ORIGINAL RESEARCH PAPER
Bosentan, an endothelin receptor antagonist, ameliorates
collagen-induced arthritis: the role of TNF-a in the induction
of endothelin system genes
Paula B. Donate•Thiago M. Cunha•Waldiceu A. Verri Jr•Cristina M. Junta•
Flavia O. Lima•Silvio M. Vieira•Rafael S. Peres•Karina F. Bombonato-Prado•
Paulo Louzada Jr•Sergio H. Ferreira•Eduardo A. Donadi•Geraldo A. S. Passos•
Fernando Q. Cunha
Received: 5 September 2011/Revised: 23 October 2011/Accepted: 5 December 2011/Published online: 17 January 2012
? Springer Basel AG 2012
inflammatory events. The present study investigated the
efficacy of bosentan, a dual ETA/ETB receptor antagonist,
in collagen-induced arthritis (CIA) in mice.
CIA was induced in DBA/1J mice. Arthritic
mice were treated with bosentan (100 mg/kg) once a day,
CIA progression was assessed by measurements
of visual clinical score, paw swelling and hypernocicep-
tion. Histological changes, neutrophil infiltration and pro-
inflammatory cytokines were evaluated in the joints. Gene
expression in the lymph nodes of arthritic mice was eval-
uated by microarray technology. PreproET-1 mRNA
Endothelins (ETs) are involved in several
expression in the lymph nodes of mice and in peripheral
blood mononuclear cells (PBMCs) was evaluated by real-
time PCR. The differences were evaluated by one-way
ANOVA or Student’s t test.
Oral treatment with bosentan markedly amelio-
rated the clinical aspects of CIA (visual clinical score, paw
swelling and hyperalgesia). Bosentan treatment also
reduced joint damage, leukocyte infiltration and pro-
inflammatory cytokine levels (IL-1b, TNFa and IL-17) in
the joint tissues. Changes in gene expression in the lymph
nodes of arthritic mice returned to the levels of the control
mice after bosentan treatment. PreproET mRNA expres-
sion increased in PBMCs from rheumatoid arthritis (RA)
patients but returned to basal level in PBMCs from patients
under anti-TNF therapy. In-vitro treatment of PBMCs with
TNFa upregulated ET system genes.
These findings indicate that ET receptor
antagonists, such as bosentan, might be useful in control-
ling RA. Moreover, it seems that ET mediation of arthritis
is triggered by TNFa.
P. B. Donate and T. M. Cunha contributed equally to this study.
Responsible Editor: John Di Battista.
Electronic supplementary material
article (doi:10.1007/s00011-011-0415-5) contains supplementary
material, which is available to authorized users.
The online version of this
P. B. Donate ? T. M. Cunha ? E. A. Donadi ?
G. A. S. Passos ? F. Q. Cunha
Post Graduation Program in Basic and Applied Immunology,
School of Medicine of Ribeira ˜o Preto, University of Sa ˜o Paulo,
Avenue Bandeirantes, 3900, Ribeira ˜o Preto,
Sa ˜o Paulo 14049-900, Brazil
T. M. Cunha ? W. A. Verri Jr ? F. O. Lima ?
S. M. Vieira ? R. S. Peres ? S. H. Ferreira ? F. Q. Cunha (&)
Department of Pharmacology, School of Medicine of Ribeira ˜o
Preto, University of Sa ˜o Paulo, Avenue Bandeirantes, 3900,
Ribeira ˜o Preto, Sa ˜o Paulo 14049-900, Brazil
W. A. Verri Jr
Department of Pathology, University of Londrina, Londrina,
Parana ´ 86051-990, Brazil
C. M. Junta
Department of Genetics, School of Medicine of Ribeira ˜o Preto,
University of Sa ˜o Paulo, Avenue Bandeirantes, 3900,
Ribeira ˜o Preto, Sa ˜o Paulo 14049-900, Brazil
K. F. Bombonato-Prado ? G. A. S. Passos
Department of Morphology, School of Dentistry,
University of Sa ˜o Paulo, Avenue Bandeirantes, 3900,
Ribeira ˜o Preto, Sa ˜o Paulo 14049-900, Brazil
P. Louzada Jr ? E. A. Donadi
Division of Clinical Immunology, School of Medicine
of Ribeira ˜o Preto, University of Sa ˜o Paulo,
Avenue Bandeirantes, 3900, Ribeira ˜o Preto,
Sa ˜o Paulo 14049-900, Brazil
Inflamm. Res. (2012) 61:337–348
Collagen-induced arthritis ? Bosentan ? Inflammation ?
Microarray ? Endothelins
Rheumatoid arthritis ?
Tumor necrosis factor
Collagen type 2
Complete Freund’s adjuvant
Significance analysis of microarrays
Peripheral blood mononuclear cells
Rheumatoid arthritis (RA) is a chronic inflammatory dis-
order of unknown cause. It is characterized by synovial
leukocyte infiltration, pannus formation, degradation of
cartilage and bone, and disabling pain [1, 2]. Increased
levels of tumor necrosis factor a (TNF-a) and the inter-
leukins IL-1b and IL-17 have been associated with variable
degrees of bone and cartilage erosion, which causes a loss
of function, disability, shortened life expectancy and con-
siderable health care costs [3, 4].
Although no effective treatment is available, there have
been major advances inthe treatment of RA, including more
aggressive use of disease-modifying antirheumatic drugs
(DMARDs) and the development of immune therapies such
as the anti-TNF agents [5, 6]. Despite these enormous
advances, the large differences in patients’ responses, inci-
dences of remission, and costs of treatments promotes a
continuous search for the development of new drugs.
Endothelins (ETs) are a family of naturally occurring
peptides  with growth-promoting, vasoactive, and noci-
ceptive properties, and they affect the function of a number
of tissues and systems . There are 3 known, 21-amino-
acid-long ETs in humans (ET-1, ET-2, and ET-3) that are
produced as preproET and then cleaved by ET-converting
enzymes to form big-ET and the active peptide .
These peptides have overlapping tissue distributions and
are synthesized by vascular endothelial and smooth muscle
cells and neural, renal, pulmonary and immune cells, such
as macrophages and leukocytes . Additionally, these
peptides seem to exert an important influence by autocrine
and paracrine actions mediated through G-protein-coupled
specific ETA and ETB receptors .
