ARTHRITIS & RHEUMATISM
Vol. 64, No. 4, April 2012, pp 1035–1045
© 2012, American College of Rheumatology
Acute-Phase Serum Amyloid A Regulates
Tumor Necrosis Factor ? and Matrix Turnover and Predicts
Disease Progression in Patients With Inflammatory Arthritis
Before and After Biologic Therapy
Mary Connolly,1Ronan H. Mullan,1Jennifer McCormick,1Clare Matthews,1Owen Sullivan,1
Aisling Kennedy,1Oliver FitzGerald,1A. Robin Poole,2Barry Bresnihan,†
Douglas J. Veale,1and Ursula Fearon1
Objective. To investigate the relationship between
acute-phase serum amyloid A (A-SAA) and joint de-
struction in inflammatory arthritis.
Methods. Serum A-SAA and C-reactive protein
(CRP) levels, the erythrocyte sedimentation rate (ESR),
and levels of matrix metalloproteinase 1 (MMP-1),
MMP-2, MMP-3, MMP-9, MMP-13, tissue inhibitor of
metalloproteinases 1 (TIMP-1), vascular endothelial
growth factor (VEGF), and type I and type II collagen–
generated biomarkers C2C and C1,2C were measured at
0–3 months in patients with inflammatory arthritis
commencing anti–tumor necrosis factor ? (anti-TNF?)
therapy and were correlated with 1-year radiographic
progression. The effects of A-SAA on MMP/TIMP ex-
pression on RA fibroblast-like synoviocytes (FLS), pri-
mary human chondrocytes, and RA/psoriatic arthritis
synovial explant cultures were assessed using real-time
polymerase chain reaction, enzyme-linked immunosor-
bent assay, antibody protein arrays, and gelatin zymog-
Results. Serum A-SAA levels were significantly
(P < 0.05) correlated with MMP-3, the MMP-3:TIMP-1
ratio, C1,2C, C2C, and VEGF. The baseline A-SAA level
but not the ESR or the CRP level correlated with the
28-joint swollen joint count and was independently
associated with 1-year radiographic progression (P ?
0.038). A-SAA increased MMP-1, MMP-3, MMP-13,
and MMP/TIMP expression in RA FLS and synovial
explants (P < 0.05). In chondrocytes, A-SAA induced
MMP-1, MMP-3, and MMP-13 messenger RNA and
protein expression (all P < 0.01), resulting in a signif-
icant shift in MMP:TIMP ratios (P < 0.05). Gelatin
zymography revealed that A-SAA induced MMP-2 and
MMP-9 activity. Blockade of the A-SAA receptor SR-B1
(A-SAA receptor scavenger receptor-class B type 1)
inhibited MMP-3, MMP-2, and MMP-9 expression in
synovial explant cultures ex vivo. Importantly, we dem-
onstrated that A-SAA has the ability to induce TNF?
expression in RA synovial explant cultures (P < 0.05).
Conclusion. A-SAA may be involved in joint de-
struction though MMP induction and collagen cleavage
in vivo. The ability of A-SAA to regulate TNF? suggests
that A-SAA signaling pathways may provide new thera-
peutic strategies for the treatment of inflammatory
Rheumatoid arthritis (RA) and psoriatic arthritis
(PsA) are inflammatory autoimmune disorders charac-
terized by joint destruction and disability (1). Synovial
inflammation, which occurs through angiogenesis, leu-
Supported by grants from the Science Foundation Ireland and
the Health Research Board of Ireland.
1Mary Connolly, PhD, Ronan H. Mullan, MBChB, PhD,
Jennifer McCormick, MPhil, Clare Matthews, MD, Owen Sullivan,
BSc, Aisling Kennedy, PhD, Oliver FitzGerald, MD, FRCPI, FRCP,
Douglas J. Veale, MD, Ursula Fearon, PhD: St. Vincent’s University
Hospital, Dublin Academic Medical Centre, and The Conway Institute
of Biomolecular and Biomedical Research, Dublin, Ireland;2A. Robin
Poole, PhD, DSc: Shriners Hospitals for Children and McGill Univer-
sity, Montreal, Quebec, Canada.
†Dr. Bresnihan is deceased.
Drs. Connolly and Mullan contributed equally to this work.
Dr. Poole has received consulting fees from Ibex and Merck
Serono (more than $10,000 each). Dr. Veale has received consulting
fees, speaking fees, and/or honoraria from Pfizer, Merck Sharp &
Dohme, Roche, Abbott, and AstraZeneca.
Address correspondence to Ursula Fearon, PhD, Department
of Rheumatology, St. Vincent’s University Hospital, Dublin Academic
Medical Centre, Elm Park, Dublin 4, Ireland. E-mail: ursula.fearon@
Submitted for publication August 30, 2010; accepted in
revised form October 25, 2011.
kocyte recruitment, and proliferation of fibroblast-like
synoviocytes (FLS), leads to the formation of a hyper-
plastic synovial pannus with locally invasive properties
(2,3). Disease progression, as measured by quantifying
the changes in joint space narrowing and erosions that
are visible on serial plain radiographs (4,5), is associated
with long-term functional disability (4). This method of
determining disease progression is the accepted approach
to monitoring the long-term response to therapy (6).
The dysregulation of cartilage metabolism in
inflammatory arthritis is measurable when cartilage me-
tabolism byproducts are released into body fluids (6–8).
Collagen biomarker and matrix metalloproteinase
(MMP) levels (9–11) predict radiographic progression in
RA (12,13). Type II collagen, the main collagen of
articular cartilage, is excessively degraded in RA, PsA,
and osteoarthritis (3,14–16). Increased levels of neo-
epitopes formed by collagenase cleavage of type II
collagen (C2C), a neoepitope formed by collagenase
cleavage of type II collagen as well as type I collagen
(C1,2C), and type II procollagen (CPII) occur in RA
(13). We previously showed that early changes in serum
C2C, C1,2C, and CPII levels following treatment with
biologic therapy predicted 12-month radiographic pro-
gression in a cohort of patients with inflammatory
During active arthritis, synovial inflammation is
mirrored by a systemic acute-phase response in which
marked biosynthetic changes in the liver alter plasma
protein concentrations (17). Previous studies demon-
strated that acute-phase serum amyloid A (A-SAA) is a
marker of disease activity that is superior to C-reactive
protein (CRP) in RA (18). We previously demonstrated
A-SAA production in RA synovial tissue, and A-SAA
induced angiogenesis, leukocyte recruitment, and
chemokine and MMP expression in RA (19). SAA exists
as constitutive and acute-phase isoforms, which have
distinct properties with respect to the development of
amyloidosis (20–22). A study by Bjorkman et al demon-
strated that A-SAA isolated from RA plasma had little
effect on neutrophil shedding compared with the effect
of recombinant A-SAA (23). However, previous work by
those investigators showed that A-SAA promoted neu-
trophil survival (24,25). Furthermore, site-specific ex-
pression of A-SAA isoforms in the synovial fluid but not
the serum of patients with inflammatory arthritis sug-
gests differential functions between systemic and locally
expressed forms (26). Our group previously demon-
strated spontaneous A-SAA secretion from RA synovial
explants and inhibition of chemokine production by
blockade of the A-SAA receptors scavenger receptor-
class B type 1 (SR-B1) and FRPL-1, suggesting that the
proinflammatory effects of endogenously produced
A-SAA are blocked (27).
