From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
The Value of Flow Cytometric Analysis of Platelet Glycoprotein Expression
on CD34+ Cells Measured Under Conditions That Prevent P-SelectinMediated Binding of Platelets
By M. Wouter Dercksen, Iris S.Weimar, Dick J. Richel, Janine Breton-Gorius, William Vainchenker,
lneke C.M. Slaper-Cortenbach, Herbert M. Pinedo, Albert E.G.Kr. von dem Borne, Winald R. Gerritsen,
and C. Ellen van der Schoot
In the present study, we show by
adhesion assays and ultrastructural studiesthat platelets can bind t o CD34’ cells from
human blood and bone marrow and that this interaction
interferes with the accurate detection of endogenously expressed plateletglycoproteins (GPs). The interaction between these cells was found t o b e reversible, dependent
on divalent cations, and mediated by P-selectin. Enzymatic
characterization showed the involvement of sialic acid residues, protein(s) containing 0-linked glycans, and elastasesensitive protein(s1. The demonstration of mRNA for the Pselectin glycoprotein ligand I(PSGL-I) in the CD34+ cells by
polymerase chain reaction (PCR) analysis suggests that this
molecule is present in these cells. Under conditions that
prevent platelet adhesion, asmall but distinctsubpopulation
of CD34+ cells diffusely expressed the platelet GPllb/llla
complex. These cells were visualized by immunochemical
studies. Furthermore, synthesis of mRNA for GPllb and
GPllla by CD34+ cells was shown using PCR analysis. The
semiquantitative PCR results show relatively
higher
amounts of GPllb mRNA than ofPF4 mRNA in CD34+CD41+
cells in comparison with this ratio in platelets. This finding
is a strong indication that the PCR results are not caused
by contaminating adhering platelets. MoAbs against GPla,
GPlba, GPV, P-selectin, and the a-chain of the vitronectin
receptor did not react with CD34’ cells.The number of
CD34+ cells expressing GPllb/llla present in peripheral blood
stem cell (PBSC)transplants was determined and was
correlated with platelet recovery after intensive chemotherapy
in 27 patients. The number of CD34+CD41+cells correlated
significantly better with the time toplatelet recovery after
PBSC transplantation ( r = -.83, P = .04) than did the total
number of CD34+ cells ( r = -.55). Statistical analysis produced a threshold value for rapid platelet recovery of 0.34
x lo6CD34+CD41+cells/kg. This study suggests that ifperformed in the presence of EDTA the flow cytometric measurement of GPllb/llla on CD34+ cells provides the most accurate indication of the platelet reconstitutive capacity of
the PBSC transplant.
0 1995 by The American Societyof Hematology.
S
the cells showed true diffuse membrane and cytoplasmic
positivity on cytospin slides.
The aims of the present study were to evaluate the expression of the different platelet GPs on hematopoietic progenitor
cells by flow cytometry and to investigate whether binding
of platelets can interfere in this assay. Subsets of PB CD34+
cells were quantified, and the numbers of reinfused subsets
of CD34+ cells were correlated with platelet recovery after
intensive chemotherapy to obtain a more accurate indication
of the reconstitutive capacity of a PBSC transplant.
EVERAL STUDIES have shown that flow cytometric
measurement of CD34+ cells can be used to establish
the reconstitutive capacity of a stem cell transp1ant.I4In
these studies, reinfusion of peripheral blood stem cells
(PBSCs) resulted in rapid granulocyte reconstitution in almost all patients but not always in a swift independence of
platelet transfusions. The observation that the time to platelet
recovery varied strongly after reinfusion of small number of
CD34+ cells suggests that an additional marker is required
to more directly measure the megakaryocytic reconstitutive
capacity of a PBSC transplant. The number of megakaryocyte progenitor cells can be determined by the use of the
clonogenic assay for megakaryocytes (colony-forming unitmegakaryocyte [Cm-MK]). However, this assay is laborious, and results are not available until after 14 days. Therefore, the flow cytometric measurement of megakaryocytic
precursor cells might be more practical and usefuL5 This
parameter could be used as an indicator for the platelet reconstitutive capacity of an autologous transplant after high-dose
chemotherapy.
A number of studies have reported on the presence of
platelet antigens on CD34+ cell^.^‘^ In these studies, a variable but, in most studies, predominant reactivity with CD41
and CD61 monoclonal antibodies (MoAbs) was found, ranging from 2% to 90% of the CD34+ cell population. These
antibodies recognize the platelet glycoproteins (GP) IIb and
GPIIIa, respectively. This high level of variation could be
due to some pitfalls. It has been shown that platelets can
adhere to monocytes, neutrophils, eosinophils, basophils,
and subpopulations of T lymphocytes, causing spurious detection of platelet antigens on these cell^.'^"^ Moreover, Betz
et all4 reported that CD41 reactivity on acute myeloid leukemic blasts was caused by adherent platelets or platelet fragments in 85% of the cases, whereas in only 15% ofthe cases
Blood, Vol86, No 10 (November 15). 1995: pp 3771-3782
MATERIALS AND METHODS
MoAbs. The following antibodies were used in this study. IgGl
and IgG2a isotype control antibodies were obtained from the CLB
From the European Cancer Centre, Amsterdam,The Netherlands;
the Central Laboratoryof the Netherlands Red Cross Blood Transfusion Service, Amsterdam.The Netherlands; the Free University Hospital, Amsterdam, The Netherlands; The Netherlands Cancer Institute/Antoni van keuwenhoekhuis, Amsterdam,The Netherlands; the
Medisch Spectrum Twente, Enschede, The Netherlands; INSERM
U.91 H6pital Henri Mondor, Crdteil, France: INSERM U.362 Institut Gustave Roussy, Villejuif; France; and the Academic Medical
Centre, University of Amsterdam, Amsterdam, The Netherlands.
Submitted September 12, 1994; accepted July 18, 1995.
Address reprint requests to C. Ellen van der Schoot, MD, PhD,
Department of Experimental Immunohematology, Central Laboratory of the Netherlands Red Cross Blood Transfusion Service, PO
Box 9190, I006 AD, Amsterdam, The Netherlands.
The publication costsof this article were defrayedin part by page
chargepayment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 1995 by The American Society of Hematology.
0006-4971/95/8610-0017$3.00/0
377 1
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
3772
(Amsterdam, The Netherlands). Fluorescein isothiocyanate (F1TC)labeled HPCA-2 (CD34), phycoerythrin (PE)-labeled rat MoAb
against the ( K ) light chain of mouse Ig, and MoAb Leu-8 (CD62L,
L-selectin) were purchased from Becton Dickinson (San Jose, CA).
MoAbs CLB-thromb/l (C17, CD61, GPIIIa), CLB-thromb/4
(CD49b, VLA-2), CLB-thromb/S (CD62P, P-selectin), CLB-thromb/
6 (CD62P, P-selectin), CLB-thrombl7 (6C9, CD41, GPIIb),
ES12FlI (CD31, GPIIa', PECAM-l), ES4C7 (CD36, GPIV), SW16
(CD42d, GPV), CLB-3B9 (CD15, Le"), CLB-grad2 (CD15, Le"),
and B13.9 (CD66b) were all produced in our laboratory (CLB) and
clustered in the International Workshops on Leukocyte Differentiation Antigens.15.1h Biotin-labeled CD41 (CLB-thrombl7) and a biotin-labeled isotype control are MoAbs produced in our laboratory
(CLB). MoAbs W6/32 (HLA class I), CSLEX-l (CDlSs), MB45
(CD42b, GPIba), Y2.51 (CD61, GPIIIa), and NKI-M7 (CD51) were
obtained from the Fourth International Workshop studies." MoAbs
lmmu-133, Immu-409, 14G3, BI3C5, CD34-9F2, HPCA2, 43A1,
581, 570, 553, 563, MD34.1, MD34.2, MD34.3, QBendlO, 4A1,
9044, and 9049, directed against CD34, were obtained from the Fifth
International Workshop studies."
Patientcharacteristics. The group of27 patients studied (median age, 35 years; range, 19 to 52 years) included 9 patients with
breast cancer, 8 patients with germ cell cancer, 4 patients with Hodgkin's disease, 3 patients with non-Hodgkin's lymphoma, 1 patient
with nasopharynx carcinoma, 1 patient with medulloblastoma, and
l patient with neuroblastoma. The patients were either in their first
chemotherapy-induced (near) complete remission (patients with
breast cancer)'* or in second partial or complete remission (remaining patients). The patients with germ cell cancer underwent a
tandem transplantation procedure with a 5-week interval. All patients
had a World Health Organization (WHO) performance status of 0
or 1, adequate renal and hepatic function (creatinine clearance, 2 5 0
(BM)
mL/min; bilirubin, S 25 pmoVL), and normal bone marrow
functions (white blood cell [WBC] count, 23.5 X IO9& platelets,
r l X~109k).
BM was aspirated from patients undergoing cardiac surgery (Dr L.
Eijsman, Department of Cardiac Surgery, Academic Medical Center,
Amsterdam, The Netherlands) and patients with malignancies without evidence of BM localization of their disease.