ETs have pathophysiological roles in pulmonary hyper-
tension, arterial hypertension, atherosclerosis, cerebral
vasospasm and inflammatory processes [9, 12], in which
they stimulate the formation of cytokines, such as interleu-
kins and TNF . In this context, high levels of ET-1 are
goutpatients.Plasma levels ofET-1inactive RAexceedthe
values in nonactive RA. Moreover, ET-1-like immunore-
activity in synovial fluid was found to be at levels several
times higher than those in plasma and was found to be
secreted from macrophage-like synoviocytes [14–16]. ETs
may also act locally; ET-1-binding sites are localized in the
synovial blood vessels, modulating synovial perfusion and
exacerbating hypoxia in chronic arthritis . Moreover,
ET-1 has been shown to increase the net metalloproteases/
tissue inhibitor of metalloproteases-1 balance and increase
collagen degradation . Recently, the participation of
ETs in the development of arthritis and arthritic pain was
demonstrated using the antigen-induced monoarthritis
(AIA) model, with methylated bovine serum albumin as the
antigen . Taking into account the evidence mentioned
above, we addressed the efficacy of the dual ET receptor
antagonist bosentan in the collagen-induced arthritis (CIA)
RA. Bosentan is the prototype of the sentan-class drugs and
was first approved by the US Food and Drug Administration
(FDA) for human use in pulmonary arterial hypertension.
We also evaluated the impact of this treatment on differ-
ential gene expression profiles of draining inguinal lymph
nodes. The modulation of ET system gene (preproET, ETA
and ETB) expression was evaluated in peripheral blood
mononuclear cells (PBMCs) from healthy donors and RA
patients, and the impact of anti-TNF-a therapy in this pro-
cess was evaluated.
Materials and methods
Male DBA/1J mice (12 weeks old; 18–22 g) were housed in
temperature-controlled rooms (22–25?C) in the animal facil-
ityofthe School ofMedicineofRibeira ˜o Preto, Universityof
Sa ˜o Paulo, Sa ˜o Paulo, Brazil, and received water and food
ad libitum. The experimental protocols were approved by the
local Ethical Committee on Animal Experimentation of the
School of Medicine of Ribeira ˜o Preto, University of Sa ˜o
Paulo, Sa ˜o Paulo, Brazil (n244/2005).
Induction and assessment of CIA
Male DBA/1J mice received 200 lg bovine type II colla-
gen (CII; Sigma) in complete Freund’s adjuvant (CFA) by
338P. B. Donate et al.
intradermal injection into the tail (day 0). CII [200 lg in
phosphate-buffered saline (PBS)] was given again on day
21 by intraperitoneal injection . Mice were monitored
daily for signs of arthritis, for which severity scores were
derived as follows: 0 = normal, 1 = erythema, 2 = ery-
thema plus swelling, 3 = extension/loss function, and total
score = sum of four limbs. Disease onset characterized by
erythema and/or paw swelling was observed between days
25 and 35. The experimental control group corresponds to
the sham-immunized mice that received the same amount
of CFA injection but without the administration of type II
collagen. For the therapeutic approach, DBA/1J mice were
treated with bosentan (Actelion Pharmaceuticals, Allsch-
wil, Switzerland) (100 mg kg-1) p.o. [21–23] once a day
for a total of 11 days. The treatment began on the day that
CIA was clinically detectable; thus, only mice that devel-
oped CIA were treated.
Measurement of mechanical hyperalgesia: electronic
The term hyperalgesia was used to define the decrease of
nociceptive withdrawal threshold . Mechanical hyper-
algesia was tested in mice as previously reported . The
test consisted of evoking the hindpaw flexion reflex with a
hand-held force transducer (electronic anesthesiometer;
IITC Life Science, Woodland Hills, CA, USA) adapted with
with a gradual increase in pressure. The endpoint was
characterized by removal of the paw followed by clear
flinchingmovements. Afterpaw withdrawal,theintensityof
pressure was automatically recorded. The value for the
response was obtained by averaging three measurements.
The results are expressed by delta (D) withdrawal threshold
(in g) calculated by subtracting zero-time mean measure-
ments from time interval mean measurements.
For histological assessment, CIA mice treated with vehicle
or bosentan were killed at the end of the experiments
(11 days after CIA was clinically detected), and the paw
joints were removed, fixed in 4% PFA buffer, and decal-
cified in EDTA. Sections (4 mm) were stained with H&E.
The joint pathology was examined and scored. Scores are
based on exudates, granulocyte infiltration, hyperplasia,
mononuclear cell infiltration, periarticular mononuclear
and granulocyte cell infiltration (each scoring 0–3), bone
and cartilage destruction (scoring 0–4) and an additional
score of 1 for fibrin deposition, resulting in a maximum
possible score of 20. The sections were scanned with a
Leica DM 4000B microscope, and the pictures were
analyzed at a magnification of 500 lm with Leica LAS
(Leica Application Suite) software.
Eleven days after CIA was clinically detected, animals
were killed by CO2inhalation, and the surrounding joint
tissue was removed and homogenized in 300 lL buffer
containing protease inhibitors. IL-17, TNF-a and IL-1b
levels were determined as described previously  by
ELISA using paired antibodies (R&D Systems). The
results are expressed as pg/joint for each cytokine. As a
control, the concentration of each cytokine was determined
in sham-immunized mice.
Myeloperoxidase activity assay
Neutrophil accumulation in the surrounding joint tissue of
mice was evaluated by assaying myeloperoxidase (MPO)
activity. Eleven days after CIA was clinically detected,
animals were killed by CO2inhalation, and the surrounding
joint tissue was removed and homogenized with a tissue
homogenizer (Power Gen 125, Fischer Scientific) in
0.2 mL pH 4.7 buffer (0.1 M NaCl, 0.02 M NaPO4,
0.015 M Na-EDTA). The suspensions were centrifuged at
3,000g for 15 min, the pellet was resuspended in lysis
buffer (0.2% NaCl) and was further centrifuged for 15 min.
The pellet cells were resuspended and homogenized in
0.5 mL H-TAB buffer (0.05 M NaPO4 buffer (pH 5.4)
containing 0.5% dexadecyltrimethylammonium bromide)
and centrifuged at 10,000g for 15 min, and the superna-
tants were used for determination of MPO activity. MPO
activity in the resuspended pellet was assayed by measur-
ing the change in absorbance at 450 nm using o-dianisidine
dihydrochloride and 1% hydrogen peroxide. The results are
reported as MPO units/mg of tissue. A unit of MPO activity
was defined as that required to convert 1 lmol hydrogen
peroxide to water in 1 min at 22?C.
