Full text of DOI:10.1359/jbmr.2001.16.5.846

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 16, Number 5, 2001
© 2001 American Society for Bone and Mineral Research
Expression of a P2X7 Receptor by a Subpopulation of
Human Osteoblasts
A. GARTLAND,1,2 R.A. HIPSKIND,3 J.A. GALLAGHER,1 and W.B. BOWLER1
ABSTRACT
There is now conclusive evidence that extracellular nucleotides acting via cell surface P2 receptors are
important local modulators of bone cell function. Multiple subtypes of P2 receptors have been localized to
bone, where their activation modulates multiple processes including osteoblast proliferation, osteoblast-
mediated bone formation, and osteoclast formation and resorptive capacity. Locally released nucleotides also
have been shown to sensitize surrounding cells to the action of systemic factors such as parathyroid hormone
(PTH). In nonskeletal tissue recent attention has focused on one particular P2 receptor, the P2X7 receptor
(previously termed P2Z), and its ability to form nonselective aqueous pores in the plasma membrane on
prolonged stimulation. Expression of this receptor originally was thought to be restricted to cells of hemo-
poietic origin, in which it has been implicated in cell fusion, apoptosis, and release of proinflammatory
cytokines. However, recent reports have indicated expression of this receptor in cells of stromal origin. In this
study, we investigated the expression of the P2X7 receptor in two human osteosarcoma cell lines, as well as
several populations of primary human bone–derived cells (HBDCs) at the levels of messenger RNA (mRNA)
and protein. We found that there is a subpopulation of osteoblasts that expresses the P2X7 receptor and that
these receptors are functional as assessed by monitoring ethidium bromide uptake following pore formation.
Inhibition of delayed lactate dehydrogenase (LDH) release in response to the specific agonist 2؅,3؅-(4-benzoyl)-
benzoyl-adenosine triphosphate (BzATP) by the nonspecific P2X receptor antagonist PPADS confirmed a
receptor-mediated event. After treatment with BzATP SaOS-2 cells exhibited dramatic morphological changes
consistent with those observed after P2X7-mediated apoptosis in hemopoietic cells. Dual staining with terminal
deoxynucleotidyl transferase–mediated deoxyuridine triphosphate–biotin nick end labeling (TUNEL) and a
P2X7-specific monoclonal antibody confirmed the induction of apoptosis in osteoblasts expressing the P2X7
receptor. These data show for the first time the expression of functional P2X7 receptors in a subpopulation of
osteoblasts, activation of which can result in ATP-mediated apoptosis. (J Bone Miner Res 2001;16:846–856)
Key words: osteoblasts, P2X7, 2؅,3؅-(4-benzoyl)-benzoyl-adenosine triphosphate, immunocytochemistry, ap-
optosis
riety of processes as diverse as neurotransmission,(1) inflam-
INTRODUCTION
mation,(2) apoptosis,(3) and bone remodeling,(4,5) is now
widely accepted. Since the early work of Burnstock et al.,(6)
HE CONCEPT that adenosine triphosphate (ATP) can act
T
as an extracellular signaling molecule via interactions the number of characterized P2 receptors responsive to
with specific purinergic receptors and mediates a wide va- extracellular nucleotides has increased dramatically. These
1Human Bone Cell Research Group, Department of Human Anatomy and Cell Biology, The University of Liverpool, Liverpool, United
Kingdom.
2Department of Human and Biological Sciences, Liverpool Hope University College, Hope Park, Liverpool, United Kingdom.
3Institut de Ge´ne´tique Mole´culaire de Montpellier, UMR 5535, Centre National de la Recherche Scientifique, 34293 Montpellier Cedex
5, France.
846

P2X7 RECEPTOR IN BONE CELLS
847
P2 receptors originally were subdivided into two families, modulated positively by shear stress(25) also would imply
P2X and P2Y, based on their pharmacologic profile and that an autocrine/paracrine signaling mechanism involving
mechanism of signal transduction. However, after the recent nucleotides may contribute to the physiological response of
cloning of genes encoding individual P2 receptors, the basis skeletal tissue to mechanical stimuli. In osteoclasts extra-
for subdivision has been superseded by the molecular struc- cellular nucleotides have been shown to induce elevations
ture of receptors.(7) The ionotropic P2X receptors are of intracellular Ca2ϩ, increase the formation and resorptive
ligand-gated cationic channels, while the metabotropic P2Y activity of osteoclasts,(5) and induce apoptosis.(26) Indeed,
receptors form a distinct family of G-protein coupled recep- we have previously reported that expression of the P2X7
tors.(7) The P2 receptor family has widespread tissue distri- receptor messenger RNA (mRNA) is found in human oste-
bution and currently is comprised of seven distinct P2X oclastoma tissue and human osteoclastoma stromal cells,
receptors and five distinct P2Y receptors.
and that the specific P2X7 receptor agonist 2Ј,3Ј-(4-
The cloning of the P2X7 receptor (previously termed benzoyl)-benzoyl ATP (BzATP) potently inhibits resorption
P2Z) has generated considerable interest in this atypical by a population of human giant cells seeded onto dentine
P2X family member.(8) The P2X7 receptor is related struc- wafers by inducing apoptosis.(27) However, in the absence
turally to the other members of the P2X family, and as such, of specific P2 receptor antagonists it is still unclear which
when stimulated briefly with low concentrations of agonist, P2 receptors specifically couple to which remodeling pro-
this receptor acts as a nonselective cationic channel. How- cesses.
ever, when stimulated repeatedly or for prolonged periods
In this study we have investigated P2X7 receptor expres-
with higher agonist concentrations, the P2X7 receptor shows sion at the mRNA and protein level in two clonal osteosar-
the unique ability among members of the P2X family to coma cell lines, as well as populations of primary human
form large aqueous pores in the plasma membrane, perme- bone–derived cells (HBDCs). In addition, we have assessed
able to hydrophilic molecules as large as 900 Da.(9) Origi- receptor function in osteoblasts by monitoring pore forma-
nally, expression of the P2X7 receptor was thought to be tion in response to the potent P2X7 receptor agonist BzATP
restricted to cells of hemopoietic origin, in which receptor and linked changes in cellular morphology resulting from
activation has been implicated in mediating aspects of the receptor activation to the induction of apoptosis.
