ARTICLE IN PRESS

Immunobiology 215 (2010) 38–52

www.elsevier.de/imbio

Characterisation of degranulation and phagocytic capacity

of a human neutrophilic cellular model, PLB-985 cells$
Christophe Pivot-Pajota,1, François C. Chouinarda, Mohammed Amine El Azreqa,
Danielle Harboura, Sylvain G. Bourgoina,b,
a
Centre de Recherche en Rhumatologie et Immunologie, Centre de Recherche du CHUL – CHUQ, Québec, Canada
b
Département d’Anatomie-Physiologie, Université Laval, Québec, Canada

Received 27 August 2008; received in revised form 12 January 2009; accepted 12 January 2009

Abstract
Exocytosis of neutrophil granules is a major event that converts circulating neutrophils into fully activated cells
capable of chemotaxis, phagocytosis and destruction of pathogens. The PLB-985 cell line is a suitable neutrophilic
cellular model which is utilised to study the different functional responses of neutrophils. In this study, we
characterised the differentiation of PLB-985 cells toward the granulocytic pathway, using three different inducing
agents: dbcAMP, DMSO and DMF. The differentiation efficiency was monitored by observation of cell morphology
with electron microscopy, and by analysis of the expression of receptors such as FPRL1 and FcgRIIA, the distribution
or release of granule markers, phagocytic capacity, as well as measurement of fMLF-induced calcium fluxes.
Exocytosis and phagocytosis in differentiated cells were weaker as compared to neutrophils. fMLF stimulated primary
granule exocytosis in cells differentiated with dbcAMP, DMSO and DMF, whereas the release of the contents
of tertiary granules, as well as that of secretory vesicles, was only observed in dbcAMP-differentiated cells.
DMSO-differentiated cells exhibited the highest phagocytic capacity. Altogether our results reinforce the fact that
depending on the differentiating agent used, PLB-985 cells represent a useful model to study neutrophil functions and
to bypass difficulties inherent to these primary cells.
r 2009 Elsevier GmbH. All rights reserved.

Keywords: Exocytosis; Granulocytic differentiation; Neutrophils; Phagocytosis; PLB-985 cells

Abbreviations: CB, cytochalasin B; dbcAMP, dibutyryl cyclic adenosine monophosphate; DMF, dimethyl formamide; DMSO, dimethyl sulfoxide;
FcgR, Fcg receptor; fMLF, N-formyl-methionyl-leucyl-phenylalanine; FPR, N-formyl peptide receptor; FPRL1, formyl peptide like receptor 1;
GAPDH, glyceraldehyde-3-phosphate dehydrogenase; LDH, lactate dehydrogenase; LF, lactoferrin; MMP-9, matrix metalloproteinase 9; NADPH,
nicotinamide adenine dinucleotide phosphate; MPO, myeloperoxidase; PMA, phorbol myristate acetate; PMN, polymorphonuclear neutrophil;
ROS, reactive oxygen species.
$
This work was supported by grants from the Canadian Institutes of Health Research and the Arthritis Society of Canada (TAS04/0061) (S.G.B).
Corresponding author at: Centre de Recherche en Rhumatologie et Immunologie, Centre de Recherche du CHUL – CHUQ, 2705 Boul. Laurier,
Local T1-49, Québec, Canada G1V 4G2. Tel.: +1418 656 4141; fax: +1418 654 2765.
E-mail address: Sylvain.bourgoin@crchul.ulaval.ca (S.G. Bourgoin).
1
The recipient of a fellowship from La Fondation pour la Recherche Médicale.

0171-2985/$ - see front matter r 2009 Elsevier GmbH. All rights reserved.
doi:10.1016/j.imbio.2009.01.007
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 39

Introduction et al., 2006; Taylor et al., 2006; van Bruggen et al.,

2004), but the other functional responses like degranula-
Among cellular mediators of inflammation, polymor- tion and phagocytosis have been less characterised,
phonuclear neutrophils (PMNs) are central effectors in or restricted to one type of differentiating agent
host defence against invading pathogens. Recruitment (Hazan-Eitan et al., 2006; Pedruzzi et al., 2002). A
of circulating neutrophils toward inflammatory foci recent paper reported that DMSO-differentiated PLB-
takes place rapidly and constitutes a major event in the 985 cells responded like blood neutrophils in terms of
neutrophil-mediated inflammatory response. Neutrophil inflammatory cytokine production and transcription
recruitment involves a series of regulated exocytic factor activation, and were readily amenable to transient
events, which gradually change the functional status of transfection (Ear and McDonald 2008). PLB-985 cells
neutrophils, converting them from non-activated circu- currently represent a suitable and very attractive
lating cells to potent effector cells of the innate immune neutrophilic cellular model.
system (Borregaard and Cowland 1997; Faurschou and In the present study, we characterised the differentia-
Borregaard 2003). Granule mobilisation and degranula- tion efficiency of PLB-985 cells toward a granulocytic
tion are important components of neutrophil functions phenotype, using three different agents: dbcAMP, DMF
in inflammation and host defence against microbial and DMSO. Granule exocytosis of these differentiated
infection. Human neutrophils contain four major types cells was compared to that of blood neutrophils.
of granules: azurophilic or primary granules, specific or Moreover, we analysed the phagocytic capacity of these
secondary granules, gelatinase or tertiary granules, and differentiated cells, using a rapid, sensitive and accurate
alkaline phosphatase-rich granules, also named secre- method for quantification of phagocytosis.
tory vesicles (Borregaard and Cowland 1997). Mechan-
isms involved in the exocytic regulation of the different
granules are complex and not yet fully understood.
Neutrophils have the ability to recognise and engulf Materials and methods
invading microorganisms, thereby eliminating patho-
gens and cell debris by a complex and tightly regulated Reagents and antibodies
process known as phagocytosis (Underhill and Ozinsky
2002). Bacterial killing is closely linked to granule fMLF, cytochalasin B (CB), phenylmethylsulfonyl
mobilisation since the phagosome builds its antimicro- fluoride (PMSF), dbcAMP, DMSO and DMF were
bial capacity through fusion with secretory vesicles and obtained from Sigma-Aldrich (Oakville, ON, Canada).
granules. Cytotoxic granule proteins thus delivered Dextran T-500 and Percoll were purchased from GE
allow the killing and the degradation of internalised Healthcare (Baie d’Urfé, QC, Canada). Ficoll-Paque,
microbes inside phagolysosomes (Lee et al., 2003; Urban phosphate buffered saline (PBS), RPMI-1640 and
et al., 2006). Mg2+-free HBSS were obtained from Wisent (St-Bruno,
Neutrophils are terminally differentiated cells that QC, Canada). Aprotinin and leupeptin were obtained
have no proliferative capacity, with a reduced lifespan from Boehringer Ingelheim (Burlington, ON, Canada).
and refractoriness to genetic manipulation, making Di-isopropylfluorophosphate (DFP) was from Serva
them difficult to use to study activation of signal (Heidelberg, Germany). Anti-actin monoclonal, anti-
transduction pathways. Several cell lines with granulo- lactoferrin (LF) and anti-CD11b antibodies were
cytic differentiating capacity have been developed. The obtained from Sigma-Aldrich. Anti-MMP-9 and anti-
human promyelocytic leukemic HL-60 and PLB-985 cell CD63 antibodies were purchased from Abcam
lines have been often studied as models of myeloid (Cambridge, MA, USA). Anti-myeloperoxidase
differentiation (Collins et al., 1977; Newburger et al., (MPO), and anti-CD16 antibodies were from Dako
1979; Tucker et al., 1987). Like HL-60 cells, PLB-985 is Canada (Mississauga, ON, Canada), and anti-lactate
a bipotential cell line that can be differentiated along the dehydrogenase (LDH) antibody was from Fitzgerald
granulocytic pathway with compounds like dimethyl (Concord, MA, USA). FITC-anti-CD63 and FITC-
sulfoxide (DMSO) (Tucker et al., 1987), dimethylfor- anti-CD66 antibodies were from Beckman Coulter
mamide (DMF) (Katschinski et al., 1999) or dibutyryl Canada (Mississauga, ON, Canada). The anti-CD32A
cyclic AMP (dbcAMP) (Hazan-Eitan et al., 2006), and monoclonal antibody IV.3 was purified from ascites
toward the monocytic pathway with 1,25-dihydroxy fluid of mice, and the anti-CD32A polyclonal antibody
vitamin D3 (Perkins et al., 1991), phorbol myristate CT10 against the cytoplasmic domain of CD32A is an
acetate (PMA) (Perkins et al., 1991) or interferon-g IgG-purified from a rabbit antiserum (Rollet-Labelle
(Hazan-Eitan et al., 2006). The PLB-985 cell line has et al., 2004). Anti-albumin antibody was purchased
been widely utilised to monitor activity of NADPH from Cedarlane. Anti-FPRL1 monoclonal antibody was
oxidase and production of reactive oxygen species obtained from R&D Systems (Mineapolis, MN, USA).
(ROS) (Boulven et al., 2006; Fay et al., 2006; Pessach Anti-Lyn antibody was purchased from Santa Cruz
 ARTICLE IN PRESS
40 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