Total RNA preparation
The DBA/1J mice were killed by CO2inhalation, and the
lymph nodes were surgically removed. To obtain sufficient
mRNA for hybridization to the glass slides and the real-
time PCR experiments, total inguinal lymph node RNA
was pooled at each time point (n = 3 mice). Total RNA
samples were prepared using Trizol?reagent according to
the manufacturer’s instructions (Invitrogen, Carlsbad, CA,
USA). For microarray and real-time PCR experiments, we
used only undegraded and DNA-, protein- and phenol-free
RNA preparations as evaluated by conventional agarose
gel electrophoresis stained with ethidium bromide and
ultraviolet spectrophotometry, respectively.
Bosentan, an endothelin receptor antagonist, ameliorates collagen-induced arthritis339
Healthy donors and RA patients, from the Clinical Hospital
of Ribeira ˜o Preto–USP–SP, with active disease under
treatment with methotrexate (MTX) or MTX plus anti-TNF
(infliximab) were recruited, and they provided written
informed consent. Peripheral blood samples from seven
healthy volunteers and 14 RA patients were collected. All
14 RA patients recruited fulfilled the 1987 revised criteria
of the American College of Rheumatology for the diagnosis
of RA . Seven of the patients were under treatment with
methotrexate (MTX) alone, and seven of the patients were
under treatment with MTX plus anti-TNF (infliximab)
(clinical information is presented in Table 1). According to
the Disease Activity Score 28 (DAS28, ), patients
treated with MTX plus anti-TNF therapy exhibited lower
levels of DAS28 (mean = 3.5) than those treated with
MTX alone (DAS28: mean = 5.6). RA patients that
received anti-TNF therapy presented a reduction of at least
1.6 in their DAS28 score, indicating the beneficial effect of
anti-TNF therapy. The healthy donors did not present any
clinical signs of disease upon anamnesis or clinical
For cell culture and RNA extraction, the PBMCs were
separated on a Ficoll-Paque PLUS density gradient (GE
Healthcare Life Sciences). For in-vitro experiments,
1 9 106PBMCs from six healthy donors and RA patients
were cultured in 10% fetal bovine serum-supplemented
RPMI 1640 medium at 37?C in a 5% CO2atmosphere in
96-well plates. PBMCs were incubated with 100 ng/mL
TNF-a for 2 h. After stimulation, total mRNA was
extracted from PBMCs using Trizol?reagent following the
manufacturer’s instructions. The integrity of the total RNA
samples was evaluated by denaturing agarose gel electro-
phoresis under standard conditions. The study was
approved by the Human Ethics Committee of the Faculty
of Medicine of Ribeirao Preto.
Gene expression analysis using microarrays
Gene expression in the lymph nodes was assessed using
glass slide microarrays prepared on silane-coated Ultra-
GAPS slides (#40015, Corning, New York, NY, USA). The
arrays contained a total of 4,500 cDNA sequences that
represented most murine tissues and organs. Sequences
were obtained from the Soares thymus 2NbMT normalized
library, which represents expressed sequence tag (EST)
cDNA clones prepared from the thymus of a C57BL/6J
4-week-old male mouse, and is available at the IMAGE
The microarrays were prepared based on published
protocols with PCR products from the cDNA clones 
using a Generation III Array Spotter (Amersham Molecular
Dynamics, Sunnyvale, CA, USA). A complete file that
provides all of the genes and ESTs present on the micro-
arrays used in this study is available on the MIAME
database under accession code E-MEXP-2404 (http://www.
Complex cDNA probe preparation and hybridization
The cDNA complex probes derived from the total RNA
obtained from the lymph nodes were prepared by reverse
transcription using 10 lg of total RNA. The cDNA sam-
ples were monocolor labeled with Cy3 fluorochrome using
the CyScribe post-labeling kit (GE Healthcare Life Sci-
ences). Samples were hybridized for 15 h and then washed
with an automatic slide processor system (ASP, Amersham
Biosciences). Microarrays were scanned using a Genera-
tion III laser scanner (Amersham Biosciences).
As a reference for the hybridization procedure, we used
equimolar quantities of cDNAs obtained from unrelated
total RNA (mouse thymus total RNA). This approach
allowed us to estimate the amount of target cDNA on each
Table 1 Clinical and
demographic features of healthy
donors and RA patients treated
with MTX or MTX ? anti-TNF
RF rheumatoid factor, Anti-
CCP anti-cyclic citrullinated
peptide antibodies, MTX
methotrexate, DAS Disease
Clinical features MTX only
(n = 7)
MTX ? infliximab
(n = 7)
(n = 7)
Mean age (years)53.8 ± 4.6 47.2 ± 4.931.9 ± 2.7
Women n, (%)5 (71.4%) 6 (85.7%)4 (57.1%)
Caucasian n, (%)7 (100%) 7 (100%) 7 (100%)
RF-positive n, (%) 6 (85.7%)6 (85.7%)0 (100%)
Anti-CCP positive n, (%) 4 (57.1%) 5 (71.4%)0 (100%)
Mean disease duration (years)3.5 ± 1.4 8.2 ± 2.4None
MTX dose (mg/week)11.215 None
MTX (months use, mean)13 ± 1.922.6 ± 1.3 None
Infliximab dose (mg/kg/dose)
Infliximab use ([6 months) n, (%)
DAS28 (mean ± SEM)
5.6 ± 0.33.5 ± 0.5 None
340P. B. Donate et al.
Microarray data analysis
Microarray images were quantified using Spotfinder soft-
ware (http://www.tm4.org/spotfinder.html) and normalized
using the R platform (http://www.r-project.org). Statistical
analyses were performed using MeV software, version 3.1
(http://www.tm4.org/mev.html). Differentially expressed
genes were identified using the significance analysis of
microarrays (SAM) program , considering only those
genes with an FDR (false discovery rate) B0.05.
To analyze the gene expression profiles from the SAM
program data set, we used a hierarchical clustering method
that grouped genes on the vertical axis and samples on the
horizontal axis using similarity in their expression patterns.
