ARTHRITIS & RHEUMATISM
Vol. 58, No. 7, July 2008, pp 2075–2087
© 2008, American College of Rheumatology
Differential Expression of GADD45?
in Normal and Osteoarthritic Cartilage
Potential Role in Homeostasis of Articular Chondrocytes
Kosei Ijiri,1Luiz F. Zerbini,1Haibing Peng,1Hasan H. Otu,1Kaneyuki Tsuchimochi,1
Miguel Otero,1Cecilia Dragomir,2Nicole Walsh,1Benjamin E. Bierbaum,3
David Mattingly,3Geoff van Flandern,3Setsuro Komiya,4Thomas Aigner,5
Towia A. Libermann,1and Mary B. Goldring1
Objective. Our previous study suggested that
growth arrest and DNA damage–inducible protein 45?
(GADD45?) prolonged the survival of hypertrophic
chondrocytes in the developing mouse embryo. This
study was undertaken, therefore, to investigate whether
GADD45? plays a role in adult articular cartilage.
Methods. Gene expression profiles of cartilage
from patients with late-stage osteoarthritis (OA) were
compared with those from patients with early OA and
normal controls in 2 separate microarray analyses.
Histologic features of cartilage were graded using the
Mankin scale, and GADD45? was localized by immu-
nohistochemistry. Human chondrocytes were trans-
duced with small interfering RNA (siRNA)–GADD45?
or GADD45?-FLAG. GADD45? and COL2A1 messen-
ger RNA (mRNA) levels were analyzed by real-time
reverse transcriptase–polymerase chain reaction, and
promoter activities were analyzed by transient transfec-
tion. Cell death was detected by Hoechst 33342 staining
of condensed chromatin.
Results. GADD45? was expressed at higher levels
in cartilage from normal donors and patients with early
OA than in cartilage from patients with late-stage OA.
All chondrocyte nuclei in normal cartilage immuno-
stained for GADD45?. In early OA cartilage, GADD45?
was distributed variably in chondrocyte clusters, in
middle and deep zone cells, and in osteophytes. In
contrast, COL2A1, other collagen genes, and factors
associated with skeletal development were up-regulated
in late OA, compared with early OA or normal cartilage.
In overexpression and knockdown experiments,
GADD45? down-regulated COL2A1 mRNA and pro-
moter activity. NF-?B overexpression increased
GADD45? promoter activity, and siRNA-GADD45? de-
creased cell survival per se and enhanced tumor necro-
sis factor ?–induced cell death in human articular
Conclusion. These observations suggest that
GADD45? might play an important role in regulating
chondrocyte homeostasis by modulating collagen gene
Dr. Zerbini’s work was supported by US Department of
Defense grant PC-051217. Dr. Aigner’s work was supported by DFG
grant 20/A.7-1. Dr. Goldring’s work was supported by NIH grant
1Kosei Ijiri, MD, PhD (current address: Kagoshima Univer-
sity, Kagoshima, Japan), Luiz F. Zerbini, PhD, Haibing Peng, MD,
Hasan H. Otu, PhD (current address: Yeditepe University, Istanbul,
Turkey), Kaneyuki Tsuchimochi, MD, PhD (current address: Hospital
for Special Surgery, and Weill College of Medicine of Cornell Uni-
versity, New York, New York), Miguel Otero, PhD (current address:
Hospital for Special Surgery, and Weill College of Medicine of Cornell
University, New York, New York), Nicole Walsh, PhD, Towia A.
Libermann, PhD, Mary B. Goldring, PhD (current address: Hospital
for Special Surgery, and Weill College of Medicine of Cornell Uni-
versity, New York, New York): Beth Israel Deaconess Medical Center,
and New England Baptist Bone and Joint Institute, Boston, Massa-
chusetts;2Cecilia Dragomir, MD: Hospital for Special Surgery, and
Weill College of Medicine of Cornell University, New York, New
York;3Benjamin E. Bierbaum, MD, David Mattingly, MD, Geoff van
Flandern, MD: New England Baptist Hospital, Boston, Massachusetts;
4Setsuro Komiya, MD, PhD: Kagoshima University, Kagoshima, Ja-
pan;5Thomas Aigner, MD, DSc: Institute of Pathology, and University
of Leipzig, Leipzig, Germany.
Dr. Mattingly has received consulting fees, speaking fees,
and/or honoraria (more than $10,000) from DePuy, a Johnson &
Johnson company, and he receives royalties from DePuy for the
S-ROM hip replacement system.
Address correspondence and reprint requests to Mary B.
Goldring, PhD, Laboratory for Cartilage Biology, Hospital for Special
Surgery, Caspary Research Building, Room 528, 535 East 70th Street,
New York, NY 10021. E-mail: firstname.lastname@example.org.
Submitted for publication March 2, 2007; accepted in revised
form February 29, 2008.
expression and promoting cell survival in normal adult
cartilage and in early OA.
Osteoarthritis (OA) is defined largely as a disease
of cartilage, since chondrocytes, which constitute the
unique cellular component of adult articular cartilage,
are able to respond to mechanical injury, joint instability
due to genetic factors, and biologic stimuli such as
cytokines and growth and differentiation factors. In
young individuals without genetic abnormalities, biome-
chanical factors due to trauma are strongly implicated in
initiating the OA lesion (1,2). Mechanical disruption of
cell–matrix interactions may lead to aberrant chondro-
cyte behavior that is reflected in the appearance of
fibrillations, cell clusters, and changes in quantity, dis-
tribution, or composition of matrix proteins (3). In the
early stages of OA, a transient increase in chondrocyte
proliferation is associated with increased synthesis of
catabolic cytokines and matrix-degrading enzymes. Lo-
cal loss of proteoglycans and cleavage of type II collagen
occur initially at the cartilage surface, resulting in an
increase in water content and a loss of tensile strength in
the cartilage matrix as the lesion progresses.
A number of studies have demonstrated en-
hanced biosynthesis and increased global gene expres-
sion of aggrecan and type II collagen in human OA
cartilage (4,5). The increased levels of anabolic factors
such as bone morphogenetic protein 2 (BMP-2) and
inhibin ?A/activin suggest a possible mechanism of
cartilage anabolism (5–7). Nevertheless, Aigner and
coworkers have shown that expression of the type II
collagen gene (COL2A1) is suppressed in upper zones
with progressing matrix destruction, whereas global
COL2A1 gene expression is increased in late-stage OA
cartilage compared with normal and early degenerative
cartilage (8–10). Phenotypic modulation, with expres-
sion of collagens normally absent in adult articular
cartilage or at atypical sites in OA cartilage, has been
proposed (11). Once the cartilage is severely degraded,
the chondrocyte is unable to replicate the complex
arrangement of collagen laid down during development.
Furthermore, the chondrocyte stress response may re-
sult in the loss of viable cells due to apoptosis or
Growth arrest and DNA damage–inducible pro-
tein 45? (GADD45?) is a member of a family of small
(18-kd) proteins that respond to genotoxic stress. Ini-
tially, GADD45?, encoded by myeloid differentiation
factor 118, was identified as a primary response gene
activated in murine myeloid leukemia cells by
interleukin-6 during terminal differentiation (13,14).