All patients gave informed consent, andthe separate protocols
were approved by the Ethical and Scientific Review Committees of
the Netherlands Cancer Institute, the Free University Hospital, and
the Academic Medical Center (Amsterdam, The Netherlands).
Mobilization procedure, P B X harvest, and reinfusion. Hematopoietic progenitor cells were mobilized by chemotherapeutic treatment followed by 300 pg of granulocyte colony-stimulating factor
(G-CSF) administered subcutaneously daily (Filgrastim; Amgen Inc,
Thousand Oaks, CA) until completion of the leukocytapheresis. In
the patients with breast cancer, the mobilizing regimen consisted of
5-fluorouracil (500 mglm'), epirubicin (120 mg/m'), and cyclophosphamide (500 mg/m2) administered on day 1,with G-CSF started
on day 2.18 In the remaining patients, PBSCs were mobilized with
ifosfamide (4 g/m2,day 1) and etoposide (100 mglm*,days 1 through
3) followed by G-CSF onward from day 4.
From day 7 of G-CSF administration, the percentage of CD34'
cells in the PB was determined daily. As soon as the WBC count
exceeded 3.0 X 10y/Land a clear increase in CD34+ cell percentage
was observed, leukocytapheresis procedures were started. Atthe
start of each leukocytapheresis procedure, the number of platelets
hadtobe 2 5 0 X 109/L. The leukocytapheresis was performed as
an outpatient procedure with a continuous-flow blood cell separator
(Fenwal CS3000; Baxter Deutschland GmbH, Munich, Germany).
One leukocytapheresis procedure per day was performed. After each
leukocytapheresis, the number of CD34+ cells was measured. Depending on the yield of CD34' cells, further leukocytaphereses were
DERCKSEN ET AL
planned. In a median of 3 leukocytapheresis procedures (range, 1 to
8 leukocytapheresis procedures) per patient, a median of 6.7 X 10"
CD34' cellskg (range, 1.6 to 42.0 X IO'kg) were procured. After
leukocytapheresis, the cells were cryopreserved in saline, containing
0.1% glucose, 0.38% trisodium citrate, 10% human serum albumin,
and 10% dimethylsulfoxide at a cell concentration of about 50 X
10' mononuclear cells (MNCs)/mL. For cryopreservation. the cell
suspensions were frozen at a controlled rate in a KryolO (Cryothec,
Schagen, The Netherlands). The frozen cells were stored in the vapor
phase of liquid nitrogen until reinfusion.
Patients with nonhematologic malignancies received high-dose
chemotherapy consisting of 1,600 mglm2 carboplatin, 480 mg/mZ
thiotepa, and 6,000 mg/m' cyclophosphamide intravenously, divided
over 4 days (CTC)." Patients with malignant lymphoma received
the BEAM regimen (300 m g h ' carmustine, 800 mg/m2 etoposide,
800 mglm2 cytarabine, and 140 mg/m2 melphalan*"). Forty-eight
hours after the last dose of chemotherapy in the CTC regimen or
24 hours after the last dose of chemotherapy in the BEAM regimen,
the cryopreserved apheresis products were thawed rapidly atthe
bedside and were reinfused via an indwelling subclavian catheter.
After transplantation, all patients received G-CSF at 5 pg/kg/d,
which was started on the day of PBSC transplantation and continued
until the WBC count in the PB exceeded 5 X 1OYLNo significant
differences in the rate of neutrophil or platelet recovery were found
with either high-dose chemotherapy regimens (CTC or BEAM) or
in patients with the different diagnoses (data not shown).
Flow cytometry. The percentage of cells expressing the CD34
antigen was determined in a sample of the leukocytapheresis product
by direct immunofluorescence just before cryopreservation. After
lysis of the erythrocytes with isomolar NH,CI buffer for 10 minutes,
1 X 10' cells were incubated with MoAb CD34-FITC. All incubations were performed at 4°C. After each incubation, the cells were
washed with phosphate-buffered saline (PBS) containing 0.2% (wt/
vol) bovine serum albumin (BSA).
For double-color immunofluorescence analysis, the cells were incubated for 30 minutes with the primary MoAb, followed by incubation with PE-labeled rat-antimouse Ig. An isotype-matched mouse
Ig served as control. Residual binding sites of rat-antimouse Ig were
blocked with a mixture of irrelevant murine MoAbs of IgGl and
IgG2a subclasses. Subsequently, the cells were incubated with MoAb
CD34-FITC. After each incubation, the cells were washed with PBS/
EDTA/BSA.
Flow cytometric analysis was performed with a FACScan flow
cytometer (Becton Dickinson Immunocytometry Systems, San Jose,
CA). Aminimum of 20,000 cells were acquired in list mode. Analysis of the five-dimensional data was performed with Consort 30
software (Becton Dickinson). The percentage of CD34' cells present
was assessed after correction for the percentage of cells reactive
with an isotype control. For the determination of the phenotype of
CD34+ cells, a minimum of 2,500 CD34' cells were analyzed. A
marker was set at the first log decade and the percentage of CD%+
cells coexpressing a specific antigen was assessed after correction
for the precentage of cells reactive with an isotype control. Absolute
numbers ofCD34' cells were calculated by multiplication of the
total amount of nucleated blood cells in the leukocytapheresis product with the percentage of CD34' cells in the total leukocytapheresis
product.
Isolation of cells. Platelet-rich plasma (PRP) was removed from
BM or leukocytapheresis samples by centrifugation at 200g for 20
minutes. The erythrocytes were lysed with isomolar N h c 1 buffer
for 10 minutes at 4°Cand subsequently washed twice with PBS
~ ) (PBSEDTM
containing 5 mmol/L EDTA and 0.2% ( w ~ / v oBSA
BSA). BM MNCs were isolated by density centrifugation Over Ficoll-Paque (Pharmacia, Uppsala, Sweden; density, 1.077 g/mL). To
facilitate further analysis of GPs on PB CD34' cells, we enriched
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
PLATELET GLYCOPROTEIN EXPRESSION
ON CD34‘CELLS
for immature cells by one or two rounds of immunomagnetic depletion of T cells and monocytes. MNCs (1 X 107/mL)containing 0.5%
to 5% CD34+ cells were incubated for 40 minutes at 4°C with a
mixture of CD2 and CD14 in the presence of DNAse (20 UL) and
10 mmol/L MgC12. The cells were washed twice and incubated
with immunomagnetic beads coated with goat-antimouse Ig (ratio
of beads:cells = 3:l; Dynal, Hamburg, Germany). This procedure
did not affect the expression of GPs on CD34’ cells (data not
shown).
Resting platelets were prepared from EDTA-anticoagulated whole
blood by direct fixation in paraformaldehyde (PFA) at a final concentration of 1% (wt/vol) for IO minutes at room temperature. Platelets
were isolated from PRP and washed twice with PBS/EDTA/BSA.
Activated platelets were prepared by treatment with 1 U/mL of
human a-thrombin (Sigma Chemical, St Louis, MO) for 10 minutes
at 37°C. The platelets were then washed once and fixed in PFA.
PB MNCs stained with CD34-FITC were sorted with anATC
3000 (Odam, Wissensoury, France) equipped with a 70-pm nozzle
or with a FACStar (Becton Dickinson) on two parameters, ie, fluorescence intensity and side-angle light scattering. More than 98%
of these sorted cells are positive for CD34 when observed with a
fluorescence microscope.
Celladhesionassay.
The interaction between CD34+ cell and
platelet cell adhesion assays was studied as described before.”
Briefly, fixed MNCs and platelets were immunostained with MoAb
CD34-FITC and MoAb W6/32 (anti-HLA class I), respectively. PElabeled rat-antimouse Ig was used as conjugate for MoAb W6/32.
Thereafter, the cells were resuspended in a buffer containing 25
mmol/L Tris-HC1, pH 7.0, 150 mmoVL NaCl, 0.2% (wt/vol) BSA,
supplemented with either 1 mmoVL CaCI2 and 1 mmoVL MgC12
(incubation buffer) or 5 mmol/L EDTA. Fifty microliters of the
platelet suspension (1 X 108/mL)was mixed with 50 pL of the MNC
suspension (1 X 107/mL), giving a platelet to MNC ratio of 1O:l.
After incubation for at least 30 minutes at room temperature, plateletcell adhesion was quantified by measuring doubly stained cells on a
FACScan cytometer (Becton Dickinson). Antibody inhibition studies
were performed by preincubating of either the platelets or the MNCs
with MoAb in optimal dilution. Then, without washing, platelets
and MNCs were gently mixed, and the FACS analysis was performed
as described above. After treatment with enzymes, the cells were
washed twice with PBS and 0.2%BSA and fixed with PFA. Subsequently, platelets and enzyme-treated MNCs were gently mixed, and
the FACS analysis was performed as described above. An isotypematched control antibody was used toset a threshold (99% of events
below threshold) for positive platelet fluorescence. The percentage
of CD34+ cells that bound platelets was expressed as the percentage
of CD34+ cells that reacted positively with antiplatelet antibodies.
In binding inhibition studies, the percentage of boundactivated platelets to CD34’ cells was set as 100% and inhibition of platelet adhesion was expressed as a percentage of this value. The mean total
platelet fluorescence (TPF) of positively reacting CD34+ cells was
used as an estimate of the number of platelets bound per CD34’
cell.
