and phagocytosis by macrophages under physiologic conditions, it was
important to distinguish them unambiguously from bare nuclei in subsequent bindingassays. To achieve this, we labeled the Eb plasma membranebefore culture with the lipophilic red fluorescent dye DiI, and nuclei were stained after
binding to macrophages with DAPI. The EEN, but not the
RESULTS
bare nuclei, acquire thefluorescent label from the Ebplasma
membrane during enucleation. In comparitive experiments,
Isolation of EEN. To obtainand isolate EEN, our aplabeling Eb with DiI had no effect on thequantity and quality
proach was to (1) purify large numbers of Eb from normal
mouse bone marrow, (2) culture them
in 20% FCS and eryth- of EEN obtained after cultivation in erythropoietin (data not
shown).This isinkeepingwith
other reportsthatDiIis
ropoietin to stimulate the terminal stages of differentiation
biologically inert after membrane labeling.*' Approximately
and enucleation, and (3) separate the relatively dense EEN
70% of the nuclei exhibited surface fluorescence after Percoll
from the remaining Eb in the cultures by density gradient
density purification. This is in agreement with data obtained
centrifugation. Immunomagneticdepletion with aMACS apby electron microscopy, in which 63% of nuclei had intact
paratus was found to be efficient
an
and rapid way of removplasma membranes (see above).
ing myeloid cells from large numbers (greater than 10') of
EEN bind avidly to RBMM. The macrophage population
bone marrow cells. The enriched Eb were furtherpurified to
that naturally binds and phagocytoses EEN is the resident
greater than 90% by density gradient centrifugation, during
macrophage network of the bone marrow. These cellswere,
which themoredense
B lymphocytes, reticulocytes, and
therefore, used in all experiments to characterize the nature
extrudednuclei were removed. Optimal numbers of EEN
of the interaction with EEN. RBMM are isolated as clusters
were obtainedaftercultivationof
Eb for 24 hours in the
of
containingacentral
macrophageand
largenumbers
presence of 1.5 U/mL of erythropoietin. Under these condiattached hematopoietic cells." After attachment of clusters
tions, approximately 10% of Eb underwent enucleation, proto glass and stripping of the attached cells, approximately
ducing nuclei and reticulocytes. The final step in the purifi50% oftheadherent cells are well-spreadstromalmacrocation was the separation of the newly generated EEN from
phages. When these cells were incubated with EEN suspenEb by density gradientcentrifugation.This resultedina
sions, EEN bound rapidly to the majority of RBMM, reachfraction that contained approximately 80% free nuclei that,
ing maximal levels of binding by 45 minutes of incubation.
by light microscopy,were small andcondensed, staining
The presence of a red rim of surface fluorescence, together
intensely with Giemsa. The remaining approximately 20%
witha condensed nuclearmorphologyvisualized
by blue
of cells were contaminating reticulocytes and erythroblasts.
DAPI fluorescence, allowed rapididentificationofbound
The overall cell yields from a typical experiment of more
EEN by fluorescencemicroscopy(Fig3).DiI-labeling
of
than 50 performed are shown in Table 1 .
EEN had no effect the
on level of binding, and EEN prepared
Previous studies with freshly prepared bone marrow susin this way were used routinely in all assays. Because of the
pensionshave shownthatEb nuclei are extrudedwitha
existence of two types of nuclei in EEN preparations, we
narrow rim ofplasmamembrane."
It was,therefore,imquantified the types of nuclei bound to RBMM and found
portant to determine ultrastructurally whether the purified
that an average of 83% i 8% (mean i SD) were authentic
nuclei obtained by our procedure were surrounded bya limEEN. The level and specificity of EEN binding were highly
iting plasma membrane (Figs 1 and 2). By electron microsreproducible. In more than 40 experiments, approximately
copy, the final EEN fraction contained 70.9% i 6.3% (mean
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
ECOGNITION MACROPHAGE
OF EXTRUDED NUCLEI
1633
Fig 1. Purification and culture of
erythroblasts.