immune response. As such, in either macrophages or mi-
croglial cells, P2X7 receptor levels are functionally up-
regulated by lipopolysaccharide (LPS) or interferon-␥,(10)
and stimulation of the receptor leads to the release of mature
interleukin-1␤ (IL-1␤).(11) Recent reports have indicated
that cells of fibroblastic origin also express the P2X7 recep-
tor,(12) in which its activation may represent a novel path-
way for fibroblast activation and their recruitment in the
inflammatory response, by triggering release of the proin-
flammatory cytokine IL-6.(13) However, the best-recognized
consequence of prolonged P2X7 receptor activation and
pore formation in hemopoietic cells is that of ATP-induced
cell death.(14)
MATERIALS AND METHODS
Materials
Dulbecco’s modified Eagle’s medium (DMEM), RPMI
medium, fetal calf serum (FCS), and Superscript II reverse
transcriptase (RT) were purchased from Gibco Life Tech-
nologies (Paisley, Scotland). Taq DNA polymerase was
purchased from Advanced Biotechnologies (Epsom, UK).
The digoxigenin (DIG) RNA labeling kit, restriction en-
zymes, deoxynucleoside triphosphates (dNTPs), RNAse in-
hibitor, and In Situ Cell Death Detection Kit were from
Roche Diagnostics (Lewes, UK). NTPs and oligo(dT) were
purchased from Pharmacia (St. Albars, UK). Hybond N
hybridization membrane was purchased from Amersham
(Little Chalfont, UK). Nucleotides, ethidium bromide,
Hoechst 33258, and tetramethylrhodamine isothiocyamate
(TRITC)-conjugated phalloidin were obtained from Sigma
(Poole, UK). Cytotox-96 nonradioactive cytotoxicity assay
kit was purchased from Promega (Madison, WI, USA).
Vectastain ABC kit was from Vector Laboratories (Burlin-
game, CA, USA). The monoclonal P2X7 antibody was a
kind gift from Dr. Iain Chessell, Glaxo-Wellcome, Cam-
bridge, UK.
Osteoblasts and osteoclasts express multiple P2 receptor
subtypes,(15,16) with receptor profile seemingly a function of
cellular differentiation.(17) Consistent with this, locally re-
leased nucleotides have been shown to have wide ranging
and often opposing roles in bone remodeling.(4,5) In osteo-
blastic cells, extracellular nucleotides induce large transient
elevations of intracellular Ca2ϩ, consistent with an increase
in inositol 1,4,5-triphosphate.(18) Functional consequences
of P2 receptor activation in osteoblasts include induction of
the immediate-early gene c-fos,(19) an important regulatory
gene in skeletal cells,(20,21) stimulation of proliferation,(22)
and inhibition of bone formation.(4) One striking conse-
quence of P2 receptor activation in osteoblasts is the poten-
tiation of parathyroid hormone (PTH)–induced responses
such as Ca2ϩ signaling(23) and c-fos gene expression.(24)
These observations and the fact that nucleotides accumulate
Cell culture
HBDCs were isolated and cultured from explants of hu-
at sites of inflammation and injury may suggest a role for man bone as previously described.(28) In brief, specimens of
nucleotides as paracrine regulators of bone cell function in human bone were finely minced with a scalpel and then
diseases such as rheumatoid arthritis. Similarly, the obser- washed free of marrow cells with several volumes of me-
vations that osteoblasts can constitutively release nucleo- dium. The minced bone was cultured in 9-cm petri dishes
tides via a nonlytic mechanism and that this release can be containing DMEM supplemented with 10% FCS, 100

848
GARTLAND ET AL.
␮g/ml of streptomycin, 100 U/ml of penicillin, and 2 mM of described.(31) The PCR products were electrophoresed on a
L-glutamine. The cells were maintained at 37°C in a humid- 1% agarose gel and transferred to a Hybond N membrane
ified atmosphere of 95% air and 5% CO2 until required for according to the manufacturer’s instructions (Amersham).
assay or preparation of total RNA. SaOS-2 and Te85 cells The membrane was hybridized with 100 ng/ml of P2X7
were maintained in DMEM supplemented as above. THP-1 riboprobe in DIG Easy Hyb overnight. After hybridization,
cells were maintained in suspension culture in RPMI me- the membrane was washed stringently in 0.1% SSC/0.1%
dium supplemented as done previously. As with HBDC, the sodium dodecyl sulfate (SDS), and detected using disodium
cells were maintained at 37°C in a humidified atmosphere of 3-(4-methoxyspiro(1,2-dioxetane-3,2Ј-(5Ј-chloro)tricyclo-
95% air and 5% CO2 until required for assay or preparation [3.3.1.13,7decan]-4-yl)phenylphosphate (CSPD), again ac-
of total RNA.
cording to the manufacturer’s instructions (Roche Diagnos-
tics).
RNA isolation and complementary DNA synthesis
Immunocytochemistry
Total RNA was isolated using the guanidine isothiocya-
nate method.(29) Before first-strand complementary DNA
(cDNA) synthesis, RNA was DNAse-treated with RNAse-
free DNAse 1 (35 U/ml) for 60 minutes at 37°C, and
phenol-chloroform was extracted. Five micrograms of
DNAse-treated total RNA was used as template for first-
strand cDNA synthesis in a 20-␮l reaction containing 0.5
mM of dNTPs, 0.5 ␮g of oligo(dT), 20 U of RNAse
inhibitor, 10 mM of dithiothreitol, 6 mM of MgCl2, 40 mM
of KCl, 50 mM of Tris-HCl, pH 8.3, and 200 U of Super-
script II RT. The reaction mix was incubated at 42°C for 50
minutes, and the reaction was stopped by heating at 70°C
for 15 minutes. cDNA was stored at Ϫ20°C until required.