Treatment of cells with enzymes. MNCs were treated with elastase (Elastin products, Pacific, MO) or trypsin (GIBCO BRL, Grand
Island, NY) as described.” Neuraminidase (Vibrio cholerae; Behringwerke, Marbur, Germany) treatment of cells was performed by
incubating 1 X lo6 cells in 100 pL of PBS and 0.2% BSA with
0.2 U/mL neuraminidase for 60 minutes at 37°C. Treatment with
glycoprotease Pasteurella haemolytica (generous gift form Dr A.
Mellors, Guelph, Ontario, Canada) was performed by incubating 1
X IO6 cells in 100 pL of PBS and 0.2%BSA with 0.05 mg glycoprotease for 60 minutes at 37°C.
Immunochemical staining. For immunocytochemistric analysis
of CD34+ cells, the fluorescence-activated sorted cells were incu-
3113
bated for 30 minutes with a biotin-labeled CD41 (CLB-thromb/7)
MoAb. A biotin-labeled isotype-matched mouse Ig served as control.
Subsequently, the cells were washed and spun on slides for 5 minutes
at 350 rpm in a Shandon cytocentrifuge (Shandon Southern Products
Ltd, Runcom, Cheshire, UK). After fixation with aceton, the cells
were incubated for 1 hour with perioxidase-conjugated avidin-biotin
complex (Brunschwig Chemicals, Amsterdam, The Netherlands) and
diluted in Tris-buffered saline (TBS; pH 7.8) to which 2.5% BSA
and 0.5% Triton X-100 was added. Peroxidase was visualized with
3,3’-diaminobenzidine-tetra-hydrochloride. All the rinsing procedures were performed with TBS. The sections were counterstained
with haematoxilin.
Ultrastructural studies. Sixty thousand fluorescence-activated
sorted cells were washed twice in Hank’s medium at 4% fixed in
1.25% glutaraldehyde in Gey’s buffe8’ for 10 minutes, washed, and
incubated in diaminobenzidine medi~m.2~
Cells were then postfixed
with osmium tetroxide, dehydrated, and embedded in epon. Thin
sections were examined with a Philips CM IO electron microscope
(Philips, Eindhoven, The Netherlands) after lead citrate staining.
RNA isolation and synthesis of cDNA. RNA of CD34+CD41’
cells was isolated from 100,OOO fluorescence-activated sorted cells
essentially as previously described by Ziegler et alZ4and Rappolee
et al.25 Briefly, cells were transferred in 100 pL of lysis buffer (4
m o m guanidine thiocyanate, 25 mmol/L sodium citrate, pH 5.0,
0.5% sodium lauroyl sarcosinate, 0.1 m o m P-mercaptoethanol) containing 20 pg of glycogen (Boehringer Mannheim, Mannheim, Germany) as carrier. The mixture was then thoroughly vortexed and
layered on top of a 100 pL 5.7 moVL CsCl cushion. Samples in 0.3
mL diethylpyrocarbonate (DEFT)-treated polycarbonate centrifuge
tubes were centrifuged for 2.5 hours at 4°C at 55,000 rpm in a TLS55 rotor on a Beckman TL-100 tabletop ultracentrifuge (Beckman
Instruments, Palo Alto, CA). The RNA pellet was resuspended in
DEPC-treated H20, precipitated with 0.1 v01of potassium acetate
and 2.5 v01 100% ethanol, washed in 70% ethanol, dried, and again
resuspended in DEPC-treated HZO.First-strand cDNA was synthesized by incubating RNA at 37°C for 1 hour in a final volume of
IO pL in RT-buffer (50 mmol/L Tris-HC1, pH 8.3, 75 mmol/L KCl,
3 mmol/L MgC12) containing 100 U of Moloney murine leukemia
virus (M-MLV) reverse transcriptase (GIBCOIBRL, Gaithersburg,
MD), 0.5 pg oligo(dT), 15 U RNase inhibitor, 10 mmol/L dithiothreitol (DTT), and 1.0 mmol/L of each deoxyribonucleoside triphosphate (dNTP).
cDNA-polymerase chain reaction (PCR) procedure. Oligonucleotide primers selected for the PCR were used to amplify those
parts ofthe cDNA that contain the sequences corresponding to
GPIIb, GPIIIa, P-selectin glycoprotein ligand 1 (PSGL-l), and Pactin, respectively. The locations of the primers in the nucleotide
sequence wereas follows: GPIIb (sense primer: positions 10251046, sequence 5’ATGGATCCACTGTATATGGAGAGAGCCG3’
in exon 12; antisense primer: positions 1580-1599, sequence 5‘TAGAATTCTTAGGGATAGCTTCTGAGG3’ in exon 16; product
length: 575 bp)? GPIIIa (sense primer: positions 379-401, sequence
S‘TAGAATTCAACAATGAGGTCATCCCTGG3‘in exon 8; antisense primer: positions 657-635, 5’ATAAGCTTCTTACACTCCACACATTCTTTC3’ in exon 10; product length: 279 bp):’ PSGL1 (sense primer: positions 76-96, sequence S‘TCCTGTTGCTGATCCTACTGG3’; antisense primer: positions 416-397, sequence
S’ATAGCTGCTGAATCCGTGG3‘; product length: 341 bp):8
platelet factor 4 (PF4; sense primer: positions 8-31, sequence
S’ATGAGCTCCGCAGCCGGGTTCTGC3’;antisense primer: positions 316-339, sequence S‘AATGCACACACGTAGGCAGCTAGT3’; product length: 332 bp), and P-actin (sense primer: positions
144-163 S‘GTGGGGCGCCCCAGGCACCA3’in exon 2; antisense
primer: positions 683-660 S‘CTCCTTAATGTCACGCACGATTTC3‘ in exon 4; product length: 540 bp).29All PCR primers were
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
DERCKSEN ET AL
3774
synthesized on a DNA synthesizer (Applied Biosystems model 392,
Palo Alto, CA) and purified with oligonucleotide purification cartridges (Applied Biosystems, Foster City, CA). A small portion of the
RT-products (1 pL) was mixed with 1.25 U of Taq DNA polymerase
(GIBCOIBRL), 1 pmol/L of each appropriate primer, 200 pmol/L
of each dNTP in a buffer containing 10 mmol/L Tris-HCI, pH 8.3,
50 mmol/L KC1, 5 mmol/L MgCI2, and 0.01% (wt/vol) BSA in a
50 pL volume. For the PCR reaction of the P-selectin ligand, 2 U
of Taq DNA polymerase from Promega (Madison, WI) and the
buffer recommended by the manufacturer was used. The PCR mixtures were overlaid with mineral oil and amplified for 35 cycles.
The cycling conditions for the PCR reactions were as follows: for
GPIIb (60 seconds at 95"C, 90 seconds at 5 0 T , and 150 seconds at
72°C); for GPIIIa, PF4, P-selectin ligand, and @-actin (60 seconds
at 9 4 T , 60 seconds at 55"C, and 180 seconds at 72°C). PCR products
were analyzed on 6% polyacrylamide gels. Fragments were visualized by UV illumination after ethidium bromide staining. One-hundred basepair ladders were used as markers.
For the semiquantitative PCR, a sample of 10 pL was taken after
21, 24, 27, and 30 cycles of the GPIIb, PF4, and @-actinPCRs. The
samples were denaturated in 0.5 mmol/L NaOH and 0.15 mmol/L
NaCl for at least 10 minutes and neutralized in 0.5 mol/L Tris, pH
8.0, and 0.15 moUL NaC1. Immediately thereafter the samples were
dot-blotted using a broad dot-blot apparatus to Hybond N+ positively
charged nylon membranes (Amersham, Buckinghamshire, UK).
After drying, the DNA was cross-linked to the membranes with UV
light for 5 minutes. The filters were then prehybridized in 6 X SSC
buffer containing 0.5% sodium dodecyl sulfate (SDS), S X Denhardt's, and 100 pg/mL denaturated salmon sperm DNA for at least
1 hour at 60°C. After 1 hour, the mixture was replaced by the
hybridization mix that contains all the ingredients of the prehybridization mix except for the Denhardt's. 32P-radioactively labeled
probes were added to the hybridization mixture. The following
sequences of probes were used: for GPIIb, TAGGATCCTGCAGGAACAAATACACACGCC; for PF4, TCCATGAAGGATGATCTGTGG; and for @-actin, GATGACCCAGATCATGTTTGAGAC. Twenty-five nanograms of probe was labeled with 3 pL
32Py-ATP(300 CUmmol) using the polynucleotide kinase 4 (PNK4;
Boehringer Mannheim, Plaats, Germany) in the PNK buffer that
belongs to the enzyme. After incubation at 37°C for 1 hour, the free
label was separated from the labeled probe using a Sephadex-G25
column. The labeled probe that gets through the sephadex column
after washing with 1X TE buffer was added to the hybridization
mixture. The dot-blots were hybridized overnight at 6 0 T , after
which they were washed two times with 2X SSC containing 0.1%
SDS for 15 minutes at 55°C. The nylon membranes were wrapped
in saran wrap and hybridized probes were detected using a X-OMAT
AR film (Eastman Kodak, Rochester, NY) that were developed after
overnight exposure. To compare the signal of the hybridized PCR
products, the blots were also exposed to Fuji imaging plates (Fuji
Photo Film Co, Fiji, Japan), that were measured by a bio imaging
analyzer (Fujix Bas 2000, Fuji, Japan). Using this method, the blackness can be expressed as a number.