(Santa Cruz, MA, USA). Alexa Fluor 488- and processed for immunoblot analysis using protein mar-
594-labelled zymosan A (S. cerevisiae) BioParticles, kers corresponding to the different subcellular compart-
ProLong Gold antifade reagent and Fura 2-AM were ments: MPO (primary granules), LF (secondary
obtained from Invitrogen (Carlsbad, CA, USA). Trypan granules), gelatinase (tertiary granules), albumin (secre-
Blue was purchased from EM Diagnostic Systems tory vesicles), CD32A (plasma membrane) and LDH
(Gibbstown, NJ, USA). Paraformaldehyde (PFA) was (cytosol).
from Laboratoire MAT (Québec, QC, Canada). Calcein
acetoxymethylester (Calcein–AM) was from Calbio- Immunoisolation
chem (San Diego, CA, USA). Plasma membranes from differentiated PLB-985 cells
were isolated using magnetic beads coated with specific
Isolation of human neutrophils antibodies. M-500 dynabeads (1.4  107) (Dynal, Oslo,
Norway) coated with a goat anti-mouse IgG specific for
Neutrophils were isolated from healthy adult volun- the Fcg fragment (Jackson Immuno Research, West
teers as previously described (Marcil et al., 1999). Grove, PA, USA) were incubated with 8.4 mg of anti-
Neutrophils were resuspended in HBSS, pH 7.4, CD32A (IV.3) or an isotype control (IgG2b) antibody
containing 1.6 mM Ca2+ but no Mg2+. Before stimula- (eBioscience, San Diego, CA, USA) in PBS/0.1% BSA
tion, the neutrophil suspensions were incubated with overnight at 4 1C under constant agitation, and then
1 mM DFP for 15 min at room temperature. rinsed with PBS/0.1% BSA three times for 5 min.
Differentiated PLB-985 cells were cavitated as described
Cell culture above. Cell lysates corresponding to 15  106 cells
were then incubated with anti-CD32- or IgG2b-coated
PLB-985 cells (German collection of microorganisms beads overnight at 4 1C on a rotator. The beads were
and cell culture) were grown in RPMI 1640 medium washed three times for 5 min in KCl-HEPES relaxation
containing 10% FBS and 2 mM L-glutamine at 37 1C, in buffer. Plasma membranes were eluted with 60 ml
a humidified atmosphere of 5% CO2. To induce a of Laemmli’s sample buffer and boiled 10 min. Proteins
neutrophil-like phenotype, cells were cultured in med- were then separated by SDS-PAGE prior to immuno-
ium supplemented either with 0.3 mM dbcAMP for blotting.
three days, or with 1.25% DMSO for five days, or with
0.5% DMF for six days. For DMSO and DMF Electrophoresis and immunoblotting
treatments, the medium was changed once on day three
of the differentiation period. Proteins were analysed on 7.5–20% gradient SDS-
polyacrylamide gels, and transferred to Immobilon
Subcellular fractionation polyvinylidene difluoride (PVDF) membranes (Milli-
pore Corporation, Bedford, MA, USA). Immunoblot-
Percoll gradient fractionation ting was performed using the indicated antibodies and
Neutrophil and PLB-985 cell subcellular organelles revealed with the HRP-conjugated secondary anti-
were isolated according to a procedure adapted from rabbit or anti-mouse antibodies (1/20,000) and the
Kjeldsen et al. (Kjeldsen et al. 1999) with modifications Renaissance detection system (NEN, Perkin Elmer Life
as previously described (Kjeldsen et al., 1999; Pivot- Sciences, Boston, MA, USA).
Pajot et al., 2008). Neutrophils or differentiated
PLB-985 cells (4  108 cells) were centrifuged and RT-PCR
resuspended in 10 ml ice-cold KCl-HEPES relaxation
buffer (100 mM KCl, 50 mM HEPES, 5 mM NaCl, Total RNA was extracted from neutrophils or PLB-
1 mM MgCl2, 0.5 mM EGTA, 1 mM DFP, 2.5 mg/ml 985 cells using TRIzol reagent (Invitrogen, Burlington,
aprotinin, 2.5 mg/ml leupeptin, and 2.5 mM PMSF, pH ON, Canada). RNA samples were subjected to DNAse
7.2). Cells were pressurised (400 psi, 10 min) in a treatment (Promega, Nepean, ON, Canada) for 30 min
nitrogen bomb (Parr Instrument Co., Moline, IL, at 37 1C (1 unit per mg of RNA) and reverse transcribed
USA). Cavitates were centrifuged at 400g for 5 min to with SuperScript II reverse transcriptase (Invitrogen)
pellet the unbroken cells. Supernatants were laid onto using random hexamer primers. The following primers
3  4.5 ml of a Percoll step gradient (1.050, 1.090 and were used to amplify human MPO, LF, MMP-9,
1.120 g/ml). After centrifugation (37,000g for 30 min), albumin and GAPDH as control: MPO, 50 primer
18 fractions were collected (1 ml each), starting from the GGGTTCCCTTCTTCTCTTCTCTCAGATGC, 30 pri-
bottom of the tube. Each fraction was centrifuged at mer GGCGGGATCTTGAGCGGGAAGC; LF, 50
100,000g for 90 min, in order to pellet the Percoll. primer AAACTTGTCTTCCTCGTCCTGCTGTTCC,
Fractions (50 ml) were aspirated with a syringe and 30 primer TGCCACAACGGCATGAGAAGGGAC;
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 41

MMP-9, 50 primer CTGGTGCTCCTGGTGCTGGGC- (107 cells/ml) in RPMI-1640. A ratio of 10 particles per