SaOS-2, Te85 and HBDC were seeded onto glass cover-
slips and incubated at 37°C in a humidified atmosphere of
95% air and 5% CO2 overnight. The media was removed
and cells were washed in phosphate-buffered saline (PBS)
before fixation by immersion in 3.7% formalin in PBS
(PBSF) for 10 minutes. The slides were then washed in PBS
and incubated for 20 minutes in blocking serum. Slides were
incubated for 1 h in either 10 ␮g/ml of monoclonal P2X7
antibody(32) or PBS. The slides were washed in PBS, and
the immunocomplexes were visualized using the ABC re-
agent and diaminobenzidine (DAB) substrate according to
the manufacturer’s protocol (Vector Laboratories and
Sigma, respectively). The sections were subsequently coun-
terstained with hematoxylin.
Polymerase chain reaction
Fifty-microliter polymerase chain reactions (PCRs) were
carried out containing 2 ␮l of the previously prepared
cDNA as template, 1 U of thermostable DNA polymerase,
50 pmol of each sense and antisense primer (see the fol-
lowing list), 0.2 mM each of deoxy-ATP (dATP),deoxycy-
tosine triphosphate (dCTP), deoxyguanosine triphosphate
(dGTP), and deoxythymidine triphosphate (dTTP), and 1.5
mM of MgCl2 in 1ϫ reaction buffer. Amplification condi-
tions were as follows: denaturation step at 94°C for 2
minutes, followed by 35 cycles at 94°C for 15 s, at 56°C for
30 s, at 72°C for 60 s, and a final 10-minute extension step
at 72°C. The following is a list of the primers:
Lactate dehydrogenase measurements
Cells were seeded at equal density into 96-well plates and
grown to 75% confluence. The media was replaced with
serum-free media overnight so cells would become quies-
cent. Antagonists were added 2 h before agonists. Treated
cells were then incubated for the required time period, after
which time the media was removed and assayed for LDH
activity using the CytoTox96 nonradioactive cytotoxicity
assay kit from Promega. To obtain total LDH values, con-
trol cells were lysed in 1% Triton X-100. Values are ex-
pressed as percentage of LDH released and represent the
fraction of LDH activity found in the supernatants with
respect to the total LDH value for control cells.
(1) P2X7 sense ϭ TGAAGGGGATAGCAGAGGTGA
(2) P2X7 antisense ϭ TGGGATGGCAGTGATGGA
(3) Nested P2X7 sense ϭ AAGGCCAAGAGCAGCGGT-
TGT
(4) Nested P2X7 antisense ϭ GCCCTGAATTGCCA-
CATCTGAAA
Ethidium bromide uptake
(5) gapdh sense-GGTGAAGGTCGGAGTCAACGG
(6) gapdh antisense-GGTCATGAGTCCTTCCACGAT
Subconfluent quiescent cells were trypsinized immedi-
ately before assay and resuspended in high potassium salt
(HKS) media (125 mM KCl, 1 mM EDTA, 5 mM glu-
cose, and 20 mM HEPES, pH 7.4) containing ethidium
bromide to a final concentration of 100 ␮M and appro-
DIG Riboprobe production and Southern analysis
A 553-base pair (bp) PCR product, amplified from mac- priate concentrations of agonist. Cells were then incu-
rophage cDNA using primer pairs 1 and 4 was confirmed by bated at 37°C with agitation for 60 minutes, after which
sequencing to be a fragment of the human P2X7 receptor. time the reaction was terminated by the addition of 0.5 ml
The 553-bp PCR product was then cloned into pBluescriptII of ice-cold 5 M NaCl. Ethidium bromide uptake was
KS using standard procedures.(30) A DIG-labeled riboprobe assessed by monitoring fluorescence intensity using a
was then generated by in vitro transcription as previously Flow Cytometer from Coulter Corp (Miami, FL, USA).

P2X7 RECEPTOR IN BONE CELLS
849
Morphological observations
Morphological changes in cells treated for LDH assays
were analyzed by phase-contrast microscopy. For immuno-
fluorescence analysis, cells were settled onto coverslips and
treated similarly. Cells were subsequently fixed in 3.7%
PBSF for 10 minutes, washed in PBS-Triton X-100 (0.1%),
and incubated at room temperature for 1 h with the DNA-
specific fluorochrome Hoechst 33258 (1 ␮g/ml) and
phalloidin-TRITC (4 ng/ml) to visualize the actin cytoskel-
eton. The terminal deoxynucleotidyl transferase–mediated
deoxyuridine triphosphate–biotin nick end labeling
(TUNEL) staining was performed in accordance with the
manufacturer’s protocol (Roche Diagnostics). The cells
were then analyzed by epifluorescence using a Leitz fluo-
rescence microscope coupled to a Princeton Instruments
Micromax CCD camera and captured using Metamorph
(Universal Imaging Systems, Inc., West Chester, PA, USA).
Images were converted and processed identically using
Adobe Photoshop.
Statistical analysis
Statistical analysis was by analysis of variance (ANOVA)
or randomized ANOVA; the significance between groups
was determined using a Tukey posttest. Logistic curve fit-
ting was performed with standard Excel software and a
modified form of the Hill equation.
RESULTS
Expression of P2X7 receptor mRNA by SaOS-2 and
primary HBDCs
FIG. 1. P2X7 mRNA expression in human bone cells. (A, top panel)
RT-PCR analysis of P2X7 mRNA in THP-1 cells, SaOS-2, and Te85
osteosarcoma cell lines and two populations of HBDCs, labeled B0020
and B0026. Thirty-five rounds of amplification with P2X7 primer pair
1 and 2 yielded an amplification product of the predicted size of 620 bp
for all cDNAs tested, unlike genomic DNA and water. (bottom panel)
Southern analysis of the amplification products shown in the top panel.
The membrane was probed with a DIG-labeled riboprobe correspond-
ing to a 553-bp fragment of the human P2X7 receptor to confirm
product specificity. (B, top panel). Reamplification of the PCR products
with nested P2X7 primer pair 3 and 4 yielded the predicted product of
412 bp, unlike the gapdh product amplification reaction. (bottom panel)
Corresponding Southern analysis of amplification products as de-
scribed in panel A. (C) PCR amplification with gapdh-specific primer
pair 5 and 6 yielded the predicted product of 519 bp for all cDNA
tested, confirming integrity and equal loading of cDNA in each sample.