The similarities and dissimilarities in gene expression were
presented as dendrograms, in which the pattern and length
of the branches reflected the relatedness of the samples
or genes, and as heat maps (http://rana.lbl.gov/Eisen
Real-time PCR was performed using a 7500 Real Time
PCR system (Applied Biosystems). The cDNA was syn-
thesized by oligo(dT) priming starting from 2.0 lg of
lymph node total RNA as previously described, amplified
using specific primers, and normalized to the amount of the
housekeeping gene glyceraldehyde-3-phosphate dehydro-
genase (Gapdh, ID: 008084). Reactions were performed in
a final volume of 20 lL containing 10 pmol of each pri-
mer, 19 SYBR Green (Applied Biosystems) and 1 lL of
first-strand cDNA. The specificity of the PCR products was
assessed by melting curve analysis for all samples, and the
products were also verified for correct size by agarose gel
electrophoresis. Each sample was assayed in triplicate, and
the mean Ctvalues were transformed into relative Amjhe-
like transcript quantities using the comparative Ctmethod
(Applied Biosystems, user bulletin #2). The primers were
identified using PRIMER3 software (http://frodo.wi.mit.
sequences retrieved from GenBank (http://www.ncbi.
nlm.nih.gov) for each gene. The following primers were
used for mice: Gapdh, ID: 008084.2, sense 50-GGGTGTG
ATGCCAAAGTT-30; preproET-1, ID: 010104.2, sense 50-
TGTGTCTACTTCTGCCACCT-30, antisense 50-CACCA
GCTGCTGATAGATAC-30; Mapk1, ID: 011949.3, sense
50-CTTCCAACCTCCTGCTGAAC-30, antisense 50-TGG
AGCTCTGTACCAACGTG-30; Ccr5, ID: 009917.3, sense
50-TTCCGAAAACACATGGTCAA-30, antisense 50-GTT
CTCCTGTGGATCGGGTA-30; Ccr2, ID: 009915.1, sense
50-ACTGGATGAAGCAGGGACAG-30, antisense 50-CCA
AAAATAAGGCGTTGGAA-30; and Ltb, ID: 008518.1,
sense 50-TATCACTGTCCTGGCTGTGC-30, antisense 50-
as follows: GAPDH, ID: 002046.3, sense 50-CTGCACC
ACCAACTGCTTA-30, antisense 50-CATGACGGCAGG
TCAGGTC-30; PREPRO ET, ID: 001955.3, sense 50-TCG
TTTTCCTTTGGGTTCAG-30, antisense 50-CTGTTTCT
GGAGCTCCTTGG-30; PRO-IL-1b, ID: 000002.11 sense
50-GCAATGAGGATGACTTGTTTTG-30, antisense 50-CA
GAGGTCCAGGTCCTGGAA-30; ETA, ID: 000004.11
000013.10 sense 50-TCATCGGGAACTCCACACTT-30,
Statistical analysis for real-time PCR and hyperalgesia
The results are presented as means ± SEM. The differ-
ences were evaluated by one-way ANOVA followed by
Bonferroni’s t test (three or more groups) or Student’s t test
(two groups). P\0.05, P\0.01, and P\0.001 were
considered statistically significant.
Effects of bosentan treatment on collagen-induced
The increasing evidence of the role of ETs in the inflam-
matory process prompted us to evaluate the impact of
bosentan treatment in a well-established model of arthritis,
CIA. DBA/1J mice were immunized with type II collagen
and therapeutically treated with bosentan for 11 days
starting on the day that arthritis was clinically detectable.
Oral treatment with bosentan ameliorated arthritis as
determined by a reduction in the clinical score, the number
of affected paws, and mechanical hyperalgesia observed
during the course of the disease (Fig. 1a–c). A represen-
tative image of paws of vehicle- and bosentan-treated mice
is shown in Fig. 1d. Infiltration of granulocytes and
mononuclear cells into the inflamed joint and bone and
cartilage destruction were lower in mice treated with bos-
entan than in the CIA vehicle-treated group (Fig. 2a).
Bosentan-treated mice also presented significantly lower
histological scores for inflammation and tissue destruction
than the CIA vehicle-treated group (Fig. 2b; Table 2).
The reduction in the disease scores in bosentan-treated
mice was also associated with a reduction in joint tissue
levels of the pro-inflammatory cytokines IL-1b, TNF-a,
and IL-17 (Fig. 3a–c). Bosentan-treated mice also pre-
sented a reduction in neutrophil accumulation in joint
tissues compared with the CIA vehicle-treated group
Bosentan, an endothelin receptor antagonist, ameliorates collagen-induced arthritis 341
PreproET mRNA expression increase in lymph nodes
of arthritic mice
ETs are produced as preproET, which is cleaved by ET-
converting enzymes to form big-ET and the active peptide
. Therefore, we next analyzed the levels of the ET
precursor preproET-1 by quantitative PCR in the inguinal
lymph nodes of sham and immunized mice. As shown in
Fig. 4, the levels of preproET-1 were higher in the lymph
nodes of arthritic mice than in those of sham mice.
Bosentan treatment modulates gene expression
in the lymph nodes
To elucidate the possible molecular mechanism associated
with bosentan’s action and consequently the role of ETs in
the CIA model, microarray technology was used to identify
significant changes in gene expression during the treat-
ment. The inguinal lymph node mRNA expression of sham
and immunized mice was initially compared, and then we
compared immunized mice before and after treatment with
Although the expression pattern remained unchanged
among the groups for a majority of the 4,500 sequences
tested [presented d(i) & dE(i)], 508 genes were differen-
tially expressed between sham and immunized mice,
whereas 127 genes were found to be significantly modu-
lated between immunized and bosentan-treated mice. With
these data, clusters of repressed and induced genes were
The Cluster-Tree View program was used to acquire
hierarchical cluster analysis of results from the SAM pro-
gram, and a comparison of the hybridization signatures
showed variability among the three experimental groups.
Based on heat map analysis, it was possible to distinguish
the expression signature of the disease state and the treat-
ment (Online Resource 1, 2). Genes that were differentially
expressed in both heat maps were selected for further
analysis according to their major biological processes. The
genes were related to cell communication, cell cycle, cell
death, response to stress and catabolic processes, among
others. The genes involved with immune response are of
particular importance, and a more careful analysis allowed
a selection of those genes related to inflammatory pro-
cesses, aside from being strictly related to CIA/RA. Among
the genes analyzed, we found a group of important genes
that exhibited an expression profile of particular interest.