TGC, 30 primer GCCGCGCCATCTGCGTTTCCAAA- cell was used to initiate phagocytosis. Synchronisation
CC; albumin, 50 primer GAACTTCGGGATGAAGG- of the interaction between cells and particles was
GAA, 30 primer GCAAGTCAGCAGGCATCTCA; achieved by rapidly sedimenting the cells together with
GAPDH, 50 primer GGGCGATGCTGGCGCTGAG- zymosan particles in a microcentrifuge tube (within
TACG, 30 primer CGCGGCCATCACGCCACAGTT- 15 s). Cells were then incubated at 37 1C, and phagocy-
TCC. PCR reactions were conducted using 5 ml of tosis was arrested by adding 300 ml of ice-cold HBSS and
cDNA and PCR products were separated on a 2% by keeping the samples on ice until flow cytometry
agarose gel. measurements were made. Two minutes prior to FACS
analysis, 0.4 mg/ml trypan blue in citrate buffer (pH 4.4)
Degranulation assay was added to quench the fluorescence of the non-
ingested Alexa 488-labelled zymosan BioParticles that
PLB-985 cells or neutrophils resuspended in HBSS remained cell associated. Samples were analysed using a
(107/ml) were stimulated with 107 M fMLF for 10 min EPICS XL flow cytometer (Beckman Coulter, Fullerton,
at 37 1C. The reaction was stopped by cooling and the CA, USA) with an argon laser operating at 488 nm. For
suspension was centrifuged for 1 min at 10,000g at 4 1C. each sample, a total of 10, 000 viable cells were counted.
To monitor primary and secondary granule exocytosis The average number of beads phagocytosed per cell
we analysed CD63 and CD66 cell surface expression by (phagocytic index) was calculated by dividing the mean
FACS using FITC-coupled anti-CD63 and FITC- fluorescence of the phagocytes by the fluorescence
coupled anti-CD66 antibodies, or the appropriate emitted by a single Alexa 488-conjugated zymosan
isotype control antibodies. Secondary and tertiary particle. The mean fluorescence intensity (MFI) multi-
granule as well as secretory vesicle exocytosis was plied by the percentage of viable cells that had ingested
analysed by monitoring CD11b cell surface expression. fluorescent particles was used to evaluate the phagocytic
Tertiary granule and secretory vesicle mobilisation was capacity of PLB-985 cells.
also monitored by measuring by ELISA the release of
MMP-9 (RayBiotech, Norcross, GA, USA) and of Phagocytosis assay and confocal microscopy
albumin, respectively (Bethyl Laboratories, Montgom- analysis
ery, TX, USA).
Neutrophils and PLB-985 cells (2  106 cells) were
Measurement of intracellular calcium resuspended in RPMI-1640 containing 2 mM Calcein-
AM and were allowed to adhere on glass coverslip for
Changes in intracellular calcium concentrations were 30 min at 37 1C in a 5% CO2 atmosphere. Alexa 594-
monitored by measuring Fura-2 fluorescence using a conjugated zymosan A BioParticles were opsonised as
Fluorolog-3 spectrofluorimeter (Horiba, NJ, USA). described above, and then added to the coverslips in a
Fluorescence was recorded at 510 nm with alternating ratio of 10 particles per cell. Incubation was performed
excitation wavelengths of 340 and 380 nm. Intracellular for 30 min at 37 1C, and phagocytosis was stopped by
calcium concentrations were calculated using the follow- washing the cells three times in ice-cold PBS. Cells were
ing formula: 224[(yFmin)(Fmaxy)1] in which y then fixed in 4% paraformaldehyde for 20 min at room
represents the fluorescence of the samples. Fmax was temperature. The coverslips were mounted using Pro-
obtained by adding 2.5 mM calcium ionophore to the cells long Gold antifade reagent and viewed under a laser
and Fmin was obtained by adding 25 mM MnCl2. scanning confocal microscope (Olympus IX-70) using a
Differentiated PLB-985 cells and blood neutrophils 60X PlanApo oil immersion objective. Images were
(107 cells/ml) were loaded with 1 mM Fura-2/AM, washed collected with Olympus FluoView FV300 (version 4.3)
twice and then transferred to the spectrofluorimeter. The acquisition software. Image analysis was performed
cell suspension, stimulated with 107 M fMLF, was using Volocity 4 (Improvision) and Adobe Photoshop
continuously stirred and maintained at 37 1C while software.
Fura-2 fluorescence measurements were made.
Electron microscopy
Phagocytosis assay and flow cytometry analysis
Cells were fixed for 1 h in 0.1 M cacodylate buffer
Alexa 488-conjugated zymosan A BioParticles were (pH 7.4) containing 2.5% glutaraldehyde and 2.5 mM
opsonised by incubation for 1 h at 37 1C in pooled and CaCl2. After washing in PBS, cells were post-fixed in 1%
decomplemented human sera diluted 50:50 with PBS. osmium tetroxide in PBS (pH 7.2), dehydrated sequen-
Particles were then washed twice in PBS, and added to tially with 30%, 50%, 70%, 95% and with 100%
25 ml of neutrophil or PLB-985 cell suspensions ethanol twice, and embedded in EPON 812. Ultrathin
 ARTICLE IN PRESS
42 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

sections were cut from hardened plastic blocks with a response to fMLF, whereas in differentiated cells, as
Sorvall M2-B ultramicrotome, stained with 3% uranyl well as in blood neutrophils, a rapid and transient
acetate and 0.1% lead citrate, and then mounted on increase in [Ca2+]c was measured. The increase in
300 mesh copper grids. The samples were observed with [Ca2+]c was stronger in dbcAMP-differentiated cells as
a JEOL-1010 electron microscope at 60 kV. compared to DMF- and DMSO-differentiated cells, but
remained lower than that of blood neutrophils. For all
three inducing agents used, longer times of differentia-
tion resulted in a lower increase in the levels of
Results intracellular calcium in response to stimulation with
fMLF. This was observed after six days of differentia-
Granulocytic differentiation of PLB-985 cells tion with dbcAMP, and after nine days of differentia-
tion with DMSO or DMF (data not shown).
Differentiation of PLB-985 toward a neutrophil-like
phenotype was induced with dbcAMP, DMF and
Expression of FccRIIA, FccRIIIB and granule
DMSO, three different agents known to promote
markers during differentiation of PLB-985 cells
granulocytic differentiation (Hazan-Eitan et al., 2006;
Katschinski et al., 1999; Tucker et al., 1987). We first
We next investigated the expression of Fcg receptors
analysed changes in cell morphology using transmission
(FcgRs) in differentiated PLB-985 cells. These FcgRs
electron microscopy. As shown in Fig. 1A, PLB-985
are known to bind the Fc domain of IgG, thereby
cells treated with dbcAMP, DMF and DMSO displayed
promoting the phagocytosis of IgG-opsonised particles.
important morphological changes as compared to
FcgRIIA (CD32A) and FcgRIII (CD16b) expression
undifferentiated cells. Undifferentiated PLB-985 cells
was monitored by Western blot. As shown in Fig. 2A,
contained a large, round nucleus with well defined
no CD32A was detected in undifferentiated cells,
nucleoli, and a cytoplasm devoid of granules. Differ-
whereas a treatment with the three differentiating agents
entiated PLB-985 cells acquired a lobular nuclear
stimulated CD32A expression. CD32A expression was
morphology, and granulations with different electron
higher in dbcAMP-differentiated cells, and lower in
densities in their cytoplasm similar to those observed in
DMF- and DMSO-differentiated cells as compared to
neutrophil cytoplasm. Differentiation with the three
neutrophils. Contrary to a previous study showing cell
different agents induced similar changes in nucleus
surface expression of CD16b in both undifferentiated
morphology and increases in granule biogenesis. Gran-
and DMF-differentiated PLB-985 cells by FACS
ulocytic maturation was also assessed by analysing the
analyses (Selmeczy et al., 2003), we found no CD16b
cell surface expression of the fMLF receptor, FPRL1,
expression in DMF-, dbcAMP- or DMSO-differentiated
by FACS. As shown in Fig. 1B, we found that FPRL1
PLB-985 cells, whereas high expression of CD16b was
expression was increased in all differentiated PLB-985
detected in blood neutrophils (Fig. 2B). In order to gain
cells as compared to undifferentiated cells. The level of
information on the granule content of PLB-985 cells, we
FPRL1 expression in blood neutrophils was higher than
monitored by RT-PCR the expression of the granule
that of DMF- and DMSO-differentiated cells, but lower
markers: MPO for primary granules, LF for secondary
than that of dbcAMP-differentiated PLB-985 cells. The
granules, gelatinase (MMP-9) for tertiary granules and
MFI multiplied by the percentage of FPRL1 positive
albumin for secretory vesicles. The levels of MPO
cells was used to evaluate and normalise the levels of
expression in dbcAMP-, DMF-, and DMSO-differen-
FPRL1 expression in differentiated PLB-985 cells. The
tiated PLB-985 cells were very similar but slightly lower
percentage of cells expressing FPRL1 after differentia-
than those of undifferentiated cells (Fig. 2C). As
tion for three days with dbcAMP, five days with DMSO
previously described for differentiated HL-60 cells,
and six days with DMF was 84%, 24.7%, and 78.5%,
which do not express secondary granule marker genes
respectively (Table 1). FPRL1 expression was maximal
(Johnston et al., 1992), no expression of LF was detected
at day three of dbcAMP differentiation and decreased
in differentiated PLB-985 cells. The levels of MMP-9
after incubation of cells with the inducing agent for six
and albumin expression in differentiated PLB-985 cells
days and nine days. Receptor expression was also
were less than those observed in blood neutrophils. As
maximal after five or six days of differentiation with
reference, GAPDH expression was also monitored.
DMSO and DMF, and declined by day nine (data not
shown). Because one of the early neutrophil responses to
chemotactic peptides like fMLF is a transient increase in Degranulation of differentiated PLB-985 cells
the levels of cytosolic calcium ([Ca2+]c), we measured
fMLF-induced calcium mobilisation in differentiated Thereafter, we monitored in differentiated PLB-985
PLB-985 cells. As shown in Fig. 1C, in undifferentiated cells the exocytosis of granules and secretory vesicles by
PLB-985 cells, no calcium mobilisation was observed in following the appearance of granule markers on the
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 43

neutrophil PLB-985 Nd

PLB-985 dbcAMP PLB-985 DMF PLB-985 DMSO

Fig. 1. Granulocytic differentiation of PLB-985 cells toward a functional neutrophil-like phenotype. Granulocytic differentiation of
PLB-985 cells was induced with 0.3 mM dbcAMP for three days, with 1.25% DMSO for five days, or with 0.5% DMF for six days.
(A) Transmission electron micrographs of undifferentiated (Nd) and differentiated PLB-985 cells, as well as blood neutrophils. Cells
were fixed and prepared as described in Materials and methods, and observed with a JEOL-1010 electron microscope at 60 kV. Bars,
2 mm (B) cell surface FPRL1 expression in blood neutrophils (PMN), undifferentiated (Nd) and differentiated PLB-985 cells, was
determined by FACS using an anti-FPRL1 monoclonal antibody or the appropriate isotype control, and a PE-conjugated goat anti-
mouse IgG. The results are representative of two independent experiments. (C) PMN, undifferentiated and differentiated PLB-985
cells were stimulated with 107 M fMLF at the time indicated by the arrow, and cytosolic calcium concentrations were monitored as
described in Materials and methods. Results shown in this panel are representative of at least three independent experiments. Means
were compared using a student’s t-test. The differences between indicated conditions were statistically significant with po0.05 (*).
 ARTICLE IN PRESS
44 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

Table 1. Comparison of the percentage of cells showing morphological changes with those of cells expressing FPRL1, CD63 and
CD11b after differentiation with dbcAMP, DMF and DMSO for three days, six days and five days, respectively.

