Download Inhibition of cyclooxygenase-2 improves cardiac - AJP

Am J Physiol Heart Circ Physiol 286: H1416–H1424, 2004;
10.1152/ajpheart.00136.2003.
Inhibition of cyclooxygenase-2 improves cardiac function after myocardial
infarction in the mouse
Margot C. LaPointe, Mariela Mendez, Alicia Leung, Zhenyin Tao, and Xiao-Ping Yang
Hypertension and Vascular Research Division, Henry Ford Hospital, Detroit, Michigan 48202
Submitted 13 February 2003; accepted in final form 5 December 2003
echocardiography; NS-398; collagen; nonsteroidal anti-inflammatory
drugs; prostanoids
CYCLOOXYGENASE (COX), also known as prostaglandin endoperoxide H synthase, exists as two isoforms, COX-1 and
COX-2 (43). COX-1 is expressed in almost all tissues, and its
prostanoid products are thought to mediate physiological responses such as vascular homeostasis and protection of the
gastric mucosa. Although COX-2 products are generally considered inflammatory mediators and thus deleterious, they can
exhibit both beneficial and deleterious effects in the same
tissue, for example, the kidney (14, 44). In addition to inflammation, COX-2 products have been associated with regulation
of cell proliferation, in particular colon cancer (12, 45). Pros-
Address for reprint requests and other correspondence: M. C. LaPointe,
Hypertension and Vascular Research Division, Henry Ford Hospital, 2799
West Grand Blvd., Detroit, MI 48202-2689 (E-mail: [email protected]).
H1416
tanoids produced by COX-2 affect a number of other biological
processes, including angiogenesis, differentiation, and apoptosis (24, 46). HEK-293 cells overexpressing both COX-2 and
PGE2 synthase release large quantities of PGE2, grow faster
than normal cells, and also exhibit altered morphology (30). In
addition, COX-2 has been detected in the heart in patients with
congestive heart failure and during allograft rejection (47, 50).
However, we know very little about its regulation in the heart
or its contribution to cardiac structure and function.
COX-1 and -2 catalyze conversion of arachidonic acid to the
prostaglandin endoperoxide PGH2 by similar mechanisms. Arachidonic acid released by proinflammatory stimuli tends to be
preferentially utilized by COX-2, and when COX-2 and COX-1
are functional in the same cell, COX-2 can form up to four times
as much PGH2 as COX-1 (40). PGH2 is converted to a variety of
prostanoids including PGE2, PGI2, PGF2␣, PGD2, and thromboxane A2 by specific synthases (12, 46), and a specific synthase
seems to be functionally coupled to either COX-1 or -2 within the
cell to produce a particular prostanoid. For example, COX-2 and
inducible PGE2 synthase (PGES) are coupled to generate PGE2 in
the inflammatory response (30). We have previously shown that
the cytokine IL-1␤ induces both COX-2 and PGES in neonatal
ventricular myocytes and that PGE2 and PGF2␣, but not a PGI2
analog, stimulate hypertrophic growth of myocytes (27).
In contrast to hormones, which have broad systemic effects,
prostanoids serve as intracrine, autocrine, and paracrine mediators, signaling changes within the immediate environment.
Upregulation of COX in the heart could produce prostanoid
products that would affect cell function. Previous studies
examining the role of COX in the heart used nonselective COX
inhibitors, and these studies produced variable results (4, 6, 18,
29). In recent studies in rodents, COX-2 selective inhibitors
were given before myocardial infarction (MI) and resulted in
decreased fibroblast proliferation and improved cardiac function (36, 38). However, the studies did not distinguish between the
effects of COX-2 inhibitors on wound healing and scar formation
as well as subsequent remodeling. In an attempt to clarify some of
the processes regulated by COX-2 in the heart, we studied the
effects of administering a COX-2 inhibitor 48 h post-MI and for
an additional 2 wk on cardiac function, fibrosis, and hypertrophy.