Electron
micrographs
showing purified Eb and an Eb in the
processof extruding its nucleus.Eb
were purified from normal mouse
bone marrow by depletion of myeloid
cells and density gradient centrifugation. Enucleation of purified Eb was
stimulated by cultivation in erythropoietin. The purified Eb (A and B) comprise a homogeneous population with
a few smallmitochondriaandsome
vesicular elements in the cytoplasm.
(C) Nuclear extrusionin a single erythroblast; the nucleus gradually moves
toward the cell surface and becomes
extruded together with a thin rim of
cytoplasm.
80% ofRBMMbound
3 to S0 EEN, whereas no binding
was seen to anyother adherentcells.including
immature
monocytes and neutrophils,fibroblastoid cells, and osteoclasts. This suggests that EENR is a macrophage-restricted
receptor.
RRMM IISC CI novel rccognirion nlechcrr~isrnt o h i r d EEN.
To determine whether the EENR
is a surface protein. RBMM
werepretreated with trypsin.BindingofEEN
to trypsintreated RBMM was inhibited in adose-dependentfashion
(Fig 4). In contrast, when EEN preparations were pretreated
with trypsin, we did not see any inhibitory effect, even at
the highest concentration tested (Fig 4). These results indicate that EENR is a trypsin-sensitive cell surface protein that
recognizes a trypsin-resistant ligand on EEN.
Binding of EEN to RBMM was independent of temperature, occurring equally at 4°C or 37°C. and was unaffccted
in the presence of sodium azide, a metabolic inhibitor, and
cytochalasin B. which disrupts actin microfilaments of the
cytoskeleton. Binding was also independent of divalent cations, being unaltered in calcium- and magnesium-free PBS
containing 0.1 mmol/L EDTA (Fig SA). EENadhesion to
RBMM is, therefore, not dependent on the EbR. a receptor
that mediates divalent cation-dependent adhesion
of Eb to
stromal macrophages (Fig SA). This also indicates that integrins.C-type Icctins. and other divalentcation-dependent
mechanismsare not required for EEN binding to macrophages.
Sialoadhesin was first identified on RBMM as a divalent
cation-independentreceptor that bindsunopsonized sheep
RBC via recognition of sialylated glycoconjugates.” EENR
is distinctfromsialoadhesinbecause
SER-4. a blocking
MoAb against sialoadhesin. had no inhibitory effect on binding of EEN at 20 pg/mL, while it was able to completely
block binding of sheep RBC to RBMM at the same concentration (Fig SB).In addition. pretreatment of EEN with sialidase had no effect on binding to RBMM. whereas parallel
treatment of sheep RBCeliminated binding (data not shown).
Some macrophage populations express the scavenger receptor. which is capable of recognizing a wide array of differcnt ligands.”’ Experiments with an inhibitory anti-mouse
macrophage scavengerreceptor MoAb. 2F8. showed that the
bindingofEEN to RRMMisunlikely to be mediated by
this receptor (Table 2 ) . This conclusion was consistent with
the failure of RBMM to endocytose Dil-labeled acetylatedlow-densitylipoprotein (LDL). aligand for the scavenger
receptor (data not shown). Finally. a variety of other MoAbs
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1634
A
QIU ET AL
.-...
-
5 00nm
that bind mouse macrophages were found to have no effect
on hintling of EEN t o R R M M (Tablc 2).
To explore thepossibility that recognitionof EEN involveslectin-likeinteractions.
we examined the effect of
PHASECONTRAST
Fig 2. Electron micrographs of purified EEN. EENwere obtained by cultivation of Eb in thepresence of erythropoietin
followed
by
density gradient
centrifugation. The purified EEN show
condensed nuclei, some of which are
bare nuclei lacking an intact membrane
(A and B), and others have intact plasma
membranes and represent authentic
EEN(C and D). Nuclear envelopes are
identified in authentic EEN (arrows).