In the present study, we investigated the relation-
ships between serum A-SAA and CRP levels and the
erythrocyte sedimentation rate (ESR) and cartilage de-
gradation biomarkers and 12-month radiographic out-
come in patients undergoing biologic therapy for inflam-
matory arthritis. Using ex vivo synovial tissue explants
and primary RA FLS and chondrocyte cultures, we
examined the effect of A-SAA on matrix turnover and
tumor necrosis factor ? (TNF?) regulation.
PATIENTS AND METHODS
Patient recruitment. The study group comprised 62
patients who were recruited from rheumatology outpatient
clinics at St. Vincent’s University Hospital and were followed
up prospectively for 1 year, as previously described (16).
Forty-five patients fulfilled the American College of Rheuma-
tology revised criteria for the classification of RA (28), and 17
patients fulfilled the classification criteria for PsA (29). The
baseline clinical and demographic characteristics of the pa-
tients are shown in Table 1. Patients with no prior exposure to
biologic therapy (30) but with a 28-joint Disease Activity Score
(DAS28) of ?3.2 despite synthetic disease-modifying antirheu-
matic drug (DMARD) therapy were offered treatment with
biologic agents. Changes in conventional therapy were permit-
ted during biologic therapy at the discretion of the patient’s
treating rheumatologist. Following institutional ethics ap-
proval and provision of fully informed written consent, patients
underwent clinical evaluation and phlebotomy before and 1, 3,
6, 9, and 12 months after the initiation of biologic therapy (16).
Paired serum samples were obtained and stored at
?80°C until used for biomarker analyses. Fifty-eight patients
received TNF-targeted therapies, as follows: 44 patients (36
with RA and 8 with PsA) received adalimumab, 11 patients
(3 with RA and 8 with PsA) received infliximab, and 3 patients
(2 with RA and 1 with PsA) received etanercept. Four pa-
tients with RA received treatment with the interleukin-1
receptor antagonist (IL-1Ra) anakinra. DMARDs were pre-
scribed to 42 patients (68%) and included methotrexate (n ?
22), leflunomide (n ? 2), sulfasalazine (n ? 2), azathioprine
(n ? 6), and hydroxychloroquine (n ? 2); 20 patients received
biologic monotherapy. Seven patients did not complete a full
year of biologic therapy (6 patients did not complete a full year
of therapy because of a lack of efficacy, and therapy was
temporarily withheld in 1 patient because of an infection).
Three patients in whom treatment with the initial TNF inhib-
itor failed were switched to a second TNF inhibitor.
Radiographic evaluation. Anteroposterior radiographs
of the hands and feet were obtained at baseline and 1 year. The
radiographs were analyzed in a blinded manner by 2 indepen-
dent observers (RM and CM) and were scored according to
the modified Sharp/van der Heijde scoring (SHS) method (5).
Radiographic progression was defined as a binomial variable
on the basis of the smallest detectable change calculated from
the scoring of radiographs between the 2 observers (SHS 1.02)
1036 CONNOLLY ET AL
(31). Seven patients (3 with RA and 4 with PsA) did not have
radiographs obtained at 1 year and were excluded from
followup radiographic analysis.
A-SAA enzyme-linked immunosorbent assay (ELISA).
A-SAA ELISAs (BioSource International) were performed
on serum samples according to the manufacturer’s instruc-
tions. The range of the assay was 4–300 ng/ml. Absorbance was
analyzed at 450 nm.
Assays for cartilage matrix molecules and markers of
inflammation. Serum biomarkers were measured using 2-step
competitive immunoassays (Ibex). The interassay reproducibil-
ity of measurements of C2C, CPII, and C1,2C concentrations
in paired sera was 9.7%, 6.4%, and 10%, respectively (13,32).
The ESR was determined using the Westergren technique.
CRP levels were measured by nephelometry.
VEGF ELISA. VEGF was measured using a specific
Quantikine ELISA according to the manufacturer’s instruc-
tions (R&D Systems). ELISA standards ranged from 31.2
mg/ml to 2,000 pg/ml. Absorbance was analyzed at 450 nm.
Isolation and culture of FLS. Arthroscopy of actively
inflamed knee joints was performed under local anesthesia,
using a Storz 2.7-mm telescope, as previously described (33).
Synovial tissue biopsy specimens were obtained from patients
with RA, using grasping forceps under direct visualization.
Selected biopsy specimens were digested with 1 mg/ml colla-
genase 1 (Worthington) in RPMI 1640 (Gibco BRL) for 4
hours at 37°C in humidified air in an atmosphere of 5% CO2.
Dissociated cells were grown to confluence in RPMI 1640,
10% fetal calf serum (FCS; Gibco BRL), 10 ml of 1 mmole/
liter HEPES (Gibco BRL), penicillin (100 units/ml; Bio-
sciences), streptomycin (100 units/ml; Biosciences), and fungi-
zone (0.25 ?g/ml; Biosciences) before passaging. Cells were
used between passages 4 and 8. RA FLS were stimulated with
A-SAA (0.1–50 ?g/ml) (PeproTech), TNF? (10 ng/ml), or
IL-1? (10 ng/ml) in serum-free RPMI 1640 for 24 hours, and
supernatants were harvested for ELISA. Endotoxin levels of
?1 endotoxin unit in A-SAA preparations were confirmed by
Limulus assay (BioWhittaker). Supernatants were harvested
and assayed using a multiplex tissue culture kit (Meso Scale
RA and PsA synovial explant cultures. Ex vivo synovial
explants from patients with RA and patients with PsA were cut
into 1-mm cubes and cultured in 96-well plates (Falcon), with
each well containing 300 ?l of serum-free RPMI 1640, strep-
tomycin (100 units/ml), and penicillin (100 units/ml) for 24
hours at 37°C in an atmosphere of 5% CO2(37). Explants were
stimulated with A-SAA (10–50 ?g/ml) in serum-free RPMI
1640 for 24 hours.
To investigate the effect of blocking A-SAA signaling
on MMP production, synovial explants were cultured in the
presence of an antibody against the A-SAA receptor SR-B1.
Synovial explant cultures (n ? 4) were sectioned as described
above and cultured immediately following arthroscopy to
maintain maximal active endogenous production of proinflam-
matory mediators such A-SAA and MMPs. Explants were
cultured in full RPMI 1640 in the presence of mouse anti–
SR-B1 (10 ?g/ml; BD Biosciences) and isotype-matched
mouse IgG control antibody (10 ?g/ml; R&D Systems) for 24
hours. Supernatants were harvested from explant cultures and
assayed for MMP and cytokine production.
TNF? assay. TNF? was assayed using a multiplex
tissue culture kit (Meso Scale Discovery) according to the
manufacturer’s instructions. Absorbance was measured using a
SECTOR Imager 2400 system (Meso Scale Discovery). TNF?
standards ranged from 2.4 pg/ml to 10,000 pg/ml.
Primary and SW1353 chondrocyte cultures. Normal
human articular cartilage was obtained as previously described,
and chondrocytes were obtained by sequential proteolysis
(10,34). Cells were cultured in Dulbecco’s modified Eagle’s
medium (DMEM) with 10% FCS, 10 ml of 1 mmole/liter
HEPES (Gibco BRL), penicillin (100 units/ml), streptomycin
(100 units/ml), and fungizone (0.25 ?g/ml) (Biosciences) for
experimentation between passages 4 and 8. Human chondro-
sarcoma cells (SW1353) purchased from American Type Cul-
ture Collection HTB-94 were cultured in DMEM–Ham’s-F12
(Invitrogen) supplemented with 1% glutamine, 1% nonessen-
tial amino acids (Invitrogen), penicillin (100 units/ml), strep-
tomycin (100 units/ml), fungizone (0.25 ?g/ml), and 10% FCS
until 80% confluency was achieved. Cells were stimulated with
A-SAA (0.1–50 ?g), TNF? (10 ng/ml), or IL-1? (10 ng/ml) in
serum-free RPMI 1640 for 3–24 hours.