METHODS
Animal Procedures
Male C57BL/6J mice from Jackson Laboratories were housed in a
temperature-controlled room with a 12:12-h light-dark cycle and
given standard chow and tap water. They were housed for at least 1
wk before the start of the experiment, which was approved by the
The costs of publication of this article were defrayed in part by the payment
of page charges. The article must therefore be hereby marked “advertisement”
in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
0363-6135/04 $5.00 Copyright © 2004 the American Physiological Society
http://www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
LaPointe, Margot C., Mariela Mendez, Alicia Leung, Zhenyin
Tao, and Xiao-Ping Yang. Inhibition of cyclooxygenase-2 improves
cardiac function after myocardial infarction in the mouse. Am J
Physiol Heart Circ Physiol 286: H1416–H1424, 2004; 10.1152/
ajpheart.00136.2003.—Cyclooxygenase (COX)-2 is expressed in the
heart in animal models of ischemic injury. Recent studies have
suggested that COX-2 products are involved in inflammatory cell
infiltration and fibroblast proliferation in the heart. Using a mouse
model, we questioned whether 1) myocardial infarction (MI) in vivo
induces COX-2 expression chronically, and 2) COX-2 inhibition
reduces collagen content and improves cardiac function in mice with
MI. MI was produced by ligation of the left anterior descending
coronary artery in mice. Two days later, mice were treated with 3
mg/kg NS-398, a selective COX-2 inhibitor, or vehicle in drinking
water for 2 wk. After the treatment period, mice were subjected to
two-dimensional M-mode echocardiography to determine cardiac
function. Hearts were then analyzed for determination of infarct size,
interstitial collagen content, brain natriuretic peptide (BNP) mRNA,
myocyte cross-sectional area, and immunohistochemical staining for
transforming growth factor (TGF)-␤ and COX-2. COX-2 protein,
detected by immunohistochemistry, was increased in MI versus sham
hearts. MI resulted in increased left ventricular systolic and diastolic
dimension and decreased ejection fraction, fractional shortening, and
cardiac output. NS-398 treatment partly reversed these detrimental
changes. Myocyte cross-sectional area, a measure of hypertrophy, was
decreased by 30% in the NS-398 versus vehicle group, but there was
no effect on BNP mRNA. The interstitial collagen fraction increased
from 5.4 ⫾ 0.4% in sham hearts to 10.4 ⫾ 0.9% in MI hearts and was
decreased to 7.9 ⫾ 0.6% in NS-398-treated hearts. A second COX-2
inhibitor, rofecoxib (MK-0966), also decreased myocyte cross-sectional
area and interstitial collagen fraction. TGF-␤, a key regulator of collagen
synthesis, was increased in MI hearts. NS-398 treatment reduced TGF-␤
immunostaining by 40%. NS-398 treatment had no effect on infarct size.
These results suggest that COX-2 products contribute to cardiac remodeling and functional deficits after MI. Thus selected inhibition of COX-2
may be a therapeutic target for reducing myocyte damage after MI.
COX-2 AND CARDIAC REMODELING
Henry Ford Hospital Institutional Animal Care and Use Committee.
Mice were ⬃12 wk of age (24–30 g) when they underwent surgery.
To induce MI, mice were anesthetized with pentobarbital sodium (50
mg/kg ip), and a left thoracotomy was performed. The left anterior
descending coronary artery (LAD) was ligated with an 8-0 silk suture
placed near its origin at the edge of the left atrium as described
previously (15). Ligation was deemed successful when the anterior
wall of the left ventricle (LV) turned pale. The ligature was positioned
so as to produce a moderate infarct (30–45%). Sham operation was
identical except that the ligature was placed and then removed without
tightening it. Mice were kept on a heating pad and monitored until
awake. The MI surgical procedure resulted in 12% mortality.
Experimental Protocols
AJP-Heart Circ Physiol • VOL
Echocardiography
Two-dimensional M-mode transthoracic echocardiography was
performed on conscious mice using an Acuson 256 system (Mountain
View, CA) with a 15-MHz linear transducer as reported previously
(15). M-mode images were used to determine LV end-diastolic
dimension (LVDd) and LV end-systolic dimension (LVDs). Diastolic
measurements were made at the maximum LV cavitary dimension,
whereas systolic parameters were measured during maximum anterior
motion of the posterior wall.
Infarct Size
IS was measured as described previously (51). For each heart,
6-␮m sections from each of the three slices of sections A, B, and D
were stained with Gomori trichome to identify fibrous tissue (infarction). IS was calculated as the ratio of infarct length to the circumference of both the endocardium and epicardium.
Measurement of ICF and MCSA
Slices from sections A, B, and D of the heart were double stained
with fluorescein-labeled peanut agglutinin to delineate the MCSA and
interstitial space and rhodamine-labeled Griffonia simplicifolia lectin
I to outline the capillaries (22). Four radially oriented microscope
fields were selected from each section, two adjacent to the infarct
(border zone) and two away from it, and photographed at a magnification of ⫻100. MCSA was measured by computer-based planimetry
(Jandel) and averaged across all four fields of the sections from the
three slices. To determine ICF, the total surface area (microscope
field), interstitial space (collagen plus capillaries), and the area occupied by capillaries alone were measured with videodensitometry
(SigmaScan, Jandel). ICF was calculated as the total surface area
occupied by the interstitial space minus the percentage of total surface
area occupied by the capillaries.
Detection of COX-2 and TGF-␤ by Immunostaining
Frozen sections (6 ␮m) of section B of the heart were fixed with
acetone for 10 min and rinsed in PBS. Tissue sections were preincubated with 0.3% hydrogen peroxide in PBS to inhibit endogenous
peroxidase activity. They were then washed twice in PBS for 5 min
each time, preincubated with blocking serum (5% normal serum) for
30 min, and then incubated with the primary antibody at 4°C overnight. For COX-2 immunostaining, sections were incubated with a
1:500 dilution of COX-2 antibody (M-19, Santa Cruz Biotechnology).
For detection of TGF-␤, a 1:500 dilution of monoclonal antibody was
used (MAB 1835, R&D Systems). After sections were washed with
PBS, species-appropriate biotinylated secondary antibodies were applied, followed by avidin-peroxidase complexes (Vector Laboratories). 3-Amino-9-ethyl carbazole was used as the substrate (Vector
Laboratories), and slides were counterstained with Gill’s hematoxylin
solution. For quantitation of TGF-␤ staining, two fields enclosing the
infarct border were photographed using a Nikon microscope attached
to a digital camera. The image was obtained using SPOT software
(version 3.4.2, Diagnostic Instruments) at ⫻400 magnification. The
area that stained positive for TGF-␤ was expressed as a percentage of
the total myocardial area. All slides were evaluated in blinded fashion.