Microarray studies have shown that another family
member, GADD45?, is expressed at higher levels in
normal cartilage than in OA (9) and is induced by
hydrostatic pressure (15). Recently, we identified
GADD45? as a BMP-2–induced early gene in chondro-
cytes and as an essential mediator of Col10a1 and
Mmp13 gene expression in late-stage hypertrophic chon-
drocytes in the mouse embryo (16). Since the hypertro-
phic zone was compressed in Gadd45b-deficient embry-
onic growth plates, we hypothesized that GADD45?
may act as a cell survival factor during terminal differ-
In 2 separate global analyses of gene expression
in cartilage, we showed in this study that chondrocytes
express GADD45? messenger RNA (mRNA) at higher
levels in cartilage from normal donors and patients with
early OA than in cartilage from patients with late-stage
OA. However, the distribution of GADD45? in chon-
drocyte clusters and in osteophytes in OA cartilage
suggests different roles of GADD45? in OA compared
with normal cartilage, where GADD45? is localized in
the nuclei of all chondrocytes. Interestingly, overexpres-
sion of GADD45? down-regulates COL2A1 promoter
activity in chondrocytes, consistent with the reciprocal
expression of the 2 genes in cartilage from patients with
early and late OA. GADD45? gene transcription is
induced by NF-?B, and endogenous levels of GADD45?
promote cell survival and protect against tumor necrosis
factor ? (TNF?)–induced cell death in chondrocytes.
These results may explain how cartilage damage and
other environmental stresses influence the survival and
anabolic activities of chondrocytes in OA cartilage.
MATERIALS AND METHODS
Cartilage samples. Human articular cartilage samples
were obtained from discarded surgical material or from cadav-
ers, with the approval of the Institutional Review Boards of the
Beth Israel Deaconess Medical Center (BIDMC), New En-
gland Baptist Hospital, the University of Erlangen–Nuremberg
(Erlangen, Germany), the Graduate School of Medicine and
Dentistry, Kagoshima University, and the Hospital for Special
RNA isolation and microarray analysis. Cartilage sam-
ples were rapidly frozen in liquid nitrogen, and total RNA was
isolated as previously described (17,18). Transcription profil-
ing of RNA samples derived from cartilage from 3 patients
with early OA (mean age 72 years) and 3 patients with late OA
(mean age 78 years) was performed at the BIDMC Genomics
Center, using the GeneChip Human Genome U133A array
(Affymetrix, Santa Clara, CA) containing 22,283 genes. Array
experiments were performed according to the recommenda-
tions of the manufacturer with 1 ?g of total RNA per sample.
2076IJIRI ET AL
After polymerase chain reaction (PCR) amplification of each
sample, complementary DNA fragments were hybridized with
a pre-equilibrated Affymetrix chip, washed, stained, and
scanned in the HP ChipScanner (Affymetrix), as previously
described (19). Microarray analysis of RNA extracts from
cartilage samples from 13 normal individuals (mean age 71.7
years [range 48–87 years]) and 12 patients with late-stage OA
(mean age 64.7 years [range 60–84 years]) diagnosed accord-
ing to the American College of Rheumatology criteria (20),
was performed at the Mu ¨nster University Genomics Center
(Mu ¨nster, Germany), using the Affymetrix GeneChip Human
Genome U133 Plus 2.0 Array for ?47,000 transcripts.
Analysis with dChip software. Scanned array images
were analyzed using dChip (21), which has been shown to be
more robust than Affymetrix Microarray Analysis Suite soft-
ware, version 5.0, in signal calculation for ?60% of genes (22).
In the dChip analysis, a smoothing spline normalization
method was applied prior to obtaining model-based gene
expression indices, also known as signal values. Single, array,
and probe outliers were interrogated as described in dChip,
where image spikes are treated as single outliers. Since no
outlier chip was identified, all samples were used for subse-
quent analysis. To compare cartilage samples from patients
with early and late-stage OA, we used the lower confidence
bound, which is a stringent estimate of the fold change and has
been shown to be the better ranking statistic (23). We used the
dChip method to assess differentially expressed genes accord-
ing to the lower confidence bound, since it is superior to other
commonly used approaches, such as the Affymetrix Microarray
Analysis Suite 5.0 algorithm and Robust Multiarray Average
(24,25). If the lower confidence bound of the fold change
between the experiment and the baseline was ?1.2, we con-
sidered the corresponding gene to be differentially expressed.
Studies using custom arrays and quantitative real-time reverse
transcriptase (RT)–PCR have suggested that Affymetrix chips
may underestimate differences in gene expression (26). Based
on this work and others (27), a criterion of selecting genes that
have a lower confidence bound ?1.2 most likely corresponds
to genes with an “actual” fold change of ?3 in gene expression.
Real-time RT-PCR. Total RNA (1.0 ?g) was reverse-
transcribed in 20 ?l containing final concentrations of 2.4
IU/?l of murine leukemia virus reverse transcriptase, 2.5 ?M
of oligo(dT)16, and 1 unit/?l of RNase inhibitor, all obtained
from Applied Biosystems (Foster City, CA). Amplifications
were carried out using SYBR Green I–based real-time PCR on
the MJ Research DNA Engine OpticonTM Continuous Fluo-
rescence Detection System (MJ Research, Waltham, MA), as
previously described (16,28). For each run, serial dilutions of
GAPDH plasmids were used as standards for quantitative
measurement of the amount of amplified DNA. All samples
were run in triplicate, and the data were calculated as the ratio
of GADD45? or COL2A1 to GAPDH. The primers used for
real-time PCR were as follows: for GADD45?, 5?-TCGGATT-
TTGCAATTTCTCC-3? (sense) and 5?-GGATGAGCGTGA-
AGTGGATT-3? (antisense); for human GAPDH, 5?-CAA-
AGTTGTCATGGATGACC-3? (sense) and 5?-CCATGGA-
GAAGGCTGGGG-3? (antisense); and for human COL2A1,
5?-CAACACTGCCAACGTCCAGAT-3? (sense) and 5?-
Immunohistochemistry. Articular cartilage samples
were obtained from hip joints from 4 asymptomatic donors
(mean age 81.5 years) undergoing joint replacement for fem-
oral neck fracture and from knee joints from 10 patients (mean
age 74 years) undergoing total knee arthroplasty because of
symptomatic OA. Multiple areas of cartilage (0.5 ? 0.5 cm2)
were fixed in 4% paraformaldehyde for 4 hours at room
temperature and embedded in paraffin. Serial sections of 8 ?m
were cut and stained with toluidine blue. Histologic features
were assessed, and sections were assigned a grade on the
Mankin scale (29). Sections of fixed and paraffin-embedded
cartilage were deparaffinized and blocked with normal horse
serum or protein block (Dako x0909; Dako, Carpinteria, CA).
Immunohistochemical analysis for GADD45? was performed
as previously described (16,30), using goat polyclonal anti-
GADD45? (sc-8776; Santa Cruz Biotechnology, Santa Cruz,
CA) in a 1:400 dilution (0.5 ?g/ml final concentration), rabbit
biotinylated anti-goat IgG (Sigma, St. Louis, MO), and the
Vectastain Elite ABC kit (Vector, Burlingame, CA.). Sections
were counterstained with hematoxylin. For negative controls,
normal goat IgG (sc-2028) was used in place of the primary
antibody against GADD45?.
Cell culture. The immortalized human chondrocyte
cell line, C-28/I2, was cultured in Dulbecco’s modified Eagle’s
medium (DMEM)–Ham’s F-12 (1/1 [volume/volume]) (In-
vitrogen, San Diego, CA) containing 10% fetal calf serum
(FCS) (Biowhittaker, Walkersville, MD), as previously de-
scribed (31,32). For experiments, subconfluent cultures were
moved to medium containing 1% Nutridoma-SP (Roche,
Indianapolis, IN) or 1% serum for 18 hours for mild starvation.
