Full text of DOI:10.1359/jbmr.2001.16.3.466

JOURNAL OF BONE AND MINERAL RESEARCH
Volume 16, Number 3, 2001
© 2001 American Society for Bone and Mineral Research
Caspase-Dependent Cleavage of Cadherins and Catenins
During Osteoblast Apoptosis
IRENE HUNTER,1,2 DUNCAN McGREGOR,1 and SIMON P. ROBINS1
ABSTRACT
As transmembrane, Ca2؉-dependent cell-cell adhesion molecules, cadherins play a central role in tissue
morphogenesis and homeostasis. Stable adhesion is dependent on interactions of the cytoplasmic domain of the
cadherins with a group of intracellular proteins, the catenins. In the present study, we have detected the
expression of ␣-, ␤-, and ␥-catenins in human osteoblasts, which assemble with cadherins to form two distinct
complexes containing cadherin and ␣-catenin, with either ␤- or ␥-catenin. In osteoblasts undergoing apopto-
sis, proteolytic cleavage of N-cadherin and ␤- and ␥- catenins but not ␣-catenin was associated with the
activation of caspase-3 and prevented by the caspase inhibitor Z-VAD-fmk. The pattern of cadherin/catenin
cleavage detected in apoptotic osteoblasts was reproduced in vitro by recombinant caspase-3. The presence of
a 90-kDa extracellular domain fragment of N-cadherin in conditioned medium from apoptotic cells indicates
that additional extracellular or membrane-associated proteases also are activated. Disruption of N-cadherin–
mediated cell-cell adhesion with function-blocking antibodies induced osteoblast apoptosis, activation of
caspases, and cleavage of ␤-catenin. These findings provide compelling evidence that N-cadherin–mediated
cell-cell adhesion promotes osteoblast survival and suggest that the underlying mechanism may involve
activation of ␤-catenin signaling. (J Bone Miner Res 2001;16:466–477)
Key words: osteoblast, apoptosis, caspase, cadherin, catenin
Apoptosis is regulated by a complex set of environmental
signals. In addition to the well-characterized growth factor
signaling,(3) recent evidence suggests that adhesion-mediated
signaling also may regulate the apoptotic process.(4–7) The
disruption of cell-matrix and cell-cell adhesion is a feature
of apoptosis in involuting tissues(8) and is an early event in
cultured cells undergoing apoptosis.(9,10) In osteoblasts,
cell rounding and detachment has been detected during
phosphatase-induced(11) and Fas-mediated(12) apoptosis. In
anchorage-dependent cells, attachment to the extracellular
INTRODUCTION
KELETAL REMODELING, the process whereby old bone is
replaced by new bone, continues throughout adult life,
S
by the concerted action of bone-resorbing osteoclasts and
bone-forming osteoblasts. On completion of their task,
some osteoblasts become embedded within the bone matrix
as osteocytes and others form flattened lining cells on the
bone surface. However, as recent studies have shown, the
fate of the majority of osteoblasts is to undergo apoptosis.(1)
Apoptosis is a controlled mechanism of cell death, neces- matrix has been shown to transmit signals that suppress
sary to maintain tissue homeostasis and characterized by apoptosis.(13,14) Recent data, linking cadherin-mediated ad-
ultrastructural changes that include membrane blebbing, hesion to survival of ovarian granuloma cells(15,16) and
chromatin condensation, nuclear fragmentation, and forma- oral squamous cell carcinoma,(17) suggest that cadherin-
tion of apoptotic bodies.(2)
dependent adhesion also could transduce antiapoptotic sig-
1Matrix Biochemistry Group, Rowett Research Institute, Aberdeen, Scotland.
2Present address: Department of Biomedical Sciences, Institute of Medical Sciences, University of Aberdeen, Foresterhill, Aberdeen,
Scotland.
466

CELL ADHESION MOLECULES IN BONE
467
nals. Previous studies have indicated that attachment to the chemicals were from Sigma (Poole, Dorset, UK), unless
extracellular matrix protein fibronectin promotes osteoblast otherwise indicated.
survival,(18) but currently there are no reports on the role of
cadherin-mediated cell-cell adhesion in the regulation of
Cell culture
osteoblast apoptosis.
The cadherin family of Ca2ϩ-dependent transmembrane
Primary osteoblast cultures were established from trabec-
proteins play an essential role in the establishment and ular bone explants as described previously.(36) Briefly, tra-
maintenance of cell-cell contacts.(19,20) The extracellular becular bone particles, obtained at orthopedic operation
domain of the cadherins mediates homophilic interactions from two donors with no known bone disease, were ex-
with cadherins on neighboring cells,(21) whereas the cyto- tracted sequentially with EDTA and trypsin and the cells
plasmic domain interacts with a group of intracellular pro- released by 0.2% collagenase in phosphate-buffered saline
teins, the catenins.(22,23) Both ␤- and ␥-catenins can interact (PBS) were maintained in ␣-modified essential medium
directly with cadherins, linking them to ␣-catenin, which in (␣-MEM) containing 10% FBS with 100 U/ml penicillin
turn interacts with the actin cytoskeleton.(24–27) In addition, and 100 ␮g/ml streptomycin at 37°C in 5% CO2 humidified
both ␤- and ␥-catenins act as intracellular signaling pro- atmosphere. Cells prepared in this way were shown to
teins, which can interact with the lymphoid-enhancer bind- exhibit an osteoblastic phenotype as judged by high levels
ing factor/T cell factor (LEF/TCF) family of transcription of production of alkaline phosphatase, collagen type I, and
factors, to regulate gene transcription.(28–32)
osteocalcin.
Osteoblasts have been reported to express a limited rep-
The human osteoblastic cell lines MG-63, TE-85, and
ertoire of cadherins,(33–35) which contribute to cell-cell ad- SaOS-2 were obtained from the American Type Culture
hesion and the development of the differentiated phenotype. Collection (ATCC) and maintained in Dulbecco’s modified
In the present study, we have examined the expression and Eagle’s medium (DMEM), supplemented with 10% FBS, 2
organization of cadherins and catenins in human osteoblasts mM L-glutamine, 100 U/ml penicillin, and 100 ␮g/ml strep-
and followed the fate of these proteins during both sponta- tomycin, at 37°C in 5% CO2 humidified atmosphere.
neous and induced apoptosis. Our results indicate that both
cadherins and catenins are targeted for cleavage during
apoptosis. In addition, the induction of apoptosis by
Reverse transcription and polymerase chain reaction
function-blocking N-cadherin antibodies suggests that
Messenger RNA (mRNA) was isolated from approxi-
N-cadherin–mediated cell-cell adhesion is a novel mecha- mately 107 cells for each cell line using the polyATract
nism whereby osteoblasts can regulate apoptosis.
1000 mRNA isolation system (Promega, Inc., Southampton,
UK). First-strand complementary DNA (cDNA) synthesis
was performed at 37°C for 1 h, with 1 ␮g of mRNA primed
with 25 pmol oligo (dT)15, in a 50-␮l reaction volume
containing 400 U Superscript II reverse transcriptase (Life
Technologies Ltd., Paisley, UK), 10 mM dithiothreitol
(DTT), 12.5 mM Tris-HCl (pH 8.3), 18.75 mM KCl, and
MATERIALS AND METHODS
Antibodies and reagents
Actinomycin D and staurosporine were from Sigma (Poole, 0.75 mM MgCl2. Cadherin cDNA was polymerase chain
Dorset, UK). Genistein was from Lancaster Synthesis, Ltd. reaction (PCR) amplified from 5 ␮l of first-strand cDNA
(Morecombe, UK). The caspase-3 substrate N-acetyl-Asp-Glu- reaction, using 25 pmol of each cadherin family degenerate
Val-Asp-p-nitroanilide (Ac-DEVD-pNA) was purchased from primer (5Ј-ACNGCNCCNCCNTAYGA and 5Ј-TCNG-
Biomol (Plymouth Meeting, PA, USA). The caspase-1 sub- CNARYTTYTTRAA)(37) in 50-␮l reaction volumes con-
strate N-acetyl-Tyr-Val-Ala-Asp-p-nitroanilide (Ac-YVAD- taining 2.5 U Taq DNA polymerase (Promega, Inc.), 100
pNA) and the caspase inhibitors N-acetyl-Asp-Glu-Val-Asp- ␮M deoxynucleoside triphosphate (dNTP), 120 ␮M MgCl2,
aldehyde (Ac-DEVD-CHO) and benzyloxycarbonyl-Val- 50 mM KCl, 10 mM Tris-HCl (pH 9.0), and 0.1% Triton
Ala-Asp-fluoromethyl ketone (Z-VAD-fmk) were from X-100. Amplification was carried out on a Techne Progene
Calbiochem (Nottingham, UK). Recombinant human thermal cycler (Scotlab, Coatbridge, UK) for 30 cycles at
caspase-3 was from Pharmingen (San Diego, CA, USA).