variousmonosaccharidcs
(Tablc 2). Thesc included the
amino sugars D-glucosnminc. D-galactosamine. and D-mannosrtminc. which have previously bccn shown to block macrophage rccognition of apoptotic neutrophils."' Nonc of the
DAPl
Fig 3. Binding of EEN t o RBMM. Eb were labeled with Diland cultured with erythropoietin, and the EEN were purified and incubated with
RBMM for 45 minutes at 37°C. Slides were washed and fixed in paraformaldehyde, and nuclei were labeled with DAPI. The photomicrographs
(A-C) show a single well-spread RBMM t o which nuclei are bound. DAPl specifically stains the nuclei, most of whichare small and condensed
(B). Staining with Dil (C) showed which of the nuclei had retained plasma membrane and, therefore, represent EEN. Bar represents 10 pm.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
MACROPHAGERECOGNITION OF EXTRUDEDNUCLEI
1635
- I
I
I
I
I
I
0
20
40
60
80
I
100 120
Trypsin concentration(pg/ml)
Fig 4. Effect of trypsin pretreatment of RBMM and EEN. RBMM
( 0 )or EEN (m) were treated with various concentrations of trypsin
for 1 hour at37°C. washed, and used in binding assays. The receptor
expressed by RBMM forEEN is trypsin-labile, whereas the ligand on
EEN is trypsin-resistant. The values show means 21 SD of data
pooled from threeexperiments.
sugars tested had any effect on EEN binding, even at concentrations up to 100 mmol/L.
Macrophages express a receptor for phosphatidylserine
thatis involved in recognition of certain tumor cells and
apoptotic thymocytes.3'.3'This receptor has not been characterized at a molecular level, but it can be blocked byglycerophosphoryl-L-serine and phospho-L-serine.32 Use of these
compounds at concentrations up to 10 mmol/L had no effect
on binding of EEN to RBMM (Table 2), again suggesting
that the EENR is distinct from previously described macrophage receptors.
Distrilx~tionqf EENR on various macrophage populations. A variety of different macrophage populations were
isolated to determine their expression of EENR. As shown
in Table 3,65% of F4/80-positive adherent cells from spleen
showed positive binding; this was most evident on the large,
well-spread macrophages compared with smaller, immature
cells. With mesenteric lymph node macrophages, binding
was again only observed on the large, well-spread macrophages, which constituted approximately 20% of the F4/80labeled cells. The most striking binding was seen with resident peritoneal macrophages, greater than 90% of which
showed intense rosetting, even higher thanthatseenwith
RBMM (Table 3). In contrast, alveolar macrophages showed
no binding whatsoever. The binding of EEN to peritoneal
cavity macrophages was independent of divalent cations and
was identical to that of RBMM by all criteria tested (data
not shown). These results suggest that EENR is expressed
differentially by both stromal tissue macrophages and peritoneal cavity macrophages. We also investigated the expression of EENR on the macrophage-like cell lines Mm l , RAW
264.8. and P388DI andthe fibroblastic cell line NIH3T3.
Binding of EEN was observed to all macrophage-like cell
lines but was restricted to the large, firmly adherent, and
well-spread population. Virtually no binding of EEN was
observed with NIH3T3 cells.
Phorbol ntyristare acetate ( P M A ) irldrtces mncropl~ages
to pkagocyrose EEN, To determine whether the binding of
EEN to macrophages was accompanied by phagocytosis,
bound EEN were removed from the macrophages by treatment with trypsin and gentle pipetting. Under the usual conditions for binding assays (45 minutes, no serum), fewer
than 10% of RBMM were seen to phagocytose EEN. This
low level of phagocytosis was increased to 20% in the presence of IO% serum, but higher levels were not observed,
even after 24 hours of culture. The failure ofRBMMto
phagocytose EEN was not due to a general defect in phagocytic activity, as large numbers of unopsonized zymosan
particles were phagocytosed by greater than 90% of the macrophages when incubated under identical conditions (data
not shown).