MMP protein antibody arrays. MMP expression pro-
files were examined using antibody arrays (RayBiotech). Array
membranes were blocked for 1 hour at room temperature prior
to incubation with 1 ml of serum-free conditioned media from
cultured cells for 2 hours. Following washing, 1 ml of biotin-
conjugated antibody was added to the membranes for 2 hours.
Following repeated washes, 2 ml of horseradish peroxidase–
conjugated streptavidin was added to each membrane and
incubated for 2 hours at room temperature prior to incubation
with detection buffer for 2 minutes. Membranes were exposed
to radiography film for 40 seconds and developed using film
developer. MMP expression induced by A-SAA (50 ?g/ml) and
TNF? (10 ng/ml) in RA FLS and chondrocytes was identified
using an antibody array map (available from the corresponding
RNA extraction from SW1353 chondrocytes. Cells
were stimulated with A-SAA (10–50 ?g/ml) in serum-free
RPMI 1640 for 3–24 hours prior to RNA extraction from
SW1353 chondrocytes, following homogenization using a
hand-held homogenizer (Glen Mills), using an RNeasy Mini
Kit according to the manufacturer’s protocol (Qiagen). The
RNA was stored at ?80°C prior to complementary DNA
Quantitative and qualitative determination of total
RNA. RNA quantity was assessed by spectrophotometry. The
integrity of RNA samples was assessed using a bioanalyzer
(Agilent). Samples with a 260:280-nm ratio of ?1.8 and an
RNA integrity number between 7 and 10 were used in subse-
quent experiments. Isolated RNA was stored at ?80°C.
Reverse transcription using extracted RNA. Total
RNA isolated from cultured SW1353 chondrocytes was reverse
transcribed into cDNA. SuperScript II RNase Reverse Tran-
scriptase (Gibco BRL) was used to synthesize first-strand
cDNA. This enzyme allows higher yields of cDNA and more
full-length products compared with other reverse transcrip-
tases. One microgram of total RNA was added to a 25-?l
reaction volume containing 200 units of SuperScript II in
reverse transcriptase buffer, 100 mM dithiothreitol (supplied
with the reverse transcriptase enzyme), 40 units of RNasin
Ribonuclease Inhibitor (Promega), 1.25 mM each dATP,
ROLE OF A-SAA IN INFLAMMATORY ARTHRITIS1037
dCTP, dGTP, and dTTP (Promega), RNase-free water, and
500 ng of oligo(dT) (Promega). Reverse transcription was
carried out at 42°C for 50 minutes. The reaction was termi-
nated by incubation at 95°C for 15 minutes. Complementary
DNA was used immediately or was frozen at ?20°C for future
Polymerase chain reactions (PCRs) for GAPDH,
MMP-1, and MMP-13. PCR primers and probes for MMP-1
and MMP-13 were designed by Applied Biosystems (Assays-
on-Demand). GAPDH and ?-actin were used as endogenous
controls. Amplification reactions contained 1 ?l of cDNA, 12.5
?l of TaqMan Fast Universal Master Mix (2?) (Applied
Biosystems), and 1.25 ?l of primer and probe mix and were
brought to a total volume of 25 ?l by the addition of RNase-
free water. All reactions/negative controls were performed in
triplicate using 96-well plates on a LightCycler 480 system
(Roche Diagnostics). Thermal cycling conditions were as rec-
ommended by the manufacturer (Applied Biosystems). Rela-
tive changes in gene expression were determined using the
MMP assays. Pro–MMP-1, pro–MMP-3, pro–MMP-
13, tissue inhibitor of metalloproteinases 1 (TIMP-1), and
TIMP-4 were measured using Quantikine ELISAs according
to the manufacturer’s instructions (R&D Systems). ELISA
standards ranged from 0.156 ng/ml to 10 ng/ml for MMP-1 and
MMP-3, from 78 pg/ml to 5,000 pg/ml for MMP-13, and from
0.156 ng/ml to 5 ng/ml for TIMP-1 and TIMP-4.
Zymography. Cultured supernatants from RA and PsA
FLS, primary human chondrocytes, and RA synovial explant
cultures incubated with A-SAA (0.1–50 ?g/ml), TNF? (10 ng/
ml), or IL-1? (10 mg/ml) for 24 hours were separated under
nonreducing conditions by sodium dodecyl sulfate (SDS)–
polyacrylamide gel electrophoresis in 10% polyacrylamide gels
copolymerized with 1% gelatin. Gels were washed twice for 25
minutes in 2.5% Triton X-100 to remove SDS, rinsed for 25
minutes in distilled H2O, and incubated overnight in 50 mM
Tris, 50 mM NaCl, 10 mM CaCl2, pH 7.5, at 37°C. Gels were
then rinsed for 5 minutes in distilled H2O before the addition
of Coomassie brilliant blue stain (30% isopropanol, 10% acetic
acid, 0.25 mg/ml Coomassie brilliant blue R) for 10 minutes.
Gels were visualized using the UVP BioImaging AutoChemi
Statistical analysis. To detect a 20% difference with
80% power at a 5% significance level allowing for a withdrawal
rate of ?20%, a minimum sample size of 40 patients (20
responders and 20 nonresponders) was calculated. Nonpara-
metric data were analyzed using Wilcoxon’s signed rank test.
Radiographic progression was examined as a categorical vari-
able (?1.5 SHS units/year [no progression] versus ?1.5 SHS
units/year [progression]) using one-way analysis of variance.
Comparisons between categorical data sets were performed
using chi-square tests. Gene expression analysis was performed
using Student’s paired t-tests. All statistical analyses were
performed using SPSS version 12 for Windows.
Clinical outcomes. The baseline clinical and de-
mographic characteristics of the patient population are
shown in Table 1. No significant differences in disease
duration, the CRP level, the Health Assessment Ques-
tionnaire (HAQ) score, the ESR, or the physician’s
global assessment score were demonstrated between
patients with RA and those with PsA. The baseline
DAS28 was significantly higher in patients with RA
compared with patients with PsA (mean ? SD 5.7 ? 1.2
versus 4.7 ? 1.5; P ? 0.05). A DAS28-based European
League Against Rheumatism (EULAR) response was
demonstrated in 76% of patients at 3 months, with no
Baseline characteristics of the study patients*
(n ? 62)
(n ? 45)
(n ? 17)
Age, mean (range) years
Female sex, %
Disease duration, mean (range) years
No. of previous DMARDs, mean (range)
DMARD use at baseline, %
Baseline corticosteroid use, %
Physician’s global assessment, 0–100 mm VAS
Modified HAQ, units
Radiographic damage, %
50 ? 23
30 ? 24
5.4 ? 1.4
33 ? 36
1.4 ? 0.7
123 ? 61
366 ? 95
1,014 ? 455
49 ? 22
30 ? 30
5.7 ? 1.2†
30 ? 31
1.5 ? 0.7
129 ? 68
381 ? 96
1,029 ? 459
51 ? 25
33 ? 29
4.7 ? 1.5
42 ? 48
1.2 ? 0.9
107 ? 30
331 ? 82
975 ? 457
* Except where indicated otherwise, values are the mean ? SD. RA ? rheumatoid arthritis; PsA ?
psoriatic arthritis; DMARDs ? disease-modifying antirheumatic drugs; VAS ? visual analog scale;
ESR ? erythrocyte sedimentation rate; DAS28 ? Disease Activity Score in 28 joints; CRP ? C-reactive
protein; HAQ ? Health Assessment Questionnaire; CPII ? type II procollagen.