94°C for 1 minute, at 50°C for 1 minute, and at 72°C for 2
Monoclonal antibodies to ␤-catenin, plakoglobin (␥- minutes. The 160-base pair (bp) reaction product was gel
catenin), desmoglein, and caspase-3 (CPP32) were from purified using a MERMAID DNA purification kit (BIO 101,
Transduction Laboratories (Lexington, KY, USA). Poly- Inc., Vista, CA, USA) according to the manufacturer’s
clonal antibodies to cadherins (pan-cadherin) and ␤-catenin, instructions and ligated overnight at 16°C into 50 ng of
a monoclonal antibody recognizing the extracellular domain pGEM-T vector (Promega, Inc.) as described by the man-
of N-cadherin (anti-A-CAM; clone GC-4) and horseradish ufacturer. Precipitated DNA was electroporated into TG-1
peroxidase (HRP)–conjugated and fluorescein isothiocya- Escherichia coli bacteria and plated onto LB-agar plates
nate (FITC)–labeled secondary antibodies were purchased containing 100 ␮g/ml ampicillin, 1 mM isopropylthio-␤-D-
from Sigma. A monoclonal antibody (3B9) recognizing the galactoside (IPTG), and 10 ␮g/ml 5-bromo-4-chloro-3-
cytoplasmic of N-cadherin was obtained from Zymed Lab- indolyl-␤-D-galactoside (X-gal) and incubated overnight at
oratories (San Francisco, CA, USA). Tissue culture media 37°C. Plasmid DNA was isolated from insert containing
and fetal bovine serum (FBS) were from Life Technologies colonies from 1.5 ml of overnight culture using a Wizard-SV
Ltd (Paisley, UK). Tissue culture supplements and all other kit (Promega, Inc.), and insert sequence was determined with

468
HUNTER ET AL.
M13FOR primer using a BigDye terminator sequencing kit at 4°C with 2.5–5 ␮g of antibody. When monoclonal anti-
(PE-Applied Biosystems, Warrington, Cheshire, UK).
bodies were used, rabbit anti-mouse immunoglobulin G
(IgG) was added, before incubation with a 50% slurry of
protein A–Sepharose. After constant mixing at 4°C for 1–2
h, samples were centrifuged and pellets were washed ex-
tensively with lysis buffer. For immunoblotting, cell lysates
were mixed with an equal volume of 2ϫ reducing sodium
dodecyl sulfate (SDS) sample buffer(38) and incubated at
100°C for 5 minutes. Equal amounts of protein were separated
by SDS-polyacrylamide gel electrophoresis (PAGE) and trans-
ferred to nitrocellulose membranes. After blocking with 5%
milk powder/10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.05%
Tween 20, membranes were incubated for 1 h at RT with
primary antibody. Specific antibody binding was detected with
HRP-conjugated secondary antibodies and developed using
enhanced chemiluminescence.
Immunofluorescence microscopy
Cells, grown on glass coverslips, were fixed in cold
(Ϫ20°C) methanol/acetone (1:1, vol:vol) for 7 minutes,
blocked with 1% bovine serum albumin (BSA) in PBS for
30 minutes, and incubated with primary antibodies for 1 h at
room temperature (RT). After three washes with 1% BSA in
PBS, cells were incubated with FITC-labeled secondary
antibodies for 1 h at RT. Coverslips were mounted on glass
slides in antifading reagent (Vectashield; Vector Laborato-
ries, Inc., Burlinghame, CA, USA) and examined using a
Zeiss Axioplan phase-epifluorescence microscope (Carl
Zeiss, Oberkochen, Germany).
Induction and detection of apoptosis
Measurement of caspase activity
Apoptosis was induced by treating cells with actinomycin
D (1 ␮g/ml) or staurosporine (1 ␮M) for 24 h or with
genistein (100 ␮M) for 72 h at 37°C. Cells showing mor-
phological features of apoptosis (cell shrinkage and mem-
brane blebbing) were found floating in the culture medium,
while the remaining adherent cells appeared normal. In
some experiments, cells were transferred to serum-free me-
dium and incubated with staurosporine (1 ␮M) for 24 h.
After removal of nonadherent cells, the supernatants were
concentrated 10-fold using Centriprep-30 concentrators
(Amicon, Beverly, MA, USA). In experiments designed to
evaluate the role of caspases, the tripeptide caspase inhibitor
Z-VAD-fmk (100 ␮M) was added to cells 1 h before the
induction of apoptosis.
Cells were lysed in caspase buffer (10 mM HEPES-KOH,
pH 7.4, 10% sucrose, 2 mM EDTA, 0.1%, 3-[(3-
cholamidopropyl)-dimethylammonio]-1-propanesulfonate
(CHAPS) 5 mM DTT, 1 mM PMSF, and 20 ␮g/ml each
pepstatin A, leupeptin, and aprotinin) for 15 minutes on ice.
After centrifugation at 15,000g for 15 minutes at 4°C,
supernatants containing 50 ␮g of protein were incubated
with either 200 ␮M Ac-DEVD-pNA (caspase-3 substrate)
or 200 ␮M Ac-YVAD-pNA (caspase-1 substrate). In some
experiments, the caspase-3 inhibitor Ac-DEVD-CHO was
included at a concentration of 0.1 ␮M. Caspase activity,
measured by the enzymatic release of the chromophore,
pNA from the labeled substrate, was monitored at 405 nm.
Adherent and nonadherent cells were collected sepa-
rately, washed with PBS, and analyzed for DNA fragmen-
tation using the Suicide-Track DNA ladder isolation kit
(Oncogene Research Products, Cambridge, MA, USA) ac-
cording to the manufacturer’s instructions. For the assess-
ment of DNA fragmentation at the single cell level, osteo-
blasts were grown on glass coverslips and treated with
apoptosis-inducing reagents as described previously. Non-
adherent cells were collected by centrifugation and cytocen-
trifuged onto glass coverslips. Apoptosis was detected using
a terminal deoxynucleotidyl transferase (TdT)-mediated de-
oxyuridine triphosphate (dUTP) nick-end labeling (TUNEL)
assay (Promega, Madison, WI, USA) based on the incorpora-
tion of fluorescein-12-dUTP into 3Ј-OH ends of cleaved
DNA. Cells were photographed using a Zeiss Axioplan
phase-epifluorescence microscope.
In vitro caspase cleavage assay
For direct cleavage of adherens junction proteins, ␤- and
␥-catenins were immunoprecipitated from 500 ␮g aliquots
of total TE-85 cell protein, washed with cleavage buffer (10
mM HEPES-KOH, pH 7.4, 10% sucrose, 2 mM EDTA,
0.1% CHAPS, and 5 mM DTT), and resuspended in 25 ␮l
of the same buffer. Recombinant human caspase-3 (20 ng)
was added and samples were incubated at 37°C for 0–60
minutes. The reaction was stopped by the addition of an
equal volume of 2ϫ reducing SDS sample buffer and the
samples were incubated at 100°C for 5 minutes. Cleavage
products were resolved by SDS-PAGE and detected by
immunoblotting as described previously.
Antibody blocking experiments
To prevent cell attachment and spreading, tissue cul-
ture plates were coated with the nonadhesive polymer
polyhydroxylethyl-methacrylate (poly-HEMA) as described
Preparation of cell extracts, immunoprecipitation,
and immunoblotting
Cells were lysed in 10 mM Tris-HCl, pH 7.4, 150 mM previously(5) and washed extensively with PBS before use.