Primary chondrocytes were isolated by sequential digestion
with Pronase and collagenase P (Invitrogen) from human
articular cartilage obtained from intact regions of femoral
condyles from patients undergoing total knee replacement
surgery, and cultured to confluence in DMEM–Ham’s F-12
containing 10% FCS.
Transient transfections using luciferase reporter con-
structs and expression plasmids. The human GADD45?
promoter fragment spanning ?1,604 to ?141 bp was prepared
by PCR using human genomic DNA (BD Clontech, Basing-
stoke, UK) as template, as previously described (16). The
?496 to ?141 bp construct was prepared by PCR using the
PfuTurbo DNA polymerase (Stratagene, La Jolla, CA) and the
?1,604 to ?141 bp construct as template. The COL2A1
sequence spanning ?577 to ?3,428 bp was cloned into the
pGL2-basic (pGL2-B) luciferase reporter gene vector (Pro-
mega, Madison, WI) to generate pGL2B-COL2 (33). The
pcDNA3-GADD45?-FLAG and pcDNA3-FLAG empty vec-
tor and the NF-?B p65 and p50 expression vectors have been
reported previously (16,34). The transduction of C-28/I2 cells
with lentiviral–small interfering RNA (siRNA)–green fluores-
cent protein (GFP) (GFP–knockdown [KD]) or lentiviral-
siRNA-GADD45? (GADD45?-KD) and tests for specificity
and efficiency of knockdown have been described previously
(16). Transient transfection experiments were carried out using
C-28/I2 cells with Plus and Lipofectamine reagents (Invitro-
gen), as previously described (16,32). Luciferase activity was
determined by the Dual Luciferase Assay using the Autolumat
LB953 luminometer (EG&G Berthold, Oak Ridge, TN). Each
experiment was repeated at least 3 times and each data point
was calculated as the mean ? SD of 3–6 wells per experiment.
GADD45? IN NORMAL AND OA CARTILAGE2077
Hoechst staining. Primary human articular chondro-
cytes, obtained from 5 patients (2 men and 3 women) with OA
(age range 53–70 years) who underwent total knee replace-
ment, were passaged at a density of 2.5 ? 104cells/cm2and
transfected with 50 nM of siRNA oligos against GFP (sense
GCAAGCUGACCCUGAAGUUCAU and antisense GAA-
CUUCAGGGUCAGCUUGCCG) or against GADD45?
(Hs_GADD45B_6_HP siRNA; Qiagen, Chatsworth, CA) us-
ing Plus and Lipofectamine reagents. Transfection efficiency
was assessed using siRNA GFP conjugated with rhodamine.
The GADD45? siRNA sequences were selected from among 5
sets tested at 48 and 72 hours, and knockdown was confirmed
by real-time PCR in the Opticon 2 Real Time PCR Detector
System (Bio-Rad, Richmond, CA) (see Supplementary Figure
1, available on the Arthritis & Rheumatism Web site at http://
suppmat/). Seventy-two hours after transfection, cells were
treated with 50 ng/ml TNF? in serum-free medium for 24
hours, and cell death was assessed by staining with 10 ?g/ml
Hoechst 33342 dye (Invitrogen) to detect chromatin conden-
sation. Cell death was defined by condensed and/or frag-
mented chromatin in the blue fluorescence–emitting nuclei. At
least 200 cells from randomly selected fields were counted in
each experiment. The significance of the differences was
estimated using analysis of variance followed by Student’s
t-test. P values less than 0.05 were considered significant.
Higher levels of GADD45? expression in early
OA than in late OA. Based on our previous study (16),
we hypothesized that GADD45? might be differentially
expressed in adult articular cartilage. Our initial screen
used total RNA isolated from cartilage samples from 3
patients with early OA and 3 patients with late OA.
Microarray analysis on Affymetrix GeneChip U133A
identified 358 transcript variants (lower confidence
bound ?1.2) that were highly expressed genes in carti-
lage from patients with early OA. GADD45? was ex-
pressed 2.66-fold higher in early OA than in late OA
(see Supplementary Table 1, available on the Arthritis &
Rheumatism Web site at http://www.mrw.interscience.
Other genes up-regulated prominently in early
OA encoded intracellular regulators of cell proliferation
(DDX11, CCNG1, HRASLS3, CDC25C, CHEK1,
H2AFN, H2AFB, RAB26), mitochondrial function
(OXA1L, TOMM20, TIMM17A, CYP4B1, ACADL,
COX7C, CTAG1), cytoskeletal function (TMOD,
(SCNN1A, SCN2B, ACADL, CLDN4, NTT5), intracell-
ular transport (LTF, GRM1, TFR2), signal transduction
(PIK3R1, MAP2K5, SWAP1, PDE9A, SSR4, PIK4CB,
GRAF, MAPK4, MSTIR, MAPK14), transcription
TLN4), ion transport
(RREB1, ETV5, FOXG1A, SAP18, ZNF144, ZNF254,
ZNF277, ZNF363, SRY, SOX17), and protein transla-
tion and modification (EIF4B, EIF4EBP2, RABBG-
GTB, MAOA, ASB13). Genes associated with inflam-
mation (SAA2, C3, TNFRSF8, IL1R1), angiogenesis
(FLT1, TIE, EPHB4, NRP2, BAI3), cell adhesion
(CDH7, GPC3, CELSR1, ITGA2, ITGA8, ITGB4,
PCDH11), matrix degradation (CSNK2A2, ADAM11,
ADAM23, MMP10), and skeletal development (FGF4,
FGF5, FGF16, BMP7, MADHIP, FST, DVL2, PTCH,
HOXA4, HOXB6, ALP1) were also up-regulated in
early OA (Supplementary Table 1).
Real-time PCR confirmed the differential ex-
pression of GADD45? mRNA in early and late, or
severe, OA cartilage samples analyzed by microarray
(Figure 1A). Interestingly, severe OA sample 3 was an
autopsy specimen with histologic features consistent
with a less severe stage of OA than sample 1 or sample
2. We decided to include these data in this report, since
they illustrate the importance of uniform handling of
material and because the Self Organizing Map analysis
also indicated sample 3 as an outlier. The Self Organiz-
ing Map algorithm groups genes that have similar ex-
pression patterns within the same cluster.
As shown in Figure 1B, 30 distinct clusters were
generated, of which 21 (c0–c11, c13–c17, and c20–c23)
contained genes that were primarily up-regulated in
severe OA sample 1 and severe OA sample 2, and 8 (c12,
c18, c19, and c24–c28) contained genes that were up-
regulated in early OA sample 1, early OA sample 2, and
early OA sample 3. The cluster c29 contained genes that
were only slightly up-regulated in early OA. In some, but
not all, of the clusters, severe OA sample 3 was an
outlier, particularly in clusters c0–c5. GADD45? clus-
tered in c25 with other intracellular mediators involved
in transport, mitochondrial function, and cell cycle reg-
ulation. The analysis of all 824 differentially expressed
genes indicated that many of the same genes were
down-regulated in severe OA samples 1, 2, and 3, as
shown for the top 54 differentially expressed genes in the
heat map in Figure 1C. In severe OA sample 3, however,
most of the genes up-regulated in severe OA sample 1
and severe OA sample 2 were down-regulated, as in the
early OA samples.