Previous studies on complement receptors have shown
A
DMEM
PBS + EDTA
a,
0
cu
80
A
0,
.G
U
K
.-
60
LI
40
5a
20
r
S
0
+ SER-4
mAb
Fig 5. Binding of EEN t o RBMM is notdependent on EbR or sialoadhesin. Binding assays of EEN (m) and Eb (B)t o RBMM (A) or EEN
( W ) and Sheep RBC (B) t o RBMM (B). (A) Binding of EEN occurs
equally in the presence or absence of divalentcations, whereas binding ofEb is greatlyreduced in their absence. (B) Binding of Sheep RBC
t o RBMM via sialoadhesin iscompletely inhibited by preincubation
of
the RBMM with MoAbSER-4 (20 pglrnLI, whereas this has no effect
on EEN binding. The values show means 21 SD of data pooled from
three experiments.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
QIU ET AL
1636
Table 2. Effect of Various Reagents on Binding of EEN t o RBMM
Reagent
MoAb/Specificity
2FWscavenger receptor
108.3
S7 c S11 + S18/leukosialin (CD43)
FA1l/macrosialin
112.1
(CD68)
5C6/Mac-1
(CD1
1b)
F4/80/macrophage antigen
ER.HR3/macrophage antigen
Monosaccharide
D-glucose
D-galactose
D-fucose
D-glucosamine
D-galactosamine
D-mannosamine
N-acetyl-D-glucosamine
N-acetyl-D-galactosamine
N-acetyl neuraminic acid
Others
Glycerophosphoryl-L-serine
Phospho-L-serine
100.6
100.8
97.8
88.0
91.1
99.4
99.3
97.2
97.7
2 9.7
2 3.2
2 2.4
2 18.7
2 11.7
2 22.3
2 5.5
2 6.2
2 0.7
97.7 -c 2.5
98.8 t 5.3
able to trigger phagocytosis in unstimulated
macrophage^.^'.'" Therefore. we performed experiments to
determine whether PMA could stimulate uptake ofEEN
(Figs 6 and 7). WhenRBMM were pretreated withPMA
for 30 minutes before the binding assay, a dose-dependent
increase in the percentage of RBMM that ingested EEN was
seen (Fig 6). An effect of PMA was observed at the lowest
concentration tested. IO ng/mL, and at 1 &mL. the highest
concentration, approximately 70% of macrophages contained 1 to 10 EEN (Figs 6 and 7). The ability of PMA to
induce phagocytosis of EEN was not enhanced when macrophages were pretreated for various periods (data not shown).
>Attemptsto stimulate phagocytosis by adherence of macrothatPMAis
Table 3. Binding of EEN t o Tissue Macrophages and
Macrophage-Like Cell Lines
7.4
3T3
Bone marrow
Spleen
Lymph node
Peritoneal cavity
Alveolar
RAW 264.7
Mm1
P388 D l
NIH
T
97.9 2 1
2 5
t 2.3
100.2 2 2
98.9 2 4
98.2 t 5.2
Binding assays were performed in the presence of MoAbs at saturating concentrations, monosaccharides at100
mmol/L, glycerophosphoryl-L-serine at 10 mmol/L, and phospho-L-serine at 10 mmoll
L. Values show mean % of control 2 1 SD of data pooled from two to
three independent experiments. Control values ranged from 72% to
88% RBMM binding of 2 3 EEN.
Macrophage Population or Cell Line
80
% Control Binding
% Cells Binding 2 3 EEN
78.6
64.8
19.0
89.8
0
2 5.3
2 9.7
-c 10.3
2 4.3
2 0
2 0.7
3.7 2 2.3
21.0 2 8.3
0.25 2 0.35
Tissue macrophages were identified by immunohistochemistryusing MoAb F4/80, and only these cells were included in the analysis.
FPl8O-negative cells did not bind EEN. Values show mean 2 1 SD of
data pooled from three experiments.
0
0
10
100
1000
PMA concentration (ng/mI)
Fig 6. Effect of PMA on the inductionof phagocytosis of EEN by
macrophages. Binding assays were performed in DMEM either with
or without FCS in the presence of the indicated amounts of PMA.
Data show the means 21 SD from t w o experiments.
phages to fibronectin-coated glass were unsuccessful (data
not shown).
DISCUSSION
In this study we have developed a method to produce and
isolate EEN in vitro and have characterized the nature of
their interaction with RBMM. Our observations provide evidence for the existence of a novel divalent cation-independent macrophage receptor that binds avidly to EEN and.
under appropriate circumstances, can lead to their phagocytosis. We suggest that this receptor is involved in recognition
and phagocytosis of EEN within hematopoietic tissues.