Morphological FPRL1 positive CD63 positive CD11b positive

changes (%) cells (%) cellsa (%) cells (%)

PLB-dbcAMP 72 83 76.5 84
PLB-DMF 35 24.7 39 41.9
PLB-DMSO 84 78.5 85.3 90
a
without CB.

cellular surface. We first measured fMLF-induced release of MMP-9 was five-fold higher in dbcAMP-
increases in cell-surface expression of specific granule differentiated cells as compared to DMF- and DMSO-
membrane markers: CD63 for primary granules and differentiated cells. Similarly, liberation of albumin was
CD66b for secondary granules. Primary granule exocy- stronger for dbcAMP-differentiated cells as compared
tosis was assessed in the presence or the absence of CB, to DMF- and DMSO-induced cells, but remained lower
an inhibitor of actin filament formation that acts as a than albumin secretion by neutrophil after stimulation
priming agent (Bentwood and Henson 1980). As shown with fMLF. Taken together, these results indicate that
in Fig. 3A, stimulation with fMLF enhanced the exocytosis of tertiary granules and secretory vesicles can
expression of CD63 in differentiated PLB-985 cells be monitored in dbcAMP-differentiated cells.
without the requirement of CB pretreatment, whereas an
increase in CD63 cell-surface expression was observed in
neutrophils only after priming with CB (Fig. 3B). The PLB-985 organelles cannot be separated on Percoll
fMLF-induced expression of CD63 at the cell surface of density gradients
dbcAMP- and DMF-differentiated cells was not in-
duced by a pretreatment with CB, whereas that of Subcellular fractionation of PLB-985 cells, or of
DMSO-differentiated PLB-985 was enhanced by prim- HL-60 cells, has never been reported. We attempted
ing with CB (Fig. 3A and B). The increase in CD63 cell to separate the different granule populations and
surface expression in DMSO-induced PLB-985 cells was membranes of differentiated PLB-985 cells using a
three-fold superior to that observed in CB-primed three-layer discontinuous Percoll gradient, as already
neutrophils stimulated with fMLF. CD66b cell surface described for neutrophils (Kjeldsen et al., 1999). As
expression was very weak in differentiated PLB-985 shown in Fig. 5A for neutrophils (left panel), four bands
cells, and no increase of expression was observed upon were formed at the interface of the gradient steps.
stimulation with fMLF (Fig. 3C), suggesting that PLB- We observed a single band after centrifugation of
985 do not possess secondary granules. We also dbcAMP-differentiated PLB-985 cells (Fig. 5A, right
monitored cell-surface expression of CD11b, a marker panel). We obtained the same results using DMF- and
located in secondary and tertiary granules as well as in DMSO-differentiated cells (data not shown). Gradient
secretory vesicles (Borregaard and Cowland 1997; fractions were collected and characterised by immuno-
Borregaard et al., 2007). An increase in CD11b blotting for granule-specific and plasma membrane
expression was only observed on the surface of markers. Fig. 5B (upper panel), shows a typical
DMSO-differentiated PLB-985 cells and blood neutro- separation profile of blood neutrophils on a three-layer
phils stimulated with fMLF (Fig. 3D). The MFI Percoll gradient. The a-band contained primary gran-
multiplied by the percentage of cells expressing CD63 ules and was enriched in MPO, the b1-band contained
or CD11b was used to evaluate and normalise the levels secondary granules and was enriched in LF, and the
of CD63 or CD11b expression in differentiated PLB-985 b2-band contained tertiary granules and was enriched in
cells. The percentage of cells expressing CD63 or CD11b MMP-9. As already reported (Pivot-Pajot et al., 2008),
after differentiation for three days with dbcAMP, five plasma membrane and secretory vesicle markers,
days with DMSO and six days with DMF was similar to CD32A and albumin, respectively, are recovered in the
that observed for FPRL1 positive cells (Table 1). same fractions (g-band), whereas LDH is recovered in
Exocytosis of tertiary granules and secretory vesicles the cytosol fractions (C). Immunoblot analyses of
was more precisely monitored by measuring, by ELISA, gradient fractions of differentiated PLB-985 cells
MMP-9 and albumin in the supernatants of DMSO- revealed that all the markers of granules, as well as of
and fMLF-treated cells. Secretion of MMP-9 in secretory vesicles and plasma membrane, were recovered
response to stimulation with fMLF was very weak in the unique g-band (Fig. 5B, lower panel). LF was not
in all differentiated PLB-985 cells when compared to detected by Western blot in these gradient fractions,
neutrophils (Fig. 4A). Nevertheless, fMLF-induced thus confirming its non-expression in PLB-985 cells.
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 45

Immunoisolation of plasma membranes from

PLB-985 PLB-985 PLB-985 differentiated PLB-985 cells
dbcAMP DMF DMSO
IB : Nd D Nd D Nd D PMN
MW(kDa) Because organelles of differentiated PLB-985 cells were
α-CD32A - 40 not separable by density gradient centrifugation, we
attempted to separate granules from the plasma membrane
α-actin - 40
in these cells using an immunoisolation method, in which
PLB-985 organelles were immunoprecipitated using magnetic beads
coated with antibodies specific for granules or plasma
DM P
M

SO
membranes. Anti-CD32A antibody (IV.3) was used to
F
cA

DM

IB : Nd PMN
db

MW(kDa) specifically immunoprecipitate plasma membranes of

dbcAMP-differentiated PLB-985 cells. Cells were cavitated
α-CD16b - 55 using a nitrogen bomb, thus allowing the preservation of
the integrity of granules and plasma membranes. Im-
α-actin - 40 munoblots with anti-CD32A (CT10) revealed that plasma
membranes were efficiently immunoprecipitated (Fig. 6).
P
M

No signal was detected in immunoprecipitates using

SO
cA

F
DM

DM

Nd

IgG2b isotype control. The Src kinase Lyn, known to be

db
B

B
PL

PL

PL

PL

PMN associated with the plasma membrane in neutrophils or

RT - + - + - + - + - + PLB-985 cells, was detected in immunoprecipitated plasma
890 bp
membranes (Rollet-Labelle et al., 2004). Purity of
MPO
immunoprecipitated plasma membranes was assessed by
monitoring the presence of MPO, contained in primary
LF 830 bp
granules, and LDH, contained in cytosol. As shown by the
immunoblots in Fig. 6, MPO and LDH were detected only
MMP-9 660 bp
in the supernatants of immunoprecipitates, demonstrating
that immunoprecipitated plasma membrane fraction was
Albumin 660 bp not contaminated with either granules or cytosol. Inter-
estingly, secretory vesicles were not isolated with plasma
GAPDH 320 bp membranes since albumin was not detected in CD32A
immunoprecipitates. Similar results were obtained using
Fig. 2. Analysis of Fcg receptors and granule markers DMF- or DMSO-differentiated PLB-985 cells (data not
expression during PLB-985 differentiation. Granulocytic shown). As expected, we failed to isolate primary granules
differentiation of PLB-985 cells was induced with 0.3 mM using magnetic beads coated with an antibody against
dbcAMP for three days, with 1.25% DMSO for five days, or CD63, due to the fact that the epitope recognised by this
with 0.5% DMF for six days. Protein samples (30 mg per lane) antibody is likely inside the granule lumen and not
of undifferentiated and differentiated PLB-985 cells, as well as accessible. Similarly, an antibody against the extracellular
neutrophils, were resolved by SDS-PAGE (7.5–20% gradient domain of the specific granule membrane marker CD35
gels) and analysed by western blotting using antibodies against
was not able to immunoisolate secretory vesicles from the
CD32A (A) or CD16b (B), and an antibody against actin
(loading control). (C) Analysis of myeloperoxidase (MPO),
plasma membrane-depleted samples, thereby confirming
lactoferrin (LF), MMP-9 and albumin expression in neutro- the integrity of secretory vesicles. Nevertheless, we provide
phils and differentiated PLB-985 cells. PCR using specific here a interesting method for isolating plasma membranes
primers for MPO, LF, MMP-9 and albumin, as well as of differentiated PLB-985 cells, without contamination of
GAPDH as control, was performed on reverse-transcription granules or secretory vesicles, and studying the presence of
products (+) obtained from total RNA of the indicated cells, associated proteins. Immunoisolation of granules or
and compared to negative controls without () the reverse secretory vesicles from the plasma-membrane-depleted
transcriptase (RT). samples would be possible using beads coated with
antibodies against the cytoplasmic tail of granule markers
(Kumar et al., 1997).