Higher levels of GADD45? expression in normal
cartilage than in cartilage from patients with late-stage
OA. To determine whether GADD45? expression in
normal cartilage was up- or down-regulated compared
with OA cartilage, we analyzed a data set from Af-
fymetrix GeneChip Whole Genome U133 Plus 2.0, in
2078IJIRI ET AL
Figure 1. Comparison of gene expression patterns in cartilage samples from patients with early osteoarthritis
(OA) and patients with late-stage OA. A, Growth arrest and DNA damage–inducible protein ? (GADD45?)
expression in cartilage samples from patients with early OA and patients with late-stage OA. RNA extracts from
cartilage samples obtained from 3 patients with early OA (eOA1–eOA3) and 3 patients with late, or severe, OA
(sOA1–sOA3) were analyzed by quantitative real-time polymerase chain reaction. Each sample was analyzed in
triplicate. Bars show the mean and SD ratio of GADD45? to GAPDH. B, Self Organizing Map analysis of gene
expression in cartilage samples from patients with early or severe OA. The algorithm generated 30 distinct
clusters (c0–c29), each representing genes with a similar expression pattern. Each square in the grid corresponds
to 1 cluster. The number in the top left of each square is the cluster ID. The yellow box around c0 marks the first
cluster obtained in the analysis. The number in the top middle of each square is the number of genes in the
cluster. Blue dots represent the average expression patterns of the genes in each cluster in (from left to right)
severe OA sample 1, severe OA sample 2, severe OA sample 3, early OA sample 1, early OA sample 2, and early
OA sample 3. Red lines represent the variation from the average expression pattern. Note that severe OA sample
3 is an outlier with respect to genes that are up-regulated in severe OA. C, Hierarchical clustering analysis of the
microarray data from cartilage samples obtained from 3 patients with early OA and 3 patients with late OA,
analyzed using Affymetrix GeneChip U133A. Up-regulated genes are shown in red; down-regulated genes are
shown in green. Black indicates no change. GADD45? gene expression was up-regulated in early OA, along with
genes involved in cell cycle regulation, such as DDX11, HRASLS3, and CCNG1, and in mitochondrial function,
such as OXA1L and TOMM20, as well as other genes involved in intracellular functions. The genes encoding
types I, II, III, and V collagen were all up-regulated in late OA compared with early OA, as were ASPN, MTN3,
and the Wnt-associated genes, WIF1 and DKK3.
GADD45? IN NORMAL AND OA CARTILAGE2079
which total RNA extracts from cartilage samples from 13
normal donors and 12 patients with late-stage OA were
compared. These data were subjected to bioinformatics
analysis by dChip in the BIDMC Genomics Center
(http://www.bidmcgenomics.org/), using lower confi-
dence bound ?1.2. As shown in the dChip analysis in
Figure 2, all 3 transcripts of GADD45? detected by the
GeneChip were markedly down-regulated in OA com-
pared with normal cartilage samples. Consistent with a
previous report by Aigner et al (9), GADD45? was also
decreased in OA relative to normal cartilage, whereas
GADD45? was expressed at low levels with a lower
confidence bound close to 1.2 (Figure 2A).
Immunohistochemical analysis of GADD45? ex-
pression in adult human articular cartilage. Since ana-
lysis of global gene expression in extracts of whole
cartilage samples does not take into account the cellular
distribution, we investigated the protein expression of
GADD45? in normal and OA cartilage at different
stages to obtain some clues about its cellular localization
and function. Immunohistochemistry demonstrated nu-
clear localization of GADD45? in all chondrocytes
where it was expressed. GADD45? was present in
chondrocytes throughout the superficial and middle
zones of normal cartilage (Figure 3A). In early OA,
some chondrocytes expressed GADD45? strongly, but
with variable distribution in the middle and superficial
zones or in the deep zones. In late OA, few chondrocytes
stained for GADD45? (Figure 3A). GADD45? was
strongly expressed in chondrocyte clusters and also
observed in cells throughout osteophytes (Figure 3B).
Up-regulation of COL2A1 and other extracellu-
lar matrix genes in cartilage from patients with late-
stage OA. Analysis using Affymetrix GeneChip U133A
identified 466 transcripts (lower confidence bound ?1.2)
that were highly expressed in late, or severe, OA (see
Supplementary Table 2, available on the Arthritis &
archal clustering analysis indicated that genes associated
with extracellular matrix synthesis and anabolism were
up-regulated in late OA (Figure 1C). COL2A1 was
among the most highly expressed genes in clusters 0 and
1. The collagen genes, COL3A1, COL1A1, COL1A2,
COL15A1, which were often represented more than
once in clusters 0, 1, 2, 3, 4, and 5, were all strongly
up-regulated in cartilage from patients with late OA
compared with cartilage from patients with early OA
(Supplementary Table 2). COL9A3, COL11A1,
COL9A2, and COL18A1 were found in clusters 6, 8, 20,
and 22, respectively. Genes involved in posttranslational
modification of collagens (PCOLCE, PLOD, P4HA2,
LOXL1, and BMP1) and other matrix-associated genes
such as those for HXB, TNXB, OGN (mimecan), BGN,
FN1, LUM, CSPG2, CHAD, SDC1, and SPARC were
also prominently represented.
Of interest were the other most highly expressed
genes, including asporin (ASPN), matrilin 3 (MTN3),
CILP, AQP1, S100A4, SGK, IGFBP7, IGFBP3,
Figure 2. Microarray analysis of normal (N) and OA cartilage. RNA
extracts from cartilage samples from 13 normal donors and 12 patients
with late-stage OA were analyzed using Affymetrix GeneChip U133
Plus 2.0, and the data were subjected to bioinformatic analysis with
dChip software, using a lower confidence bound (LCB) of ?1.2. A,
Fold change (FC) in expression of GADD45?, GADD45?, and
GADD45? transcripts in normal compared with OA cartilage. Values
are the mean ? SEM. The probe ID is the Affymetrix identifier for
separate probes on the microarray. B–D, Individual values in each
normal and OA cartilage sample for each of the 3 GADD45? probes.
See Figure 1 for other definitions.
2080IJIRI ET AL
Figure 3. Immunohistochemical analysis of GADD45? expression in representative specimens of
human adult articular cartilage, including those used in the microarray analyses. A, Cartilage
sections from a normal donor (Mankin grade 0) (a–c), a patient with early OA (Mankin grade 4)
(d–f), and a patient with late-stage OA (Mankin grade 8) (g–i) were stained with toluidine blue or
subjected to immunohistochemistry using anti-GADD45? or normal goat IgG (negative control).
Arrows in b and e show single cells with nuclear staining (original magnification ? 40). B, Sections
with chondrocyte clusters in the middle zone from a patient with early OA (Mankin grade 4) (a–c)
and an osteophyte from a patient with late OA (Mankin grade 8) (d–f) were stained with toluidine
blue or immunostained with anti-GADD45? or normal goat IgG. Boxed areas in a and d are shown
at higher magnification in b and e. See Figure 1 for other definitions.
GADD45? IN NORMAL AND OA CARTILAGE2081
ANXA2, DKK3, WIF1, CTKSF1B1, POSTN, and
LTBP1, that play a role in cartilage development and
homeostasis. Other genes potentially involved in inflam-
mation, angiogenesis, and stress responses, including
PTGES, VCAM1, TNFSF11, TNFSF10, TNFRSF11B
(osteoprotegerin), ANGPTL2, PLA2G4A, PRKCA,
AXL, and GAS7, heat-shock proteins 22 (HSPB8), 27
(HSPB1), 47 (SERPINH1), 70 (HSPA1B), and 90
(HSPCA), and genes involved in tissue catabolism
(SERPINA5, SERPINE1, PRSS23, CTSK, and
MMP13) were also up-regulated. SDRG1, CRTAC1,
FAP, MIA/CD-RAP, and NOTCH2 are of potential
interest as markers in cartilage from patients with late
OA (Supplementary Table 2).