Staristical analysis. For nonnormal distributed values, data were
summarized by means of median and ranges; otherwise, the arithmetic mean and standard deviation were used. Differences were calculated by means of the Mann-Whitney U test. The correlations are
Spearman Rank correlations. Differences were calculated by means
of a z-test. A P value less than LY = .05 was considered significant.
To determine the relative influence of parameters (multivariate analysis), a Cox's proportional hazards model was used. The models
were built in a stepwise procedure.
Thresholds for rapid platelet recovery (arbitrarily set at as the
time to platelet transfusion independence within 14 days after PBSC
transplantation) were defined by the optimum of both sensitivity and
specificity of a tested parameter in the receiver operating characteristic curve.
RESULTS
Flow cytometric assay for adhesion of thrombin-activated
platelets to CD34' cells. A double-color immunofluorescence assay was used to investigate whether platelets can
adhere to CD34' cells. Platelets were thrombin-activated
and PE-labeled and then added to FITC-labeled CD34' cells
in a buffer containing Ca" and Mg2'.'o Without additional
washing, platelet-cell adhesion was quantified using flow
cytometry.
It appeared that nearly half of the BM CD34+ cells bound
thrombin-activated platelets, whereas directly fixed nonactivated platelets bound to only low numbers of CD34' cells
(Table 1). Similar results were obtained with PB CD34'
cells.
The involvement of divalent cations in the interaction of
platelets and CD34' cells was studied by incubation of
MNCs and platelets in the presence of EDTA. Binding of
activated platelets to BM and PB CD34' cells was almost
completely abolished by 5 mmol/L EDTA. A slight reduction
of binding of nonactivated platelets to CD34' cells was also
observed (Table 1).
As an estimate for the number of platelets bound per
CD34' cell, the TPF per CD34+ cell was also determined.
The mean TPF for activated platelets bound to BM CD34'
cells was 5 times as high as for nonactivated platelets. Similar results were obtained with PB CD34' cells (Table 1).
To investigate whether platelet-CD34+ cell interaction is
reversible, activated platelets were bound to CD34+ cells
and, after incubation for 30 minutes, the cells were washed
twice in the presence of 5 mmol/L EDTA. These wash steps
abolished platelet-CD34+ cell interaction and reduced the
percentage of cells with bound platelets and the TPF to
values comparable with those found when MNCs and platelets were incubated in the presence of EDTA (Table 1).
P-selectin-mediated adhesion of thrombin-activated
platelets to CD34' cells. Platelets were preincubated with
MoAbs against P-selectin before incubation with MNCs to
study the involvement of P-selectin in the interaction of
activated platelets and CD34' cells. Antibody CLB-thromb/
6 almost completely blocked adhesion of activated platelets
to CD34' cells from BM or PB, reducing the mean TPF to
6.1 ? 3.6 and 5.7 ? 4.5 (on a linear arbitrary scale), respectively (Table 1). Antibody CLB-thromb/S, which recognizes
a nonfunctional epitope of P-selectin, did not inhibit the
platelet-CD34' cell interaction. The number of bound activated platelets as measured by the TPF was then the same
as in the presence of control MoAbs. Thus, P-selectin is
involved in the platelet-CD34' cell interaction.
Putative CD34'
cell receptors for binding platelets.
MNCs were preincubated with MoAbs or enzymatically
treated to characterize the P-selectin ligand present on
CD34' cells. When MoAbs against CD15 antigen (CLB3B9" and CLB-gran/2") were preincubated with the MNCs,
binding to activated platelets was not inhibited (Table 1).
Similar results were obtained with anti-sialyl Lewis" MoAb
CSLEX- 1 in different concentrations (range, 1.5 to 300 pg/
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
PLATELET GLYCOPROTEIN EXPRESSION ON CD34' CELLS
3775
Table 1. Effect of EDTA and MoAbs Against P-Selectin on Platelet Adhesion to CD34' Cells
BM I n = 5)
% Adhesion ISD)
TPF
Leukocytapheresis Samples I n = 10)
% Inhibition"
% Adhesion ISD)
TPF
% Inhibition
Resting platelets
EDTA(2.4)
Activated platelets
5.2
-
7.0
EDTA
CLB-thromb/G (CD62P)
CLB-thromb/5 (CD62P)
CD15
CSLEX-1 (CD15s) (n = 2)
Control MoAbs
Activated platelets
Two 85
EDTA wash8.1
steps
(4.0)
~
~
8.5 (4.3)
5.6 (3.4)
5.8
5.2
47.9 (9.6)
6.4 (3.4)
6.2 (2.7)
44.5 (11.6)
46.9 (10.4)
50.7
48.3 (8.7)
31.2
5.8
6.1
28.4
29.7
30.8
32.1
0
87
87
7
2
0
0
7.2
8.1 (2.8)
5.6
4.8
51.6 (11.3)
5.2 (3.1)
6.9 (4.5)
49.0 (10.2)
52.8 (9.7)
50.5
50.2 (10.7)
39.3
5.6
5.7
35.7
36.3
35.0
37.2
0
90
87
5
0
2
3
6.1
86
(3.9)
~~
Abbreviation: SD,standard deviation.
* In binding inhibition studies, the percentage of CD34' cells with adherent activated platelets was setas loo%, and inhibition of platelet
adhesion was expressed as a percentage of this value.
mL). Furthermore, the TPF of activated platelets bound to
CD34+ cells was not decreased in the presence of CD15 or
CSLEX- 1.
Vascular CD34 is a ligand for L-selectin3*and could be
a ligand for P-selectin as well. To study the involvement of
CD34 in the interaction of activated platelets and CD34+
cells, MNCs were preincubated witha panel of MoAbs
against all three classes of CD34 epitopes before incubation
with platelets. No blocking effects of these MoAbs was observed on platelet-CD34+ cell interaction (data not shown).
The involvement of carbohydrates was studied further by
enzymatic treatment of the MNCs with neuraminidase from
Vibrio cholerae, with glycoprotease derived from Pasteurella haemolytica, or by treatment with the proteolytic enzymes trypsin or elastase (Table 2). Treatment with neuraminidase, which hydrolyses terminal N-acetylneuraminic
acids, partially inhibited binding of activated platelets to BM
CD34' cells and reduced the number of bound platelets to
CD34+ cells as the TPF decreased by54% +- 8%. This
finding indicates that sialic acid residues are present on the
P-selectin ligand of CD34+ cells. The efficacy of the treatment with neuraminidase was established by the complete
loss of reactivity of anti-sialyl LewisXantibody CSLEX-1
with CD34+ cells and loss of binding of activated platelets
to neutrophils present in the same cell sample as the CD34+
cells (data not shown).
Treatment of BM CD34+ cells with glycoprotease derived
from Pasteurella haemolyticu, which cleaves cell surface
proteins containing 0-linked s ~ g a r s ,almost
~ ~ . ~completely
~
abolished binding of activated platelets (Table 2). Treatment
with the proteolytic enzyme trypsin partially inhibited binding of activated platelets (mean percentage of inhibition,
44% -C 9%), whereas elastase almost completely inhibited
binding to values seen on resting platelets (mean percentage
of inhibition, 88% t 5%). Similar results were obtained
by the enzymatic treatment of PB CD34+ cells (Table 2).
Reactivity of HPCA-2 (reactive with a class I11 epitope of
the CD34 antigen), Leu-8 (L-selectin), and CSLEX-1 (sialyl
Lewis") remained unaltered by treatment with these proteolytic enzymes (data not shown).
To obtain evidence for the synthesis of PSGL-l by CD34+
progenitor cells, messenger ribonucleic acid (mRNA) was isolated from fluorescence-activated cell sorted CD34+ cells and
PCR analysis wasperformed.The analysisof RT-PCR products
showed the expression of mRNA for PSGL-1 (Fig 1).
Platelet GP expression on CD34+ cells from BM and PB.
Table 2. Effect of Enzymatic Treatment on Platelet Adhesion to CD%+ Cells
BM (n = 5)
% Adhesion (SD)
TPF
Leukocytapheresis Samples I n = 10)
% Inhibition*
% Adhesion (SD)
TPF
% Inhibition
18.1
7.5
20.5
5.3
58
80
40
86
Activated platelets
Neuraminidase
5.8 Glycoprotease
(4.2)
Trypsin
Elastase
47.9 (9.6)
24.4 (6.3)
8.3
26.8 (8.3)
(3.7)
5.8 (3.8)
39.3
31.2
16.1
18.3
7.2 4.9
(11.3)
0
49
44
88
51.6
21.7 (5.9)
10.2 (4.7)
31.1 (7.4)
0
Abbreviation: SD, standard deviation.