NaCl, 0.5% Triton X-100, 1 mM phenylmethylsulfonyl Primary human osteoblasts (HOBs) were plated on poly-
fluoride (PMSF), 2 mM EDTA, 1 mM Na3VO4, 10 mM HEMA–coated 24-well plates, in the presence of anti-N-
NaF, and 20 ␮g/ml each leupeptin, pepstatin A, and apro- cadherin antibody (A-CAM, clone GC-4) or an isotype-
tinin for 1 h on ice. Samples were centrifuged at 15,000g for matched control antibody at a concentration of 800 ␮g/ml or
15 minutes at 4°C and the supernatants were retained. For with 5 mM EDTA. After overnight incubation, cells were
immunoprecipitation, cell lysates were incubated overnight cytocentrifuged onto glass coverslips and assessed for apo-

CELL ADHESION MOLECULES IN BONE
469
cadherin-4 in these cells. HOBs, SaOS-2, and bone marrow
stromal cells recently have been reported to lack cadherin-
6(34) while cadherin-6 transcripts have been detected in
human stromal cells, the MG-63, and TE-85 osteoblastic
cell lines and in osteoclast-like cells.(39) Using a specific
antibody, we have confirmed the expression of cadherin-6 in
MG-63 cells but not in TE-85 or SaOS-2 cells (Fig. 1). It
has been suggested that cadherin-6 may mediate heterotypic
interactions between osteoblast and osteoclast precursors,
important in the generation of osteoclasts.(39) In the present
study, the finding that cadherin-6 also is expressed in HOBs
(Fig. 1) indicates that it is likely to have specific functions
in differentiated osteoblasts.
Desmogleins, members of the cadherin superfamily nor-
mally present in cell-cell desmosomal junctions,(40,41) also
were expressed in osteoblasts (Fig. 1). RT-PCR analysis
revealed that osteoblasts expressed desmoglein-2, a widely
distributed desmoglein and the only isoform expressed in
nonepithelial tissues.(42) The detection of an immunoreac-
tive band of the correct size (165 kDa) in HOBs, as well as
in osteoblastic cell lines, suggested that desmoglein expres-
sion was not an artifact of cultured osteosarcoma-derived
cell lines. To our knowledge, neither desmosomes nor their
constituent proteins have been detected previously in osteo-
blastic cells and the functional significance of this observa-
tion remains to be elucidated.
FIG. 1. Expression of cadherins and catenins in osteoblastic cells and
HOBs. Extracts of the osteoblastic cell lines MG-63 (lane 1), TE-85
(lane 2), and SaOS-2 (lane 3) and primary human osteoblasts (HOBs;
lane 4) were fractionated by SDS-PAGE and immunoblotted with
antibodies against cadherins and catenins.
ptosis by TUNEL staining or extracted and analyzed by
immunoblotting as described previously. Analysis of cells
fixed before cytocentrifugation confirmed that this proce-
dure did not induce cell damage.
In TE-85 cells, which have an epithelial-like morphol-
ogy, forming closely associated arrays of polygonal cells
at high density, cadherins and catenins were localized to
lateral cell borders at sites of cell-cell contact (Fig. 2). An
identical distribution of cadherins/catenins also was de-
RESULTS
Expression and distribution of cadherins
in osteoblastic cells
We have detected the expression of two type I (N- tected in MG-63 and SaOS-2 cells (data not shown). In
cadherin and cadherin-4) and two type II (cadherin-6 and HOBs, which have a more spread morphology, cadherins
cadherin-11) classical cadherins in human osteoblasts. and catenins were localized similarly to cell-cell con-
Using RT-PCR with specific primers, cadherin-11 and tacts, both at lateral cell borders and where fingerlike
N-cadherin were identified in MG-63, TE-85, and SaOS-2 processes contacted an adjacent cell (Fig. 2), as has
cells, confirming a recent report on the expression of these recently been described.(34)
cadherins in osteoblastic cells.(34) Immunoblotting, using an
antibody that recognizes a highly conserved sequence in the
cytoplasmic domain of all classical cadherins, revealed that
the major reactive species in both HOBs and osteoblastic
Organization of cadherin/catenin complexes
in osteoblastic cells
cell lines had an apparent molecular mass of 135 kDa (Fig.
The interaction of the cytoplasmic domain of the cad-
1, pan-cadherin), identical to that of N-cadherin (Fig. 1) but herins with a group of intercellular proteins, the catenins,
significantly higher than that reported for cadherin-11 (120 and subsequent linkage to the actin cytoskeleton is essential
kDa).(33,34) These results suggest that the major classical for stable cell-cell adhesion.(24–27) Although it has been
cadherin expressed in human osteoblasts is likely to be reported that ␤-catenin interacts with cadherins in bone
N-cadherin and not cadherin-11 (120 kDa) as previously marrow stromal cells(34) and with cadherin-6 in MG-63
reported.(34) This conclusion also is consistent with the cells,(39) neither the presence nor the organization of other
finding that a synthetic peptide containing the His-Ala-Val catenins has been described in osteoblasts.
(HAV) adhesion motif, found in the extracellular domain of
In the present study, we have detected the expression of
type I (e.g., N-cadherin) but not type II (e.g., cadherin-11) ␣-, ␤-, and ␥-catenins in both osteoblastic cells and HOBs,
classical cadherins, was effective in inhibiting cell-cell ad- in which their colocalization with cadherins at sites of
hesion in osteoprogenitor cells by Ͼ70%.(34)
cell-cell contact (Fig. 2) is consistent with the assembly of
Cadherin-4, present in HOBs, was reportedly absent from these proteins into complexes. To define in detail the mo-
osteoblastic cell lines.(34) However, using cadherin-4– lecular organization of cadherins and catenins in osteo-
specific primers, we have obtained PCR products of the blasts, extracts of TE-85 cells were immunoprecipitated
correct size in MG-63, SaOS-2, and TE-85 cells. In addi- with ␤-catenin and the presence of additional proteins was
tion, sequence analysis of cDNA clones from TE-85 cells, determined by immunoblotting. The results clearly show
generated with primers to a highly conserved region in the both the effective precipitation of ␤-catenin (Fig. 3A, ␤-cat)
cytoplasmic domain of cadherins, confirmed the presence of and the coprecipitation of cadherins (Fig. 3A, pan-cad),

470
HUNTER ET AL.
FIG. 3. The composition of cadherin/catenin complexes in osteoblas-
tic cells. (A) Extracts of TE-85 cells were immunoprecipitated (IP) with
antibodies against ␤-catenin, resolved by SDS-PAGE, and immuno-
blotted (IB) with antibodies recognizing pan-cadherin (pan-cad),
N-cadherin (N-cad), ␣-catenin (␣-cat), ␤-catenin (␤-cat), and ␥-catenin
(␥-cat). Cadherin/catenin complexes of identical composition also were
detected in MG-63 and SaOS-2 cells. (B) TE-85 cell extracts, immu-
noprecipitated with antibodies recognizing either pan-cadherin (pan-
cad) or ␣-catenin (␣-cat) were immunoblotted with anti–␥-catenin
(␥-cat) antibodies.
FIG. 2. Immunolocalization of cadherins and catenins in osteoblastic
TE-85 cells and HOBs. Cells were grown to confluence on glass
coverslips, fixed and incubated with antibodies recognizing (A and B)
pan-cadherin, (C and D) N-cadherin, (E and F) ␣-catenin, (G and H)
␤-catenin, and (I and J) ␥-catenin. For each cell type, antibody distri-
bution was assessed in multiple cell monolayers, in at least three
independent experiments. Bar ϭ 25 ␮m.
FIG. 4. SaOS-2 cells undergo spontaneous apoptosis. During routine
culture, a small number of SaOS-2 cells are found as single cells,
floating in the culture medium. (A) DNA isolated from adherent (ad)
and nonadherent (na) SaOS-2 cells was electrophoresed through 2%
agarose gels containing ethidium bromide and DNA laddering visual-
ized with UV light. DNA molecular weight markers, expressed as base
pairs (bp), are indicated to the right. Extensive cleavage of DNA
indicated that nonadherent cells had undergone apoptosis. (B) Adherent
(a) and nonadherent (b) SaOS-2 cells, were fixed and permeabilized
and apoptosis was detected by positive TUNEL staining.