The Mapk1 (Mitogen-activated protein kinase 1), Ccr2
Fig. 1 Bosentan (dual antagonist of ETA/ETB receptors) treatment
ameliorates collagen-induced arthritis (CIA) in mice. DBA/1J mice
were immunized with collagen, and the appearance of arthritis was
monitored daily. Treatment with bosentan (100 mg kg-1once a day)
was started 1 day after CIA was clinically significant and lasted for a
total of 11 days. The clinical score (a), number of affected paws
(b) and hyperalgesia (c) were evaluated daily. (d) Representative
pictures of mice paws showing the visual aspect of CIA development
in vehicle- or bosentan-treated groups. The data are from one
experiment with eight mice per group (error bars, SEM). *P\0.05
compared with the CIA control group (Student’s t test)
342 P. B. Donate et al.
(Chemokine (C–C motif) receptor 2), Ccr5 (Chemokine
(C–C motif) receptor 5) and Ltb (Lymphotoxin B) genes are
involved in the arthritis process and are highly expressed
during CIA development. On the other hand, during
treatment with bosentan, their expressions were similar to
Real-time PCR confirmed the microarray data, showing
increased transcription levels during disease in immunized
mice and lower expression during bosentan treatment, with
levels similar to control values (Fig. 5a–d).
PreproET mRNA is increased in RA-patient-derived
The level of the ET precursor preproET was elevated in
PBMCs from patients under treatment with MTX com-
pared to healthy donors (Fig. 6a). On the other hand,
PBMCs from patients receiving infliximab showed similar
levels of preproET-1 compared with healthy donors
(Fig. 6a). These results suggest that TNF-a could be
involved in the upregulation of the ET system during RA.
In an attempt to investigate this hypothesis, we performed
in-vitro experiments using PBMCs from healthy donors
and RA patients. PBMCs were cultured with TNF-a
(100 ng/ml) for 2 h, followed by mRNA extraction. mRNA
expression of preproET and ETA and ETB receptors
increased in PBMCs from healthy donors and RA patients
after TNF-a (Fig. 6b). However, this effect was more
prominent in cells isolated from the RA patients (Fig. 6b).
Interestingly, TNF-a also upregulated IL-1b mRNA
expression in PBMCs from RA patients, but its expression
was no different from cells from healthy donors.
The pro-inflammatory properties of ETs were described
immediately after their discovery, and they explain why
these peptides contribute to the progression of a wide range
of diseases that present inflammatory components, such as
lupus erythematosus, systemic sclerosis and scleroderma
[32, 33]. Our group and others have shown that ETs might
participate in the pathophysiology of RA [14–17, 19].
Here, we extended these studies showing that bosentan,
which is a dual ET receptor antagonist and was recently
approved by the FDA for human use, ameliorated the
Fig. 2 Effect of bosentan
treatment on the
of CIA in mice. CIA mice were
treated with vehicle or bosentan
(100 mg kg-1once a day) for
11 days, and treatment started
1 day after CIA was clinically
detectable. After treatment,
tibio-tarsal joints were removed
and processed followed by
(a) hematoxylin and eosin
staining of joint sections.
(b) Histological scores are
based on HE section analyses.
Each symbol represents one
paw. The data are from one
experiment with six mice per
group (error bars, SEM).
**P\0.01 compared with the
CIA group treated with vehicle
(Student’s t test)
Bosentan, an endothelin receptor antagonist, ameliorates collagen-induced arthritis 343
clinical and pathological inflammatory signals of CIA.
Moreover, the bosentan treatment also prevented CIA-
induced changes in the expression of important genes that
encode key inflammatory mediators in the lymph nodes. In
addition, the expression of ET precursor, preproET-1, was
enhanced in the lymph nodes of CIA mice. From a clinical
perspective, we also observed that the levels of ET pre-
cursors were elevated in PBMCs from RA patients who had
undergone conventional treatment and were reduced by
treatment with anti-TNF therapy. In vitro, TNF-a was able
to enhance gene expression of the ET system (preproET-1,
ETA and ETB) in PBMCs from RA patients compared
with healthy donors.
Since their discovery, it has been clear that ETs are not
merely vasoconstrictors but multifunctional peptides with
cytokine-like activity, and they interfere with almost all
aspects of cell function. Clinical studies have found ele-
vated levels of ET-1 in synovial fluid and plasma of RA
patients [14–16]. The strong association between these
peptides and inflammation led us to investigate the efficacy
of bosentan in CIA, which is a well-established experi-
mental model for RA. CIA is characterized by acute and
chronic inflammation of the joints and resembles most
features of human RA, such as swelling, cartilage degra-
dation, pain and loss of joint function . Bosentan-treated
mice presented a reduction in the following inflammatory
parameters: edema, pain, joint movement, granulocyte and
mononuclear cell infiltration into the joint tissue, bone and
cartilage destruction, the production of pro-inflammatory
cytokines and neutrophil migration. These results are in
accordance with previous results that demonstrated that
ET-1 mediates inflammatory events of zymosan-induced
articular inflammation, including edema and neutrophil
infiltration . Moreover, ETs seem to be involved in pain
in antigen-induced arthritis in mice [22, 36].
RA is a complex disease that results from an uncon-
trolled attack of the adaptive and innate immune system on
joint structures. Although tissue and cells obtained from the
Table 2 Histological scores for collagen-induced arthritis (CIA)
# Exudate Granulocyte
Bone and cartilage
CIA #31212231 13
CIA #43202131 12
CIA #61302211 10
CIA #82303221 13
CIA #91303211 11
Mean ± SD 1.08 ± 0.90 1.9 ± 0.990.08 ± 0.28 1.66 ± 0.981.41 ± 0.791.41 ± 0.99 0.83 ± 0.388.5 ± 4.3
Bosentan #71202221 10
Bosentan #10 11011105
Bosentan #11 11011105
Bosentan #12 11012117
Mean ± SD 0.75 ± 0.45 1.08 ± 0.7901.08 ± 0.79 1.0 ± 0.850.75 ± 0.620.33 ± 0.495.0 ± 3.4
344P. B. Donate et al.
inflamed joint might more accurately reflect the ongoing
pathogenetic features of chronic diseases, lymph nodes and
circulating immune cells are used to identify novel disease
mediators, treatment responses and genetic disease variants
[37, 38]. The use of large-scale analyses of gene expression
patterns is increasing in many fields, including rheumatol-
interest in the genetic study of heterogeneous autoimmune
possible molecular mechanisms behind the role of ETs in
CIA, the microarray technique was used to discover differ-
ences in lymph node gene profiles during CIA and bosentan
treatment. Gene expression in the lymph nodes of arthritic
mice was analyzed 11 days after bosentan treatment started
most prominent. In our study, 4,500 sequences were ana-
508 genes was found between sham and immunized mice,
and a set of 127 genes was found between immunized mice
treated with vehicle and those treated with bosentan. More-
over, a hierarchical cluster analysis of the results from SAM
using the Cluster-Tree View program showed that the three
experimental groups analyzed (sham, CIA vehicle- or bos-
entan-treated) had distinct profiles.