* In binding inhibition studies, the percentage of CD34+ cells with adherent activated platelets was setas loo%, and inhibition of platelet
adhesion was expressed as a percentage of this value.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
3776
DERCKSEN ET AL
M
1
2
3
4
5
-
575
540-
341
279
Fig 1. PCR products of CD34+CD41+fluorescence-activatedcell
sorted cells analyzedon a 10% polyacrylamidegel stained with ethidium bromide. Lane 1, GPllb; lane 2, GPllla; lane 3, PSLG-1; lane 4, pactin; lane 5, water. M represents the 100-bp marker ladder.
Fig 2. GPllb/llla expression onsorted CD34' cells after removal of
platelets by density centrifugation and washing in EDTA-containing
buffer.
and distinct subset of CD34' cells (2% to 12%) still reacted
with MoAbs against the GPIIb (CD41) and GPIIIa (CD61)
(Fig 2), whereas virtually no staining was seen with MoAbs
recognizing GPIa, GPIba, GPV, VNRa, and P-selectin either in BM or in PB. The expression of platelet GPs was
not significantly different on CD34' cells obtained from BM
or PB.
Immunochemical and ultrastructural studies. Immunochemical and ultrastructural studies were performed to confirm that, when EDTA is omitted from the buffers, platelets
attached to CD34' cells could attribute to the detection of
platelet GPs on CD34' cells from PB. When cytocentrifuge
preparations of fluorescence-activated sortedCD34' cells,
stained in the absence of EDTA, were labeled with CD41,
some diffusely stained CD34' cells werevisualized (Fig
3A), but, in addition to these cells, several CD34' cells
Expression of GPs on CD34' cells from BM and PB was
examined in a double-color fluorescence assay with a panel
of MoAbs recognizing platelet membrane GPs Ia, Iba, IIa',
IIIa, IIblIIIa, IV, and V (Table 3). GP expression was measured under two conditions, ie, either in the absence of
EDTA or after removal of platelets by density centrifugation
and washing in EDTA-containing buffer. In the absence of
EDTA, there was significant reactivity of BM-andPBCD34' cells with the platelet-reactive MoAbs against GPIIa'
(CD31), GPIIb (CD41), GPIIIa (CD61), and GPIV (CD36),
whereas reactivity with MoAbs against GPIa, GPIba,
VNRa, and P-selectin was weak. Binding of all these MoAbs
was reduced whenantigen expression was measured after the
removal of platelets by density centrifugation and washing in
EDTA-containing buffer. Under these conditions, a small
Table 3. Expression of Platelet Antigens on PB and BM CD34* Cells in the Absence or Presence of EDTA
BM (n = 5)
GP
CD
MoAb
la
Iba
Ila'
llbfllla
llla
llla
IV
V
VNRa
P-selectin
49b
42b
31
41
61
61
36
42d
51
62P
CLB-thromb/4
MB45
ES12F11
CLB-thrombn
CLB-thrombfl
Y2.51
ES4C7
SW16
NKI-M7
CLB-thromb/G
EDTA
-
2.1 (0.1-5.3)
2.5 (1.4-7.2)
89.6 (86.0-97.9)
14.3 (6.7-18.4)
7.2 (3.8-11.8)
6.8 (2.9-11.7)
12.2 (6.8-15.2)
1.2 (0.1-2.1)
1.9 (0.9-5.1)
5.7 (2.5-8.6)
Leukocytapheresis Samples In = 5)
EDTA
0.6
1.3
86.1
6.8
5.9
4.9
11.5
0.3
1.1
0.9
+
(0.1-1.5)
(0.2-2.6)
(84.3-94.2)
(3.7-12.7)
(3.3-9.3)
(1.9-8.6)
(5.7-13.3)
(0.1-0.5)
(0.1-2.3)
(0.1-1.2)
EDTA
-
4.2 (1.5-6.3)
6.2 (2.7-9.5)
94.5 (84.7-95.4)
19.5 (8.3-24.5)
14.2 (4.7-18.4)
11.8 (2.3-13.3)
16.7 (12.0-19.9)
4.5 (0.9-7.7)
2.4 (1.4-5.1)
7.9 (4.9-12.4)
EDTA +
1.5 (1.2-3.0)
1.9 (0.2-7.5)
91.3 (85.6-94.2)
7.3 (4.0-12.1)
6.4 (2.1-10.7)
4.9 (2.1-9.3)
15.0 (12.1-18.1)
1.1(0.1-2.7)
1.4 (0.4-2.4)
0.8 (0.1-1.2)
Values are the median percentage of cells expressing platelet antigen as determined in five experiments, with the range of platelet antigen
expression in parentheses.
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
SSION
GLYCOPROTEIN
PLATELET
CELLS
3777
ON CD34'
'l
t
C
I
A
Fig 3. Imrnuno-peroxidase staining of cytocentrifuge preparations of fluorescence-activated sorted CD34' cells stained in the absence of
EDTA (A) or under conditions that prevent platelet
adhesion (B). Thrombin-activated platelets incubated
with CD34' cells in a buffer containing
either Ca" and Mgz+(C) or EDTA ID). All preparations werelabeled with biotin-labeled CD41 MoAb and were
counterstained with hematoxilin.
In addition t o strongly stained CD34' cells, some CD34' cells with a patchy localized staining were identified (A, arrows).
with a patchy localized staining were identified, strongly
suggesting the appearance of platelets and platelet particles
adhering to the cell membrane. Similar cells were found
when thrombin-activated platelets were added to the sorted
CD34' cells (Fig 3C). In the CD34' cell preparation, labeled
in the presence of EDTA after removal of platelets by density
centrifugation, only diffusely stained cells and no cells with
a patchy appearance were detected (Fig 3B). In addition, the
interaction between thrombin-activated platelets and sorted
CD34' cells was blocked by EDTA (Fig 3D).
Electron microscopy on CD34+CD4If sorted cells,
stained in the absence of EDTA, showed numerous platelet
membrane fragments adhering to cell membrane. Large
nucleoli were observed in the nuclei of these cells and
all granules showed strong peroxidase staining. Therefore,
these cells were identified as promonocytes (Fig 4A). In
contrast, when CD41-labeling of CD34' cells was performed under conditions that prevent platelet adherence,
no membrane fragments attached to CD34+CD41' sorted
cells were detected. Megakaryocytic cells were identified
in this fraction by platelet-peroxidase staining of the endoplasmatic reticulum and of the area around the nuclear envelope (Fig 4B). Platelet organelles were not observed in
these cells. In addition to these promegakaryoblasts. pro-
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
DERCKSENET AL
3778
Fig
4.
(A) Electron
microscopy
of a promonocyte from
CD34*GPllb/llla+sorted cells (without EDTA). A big nucleolus (Nu) is
present in the nucleus. Strong peroxidase activity is present in all
granules. Note thepresence of numerous membrane fragments that
adhere to thepromonocytemembrane (arrows). (Original magnification x 13,300.) (B) Electron microscopy ofa megakaryocyteprogenitor in the CD34'GPllb/llla' cell fraction (with EDTA). This cell, which
expresses peroxidaseactivity (arrows) around the nuclear envelope
and in the endoplasmic reticulum, has anirregular nucleus,few mitochondria, and noplatelet organelle. (Originalmagnification x 17,400.)
monocytes with small peroxidase-positive granules were
still found to be present in this fraction. However, all of
these cells were free of attached membranes or membrane
fragments of platelets.
Detection of GPIIb and GPIIIa by PCR. To obtain evidence for the synthesis of platelet GPs by progenitor cells,
PCR analysis was performed on mRNAisolated
from
CD34'CD41+ fluorescence-activated cell sorted cells that
were stained under conditions that preventplatelet adherence
to CD34' cells. The analysis of RT-PCR products showed
the expression of mRNA for GPIIb and GPIIIa (Fig l ) . The
same products were detected in CD34'CD41- fluorescenceactivated cell sorted cells (data not shown). To exclude the
possibility that the expression of GPIIb and GPIIIa was
mainly due to platelets adhering to the CD34TD4 1 sorted
cells, the mRNA expression of GPIlb and the platelet-spe+
cific protein PF4 were compared in a semiquantitative PCR
assay by measuring the obtained PCR products after 2 I , 24,
27, and 30 PCR cycles in platelets and sorted CD34TD41'
cells (Fig 5). In CD34'CD4Ic cells, relatively more GPIIb
mRNA was presentthanPF4mRNA
in comparison with
the expression in platelets, eg, at 24 cycles the ratio of GPIlb/
PF4 PCR products in CD34TD41' cells as measured by
fosfoimaging analysis was 2.6, whereas in platelets this ratio
was 0.1. In all cell types, similar amounts of &actin mRNA
were present (data not shown). These results indicate that
mRNA for GPIIb is mainly synthesized by CD34' progenitor cells. Also, in CD34TD41- cells, some GPIIb mRNA
expression was detected, although at a lower level, whereas
PF4 mRNA could not be detected.
Correlation with platelet recovery. After high-dose chemotherapy, a median number of 6.0 X IO" CD34' cellskg
(range, 1.6 to 39.4 X I0"kg) was reinfused in 27 patients.
This resulted in a rapid platelet recovery ( Sl4 days) to
platelet transfusion independence (defined as the platelet
count remaining 2 2 0 X 10y/Lwithout platelet transfusions)
in 15 of 27 patients (n = 27; median, 14 days; range, 7 to
37 days).