Similarly to neutrophil fractionation, LDH was recov-

ered in the cytosol fractions (C). These results demon- PLB-985 cells differentiated with dbcAMP, DMF
strate that granules, secretory vesicles and plasma and DMSO have different phagocytic capacities
membranes probably have the same density in differ-
entiated PLB-985 cells, making them difficult to separate To get further insight into the functional responses of
by this method. differentiated PLB-985 cells relative to mature blood
 ARTICLE IN PRESS
46 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

Fig. 3. Analysis of fMLF-induced degranulation of differentiated PLB-985 cells. Granulocytic differentiation of PLB-985 cells was
induced with 0.3 mM dbcAMP for three days, with 1.25% DMSO for five days, or with 0.5% DMF for six days. Degranulation of
PLB-985 cells was monitored by measuring fMLF-induced increase in cell-surface expression of specific granule membrane markers.
Blood neutrophils and differentiated PLB-985 cells were resuspended in HBSS (1  106/ml) and stimulated for 10 min with 107 M
fMLF. All degranulation experiments were performed without cytochalasin B (CB), excepted for the experiment in (B) where cells
were preincubated for 5 min at 37 1C in presence of 10 mM CB. Degranulation was analysed by measuring CD63 (A, B) and CD66b
(C) cell-surface expression by FACS using FITC-coupled anti-CD63 and anti-CD66b antibodies or the appropriate isotype controls.
(D) CD11b cell-surface expression was also monitored using a mouse anti-CD11b antibody or the appropriate isotype control, and
an Alexa 488-conjugated goat anti-mouse IgG. The results are representative of three separate experiments and are the means7SD
from at least three experiments.

neutrophils, we analysed the phagocytic capacity of for differentiated PLB-985 cells, whereas neutrophils
these cells using confocal microscopy. Cells labelled with displayed a continuing increase in their phagocytic
the cytosolic fluorophore Calcein-AM were incubated capacity, and no levelling off tendency was observed,
with Alexa 594-labelled zymosan A opsonised particles. even at 10 min. Phagocytosis of differentiated PLB-985
As shown in Fig. 7A, differentiated PLB-985 cells, as cells was markedly weaker than that of neutrophils,
well as neutrophils, were able to internalise opsonised except for DMSO-induced cells, which was almost two-
zymosan particles. Typical horizontal (x versus y) as fold higher as compared to dbcAMP- and DMF-
well as vertical (y versus z) sections of these cells are induced cells. As observed by confocal microscopy,
shown, illustrating the distribution of the internalised undifferentiated PLB-985 cells displayed no or very
particles. In contrast, undifferentiated PLB-985 cells limited phagocytic capacity. Because the fluorescence of
failed to internalise opsonised zymosan particles despite Alexa 488-labelled zymosan particles is not modified by
their marked presence in the vicinity of the cells. Next, intraphagosomal pH (Lehmann et al., 1998; Panchuk-
we quantified phagocytosis by FACS analysis. After Voloshina et al., 1999), we were also able to quantify the
phagocytosis, 0.04% trypan blue was added to distin- mean number of internalised fluorescent particles per
guish between ingested and membrane-bound opsonised phagocytic cell (phagocytic index) by dividing the MFI
particles. We quantified the phagocytic capacity by associated with these cells by the fluorescence emitted
multiplying the percentage of phagocytosing cells by from single particle. For all cells, the patterns of the
their associated MFI. As illustrated in Fig. 7B, phagocytic index closely matched those of the phagocy-
phagocytic capacity reached a maximal level at 5 min tic capacity (Fig. 7C). At 10 min, the mean number of
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 47

3.0×107 DMSO
2.0×107 fMLF
MMP-9 release (pg/ml)

1.0×107

6000

4000

2000

0
PLB dbcAMP PLB DMF PLB DMSO PMN

8000 DMSO
5500 fMLF
Albumin release (pg/ml)

3000
500
200

100

0
PLB dbcAMP PLB DMF PLB DMSO PMN

Fig. 4. Analysis of fMLF-induced secretion of MMP-9 and

albumin by differentiated PLB-985 cells. Tertiary granule and
secretory vesicle exocytosis was analysed by measuring MMP-
9 and albumin secretion, respectively. Blood neutrophils and
differentiated PLB-985 cells were resuspended in HBSS
(1  106/ml) and stimulated for 10 min with 107 M fMLF.
The reaction was stopped by cooling and the suspension was
quickly centrifuged for 1 min at 10,000g at 4 1C. The cell-free
supernatants were collected and analysed for MMP-9 (A) and
albumin (B) by ELISA. The results are representative of three
separate experiments and are the mean7SD from at least three
experiments.

Fig. 5. Subcellular fractionation of nitrogen-cavitated neutro-

opsonised zymosan particles ingested per neutrophil was phils and differentiated PLB-985 cells on three-layer Percoll
5.670.4, while that of DMSO-induced cells reached gradients. After nitrogen cavitation, neutrophils and dbcAMP-
4.070.1 (Fig. 7C). Phagocytic index of DMF- and differentiated PLB-985 cell postnuclear supernatants were
dbcAMP-differentiated cells was markedly lower as fractionated on a three-layer Percoll gradient. (A) Photos of
compared to DMSO-induced cells and neutrophils. the three-layer gradients after centrifugation. The bands
Indeed, at 10 min the phagocytic index of dbcAMP- formed after centrifugation are marked. (B) The gradient
differentiated cells almost matched that of DMF- fractions (identified from 1 to 18) were collected and analysed
differentiated PLB-985 cells (2.170.1 and 2.370.1, for MPO, LF, MMP-9, albumin, CD32A (FcgRIIA) and LDH
respectively). Altogether, these results indicate that by immunoblotting. a-band corresponds to enrichment in
amongst the differentiating agents utilised, DMSO- primary granules, b1-band to secondary granules, and b2-band
to tertiary granules. For neutrophils, the g-band contains
differentiated cells show the strongest phagocytic
plasma membranes as well as secretory vesicles, whereas for
capacity. dbcAMP-differentiated PLB-985 cells g-band includes all the
granules and plasma membranes. Cytosol components were
located in the upper fraction of the gradient (C).
Discussion
studies, some of the functional responses of differen-
The human pro-myeloid cell line PLB-985 is a useful tiated PLB-985 cells have been analysed using one type
system for biological and molecular analysis of myeloid of differentiating agent. In this work, we have differ-
differentiation, and represents a suitable and very entiated PLB-985 cells along the granulocytic pathway
attractive model of neutrophil-like cells. In many using three different agents: dbcAMP, DMF or DMSO,
 ARTICLE IN PRESS
48 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