Analysis of Brain Natriuretic Peptide and COX-2 mRNA
by Real-Time PCR
Total RNA was extracted by homogenization of myocardial samples in Tri-Reagent (Molecular Research Center; Cincinnati, OH)
following the manufacturer’s instructions. For brain natriuretic peptide (BNP) mRNA, the section of the mouse heart homogenized
included a portion of the noninfarcted LV free wall and the septum
(section C from protocol 2). For COX-2 mRNA, the LV was divided
into infarcted and noninfarcted regions, which were extracted sepa-
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
Protocol 1: effect of NS-398 on cardiac PGE2. Mice were subjected
to MI to document an increase in cardiac COX-2 and PGE2 and to test
the ability of NS-398 (NS) to inhibit COX-2-derived PGE2. After
24 h, the LV was divided into infarcted and noninfarcted regions, and
RNA was isolated for real-time RT-PCR analysis of COX-2 mRNA.
A second group of mice was subjected to MI, and, after 24 h, hearts
were extracted for determination of PGE2 levels. To insure that NS
inhibited MI-induced PGE2, a third group of MI mice was treated
three times with 3 mg/kg NS (on the day before surgery, 2 h before
surgery, and 2 h before death). Twenty-four hours after surgery, hearts
were removed and extracted for PGE2 analysis.
Hearts were homogenized in 1 ml of methanol containing 10 ␮g/ml
indomethacin. The volume of methanol was adjusted to 4 ml, and
PGE2 was extracted by repeated vortexing over a 30-min period. After
centrifugation at 4,500 rpm at 4°C, the supernatant was purified
through PGE2 affinity columns according to the manufacturer’s instructions (Cayman Chemicals; Ann Arbor, MI). After the sample was
eluted from the column, the sample was dried in a Savant and then
resuspended in 0.4 ml assay buffer. One-half of the sample was
frozen, whereas the other half was analyzed for PGE2 with an
enzyme-linked immunoassay kit from Cayman Chemicals according
to the manufacturer’s protocol. Results were recorded as picograms
per sample. PGE2 (200 pg) was added to an affinity column to
measure recovery. Recovery of PGE2 from the columns averaged
⬎95%.
Protocol 2: effect of NS on cardiac structure and function. Mice
with MI were divided into two groups and started on 3
mg䡠kg⫺1 䡠day⫺1 NS (Cayman Chemicals) or vehicle in drinking water
48 h after MI and lasting for 2 wk. Sham-operated mice received tap
water. In the first 5–6 days of treatment, there was no apparent
difference in survival of vehicle-treated and NS-treated mice.
After the treatment period, mice were subjected to two-dimensional
M-mode echocardiography to determine cardiac function, and mice
were then euthanized. Hearts were removed and sectioned transversely into four slices from apex to base. Three of the sections
(sections A, B, and D) were frozen in isopentane and stored at ⫺70°C
for determination of infarct size (IS), interstitial collagen fraction
(ICF), myocyte cross-sectional area (MCSA), and immunohistochemical staining for transforming growth factor (TGF)-␤ and COX-2. The
fourth section (section C) was stored in RNALater (Ambion) at
⫺70°C for subsequent extraction of RNA. Section A contained mostly
infarcted tissue, and section B was adjacent to the infarct, whereas
sections C and D contained noninfarcted myocardium.
Protocol 3: effect of rofecoxib on MCSA and ICF. To verify the
histopathology studies described in protocol 2, we tested a second
COX-2 inhibitor, rofecoxib (MK-0966; kindly provided by Dr. Robert
Young, Merck Frosst, Quebec, Canada). The drug was administered
in chow [0.0075% rofecoxib (wt/wt) in chow ⫽ 15 mg䡠kg⫺1 䡠day⫺1]
beginning 2 days after MI and continuing for 2 wk. Plasma samples
from 11 mice fed this dose for 2 wk gave rofecoxib values of 109 ⫾
16 ng/ml, consistent with therapeutic levels in humans (values kindly
determined by Pauline Luk, Merck Frosst, Quebec, Canada).
H1417
H1418
COX-2 AND CARDIAC REMODELING
Statistical Analysis
Data were tested for normality and equality of variance and
adjusted appropriately. If the data were not normally distributed, then
the nonparametric Wilcoxon rank-sum method was used. The Satterthwaite adjustment was used for data with unequal variances. BNP
mRNA data were analyzed by the nonparametric Wilcoxon rank-sum
method. All other data are expressed as means ⫾ SE and were
analyzed by t-test or one-way ANOVA, with multiple pairwise comparisons made by the Student-Newman-Keuls method. TGF-␤ data
were analyzed by t-test with the Satterthwaite adjustment for unequal
variances. P ⬍ 0.05 was considered significant.
RESULTS
Effect of MI on COX-2 and PGE2 in the Heart
To test the acute effect of MI on COX-2 mRNA and PGE2
levels in the heart, we analyzed total RNA from the infarcted
and noninfarcted regions of the LV free wall 24 h post-MI.