Negative regulation of COL2A1 mRNA and pro-
moter activity by GADD45?. To further explore the
reciprocal relationship between GADD45? and matrix
gene expression observed in our microarray analyses, we
small interfering RNA (siRNA) blockade of endogenous growth arrest
and DNA damage–inducible protein ? (GADD45?). A, Total RNA
was extracted from cultures of C-28/I2 cells transduced with lentiviral
siRNA–green fluorescent protein (GFP) (GFP–knockdown [KD]) or
siRNA-GADD45? (GADD45?-KD) (left). Quantitative real-time re-
verse transcriptase–polymerase chain reaction analysis was performed
to determine the levels of COL2A1 mRNA normalized to GAPDH
mRNA. GFP-KD and GADD45?-KD C-28/I2 cells were transfected
with pGL2-B/4.0, and COL2A1 promoter–driven luciferase activity
was measured (right). B, C-28/I2 cells were cotransfected with pGL2-
B/4.0 (COL2A1, spanning ?577 to ?3,428 bp) and pcDNA3-FLAG
(control) or increasing amounts of GADD45?-FLAG, as indicated.
Bars show the mean and SD of triplicate wells.
Increased COL2A1 mRNA and promoter activity after
Figure 5. Induction of growth arrest and DNA damage–inducible
protein ? (GADD45?) by NF-?B in chondrocytes. C-28/I2 cells were
cotransfected with pGL2-GADD45? promoter constructs ?1,604 to
?141 bp (A) and ?496 to ?141 bp (B), and the empty vector, pCI, or
expression vector encoding the NF-?B subunit p65 alone, the NF-?B
subunit p50 alone, or both p65 and p50. Bars show the mean and SD.
2082 IJIRI ET AL
analyzed the influence of endogenous or overexpressed
GADD45? on COL2A1 mRNA levels and promoter
activity in the human chondrocyte C-28/I2 cells trans-
duced with lentiviral siRNA-GFP (GFP-KD control) or
siRNA-GADD45? (GADD45?-KD) (16). The levels
of COL2A1 mRNA, analyzed by quantitative real-time
RT-PCR, and COL2A1 promoter activity were in-
creased by GADD45? silencing (Figure 4A). Over-
expression of GADD45? suppressed COL2A1 promoter
activity in a dose-dependent manner. Cotransfection
with GADD45?-FLAG inhibited COL2A1 promoter
activity compared with cotransfection with the control
pcDNA3-FLAG vector (Figure 4B). These results sug-
gest that GADD45? expression may be associated
with low collagen turnover in cartilage in physiologic
GADD45? is a survival factor in human articular
chondrocytes. Since NF-?B is known to induce
GADD45? during genotoxic and oxidative stress (28,35)
and to play an important role in TNF?-induced apopto-
sis in chondrocytes, we hypothesized that the induction
of GADD45? by NF-?B could represent a protective
feedback mechanism for cell survival, as shown in pre-
vious studies of other cell types (36,37). In addition,
several genes associated with apoptosis were differen-
tially expressed in early and late OA, including apoptosis
antagonizing transcription factor (AATF), caspase 10
(CASP10), BCL2-associated athanogene 3 (BAG3),
apoptosis-related protein (APR), and apoptosis-related
cysteine proteinase (Supplementary Tables 1 and 2).
As shown in Figure 5, cotransfection of the
NF-?B subunits p65 and p50 up-regulated GADD45?
promoter activity in C-28/I2 chondrocytes. Compared
with the empty vector, p65 alone, but not p50, stimulated
the activity of both promoter constructs. In cotransfec-
tions with both p65 and p50, the level of ?1,604 to ?141
bp promoter activity was ?2-fold greater than in the
presence of p65 or p50 alone. Of note, the activity of the
shorter ?496 to ?141 bp promoter in the presence of
p65 alone was not significantly different from that in the
presence of both p65 and p50, consistent with the
location of functional NF-?B sites described previously
(35). We then examined whether GADD45? could serve
as a cell survival signal, as reported for the NF-?B
response in other cell types (28,36,37). Knockdown of
GADD45? expression after transfection of siRNA-
GADD45? in cultured human articular chondrocytes
isolated from 5 patients increased cell death, measured
by Hoechst staining of condensed chromatin, in both the
absence and the presence of TNF? (Figure 6). These
results indicated that GADD45? was a prosurvival signal
in chondrocytes and that endogenous levels were suffi-
cient to protect against cell death.
Based on our recent findings that highlighted a
new function of GADD45? in hypertrophic chondro-
cytes during endochondral ossification (16), we investi-
gated whether this stress response factor might play a
Figure 6. Growth arrest and DNA damage–inducible protein ?
(GADD45?) is a cell survival factor in human articular chondrocytes.
Primary cultures were passaged at 2.5 ? 104cells/cm2, grown to
confluence, transfected with small interfering RNA (siRNA)–GFP
(siGFP) or siRNA-GADD45? oligos, and treated 72 hours later with
tumor necrosis factor ? (TNF?) for 24 hours. A, Cell death was
evaluated by staining with Hoechst 33342 and counting the number of
nuclei with brightly stained condensed chromatin. Bars show the
mean ? SD percentage of cells showing nuclear condensation and
fragmentation relative to the total number of cells counted (?200 cells
per data point). Results are from ?4 independent experiments, each
performed in triplicate. B, Representative photomicrographs of
Hoechst-stained cells from 1 set of cultures isolated from a single
patient are shown.
GADD45? IN NORMAL AND OA CARTILAGE2083
role in chondrocyte survival in adult articular cartilage.
In our initial screen of a small number of samples of
cartilage with well-characterized pathology, we found,
surprisingly, a clear distinction between genes up-
regulated in early and late OA, using GADD45? and
COL2A1, respectively, as “marker” genes. The down-
regulation of GADD45? in late (moderate to severe)
OA was confirmed using a much larger data set from the
U133 Plus 2.0 array containing the whole human ge-
nome. Along with the Self Organizing Map analysis
showing that one of the late OA samples was an outlier,
these data demonstrated the value of comparisons
among individual samples. The consistency and statisti-
cal significance of differences in global gene expression
were surprising because of the heterogeneity in prolif-
erative and synthetic activities that we would expect at
any given time among cells within individual human
cartilage samples. Immunohistochemical analysis of in-
dividual cartilage samples also supported the assump-
tions derived from the microarray data. The expression
of GADD45? in all chondrocytes throughout normal
articular cartilage, in addition to its localization in early
OA cartilage at sites peripheral to the lesion and in
chondrocyte clusters and osteophytes, along with our
finding that endogenous levels protect against cell death
in articular chondrocyte cultures, suggests a role of
GADD45? as a survival factor in chondrocytes in both
health and disease.
Consistent with the low levels of COL2A1 in
normal and early OA cartilage, GADD45? down-
regulated COL2A1 promoter activity and mRNA levels
in chondrocyte cultures. In the mouse embryo,
GADD45? is required for expression of the type X
collagen gene during chondrocyte hypertrophy (16), and
our present results suggest that it may also be associated
with decreased COL2A1 transcription during terminal
differentiation. The increased expression of the vascular
endothelial growth factor receptors, FLT1 and NRP2,
which mediate angiogenesis during endochondral ossifi-
cation, and other genes associated with skeletal devel-
opment in cartilage samples from patients with early
OA, supports this concept. Thus, GADD45? may be an
important signal that serves as the molecular link com-
mon to pathways of cartilage remodeling in endochon-
dral ossification and in OA.