Sizes on left refer to marker bands.
Osteoblasts have been described to express multiple P2
receptors,(15) and because of the recent report that cells of
stromal origin express the P2X7 receptor,(12) we used RT-
PCR to investigate whether P2X7 receptor transcripts were
expressed by the human osteosarcoma cell lines SaOS-2 and
Te85, as well as populations of primary osteoblastic cells
derived from explants of human bone. First-round RT-PCR
using P2X7-specific primer pair 1 and 2 generated the ex-
pected 620-bp amplification product in bone cells, albeit to
lower levels than the THP-1–positive control (Fig. 1A).
Reamplification of the first-round PCR products with the
nested P2X7 primer pair 3 and 4 generated the expected
412-bp amplification product (Fig. 1B). The housekeeping
gene gapdh was used to confirm the integrity and equal
loading of cDNA in each sample (primer pairs 5 and 6; Fig.
1C). Genomic DNA, water, and the gapdh PCR product
served as negative controls for the P2X7 amplification re-
actions. The PCR products hybridized strongly with a P2X7
DIG-labeled riboprobe showing that they arose specifically
from P2X7 cDNA.
performed immunocytochemistry on the same cells using
a specific P2X7 monoclonal antibody. In SaOS-2 cells a
distinct subpopulation of cells expressed high levels of
the P2X7 receptor (Fig. 2A, black arrows), and two
populations of HBDC tested showed varying degrees of
receptor expression (Figs. 2C and 2D). Interestingly, we
were repeatedly unable to detect receptor expression in
Immunolocalisation of P2X7 receptors
To confirm that the P2X7 receptor mRNA levels de- populations of Te85 cells (Fig. 2B). No staining was
scribed previously reflect receptor protein expression, we observed when the primary antibody was omitted or

850
GARTLAND ET AL.
FIG. 2. Immunocytochemical
localization of P2X7 receptor in
human osteoblasts. SaOS-2 cells,
Te85 cells, and HBDCs cultured
on glass coverslips were fixed in
PBSF and incubated with
a
monoclonal antibody raised
against the P2X7 receptor (10␮g/
ml). Immunocomplexes were re-
vealed using ABC/DAB staining,
and the sections were counter-
stained with hematoxylin. (A) A
distinct subpopulation of SaOS-2
cells, indicated by black arrows,
shows P2X7 receptor immuno-
staining. (B) Te85 cells show no
P2X7 receptor immunostaining.
(C and D) Two populations of
primary HBDCs show varying
degrees of P2X7 receptor immu-
nostaining. Original magnifica-
tion ϫ400 in all panels.
when a control osteoclast–specific antibody was used ity, a time course for LDH release was performed. Only
(data not shown).
after 24 h did 100 ␮M of BzATP induce LDH release (85%
relative release at 24 h compared with 2% and 10% at 6 h
and 12 h, respectively), while 1 mM of BzATP induced
20% and 37% relative LDH release from SaOS-2 cells at 6 h
and 12 h, respectively, with the maximal release of 87%
occurring at 24 h (Fig. 3C). Te85 cells showed no increased
release of LDH at any of the time points tested (data not
shown).
P2X7 receptor-induced responses in osteoblasts
A major consequence of P2X7 receptor activation in
hemopoietic cells is the formation of large nonselective
membrane pores and induction of apoptosis. Therefore, we
investigated the responses elicited by osteoblasts after P2X7
receptor agonist treatment. We used several criteria, includ-
ing the delayed release of the cytoplasmic enzyme LDH, the
uptake of the fluorescent dye ethidium bromide, and apo-
ptotic morphological changes.
SaOS-2 and Te85 cells were treated with the potent P2X7
receptor agonist BzATP, the putative partial P2X7 agonist
ATP, and the ineffective P2X7 agonist uridine triphosphate
(UTP), all at a concentration of 1 mM for 24 h. The P2X7
agonists significantly induced LDH release in SaOS-2 cells,
whereas UTP had no effect on LDH release (Fig. 3A). The
nonspecific antagonist PPADS,(33) at a concentration of 30
␮M, significantly inhibited LDH release from SaOS-2 cells
induced by 1 mM of BzATP, 0.5 mM of BzATP, and 1 mM
of ATP by 50, 56, and 75%, respectively. In Te85 cells, a
miniscule effect on LDH release was observed with BzATP,
which was slightly larger with ATP and UTP. Pretreatment
with PPADS appeared to augment rather than inhibit this
release (Fig. 3B).
To further assess pore formation, SaOS-2 and Te85 cells
were treated with BzATP, ATP, and UTP in a low divalent
ion solution containing ethidium bromide for 60 minutes at
37°C. Flow cytometry revealed increased uptake of
ethidium bromide in a subpopulation of agonist-treated
SaOS-2 cells (Fig. 4A). In contrast, no significant ethidium
bromide uptake was observed in Te85 cells studied (Fig.
4A). Although the response observed for individual HBDC
populations was somewhat variable (Fig. 4B), statistical
analysis of the mean data (randomized ANOVA) showed
that BzATP significantly increased ethidium bromide up-
take in HBDC (Fig. 4C). A typical P2X7 receptor agonist
potency order curve was obtained for a further HBDC
population B0023. As with the cell line THP-1, which is
known to homogeneously express high levels of the P2X7
receptor, BzATP was a more potent agonist than ATP (note
the right shift of the response curve from BzATP to ATP).
The maximal response of B0023 to BzATP was 60%, in-
dicative of a subpopulation expressing the P2X7 receptor.
B0023 P[A]50 values were BzATP, 5.5, and ATP, 4.13 (Fig.
4D).