There were four main genes involved in the immune
response whose expressions were increased during CIA
and repressed by bosentan treatment. One gene was Mito-
gen-activated protein kinase 1 (Mapk1), a signaling
molecule that regulates the production of cytokines and
cytotoxic enzymes that have been implicated in RA path-
ogenesis. In agreement with our results, Mapk1 is induced
through ETA and ETB receptors, and it mediates important
peripheral functions of ETs, including DNA synthesis,
mitogenesis, and the activation of vascular smooth muscle
[40, 41]. Chemokine (C–C motif) receptor 2 (Ccr2) and
chemokine (C–C motif) receptor 5 (Ccr5), which along
Fig. 3 Effect of bosentan
treatment on the levels of pro-
inflammatory cytokines in joint
tissues of arthritic mice. CIA
mice were treated with vehicle
or bosentan (100 mg kg-1once
a day) for 11 days. After the
treatment period, the mice were
killed, the surrounding tissue of
the tibio-tarsal joint was
removed and the concentrations
of IL- 1b (a), TNF-a (b), and
IL-17 (c) and MPO activity
(d) were determined. The levels
of these cytokines and MPO
activity in samples from sham-
immunized mice were used as a
control. The results are
presented as mean ± SEM from
one experiment with six mice
per group. *P\0.05 compared
with CIA group treated with
vehicle (Student’s t test)
Fig. 4 PreproET mRNA expression in lymph nodes during CIA.
Real-time PCR was used to compare the mRNA levels of preproET
genes in inguinal lymph nodes of sham and immunized mice during
CIA. Lymph nodes were removed 11 days after CIA was clinically
detected. The expression levels were normalized to Gapdh expression.
The results are presented as mean ± SEM from one experiment with
eight mice per group. *P\0.05 compared with sham-immunized
Bosentan, an endothelin receptor antagonist, ameliorates collagen-induced arthritis345
with their cognate ligands have been detected in the
synovial fluid of patients with RA and in inflamed joints in
CIA, were also affected by CIA and bosentan treatment
[42–44]. Another affected gene was Lymphotoxin (LT)
Beta, which is predominantly expressed in a variety of
immune cells [45, 46] and is highly expressed in the RA
synovial lymphoid follicles. In addition, treatment with
LT-BetaR-Ig prevents the induction of murine CIA [47,
48]. It is important to mention that the expression of these
genes described above was also evaluated by real-time
PCR (Fig. 3), which confirmed the participation and
modulation of these transcripts in the studied model, sug-
gesting that they might have importance in the cascade of
ET effects. Corroborating the participation of endogenous
ETs in the modulation of these transcripts in peripheral
lymph nodes, we detected an increased expression of
preproET during CIA (Fig. 4). Therefore, it seems that ETs
may have an immunomodulatory role by acting in the
lymph nodes during experimental arthritis. This effect
might contribute to the exacerbation of the immune
response in RA. In agreement with this hypothesis, there is
evidence that dendritic cells produce large amounts of
Fig. 5 Confirmation of repressed gene expression in lymph nodes
during bosentan treatment in CIA. Real-time PCR was used to
confirm the mRNA levels of Mapk1 (a), Ltb (b), Ccr5 (c) and Ccr2
(d) in inguinal lymph nodes from sham, immunized and bosentan-
treated DBA/1J mice. Lymph nodes were removed 11 days after CIA
was clinically detected. Expression levels were normalized to Gapdh
expression. The results are presented as mean ± SEM from one
experiment with six mice per group. The differences between groups
were evaluated by one-way ANOVA followed by Bonferroni’s t test.
*P\0.05 compared with sham-immunized mice;
pared with CIA mice treated with vehicle
Fig. 6 TNF-a induces the expression of ET system genes in PBMCs
of healthy individuals and RA patients. Real-time PCR was used to
detect the mRNA levels of (a) preproET in PBMCs from healthy
individuals and RA patients and (b) IL-1b, preproET, and ETA and
ETB receptors in PBMCs cultured with 100 ng/ml TNF-a or medium
for 2 h. The expression levels were normalized to GAPDH
expression. The results are expressed as mean ± SEM. The differ-
ences between responses were evaluated by one-way ANOVA
followed by Bonferroni’s t test (three or more groups). *P\0.05
treated patients group, respectively (a); *P\0.05 and **P\0.01
compared with control group (b)
#P\0.05 compared with healthy individuals and the MTX-
346 P. B. Donate et al.
ET-1 and significantly increase the expression of ET
receptors upon activation. Furthermore, selective blockade
of the ETA receptor significantly reduced expression of the
mature dendritic cell marker CD83, decreased the pro-
duction of IL-12, downregulated the ability of dendritic
cells to stimulate T cells, and promoted dendritic cell
Once the effectiveness of bosentan was demonstrated in
the CIA model, suggesting the participation of the ET
system in the pathophysiology of CIA, we then evaluated
the expression of the ET precursor in immune cells from
RA patients. Our results showed increased levels of prep-
roET-1 in PBMCs from RA patients who were under MTX
treatment compared with levels in cells from healthy
donors. With a combination of MTX and anti-TNF (inf-
liximab) therapy, preproET-1 gene expression level did not
differ from the expression in healthy donors. These results
suggest that upregulation of ET gene expression in PBMCs
from RA patients seems to be controlled by pro-inflam-
matory cytokines, specifically TNF-a. There is a large body
of evidence showing that TNF and other cytokines promote
ET gene expression in a variety of cell types [50–52]. On
the other hand, ETs can also contribute to the production of
pro-inflammatory mediators such TNF, IL-1, IL-6, IL-8,
monocyte chemotactic protein-1 and granulocyte/macro-
phage colony-stimulating factor [13, 52]. Aside from the
possible immunomodulatory role of ETs in RA, we could
not discount their local involvement in several inflamma-
tory events in the joints. Indeed, as mentioned above, the
levels of ET-1 were higher in RA patients than in osteo-
arthritic patients [15, 16]. Furthermore, an increased
expression of preproET in zymosan- and antigen-induced
articular inflammation was found in mice [22, 35]. Addi-
expressed in synovial tissue [23, 34]. Further supporting
the involvement of TNF-a in the upregulation of ET system
genes, we showed that TNF-a increased the expression of
preproET-1, ETA and ETB in PBMCs from healthy donors
and RA patients. TNF-a also increased the expression of
IL-1b mRNA in PBMCs. Interestingly, the effect of TNF-a
on the ET system genes was more prominent in cells from
RA patients than in cells from healthy donors; however,
this effect was not observed for IL-1b expression, sug-
gesting a specific effect of TNF-a on the ET system. The
intracellular mechanism by which TNF-a upregulates the
gene expression of ETs in cells from RA patients was not
addressed. Further studies are required to clarify the reg-
ulation of the TNF-a/ET axis, and a higher number of
patients is necessary to achieve a definitive conclusion.