COX-2 mRNA normalized to ␤-actin mRNA was increased
4.4-fold in the infarcted LV (Fig. 1A). Cardiac PGE2 levels
were increased 3.7-fold after MI (Fig. 1B).
Doses of NS ranging from 1 to 10 mg䡠kg⫺1 䡠day⫺1 have been
shown to be efficacious in several different models (13, 25, 37,
39). PGE2 in sham hearts was 6.9 pg/sample (n ⫽ 8), which
increased to 26.52 pg/sample (n ⫽ 12) in MI hearts. Treatment
of mice with 3 mg䡠kg⫺1 䡠day⫺1 NS reduced cardiac PGE2
levels to 12.5 pg/sample (n ⫽ 9) (Fig. 1B). As a control for the
PGE2 assay, we used indomethacin (10 mg/kg), a COX-1/
COX-2 inhibitor, and found that PGE2 was decreased to levels
below the lower limit of detection of the assay.
To study chronic induction of COX-2 in MI hearts, we
stained sections of hearts from a separate group of animals for
COX-2 immunoreactivity. Part of section B (adjacent to the
infarct) from sham and MI hearts was immunostained using a
polyclonal COX-2 antibody. Sections from sham hearts
showed no staining, but sections from the heart 16 days
post-MI showed very intense localized brown staining in the
myocytes (Fig. 2).
Effect of COX-2 Inhibition on Cardiac Function
We used two-dimensional M-mode echocardiography to
evaluate cardiac function in untreated and NS-treated mice.
Ejection fraction, shortening fraction, stroke volume, and cardiac output decreased in MI versus sham mice (Fig. 3). In all
cases, COX-2 inhibition partly reversed the detrimental effects
of MI. COX-2 inhibition had no effect on heart rate (data not
shown). MI also increased both LVDs and LVDd (Fig. 4, A and
B). COX-2 inhibition significantly reduced LVDs (P ⬍ 0.01)
and tended to reduce LVDd (P ⫽ 0.09, NS-treated MI mice
versus vehicle-treated MI mice).
Effect of COX-2 Inhibition on Hypertrophy and Fibrosis
Using the heart sections, we next measured MCSA as an
index of hypertrophy. MCSA increased 2-fold in MI versus
sham hearts, and COX-2 inhibition reduced this to 1.7-fold
(sham: 149 ␮m2, MI: 302 ␮m2, MI ⫹ NS: 257 ␮m2; P ⬍ 0.05,
MI vs. NS; Fig. 5A). Because increases in the expression of
Fig. 1. Analysis of cyclooxygenase (COX)-2
mRNA and cardiac PGE2. A: COX-2 mRNA
was evaluated in the infarcted (INF) and noninfarcted (non-INF) regions of the left ventricle (LV) and normalized to ␤-actin mRNA;
n ⫽ 4 for both bars. B: PGE2 was extracted
from sham or myocardial infarcted (MI) hearts
and expressed as picograms per sample; n ⫽ 8
for sham; n ⫽ 12 for MI; n ⫽ 9 for MI/NS-398
(NS); n ⫽ 7 for MI/indomethacin (Indo). ND,
not detectable. Statistical significance is indicated in the graph. The values for sham versus
MI/NS were not statistically different.
AJP-Heart Circ Physiol • VOL
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
rately. At the end of the procedure, samples were treated with DNAse
I from Ambion’s RNAqueous-4PCR kit. RT was performed using the
reagents of the Omniscript RT kit (Qiagen) with 2 ␮g total RNA, 1 ␮g
random primer (Invitrogen), and 10 U/␮l RNAsin (Promega) in a
20-␮l volume for 1 h at 37°C.
Real-time PCR was performed using the QuantiTect Probe PCR kit
(Qiagen) and a Roche LightCycler (version 3.5). Fluorescence resonance energy transfer probes and primers were designed by TIB
MolBiol (Adelphia, NJ). Data analysis was performed with LightCycler software (version 3.5.28). The reaction volume was 20 ␮l and
contained 2 ␮l cDNA, 0.5 ␮M sense and antisense primers, 0.2 ␮M
each of the LC-640 red-labeled and fluorescein-labeled probes, Taq
polymerase, and dNTPs for 45 cycles. For analysis of BNP (241 bp)
and ␤-actin (342 bp) cDNAs, PCR conditions were as follows: 0 s at
95°C for denaturation, 40 s at 58°C for annealing, and 40 s at 72°C for
extension. For COX-2 (448 bp), PCR conditions were as follows: 0 s
at 95°C for denaturation, 45 s at 58°C for annealing, and 90 s at 72°C
for extension. The sequences of the primers/probes were as follows:
BNP sense, 5⬘-ATGGATCTCCTGAAGGTGCTG-3⬘; BNP antisense,
5⬘-GTGCTGCCTTGAGACCGAA-3⬘; BNP fluorescein probe, 5⬘TGGCTAGGACTTCCCAGAGGATAGG-x-3⬘; BNP LC-640 probe,
5⬘-LC red 640-TGACCTCCCAGCGGTGACAGATAAA-p-3⬘; ␤-actin sense, 5⬘-ACCCACACTGTGCCCATCTA-3⬘; ␤-actin antisense,
5⬘-GCCACAGGATTCCATACCCA-3⬘; ␤-actin fluorescein probe,
5⬘-GCCACGCTCGGTCAGGATCTTCAT-x-3⬘; ␤-actin LC-705
probe, 5⬘-LC red 705-AGGTAGTCTGTCAGGTCCCGGCCA-p-3⬘;
COX-2 sense, 5⬘-GTGACTGTACCCGGACTGGAT-3⬘; COX-2
antisense, 5⬘-AAGTGCTGGGCAAAGAATGC-3⬘; COX-2 fluorescein probe, 5⬘-CTGGGAAGCCTTCTCCAACCTCTCC-x-3⬘; and
COX-2 LC-640 probe, 5⬘-ACTACACCAGGGCCCTTCCTCCC-p3⬘. Gene-specific standard curves were generated using a linearized
plasmid containing the cDNA of interest and serially diluting it to
generate concentrations ranging over two to four orders of magnitude.