N-cadherin (Fig. 3A, N-cad), and ␣-catenin (Fig. 3A,
␣-cat). In contrast, ␥-catenin was not coprecipitated with
␤-catenin from these cells (Fig. 3A, ␥-cat) but was present
in both pan-cadherin and ␣-catenin immunoprecipitates
(Fig. 3B). These results show that at least two distinct
complexes are present in osteoblasts, one containing
cadherin/␣-catenin/␤-catenin and one containing cadherin/
␣-catenin/␥-catenin. Although ␥-catenin is the only catenin
reported to interact with both classical and desmosomal
cadherins,(22,43) we have been unable to coprecipitate
␥-catenin and desmogleins. Because the desmoglein and
␣-catenin binding sites on ␥-catenin overlap,(44) it is possi-
ble that the predominant interaction of ␥-catenin in these
cells is with ␣-catenin.
Fate of cadherins/catenins during osteoblast apoptosis
During routine culture, we noticed that a small number of
SaOS-2 cells are found as single cells in the culture me-
dium. The detached cells were shrunken in appearance, with
membrane blebbing and nuclear condensation, morpholog-
ical features suggestive of apoptotic cell death. Agarose gel
electrophoresis revealed a ladder of 180–200-bp multiple
DNA fragments in nonadherent (Fig. 4A, na) but not in
adherent (Fig. 4A, ad) cells. Laddering reflects internucleo-

CELL ADHESION MOLECULES IN BONE
471
FIG. 5. Expression of cadherins and catenins in osteoblasts under-
going spontaneous and induced apoptosis. (A) Extracts of adherent (ad)
and nonadherent (na) SaOS-2 cells were resolved by SDS-PAGE and
immunoblotted with antibodies against cadherins and catenins. (B)
FIG. 6. Caspase-3 is activated in apoptotic osteoblasts. (A) Direct
measurement of caspase-3 activity in extracts of adherent (ad) and
Apoptosis was induced by treating MG-63 cells for 72 h with 100 ␮M nonadherent (na) SaOS-2 cells. Hydrolysis of the caspase-3 substrate
genistein, TE-85 cells with 1 ␮M staurosporine for 24 h, and HOBs Ac-DEVD-pNA was measured at 405 nm over a period of 8 h, in the
with 1 ␮g/ml actinomycin D for 24 h. Adherent (ad) and nonadherent
(na) cell extracts were immunoblotted with antibodies raised against
N-cadherin, ␤-catenin, and ␥-catenin.
absence and presence of the caspase-3 inhibitor Ac-DEVD-CHO at a
concentration of 0.1 ␮M (ϫ, ad; f, ad ϩ inhib; Œ, na; ࡗ, na ϩ inhib).
(B) Osteoblast apoptosis was induced as described in .Fig. 5B. Extracts
of adherent (ad) and nonadherent (na) cells were immunoblotted with
an antibody recognizing the inactive proenzyme form of caspase-3.
Activation of caspase-3 is measured by a decrease in the expression of
the intact 32-kDa proenzyme.
somal DNA cleavage and is considered a hallmark of apo-
ptosis.(2) Positive TUNEL staining of nonadherent (Fig. 4B,
b) but not adherent (Fig. 4B, a) SaOS-2 cells confirmed that
the detached cells had undergone apoptosis.
In apoptotic SaOS-2 cells, immunoblotting revealed that TE-85 cells and in HOBs resulted in the loss of N-cadherin
cellular cadherins, including N-cadherin, were significantly and the cleavage of ␤-catenin into 70- to 75-kDa peptides
decreased in amount and that ␤-catenin was cleaved into (Fig. 5B), identical to those detected in apoptotic SaOS-2
three major polypeptides of 70- to 75-kDa and minor 85- to cells. In TE-85 cells, which express high levels of ␥-catenin
90-kDa fragments (Fig. 5A). An antibody recognizing (Fig. 1), major fragments of 75, 72, and 65 kDa and a minor
amino acids 768–781 of ␤-catenin failed to detect any of the fragment of about 55 kDa were detected in nonadherent but
cleavage products in apoptotic cells (data not shown), indi- not adherent cells (Fig. 5B).
cating that these fragments lack the COOH-terminal do-
main. Cleavage of ␥-catenin also was apparent but was
more clearly seen in TE-85 cells (Fig. 5B), which express
higher levels of this protein. By contrast, there was no
Caspase-dependent cleavage of cadherins/catenins
in apoptotic osteoblasts
significant difference in the level of ␣-catenin in adherent
Because activation of the caspase family of cysteine
and nonadherent cells (Fig. 5A). We also detected a dra- proteases and their cleavage of key cellular substrates is a
matic reduction in the expression of desmogleins in nonad- crucial step in the apoptotic process,(48) the presence of
herent SaOS-2 cells (Fig. 5A). To determine if these effects caspase-1 and caspase-3 activity in extracts of SaOS-2 cells
were a common feature of osteoblast apoptosis, the expres- was investigated by measuring the cleavage of their respec-
sion of cadherins/catenins examined in osteoblasts was tive substrates Ac-YVAD-pNA and Ac-DEVD-pNA.
treated with genistein, staurosporine, or actinomycin D, all Caspase-3 activity, barely detectable in adherent cells, was
known inducers of apoptosis.(45–47) Consistent with the strongly activated in nonadherent, apoptotic cells (Fig. 6A),
induction of apoptosis, osteoblasts detached from the cul- reaching a maximal activity 11-fold higher than basal levels.
ture dish by treatment with these reagents showed positive In the presence of the specific peptide inhibitor Ac-DEVD-
TUNEL staining (data not shown). Apoptosis in MG-63 and CHO caspase-3 activity was reduced to basal levels (Fig.

472
HUNTER ET AL.
FIG. 7. N-cadherin and ␤- and ␥-catenins but not ␣-catenin are cleaved by caspases. (A) TE-85 cells were grown without additions (Ϫ) or with
(ϩ) 1 ␮M staurosporine for 24 h to induce apoptosis, in the absence (Ϫ) or presence (ϩ) of the caspase inhibitor Z-VAD-fmk at a concentration
of 100 ␮M. Cell extracts from control and treated cells were resolved by SDS-PAGE and immunoblotted with antibodies against N-cadherin,
␤-catenin, and ␥-catenin. (B) Extracts of untreated TE-85 cells were immunoprecipitated with an antibody to ␤-catenin and incubated with
recombinant human caspase-3 at 37°C for 0–60 minutes. Cell extracts were fractionated by SDS-PAGE before immunoblotting with antibodies
against ␤-catenin, ␣-catenin, and either the cytoplasmic (cyto) or the extracellular (EC) domains of N-cadherin. (C) Conditioned medium from
TE-85 cells grown in the absence (Ϫ) or presence (ϩ) of 1 ␮M staurosporine for 24 h was concentrated 10-fold and analyzed by immunoblotting
with an antibody raised to the extracellular domain of N-cadherin.