Previous studies have shown that gene expression
and synthesis of type II collagen are up-regulated in OA
cartilage even though catabolic activity leads finally to
the destruction of the matrix (38–40). Thus, the endog-
enous levels of GADD45? in normal cartilage may
reflect the low matrix synthetic activity in quiescent
chondrocytes under low turnover conditions. The up-
regulation of other matrix genes, including those for
types I, III, and V collagens and MTN3, in late OA may
reflect, however, an aberrant attempt at repair. Despite
the high levels of COL2A1 in late OA cartilage, SOX9,
the master switch for the COL2A1 phenotype, was not
differentially expressed, consistent with findings that
SOX9 mRNA is decreased near the lesions in OA
cartilage (41), and that SOX9 and COL2A1 expression
do not colocalize in adult articular cartilage (42). The
differential up-regulation of transcription factors, such
as MAF, SOX4, PMX1, FosB, RUNX1, DLX4, ELF2,
NFAT5, and zinc-finger protein 42 (ZNF42), ZNF131,
and ZNF137, suggests alternative mechanisms for regu-
lating the other differentially expressed genes in carti-
lage in patients with late OA, whereas distinct transcrip-
tion factors, including RREB1, ETV5, FOXG1A,
SAP18, ZNF144, ZNF254, ZNF277, ZNF363, SRY, and
SOX17, were up-regulated in early OA.
Although this is the first report of differential
expression of GADD45? in normal and OA human
cartilage, our results confirmed previous findings of
microarray studies of other genes, particularly matrix
genes (10,43). Genome-wide scans in OA patients have
identified polymorphisms in MTN3 (44), ASPN, which
inhibits cartilage anabolism by binding to transforming
growth factor ? (45), and other genes (46) that were also
up-regulated in late OA in our microarray analysis. Of
interest were genes with prominent roles in anabolism,
including CILP, AQP1, S100A4, IGFBP7, and ANXA2.
The increased expression of the Wnt-associated genes,
including DKK3, WIF1, and SFRP4, as well as
CKTSF1B1, POSTN, LTBP1, TGFB1I4, FGF2, FGF18,
Gli3, DLX4, and ACVR1B, suggests additional mecha-
nisms due to modified biologic activities in the calcified
cartilage and subchondral bone (47–49).
Other genes involved in regulating skeletal devel-
opment were differentially up-regulated in early OA.
We reported previously that RUNX2 is involved in the
induction of GADD45?, resulting in induction of
MMP13 in hypertrophic chondrocytes (16). RUNX2
(data not shown) and MMP13 were expressed in early
OA, although at higher levels in late OA, suggesting that
RUNX2 is not the only requirement for GADD45?
expression in adult articular cartilage. The presence of
GADD45? in osteophytes, outgrowths from the perios-
teum that undergo endochondral ossification, also sug-
gests its involvement in unsuccessful repair.
Regarding OA pathogenesis, few of the obvious
candidates were identified. Genes associated with in-
flammation, angiogenesis, cell adhesion, and matrix
2084 IJIRI ET AL
degradation were up-regulated in early OA. Many of
these genes were differentially expressed with relatively
low fold change and lower confidence bound, reflected
in the cluster analysis, suggesting that they may be
expressed at early stages of OA and throughout the
progression of the disease. For example, MMP10 has
been implicated as a collagenase activator during carti-
lage degradation in arthritis (50). In late-stage OA, TNF
superfamily members and heat-shock proteins may rep-
resent responses to stress and inflammation. Although
serine proteinase inhibitors, SERPINA5 and SERPINE1,
are expressed along with few proteinases other than
PRSS23 (serine protease 23), cathepsin K (a cysteine
proteinase), and MMP13, the level of catabolic activity
in late-stage OA cannot be predicted by these results.
Other candidates for further study for which novel roles
have been identified in cartilage include SDRG1 (51),
CRTAC1 (52), FAP (53), MIA/CD-RAP (54), and
In adult articular cartilage, chondrocytes are qui-
escent, and cell survival and anabolic activity are essen-
tial for maintenance of this avascular tissue. Thus,
chondrocyte apoptosis may have significant conse-
quences in the pathogenesis of OA, although its impor-
tance in the slowly progressing disease in humans is
controversial (12). Studies showing increased chondro-
cyte death in response to injurious compression support
the idea of an association between apoptosis and in-
creased stress (56). The presence of GADD45? in the
nuclei of nonmitotic chondrocytes and the differential
expression of several apoptosis-related genes in OA
cartilage are consistent with its described role in other
tissues as an antiapoptotic molecule. GADD45? is asso-
ciated with G2/M cell cycle arrest (57), and its interac-
tion with MTK1/MEKK4 promotes activation of JNK, as
well as p38 MAPK (58). However, GADD45? interac-
tion with MKK7, resulting in inhibition of TNF?-
induced JNK activity by engaging the ATP-binding site,
is responsible for the antiapoptotic activity of NF-?B
(36,37,59). Thus, inadequate levels of GADD45? and
sustained activation of JNK by mediators of inflamma-
tion may explain the proapoptotic response to NF-?B in
some cell types (60). Our finding that endogenous levels
of GADD45? protect against TNF?-induced cell death
suggests a role for this survival factor as a feedback
regulatory molecule during acute cell stress. Interest-
ingly, the tumor suppressor CYLD, which is induced by
TNF? and negatively regulates NF-?B signaling (61)
and adlican, which is involved in colon cancer progres-
sion (62), were both up-regulated in cartilage from
patients with late-stage OA.
Overall, our results indicate that GADD45? is a
key factor contributing to physiologic cartilage ho-
meostasis and to the imbalance in matrix remodeling in
OA cartilage, as well as to chondrocyte survival after cell
We are grateful to Dr. Thomas P. Sculco (Hospital for
Special Surgery) for providing cartilage samples.
Dr. Goldring had full access to all of the data in the study and
takes responsibility for the integrity of the data and the accuracy of the
Study design. Ijiri, Zerbini, Walsh, Mattingly, Aigner, Libermann,
Acquisition of data. Ijiri, Zerbini, Peng, Tsuchimochi, Otero, Drag-
omir, Bierbaum, Mattingly, Aigner.
Analysis and interpretation of data. Ijiri, Zerbini, Peng, Otu, Tsuchi-
mochi, Otero, Dragomir, Walsh, Mattingly, Komiya, Aigner, Goldring.
Manuscript preparation. Zerbini, Otero, Aigner, Libermann, Gold-
Statistical analysis. Otu, Otero, Dragomir.
Surgical retrieval. Mattingly, van Flandern.
1. Hunter DJ, Zhang Y, Niu J, Tu X, Amin S, Goggins J, et al.
Structural factors associated with malalignment in knee osteoar-
thritis: the Boston osteoarthritis knee study. J Rheumatol 2005;32:
2. Alexopoulos LG, Williams GM, Upton ML, Setton LA, Guilak F.
Osteoarthritic changes in the biphasic mechanical properties of
the chondrocyte pericellular matrix in articular cartilage. J Bio-
3. Pritzker KP, Gay S, Jimenez SA, Ostergaard K, Pelletier JP,
Revell PA, et al. Osteoarthritis cartilage histopathology: grading
and staging. Osteoarthritis Cartilage 2006;14:13–29.