LDH release has been used classically as a marker of
cytotoxicity, in which the early damage of the plasma mem-
brane would result in the fast release of the cells’ cytosolic
components into the culture medium. To confirm that the
release of the LDH after BzATP treatment occurred because
of slow release via the P2X7 receptor pores and/or because
of secondary necrosis accompanying P2X7 receptor–
BzATP treatment of osteoblasts also induced dramatic
changes to several aspects of cell morphology that are
consistent with P2X7 receptor–induced apoptosis. In
induced apoptosis as opposed to general agonist cytotoxic- SaOS-2 cells and primary HBDC, but not in Te85 cells,

P2X7 RECEPTOR IN BONE CELLS
851
extensive blebbing of the plasma membrane was clearly
visible after 40 minutes of treatment with 100 ␮M of
BzATP, which was prevented by pretreatment with the
antagonist PPADS (Fig. 5). Three hours post-BzATP treat-
ment, cells were fixed and stained with Hoechst 33258 and
phalloidin-TRITC to reveal nuclear and cytoskeleton
changes, respectively. Severe disruption of the actin cy-
toskeleton and nuclear condensation were evident in treated
SaOS-2 cells (Fig. 6, compare panel A with panel B and
panel C with panel D) but not in Te85 cells (Figs. 6E and
6F). After 12 h in the presence of BzATP, the majority of
SaOS-2 cells had rounded up and detached from the sub-
strate, whereas Te85 cells remained unaffected, even when
treated with 1 mM of BzATP (data not shown). ATP in-
duced similar changes, whereas UTP had no effect on cell
morphology (data not shown).
To further confirm that these observed morphological
changes were caused by P2X7 receptor–induced apoptosis,
dual staining with the P2X7 receptor antibody and the In
Situ Cell Death Detection Kit (TUNEL staining) was per-
formed. In a population of cells highly immunopositive for
the P2X7 receptor, positive TUNEL staining was observed
in all cells 6 h post-BzATP treatment (Fig. 7 compare panel
B with panel A and panel E with panel D). Pretreatment
with PPADS before BzATP treatment resulted in levels of
TUNEL staining observed with untreated cells (Fig. 7,
compare panel C with panel A and panel F with panel D).
DISCUSSION
In this study we present for the first time molecular,
immunocytochemical, and pharmacologic evidence for the
functional expression of the bifunctional P2X7 receptor by
human osteoblasts. Using RT-PCR and Southern analysis,
we were able to detect P2X7 mRNA transcripts in the
human osteosarcoma cell lines SaOS-2 and Te85, as well as
several populations of primary HBDCs obtained from pa-
tients of varying age, sex, and skeletal pathology. We then
used a P2X7-specific monoclonal antibody to confirm that
these mRNAs led to the expression of the P2X7 receptor
protein. In SaOS-2 cultures, a distinct subpopulation of cells
was found to be highly immunoreactive to the P2X7 anti-
body. This observation is consistent with our previous stud-
ies describing the heterogeneous nature of these cells with
FIG. 3. LDH release in response to P2X7 receptor activation. Cells
were seeded equally into 96-well plates and grown to 75% confluence. respect to expression of other P2 receptors.(34) Surprisingly,
Quiescent cells were then incubated in serum-free DMEM supple-
mented with different agonists and the antagonist PPADS as indicated.
After the indicated time period, the supernatant was assayed for LDH
activity. (A) LDH release from SaOS-2 cells in response to P2X7
receptor agonist stimulation. (B) LDH Release in Te85 cells in re-
sponse to P2X7 receptor agonist stimulation. Values are expressed as a
percentage of the total cellular LDH. Experiments were repeated three
times for both cell types. Data represent means Ϯ SEM; n ϭ 8; ***p Ͻ
0.001; significant increase in experimental values compared with MEM
only control, or significant inhibition in comparison to the treated
despite the expression of receptor mRNA, no P2X7 immu-
noreactivity could be detected in Te85 cultures. Although
the exact reason for this is unclear, it may reflect the
differentiation state of these osteoblasts before transforma-
tion. This lack of P2X7 receptor protein expression by Te85
cells refutes arguments that P2X7 receptor expression in
clonal cells is merely a consequence of cellular transforma-
tion and a by-product of in vitro culture (Dr. Chessell,
personal communication, 1999). The observation that cul-
sample without PPADS. (C) Time course of LDH release from SaOS-2 tured, nontransformed primary human osteoblasts showed
cells in response to P2X7 receptor agonist stimulation. Data represent
means Ϯ SEM; n ϭ 3.
varying degrees of P2X7 immunoreactivity provides further
support for the notion of differentiation-dependent P2X7
receptor expression in osteoblast populations.

852
GARTLAND ET AL.
To assess whether P2X7 receptors expressed by osteo-
blasts are functional, we next exploited the unique proper-
ties of this receptor to form large aqueous pores in the
plasma membrane when stimulated with repeated or pro-
longed applications of higher concentrations of agonist and
also to induce apoptosis. Consistent with expression of the
P2X7 receptor by SaOS-2 cells, 1 mM of BzATP and 1 mM
of ATP induced high levels of LDH release 24 h posttreat-
ment. Conversely, Te85 cells, which do not express the
P2X7 receptor protein, showed low levels of LDH release in
response to treatment with either agonist. The observed
release of LDH probably is caused by the gradual release
via P2X7 receptor pores and/or secondary necrosis known to
accompany apoptosis and as such is a receptor-mediated
event.(35) The fact that the cytosolic enzyme LDH was
retained by SaOS-2 cells for up to 6 h after BzATP treat-
ment would suggest an intact plasma membrane and the
release was not merely a consequence of general agonist
cytotoxicity. In agreement with this, the nonspecific P2X
antagonist PPADS reduced nucleotide-induced LDH release
from SaOS-2 cells by as much as 75%. Similarly, millimo-
lar concentrations of UTP, which is an ineffective P2X7
receptor agonist, did not induce LDH release from SaOS-2
cells and no significant LDH release could be detected at
earlier time points.
Another method of determining P2X7 receptor activation
and formation of transmembrane pores is to monitor the
uptake of the dye ethidium bromide, in which its fluorescent
properties change when bound to DNA. When combined
with flow cytometry, this method provided data on the
proportion of cells within populations that form transmem-
brane pores. As established by immunocytochemistry, there
appeared to be a distinct subpopulation of SaOS-2 cells that
formed pores when stimulated with BzATP. Similarly, three
populations of HBDC tested showed an ethidium bromide
uptake between 20 and 60%. This varied response by dif-
ferent populations of HBDC further supports our observa-
tions that these cells are heterogeneous with respect to P2
receptor expression.(34) These varied responses also could
reflect the expression of other P2 receptors by these cells, in
particular other P2X receptors. As with hemopoietic cells,
the ability of P2X receptors to form heterodimers with the
P2X7 receptor then would have both positive(36) and nega-
tive(37) effects on pore formation.