As the nuclei extruded from Eb are rapidly digested by
macrophages, it is difficult to isolate EEN directly from hematopoietic tissues. However, previous work3 demonstrated
that when Eb were cultured in the presence of a physiologic
concentration of erythropoietin, they could undergo terminal
differentiation to yield nuclei and reticulocytes in a manner
very similar to thatseen in vivo. In agreement withthat
study, we obtained and isolated a workable number of EEN
from erythropoietin-stimulated Eb in vitro. Because mechanical disruption of membranes may occur during cell separation, Eh nuclei may be released from lysed cells in a different
manner from the naturally occurring enucleation process,
in which nuclei are extruded with an intact rimof plasma
membrane. This distinction was observed in the present
study during ultrastructural analysis of isolated Eb nuclei in
which two types of free nuclei could be distinguished. EEN
with an intact plasma membrane were contaminated by bare
nuclei without a plasma membrane. To distinguish extruded
from bare nuclei in subsequent experiments. we labeled the
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
MACROPHAGERECOGNITION OF EXTRUDEDNUCLEI
1637
c.
.
"
Fig 7. Phagocytosisof €EN
by RBMM stimulated with PMA.
Binding assays were performed
in DMEM plus 10% FCS in the
absence (AI orpresence (B) of
100 nmol/L PMA. After 45 min€EN were reutes,
adherent
moved by trypsinization and pipetting and were examined by
phase-contrastmicroscopy.Bar
represents 10 pm.
Eb plasma membrane with a fluorescent dye, DiI, before
culture. The EEN could then be identified by the presence
of a weak but clear rim of red DiI fluorescence.
The susceptibility of EENRto treatment withtrypsin
shows that it is a cell-surface protein. However, several lines
of evidence indicate thatthe binding ofEEN to RBMM
is not mediated by a previously characterized macrophage
receptor. The pattern of expression of EENR is distinct from
both EbR and sialoadhesin, which are expressed on the majority of stromal macrophages but not on normal peritoneal
macrophages. In addition. EbR is a divalent cation-dependent receptor, and sialoadhesin-mediated binding is sensitive
to sialidase treatment of ligand-bearing cells.''"5~"~'sIt was
also shown in the present study that the binding of EEN to
RBMM is not inhibited by a sialoadhesin-specific blocking
MoAb. EENRis distinct from the macrophage scavenger
receptor,"." as evidenced by its specific tissue distribution
andthe failure of a scavenger receptor-specific MoAb to
block its activity.
Previous studies demonstrated a reduced negative charge
on the plasma membrane surrounding the extruded nucleus
compared with that of the reticulocyte.'' It was, therefore,
assumed that this maybe a mechanism by which macrophages recognize EEN as particles destined for phagocytosis.
Our studies suggest, however, that the interaction is not mediated simply by a charge effect, because N-acetyl neuraminic acid and several cationic sugars did not inhibit binding,
even up to a concentration of 100 mmol/L.
Macrophage recognition and ingestion of EEN is a swift.
efficient way of removing effete nuclei from hematopoietic
tissues without the release of potentially toxic substances
that mightdamage neighboring cells. This function is similar
to that of macrophage recognition and removal of apoptotic
cells, a process that is believedto involve atleast three
recognitionpathways.3".z~.'".'x Depending on the cell type,
the surface exposure of thrombospondin, specific carbohydrate, or phosphatidylserine may lead to macrophage recognitionand phagocytosis via the vitronectin receptor and
CD36. lectin-like receptors. and a phosphatidylserine receptor, respectively. As binding to the EENR was divalent cation-independent, integrins like the vitronectin receptor and
C-type lectins such as the mannose receptor are unlikely to
play a dominant role. In addition, amino sugars that block
the thrombospondin-dependent recogniton of apoptotic neutrophils had no effect on EEN binding, nor did phosphoglyceroserine. whichblocksuptake
of apoptotic thymocytes
through the phosphatidylserine receptor.'".'' Taken together,
our results argue that the EENR isa novel macrophage receptor.