CD32A IgG2b criteria, the percentage of differentiated cells was 35%

IB : who IP SN IP SN
and 84% after treatment with DMF and DMSO,
MW(kDa) respectively. Using DMF, Pedruzzi et al. (2002)
reported, based on morphological criteria after Giemsa
α-CD32A - 43 staining, such as nucleus lobularity and cytoplasm
volume, that 91% of PLB-985 cells had undergone
granulocytic differentiation. Another study also demon-
strated that 80% of DMF-differentiated cells, in the
α-Lyn presence of 10% FCS, acquired respiratory burst
- 55
activity (Zhen et al., 1993). The discrepancy we observed
could be due to the fact that we conducted our
experiments in the presence of 10% FCS in the culture
α-MPO
- 55 medium, which has been shown to inhibit DMF-induced
differentiation of PLB-985 cells (Pedruzzi et al., 2002),
or depends on the functional response studied. Alto-
α-LDH - 34
gether, these results indicate that morphological criteria
are not sufficient to monitor the efficiency of PLB-985
α-albumin cell differentiation toward neutrophil-like cells. Inter-
- 70
estingly, elevation of cytosolic calcium under fMLF
Fig. 6. Immunoisolation of plasma membranes from differ- stimulation was higher in DMF-differentiated PLB-985
entiated PLB-985 cells. Immunoprecipitations using magnetic cells as compared to DMSO-differentiated cells, and the
beads coated with a monoclonal antibody against CD32A percentage of phagocytosing dbcAMP- and DMF-
(IV.3) or the appropriate isotype control (IgG2b) were carried induced cells was similar. Depending on the agent used,
out using dbcAMP-differentiated PLB-985 cells disrupted by PLB-985 cells were differentiated for various time. The
nitrogen cavitation as described in Materials and methods. chosen times of differentiation (three days for dbcAMP,
Following SDS-PAGE, proteins were transferred to PVDF five days for DMSO and six days for DMF) were
membranes which were then immunoblotted with anti- optimal regarding morphologic traits, fMLF-induced
CD32A, anti-Lyn, anti-MPO, anti-LDH and anti-albumin
calcium mobilisation and FPRL1 expression. Differen-
antibodies. These results are representative of three indepen-
tiation for up to nine days with all three inducing agents
dent experiments.
did not increase the percentage of differentiated cells,
and resulted in lower FPRL1 expression and reduction
and we have compared their capacity to mobilise of fMLF-induced calcium fluxes (data not shown).
intracellular calcium, to exocytose granule and secretory The expression of several granulocyte receptors such
vesicle content, and to phagocytose serum-opsonised as FPRL1 and FcgRIIA was upregulated during
particles. Differentiated PLB-985 cells display many of PLB-985 cell differentiation. Neutrophils express at
the functional characteristics associated with normal least two N-formyl peptide receptors, FPR and its
peripheral blood granulocytes such as calcium mobilisa- closely related homolog FPRL1 (Fu et al., 2008). These
tion and granule exocytosis upon stimulation with two receptors are not present in promeylocytes, suggest-
fMLF, as well as phagocytosis. Nevertheless, the ing that both are expressed fairly late during neutrophil
functional responses of differentiated PLB-985 cells maturation (Dahlgren et al., 2000). FPRL1 expression
were generally weaker than those of blood neutrophils. in PLB-985 cells treated with inducing agents indicates
As previously described for dbcAMP- and DMSO- that they have undergone myeloid maturation to
induced differentiation (Tucker et al., 1987), we found an advanced stage of neutrophil differentiation.
that PLB-985 cells, after treatment with the inducing DMF-differentiated PLB-985 cells and neutrophils were
agents described above, developed a characteristic shown to bind FITC-formyl peptide to similar levels
neutrophilic morphology such as the appearance of (Pedruzzi et al., 2002). Nevertheless, elevation of
numerous cytoplasmic and electron dense granules. cytosolic calcium after stimulation with fMLF was
Nevertheless, with the three types of differentiating weaker in differentiated PLB-985 cells compared to
agents, the cells showed unsegmented nuclei. The neutrophils, suggesting no direct relationship between
horseshoe-shaped appearance of the nuclei, suggest that the amounts of FPRs expressed at the cellular surface
differentiated PLB-985 cells have undergone differentia- and fMLF-induced cytosolic calcium mobilisation in
tion to metamyelocyte or band cell neutrophil matura- granulocytic cells. FPRL1 and FcgRIIA expression was
tion stage. After Wright’s staining, and based on nucleus markedly higher in dbcAMP-differentiated cells than in
morphological changes, we estimated that about 72% of neutrophils. dbcAMP is a cell-permeant cyclic nucleo-
PLB-985 cells induced with dbcAMP had undergone tide, and its intracellular hydrolysis leads to the
granulocytic differentiation. According to the same production of butyrate and cAMP. A previous study
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 49

PLB-985 Nd PLB-985 dbcAMP PLB-985 DMF

Scan: XZ XY XZ XY XZ XY
PLB-985 DMSO PMN

Scan: XZ XY XZ XY

6
Phagocytic index (particles per cell)

45
5
PMN PMN
(MFI x positive cells)
Phagocytic capacity

PLB Nd 4 PLB Nd
30 PLB dbcAMP PLB dbcAMP
PLB DMSO 3 PLB DMSO
PLB DMF PLB DMF

15 2

0 0
0 2 4 6 8 10 0 2 4 6 8 10
Time (min) Time (min)

Fig. 7. Analysis and quantification of phagocytosis by differentiated PLB-985 cells. Granulocytic differentiation of PLB-985 cells
was induced with 0.3 mM dbcAMP for three days, with 1.25% DMSO for five days, or with 0.5% DMF for six days. (A) Confocal
scanning laser microscopy of PLB-985 cells and neutrophils phagocytosing Alexa 594-conjugated opsonised zymosan particles (red).
Cells labelled with Calcein-AM (green) were allowed to sediment on glass coverslips for 30 min at 37 1C. Phagocytosis was initiated
by adding fluorescent particles to the coverslips in a ratio of 10 particles per cell. Confocal images were acquired after 30 min of
phagocytosis. For each cell type, vertical sections (y versus z) are shown in the left panel, and horizontal sections (x versus y) in the
right panel. Images are representative of three independent experiments. Bars, 6 mm. (B) FACS analysis of phagocytic capacity of
PLB-985 cells and neutrophils. Phagocytosis of Alexa 488-conjugated opsonised zymosan particles, in a ratio of 10 particles per cell,
was performed using cell suspensions, as indicated in Materials and methods. Phagocytic capacity was determined by multiplying the
number of positive phagocytosing cells by the mean fluorescence intensity (MFI). (C) Phagocytic index was calculated by dividing
the MFI obtained in (B) by the fluorescence emitted from single Alexa 488-conjugated particle. Each value represents the mean
value7SEM of four experiments each performed in duplicate. Each value was the result of counting 10,000 viable cells.

has shown that agents which increase intracellular dbcAMP- or DMSO-treated cells and neutrophils. As
cAMP, such dbcAMP, are inducers of a modified suggested by Chaplinski and colleagues, increases in
program of differentiation in HL-60 cells (Chaplinski intracellular butyrate and cAMP may have a positive
and Niedel 1982). Formyl-peptide receptor expression effect on chemotactic receptor expression. Likewise, we
was about three to four-fold higher in dbcAMP- also observed that CD32A expression was higher in
differentiated HL-60 as compared to DMSO-induced dbcAMP-treated PLB-985 cells suggesting that
HL-60 cells (Chaplinski and Niedel 1982). We observed dbcAMP hydrolysis product may upregulate the ex-
similar discrepancies in FPRL1 expression between pression of various receptors.
 ARTICLE IN PRESS
50 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