The P2X7 receptor expressed by osteoblasts has a some-
what atypical pharmacology in that the formation of the
bromide–positive SaOS-2 and Te85 cells after P2X7 receptor agonist
stimulation. (B) Percentage ethidium bromide–positive cells in several
populations of HBDC cells in response to P2X7 receptor agonist
stimulation. (C) Mean percentage ethidium bromide–positive cells in
FIG. 4. Ethidium bromide uptake in response to P2X7 receptor several populations of HBDC cells in response to P2X7 receptor
activation. Subconfluent, quiescent cells were trypsinized and resus-
pended in HKS buffer containing 100 ␮m ethidium bromide and the
indicated concentration of nucleotide. After incubation at 37°C for 60
minutes, reactions were terminated by the addition of 0.5 ml ice-cold 5
M NaCl, and the percentage of cells that were ethidium bromide
agonist stimulation. Data represent means Ϯ SEM; n ϭ 3; **p Ͻ 0.01,
*p Ͻ 0.5 significant increase in experimental values compared with
HKS only control. (D) Dose-response curve for BzATP or ATP-
induced ethidium bromide uptake in THP-1 and B0023 cells. P[A]50
values were determined using standard logistic curve fitting and a
positive was determined by flow cytometry. (A) Percentage of ethidium modified form of the Hill equation.

P2X7 RECEPTOR IN BONE CELLS
853
FIG. 5. Morphological changes associated with P2X7 activation in human osteoblasts. SaOS-2 cells and HBDCs were cultured onto glass
coverslips and treated with 100 ␮M of BzATP with and without pretreatment for 1 h with 30 ␮M of PPADS. Cells were then examined by
phase-contrast microscopy 40 minutes posttreatment. (A) BzATP-treated SaOS-2 cells; (B) PPADS-pretreated and BzATP-treated SaOS-2 cells;
(C) BzATP-treated HBDCs; (D) PPADS-pretreated and BzATP-treated HBDCs. Original magnification ϫ200.
FIG. 6. Morphological changes
associated with P2X7 activation
in human osteosarcoma cell lines.
SaOS-2 and Te85 were seeded
onto glass coverslips and treated
with 100 ␮M of BzATP. Three
hours posttreatment cells were
fixed in 3.7% PBSF, washed in
PBS-Triton X-100 (0.1%) and in-
cubated at room temperature for
1 h with (A–C) phalloidin-TRITC
and (D–F) the DNA-specific flu-
orochrome Hoechst 33258. The
cells were then analyzed by epi-
fluorescence using a Leitz IF mi-
croscope and appropriate filters.
(A and D) Untreated SaOS-2
cells; (B and E) BzATP-treated
SaOS-2 cells; (C and F) BzATP-
treated Te85 cells. Original mag-
nification ϫ600.
trademark transmembrane pore is delayed. A prolonged as a means of eliminating cells of hemopoietic origin
incubation of 60 minutes, as compared with 5 minutes with from isolated bone cell populations and fetal bone rudi-
hemopoietic cells, is required before ethidium bromide up- ments to leave an intact stromal cell population.(26) These
take can be detected in these cells. This delayed onset of findings imply that osteoblasts do not express functional
pore formation is a phenomenon that also has been reported P2X7 receptors; however, the data presented in this study
in human lymphocytes(38) and may account for previous clearly show that there is a distinct subpopulation of
observations that osteoblasts are insensitive to ATP4Ϫ in osteoblasts that do express functional P2X7 receptors.
this respect.(26)
The exact role of this receptor in modulating osteoblast
Although the expression of this pore-forming receptor cell function is unclear. What is clear is that repeated or
on cells of hemopoietic origin is well established and has prolonged exposure to higher concentrations of agonist
been implicated in the immune response, its expression leads to the formation of transmembrane pores and re-
by cells of stromal origin, in particular osteoblasts, has sults in eventual cell death. The process of bone remod-
not been investigated in detail. One report describes the eling is essential for the anatomical integrity and ho-
use of ATP4Ϫ, the putative agonist at the P2X7 receptor, meostasis of the skeleton, and critical to this is the

854
GARTLAND ET AL.
FIG. 7. P2X7 receptor –induced TUNEL staining in SaOS-2 cells. SaOS-2 were seeded onto glass coverslips and treated with BzATP, with and
without pretreatment for 1 h with 30 ␮M of PPADS. Six hours posttreatment cells were fixed in 3.7% PBSF, washed in PBS-Triton X-100 (0.1%),
and incubated at 37°C for 1 h with the TUNEL mix. The cells were subsequently washed in PBS-Triton X-100 (0.1%), incubated in blocking
serum, and then incubated with the P2X7 receptor antibody (10 ␮g/ml). After incubation with fluorescein isothiocyanate (FITC)–labeled secondary
antibody and the DNA-specific fluorochrome Hoechst 33258, the coverslips were rinsed in water, mounted in Mowiol, and then analyzed by
epifluorescence using a Leitz IF microscope and appropriate filters. (A and D) Untreated SaOS-2 cells; (B and E) BzATP-treated SaOS-2 cells;
(C and F) PPADS-pretreated and BzATP-treated SaOS-2 cells. Green ϭ FITC labeled P2X7 monoclonal antibody (mAb); blue ϭ Hoechst nuclear
stain; purple ϭ TUNEL-positive nuclei; yellow ϭ TUNEL-positive P2X7-positive cells. Original magnification ϫ400.
orderly genesis of osteoblasts and osteoclasts from their centrations of ATP in very localized pockets of the bone
precursors, as well as the removal of existing osteoblasts microenvironment, and as such could facilitate targeted
and osteoclasts from the bone surface.(39) Apoptosis, or osteoblast apoptosis.