The LightCycler software calculated the amount of cDNA in a given
sample (copies/␮l) compared with the standard curve. BNP and
COX-2 mRNA were normalized to ␤-actin mRNA.
H1419
COX-2 AND CARDIAC REMODELING
atrial natriuretic peptide and BNP genes are markers of hypertrophy, we used real-time PCR to quantitate changes in BNP
gene expression in sham, MI, and NS-treated mouse hearts.
There was a 3.6-fold increase in BNP mRNA in MI versus
sham hearts, but COX-2 inhibition had no significant effect on
BNP mRNA (Fig. 5B).
We measured changes in ICF in sections from sham, MI, and
MI ⫹ NS hearts as a measure of collagen in noninfarcted areas
of the heart. The ICF of sham hearts was 5.4% and increased
twofold to 10.4% in MI hearts (Fig. 6). In NS-treated hearts,
ICF was decreased by 50% to 7.9%, indicating decreased
collagen production (P ⬍ 0.05, MI vs. NS treated).
To confirm the MCSA and ICF data, studies were repeated
using a second COX-2 inhibitor, rofecoxib (MK-0966 or
Vioxx). Treatment with rofecoxib for 2 wk decreased MCSA
(Fig. 7A) and ICF (Fig. 7B) by 65% and 48%, respectively.
Additionally, if mice were treated for 4 wk beginning 2 wk
after MI, MCSA and ICF were similarly decreased (Fig. 7, C
and D).
Effect of COX-2 Inhibition on TGF-␤
After MI, there is a sequential infiltration of the heart by
inflammatory cells, such as neutrophils, macrophages, and
lymphocytes (49), which contribute to removal of necrotic
tissue, scar formation, and interstitial fibrosis. TGF-␤, which is
synthesized and released by macrophages and fibroblasts, results in fibroblast proliferation and collagen synthesis (for a
review, see Ref. 23). We examined slices of section B of the
heart (adjacent to the scar) for TGF-␤ (Fig. 8A) and quantified
the immunostained area as a percentage of the entire section
examined. TGF-␤ staining increased from 0.4% in sham hearts
Fig. 3. Effect of NS on cardiac function as measured by
two-dimensional (2-D) M-mode echocardiography. A:
ejection fraction (EF); B: shortening fraction (SF); C:
cardiac output (CO); D: stroke volume (SV). Sham (n ⫽
8) ⫽ sham operation; MI/Cont (n ⫽ 9) ⫽ vehicletreated MI mice; MI/NS (n ⫽ 8) ⫽ NS-treated MI mice.
AJP-Heart Circ Physiol • VOL
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
Fig. 2. Immunohistochemical detection of COX-2 in the mouse heart post-MI. A: absence of specific staining in the sham heart.
The reddish-brown staining in B demonstrates the presence of COX-2. Magnification ⫻400. The micrographs are representative
of results from 8 sections of sham and MI hearts.
H1420
COX-2 AND CARDIAC REMODELING
Fig. 4. Effect of NS on LV shape as measured
by 2-D M-mode echocardiography. A: LV systolic dimension (LVDs). B: LV diastolic dimension (LVDd) in sham (n ⫽ 8), MI/Cont (n ⫽ 9),
and MI/NS mice (n ⫽ 8).
Effect of COX-2 Inhibition on IS
To test whether the cardioprotective effects of COX-2 inhibition were due to a decrease in IS, we measured IS in tissue
sections from MI hearts treated with vehicle and MI hearts
treated with NS. We found that IS as a percentage of the entire
LV was unchanged by COX-2 inhibition [IS in the MI/vehicle
group ⫽ 40 ⫾ 3% (n ⫽ 9) and in the MI/NS group ⫽ 37 ⫾ 4%
(n ⫽ 8)]. The same was true for rofecoxib treatment [IS in the
MI/vehicle group ⫽ 37.1 ⫾ 1.5% (n ⫽ 12) and in the
MI/rofecoxib group ⫽ 37.8 ⫾ 1.6% (n ⫽ 12)].