The respiratory burst is one well studied functional CB alone did not induce secretion of albumin, indicating
response of PLB-985 cells differentiated toward the that secretory vesicles of PLB-985 cells are less easily
granulocytic pathway (Bissonnette et al., 2008; Pedruzzi mobilisable, compared to those of neutrophils. Our data
et al., 2002; Pessach and Levy 2000). CGD-mutant PLB- suggest that dbcAMP-induced PLB-985 cells represent
985 cells, in which the NADPH oxidase complex is not the best mode of differentiation to study exocytosis of
functional, have been widely utilised as a useful model to tertiary granules and the release of albumin stored in
study the NADPH complex and associated ROS secretory vesicles. DMSO differentiation appears more
production (Bionda et al., 2004; Verchier et al., 2007). appropriate to monitor CD11b cell-surface expression in
More recently, Ear et al. (2008) reported that DMSO- PLB-985. With respect to priming by CB of fMLF-
differentiated PLB-985 cells respond like primary induced cell surface expression of the primary granule
neutrophils in terms of inflammatory cytokine produc- marker CD63, DMSO-differentiated PLB-985 cells
tion and transcription factor activation profile. Degra- appear to behave more like blood neutrophils.
nulation of PLB-985 cells has been studied in a Subcellular fractionation is an important tool in the
fragmentary way. Primary granule exocytosis was investigation of neutrophil granule composition and
reported in DMF- and DMSO-differentiated cells by mobilisation. Unlike neutrophils (Kjeldsen et al., 1999),
analyzing the release of b-glucuronidase or b-hexosami- fractionation of granule and intracellular membranes of
nidase after stimulation with fMLF or LTB4, respec- HL-60 or PLB-985 cells has never been reported.
tively (Gaudreault et al., 2005; Kaldi et al., 2003). However, granules and plasma membranes of differ-
Pedruzzi et al. (2002) described the functional degranu- entiated PLB-985 cells were not separable by classical
lation of specific granules and secretory vesicles in density centrifugation on a Percoll gradient due to the
DMF-induced PLB-985 cells, by analysing the fMLF- fact that all of these organelles displayed similar or very
induced increase in cell-surface expression of CD11b close densities. Despite numerous attempts, even by
and CD35. In this study, we report on the exocytosis modifying the Percoll density or using alternative
capacity of dbcAMP-, DMF- and DMSO-differentiated methods like differential centrifugation or Optiprep
PLB-985 cells in response to fMLF. Primary granule density gradient ultracentrifugation, we failed to isolate
exocytosis in all of the induced cells was comparable to different populations of granules from PLB-985 cells.
that of neutrophils primed with CB. Interestingly, Nevertheless, we provide an efficient method to isolate
treatment with CB, which is a prerequisite to observe plasma membranes based on immunoprecipitation with
primary granule mobilisation in neutrophils (Bentwood anti-FcgRIIA antibody. Proteomic analysis of plasma
and Henson 1980), was not necessary in differentiated membranes and secretory vesicle proteins in neutrophils
PLB-985 cells, suggesting that this compartment is more indicated that FcgRIIA is expressed only on plasma
easily mobilisable in these cells upon stimulation with membranes (Uriarte et al., 2008). Taking advantage of
fMLF. No increase in cell surface expression of the this, we propose a useful approach to separate plasma
specific secondary granule marker CD66b was observed membranes from other secretory granules, especially
after stimulation of differentiated PLB-985 cells with from secretory vesicles which were known to be
fMLF. Moreover, LF mRNA transcript was not frequently found in fractions enriched in plasma
expressed in PLB-985 cells, thereby suggesting, as membrane after subcellular fractionation. This method
reported for HL-60 cells, a lack of secondary granules can be of great interest for the study of receptors or
(Rado et al., 1987). LF regulates several neutrophil proteins associated with the plasma membrane.
functional responses like adhesion, oxidative metabo- Phagocytosis, a process by which microorganisms are
lism, and exocytosis of specific granule contents and engulfed and delivered to a digestive compartment,
appears to be important for the amplification of the represents a crucial event in the bactericidal function
initial and subsequent phases of the inflammatory of neutrophils (Lee et al., 2003). Neutrophils can
response. The impact of the apparent lack of secondary internalise opsonised particles which are recognised by
granules on PLB-985 cell functions has not been Fcg receptors (Lin et al., 1994). Although phagocytosis
investigated but a possible outcome could be reduced of opsonised zymosan or latex beads by DMF- and
bacterial killing. We can only conclude that PLB-985 is DMSO-differentiated PLB-985 cells has been analysed
not the appropriate cellular model to study this type of by confocal or phase contrast microscopy, quantifica-
granule, or that the cocktail of differentiating agents tion of phagocytosis has not been reported in these cells
used is not adapted to induce the biogenesis of (Perkins et al., 1991; Suh et al., 2006; van Bruggen et al.,
secondary granules. Secretory vesicles are the most 2004). PLB-985 cells differentiated with dbcAMP,
responsive and are secreted first, even before the DMF, and DMSO were able to phagocytose serum-
neutrophil reaches the site of infection (Sengelov et al., opsonised zymosan A particles, but their phagocytic
1995). We observed that treatment with CB alone led to capacity was weaker than that of neutrophils. In our
exocytosis of secretory vesicles in neutrophils (data not study, and as previously reported, we found that
shown). Interestingly, in differentiated PLB-985 cells, FcgRIIA (CD32) expression was markedly upregulated
 ARTICLE IN PRESS
C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52 51

during granulocytic differentiation of PLB-985 cells matous disease by means of transgenic PLB-985 cells. Hum.
(Hazan-Eitan et al., 2006). Despite high expression of Genet. 115, 418–427.
FcgRIIA in dbc-AMP induced PLB-985 cells, as Bissonnette, S.A., Glazier, C.M., Stewart, M.Q., Brown, G.E.,
compared to other differentiated PLB-985 cells or even Ellson, C.D., Yaffe, M.B., 2008. Phosphatidylinositol
blood neutrophils, their phagocytic capacity was among 3-phosphate-dependent and -independent functions of
the weakest, indicating that amount of FcgRIIA does p40phox in activation of the neutrophil NADPH oxidase.
J. Biol. Chem. 283, 2108–2119.
not correlate entirely with the phagocytic capacity of the
Borregaard, N., Cowland, J.B., 1997. Granules of the human
cells. We cannot exclude that other FcgR may be
neutrophilic polymorphonuclear leukocyte. Blood 89,
involved in phagocytosis in PLB-985 cells, and therefore 3503–3521.
PLB-985 cells represent an attractive model to study Borregaard, N., Sorensen, O.E., Theilgaard-Monch, K., 2007.
FcgR-mediated phagocytosis. A previous study re- Neutrophil granules: a library of innate immunity proteins.
ported, by FACS analysis of DMF-differentiated PLB- Trends Immunol. 28, 340–345.
985 cells, the expression of FcgRIIIb (CD16) in these Boulven, I., Levasseur, S., Marois, S., Pare, G., Rollet-Labelle,
cells (Selmeczy et al., 2003), but we failed to detect its E., Naccache, P.H., 2006. Class IA phosphatidylinositide
expression by Western blot in any of the differentiated 3-kinases, rather than p110 gamma, regulate formyl-
cells. High-affinity FcgRI (CD64) is expressed predo- methionyl-leucyl-phenylalanine-stimulated chemotaxis and
minantly after neutrophils have been primed with superoxide production in differentiated neutrophil-like
interferon (McKenzie and Schreiber 1998). Whether PLB-985 cells. J. Immunol. 176, 7621–7627.
FcgRI is expressed by differentiated PLB-985 cells, and Chaplinski, T.J., Niedel, J.E., 1982. Cyclic nucleotide-induced
contributes to phagocytosis, remains to be investigated. maturation of human promyelocytic leukemia cells. J. Clin.
Invest. 70, 953–964.
Undoubtedly, PLB-985 cells differentiated with DMSO
Collins, S.J., Gallo, R.C., Gallagher, R.E., 1977. Continuous
represent the best model the study phagocytosis since
growth and differentiation of human myeloid leukaemic
their phagocytic capacity and their phagocytic index are cells in suspension culture. Nature 270, 347–349.
closer to that of neutrophils. Dahlgren, C., Christophe, T., Boulay, F., Madianos, P.N., Rabiet,
In summary, we have demonstrated that differences M.J., Karlsson, A., 2000. The synthetic chemoattractant Trp-
exist in the efficiency of granulocytic differentiation, Lys-Tyr-Met-Val-DMet activates neutrophils preferentially
depending on the type of inducing agent used. The through the lipoxin A(4) receptor. Blood 95, 1810–1818.
functional responses that we monitored appear to be Ear, T., McDonald, P.P., 2008. Cytokine generation, promoter
strongly dependent on the type of the differentiating activation, and oxidant-independent NF-kappaB activation
agent used. DMSO-differentiated PLB-985 cells appear in a transfectable human neutrophilic cellular model. BMC
more suitable for the study of FcgR-mediated phago- Immunol. 9, 14.
cytosis and exocytosis of primary granules. In contrast, Faurschou, M., Borregaard, N., 2003. Neutrophil granules
dbcAMP-differentiated PLB-985 cells could be used to and secretory vesicles in inflammation. Microbes Infect. 5,
1317–1327.
study exocytosis of both tertiary granules and secretory
Fay, A.J., Qian, X., Jan, Y.N., Jan, L.Y., 2006. SK channels
vesicles.
mediate NADPH oxidase-independent reactive oxygen
species production and apoptosis in granulocytes. Proc.
Natl. Acad. Sci. USA 103, 17548–17553.
Acknowledgements Fu, H., Karlsson, J., Bjorkman, L., Stenfeldt, A.L., Karlsson,
A., Bylund, J., Dahlgren, C., 2008. Changes in the ratio
We thank Marie Duval for her precious technical between FPR and FPRL1 triggered superoxide production
in human neutrophils-a tool in analysing receptor specific
assistance for electron microscopy. We are grateful to
events. J. Immunol. Methods 331, 50–58.
Dr. Caroline Gilbert for her helpful comments, and to
Gaudreault, E., Thompson, C., Stankova, J., Rola-Pleszczyns-
Dr. Stéphanie Gout and Lynn Davis for their critical
ki, M., 2005. Involvement of BLT1 endocytosis and Yes
reading of this manuscript. We thank Vivien Berny for kinase activation in leukotriene B4-induced neutrophil
his kind assistance during manuscript submission. degranulation. J. Immunol. 174, 3617–3625.
Hazan-Eitan, Z., Weinstein, Y., Hadad, N., Konforty, A.,
Levy, R., 2006. Induction of Fc gammaRIIA expression in
References myeloid PLB cells during differentiation depends on
cytosolic phospholipase A2 activity and is regulated via
Bentwood, B.J., Henson, P.M., 1980. The sequential release of activation of CREB by PGE2. Blood 108, 1758–1766.
granule constitutents from human neutrophils. J. Immunol. Johnston, J.J., Rintels, P., Chung, J., Sather, J., Benz Jr., E.J.,
124, 855–862. Berliner, N., 1992. Lactoferrin gene promoter: structural
Bionda, C., Li, X.J., van Bruggen, R., Eppink, M., Roos, D., integrity and nonexpression in HL60 cells. Blood 79,
Morel, F., Stasia, M.J., 2004. Functional analysis of two- 2998–3006.
amino acid substitutions in gp91 phox in a patient with Kaldi, K., Kalocsai, A., Rada, B.K., Mezo, G., Molnar, G.Z.,
X-linked flavocytochrome b558-positive chronic granulo- Bathori, G., Ligeti, E., 2003. Degranulation and superoxide
 ARTICLE IN PRESS
52 C. Pivot-Pajot et al. / Immunobiology 215 (2010) 38–52