Despite the dissimilarities of EENR to previously characterized macrophage receptors involved in recognition of
apoptotic cells. it is possible that EENR is involved in this
process as well as others. such as recognition of senescent
RBC. In the bone marrow, for example. RBMM are believed
to ingest large numbers of apoptosing B cells, but the receptors and ligands involved in their specific recognition have
not yet been characterized. In addition, the high expression
ofEENRon
peritoneal cavity macrophages suggests that
this receptor is involved in functions other than recognition
of EEN.
Previous studies have demonstrated that as EEN are extruded, they are rapidly phagocytosed by macrophages both
in vivo and in long-term bonemarrow cultures.h.i4Some
investigators have estimated thatupto
40 EENmaybe
phagocytosed by a single macrophage in a day.' However.
under our initial experimental conditions, we did not observe
significant RBMM phagocytosis of EEN for culture periods
of up to 24 hours. Control experiments with zymosan showed
that the macrophages used in this study had a normal capacity for phagocytosis. Our original explanation for this disparity wasthat phagocytosis ofEEN in vivo maybemore
complicated, perhaps requiring an extra inducing factor or
specific stromal components that are not present in vitro.
We were. therefore, interested in searching for compounds
or conditions that stimulate phagocytosis of EEN. Previous
studies showed that both C3b and C3b' receptors on human
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
OIU ET AL
1638
monocytes efficiently mediate adherenceof RBC coatedwith
the corresponding ligands but not
do promote their ingestion.
However, these cells become capable
of ingesting C3-coated
RBC after treatment with PMA.33-34
These findings prompted
us to test whether PMA wassimilarly capable of stimulating
the phagocytosis of EEN by RBMM. Indeed, thecapacity of
RBMM to phagocytose EEN increased in a dose-dependent
fashion in the presence of PMA. These findingsestablish
the important principle that RBMM are capableof^ phagocytosing EEN in vitro and that this can be regulated. However,
in the present experiments, it was not clear whetherthe effect
of PMA isspecific forEENRor only reflects ageneral
enhancement of cellular phagocytic capacities. The fact that
only arelativelysmallproportion
of EEN thatbound to
RBMM were phagocytosed, even at high concentrations of
PMA, suggests thatspecific components of the microenviroment may be required for efficient phagocytosis of EEN in
vivo.
The expression of EENR is restricted to mature macrophages; nobinding of EEN to contamining monocytes or
neutrophils in macrophage preparations was observed. The
heterogeneity of EENR distribution between various macrophage populations may reflect the effects of local and systemic regulation on EENRexpression. Due to lackof specific
MoAbs orinhibitors of EbR and EENR, we have
been unable
t o determine whether, in the presence
of divalent cations,
the binding of EEN to RBMM can be mediated by EbR as
well as EENR. Wehave not characterized the EENR ligands
and, therefore, cannot determine whether preexisting ligands
become modified or whether EEN acquire a novel ligand(s)
duringtheenucleation
process. Previousstudiesdemonstrated that glycoconjugates recognized by Con A were enriched in the EEN plasma membrane duringenucleation and
thatthis processwasaccompanied by areduction in the
negative charge on the surface of EEN.5."'." It is possible
that molecular remodeling of the EEN plasma membrane
results in two important events with respect to macrophage
recognition.First, if potentialligands forEENRbecome
concentrated on the EEN plasma membrane during enucleation, this could trigger binding and phagocytosis of EEN.
Second, concomitant depletion of ligands for EbR from the
reticulocyte membrane could promote detachmentof reticulocytes from the RBMM surface.
ACKNOWLEDGMENT
We thank lain Fraser for the gift
of 2F8 MoAb, Pieter Leenen
for providing MoAb ER.HR3, Laurence Turley for the gift of 5C6
MoAb, and Adele Hartnell and Deepa Nath for critical reading of
the manuscript.
REFERENCES
1 , Bessis M, Bricka M: Aspect dynamique des cellules du sang.