DISCUSSION
As an early inflammatory response gene, COX-2 is induced
by a variety of stimuli. In the myocardium it is stimulated by
cardiac allograft rejection (50) and MI (38) as well as during
the development of heart failure (47). Our data show that 1) MI
induced COX-2 in the mouse heart, 2) COX-2 inhibition with
NS improved cardiac function, and 3) both NS and rofecoxib
reduced MCSA and interstitial collagen. These data thus implicate COX-2 products in cardiac dysfunction after MI. Induction of COX-2 by injury has also been reported in other
tissues. COX-2 expression was induced in the brain in different
models of injury (11, 17, 19, 31), and inhibition of COX-2 by
NS or gene knockout improved function after injury (17, 31).
In mice, LAD ligation causes ischemia of the LV free wall
and necrosis of infarcted tissue, followed by scar formation and
remodeling of the noninfarcted myocardium. In this mouse
model, LAD ligation results in rapid increases in LVDs and
LVDd and decreased ejection fraction and cardiac output (51).
We found that COX-2 inhibition with NS improved cardiac
function, as evidenced by increased ejection fraction, shortening fraction, cardiac output, and stroke volume versus untreated mice. It is likely that the increases in ejection fraction
and shortening fraction were due to improved cardiac contractility, because LVDd was not changed, whereas LVDs decreased significantly. Two other reports have focused on the
role of COX-2 in the heart post-MI, but in both studies the
COX-2 inhibitor was given before MI. Scheuren et al. (38)
treated female rats with rofecoxib one day before MI and then
for 4 days afterward. Their acute studies revealed decreased
infiltration of inflammatory cells and decreased fibroblast proliferation and invasion of the infarct border. No functional
studies were done, and no conclusions were drawn concerning
the potential beneficial or deleterious effects of the acute
reduction in inflammatory processes. In contrast, Saito et al.
(36) administered the COX-2 selective inhibitor 5,5-dimethyl3-(3-fluorophenyl)-4-(4-methylsulfonyl)phenyl-2 (5H)-furanone
to male rats 30 min before MI and continued treatment for 2
wk. After another 2 wk, hemodynamic studies indicated that
COX-2 inhibition reduced LV end-diastolic pressure and improved cardiac compliance, resulting in an overall increase in
cardiac function. This supports our finding of an improvement
in LVDs subsequent to NS treatment.
Our studies also indicate that COX-2 inhibition reduced
collagen deposition in the heart after MI. This beneficial
alteration in the remodeling process probably reduced cardiac
stiffness, thus accounting for the improved cardiac function.
The decrease in collagen likely resulted from decreased TGF-␤
release by either macrophages, myocytes, or fibroblasts, as we
detected less TGF-␤ immunostaining in NS-treated hearts. Our
result is consistent with reports using two different models of
renal injury in the rat. In these studies, chronic COX-2 inhibition
reduced the extracellular matrix, TGF-␤ production, glomerular
injury, and proteinuria, thus improving renal function (8, 44).
Fig. 5. Effect of NS on cardiac hypertrophy.
A: myocyte cross-sectional area (MCSA);
n ⫽ 8 sham, 9 MI/Cont, and 8 MI/NS mice.
B: brain natriuretic peptide (BNP) mRNA
(copies/␮l) as determined by real-time PCR
and normalized to ␤-actin in sham (n ⫽ 6),
MI/Cont (n ⫽ 6), and MI/NS mice (n ⫽ 8).
The difference between MI/NS and MI/vehicle for BNP mRNA was not statistically
significant as determined by the nonparametric Wilcoxon rank-sum method.
AJP-Heart Circ Physiol • VOL
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
to 1.2% in MI hearts and decreased by 50% to 0.7% in
response to COX-2 inhibition by NS (Fig. 8B).
COX-2 AND CARDIAC REMODELING
The decrease in collagen content in our study most likely
resulted from decreased proliferation of fibroblasts and decreased infiltration of inflammatory cells, which synthesize and
release cytokines and growth factors involved in extracellular
matrix formation. This was demonstrated by Scheuren et al.
(38), who noted decreased numbers of macrophages and fibroblasts around the infarct area of rofecoxib-treated rats. Preliminary data from our laboratory suggest that the PGE2 analog
sulprostone and PGF2␣ increase fibroblast proliferation in vitro
(M. Mendez and M. C. LaPointe, unpublished observations). In
addition to affecting fibroblast proliferation, prostanoids could
participate in the acute inflammatory response by upregulation
of the adhesion molecules necessary for infiltration and activation of neutrophils. The prostanoid PGF2␣ has been shown to
stimulate ICAM-1 expression in gingival fibroblasts (32).
COX-2 inhibitors have also been shown to decrease plasma
levels of lipid peroxides, reduce cytokine release and/or action,
and decrease production of superoxide anions by neutrophils
(2, 3, 41). Thus the effect of COX-2 inhibition in reducing
collagen production could result from a combination of direct
effects on cardiac prostanoid formation, which would influence
neutrophil and macrophage infiltration, fibroblast proliferation,
and TGF-␤ and collagen synthesis.
COX-2 inhibition also reduced cardiac hypertrophy, as determined by measurements of MCSA. It has been reported that
the nonsteroidal anti-inflammatory drug fenbufen prevented
cardiac hypertrophy induced by clenbuterol in rats (35). Enhanced COX-2 expression in the heart would increase prostanoid production. We have previously shown that PGE2 acting
through the EP receptor subtype EP4 stimulates hypertrophy of
cardiac myocytes in vitro (27). The presence of EP4 receptors
in the heart would allow locally produced PGE2 to induce
hypertrophy of myocytes. Other prostanoids, including PGF2␣
and thromboxane A2, can also stimulate hypertrophy (1, 20,
27) as well as stimulate BNP expression (M. C. LaPointe,
unpublished observations).