† P ? 0.05 versus PsA.
1038CONNOLLY ET AL
significant difference in DAS28-based EULAR response
rates between patients with RA (74%) and patients with
PsA (81%). No difference in the expression of cartilage
degradation/turnover markers (C2C, C1,2C, and CPII)
between patients with RA and patients with PsA was
observed at baseline (Table 1) or at 3 months.
Forty-seven patients (85%) had baseline radio-
graphic damage, with a median baseline SHS of 43
(interquartile range [IQR] 16.7–93.1) and a baseline
progression rate per year since diagnosis of 4.82 (IQR
2.8–10.9). No significant difference was observed in the
percentage of patients with baseline radiographic dam-
age (86% of patients with RA and 85% of patients with
PsA). Radiographic progression occurred in 24 (44%) of
55 patients with available 1-year followup radiographs,
all of whom had radiographic damage at baseline, with a
median increase in the SHS of 2.75 (IQR 2.0–6.125);
this is consistent with previous studies showing that
inhibition of radiographic progression is reduced in
patients with established versus early disease (35). No
significant difference was observed in 1-year radio-
graphic progression between patients with RA (46%)
and patients with PsA (43%).
Baseline A-SAA associations. Baseline A-SAA
levels were significantly correlated with the baseline
28-joint swollen joint counts (SJCs) (r ? 0.26, P ?
0.048); this association was strengthened by the exclu-
sion from analysis of patients with an SJC in the lowest
quartile (?5) (r ? 0.4, P ? 0.009). No associations
between the SJC and the CRP level or the ESR were
observed. Elevations of the serum A-SAA level (mean ?
SD 420 ? 96 ?g/ml [normal ?10 ?g/ml]), the CRP level
(33.2 ? 3 mg/liter), and the ESR (30.4 ? 4.6 mm/hour)
were observed. After 1 month of biologic therapy,
significant reductions were demonstrated for the A-SAA
level (244 ? 76 ?g/ml; P ? 0.001), the CRP level (15 ?
4 mg/liter; P ? 0.001), and the ESR (19 ? 3 mm/hour;
P ? 0.001); these reductions were maintained after 3
months (for A-SAA, 246 ? 76 ?g/ml [P ? 0.001]; for
CRP, 14 ? 3 mg/liter [P ? 0.001]; for ESR, 17 ? 3
mm/hour [P ? 0.001]).
Associations of A-SAA with matrix degradation.
Baseline A-SAA levels were significantly corrected with
baseline C1,2C levels (P ? 0.01), but no correlation was
observed for the CRP level or the ESR. Baseline C2C
levels correlated significantly with A-SAA (P ? 0.01) but
not with CRP or the ESR. The baseline A-SAA level was
significantly correlated with 1-year radiographic progres-
sion (r ? 0.408, P ? 0.002). The ESR (r ? 0.512, P ?
0.001) and CRP (r ? 0.4, P ? 0.002) were also correlated
with 1-year radiographic progression.
Baseline MMP-3 levels and the MMP-3:TIMP-1
ratio were significantly correlated with A-SAA (r ?
0.354, P ? 0.02 and r ? 0.488, P ? 0.005, respectively),
CRP (r ? 0.419, P ? 0.01 and r ? 0.424, P ? 0.01,
respectively), and the ESR (r ? 0.333, P ? 0.023 and r ?
0.53, P ? 0.001, respectively). A-SAA levels were signif-
icantly correlated with C1,2C at 1 month (r ? 0.358, P ?
0.006), with C2C at 1 month (r ? 0.408, P ? 0.002) and
3 months (r ? 0.327, P ? 0.012), and with MMP-3 (r ?
0.405, P ? 0.01) and the MMP-3:TIMP-1 ratio (r ?
0.589, P ? 0.01) at 3 months.
To further characterize the relationship between
markers of inflammation, collagen cleavage biomarkers,
and MMPs, patients were grouped according to the
highest and lowest quartiles for the baseline A-SAA,
ESR, and CRP values, as shown in Figures 1A and B.
C1,2C levels were significantly higher in patients with
high serum A-SAA concentrations compared with pa-
tients with low serum A-SAA concentrations (mean ?
SEM 434 ? 38 ng/ml versus 310 ? 24 ng/ml; P ? 0.004)
and to a lesser extent in those with high ESRs compared
with low ESRs (410 ? 27 ng/ml versus 347 ? 28 ng/ml;
P ? 0.049), with no difference observed for CRP (Figure
1A). C2C levels were also significantly higher in patients
with high A-SAA expression compared with patients
with low A-SAA expression (mean ? SEM 155 ? 25
ng/ml versus 98 ? 7 ng/ml; P ? 0.017) (Figure 1B). No
difference was observed for C2C levels when the highest
and lowest quartiles for the ESR or CRP were exam-
ined. The baseline MMP-3:TIMP-1 ratio was signifi-
cantly higher in patients with high baseline A-SAA levels
(0.22 versus 0.13; P ? 0.04), high baseline CRP levels
(0.27 versus 0.12; P ? 0.007), and high ESRs (0.26 versus
0.1; P ? 0.01) (Figure 1C).
Associations of A-SAA with radiographic pro-
gression. Baseline serum levels of inflammation markers
were compared between radiographic nonprogressors
(SHS increase ?1.02) and radiographic progressors
(SHS increase ?1.02) (Figures 1D–F). A-SAA levels
were significantly higher in radiographic progressors
both at baseline (mean ? SEM 664 ? 185 ?g/ml versus
216 ? 66 ?g/ml in nonprogressors; P ? 0.002) and
following 1 month of therapy (359 ? 136 ?g/ml versus
99 ? 29 ?g/ml; P ? 0.031). The ESR was higher in
progressors at both baseline (44 ? 4 mm/hour versus
22 ? 4 mm/hour in nonprogressors; P ? 0.001) and 3
months (25 ? 5 mm/hour versus 16 ? 4 mm/hour; P ?
0.007). The CRP level was significantly higher in pro-
gressors only at baseline (50 ? 9 mg/liter versus 21 ? 4
mg/liter in nonprogressors; P ? 0.003).
To investigate the predictive relationship be-
ROLE OF A-SAA IN INFLAMMATORY ARTHRITIS1039
tween baseline A-SAA levels and subsequent radio-
graphic progression, predictive factors identified for
radiographic progression during univariate analysis were
examined using binary logistic multivariate regression
analysis. Baseline A-SAA levels were independently
associated with radiographic progression (P ? 0.038).
Neither the baseline ESR (P ? 0.1) nor the baseline
CRP level (P ? 0.41) was associated with radiographic
progression (Table 2).
Effect of A-SAA on FLS. The effect of recombi-
nant A-SAA and TNF? on MMP production by primary
FLS was examined using an MMP antibody array.
MMP-1, MMP-3, and MMP-13 were up-regulated fol-
lowing stimulation with A-SAA (50 ?g/ml) and TNF?