6A). Caspase-1 was not activated in apoptotic osteoblasts ognizes the cytoplasmic domain of N-cadherin, caspase-3
(data not shown). Activation of caspase-3 in apoptotic cells incubation caused a dramatic decrease in the amount of
was further confirmed by the disappearance of the inactive, intact N-cadherin (Fig. 7B), consistent with the pattern seen
32-kDa proenzyme in nonadherent SaOS-2 cells undergoing in apoptotic cells (Fig. 5). Immunoblotting with an antibody
spontaneous apoptosis (Fig. 6B) and in nonadherent MG-63 against the extracellular domain indicated that, in the pres-
and TE-85 cells and in HOBs (Fig. 6B), induced to undergo ence of caspase-3, N-cadherin was reduced in apparent
apoptosis with staurosporine. The ability of the tripeptide molecular mass from 135 to 120 kDa (Fig. 7B). A fragment
caspase inhibitor Z-VAD-fmk to inhibit cleavage of both ␤- of this size has been detected in some apoptotic cells, but its
and ␥-catenins and to abrogate the decrease in N-cadherin absence from others suggests that additional proteolysis
expression in staurosporine-treated TE-85 cells (Fig. 7A) may occur (data not shown). The recent demonstration that
strongly implicated activated caspases in the cleavage of N-cadherin can be cleaved extracellularly during ovarian
these proteins.
granuloma cell apoptosis(16) led us to examine whether a
To assess if cadherins/catenins were direct substrates for similar cleavage might occur during osteoblast apoptosis.
caspases, in the absence of endogenous cellular proteases, Immunoblotting with an antibody against the extracellular
␤-catenin was immunoprecipitated from untreated TE-85 domain of N-cadherin did indeed reveal the presence of a
cell extracts before incubation with recombinant human 90-kDa fragment in conditioned medium from staurosporine-
caspase-3. Initially, ␤-catenin was cleaved into several treated but not control TE-85 cells, consistent with extracellu-
polypeptides ranging in size from 70 to 90 kDa (Fig. 7B). lar domain cleavage in apoptotic cells (Fig. 7C). Caspase-3
The higher molecular-mass species gradually disappeared also was able to cleave immunoprecipitated ␥-catenin, pro-
with time and after 1 h, the predominant ␤-catenin frag- ducing the 75-kDa and 65-kDa fragments, but the additional
ments had a molecular mass of 70–75 kDa (Fig. 7B), 72-kDa and 55-kDa fragments seen in apoptotic osteoblasts
identical to those found in apoptotic cells (Fig. 4). We have were not detected (data not shown).
taken advantage of the fact that both N-cadherin and
␣-catenin coprecipitate with ␤-catenin (Fig. 3) to analyze
the effect of caspase-3 on these proteins present in ␤-catenin
immunoprecipitates. In agreement with observations in
Effect of caspase cleavage on cadherin/catenin
interactions
apoptotic cells (Fig. 4), ␣-catenin was not a substrate for
In osteoblasts, we have shown the presence of two dis-
caspase-3 in vitro (Fig. 7B), despite the presence of three tinct complexes containing cadherin and ␣-catenin, together
potential caspase-3 cleavage sites. Using an antibody that rec- with either ␤- or ␥-catenin (Fig. 3). Because the adhesive

CELL ADHESION MOLECULES IN BONE
473
FIG. 9. Disruption of N-cadherin–mediated adhesion induces apo-
ptosis. (A) HOBs were cultured on poly-HEMA–coated dishes in the
presence of control Ig (a), a function-blocking antibody recognizing the
extracellular domain of N-cadherin (b), or 5 mM EDTA (c) for 24 h.
Cells were cytocentrifuged onto glass coverslips and subjected to
TUNEL staining. (B) HOBs treated with control (lane 1), anti–N-
cadherin (lane 2) or EDTA (lane 3) were extracted, resolved by
SDS-PAGE, and immunoblotted with an antibody recognizing
␤-catenin.
FIG. 8. Cadherin/catenin complex assembly in apoptotic osteoblasts.
Extracts of apoptotic SaOS-2 cells were left untreated (Ϫ) or immu-
noprecipitated (IP) with antibodies recognizing pan-cadherin (pan-cad)
or ␣-catenin (␣-cat). After fractionation by SDS-PAGE, the composi-
tion of cadherin/catenin complexes was assessed by immunoblotting
(IB) with antibodies against (A) ␤-catenin or (B) ␥-catenin.
function of cadherins is dependent on interactions with the
intracellular catenins,(24–27,49) we have examined the effect
of caspase cleavage on complex assembly. All three major
␤-catenin cleavage products present in apoptotic SaOS-2
cells retained the ability to interact with cadherins, whereas
only the two higher-mass species were able to interact with
␣-catenin (Fig. 8A). Both cadherins and ␣-catenin inter-
acted with the major cleavage products of ␥-catenin, al-
though they interacted poorly with the 55-kDa fragment
(Fig. 8B). An increase in the relative amount of intact
␥-catenin coprecipitated with ␣-catenin (Fig. 8B) may be
indicative of a stronger interaction between the uncleaved
forms of these proteins. Taken together, caspase-dependent
cleavage of ␤- and ␥-catenins appears to have little effect on
their ability to form adhesion complexes, although the in-
ability of the smallest ␤-catenin fragment to associate with
␣-catenin may lead to a decrease in adhesion.
associated and contained many apoptotic, TUNEL-positive
cells (Fig. 9A, b). The inability of anti–N-cadherin to dis-
rupt completely aggregate formation, strongly indicates that
cadherins-4, -6, and -11, present in HOBs, also contribute to
cell-cell adhesion. Cadherin-mediated cell adhesion is a
Ca2ϩ-dependent process,(19) and in the presence of the cal-
cium chelator, EDTA cells failed to form cell-cell contacts
and showed extensive TUNEL-positive staining (Fig. 9A, c).
As a consequence of disrupting N-cadherin–mediated
cell-cell adhesion, ␤-catenin was cleaved (Fig. 9B), giving
rise to a number of lower molecular mass species, including
the 70- to 75-kDa polypeptides characteristic of apoptotic
osteoblasts (Fig. 5) and generated by caspase-3 activity in
vitro (Fig. 7). In cells treated with EDTA, in which cell-cell
adhesion was almost completely disrupted, ␤-catenin was
cleaved more extensively to the 70- to 75-kDa fragments
(Fig. 9B). Taken together, these results strongly suggest that
disruption of N-cadherin–mediated cell-cell adhesion in os-
teoblasts leads to the activation of caspase-3, cleavage of
␤-catenin, and induction of apoptosis in human osteoblasts.
Disruption of N-cadherin–mediated adhesion induces
osteoblast apoptosis
During osteoblast apoptosis, the detachment of cells not
only from each other but from the culture dish indicates that
both cell-cell and cell-matrix interactions are disrupted.
Because loss of cell-matrix adhesion has been shown to lead
to apoptosis,(4) we have attempted to distinguish between
DISCUSSION
Although recent data(1) indicate that the majority of os-
these adhesive processes by culturing HOBs on dishes teoblasts are destined to die by apoptosis, the molecular
coated with poly-HEMA, a nonionic polymer, which pre- events underlying and regulating osteoblast-programmed
vents matrix deposition, cell attachment, and spreading.(4,50) cell death remain to be elucidated. In the present study, we
HOBs, grown on poly-HEMA–coated dishes in the pres- have identified proteins associated with the intercellular
ence of control antibody formed large, compact multicellu- adherens junctions as molecular targets during osteoblast
lar aggregates, with no detectable apoptosis as judged by the apoptosis and provide evidence that cadherin-mediated cell-
absence of TUNEL staining (Fig. 9A, a). In contrast, when cell adhesion acts to inhibit the apoptotic process.
cultured in the presence of a function-blocking N-cadherin
While studying the expression and organization of adhe-
antibody, HOBs formed aggregates that were only loosely rens junction proteins in osteoblasts, we detected a signifi-

474
HUNTER ET AL.
cant reduction in the expression of classical cadherins, in- protein-protein interactions required for adhesion complex
cluding N-cadherin, and the cleavage of ␤- and ␥- but not assembly remain essentially intact in apoptotic osteoblasts.