Nevertheless, these results might provide an important
novel mechanism by which anti-TNF-a therapy amelio-
rates inflammation via downregulation of ET system gene
expression in leukocytes in RA.
The results from this investigation contribute to a better
understanding of the mechanisms involved in CIA and the
participation of ETs in CIA. These results could lead to
future associated therapies for RA, including the use of ET
dac ¸a ˜o de Amparo a ` Pesquisa do Estado de Sa ˜o Paulo (FAPESP) and
Conselho Nacional de Desenvolvimento Cientı ´fico e Tecnolo ´gico
(CNPq) (Sa ˜o Paulo, Brazil). qRT-PCRs were carried out in Dr. Zila ´ L.
P. Simo ˜es’s Laboratory, University of Sa ˜o Paulo at Ribeira ˜o Preto.
We thank Dr. Catherine Nguyen, from the Institut National de la
Sante ´ et de la Recherche Me ´dicale (INSERM, Marseille, France) for
the cDNA clones used in the preparation of the microarrays and
acknowledge the excellent technical assistance of Ieda Regina dos
Santos Schivo, Se ´rgio Roberto Rosa, Fabiola Mestriner, Giuliana
Bertozi Francisco and Vani Maria Alves.
This work was supported by grants from Fun-
1. Arend WP. Physiology of cytokine pathways in rheumatoid
arthritis. Arthritis Rheum. 2001;45:101–6.
2. Yamamura YR, Gupta Y, Morita X, He R, Pai J, Endres A, et al.
Effector function of resting T cells: activation of synovial fibro-
blasts. J Immunol. 2001;166:2270–5.
3. Feldmann M, Brennan FM, Maini RN. Role of cytokines in
rheumatoid arthritis. Annu Rev Immunol. 1996;14:397–440.
4. Lajas C, Abasolo L, Bellajdel B. Costs and predictors of costs in
rheumatoid arthritis: a prevalence-based study. Arthritis Rheum.
5. Christopher JE. Immunological therapies for rheumatoid arthritis.
Br Med Bull. 2005;73 and 74:71–82.
6. Feldmann M, Maini RN. Lasker clinical medical research award:
TNF defined as a therapeutic target for rheumatoid arthritis and
other autoimmune diseases. Nat Med. 2003;9:1245–50.
7. Yanagisawa M, Kurihara H, Kimura S, Goto K, Masaki T. A
novel peptide vasoconstrictor, endothelin, is produced by vas-
cular endothelium and modulates smooth muscle Ca2channels.
J Hypertens. 1988;6:188–91.
8. Rubanyi GM, Polokoff MA. Endothelins: molecular biology,
biochemistry, pharmacology, physiology, and pathophysiology.
Pharmacol Rev. 1988;46:325–415.
9. Dhaun N, Pollock DM, Goddard J, Webb DJ. Selective and
mixed endothelin receptor antagonism in cardiovascular disease.
Trends Pharmacol Sci. 2007;28:573–9.
10. Inoue A, Yanagisawa M, Kimura S. The human endothelin
family: three structurally and pharmacologically distinct isopep-
tides predicted by three separate genes. Proc Natl Acad Sci USA.
11. Webb DJ, Monge JC, Rabelink TJ, Yanagisawa M. Endothelin:
new discoveries and rapid progress in the clinic. Trends Phar-
macol Sci. 1998;19:5–8.
12. Ferreira SH, Romitelli M, de Nucci G. Endothelin -1 participation
in overt and inflammatory pain. J Cardiovasc Pharmocol. 1989;
13. Rae GA, Henriques MG. Endothelins in inflammation: pro-
inflammatory and anti-inflammatory peptides. Marcel Dekker:
New York; 1998. p. 163–202.
14. Haq A, El-Ramahi K, Al-Dalaan A, Al-Sedairy ST. Serum and
synovial fluid concentrations of endothelin-1 in patients with
rheumatoid arthritis. J Med. 1999;30(1–2):51–60.
Bosentan, an endothelin receptor antagonist, ameliorates collagen-induced arthritis347
15. Miyasaka N, Hirata Y, Ando K, Sato K, Morita H, Shichiri M, Download full-text
et al. Increased production of endothelin-1 in patients with
inflammatory arthritis. Arthritis Rheum. 1992;35:397–400.
16. Yoshida H, Imafuku Y, Ohhara M, Miyata M, Kasukawa R,
Ohsumi K, et al. Endothelin-1 production by human synovio-
cytes. Ann Clin Biochem. 1998;35(2):290–4.
17. Wharton J, Rutherford RA, Walsh DA, Mapp PI, Knock GA,
Blake DR, et al. Autoradiographic localization and analysis of
endothelin-1 binding sites in human synovial tissue. Arthritis
18. Marjolaine RB, Johanne MP, Jean-Pierre P, Khatija NM, Stephan
C, Fazool S, et al. Endothelin 1 promotes osteoarthritic cartilage
degradation via matrix metalloprotease 1 and matrix metallo-
protease 13 induction. Arthritis Rheum. 2003;48(10):2855–64.
19. Imhof AK, Gluck L, Gajda M, Brauer R, Schaible H, Schulz S.
Potent antiinflammatory and antinociceptive activity of the
endothelin receptor antagonist bosentan in monoarthritic mice.
Arthritis Res Ther. 2001;13(3):R97.
20. Leung BP, Sattar N, Crilly A, Prach M, McCarey DW, Payne H,
et al. A novel anti-inflammatory role for simvastatin in inflam-
matory arthritis. J Immunol. 2003;170:1524–30.
21. Lagares D, Garcı ´a-Ferna ´ndez RA, Jime ´nez CL, Maga ´n-Marchal
N, Busnadiego O, Lamas S, et al. Endothelin 1 contributes to the
effect of transforming growth factor b1 on wound repair and skin
fibrosis. Arthritis Rheum. 2010;62(3):878–89.