production depend on cholesterol in PLB-985 cells. Pivot-Pajot, C., Varoqueaux, F., de Saint Basile, G., Bourgoin,
Biochem. Biophys. Res. Commun. 310, 1241–1246. S.G., 2008. Munc13-4 regulates granule secretion in human
Katschinski, D.M., Robins, H.I., Schad, M., Frede, S., neutrophils. J. Immunol. 180, 6786–6797.
Fandrey, J., 1999. Role of tumor necrosis factor alpha in Rado, T.A., Wei, X.P., Benz Jr., E.J., 1987. Isolation of
hyperthermia-induced apoptosis of human leukemia cells. lactoferrin cDNA from a human myeloid library and
Cancer Res. 59, 3404–3410. expression of mRNA during normal and leukemic myelo-
Kjeldsen, L., Sengelov, H., Borregaard, N., 1999. Subcellular poiesis. Blood 70, 989–993.
fractionation of human neutrophils on percoll density Rollet-Labelle, E., Marois, S., Barbeau, K., Malawista, S.E.,
gradients. J. Immunol. Methods 232, 131–143. Naccache, P.H., 2004. Recruitment of the cross-linked
Kumar, A., Wetzler, E., Berger, M., 1997. Isolation and opsonic receptor CD32A (FcgammaRIIA) to high-density
characterization of complement receptor type 1 (CR1) detergent-resistant membrane domains in human neutro-
storage vesicles from human neutrophils using antibodies to phils. Biochem. J. 381, 919–928.
the cytoplasmic tail of CR1. Blood 89, 4555–4565. Selmeczy, Z., Szelenyi, J., Nemet, K., Vizi, E.S., 2003. The
Lee, W.L., Harrison, R.E., Grinstein, S., 2003. Phagocytosis inducibility of TNF-alpha production is different in the
by neutrophils. Microbes Infect. 5, 1299–1306. granulocytic and monocytic differentiated forms of wild
Lehmann, A.K., Halstensen, A., Bassoe, C.F., 1998. Flow type and CGD-mutant PLB-985 cells. Immunol. Cell Biol.
cytometric quantitation of human opsonin-dependent 81, 472–479.
phagocytosis and oxidative burst responses to meningo- Sengelov, H., Follin, P., Kjeldsen, L., Lollike, K., Dahlgren,
coccal antigens. Cytometry 33, 406–413. C., Borregaard, N., 1995. Mobilization of granules and
Lin, C.T., Shen, Z., Boros, P., Unkeless, J.C., 1994. Fc secretory vesicles during in vivo exudation of human
receptor-mediated signal transduction. J. Clin. Immunol. neutrophils. J. Immunol. 154, 4157–4165.
14, 1–13. Suh, C.I., Stull, N.D., Li, X.J., Tian, W., Price, M.O.,
Marcil, J., Harbour, D., Houle, M.G., Naccache, P.H., Grinstein, S., Yaffe, M.B., Atkinson, S., Dinauer, M.C.,
Bourgoin, S., 1999. Monosodium urate-crystal-stimulated 2006. The phosphoinositide-binding protein p40phox acti-
phospholipase D in human neutrophils. Biochem. J. 337, vates the NADPH oxidase during FcgammaIIA receptor-
185–192. induced phagocytosis. J. Exp. Med. 203, 1915–1925.
McKenzie, S.E., Schreiber, A.D., 1998. Fc gamma receptors in Taylor, R.M., Baniulis, D., Burritt, J.B., Gripentrog, J.M.,
phagocytes. Curr. Opin. Hematol. 5, 16–21. Lord, C.I., Riesselman, M.H., Maaty, W.S., Bothner, B.P.,
Newburger, P.E., Chovaniec, M.E., Greenberger, J.S., Cohen, Angel, T.E., Dratz, E.A., Linton, G.F., Malech, H.L.,
H.J., 1979. Functional changes in human leukemic cell line Jesaitis, A.J., 2006. Analysis of human phagocyte flavocy-
HL-60. A model for myeloid differentiation. J. Cell. Biol. tochrome b(558) by mass spectrometry. J. Biol. Chem. 281,
82, 315–322. 37045–37056.
Panchuk-Voloshina, N., Haugland, R.P., Bishop-Stewart, J., Tucker, K.A., Lilly, M.B., Heck Jr., L., Rado, T.A., 1987.
Bhalgat, M.K., Millard, P.J., Mao, F., Leung, W.Y., Characterization of a new human diploid myeloid leukemia
Haugland, R.P., 1999. Alexa dyes, a series of new cell line (PLB-985) with granulocytic and monocytic
fluorescent dyes that yield exceptionally bright, photostable differentiating capacity. Blood 70, 372–378.
conjugates. J. Histochem. Cytochem. 47, 1179–1188. Underhill, D.M., Ozinsky, A., 2002. Phagocytosis of microbes:
Pedruzzi, E., Fay, M., Elbim, C., Gaudry, M., Gougerot- complexity in action. Annu. Rev. Immunol. 20, 825–852.
Pocidalo, M.A., 2002. Differentiation of PLB-985 myeloid Urban, C.F., Lourido, S., Zychlinsky, A., 2006. How do
cells into mature neutrophils, shown by degranulation of microbes evade neutrophil killing? Cell Microbiol. 8,
terminally differentiated compartments in response to 1687–1696.
N-formyl peptide and priming of superoxide anion Uriarte, S.M., Powell, D.W., Luerman, G.C., Merchant, M.L.,
production by granulocyte-macrophage colony-stimulating Cummins, T.D., Jog, N.R., Ward, R.A., McLeish, K.R.,
factor. Br. J. Haematol. 117, 719–726. 2008. Comparison of proteins expressed on secretory vesicle
Perkins, S.I., Link, D.C., Kling, S., Ley, T.J., Teitelbaum, membranes and plasma membranes of human neutrophils.
S.L., 1991. 1,25-Dihydroxyvitamin D3 induces monocytic J. Immunol. 180, 5575–5581.
differentiation of the PLB-985 leukemic line and van Bruggen, R., Anthony, E., Fernandez-Borja, M., Roos,
promotes c-fgr mRNA expression. J. Leukoc. Biol. 50, D., 2004. Continuous translocation of Rac2 and the
427–433. NADPH oxidase component p67(phox) during phagocy-
Pessach, I., Levy, R., 2000. The NADPH oxidase diaphorase tosis. J. Biol. Chem. 279, 9097–9102.
activity in permeabilized human neutrophils and granulo- Verchier, Y., Lardy, B., Nguyen, M.V., Morel, F., Arbault, S.,
cytic like PLB-985 cells. Adv. Exp. Med. Biol. 479, Amatore, C., 2007. Concerted activities of nitric oxide
107–114. synthases and NADPH oxidases in PLB-985 cells. Bio-
Pessach, I., Shmelzer, Z., Leto, T.L., Dinauer, M.C., Levy, R., chem. Biophys. Res. Commun. 361, 493–498.
2006. The C-terminal flavin domain of gp91phox bound to Zhen, L., King, A.A., Xiao, Y., Chanock, S.J., Orkin, S.H.,
plasma membranes of granulocyte-like X-CGD PLB-985 Dinauer, M.C., 1993. Gene targeting of X chromosome-
cells is sufficient to anchor cytosolic oxidase components linked chronic granulomatous disease locus in a human
and support NADPH oxidase-associated diaphorase activ- myeloid leukemia cell line and rescue by expression of
ity independent of cytosolic phospholipase A2 regulation. recombinant gp91phox. Proc. Natl. Acad. Sci. USA 90,
J. Leukoc. Biol. 80, 630–639. 9832–9836.