programmed cell death, is a widely recognized method of
With a low prevalence of apoptosis in osteoblasts,(41)
conservation of a receptor in which its only function was
limiting the life span of cells in regenerative tissues,(40)
a
process that recently has been shown in osteoblasts both to induce cell death would appear extravagant and as
in vitro and in vivo.(41) Prolonged exposure of SaOS-2 such would evolve to redundancy. A much more perti-
cells to BzATP results in several morphological changes nent function of this receptor therefore would seem to be
typical of apoptosis, such as membrane blebbing, nuclear the controlled release of cytoplasmic components via
condensation, and cytoskeletal alterations. Apoptosis in rapid reversible pore formation in response to transient
P2X7 receptor–positive osteoblast cells was further high concentrations of ATP. Such transient high ATP
shown by dual labeling with the P2X7 monoclonal anti- concentrations would exist more readily throughout the
body and the apoptotic marker TUNEL. Therefore, as bone microenvironment, and the release of proresorption
observed with cells of hemopoietic origin, prolonged cytokines such as IL-1␤ and IL-6, which has been re-
P2X7 activation in osteoblasts would appear to be a ported to occur as a consequence of P2X7 receptor acti-
suicide response. The sustained high in vivo concentra- vation in other cell types abundantly expressing this
tions of ATP, in particular the tetrabasic ATP4Ϫ species, receptor,(11,13) would prove to be a more physiological
required for receptor activation and induction of apopto- relevant consequence of receptor activation in skeletal
sis in osteoblasts would appear unlikely in the bone tissues.
microenvironment. However, release of ATP from intra-
These data provide compelling evidence for the expres-
cellular stores in response to trauma or stress or by sion of the P2X7 receptor in a subpopulation of human
neighboring osteoblasts via a constitutive nonlytic mech- osteoblasts and suggests a possible role for this receptor in
anism,(25) could result in the accumulation of high con- osteoblast apoptosis. However, P2X7 receptor activation

P2X7 RECEPTOR IN BONE CELLS
855
osteoclastoma, bone and osteoblasts. J Bone Miner Res 10:
1137–1145.
and pore formation is almost certainly much more complex
and subtle in nature and as such requires further investiga-
tion.
16. Weidema AF, Dixon SJ, Sims SM 1998 Extracellular nucle-
otides activate P2X and P2Y purinoceptors on mammalian
osteoclasts. J Dent Res 77:B2178 (meeting abstract).
17. Bowler WB, Maier R, Bilbe G, Gallagher JA 1998 Osteoblast
express multiple G-protein coupled P2Y receptor subtypes.
Bone 22:2S. (meeting abstract).
ACKNOWLEDGMENTS
18. Scho¨fl C, Cuthbertson KSR, Walsh CA, Mayne C, Cobbold P,
Mu¨hlen AVZ, Hesch RD, Gallagher JA 1992 Evidence for P2
receptor on human osteoblast-like cells. J Bone Miner Res
7:485–491.
19. Dixon SJ, Sims SM 2000 P2 purinergic receptors on osteo-
blasts and osteoclasts: Potential targets for drug development.
Drug Dev Res 49:187–200.
20. Grigoriadis AE, Schellander K, Wang ZQ, Wagner EF 1993
Osteoblasts are target cells for transformation in c-fos trans-
genic mice. J Cell Biol 122:685–701.
The authors thank Dr. Chessell for the kind gift of the
P2X7 monoclonal antibody, Mandy Lawson for help with
flow cytometry, and Jane Dillon for assistance with HBDC
culture. This work was supported by the University of
Liverpool, Liverpool Hope University, and Association of
Clinical Biochemists (A.G.); the Association pour la Re-
cherche sur le Cancer (R.A.H.); and the Arthritis Research
Campaign (W.B.B.).
21. Hipskind RA, Bilbe G 1998 MAP kinase signaling cascades
and gene expression in osteoblasts. Front Biosci 3:D804–
D816.
22. Shimegi S 1996 ATP and adenosine act as mitogen for
osteoblast-like cells (MC3T3–E1). Calcif Tissue Int 58:108–
113.
23. Kaplan AD, Reimer WJ, Feldman RD, Dixon SJ 1995 Extra-
cellular nucleotides potentate the cytosolic Ca2ϩ, but not cy-
clic adenosine 3Ј,5Ј-monophosphate response to parathyroid
hormone in rat osteoblastic cells. Endocrinoclogy 136:1674–
1685.
24. Bowler WB, Dixon CJ, Halleux C, Maier R, Bilbe G, Fraser
WD, Gallagher JA, Hipskind RA 1999 Signaling in human
osteoblasts by extracellular nucleotides. Their weak induction
of the c-fos proto-oncogene via Ca2ϩ mobilization is strongly
potentiated by a parathyroid hormone/cAMP-dependent pro-
tein kinase pathway independently of mitogen-activated pro-
tein kinase. J Biol Chem 274:14315–14324.
25. Bowler WB, Tattersall JA, Hussein R, Dixon CJ, Cobbold PH,
Gallagher JA 1998 Release of ATP by osteoblasts: Modulation
by fluid shear forces. Bone 22:3S (meeting abstract).
26. Modderman WE, Weidema AF, Vrijheid-Lammers T, Was-
senaar AM, Nijweide PJ 1994 Permeabilization of cells of
hemopoietic origin by extracellular ATP4Ϫ: Elimination of
osteoclasts, macrophages, and their precursors from isolated
bone cell populations and fetal bone rudiments. Calcif Tissue
Int 55:141–150.
27. Gartland A, Ginty AF, Gallagher JA, Bowler WB 1999 Acti-
vation of P2X7 receptors expressed by human osteoclastoma
modulates bone resorption. Calcif Tissue Int 64(Suppl 1):S56
(meeting abstract).
28. Gallagher JA, Gundle R, Beresford JN 1996 Isolation and
culture of bone-forming cells (osteoblasts) from human bone.
In: Jones GE (ed.) Methods in Molecular Medicine: Human
Cell Culture Protocols. Humana Press Inc., Totowa, NJ, USA,
pp. 233–262.
29. Chomczynski P, Sacchi N 1987 Single-step method of RNA
isolation by acid guanidium thiocyanate-phenol-chloroform
extraction. Anal Biochem 162:156–161.