4. Bau B, Gebhard PM, Haag J, Knorr T, Bartnik E, Aigner T.
Relative messenger RNA expression profiling of collagenases and
aggrecanases in human articular chondrocytes in vivo and in vitro.
Arthritis Rheum 2002;46:2648–57.
5. Hermansson M, Sawaji Y, Bolton M, Alexander S, Wallace A,
Begum S, et al. Proteomic analysis of articular cartilage shows
increased type II collagen synthesis in osteoarthritis and expres-
sion of inhibin ?A (activin A), a regulatory molecule for chondro-
cytes. J Biol Chem 2004;279:43514–21.
6. Fukui N, Zhu Y, Maloney WJ, Clohisy J, Sandell LJ. Stimulation
of BMP-2 expression by pro-inflammatory cytokines IL-1 and
TNF-? in normal and osteoarthritic chondrocytes. J Bone Joint
Surg Am 2003;85-A Suppl 3:59–66.
7. Nakase T, Miyaji T, Tomita T, Kaneko M, Kuriyama K, Myoui A,
et al. Localization of bone morphogenetic protein-2 in human
osteoarthritic cartilage and osteophyte. Osteoarthritis Cartilage
8. Aigner T, Vornehm SI, Zeiler G, Dudhia J, von der Mark K,
Bayliss MT. Suppression of cartilage matrix gene expression in
upper zone chondrocytes of osteoarthritic cartilage. Arthritis
9. Aigner T, Zien A, Gehrsitz A, Gebhard PM, McKenna L. Ana-
bolic and catabolic gene expression pattern analysis in normal
GADD45? IN NORMAL AND OA CARTILAGE2085
versus osteoarthritic cartilage using complementary DNA–array
technology. Arthritis Rheum 2001;44:2777–89.
10. Aigner T, Fundel K, Saas J, Gebhard PM, Haag J, Weiss T, et al.
Large-scale gene expression profiling reveals major pathogenetic
pathways of cartilage degeneration in osteoarthritis. Arthritis
11. Sandell LJ, Aigner T. Articular cartilage and changes in arthritis:
cell biology of osteoarthritis. Arthritis Res 2001;3:107–13.
12. Aigner T, Kim HA, Roach HI. Apoptosis in osteoarthritis. Rheum
Dis Clin North Am 2004;30:639–53, xi.
13. Abdollahi A, Lord KA, Hoffman-Liebermann B, Liebermann DA.
Sequence and expression of a cDNA encoding MyD118: a novel
myeloid differentiation primary response gene induced by multiple
cytokines. Oncogene 1991;6:165–7.
14. Selvakumaran M, Lin HK, Sjin RT, Reed JC, Liebermann DA,
Hoffman B. The novel primary response gene MyD118 and the
proto-oncogenes myb, myc, and bcl-2 modulate transforming
growth factor ?1-induced apoptosis of myeloid leukemia cells. Mol
Cell Biol 1994;14:2352–60.
15. Sironen RK, Karjalainen HM, Elo MA, Kaarniranta K, Torronen
K, Takigawa M, et al. cDNA array reveals mechanosensitive genes
in chondrocytic cells under hydrostatic pressure. Biochim Biophys
16. Ijiri K, Zerbini LF, Peng H, Correa RG, Lu B, Walsh N, et al. A
novel role for GADD45? as a mediator of MMP-13 gene expres-
sion during chondrocyte terminal differentiation. J Biol Chem
17. McKenna LA, Gehrsitz A, Soder S, Eger W, Kirchner T, Aigner T.
Effective isolation of high-quality total RNA from human adult
articular cartilage. Anal Biochem 2000;286:80–5.
18. Bau B, Haag J, Schmid E, Kaiser M, Gebhard PM, Aigner T. Bone
morphogenetic protein-mediating receptor-associated Smads as
well as common Smad are expressed in human articular chondro-
cytes but not up-regulated or down-regulated in osteoarthritic
cartilage. J Bone Miner Res 2002;17:2141–50.
19. Spentzos D, Levine DA, Ramoni MF, Joseph M, Gu X, Boyd J, et
al. Gene expression signature with independent prognostic signif-
icance in epithelial ovarian cancer [published erratum appears in
J Clin Oncol 2005;23:248]. J Clin Oncol 2004;22:4700–10.
20. Altman R, Alarcon G, Appelrouth D, Bloch D, Borenstein D,
Brandt K, et al. The American College of Rheumatology criteria
for the classification and reporting of osteoarthritis of the hip.
Arthritis Rheum 1991;34:505–14.
21. Li C, Wong WH. Model-based analysis of oligonucleotide arrays:
expression index computation and outlier detection. Proc Natl
Acad Sci U S A 2001;98:31–6.
22. Barash Y, Dehan E, Krupsky M, Franklin W, Geraci M, Friedman
N, et al. Comparative analysis of algorithms for signal quantitation
from oligonucleotide microarrays. Bioinformatics 2004;20:839–46.
23. Li C, Wong WH. Model-based analysis of oligonucleotide arrays:
model validation, design issues and standard error application.
Genome Biol 2001;2:RESEARCH0032.
24. Irizarry RA, Bolstad BM, Collin F, Cope LM, Hobbs B, Speed TP.
Summaries of Affymetrix GeneChip probe level data. Nucleic
Acids Res 2003;31:e15.
25. Shedden K, Chen W, Kuick R, Ghosh D, Macdonald J, Cho KR,
et al. Comparison of seven methods for producing Affymetrix
expression scores based on False Discovery Rates in disease
profiling data. BMC Bioinformatics 2005;6:26.
26. Yuen T, Wurmbach E, Pfeffer RL, Ebersole BJ, Sealfon SC.
Accuracy and calibration of commercial oligonucleotide and cus-
tom cDNA microarrays. Nucleic Acids Res 2002;30:e48.
27. Ramalho-Santos M, Yoon S, Matsuzaki Y, Mulligan RC, Melton
DA. “Stemness”: transcriptional profiling of embryonic and adult
stem cells. Science 2002;298:597–600.
28. Zerbini LF, Wang Y, Czibere A, Correa RG, Cho JY, Ijiri K, et al.
NF-?B-mediated repression of growth arrest- and DNA-damage-
inducible proteins 45? and ? is essential for cancer cell survival.
Proc Natl Acad Sci U S A 2004;101:13618–23.
29. Mankin HJ, Dorfman H, Lippiello L, Zarins A. Biochemical and
metabolic abnormalities in articular cartilage from osteo-arthritic
human hips. II. Correlation of morphology with biochemical and
metabolic data. J Bone Joint Surg Am 1971;53:523–37.
30. Qiu W, David D, Zhou B, Chu PG, Zhang B, Wu M, et al.
Down-regulation of growth arrest DNA damage-inducible gene
45? expression is associated with human hepatocellular carci-
noma. Am J Pathol 2003;162:1961–74.
31. Goldring MB, Birkhead JR, Suen LF, Yamin R, Mizuno S,
Glowacki J, et al. Interleukin-1?-modulated gene expression in
immortalized human chondrocytes. J Clin Invest 1994;94:2307–16.
32. Tan L, Peng H, Osaki M, Choy BK, Auron PE, Sandell LJ, et al.
Egr-1 mediates transcriptional repression of COL2A1 promoter
activity by interleukin-1?. J Biol Chem 2003;278:17688–700.