22. Verri WA Jr, Cunha T, Magro DA, Guerrero ATG, Vieira SM,
Carregaro V, et al. Targeting endothelin ETA and ETB receptors
inhibits antigen-induced neutrophil migration and mechanical
hypernociception in mice. Naunyn-Schmiedeberg’s Arch Phar-
23. Verri WA Jr, Guerrero AT, Fukada SY, Valerio DA, Cunha TM,
Xu D, et al. IL-33 mediates antigen-induced cutaneous and
articular hypernociception in mice. Proc Natl Acad Sci USA.
24. Cunha TM, Verri WA Jr, Poole S, Parada CA, Cunha FQ,
Ferreira SH. Pain facilitation by proinflammatory cytokine
actions at peripheral nerve terminals. In: DeLeo J, Sorkin L,
Watkins L, editors. Immune and glial regulation of pain. Seattle:
IASP; 2007. p. 67–83.
25. Cunha TM, Verri WA Jr, Vivancos GG, Moreira IF, Reis S,
Parada CA. An electronic pressure-meter nociception paw test for
mice. Braz J Med Biol Res. 2004;37:401–7.
26. Cunha TM, Verri WA Jr, Silva JS, Poole S, Cunha FQ, Ferreira
SH. A cascade ofcytokines mediates mechanical inflammatory
hypernociception in mice. Proc Natl Acad Sci USA. 2005;
27. 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.
28. 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
arthritis. Arthritis Rheum. 1995;38:44–8.
29. Hegde P, Qi R, Abernathy K, Gay C, Dharap S, Gaspard R, et al.
A concise guide to cDNA microarray analysis. Biotechniques.
30. Tusher VG, Tibshirani R, Chu G. Significance analysis of
microarrays applied to the ionizing radiation response. Proc Natl
Acad Sci USA. 2001;98(9):5116–21.
31. Eisen MB, Spellmam PT, Brown PO, Botstein D. Cluster analysis
and display of genome-wide expression patterns. Proc Natl Acad
Sci USA. 1998;95:14863–8.
32. Denton CP. Therapeutic targets in systemic sclerosis. Arthritis
Res Ther. 2007;9(2):S6.
33. Mayes MD. Endothelin and endothelin receptors antagonists in
systemic rheumatic disease. Arthritis Rheum. 2003;48:1190–9.
34. Bendele A. Animal models of rheumatoid arthritis. J Musculosk-
elet Neuronal Interact. 2001;1:377–85.
35. Conte FP, Barja-Fidalgo C, Verri WA Jr, Cunha FQ, Rae GA,
Penido C, et al. Endothelins modulate inflammatory reaction in
Zymosan induced arthritis: participation of LTB4, TNFa, and
CXCL-1. J Leukocyte Biol. 2008;84:652–60.
36. Pinto LG, Cunha TM, Vieira SM, Lemos HP, WA VerriJ, Cunha
FQ, et al. IL-17 mediates articular hypernociception in antigen-
induced arthritis in mice. Pain. 2010;148(2):247–56.
37. Shaffer AL, Wright G, Yang L, Powell J, Ngo V, Lamy L, Lam
LT, et al. A library of gene expression signatures to illuminate
normal and pathological lymphoid biology. Immunol Rev.
38. Yamagata T, Benoist C, Mathis DA. Shared gene-expression
signature in innate-like lymphocytes. Immunol Rev. 2006;210:
39. Lequerre T, Coulouarn C, Derambure C, Lefebvre G, Vittecoq O,
Daveau M. A new tool for rheumatology: large-scale analysis of
gene expression. Joint Bone Spine. 2003;70:248–56.
40. Barman SA, Marrero MB. Mechanism of endothelin-1 activation
of map kinases in neonatal pulmonary vascular smooth muscle.
41. Khodorova A, Montmayeur JP, Strichartz G. Endothelin recep-
tors and pain. J Pain. 2009;10:4–28.
42. Patel DD, Zachariah JP, Whichard LP. CXCR3 and CCR5
ligands in rheumatoid arthritis synovium. Clin Immunol. 2001;
43. Robinson E, Keystone EC, Schall TJ, Gillett N, Fish EN. Che-
mokine expression in rheumatoid arthritis (RA): evidence of
RANTES and macrophage inflammatory protein (MIP)-1 beta
production by synovial T cells. Clin Exp Immunol. 1995;101:
44. Thornton S, Duwel LE, Boivin GP, Ma Y, Hirsch R. Association
of the course of collagen-induced arthritis with distinct patterns
of cytokine and chemokine messenger RNA expression. Arthritis
45. Agyekum S, Church A, Sohail M, Krausz T, Van Noorden S,
Polak J. Expression of lymphotoxin-beta (LT-Beta) in chronic
inflammatory conditions. J Pathol. 2002;199:115–21.
46. Lu G, Janjic BM, Janjic J, Whiteside TL, Storkus WJ, Vujanovic
NL. Innate direct anticancer effector function of human immature
dendritic cells. II. Role of TNF, lymphotoxin-alpha(1)beta(2), Fas
ligand, and TNF-related apoptosis-inducing ligand. J Immunol.
47. Rennert PD, James D, Mackay F, Browning JL, Hochman PS.
Lymph node genesis is induced by signaling through the lym-
photoxin beta receptor. Immunity. 1998;9:71–9.
48. Takemura S, Braun A, Crowson C, Kurtin PJ, CoWeld RH,
O‘Fallon WM. Lymphoid neogenesis in rheumatoid synovitis.
J Immunol. 2001;167:1072–80.
49. Guruli G, Pflug BR, Pecher S, Makarenkova V, Shurin MR,
Nelson JB. Function and survival of dendritic cells depend on
endothelin-1 and endothelin receptor autocrine loops. Blood.
50. Kurihara H, Yoshizumi M, Sugiyama T, Takaku F, Yanagisawa
M. Transforming growth factor-beta stimulates the expression of
endothelin mRNA by vascular endothelial cells. Biochem Bio-
phys Res Commun. 1989;159:1435–40.
51. Lattmann T, Hein M, Horber S, Ortmann J. Activation of pro-
inflammatory and anti-inflammatory cytokines in host organs
during chronic allograft rejection: role of endothelin receptor
signaling. Am J Transplant. 2005;5:1042–9.
52. Marsden PA, Brenner BM. Regulation of ET-1 gene by TNF-a.
Am J Physiol. 1992;262:C854–61.
348 P. B. Donate et al.