BNP mRNA, a marker of hypertrophy and LV dysfunction,
was induced by MI, but its levels were unchanged by COX-2
Fig. 7. Effect of 2- and 4-wk rofecoxib [MK-0966 (MK)] treatment on MCSA and ICF; n ⫽ 6 for each bar. A: MCSA after 2-wk
treatment; B: ICF after 2-wk treatment; C: MCSA after 4-wk treatment; D: ICF after 4-wk treatment. Statistical significance is noted
in the graph.
AJP-Heart Circ Physiol • VOL
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
Fig. 6. Effect of NS on interstitial collagen fraction (ICF) in sham (n ⫽ 8),
MI/Cont (n ⫽ 9), and MI/NS mice (n ⫽ 8). Statistical significance is noted in
the graph.
H1421
H1422
COX-2 AND CARDIAC REMODELING
inhibition. The uncoupling of BNP expression from regression
of hypertrophy has been previously reported. Using a model of
suprarenal aortic banding, Ogawa et al. (33) demonstrated that
when both hypertrophy and blood pressure were decreased by
high-dose ramipril (an angiotensin-converting enzyme inhibitor), LV BNP mRNA was reduced. However, when hypertrophy was regressed without decreased blood pressure by lowdose ramipril, LV BNP mRNA remained elevated. Given that
BNP decreases collagen synthesis and increases matrix metalloproteinase in fibroblasts in vitro (42), its continued synthesis
in vivo might counter collagen accumulation in the setting of
COX-2 inhibition.
We found that COX-2 inhibition with either NS or rofecoxib
(MK-0966) beginning 2 days post-MI had no effect on IS
measured 14 days later. Scheuren et al. (38) also found that
rofecoxib did not influence IS acutely in the rat. Although we
did not detect any changes in overall IS in our study, we cannot
exclude the possibility that COX-2 inhibition affected wound
repair by altering infarct expansion or infarct thinning. A
decrease in collagen in the scar per se could predispose the
heart to rupture. Future studies will examine COX-2 inhibition
in the chronic phases of remodeling so as to avoid potential
effects on wound healing.
There is some controversy regarding the role of COX-2 in
the heart post-MI. Recent studies have indicated that rofecoxib
may increase the risk of cardiovascular events in patients (5,
28). Cheng et al. (9) demonstrated that the vasodilator PGI2 is
critical to oppose the actions of platelet thromboxane A2, and
McAdam et al. (26) indicated that COX-2 is important for
systemic synthesis of PGI2. If COX-2 is important for production of PGI2 in the vessel wall, then COX-2 inhibition might
AJP-Heart Circ Physiol • VOL
favor thrombosis. On the other hand, several studies have
implicated COX-2 and its products in cardiovascular disease.
Chenevard et al. (7) found that the COX-2 inhibitor celecoxib
improved flow-mediated dilatation and decreased markers of
inflammation and oxidative stress in male patients with severe
coronary artery disease. COX-2 overexpression has also been
associated with atherosclerotic plaque instability, which would
lead to acute ischemic syndromes (10). Moreover, COX-2
expression is strongly induced during LPS-induced septicemia
in rats, and treatment with the COX-2 inhibitor rofecoxib
reversed the effects of LPS on systolic arterial pressure and
heart rate as well as liver damage (16). It has also been shown
that COX-2 is involved in the inflammatory process after MI.
COX-2 inhibition reduced macrophage infiltration into the
necrotic area and border zone of the infarcted rat heart as well
as fibroblast proliferation (38, 48). Thus additional studies will
be required to evaluate whether COX-2 inhibitors as a class
have prothrombotic properties under certain circumstances, or
if there are specific properties of the different COX-2 inhibitors
that render them prothrombotic, as well as determine the
different effects of COX-2 and prostanoid production on myocyte, fibroblast, and endothelial cell function. Because different
cell types in the heart, e.g., myocytes, fibroblasts, and endothelial cells, produce different prostanoid profiles (21, 27, 34),
the effect of induction of COX-2 will depend on the cell in
which it is expressed, the prostanoids produced, and the ability
of surrounding cells to respond to the prostanoids.
In summary, our findings suggest that COX-2 plays a deleterious role in the heart after MI caused by chronic LAD
occlusion. MI induces COX-2, with concomitant increases in
PGE2, collagen content, and hypertrophy, and decreases in
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
Fig. 8. Effect of NS on transforming growth factor (TGF)-␤ immunostaining. A: sham heart (a), peri-infarct region of MI heart (b),
and peri-infarct region of MI ⫹NS heart (c). The reddish-brown staining demonstrates the presence of TGF-␤. Magnification ⫽
⫻400. B: quantification of the effect of NS on TGF-␤ immunostaining in sham (n ⫽ 6), MI/Cont (n ⫽ 9), and MI/NS hearts (n ⫽
7). The difference between MI/NS and MI/vehicle was statistically significant (P ⬍ 0.05).