␣-catenin in MG-63, TE-85, and SaOS-2 osteoblastic cells
In apoptotic osteoblasts, we have detected a decrease in
as well as in primary osteoblasts (HOBs) undergoing apo- the expression of cadherins, which we now report appears,
ptosis. The ability of the tripeptide caspase inhibitor at least in part, to be the result of caspase-dependent cleav-
Z-VAD-fmk to suppress cleavage of ␤- and ␥-catenins and age. N-cadherin was cleaved by caspase-3 in vitro, to give
to abrogate the decrease in N-cadherin expression, sug- a polypeptide of 120 kDa, recognized by an antibody
gested that the caspase family of cysteine proteases, acti- against the extracellular domain but not by an antibody
vated during apoptosis,(48) may play an important role in the against the cytoplasmic domain of the protein. Of the three
cleavage of these proteins. In support of this notion, caspase-3 consensus cleavage sites identified within the
caspase-3, a major effector caspase responsible for the cytoplasmic domain of N-cadherin,(52) cleavage at the most
cleavage of a number of key cellular proteins during apo- COOH-terminal site (830Asp-Ile-Gly-Asp833) is inconsistent
ptosis,(48) was activated in both apoptotic osteoblastic cell with the size of the cleaved fragment. However, cleavage at
lines and HOBs. Cleavage of N-cadherin and ␤- and either of the two immediately adjacent, membrane proximal
␥-catenins, but not ␣-catenin, by caspase-3 in vitro to gen- sites (760Asp-Pro-Glu-Asp763 and 764Asp-Val-Arg-Asp767
)
erate fragments identical to those seen in apoptotic cells would remove 138–142 amino acids and cause a reduction
further confirmed the central role of caspase-3 in the cleav- in the apparent molecular mass of N-cadherin (135 kDa) of
age of adherens junction proteins during osteoblast apopto- about 15 kDa, consistent with the 120-kDa fragment de-
sis.
tected after caspase-3 treatment. A fragment of this size has
Cleavage of ␤-catenin has been reported in apoptotic been detected in apoptotic osteoblasts, although its absence
endothelial cells(10) and could be reproduced by the addition from some extracts may indicate subsequent cleavage by
of caspase-3 to endothelial cell extracts but not to immuno- caspases or other proteases. Interestingly, by comparing the
precipitated ␤-catenin. The authors’ conclusion that com- sequences of many different classical cadherins, we have
plete cleavage of ␤-catenin is dependent on caspase-3– found that one of the two membrane proximal sites (764Asp-
dependent activation of additional cellular proteases(10) Val-Arg-Asp767) is conserved in type I but not type II
contrasts sharply with our results, which clearly show that proteins, suggesting that it may define a function that differs
caspase-3 activity alone is sufficient to generate the between the two types of cadherins.
␤-catenin cleavage pattern found in apoptotic cells. In sup-
port of this conclusion, in vitro translated ␤-catenin has caspases in the cleavage of cadherins. In that study,(53)
been shown to be cleaved efficiently by caspase-3 to give a role for caspases was implied from the ability of caspase
pattern resembling that in apoptotic cells.(51)
inhibitors to prevent cleavage of E- and P-cadherins in
We are aware of only one previous report implicating
a
Caspase-3–dependent cleavage of immunoprecipitated apoptotic cells, but direct cleavage of cadherins was not
␥-catenin generated the major 75-kDa and 65-kDa polypep- shown. The major cleavage product of P-cadherin in apo-
tides seen in apoptotic cells but failed to generate the 72-kDa ptotic cells had an apparent molecular mass of 104 kDa,
and 55-kDa fragments. Caspase-3–dependent cleavage of consistent with cleavage at the site identified by us to be
either immunoprecipitated(10) or in vitro translated(52) ␥-catenin conserved in type I cadherins. In apoptotic cells, E-cadherin
also has failed to generate the pattern of fragments produced was cleaved to a 48-kDa product, which was suggested to be
during apoptosis. Thus, although ␥-catenin can clearly act the result of caspase cleavage at amino acids 476–479.(53)
as caspase-3 substrate, it remains to be shown whether However, this sequence, which forms part of a putative
additional proteases are involved in its cleavage in vivo.
Ca2ϩ-binding site,(54) is located within the extracellular
In epithelial cells, both ␤- and ␥-catenins interact directly domain of E-cadherin, while the caspases are intracellular
with the cytoplasmic domain of the cadherins and with enzymes.
␣-catenin to form a supramolecular complex that is re-
In addition to cleavage of the intracellular domain, the
garded as the functional unit of cell adhesion.(24–27,49) In the presence of a 90-kDa N-cadherin fragment in conditioned
present study, the identification in osteoblasts of two dis- medium from apoptotic but not control cells indicated that
tinct complexes containing cadherin and ␣-catenin with N-cadherin also could be cleaved extracellularly. Although
either ␤- or ␥-catenin allowed us to assess the impact of the nature of the proteases responsible for extracellular
apoptosis on complex assembly. Cleavage of ␤- and domain cleavage remain to be identified, surface-associated
␥-catenins during apoptosis had little effect on their ability metalloproteinases have been implicated in the cleavage of
to interact with the cadherins. The major cleavage products N-cadherin during retinal development(55) and ovarian gran-
of ␥-catenin retained the ability to interact with ␣-catenin as uloma cell apoptosis(16) and in the release of a surface-
did the two higher molecular mass cleavage products of associated fragment of VE-cadherin during endothelial
␤-catenin. The 55-kDa fragment of ␥-catenin showed re- apoptosis.(10)
duced binding to ␣-catenin, and there was no detectable
Recent studies have provided evidence that cell-cell in-
interaction between the 70-kDa fragment of ␤-catenin and teractions may be one mechanism by which cells can reg-
␣-catenin. Because the association of the cadherins with ulate apoptosis.(6,7) In the present study, the induction of
␣-catenin and the actin cytoskeleton is essential for their apoptosis in HOBs by a function-blocking N-cadherin an-
adhesive function, a reduction in ␣-catenin binding by tibody, but not by an isotype-matched control, indicated that
cleaved ␤- and ␥-catenins may result in reduced cell-cell N-cadherin–mediated cell-cell adhesion supports osteoblast
adhesion. However, taken together, our results indicate that survival. N-cadherin also has been linked recently to the

CELL ADHESION MOLECULES IN BONE
475
survival of ovarian granulosa cells,(15,16) and an association formation is further suggested by the observation that para-
between E-cadherin–mediated aggregation and survival of thyroid hormone (PTH), which increases bone formation by
squamous epithelial cells has been reported.(17)
preventing osteoblast apoptosis,(62) also increases the ex-
The mechanisms whereby cadherin-mediated cell-cell ad- pression of cadherins in osteoblasts and osteoprogenitor
hesion generates cell survival signals remain poorly under- cells.(63,64)
stood. N-cadherin–mediated adhesion of ovarian granulosa
In summary, we have identified N-cadherin–mediated
cells has been linked to tyrosine phosphorylation of the cell-cell adhesion as a molecular mechanism for suppress-
fibroblast growth factor (FGF) receptor,(56) while E-cadherin– ing apoptosis in human osteoblasts, which we suggest is
dependent aggregation of epithelial cells results in activa- mediated by a ␤-catenin signaling pathway.
tion of the retinoblastoma protein and G1 cell cycle arrest.(57)
In the present study, apoptosis induced by the disruption of
N-cadherin–mediated intercellular adhesion resulted in the
cleavage of ␤-catenin to give a pattern indicative of caspase-3
activation. Because caspase-dependent cleavage of ␤-catenin
has a limited effect on its ability to interact with its cellular
binding partners, we think it unlikely that cleavage of this
molecule plays a major role in the direct dismantling of
cell-cell contacts during osteoblast apoptosis. However, in
addition to a structural role in adherens junctions, ␤-catenin
is an important transducer of cellular signals and can translo-
cate to the nucleus, where it interacts with the LEF/TCF family
of transcription factors to regulate gene expression.(28–32) We
suggest, therefore, that ␤-catenin is involved in antiapoptotic
signaling and that this is disrupted by caspase-dependent
cleavage during osteoblast apoptosis. The ability of the
cleavage products of ␤-catenin to interact with cadherins
indicates that the LEF/TCF binding site is also intact, be-
cause these transcription factors share a large, overlap-
ping binding domain with the cadherins in the center of
␤-catenin.(58) Within the LEF/TCF–␤-catenin complex, the
LEF/TCF transcription factors provide a DNA binding do-
main, while the transactivation function resides within the
COOH-terminal domain of ␤-catenin.(32) In apoptotic os-
teoblasts, cleavage of ␤-catenin resulted in the loss of the
COOH-terminal domain, as judged by the inability of an
antibody to amino acids 768–781 of ␤-catenin to recognize
the cleaved protein. Taken together, our results indicate that
although cleaved ␤-catenin would retain the ability to in-
teract with LEF/TCF, transcriptional activation and thus
antiapoptotic signaling would be compromised. Although
additional studies will be required to validate this mecha-
nism, the recent demonstration that deletion constructs of
␤-catenin, resembling the proteolytic fragments seen in
apoptotic cells, have strongly reduced transcriptional activ-
ity(59) is consistent with our hypothesis.