(10 ng/ml). The results of ELISA showed that A-SAA
significantly increased MMP-1, MMP-3, and MMP-13
expression in RA (P ? 0.05) (Figures 2A–C) and
increased MMP-3 expression in PsA synovial explant
cultures (Figure 2D). Moreover, MMP expression in-
duced by A-SAA was comparable with that induced by
IL-1? (10 ng/ml) and TNF? (10 ng/ml) (P ? 0.05).
A-SAA did not significantly increase TIMP expression
levels (data not shown), which is consistent with previous
studies (10,39). A-SAA markedly induced pro–MMP-2
activity in RA FLS (Figure 2E) and pro and active forms
of MMP-2 and MMP-9 in RA synovial explants (Figure
2F) and PsA synovial explants (Figure 2G).
To further assess the effects of A-SAA on the
regulation of matrix turnover, we examined the effect of
blocking a specific A-SAA receptor, SR-B1, using the
synovial explant model. Explants were cultured with
mouse anti–SR-B1 (10 ?g/ml) and matched IgG isotype
control antibody (10 ?g/ml). Anti–SR-B1 inhibited
spontaneous secretion of MMP-3 and MMP-2/9 in 3 of
4 synovial explant experiments, suggesting that SR-B1
blockade inhibits the effects of endogenously produced
A-SAA. The different responses to SR-B1 blockade
reflect the heterogeneity of patients and are consis-
tent with previous findings showing that differential
Figure 1. A–C, Serum amyloid A (SAA) correlates with collagen cleavage and radiographic progression. Serum levels of C1,2C (A), C2C (B), and
matrix metalloproteinase 3 (MMP-3)/tissue inhibitor of metalloproteinases 1 (TIMP-1) (C) separated according to the lowest (open bars) and
highest (solid bars) quartiles for the erythrocyte sedimentation rate (ESR) and serum C-reactive protein (CRP) and acute-phase SAA (A-SAA)
levels. D–F, Serum levels of A-SAA (D) and CRP (E) and the ESR (F) at baseline, 1 month, and 3 months in patients for whom followup data on
radiographic progression were available (n ? 55), separated according to nonprogressors (open bars; n ? 24) and progressors (solid bars; n ? 31)
Values are the mean ? SEM. ? ? P ? 0.05.
Binarylogisticregressionanalysis for radiographic
Baseline input covariateOR (95% CI)P
SHS (per unit)
SAA, ?g/ml (per unit)
ESR, mm/hour (per unit)
CRP, mg/liter (per unit)
* Analyses were controlled for age, sex, and diagnosis. P values less
than 0.05 indicate that the covariate was independently associated with
radiographic progression. OR ? odds ratio; 95% CI ? 95% confi-
dence interval; SHS ? Sharp/van der Heijde score; SAA ? serum
amyloid A; ESR ? erythrocyte sedimentation rate; CRP ? C-reactive
1040 CONNOLLY ET AL
expression of SR-B1 is correlated with the responsive-
ness of RA synovial fibroblasts (RASFs) to A-SAA in
Effect of A-SAA on MMP expression in chondro-
cytes. Real-time PCR using RNA from A-SAA–treated
SW1353 chondrocytes was used to examine MMP-1 and
MMP-13 induction (Figures 3A and B). A-SAA (50
?g/ml) significantly increased expression of both genes
at all time points. Maximal MMP-1 induction occurred
at 24 hours, when a 6.5-fold increase was observed (P ?
0.03). Furthermore, stimulation of chondrocytes with
TNF? resulted in a 99-fold up-regulation of MMP-1
(data not shown). Similarly, A-SAA (50 ?g/ml) stimu-
lation of SW1353 cells resulted in up-regulation of
Figure 2. A-SAA–induced MMP production in fibroblast-like synoviocytes (FLS) from patients with rheumatoid arthritis (RA) and patients with
psoriatic arthritis (PsA). A–D, A-SAA significantly increased MMP-1, MMP-3, and MMP-13 expression in RA FLS (A–C) and PsA FLS (D) in a
dose-dependent manner. Bars show the mean ? SD (n ? 3). ? ? P ? 0.05 versus basal (B) control. E–G, Stimulation with A-SAA and tumor necrosis
factor ? (TNF?) significantly increased expression of the pro and active forms of MMP-2 and MMP-9 in RA FLS (n ? 3) (E), RA explant cultures
(n ? 3) (F), and PsA explant cultures (n ? 5) (G). A-SAA stimulation increased expression of the pro and active forms of MMP-2 and MMP-9 to
levels greater than or comparable with those induced by stimulation with interleukin-1? (IL-1?) or TNF?. Representative gelatin zymograms are
shown. See Figure 1 for other definitions.
Figure 3. A-SAA–induced MMP production in chondrocyte cultures. SW1353 cells were cultured for the 3 hours, 10 hours, or 24 hours with
medium alone (control; solid bars), A-SAA 10 ?g/ml (shaded bars), and A-SAA 50 ?g/ml (open bars). A and B, Total RNA was isolated and analyzed
by real-time polymerase chain reaction for expression of MMP-1 and MMP-13, respectively. GAPDH and ?-actin were used as endogenous controls.
C–E, Primary chondrocytes were analyzed for MMP activity by enzyme-linked immunosorbent assay and zymography following 24-hour stimulation
with A-SAA, interleukin-1? (IL-1?), or tumor necrosis factor ? (TNF?). A-SAA significantly increased MMP-1 (C), MMP-13 (D), and MMP-3 (E)
expression in primary chondrocyte cultures in a dose-dependent manner. ? ? P ? 0.05 versus control or basal (B). F, A-SAA induced MMP-2 activity
in primary human chondrocytes, as shown in a representative zymogram. Values in A–E are the mean ? SD. See Figure 1 for other definitions.
ROLE OF A-SAA IN INFLAMMATORY ARTHRITIS1041
MMP-13 gene expression, with maximal levels observed
between 3 hours and 10 hours (5.8-fold and 5.3-fold
increases, respectively, compared with basal control
[P ? 0.02 and P ? 0.01, respectively]). A 74-fold increase
in MMP-13 expression was observed following 10 hours
of TNF? stimulation (P ? 0.03).
In addition, A-SAA (50 ?g/ml) significantly in-
duced MMP-1 protein expression (P ? 0.001) to a level
equivalent to that induced by stimulation with IL-1? (10
ng/ml) or TNF? (10 ng/ml) (Figure 3C). MMP-13 was
significantly induced by both IL-1? and A-SAA (50
?g/ml) (Figure 3D). There was no significant effect on
TIMP-1 levels following stimulation (data not shown).
Maximal expression of MMP-3 was observed with
A-SAA (50 ?g/ml) (Figure 3E). However, when the
MMP:TIMP ratios were analyzed, significant increases
for A-SAA (50 ?g/ml), IL-1?, and TNF? (10 ng/ml)
were demonstrated, suggesting a shift in balance of
MMP activity. The MMP-1:TIMP-1 ratio significantly
increased in response to stimulation with A-SAA 10
?g/ml, A-SAA 50 ?g/ml, and TNF?, from a mean ? SD
level of 2.01 ? 0.8 ng/ml to 5.71 ? 1.0 ng/ml (P ? 0.05),
8.34 ? 1.5 ng/ml (P ? 0.05), and 10.09 ? 1.1 ng/ml (P ?
0.05), respectively. Similarly, A-SAA, IL-1?, and TNF?
all significantly increased the ratio of MMP-3 to TIMP-1
(P ? 0.05). As shown in Figure 3F, increased MMP-2
expression was demonstrated in response to stimulation
with A-SAA (10–50 ?g/ml), IL-1? (10 ng/ml), and TNF?