33. Goldring MB, Fukuo K, Birkhead JR, Dudek E, Sandell LJ.
Transcriptional suppression by interleukin-1 and interferon-? of
type II collagen gene expression in human chondrocytes. J Cell
34. Zerbini LF, Wang Y, Cho JY, Libermann TA. Constitutive
activation of nuclear factor ?B p50/p65 and Fra-1 and JunD is
essential for deregulated interleukin 6 expression in prostate
cancer. Cancer Res 2003;63:2206–15.
35. Jin R, De Smaele E, Zazzeroni F, Nguyen DU, Papa S, Jones J, et
al. Regulation of the gadd45? promoter by NF-?B. DNA Cell Biol
36. De Smaele E, Zazzeroni F, Papa S, Nguyen DU, Jin R, Jones J, et
al. Induction of gadd45? by NF-?B downregulates pro-apoptotic
JNK signalling. Nature 2001;414:308–13.
37. Papa S, Zazzeroni F, Bubici C, Jayawardena S, Alvarez K,
Matsuda S, et al. Gadd45? mediates the NF-?B suppression of
JNK signalling by targeting MKK7/JNKK2. Nat Cell Biol 2004;6:
38. Gebhard PM, Gehrsitz A, Bau B, Soder S, Eger W, Aigner T.
Quantification of expression levels of cellular differentiation
markers does not support a general shift in the cellular phenotype
of osteoarthritic chondrocytes. J Orthop Res 2003;21:96–101.
39. Ronziere MC, Ricard-Blum S, Tiollier J, Hartmann DJ, Garrone
R, Herbage D. Comparative analysis of collagens solubilized from
human foetal, and normal and osteoarthritic adult articular carti-
lage, with emphasis on type VI collagen. Biochim Biophys Acta
40. Lorenzo P, Bayliss MT, Heinegard D. Altered patterns and
synthesis of extracellular matrix macromolecules in early osteoar-
thritis. Matrix Biol 2004;23:381–91.
41. Tchetina EV, Squires G, Poole AR. Increased type II collagen
degradation and very early focal cartilage degeneration is associ-
ated with upregulation of chondrocyte differentiation related
genes in early human articular cartilage lesions. J Rheumatol
42. Aigner T, Gebhard PM, Schmid E, Bau B, Harley V, Poschl E.
SOX9 expression does not correlate with type II collagen expres-
sion in adult articular chondrocytes. Matrix Biol 2003;22:363–72.
43. Sato T, Konomi K, Yamasaki S, Aratani S, Tsuchimochi K,
Yokouchi M, et al. Comparative analysis of gene expression
profiles in intact and damaged regions of human osteoarthritic
cartilage. Arthritis Rheum 2006;54:808–17.
44. Stefansson SE, Jonsson H, Ingvarsson T, Manolescu I, Jonsson
HH, Olafsdottir G, et al. Genomewide scan for hand osteoarthri-
tis: a novel mutation in matrilin-3. Am J Hum Genet 2003;72:
45. Kizawa H, Kou I, Iida A, Sudo A, Miyamoto Y, Fukuda A, et al.
An aspartic acid repeat polymorphism in asporin inhibits chondro-
genesis and increases susceptibility to osteoarthritis. Nat Genet
46. Loughlin J. Polymorphism in signal transduction is a major route
2086 IJIRI ET AL
through which osteoarthritis susceptibility is acting. Curr Opin Download full-text
47. Bailey AJ, Mansell JP, Sims TJ, Banse X. Biochemical and
mechanical properties of subchondral bone in osteoarthritis. Bio-
48. Loughlin J, Dowling B, Chapman K, Marcelline L, Mustafa Z,
Southam L, et al. Functional variants within the secreted frizzled-
related protein 3 gene are associated with hip osteoarthritis in
females. Proc Natl Acad Sci U S A 2004;101:9757–62.
49. Lane NE, Lian K, Nevitt MC, Zmuda JM, Lui L, Li J, et al.
Frizzled-related protein variants are risk factors for hip osteoar-
thritis. Arthritis Rheum 2006;54:1246–54.
50. Barksby HE, Milner JM, Patterson AM, Peake NJ, Hui W,
Robson T, et al. Matrix metalloproteinase 10 promotion of
collagenolysis via procollagenase activation: implications for car-
tilage degradation in arthritis. Arthritis Rheum 2006;54:3244–53.
51. Ochi K, Derfoul A, Tuan RS. A predominantly articular cartilage-
associated gene, SCRG1, is induced by glucocorticoid and stimulates
chondrogenesis in vitro. Osteoarthritis Cartilage 2006;14:30–8.
52. Steck E, Braun J, Pelttari K, Kadel S, Kalbacher H, Richter W.
Chondrocyte secreted CRTAC1: a glycosylated extracellular ma-
trix molecule of human articular cartilage. Matrix Biol 2007;26:
53. Milner JM, Kevorkian L, Young DA, Jones D, Wait R, Donell ST,
et al. Fibroblast activation protein ? is expressed by chondrocytes
following a pro-inflammatory stimulus and is elevated in osteoar-
thritis. Arthritis Res Ther 2006;8:R23.
54. Tscheudschilsuren G, Bosserhoff AK, Schlegel J, Vollmer D,
Anton A, Alt V, et al. Regulation of mesenchymal stem cell and
chondrocyte differentiation by MIA. Exp Cell Res 2006;312:63–72.
55. Hayes AJ, Dowthwaite GP, Webster SV, Archer CW. The distri-
bution of Notch receptors and their ligands during articular
cartilage development. J Anat 2003;202:495–502.
56. Kurz B, Lemke AK, Fay J, Pufe T, Grodzinsky AJ, Schunke M.
Pathomechanisms of cartilage destruction by mechanical injury.
Ann Anat 2005;187:473–85.
57. Vairapandi M, Balliet AG, Hoffman B, Liebermann DA.
GADD45b and GADD45g are cdc2/cyclinB1 kinase inhibitors
with a role in S and G2/M cell cycle checkpoints induced by
genotoxic stress. J Cell Physiol 2002;192:327–38.
58. Takekawa M, Saito H. A family of stress-inducible GADD45-like
proteins mediate activation of the stress-responsive MTK1/
MEKK4 MAPKKK. Cell 1998;95:521–30.
59. Papa S, Monti SM, Vitale RM, Bubici C, Jayawardena S, Alvarez
K, et al. Insights into the structural basis of the GADD45?-
mediated inactivation of the JNK kinase, MKK7/JNKK2. J Biol
60. Larsen CM, Dossing MG, Papa S, Franzoso G, Billestrup N,
Mandrup-Poulsen T. Growth arrest- and DNA-damage-inducible
45? gene inhibits c-Jun N-terminal kinase and extracellular signal-
regulated kinase and decreases IL-1?-induced apoptosis in insulin-
producing INS-1E cells. Diabetologia 2006;49:980–9.
61. Jono H, Lim JH, Chen LF, Xu H, Trompouki E, Pan ZK, et al.
NF-?B is essential for induction of CYLD, the negative regulator
of NF-?B: evidence for a novel inducible autoregulatory feedback
pathway. J Biol Chem 2004;279:36171–4.
62. Zou TT, Selaru FM, Xu Y, Shustova V, Yin J, Mori Y, et al.
Application of cDNA microarrays to generate a molecular taxon-
omy capable of distinguishing between colon cancer and normal
colon. Oncogene 2002;21:4855–62.
GADD45? IN NORMAL AND OA CARTILAGE2087