The factors responsible for disruption of N-cadherin–
mediated adhesion during bone remodeling remain to be
elucidated. However, the cytokines tumor necrosis factor ␣
(TNF-␣) and interleukin-1 (IL-1), present in the bone mi-
croenvironment and recognized as powerful stimulants of
bone resorption and inhibitors of bone formation,(60) re-
cently have been shown to down-regulate the expression of
N-cadherin in the mouse osteoblastic cell line MC3T3-
E1.(61) Thus, in conditions where the levels of these cyto-
kines are increased, such as during postmenopausal osteo-
porosis,(60) disruption of N-cadherin–mediated cell-cell
adhesion and induction of osteoblast apoptosis may be one
mechanism whereby access of osteoclasts to the bone sur-
face and bone resorption is facilitated. That cadherin-
mediated adhesion may play a role in promoting bone
ACKNOWLEDGMENTS
We are indebted to the Ministry of Agriculture, Fisheries
and Food and to the Scottish Executive Rural Affairs De-
partment for support.
REFERENCES
1. 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.
2. Wyllie AH, Kerr JF, Currie AR 1980 Cell death: The signif-
icance of apoptosis. Int Rev Cytol 68:251–306.
3. Collins MK, Lopez Rivas A 1993 The control of apoptosis in
mammalian cells. Trends Biochem Sci 18:307–309.
4. Frisch SM, Francis H 1994 Disruption of epithelial cell-matrix
interactions induces apoptosis. J Cell Biol 124:619–626.
5. Re F, Zanetti A, Sironi M, Polentarutti N, Lanfrancone L,
Dejana E, Colotta F 1994 Inhibition of anchorage-dependent
cell spreading triggers apoptosis in cultured human endothelial
cells. J Cell Biol 127:537–546.
6. Bates RC, Buret A, van Helden DF, Horton MA, Burns GF
1994 Apoptosis induced by inhibition of intercellular contact.
J Cell Biol 125:403–415.
7. Hermiston ML, Gordon JI 1995 In vivo analysis of cadherin
function in the mouse intestinal epithelium: Essential roles in
adhesion, maintenance of differentiation, and regulation of
programmed cell death. J Cell Biol 129:489–506.
8. Potter SW, Gaza G, Morris JE 1996 Estradiol induces
E-cadherin degradation in mouse uterine epithelium during the
estrous cycle and early pregnancy. J Cell Physiol 169:1–14.
9. Brancolini C, Lazarevic D, Rodriguez J, Schneider C 1997
Dismantling cell-cell contacts during apoptosis is coupled to a
caspase-dependent proteolytic cleavage of beta-catenin. J Cell
Biol 139:759–771.
10. Herren B, Levkau B, Raines EW, Ross R 1998 Cleavage of
beta-catenin and plakoglobin and shedding of VE-cadherin
during endothelial apoptosis: Evidence for a role for caspases
and metalloproteinases. Mol Biol Cell 9:1589–1601.
11. Morimoto Y, Ohba T, Kobayashi S, Haneji T 1997 The protein
phosphatase inhibitors okadaic acid and calyculin A induce
apoptosis in human osteoblastic cells. Exp Cell Res 230:181–
186.
12. Kawakami A, Eguchi K, Matsuoka N, Tsuboi M, Koji T,
Urayama S, Fujiyama K, Kiriyama T, Nakashima T, Nakane
PK, Nagataki S 1997 Fas and Fas ligand interaction is neces-
sary for human osteoblast apoptosis. J Bone Miner Res 12:
1637–1646.
13. Frisch SM, Vuori K, Ruoslahti E, Chan-Hui PY 1996 Control
of adhesion-dependent cell survival by focal adhesion kinase.
J Cell Biol 134:793–799.
14. Hungerford JE, Compton MT, Matter ML, Hoffstrom BG,
Otey CA 1996 Inhibition of pp125FAK in cultured fibroblasts
results in apoptosis. J Cell Biol 135:1383–1390.

476
HUNTER ET AL.
15. Peluso JJ, Pappalardo A, Trolice MP 1996 N-cadherin– 34. Cheng SL, Lecanda F, Davidson MK, Warlow PM, Zhang SF,
mediated cell contact inhibits granulosa cell apoptosis in a
progesterone-independent manner. Endocrinology 137:1196–
1203.
Zhang L, Suzuki S, St John T, Civitelli R 1998 Human
osteoblasts express a repertoire of cadherins, which are critical
for BMP-2-induced osteogenic differentiation. J Bone Miner
Res 13:633–644.
16. Makrigiannakis A, Coukos G, Christofidou-Solomidou M,
35. Ferrari SL, Traianedes K, Thorne M, Lafage-Proust MH, Ge-
never P, Cecchini MG, Behar V, Bisello A, Chorev M, Rosen-
blatt M, Suva LJ 2000 A role for N-cadherin in the develop-
ment of the differentiated osteoblastic phenotype. J Bone
Miner Res 15:198–208.
36. Scheven BA, van der Veen MJ, Damen CA, Lafeber FP, Van
Rijn HJ, Bijlsma JW, Duursma SA 1995 Effects of methotrex-
ate on human osteoblasts in vitro: Modulation by 1,25-
dihydroxyvitamin D3. J Bone Miner Res 10:874–80.
37. Pytela R, Suzuki S, Breuss J, Erle DJ, Sheppard D 1994
Polymerase chain reaction cloning with degenerate primers:
Homology-based identification of adhesion molecules. Meth-
ods Enzymol 245:420–451.
38. Laemmli UK 1970 Cleavage of structural proteins during the
assembly of the head of bacteriophage T4. Nature 227:680–
685.
39. Mbalaviele G, Nishimura R, Myoi A, Niewolna M, Reddy SV,
Chen D, Feng J, Roodman D, Mundy GR, Yoneda T 1998
Cadherin-6 mediates the heterotypic interactions between the
hemopoietic osteoclast cell lineage and stromal cells in a
murine model of osteoclast differentiation. J Cell Biol 141:
1467–1476.
40. Koch PJ, Franke WW 1994 Desmosomal cadherins: Another
growing multigene family of adhesion molecules. Curr Opin
Cell Biol 6:682–687.
41. Garrod D, Chidgey M, North A 1996 Desmosomes: Differen-
tiation, development, dynamics and disease. Curr Opin Cell
Biol 8:670–678.
42. Schafer S, Koch PJ, Franke WW 1994 Identification of the
ubiquitous human desmoglein, Dsg2, and the expression cat-
alogue of the desmoglein subfamily of desmosomal cadherins.
Exp Cell Res 211:391–399.
43. Mathur M, Goodwin L, Cowin P 1994 Interactions of the
cytoplasmic domain of the desmosomal cadherin Dsg1 with
plakoglobin. J Biol Chem 269:14075–14080.
44. Witcher LL, Collins R, Puttagunta S, Mechanic SE, Munson
M, Gumbiner B, Cowin P 1996 Desmosomal cadherin binding
domains of plakoglobin. J Biol Chem 271:10904–10909.
45. Naora H, Nishida T, Shindo Y, Adachi M, Naora H 1996
Constitutively enhanced nbl expression is associated with the
induction of internucleosomal DNA cleavage by actinomycin
D. Biochem Biophys Res Commun 224:258–264.
46. Janicke RU, Sprengart ML, Wati MR, Porter AG 1998
Caspase-3 is required for DNA fragmentation and morpholog-
ical changes associated with apoptosis. J Biol Chem 273:
9357–9360.
47. Li Y, Upadhyay S, Bhuiyan M, Sarkar FH 1999 Induction of
apoptosis in breast cancer cells MDA-MB-231 by genistein.
Oncogene 18:3166–3172.
48. Thornberry NA, Lazebnik Y 1998 Caspases: Enemies within.
Science 281:1312–1316.
49. Hinck L, Nathke IS, Papkoff J, Nelson WJ 1994 Beta-catenin:
A common target for the regulation of cell adhesion by Wnt-1
and Src signaling pathways. Trends Biochem Sci 19:538–542.