Regulation of TNF? by A-SAA in synovial tissue.
Finally, we examined whether A-SAA can directly alter
TNF? expression in synovial explant cultures. A-SAA 50
?g/ml dramatically up-regulated TNF? production by
18-fold, from a mean ? SD level of 280 ? 22.8 pg/ml to
5,085 ? 1,336 pg/ml (P ? 0.05) (Figure 4). These results
highlight an important role for A-SAA in driving pro-
inflammatory processes and suggest that blockade of
proinflammatory SAA pathways may be efficacious in
controlling inflammatory arthritis.
This study is the first to show in vivo evidence of
significant relationships between A-SAA and measures
of disease activity, cartilage destruction, and radio-
graphic progression, through a subanalysis of a previ-
ously published cohort of patients with inflammatory
arthritis (16). Here, we identified associations between
A-SAA and cartilage and noncartilage collagen cleavage
by MMP, collagenases, C2C, and C1,2C in serum and
radiographic progression in vivo. Furthermore, we also
demonstrated that A-SAA increased matrix turnover in
favor of cartilage degradation in cartilage cell cultures
and stimulated MMP production and activity in synovial
cell cultures in vitro and in RA and PsA synovial explant
cultures ex vivo.
We previously showed that the collagen degrada-
tion biomarkers C2C and C1,2C can identify radio-
graphic progression in this clinical cohort (16). The
strongest associations observed in this study were those
between A-SAA and the MMP-3:TIMP-1 ratio and
between A-SAA and 1-year radiographic progression,
which is consistent with previous studies relating MMP
and TIMP to RA disease activity (9,11). The weaker
associations observed between A-SAA and cartilage
neoepitopes may be indirect, through induction of de-
structive MMPs. We also demonstrated that inhibition
of the A-SAA receptor SR-B1 inhibits MMP-3, MMP-2,
and MMP-9, suggesting that they mediate the effects
of endogenously produced A-SAA. Finally, we demon-
strated that A-SAA dramatically increased TNF? ex-
pression in whole tissue synovial explants, suggesting
that A-SAA is a key mediator of pathology in inflamma-
tory arthritis in vivo.
In this study, we demonstrated a specific relation-
ship between A-SAA and clinical disease activity in
inflammatory arthritis, as measured by the SJC. Hepatic
induction of acute-phase proteins including CRP is a
nonspecific feature of systemic inflammation in response
to diverse causes, including trauma, infection, and auto-
immune disease (36). In contrast to other hepatic-
derived acute-phase proteins, A-SAA at concentrations
of up to 1,000 ?g/ml has previously been reported in
Figure 4. Acute-phase serum amyloid A (A-SAA)–induced produc-
tion of tumor necrosis factor ? (TNF?) in ex vivo rheumatoid arthritis
(RA) synovial explant cultures. Whole-tissue synovial explants from
patients with RA were cultured for 24 hours with medium alone
(control) or A-SAA (10 ?g/ml or 50 ?g/ml). Supernatants were
harvested and analyzed for TNF? by multiplex assay. Bars show the
mean ? SD (n ? 8). ? ? P ? 0.05 versus control.
1042 CONNOLLY ET AL
RASFs, which may exceed serum levels obtained in the
same subjects (37). Measurement of serum A-SAA
levels, which may be partly SF-derived, may therefore be
a more accurate indicator of radiographic progression
in clinical practice, although further direct comparison
with high-sensitivity CRP may be useful. Although het-
erogeneous by diagnosis, the patients with RA and the
patients with PsA in this cohort were indistinguishable
in terms of disease duration and baseline radiographic
damage prior to biologic therapy. The relatively high
incidence of radiographic progression observed in both
patients with RA and patients with PsA prior to biologic
therapy is likely to be due to a combination of factors,
including the high incidence of baseline structural
damage, the long history of DMARD-resistant disease,
incomplete adherence to therapy, and a relatively high
rate of biologic monotherapy, which has been shown to
be associated with worse radiographic outcomes (38).
Using in vitro and ex vivo culture models, we
demonstrated the potent ability of A-SAA to induce
collagenases MMP-1, MMP-3, and MMP-13 and gelati-
nases MMP-2 and MMP-9 in human chondrocytes, RA
FLS, and RA and PSA explant cultures, with potency
similar to that of IL-1? and TNF?. Previous studies have
shown the importance of MMPs to the destructive
process in inflammatory arthritis, with many studies
showing elevated levels of MMPs and TIMP in both RA
and PsA serum and synovial fluid (3,39). MMP-3 corre-
lates with progression of erosion in RA (9), and suppres-
sion of MMP-3 in PsA following treatment with adali-
mumab has been demonstrated (40). Furthermore, we
demonstrated A-SAA induction and transcription of
MMP-1 and MMP-13 in chondrocytes, confirming both
transcriptional and posttranslational regulation. A sig-
nificant shift in the MMP:TIMP ratio in response to
A-SAA suggests that high A-SAA levels in the joint
favor destructive pathways. Increased collagen cleavage
is dependent on an increased ratio of collagenase in
relation to its natural inhibitors, TIMPs (10). Treatment
with anti-TNF? has been shown to reduce the MMP-1:
TIMP-1 ratio (41).
The level of response to A-SAA stimulation was
greater for MMP-3 than for MMP-1, which is consistent
with the findings of previous studies in RA FLS (42).
The induction of MMP-2 and MMP-9 by A-SAA sup-
ports an angiogenic role for A-SAA in RA synovial
tissue and endothelial cells and in an RA synovium–
engrafted SCID mouse chimera model (19,42,43), pro-
moting the invasion of the adjacent cartilage by the
proliferating synovial pannus. The role of A-SAA in
joint destruction is further supported by the effect of
SR-B1 blockade on MMP expression in synovial explant
cultures, suggesting inhibition of the regulatory effects
of endogenously secreted A-SAA. SR-B1 is expressed in
RA synovial perivascular and lining layer regions and
mediates A-SAA–induced proinflammatory and angio-
genic effects (27,44). Therefore, the spontaneous secre-
tion of A-SAA from RA and PsA explants (43), the
association of A-SAA with radiographic progression,
and the ability of A-SAA to regulate MMP/TIMP
balance in primary synovial cell provide further evidence
for a pathogenic role of A-SAA in collagenase-mediated
joint destruction in inflammatory arthritis.
Finally, using an ex vivo synovial explant tissue
model of inflammatory arthritis, we demonstrated for
the first time that A-SAA dramatically up-regulates
TNF? itself, suggesting that A-SAA may be a potent
upstream mediator of TNF?, and that A-SAA could
influence patient responses to existing anti-TNF? ther-
apies. Anti-TNF? therapies represent a major advance
in the treatment of inflammatory arthritis; however, a
significant proportion of patients fail to experience a
response following treatment (45–47), indicating diver-
gence in the hierarchical control of inflammation by
cytokine mediators. Similar to TNF?, A-SAA has been
identified as pathogenic in other chronic diseases with
an inflammatory component, including atherosclerosis
(48), metabolic syndrome (49), and cancer (50). Here,
we provide evidence that targeting A-SAA may have
efficacy in the treatment of inflammatory arthritis.
In conclusion, the results of this study provide
evidence that A-SAA is a sensitive marker of synovitis
and joint destruction in a cohort of patients with inflam-
matory arthritis receiving biologic therapy. We showed
that A-SAA may be directly involved in joint destruction
through activation of MMP/TIMP activity and increased
TNF? production in RA synovial tissue and through its
association with cartilage collagen cleavage biomarkers
All authors were involved in drafting the article or revising it
critically for important intellectual content, and all authors approved
the final version to be published. Dr. Fearon 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 data analysis.