Rev Haematol 7:407, 1952
2. Skutelsky E, Danon D: An electron microscopic study of nuclear elimination from the late erythroblast. J Cell Biol33:62S, 1967
3. Koury ST, Koury MJ, Bondurant MC: Morphological changes
in erythroblasts during erythropoietin-induced terminal differentiation in vitro. Exp Hematol 16:7S8, 1988
4. Koury ST, Koury MJ, Bondurant MC: Cytoskeletal distribution
and function during the maturation and enucleation of mammalian
erythroblasts. J Cell Biol 109:3005, 1989
S. GeiduschekJB,SingerSJ:Molecularchanges
in the membranes of mouse erythroid cells accompanying differentiation. Cell
16:146, 1979
6. Bessis M, Mize C, Prenant M: Erythropoiesis: Comparison of
in vivo and in vitro amplification. Blood Cells 4:1S5, 1978
7. Yoffey JM, Yaffe P The phagocytic central reticular cell of
the erythroblastic island in rat bone marrow: Changes in hypoxia
and rebound. J Reticuloend Soc 28:37. 1980
8. BessisM:L'iloterythroblastique,
unite fonctionelle de la
moelle osseuse. Rev Haematol 13:8, 1958
9. Bessis M, Breton-Conus J: Iron metabolism in the bone marrow as seen by electron microscopy. Blood 19:63S, 1962
IO. Skutelsky E, Farquhar M: Variations in distribution of Con
A receptor sites and anionic groups during
red blood cell differentiation in the rat. J Cell Biol 71:218. 1976
11. Skutelsky E, DanonD:Reduction
in surfacecharge as an
explanation of the recognition by macrophages of nuclei expelled
from normoblasts. J Cell Biol 43:8, 1969
12. Morris L, Crocker PR, Gordon S: Murine fetal liver macrophages bind developing erythroblasts by a divalent cation-dependent
hemagglutinin. J Cell Biol 106:649, 1988
13. Crocker PR, Kelm S, Dubois C, Martin B, McWilliam AS,
Shotton DM, Paulson JC, Gordon S: Purification and properties of
sialoadhesin, a sialic acid-binding receptor of murine tissue macrophages. EMBO J 10:1661, 1991
14. Morris L, Crocker PR, FraserI, Hill M, Gordon S : Expression
of adivalentcation-dependenterythroblastadhesionreceptor
by
stromal macrophages from murine bone marrow. J Cell Sci 99:141.
1991
I 5. Crocker PR, GordonS: Properties and distribution ofa lectinlike hemagglutinin differentially expressed by murine stromal tissue
macrophages. J Exp Med 164:1862, 1986
16. Crocker PR, Werb Z , Gordon S, Bainton DF: Ultrastructural
localization of a macrophage-restricted sialic acid binding hemagglutinin,
SER,
in macrophage-hematopoietic cell clusters. Blood
76:1131,1990
17. Crocker PR, Mucklow S, Bouckson V, McWilliam A, Willis
A, Gordon S, Milon G, Kelm S, Bradfield P: Sialoadhesin, a macrophagesialicacid-bindingreceptorforhaemopoieticcells
with 17
immunoglobulin-like domains. EMBO J 13:4490, 1994
18. Kyoizumi S, Masuda T: Alectin-likereceptoronmurine
macrophagecelllines,
Mml: Involvement of sialicacid-binding
sites in opsonin-independent phagocytosis for xenogeneic red cells.
J Leukoc Biol 37:289, 1985
19. Austyn JM, GordonS: F4/80, a monoclonal antibody directed
specifically against the mouse macrophage. Eur J Immunol 11:80S,
19x1
20. Rosen H. Gordon S: Monoclonal antibody to the murine type
3 complement receptor inhibits adhesion of myelomonocytic cells
in vitro andinflammatorycellrecruitment
in vivo. J Exp Med
169: 1685, I987
21. CrockerPR,Gordon
S: Mousemacrophagehemagglutinin
(sheep erythrocyte receptor) withspecificity for sialylated glycoconjugates
characterized
by a monoclonal
antibody.
J Exp Med
169:1333, 1989
22. Fraser I, Hughes D, Gordon S: Divalent cation-independent
macrophage adhesion inhibited by monoclonal antibody t o murine
scavenger receptor. Nature 364:343, 1993
23. Kemp JD: Resolution and characterisation of pro-B and preB cell stages in normal mouse bone marrow. J Exp Med 173:1213.