The expression of CD41 and CD61 antigens on CD34'
cells was measured in the presence of EDTA on all 92 leukocytapheresis samples from these patients. Under these conditions, a median of 6.3% of the CD34' cells expressed the
CD41 antigen (range, 0% to 18.1%). whereas the CD61
antigen was detected on a median of 3.5% of CD34' cells
(range, 0%to 10.3%). These percentages were used tocalculate the absolute number of CD34' cell subsets per kilogram
of body weight that were reinfused into a patient.
The correlation betweeen the total number of CD34' cells
per killogram and the time to platelet transfusion independence was r = - S 5 (95% confidence interval [CI], -0.68
to -0.34). When subsets of CD34' cells, defined by the
expression of the megakaryocytic lineage-associated antigens CD41 or CD61, were correlated with the time to platelet
recovery, both subsets correlated better (CD34TD41' cells,
r = -.S3 [95% CI, -0.89 to -0.731; CD34TD61' cells, r
= -.78 [95% CI, -0.85 to -0.651) than did the total number
of CD34' cells. This difference was significant only for the
subset of CD34' cells defined by CD41 ( P = .04),but failed
to reach significance for the subset defined by the expression
of CD61 ( P = .09). Using a stepwise procedure in a Coxregression analysis with the numbers ofCD34'
cells,
CD34TD41' cells, and CD34TD61' cells, respectively,
as covariates, only the number of CD34TD41' cells was a
significant covariable (p = .0012, SE = .003,P = .0005).
To assess a threshold for rapid platelet recovery the optimal sensitivity and specificity of the number of reinfused
CD34' cells per killogram of body weight was calculated.
The threshold of CD41-expressing CD34' cells for rapid
platelet recovery was calculated be
to
0.34 X IO"
CD34'CD41+ cellskg. In the 17 patients who received more
than the threshold of 0.34 X IO6 CD34'CD41+ cellskg, the
time to platelet recovery was significantly shorter (n = 17;
median, 11 days; range, 7 to 16 days) as compared with the
time to platelet recovery of the patients who received fewer
CD34TD41' cells (n = IO; median, 20 days; range, 13 to
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
SSION
GLYCOPROTEIN
PLATELET
CELLS
ON CD34'
3779
GP l l b
PF4
1
2
3
Fig 5. Dot-blots of PCR productsof GPllb and PF4
hybridizedto '*P-labeled probes as detected after 21,
24, 27, and 30 cycles. In all three cell types similar
amounts of p-actin PCR products were detected
after 24,27, and 30 cycles,whereas no expression of
p-actin was detected after 21 cycles (data not
shown). Lane
1,
CD34+CD41- cells; lane 2,
CD34'CD41' cells; lane 3, platelets; lane 4, H,O.
4
2 42 1
37days; P < . O O O I ; Fig 6). None of these I O patients had
recovered within 2 weeks.
DISCUSSION
The expression of platelet GPs on CD34' cells was analyzed to develop a simple flow cytometric assay for the
number of platelet precursor cells. This study shows that the
detection of endogenously expressed platelet GPs can be
hampered by the adherence of activated platelets to CD34'
cells. Firstly, we observed that binding of the platelet-reactive MoAbs was reduced when antigen expression was measured after the removal of platelets by density centrifugation
and washing in EDTA-containing medium. Secondly, both
immunochemical and ultrastructural studies showed the
presence of platelet membranes and membrane fragments
attached to CD34' cells when EDTA is omitted from the
40
I
*I
r=-0.83
""""""""-"""-
15t"-+*-;.;
- I
t
lo
5 '
0
'
.m
-
*t
*I
; * *.
I
1
i
2
I
3
CD34+CD41+ cells/kg (x 1 06)
Fig 6. Correlation between reinfusionof CD34TD41' cellsper
killogram and time to platelet transfusion independence. Calculated
threshold of reinfused CD34+CD41+cells per killogram (0.34 x 10'
CD34'CD41' cells/kg) for rapid platelet recovery and the time for
rapid recovery are depicted as dashed lines.
27
30
21
24
27
30
buffers. Most of these cells were identified as promonocytes.
In contrast, when CD41-labeling of CD34' cells was performed after removal of platelets by density centrifugation
and washing in EDTA-containing buffer, membrane fragments attached to the sorted cells werenot detected. We
have quantified the platelet-cell interaction and showed that,
when thrombin-activated platelets were added to the MNC
fraction of BM or PB, about 50% of the CD34' cells bound
activated platelets. This binding of platelets to CD34' cells
is dependent on divalent cations and is reversible, because
thrombin-activated platelets adherent to CD34' cells could
be removed by subsequent washing with EDTAin high concentration. Therefore, these studies indicate that platelet adherence could be attributed to the detection of platelet GPs
on CD34' cells and could explain why, when measured in
the absence of EDTA, strong expression of platelet antigens
on CD34' cells has been found by others."'"
Furthermore, the platelet interaction with CD34' cells derived from BM or PBwas characterized. This interaction
was shown to be mediated by P-selectin, because MoAbs
against P-selectin almost completely blocked this interaction.
Platelets adherence was similar to CD34' from BM or PB.
P-selectin isknown to interact with carbohydrate ligands
such as sialyl Lewis" and relatedcarbohydrates presented on
a polypeptide backbone.2R.35
Recently, a specific GP (PSGL1) on myeloid cells has been recognized as a ligand for Pselectin.36 This P-selectin ligand carries large numbers of
sialylated 0-linked oligosaccharides and is functionally active only when it is appropriately glycosylated. The sialic
acid(s) on this ligand is required for binding to P-selectin
but appears to be somewhat resistant to sialidase treatment.3s
Our results are in accordance with these findings. We showed
that treatment of CD34' cells with glycoprotease derived
from Pasteurella haemolytica, which cleaves cell surface
proteins containing 0-linked glycans, completely inhibited
the binding of activated platelets to CD34' cells. In addition,
this binding was only partially inhibited by neuraminidase,
as was previously found for monocytes, lymphocytes, eosinophils, and basophils."' Treatment with the proteolytic en-
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
3780
zyme trypsinpartially inhibited and with elastase completely
inhibited the binding of activated platelets to CD34+ cells,
although the amount of sialyl Lewis"was unaffected. In
addition to the enzymatic characterization of the P-selectin
ligand on CD34+ cells, our PCR results suggest the expression of mRNA for PSGL-1 by the CD34+ progenitor cells.
Our data do not exclude the presence of P-selectin ligands
other than PSGL-1 on CD34' cells. L-selectin, which has
multiple ligands, including other selectins, may function as
a ligand for P-~electin.~'
In addition, it was recently reported
that vascular CD34 functions as a ligand for L-selectin.32
Analogous with this report, hematopoietic CD34 could function as a ligand for P-selectin. In our study, no evidence was
found for a role for either L-selectin or hematopoietic CD34
as ligand for P-selectin. The reactivity of the L-selectin reactive MoAb Leu-8 with CD34+ cells was not decreased by
elastase treatment, whereas the binding of activated platelets
to CD34' cells was completely inhibited. In addition, no
blocking effect of platelet CD34' cell interaction by a panel
of anti-CD34 MoAbs was observed, suggesting that hematopoietic CD34 is not a major ligand in the P-selectin-mediated interaction of activated platelets and CD34+ cells.
After showing that the binding of platelets to CD34' cells
is reversible and dependent on divalent cations, we were
able to measure platelet GP expression on CD34' cells using
double-color flow cytometry after the removal of platelets
by density centrifugation and washing in EDTA-containing
buffer. Under these conditions, a small but distinct population of CD34' cells, ranging from 2% to 12%, reacted with
CD41 and CD61 MoAbs, which recognize GPs IIb and IIIa.
The percentage of cells reactive with CD41 MoAbwas
higher than that reactive with CD61 MoAb. This discrepancy
is probably caused by a higher affinity of the CD41 MoAb.
We cannot rule out the possibility that, in the presence of
E D T A , a s m a l lp e r c e n t a g eo fC D 3 4 ' C D 4 1 +o r
CD34'CD61' cells are cells to which platelet fragments are
still bound, because some contaminating promonocytes were
still present in the CD34'CD41+ sorted cell fraction as observed by electron microscopy. However, several lines of
evidence indicate that the number ofCD34'
cells with
attached platelets in this population is negligible and that the
reactivity with the GPIIbDIIa antibodies is caused by protein
synthesis by CD34+ cells. Firstly, no CD34' cells reacted
withMoAbs recognizing GPIba or withMoAbs against
GPIa and GPV, GPs known to be present on platelets. Secondly, using a semiquantitative PCR analysis, we showed
that the expression of mRNA for GPIIb cannot be due only
to adhering platelets because, in that case, the ratio of GPIIb
mRNAand PF4 mRNAwouldbe similar to the ratio of
these products in platelets, whereas we clearly found a preferentially expression of GPIIb in CD34' cells. Therefore,
these results indicate that mRNAfor GPIIb is mainlysynthesized by CD34' progenitor cells. The positive results of the
PF4 PCR may indicate that this protein has already begun
to be synthesized at the progenitor cell level, although we
cannot exclude the possibility that PF4 mRNA is solely derived from adhering platelets. We found GPIIblIIIa mFWA
also in the CD41- population, although at a lower level.