Study conception and design. Connolly, Mullan, Matthews, FitzGerald,
Poole, Bresnihan, Veale, Fearon.
Acquisition of data. Connolly, Mullan, McCormick, Matthews, Sullivan,
Kennedy, Veale, Fearon.
Analysis and interpretation of data. Connolly, Mullan, McCormick,
Kennedy, Poole, Veale, Fearon.
1. Harris ED Jr. Rheumatoid arthritis: pathophysiology and implica-
tions for therapy. N Engl J Med 1990;322:1277–89.
ROLE OF A-SAA IN INFLAMMATORY ARTHRITIS1043
2. Fearon U, Reece R, Smith J, Emery P, Veale DJ. Synovial
cytokine and growth factor regulation of MMPs/TIMPs: implica-
tions for erosions and angiogenesis in early rheumatoid and
psoriatic arthritis patients. Ann N Y Acad Sci 1999;878:619–21.
3. Fraser A, Fearon U, Billinghurst RC, Ionescu M, Reece R,
Barwick T, et al. Turnover of type II collagen and aggrecan in
cartilage matrix at the onset of inflammatory arthritis in humans:
relationship to mediators of systemic and local inflammation.
Arthritis Rheum 2003;48:3085–95.
4. Drossaers-Bakker KW, de Buck M, van Zeben D, Zwinderman
AH, Breedveld FC, Hazes JM. Long-term course and outcome of
functional capacity in rheumatoid arthritis: the effect of disease
activity and radiologic damage over time. Arthritis Rheum 1999;
5. Van der Heijde D. How to read radiographs according to the
Sharp/van der Heijde method [corrected and republished in
J Rheumatol 2000;27:261–3]. J Rheumatol 1999;26:743–5.
6. Felson DT, Anderson JJ, Boers M, Bombardier C, Chernoff M,
Fried B, et al. The American College of Rheumatology prelimi-
nary core set of disease activity measures for rheumatoid arthritis
clinical trials. Arthritis Rheum 1993;36:729–40.
7. Poole AR, Alini M, Hollander AP. Cell biology of cartilage
degradation. In: Henderson B, Edwards JC, Pettipher ER, editors.
Mechanisms and models in rheumatoid arthritis. London: Aca-
demic Press; 1995. p. 163–204.
8. Poole AR. Can serum biomarker assays measure the progression
of cartilage degeneration in osteoarthritis? [editorial]. Arthritis
9. Green MJ, Gough AK, Devlin J, Smith J, Astin P, Taylor D, et al.
Serum MMP-3 and MMP-1 and progression of joint damage in
early rheumatoid arthritis. Rheumatology (Oxford) 2003;42:83–8.
10. Moran EM, Mullan R, McCormick J, Connolly M, Sullivan O,
FitzGerald O, et al. Human rheumatoid arthritis tissue production
of IL-17A drives matrix and cartilage degradation: synergy with
tumour necrosis factor-?, Oncostatin M and response to biologic
therapies. Arthritis Res Ther 2009;11:R113.
11. Yoshihara Y, Obata K, Fujimoto N, Yamashita K, Hayakawa T,
Shimmei M. Increased levels of stromelysin-1 and tissue inhibitor
of metalloproteinases–1 in sera from patients with rheumatoid
arthritis. Arthritis Rheum 1995;38:969–75.
12. Landewe R, Geusens P, Boers M, van der Heijde D, Lems W,
te Koppele J, et al. Markers for type II collagen breakdown predict
the effect of disease-modifying treatment on long-term radio-
graphic progression in patients with rheumatoid arthritis. Arthritis
13. Verstappen SM, Poole AR, Ionescu M, King LE, Abrahamowicz
M, Hofman DM, et al. Radiographic joint damage in rheumatoid
arthritis is associated with differences in cartilage turnover and can
be predicted by serum biomarkers: an evaluation from 1 to 4 years
after diagnosis. Arthritis Res Ther 2006;8:R31.
14. Dodge GR, Poole AR. Immunohistochemical detection and im-
munochemical analysis of type II collagen degradation in human
normal, rheumatoid, and osteoarthritic articular cartilages and in
explants of bovine articular cartilage cultured with interleukin 1.
J Clin Invest 1989;83:647–61.
15. Hollander AP, Pidoux I, Reiner A, Rorabeck C, Bourne R, Poole
AR. Damage to type II collagen in aging and osteoarthritis starts
at the articular surface, originates around chondrocytes, and
extends into the cartilage with progressive degeneration. J Clin
16. Mullan RH, Matthews C, Bresnihan B, FitzGerald O, King L,
Poole AR, et al. Early changes in serum type II collagen biomark-
ers predict radiographic progression at one year in inflammatory
arthritis patients after biologic therapy. Arthritis Rheum 2007;56:
17. Baumann H, Gauldie J. The acute phase response. Immunol
18. Cunnane G, Grehan S, Geoghegan S, McCormack C, Shields D,
Whitehead AS, et al. Serum amyloid A in the assessment of early
inflammatory arthritis. J Rheumatol 2000;27:58–63.
19. Mullan RH, Bresnihan B, Golden-Mason L, Markham T, O’Hara
R, FitzGerald O, et al. Acute-phase serum amyloid A stimulation
of angiogenesis, leukocyte recruitment, and matrix degradation in
rheumatoid arthritis through an NF-?B–dependent signal trans-
duction pathway. Arthritis Rheum 2006;54:105–14.
20. Moriguchi M, Terai C, Kaneko H, Koseki Y, Kajiyama H, Uesato
M, et al. A novel single-nucleotide polymorphism at the 5?-
flanking region of SAA1 associated with risk of type AA amyloid-
osis secondary to rheumatoid arthritis. Arthritis Rheum 2001;44:
21. Strachan AF, de Beer FC, van der Westhuyzen DR, Coetzee GA.
Identification of three isoform patterns of human serum amyloid
A protein. Biochem J 1988;250:203–7.
22. Uhlar CM, Whitehead AS. Serum amyloid A, the major vertebrate
acute-phase reactant. Eur J Biochem 1999;265:501–23.
23. Bjorkman L, Raynes JG, Shah C, Karlsson A, Dahlgren C, Bylund
J. The proinflammatory activity of recombinant serum amyloid A
is not shared by the endogenous protein in the circulation.
Arthritis Rheum 2010;62:1660–5.
24. Bjorkman L, Karlsson J, Karlsson A, Rabiet MJ, Boulay F, Fu H,
et al. Serum amyloid A mediates human neutrophil production of
reactive oxygen species through a receptor independent of formyl
peptide receptor like-1. J Leukoc Biol 2008;83:245–53.
25. Christenson K, Bjorkman L, Tangemo C, Bylund J. Serum amyloid
A inhibits apoptosis of human neutrophils via a P2X7-sensitive
pathway independent of formyl peptide receptor-like 1. J Leukoc
26. Kjelgaard-Hansen M, Christensen MB, Lee MH, Jensen AL,
Jacobsen S. Serum amyloid A isoforms in serum and synovial fluid
from spontaneously diseased dogs with joint diseases or other
conditions. Vet Immunol Immunopathol 2007;117:296–301.