REFERENCES
1. Edwards FA, Gibb AJ, Colquhoun D 1992 ATP receptor-
mediated synaptic currents in the central nervous system.
Nature 359:144–147.
2. Perregaux D, Gabel CA 1994 Interleukin-1␤ maturation and
release in response to ATP and nigericin. J Biol Chem 21:
15195–15203.
3. Chow SC, Kass GEN, Orrenius S 1997 Purines and their role
in apoptosis. Neuropharmacology 36:1149–1156.
4. Jones SJ, Gray C, Boyde A, Burnstock G 1997 Purinergic
transmitters inhibit bone formation by cultured osteoblasts.
Bone 21:393–399.
5. Morrison MS, Turin L, King BF, Burnstock G, Arnett TR
1998 ATP is a potent stimulator of the activation and forma-
tion of rodent osteoclasts. J Physiol 511:495–500.
6. Burnstock G 1972 Purinergic nerves. Pharmacol Rev 24:509–
581.
7. Fredholm BB, Abbracchio MP, Burnstock G, Daly JW,
Harden TK, Jacobson KA, Leff P, Williams M 1994 Nome-
clature and classification of purinoceptors. Pharmacol Rev
64:143–156.
8. Suprenant A, Rassendren F, Kawashima E, North RA Buell G
1996 The cytolytic P2Z receptor for extracellular ATP identi-
fied as a P2X receptor (P2X7). Science 272:735–738.
9. Cockcroft S, Gomperts BD 1979 ATP induces nucleotide
permeability in rat mast cells. Nature 279:541–542.
10. Humphreys BD, Dubyak GR 1996 Induction of the P2Z/P2X7
nucleotide receptor and associated phospholipase D activity by
lipopolysaccharide and INF␥ in the human THP-1 monocytic
cell line. J Immunol 157:5627–5637.
11. Ferrari D, Chiozzi P, Falzoni S, Dal Susino M, Melchiorri L,
Baricordi OR, Di Virgilio F 1997 Extracellular ATP triggers
IL-1␤ release by activating the purinergic P2Z receptor in
human macrophages. J Immunol 159:1451–1458.
12. Groschel-Stewart U, Bardini M, Robson T, Burnstock G 1999
Localisation of P2X5 and P2X7 receptors by immunohisto-
chemistry in rat stratified squamous epithelia. Cell Tissue Res
296:599–605.
13. Solini A, Chiozzi P, Morelli A, Fellin R, Di Virgilio F 1999
Human primary fibroblast in vitro express a purinergic P2X(7)
receptor counted to ion fluxes, microvesicle formation and
IL-6 release. J Cell Sci 112:297–305.
30. Sambrook J, Fritsch EF, Maniatis T 1989 Plasmid vectors. In:
Molecular Cloning: A Laboratory Manual. Cold Spring Harbor
Laboratory Press, Cold Spring Harbor, NY, USA, p. 1.21.
31. Rogers M, Sun M, Lerner MR, Vogt RG 1997 Snmp-1, a novel
membrane protein of olfactory neurons of the silk moth An-
theraea polyphemus with homology to the CD36 family of
membrane proteins. J Biol Chem 272:14792–14804.
14. DiVirgilio F 1995 The P2Z purinoceptor: An intriguing role in
immunity, inflammation and cell death. Immunol Today 16: 32. Buell G, Chessell IP, Michel AD, Collo G, Salazzo M, Herren
524–528.
S, Gretener D, Grahames C, Kaur R, Kosco-Vilbios MH,
Humphrey PPA 1998 Blockade of human P2X7 receptor func-
tion with a monoclonal antibody. Blood 92:3521–3528.
15. Bowler WB, Birch MA, Gallagher JA, Bilbe G 1995 Identifi-
cation and cloning of human P2U purinoceptor present in

856
GARTLAND ET AL.
33. Chessell IP, Michel AD, Humphrey PPA 1998 Effects of 39. Gallagher JA, Dillon JP 1997 Bone remodeling. In: Dulbecco
R (ed.) Encyclopedia of Human Biology, 2nd ed., vol. 2.
Academic Press, San Diego, CA, USA, pp. 135–149.
40. Steller H 1995 Mechanisms and genes of cellular suicide.
Science 267:1445–1449.
41. Jilka RL, Weinstein RS, Bellido T, Parfitt AM, Manolagas SC
1998 Osteoblast programmed cell death (apoptosis): Modula-
tion by growth factors and cytokines. J Bone Miner Res
13:793–802.
antagonists at the human recombinant P2X7 receptor. Br J
Pharmacol 124:1314–1320.
34. Dixon CJ, Bowler WB, Walsh CA, Gallagher JA 1997 Effects
of extracellular nucleotides on single cells and populations of
human osteoblasts: Contribution of cell heterogeneity to rela-
tive potencies. Br J Pharmacol 120:777–780.
35. Schulze-Lohoff E, Hugo C, Rost S, Arnold S, Gruber A, Bru¨ne
B, Bernd Sterzel R 1998 Extracellular ATP causes apoptosis
and necrosis of cultured mesangial cells via P2Z/P2X7 recep-
tors. Am J Physiol 275:F692–F971.
36. Chessell IP, Michel AD, Humphrey PPA 1997 Properties of
the pore forming P2X7 purinoceptor in mouse NTW8 micro-
glial cells. Br J Pharmacol 121:1429–1437.
37. Petrou S, Ugur M, Drummond RM, Singer JJ, Walsh JV 1997
P2X7 purinoceptor expression in Xenopus oocytes is not suf-
ficient to produce a pore-forming P2Z-like phenotype. FEBS
Lett 411:339–345.
38. Fernando KC, Gargett CE Wiley JS 1999 Activation of the
P2Z/P2X7 receptor in human lymphocytes produces a delayed
permeability lesion: Involvement of phospholipase D. Arch
Biochem Biophys 362:197–202.
Address reprint requests to:
Alison Gartland
Human Anatomy and Cell Biology Group
Department of Human Anatomy and Cell Biology
The University of Liverpool
Ashton Street
Liverpool L69 3GE, UK
Received in original form January 19, 2000; in revised form
October 17, 2000; accepted October 25, 2000.