This may becaused by a lower sensitivity of flow cytometric
DERCKSEN ET AL
analysis. Debili et aly haveshownthat
the moremature
megakaryocyte progenitors are present in the CD34 'CD4 1
cell population, whereas immature megakaryocyte progenitors are present in the CD34'CD41- cell population.
No staining of CD34+ cells was seen with MoAbs recognizing GPIa, GPIba, or GPV. GPIIIa can also be associated
withthe a-chain of the vitronectin receptor (CDSl).'X.'"
CD34+ cells did not react with CD51 antibodies, indicating
that most if not all GPIIIa on CD34' cells is associated with
the platelet-specific GP GPIIb. CD34' cells also reacted with
MoAbs against GPIIa' (CD31) and GPIV (CD36). However,
these MoAbs cannot be used for determining the number of
platelet progenitor cells, because GPIIa' was found tobe
also present on myeloidprogenitors4"and GPIV on erythroid
precursors."
Current evidence indicates that GPIIb and GPIIIa are detected first during megakaryocytic differentiation, whereas
the platelet GP Iba appears to be a slightly later marker of
differentiati~n.~.~'"
Usingpurified CD34+ cells, Debili et
al' showed that, in culture, a very small population of CD34'
express GPIba. These cells are extremely rare and might not
be detectable using the present technique. Moreover, among
acute megakaryoblastic leukemic cells, cells showing reactivity with both GPIIb/IIIa and GPIba antibodies as well as
cells expressing only GPIIbDIIa have been f o ~ n d . The
~ ~ ~ ~ " ~ ~
results of our experiments are concordant with those of these
earlier reports, because we show that only GPIIb/IIIa is present on CD34' cells, indicating that these cells represent an
early stage in the megakaryocyte lineage.
We assumed that, if performed in the presence of EDTA,
the flow cytometric measurement of CD34' cells expressing
platelet antigens could be used as an indicator for the number
of megakaryocytic progenitors. Indeed, we found that
CD34'CD41' cells correlated significantly better with the
time to platelet recovery after autologous PBSC transplantation than the total number of CD34' cells. This observation
suggests that the CD34TD41' cells represent megakaryocytic precursors that are responsible for platelet recovery
after PBSC transplantation. An important clinical issue is to
establish a practical minimum number of cells required for
rapid platelet recovery. In this study, the threshold for rapid
platelet recovery
was
calculated be
to
0.34 X IOh
CD34'CD41+ cellskg and reinfusion of morethanthe
threshold of CD34+CD41 cells per killogram resulted in a
significantly faster recovery. Together with the significantly
better correlation of the number of reinfused CD34+CD41'
cells per killogram with platelet recovery than of the total
number of CD34+ cells per killogram, this study indicates
that determining the number of megakaryocytic precursor
cells by flow cytometry most accurately represents the platelet reconstitutive capacity of the PBSC transplant. The value
of determining the number of CD34'CD41'
cells by flow
cytometry has to be confirmed in studies with a larger number of patients receiving different types of high-dose chemotherapy.
+
+
ACKNOWLEDGMENT
We thank Dr Suat Simsek, Lucia de Bruijne-Admiraal, and Onno
Verhagen for technical assistance; Wim Schaasberg for statistical
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
PLATELETGLYCOPROTEIN EXPRESSION ON CD34’CELLS
analyses; Dr Najet Debili for helpful discussion; and Dr Dirk Roos
and Dr Paul Engelfriet for critically reading the manuscript.
REFERENCES
1. Siena S, Bregni M, Brando B, Belli N, Ravagnani F, Gandola
L, Stern AC, Lansdorp PM, Bonadonna G, Gianni AM: Flowcytometry for clinical estimation of circulating hematopoietic progenitors
for autologous transplantation in cancer patients. Blood 77:400, 1991
2. Haas R, Mohle R, Friihauf S, Goldschmidt H, Witt B, Flentje
M, Wannenmacher M, Humstein W: Patient characteristics associated with successful mobilizing and autografting of peripheral blood
progenitor cells in malignant lymphoma. Blood 83:3787, 1994
3. van der Wall E, Richel DJ, Holtkamp MJ, Slaper-Cortenbach
K M , van der Schoot CE, Dalesio 0, Nooijen WJ, Schornagel JH,
Rodenhuis S: Bone marrow reconstitution after high-dose chemotherapy and autologous peripheral stem cell transplantation: Effect
of graft size. Ann Oncol 5:795, 1994
4. Chao NJ, Schriber JR, Grimes K, Long GD, Negrin RS,
Raimondi CM, HomingSJ, Brown SL, Miller L, Blume KG: Granulocyte colony-stimulating factor “mobilized” peripheral blood progenitor cells accelerate granulocyte and platelet recovery after highdose chemotherapy. Blood 81:2031, 1993
S. Zucker Franklin D, Yang JS, Grusky G: Characterization of
glycoprotein IIb/IIIa-positive cells in human umbilical cord blood:
Their potential usefulness as megakaryocyte progenitors. Blood
79:347, 1992
6. Smeland EB, Funderud S, Kvalheim G, Gaudernack G, Rasmussen AM, Rusten L, Wang MY,Tindle RW, Blomhoff HK, Egeland T: Isolation and characterization of human hematopoietic progenitor cells: An effective method for positive selection of CD34+
cells. Leukemia 6:845, 1992
7. Lansdorp PM: Subpopulations of cells that express CD34.
Blood 77:13a, 1991
8. Bender JG, Williams SF, Myers S, Nottleman D, Lee WJ,
Unverzagt KL, Walker D, To LB, Van Epps DE: Characterization
of Chemotherapymobilized peripheral blood progenitor cells for use
in autologous stem cell transplantation. Bone Marrow Transplant
10:281,1992
9. Debili N. Issaad C, Masse JM, Guichard J, Katz A, BretonGorius J, Vainchenker W: Expression of CD34 and platelet glycoproteins during human megakaryocytic differentiation. Blood
80:3022, 1992
10. de Bruijne-Admiraal LG, Modderman PW, von dem Borne
AEGKr, Sonnenberg A: P-selectin mediates Ca*+-dependentadhesion of activated platelets to many different types of leukocytes:
Detection by flow cytometry. Blood 80:134, 1992
1 1 . Jungi TW, Spycher MO, Nydegger UE, Barandun S: Plateletleukocyte interactions: Selective binding of thrombin-stimulated
platelets to human monocytes, polymorphonuclear leukocytes and
related cell lines. Blood 67:629, 1986
12. Perussia B, Jankiewicz J, Trinchieri G : Binding of platelets
to human monocytes: A source of artifacts in the study of the specificity of antileukocyte antibodies. J Immunol Methods SO:269, 1982
13. Breton-Gorius J, Lewis JC, Guichard J, Kieffer N, Vainchenker W: Monoclonal antibodies specific for human platelet membrane glycoproteins bind tomonocytes by focal absorption of platelet
membrane fragments: An ultrastructural immunogold study. Leukemia 1:131,1987
14. Betz SA, Foucar K, Head DR, Chen IM, Willman CL: Falsepositive flow cytometric platelet glycoprotein IIbAIIa expression in
myeloid leukemias secondary to platelet adherence to blasts. Blood
79:2399, 1992
15. ffiapp W, Bettelheim P, Gadd S, Koller U, Majdic 0,Peschel
C, Radaszkiewicz T, Stockinger H, Tetteroo PAT: Myeloid section
report. White cell differentiation antigens, part 5, in Knapp W, Dor-
3781
ken B, G i b WR, Rieben EP, Schmidt R E , Stein H, Von Dem Borne
AEGKr (eds): Fourth International Workshop and Conference on
Human Leukocyte Differentiation Antigens, Leukocyte Typing IV.
New York, NY, Oxford, 1989, p 747
16. von dem Borne AEGKr, Modderman PW, Admiraal LG,
Nieuwenhuis HK: Joint report of the platelet section. White cell
differentiation antigens, part 6, platelet antigens, in Knapp W, Dorken B, Gilks WR, Rieben EP, Schmidt RE, Stein H, Von DemBorne
AEGKr (eds): Fourth International Workshop and Conference on
Human Leukocyte Differentiation Antigens, Leukocyte Typing IV.