27. Mullan RH, McCormick J, Connolly M, Bresnihan B, Veale DJ,
Fearon U. A role for the high-density lipoprotein receptor SR-B1
in synovial inflammation via serum amyloid-A. Am J Pathol;176:
28. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF,
Cooper NS, et al. The American Rheumatism Association 1987
revised criteria for the classification of rheumatoid arthritis.
Arthritis Rheum 1988;31:315–24.
29. Veale D, Rogers S, FitzGerald O. Classification of clinical subsets
in psoriatic arthritis. Br J Rheumatol 1994;33:133–8.
30. Prevoo ML, van ’t Hof MA, Kuper HH, van Leeuwen MA, van de
Putte LB, van Riel PL. Modified disease activity scores that
include twenty-eight–joint counts: development and validation in a
prospective longitudinal study of patients with rheumatoid arthri-
tis. Arthritis Rheum 1995;38:44–8.
31. Bruynesteyn K, Boers M, Kostense P, van der Linden S, van der
Heijde D. Deciding on progression of joint damage in paired films
of individual patients: smallest detectable difference or change.
Ann Rheum Dis 2005;64:179–82.
32. Mazzuca SA, Poole AR, Brandt KD, Katz BP, Lane KA, Lobanok
T. Associations between joint space narrowing and molecular
markers of collagen and proteoglycan turnover in patients with
knee osteoarthritis. J Rheumatol 2006;33:1147–51.
33. Youssef PP, Kraan M, Breedveld F, Bresnihan B, Cassidy N,
Cunnane G, et al. Quantitative microscopic analysis of inflamma-
tion in rheumatoid arthritis synovial membrane samples selected
at arthroscopy compared with samples obtained blindly by needle
biopsy. Arthritis Rheum 1998;41:663–9.
34. Fearon U, Mullan R, Markham T, Connolly M, Sullivan S, Poole
AR, et al. Oncostatin M induces angiogenesis and cartilage
degradation in rheumatoid arthritis synovial tissue and human
cartilage cocultures. Arthritis Rheum 2006;54:3152–62.
35. Emery P, Breedveld FC, Hall S, Durez P, Chang DJ, Robertson D,
1044 CONNOLLY ET AL
et al. Comparison of methotrexate monotherapy with a combina-
tion of methotrexate and etanercept in active, early, moderate to
severe rheumatoid arthritis (COMET): a randomised, double-
blind, parallel treatment trial. Lancet 2008;372:375–82.
36. Chait A, Han CY, Oram JF, Heinecke JW. Thematic review series:
the immune system and atherogenesis. Lipoprotein-associated
inflammatory proteins: markers or mediators of cardiovascular
disease? J Lipid Res 2005;46:389–403.
37. Kumon Y, Suehiro T, Hashimoto K, Nakatani K, Sipe JD. Local
expression of acute phase serum amyloid A mRNA in rheumatoid
arthritis synovial tissue and cells. J Rheumatol 1999;26:785–90.
38. Breedveld FC, Weisman MH, Kavanaugh AF, Cohen SB, Pavelka
K, van Vollenhoven R, et al, for the PREMIER Investigators. The
PREMIER study: a multicenter, randomized, double-blind clinical
trial of combination therapy with adalimumab plus methotrexate
versus methotrexate alone or adalimumab alone in patients with
early, aggressive rheumatoid arthritis who had not had previous
methotrexate treatment. Arthritis Rheum 2006;54:26–37.
39. Myers A, Lakey R, Cawston TE, Kay LJ, Walker DJ. Serum
MMP-1 and TIMP-1 levels are increased in patients with psoriatic
arthritis and their siblings. Rheumatology (Oxford) 2004;43:272–6.
40. Van Kuijk AW, DeGroot J, Koeman RC, Sakkee N, Baeten DL,
Gerlag DM, et al. Soluble biomarkers of cartilage and bone
metabolism in early proof of concept trials in psoriatic arthritis:
effects of adalimumab versus placebo. PLoS One 2010;5:e12556.
41. Catrina AI, Lampa J, Ernestam S, af Klint E, Bratt J, Klareskog L,
et al. Anti-tumour necrosis factor (TNF)-? therapy (etanercept)
down-regulates serum matrix metalloproteinase (MMP)-3 and
MMP-1 in rheumatoid arthritis. Rheumatology (Oxford) 2002;41:
42. O’Hara R, Murphy EP, Whitehead AS, FitzGerald O, Bresnihan
B. Local expression of the serum amyloid A and formyl peptide
receptor–like 1 genes in synovial tissue is associated with matrix
metalloproteinase production in patients with inflammatory ar-
thritis. Arthritis Rheum 2004;50:1788–99.
43. Connolly M, Marrelli A, Blades M, McCormick J, Maderna P,
Godson C, et al. Acute serum amyloid A induces migration,
angiogenesis, and inflammation in synovial cells in vitro and in a
human rheumatoid arthritis/SCID mouse chimera model. J Immu-
44. Baranova IN, Vishnyakova TG, Bocharov AV, Kurlander R, Chen
Z, Kimelman ML, et al. Serum amyloid A binding to CLA-1
(CD36 and LIMPII analogous-1) mediates serum amyloid A
protein-induced activation of ERK1/2 and p38 mitogen-activated
protein kinases. J Biol Chem 2005;280:8031–40.
45. Bathon JM, Martin RW, Fleischmann RM, Tesser JR, Schiff MH,
Keystone EC, et al. A comparison of etanercept and methotrexate
in patients with early rheumatoid arthritis. N Engl J Med 2000;
46. Lipsky PE, van der Heijde DM, St Clair EW, Furst DE, Breedveld
FC, Kalden JR, et al, for the Anti-Tumor Necrosis Factor Trial in
Rheumatoid Arthritis with Concomitant Therapy Study Group.
Infliximab and methotrexate in the treatment of rheumatoid
arthritis. N Engl J Med 2000;343:1594–602.
47. Weinblatt ME, Keystone EC, Furst DE, Moreland LW, Weisman
MH, Birbara CA, et al. Adalimumab, a fully human anti–tumor
necrosis factor ? monoclonal antibody, for the treatment of
rheumatoid arthritis in patients taking concomitant methotrexate:
the ARMADA trial. Arthritis Rheum 2003;48:35–45.
48. Artl A, Marsche G, Pussinen P, Knipping G, Sattler W, Malle E.
Impaired capacity of acute-phase high density lipoprotein particles
to deliver cholesteryl ester to the human HUH-7 hepatoma cell
line. Int J Biochem Cell Biol 2002;34:370–81.
49. Berger JJ, Barnard RJ. Effect of diet on fat cell size and
50. Chan DC, Chen CJ, Chu HC, Chang WK, Yu JC, Chen YJ, et al.
Evaluation of serum amyloid A as a biomarker for gastric cancer.
Ann Surg Oncol 2007;14:84–93.
J Appl Physiol1999;87:
In the article by Emery et al in the May 2011 issue of Arthritis & Rheumatism (pages 1200–1210), some of the
data in the GO-FORWARD study portion of Table 3 were inadvertently transposed. The mean ? SD change
in the erosion score from baseline to week 52 should have been shown as ?0.04 ? 1.09 in group 4 (median
[interquartile range 0.00 [0.00, 0.00]) and 0.14 ? 1.41 in groups 3 and 4 combined (median [interquartile
range] 0.00 [0.00, 0.50]). The mean ? SD change in the joint space narrowing score from baseline to week 52
should have been shown as 0.24 ? 0.97 in group 4 and 0.23 ? 0.94 in groups 3 and 4 combined.
We regret the error.
ROLE OF A-SAA IN INFLAMMATORY ARTHRITIS1045