50. Folkman J, Moscona A 1978 Role of cell shape in growth
control. Nature 273:345–349.
51. Brancolini C, Sgorbissa A, Schneider C 1998 Proteolytic pro-
cessing of the adherens junction components beta-catenin and
gamma-catenin/plakoglobin during apoptosis. Cell Death Dif-
fer 5:1042–1050.
Gour BJ, Radice GL, Blaschuk O, Coutifaris
C 1999
N-cadherin–mediated human granulosa cell adhesion prevents
apoptosis: A role in follicular atresia and luteolysis? Am J
Pathol 154:1391–1406.
17. Kantak SS, Kramer RH 1998 E-cadherin regulates anchorage-
independent growth and survival in oral squamous cell carci-
noma cells. J Biol Chem 273:16953–16961.
18. Globus RK, Doty SB, Lull JC, Holmuhamedov E, Humphries
MJ, Damsky CH 1998 Fibronectin is a survival factor for
differentiated osteoblasts. J Cell Sci 111:1385–1393.
19. Takeichi M 1991 Cadherin cell adhesion receptors as a mor-
phogenetic regulator. Science 251:1451–1455.
20. Kemler R 1993 From cadherins to catenins: Cytoplasmic pro-
tein interactions and regulation of cell adhesion. Trends Genet
9:317–321.
21. Nose A, Tsuji K, Takeichi M 1990 Localization of specificity
determining sites in cadherin cell adhesion molecules. Cell
61:147–155.
22. Ozawa M, Baribault H, Kemler R 1989 The cytoplasmic
domain of the cell adhesion molecule uvomorulin associates
with three independent proteins structurally related in different
species. EMBO J 8:1711–1717.
23. Ozawa M, Kemler R 1992 Molecular organization of the
uvomorulin-catenin complex. J Cell Biol 116:989–996.
24. Aberle H, Butz S, Stappert J, Weissig H, Kemler R, Hos-
chuetzky H 1994 Assembly of the cadherin-catenin complex in
vitro with recombinant proteins. J Cell Sci 107:3655–3663.
25. Hinck L, Nathke IS, Papkoff J, Nelson WJ 1994 Dynamics of
cadherin/catenin complex formation: Novel protein interac-
tions and pathways of complex assembly. J Cell Biol 125:
1327–40.
26. Jou TS, Stewart DB, Stappert J, Nelson WJ, Marrs JA 1995
Genetic and biochemical dissection of protein linkages in the
cadherin-catenin complex. Proc Natl Acad Sci USA 92:5067–
5071.
27. Rimm DL, Koslov ER, Kebriaei P, Cianci CD, Morrow JS
1995 Alpha 1(E)-catenin is an actin-binding and -bundling
protein mediating the attachment of F-actin to the membrane
adhesion complex. Proc Natl Acad Sci USA 92:8813–8817.
28. Molenaar M, van de Wetering M, Oosterwegel M, Peterson-
Maduro J, Godsave S, Korinek V, Roose J, Destree O, Clevers
H 1996 XTcf-3 transcription factor mediates beta-catenin–
induced axis formation in Xenopus embryos. Cell 86:391–399.
29. Behrens J, von Kries JP, Kuhl M, Bruhn L, Wedlich D,
Grosschedl R, Birchmeier W 1996 Functional interaction of
beta-catenin with the transcription factor LEF-1. Nature 382:
638–42.
30. Berens ME, Rief MD, Shapiro JR, Haskett D, Giese A, Joy A,
Coons SW 1996 Proliferation and motility responses of pri-
mary and recurrent gliomas related to changes in epidermal
growth factor receptor expression. J Neurooncol 27:11–22.
31. Huber O, Korn R, McLaughlin J, Ohsugi M, Herrmann BG,
Kemler R 1996 Nuclear localization of beta-catenin by inter-
action with transcription factor LEF-1. Mech Dev 59:3–10.
32. van de Wetering M, Cavallo R, Dooijes D, van Beest M, van
Es J, Loureiro J, Ypma A, Hursh D, Jones T, Bejsovec A,
Peifer M, Mortin M, Clevers H 1997 Armadillo coactivates
transcription driven by the product of the Drosophila segment
polarity gene dTCF. Cell 88:789–799.
33. Okazaki M, Takeshita S, Kawai S, Kikuno R, Tsujimura A,
Kudo A, Amann E 2098 1994 Molecular cloning and charac-
terization of OB-cadherin, a new member of cadherin family
expressed in osteoblasts. J Biol Chem 269:12092–12098.
52. Thornberry NA, Rano TA, Peterson EP, Rasper DM, Timkey
T, Garcia-Calvo M, Houtzager VM, Nordstrom PA, Roy S,
Vaillancourt JP, Chapman KT, Nicholson DW 1997 A com-
binatorial approach defines specificities of members of the

CELL ADHESION MOLECULES IN BONE
477
caspase family and granzyme B. Functional relationships es-
tablished for key mediators of apoptosis. J Biol Chem 272:
17907–17911.
ments with reduced transactivation potential. J Biol Chem
275:16345–16353.
60. Pacifici R 1996 Estrogen, cytokines, and pathogenesis of post-
menopausal osteoporosis. J Bone Miner Res 11:1043–1051.
53. Schmeiser K, Grand RJ 1999 The fate of E- and P-cadherin
during the early stages of apoptosis. Cell Death Differ 6:377– 61. Tsutsumimoto T, Kawasaki S, Ebara S, Takaoka K 1999
386.
TNF-(alpha) and IL-1(beta) suppress N-cadherin expression in
54. Berndorff D, Gessner R, Kreft B, Schnoy N, Lajous-Petter
MC3T3–E1 cells. J Bone Miner Res 14:1751–1760.
AM, Loch N, Reutter W, Hortsch M, Tauber R 1994 Liver- 62. Jilka RL, Weinstein RS, Bellido T, Robertson P, Parfitt AM,
intestine cadherin: Molecular cloning and characterization of a
novel Ca(2ϩ)-dependent cell adhesion molecule expressed in
liver and intestine. J Cell Biol 125:1353–1369.
Manolagas SC 1999 Increased bone formation by prevention
of osteoblast apoptosis with parathyroid hormone. J Clin In-
vest 104:439–446.
55. Paradies NE, Grunwald GB 1993 Purification and character- 63. Babich M, Foti LR 1994 E-cadherins identified in osteoblastic
ization of NCAD90, a soluble endogenous form of N-cadherin,
which is generated by proteolysis during retinal development
cells: Effects of PTH and extracellular calcium on localization.
Life Sci 54:PL201–PL208.
and retains adhesive and neurite-promoting function. J Neuro- 64. Liu P, Oyajobi BO, Croucher P, Stringer B, Russell RGG,
sci Res 36:33–45.
Scutt A 1997 Regulation of cadherin expression in human
osteoprogenitor cells by osteotropic hormones and growth
factors. Bone 20:73S (abstract).
56. Trolice MP, Pappalardo A, Peluso JJ 1997 Basic fibroblast
growth factor and N-cadherin maintain rat granulosa cell and
ovarian surface epithelial cell viability by stimulating the ty-
rosine phosphorylation of the fibroblast growth factor recep-
tors. Endocrinology 138:107–113.
57. Day ML, Zhao X, Vallorosi CJ, Putzi M, Powell CT, Lin C,
Day KC 1999 E-cadherin mediates aggregation-dependent sur-
vival of prostate and mammary epithelial cells through the
retinoblastoma cell cycle control pathway. J Biol Chem 274:
9656–9664.
Address reprint requests to:
Irene Hunter
Department of Biomedical Sciences
Institute of Medical Sciences
University of Aberdeen
Foresterhill
Aberdeen AB25 2ZD, Scotland
58. Huber O, Bierkamp C, Kemler R 1996 Cadherins and catenins
in development. Curr Opin Cell Biol 8:685–691.
59. Steinhusen U, Badock V, Bauer A, Beherens J, Wittman-
Leibold B, Dorken B, Bommer K 2000 Apoptosis-induced Received in original form June 8, 2000; in revised form September
cleavage of ␤-catenin by caspase-3 results in proteolytic frag- 21, 2000; accepted October 30, 2000.