1991
24. Holness CL, Silva RPD, Fawcett J, Gordon S, Simmons DL:
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
MACROPHAGERECOGNITION OF EXTRUDEDNUCLEI
Macrosialin, a mouse macrophage-restricted glycoprotein, is a member of the IampAgp family. J Biol Chem 268:9661, 1993
25. Jong JPD, Nikkels PGJ, Piersma AH, Ploemacher RE: Erythropoiesis and macrophage subsets in medullary and extramedullary
sites, in Rich IN (ed): Molecular and Cellular Aspects of Erythropoietin and Erythropoiesis. Berlin, Germany, Springer Verlag, p 237,
1987
26. Crocker PR, Gordon S: Isolation and characterization of resident stromal macrophages and hematopoietic cell cluster from mouse
bone marrow. J Exp Med 162:993, 1985
27. Honig MG, Hume RI: Fluorescent carbocyanine dyes allow
living neurons of identified origin to be studied in long-term cultures.
J Cell Biol 103:171, 1986
28. Hsu SM, Raine L, Fanger H: The use of avidin-biotin-peroxidase complex (ABC) in immunoperoxidase techniques: A comparison between ABC and unlabelled antibody (PAP) procedures. J Histochem Cytochem 29:577, 1981
29. Krieger M: Molecular flypaper and atherosclerosis: Structure
of the macrophage scavenger receptor. Trends Biochem Sci 17:141,
1992
30. Savill J, Fadok V, Henson P, Haslett C: Phagocyte recognition
of cells undergoing apoptosis. Immunol Today 14:131, 1993
31. Fidler IJ: Recognition and destruction of neoplastic cells by
activated macrophages: Discrimination of altered self. Biochim Biophys Acta 948:151, 1988
1639
32. Fadok VA, Voelker DR, Campbell PA, Cohen JJ, Bratton
DL, Henson PM: Exposure of phosphatidylserine on the surface of
apoptotic lymphocytes triggers specific recognition and removal by
macrophages. J Immunol 148:2207, 1992
33. Newman SL, Musson RA, Henson PM: Development of functional complement receptors during in vitro maturation of human
monocytes into macrophages. J Immunol 1252236, 1980
34. Wright SD, Silverstein SC: Tumor-promoting phorbol esters
stimulate C3b and C3b' receptor-mediated phagocytosis in cultured
human monocytes. J Exp Med 156:1149, 1982
35. Crocker PR, Morris L, Gordon S: Novel cell surface adhesion
receptors involved in interaction between stromal macrophages and
hematopoietic cells. J Cell Sci 9:185, 1988 (suppl)
36. Savill J, Dransfield I, Hogg N, Haslett C: Vitronectin receptormediated phagocytosis of cells undergoing apoptosis. Nature
343:170, 1990
37. Savill J, Hogg N, Ren Y, Haslett C: Thrombospondin cooperates with CD36 and the vitronectin receptor in macrophage recognition of neutrophils undergoing apoptosis. J Clin Invest 90: 1513, 1992
38. Duvall E, Wyllie AH, Moms RG: Macrophage recognition
of cells undergoing programmed cell death (apoptosis). Immunology
56:351, 1985
39. Allen TD, Testa NG: Cellular interactions in erythroblastic
islands in long-term bone marrow cultures, as studied by time-lapse
video. Blood Cells 17:29, 1991
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1995 85: 1630-1639
Extruded erythroblast nuclei are bound and phagocytosed by a novel
macrophage receptor
LB Qiu, H Dickson, N Hajibagheri and PR Crocker
Updated information and services can be found at:
http://www.bloodjournal.org/content/85/6/1630.full.html
Articles on similar topics can be found in the following Blood collections
Information about reproducing this article in parts or in its entirety may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#repub_requests
Information about ordering reprints may be found online at:
http://www.bloodjournal.org/site/misc/rights.xhtml#reprints
Information about subscriptions and ASH membership may be found online at:
http://www.bloodjournal.org/site/subscriptions/index.xhtml
Blood (print ISSN 0006-4971, online ISSN 1528-0020), is published weekly by the American
Society of Hematology, 2021 L St, NW, Suite 900, Washington DC 20036.
Copyright 2011 by The American Society of Hematology; all rights reserved.