New York, NY, Oxford, 1989, p 952
17. Greaves MF, Titley I, Colman SM, Buhring HJ, Campos L,
Castoldi GL, Garrido F, Gaudernack G, Girard JP, Ingles-Esteve J,
Invernizzi R, Knapp W, Lansdorp PM, Lanza F, Merle-Beral H,
Parravicini C, Razak K, Ruiz-Cabello F, Springer TA, van der
Schoot CE, Sutherland D R Report on the CD34 Cluster Workshop,
in Greaves MF (ed): Fifth International Workshop and Conference
on Human Leukocyte Differentiation Antigens, Leukocyte Typing
V. New York, NY, Oxford, 1995, p 840
18. van der Wall E, Richel DJ, Kusumanto YH, Rutgers EJT,
Schornagel JH, Schaake-Koning CCE, Peterse JL, Rodenhuis S:
Feasibility study of FEC-chemotherapy with dose-intensive epirubicin as initial treatment in high-riskbreast cancer. Ann Oncol 4:791,
1993
19. Rodenhuis S, Baars JW, Schornagel JH, Vlasveld LT, Mandjes I, Pinedo HM, Richel DJ: Feasibility and toxicity study of a
high-dose chemotherapy regimen for autotransplantation incorporating carboplatin, cyclophosphamide and thiotepa. Ann Oncol 32355,
1992
20. Biron P, Goldstone A, Colombat P: A new cytoreductive
conditioning regimen before ABMT in lymphomas: the BEAM protocol. A phase I1 study in autologous bone marrow transplantation,
in Dicke KA, Spitzer G, Zander AR (eds): Proceedings of the 2nd
International Symposium on Autologous Bone Marrow Transplantation. Houston, TX, University of Texas MD Anderson Hospital and
Tumor Institute, 1987, p 593
21. de Bruijne-Admiraal LG, Daams GM, Bos MJE, von dem
Borne AEGKr: Platelet activation detected by two groups of antibodies from the Platelet Workshop and by two unique antibodies. White
cell differentiation antigens, part 6, platelets, in Knapp W, Dorken
B, Gilks WR, Rieben EP, Schmidt RE, Stein H, Von Dem Bome
AEGKr (eds): Fourth International Workshop and Conference on
Human Leukocyte Differentiation Antigens, Leukocyte Typing IV.
New York, NY, Oxford, 1989, p 1041
22. Breton-Gonus J, Vanhaeke D, Pryzwansky KB, Guichard J,
Tabilio A, Vainchenker W, Camel R: Simultaneous detection of
membrane markers with monoclonal antibodies and peroxidatic activities in leukaemia: Ultrastructural analysis using a new method
of fixation preserving the platelet peroxidase. Br J Haematol 58:447,
1984
23. Graham RC, Karnovsky MJ: The early stages of absorption
of injected horseradish peroxidase in proximal tubules ofmouse
and kidney. Ultrastructural cytochemistry by a new technique. J
Histochem Cytochem 14:291, 1966
24. Ziegler BL, Lamping C, Thoma S, Thomas CA: Single-cell
cDNA-PCR. Removal of contaminating genomic DNA from total
RNA using immobilized DNase I. Biotechniques 13:726, 1992
25. Rappolee DA, Wang A, Mark D, Werb 2: Novel method for
studying mRNA phenotypes in single or small numbers of cells. J
Cell Biochem 39:1, 1989
26. Poncz M, Eisman R, Heidenreich R, Silver SM, Vilaire G ,
Surrey S, Schwartz E, Bennett JS: Structure of the platelet membrane
glycoprotein IIb. Homology to the alpha subunits of the vitronectin
and fibronectin membrane receptors. J Biol Chem 262:8476, 1987
27. Zimrin AB, Gidwitz S, Lord S, Schwartz E, Bennett JS, White
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
3782
GC, Poncz M: The genomic organisation of platelet glycoprotein
IIIa. J Biol Chem 265:8590, 1990
28. Sako D, Chang X, Barone KM, Vachino G, White HM, Shaw
G, Veldman GM, Bean KM, Ahem TJ, Furie B, Cumming DA,
Larsen GR: Expression cloning of a functional glycoprotein ligand
for P-selectin. Cell 75: 1179, 1993
29. Ponte P, Ng SY, Engel J, Gunning P, Kedes L: Evolutionary
conservation in the untranslated regions ofactinmRNAs:DNA
sequence of a human beta-actin cDNA. Nucleic Acids Res 12:1687,
1984
30. Skubitz K, Balke J, Ball E, Bridges R, Buescher ES, Campos
L, Harvath L, Kerr M, Kniep B, Spitalnik S, Skubitz A, Thomson
J, Wick M, William L: Report on CD15 Cluster Workshop. White
cell differentiation antigens, part 5 , myeloid antigens, in Knapp W,
Darken B, Gilks WR, Rieben EP, Schmidt RE, Stein H, Von Dem
Borne AEGKr (eds): Fourth International Workshop and Conference
on Human Leukocyte Differentiation Antigens, Leukocyte Typing
IV. New York, NY, Oxford, 1989, p 800
31. Tetteroo PA, Mulder A, Lansdorp PM, Zola H, Baker DA,
Visser FJ, von dem Borne AE: Myeloid-associated antigen 3-alphafucosyl-N-acetyllactosamine (FAL): Location on various granulocyte membrane glycoproteins and masking upon monocytic differentiation. Eur J Immunol 14:1089, 1984
32. Baumhueter S, Singer MS, Henzel W, Hemmerich S, Renz
M,Rosen SD, Lasky LA: Binding of L-selectin to the vascular
sialomucin CD34. Science 262:436, 1993
33. Abdullay KM, Udoh EA, Shewen PE, Mellors A: A neutral
glycoprotease of Pasteurella haemolytica A1 specifically cleaves 0sialoglycoproteins. Infect Immun 6056, I992
34. Sutherland DR, Abdullah KM, Cyopick P, Mellors A: Cleavageof the cell-surface 0-sialoglycoproteins CD34, CD43, CD44,
and CD45 by a novel glycoprotease from Pasteurella haemolytica.
J Immunol 148:1458, 1992
35. Norgard KE, Moore KL, Diaz S, Stults NL, Ushiyama S,
McEver RP, Cummings RD, Varki A: Characterization of a specific
ligand for P-selectin on myeloid cells. J Biol Chem 268:12764, 1993
36. Moore KL, Stults NL,Diaz S, Smith DF, Cummings RD.
Varki A, McEver RP: Identification of a specific glycoprotein ligand
for P-selectin (CD62) on myeloid cells. J Cell Biol 118:445, 1992
37. Picker LJ, Warnock RA, Bums AR, Doerschuk CM, Berg
EL, Butcher EC: The neutrophil selectin LECAM-l presents carbo-
DERCKSEN ET AL
hydrate ligands tothe vascular selectins ELAM-l and GMP-140.
Cell 66:921. 1991
38. Ginsberg MH, Loftus J, Ryckwaert JJ, Pierschbacher M. Pytela R, Ruoslahti E, Plow EF: Immunochemical and amino-terminal
sequence comparison of two cytoadhesins indicates theycontain
similar or identical beta subunits and distinct alpha subunits. J Biol
Chem 262:5437, 1987
39. Leeksma OC, Zandbergen Spaargaren J, Giltay JC, van
Mourik JA: Cultured human endothelial cells synthesize a plasma
membrane protein complex immunologically related to the platelet
glycoprotein IIb/IIIa complex. Blood 67: 1176, 1986
40. Watt SM, Williamson J, Genevier H, Fawcett J, Simmons
DL, Hatzfeld A, Nesbitt SA. Coombe DR: The heparinbinding
PECAM-I adhesion molecule i s expressed by CD34+ hematopoietic
precursor cells with early myeloid and B-lymphoid cell phenotypes.
Blood 82:2649, 1993
41. Edelman P. Vinci G, Villeval JL, Vainchenker W, Henri A,
Miglierina R, Rouger P, Reviron J, Breton-Gorius J, Sureau C, Edelman L: A monoclonal antibody against an erythrocyte ontogenic
antigen identifies fetal and adult erythroid progenitors. Blood 67:56,
1986
42. Vainchenker W, Deschamps JF, Bastin JM, Guichard J, Titeux M, Breton-Gorius J, McMicbael AJ: Two monoclonal antiplatelet antibodies as markers of human megakaryocyte maturation: immunofluorescent staining and platelet peroxidase detection in
megakaryocyte colonies and in in vivo cells from normal and leukemic patients. Blood 59:514, 1982
43. Vinci G, Tabilio A, Deschamps JF, Van Haeke D, Henri A,
Guichard J, Tetteroo P, Lansdorp PM, Hercend T, Vainchenker W,
Breton-Gorius J: lmmunological study of in vitro maturation of human megakaryocytes. Br J Haematol 56589, 1984
44. Breton-Gorius J, Vainchenker W: Expression of platelet proteins during the in vitro and invivo differentiation of megakaryocytes
and morphological aspects of their maturation. Semin Hematol
23:43, 1986
45. Imamura N, Kajihara H, Kuramoto A: Flow cytometric analysis of peroxidase negative acute leukemias by monoclonal antibodies-11. Acute megakaryoblastic and acute pro-megakaryocytic leukemias. Leuk Res 12:279, 1988
46. Koike T: Megakaryoblastic leukaemia: The characterization
and identification of megakaryoblasts. Blood 64:683, 1984
From www.bloodjournal.org by guest on June 17, 2017. For personal use only.
1995 86: 3771-3782
The value of flow cytometric analysis of platelet glycoprotein
expression of CD34+ cells measured under conditions that prevent Pselectin-mediated binding of platelets
MW Dercksen, IS Weimar, DJ Richel, J Breton-Gorius, W Vainchenker, CM Slaper- Cortenbach,
HM Pinedo, AE von dem Borne, WR Gerritsen and CE van der Schoot
Updated information and services can be found at:
http://www.bloodjournal.org/content/86/10/3771.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.