COX-2 AND CARDIAC REMODELING
cardiac function. COX-2 inhibition partly reverses these effects. Thus selective and timely use of COX-2 inhibitors may
improve cardiac function after ischemic injury.
ACKNOWLEDGMENTS
The authors thank Dr. Yun-He Liu, Lisa Li, and Nicole Bart for assistance
with morphometry.
GRANTS
This study was supported by National Heart, Lung, and Blood Institute
Grant P01 HL-28982.
REFERENCES
AJP-Heart Circ Physiol • VOL
17. Iadecola C, Niwa K, Nogawa S, Zhao X, Nagayama M, Araki E,
Morham S, and Ross ME. Reduced susceptibility to ischemic brain
injury and N-methyl-D-aspartate-mediated neurotoxicity in cyclooxygenase-2-deficient mice. Proc Natl Acad Sci USA 98: 1294–1299, 2001.
18. Jugdutt BI, Hutchins GM, Bulkley BH, and Becker LC. Salvage of
ischemic myocardium by ibuprofen during infarction in the conscious dog.
Am J Cardiol 46: 74–82, 1980.
19. Koistinaho J and Chan PH. Spreading depression-induced cyclooxygenase-2 expression in the cortex. Neurochem Res 25: 645–651, 2000.
20. Lai J, Jin H, Yang R, Winer J, Li W, Yen R, King KL, Zeigler F, Ko
A, Cheng J, Bunting S, and Paoni NF. Prostaglandin F2␣ induces cardiac
myocyte hypertrophy in vitro and cardiac growth in vivo. Am J Physiol
Heart Circ Physiol 271: H2197–H2208, 1996.
21. Linssen MC, Engels W, Lemmens PJ, Heijnen VV, Van Bilsen M,
Reneman RS, and van der Vusse GJ. Production of arachidonic acid
metabolites in adult rat cardiac myocytes, endothelial cells, and fibroblastlike cells. Am J Physiol Heart Circ Physiol 264: H973–H982, 1993.
22. Liu YH, Yang XP, Sharov VG, Nass O, Sabbah HN, Peterson E, and
Carretero OA. Effects of angiotensin-converting enzyme inhibitors and
angiotensin II type 1 receptor antagonists in rats with heart failure. Role of
kinins and angiotensin II type 2 receptors. J Clin Invest 99: 1926–1935,
1997.
23. Manabe I, Shindo T, and Nagai R. Gene expression in fibroblasts and
fibrosis: involvement in cardiac hypertrophy. Circ Res 91: 1103–1113,
2002.
24. Marnett LJ. Aspirin and the potential role of prostaglandins in colon
cancer. Cancer Res 52: 5575–5589, 1992.
25. Masferrer JL, Zweifel BS, Manning PT, Hauser SD, Leahy KM,
Smith WG, Isakson PC, and Seibert K. Selective inhibition of inducible
cyclooxygenase 2 in vivo is antiinflammatory and nonulcerogenic. Proc
Natl Acad Sci USA 91: 3228–3232, 1994.
26. McAdam BF, Catella-Lawson F, Mardini IA, Kapoor S, Lawson JA,
and FitzGerald GA. Systemic biosynthesis of prostacyclin by cyclooxygenase (COX)-2: the human pharmacology of a selective inhibitor of
COX-2. Proc Natl Acad Sci USA 96: 272–277, 1999.
27. Mendez M and LaPointe MC. Trophic effects of the cyclooxygenase-2
product prostaglandin E2 in cardiac myocytes. Hypertension 39: 382–388,
2002.
28. Mukherjee D, Nissen SE, and Topol EJ. Risk of cardiovascular events
associated with selective COX-2 inhibitors. JAMA 286: 954–959, 2001.
29. Mullane KM, Read N, Salmon JA, and Moncada S. Role of leukocytes
in acute myocardial infarction in anesthetized dogs: relationship to myocardial salvage by anti-inflammatory drugs. J Pharmacol Exp Ther 228:
510–522, 1984.
30. Murakami M, Naraba H, Tanioka T, Semmyo N, Nakatani Y, Kojima
F, Ikeda T, Fueki M, Ueno A, Oh S, and Kudo I. Regulation of
prostaglandin E2 biosynthesis by inducible membrane-associated prostaglandin E2 synthase that acts in concert with cyclooxygenase-2. J Biol
Chem 275: 32783–32792, 2000.
31. Nagayama M, Niwa K, Nagayama T, Ross ME, and Iadecola C. The
cyclooxygenase-2 inhibitor NS-398 ameliorates ischemic brain injury in
wild-type mice but not in mice with deletion of the inducible nitric oxide
synthase gene. J Cereb Blood Flow Metab 19: 1213–1219, 1999.
32. Noguchi K, Iwasaki K, and Ishikawa I. Prostaglandin F2 alpha upregulates intercellular adhesion molecule-1 expression in human gingival
fibroblasts. J Periodontal Res 34: 277–281, 1999.
33. Ogawa T, Linz W, Stevenson M, Bruneau BG, Kuroski de Bold ML,
Chen JH, Eid H, Scholkens BA, and de Bold AJ. Evidence for
load-dependent and load-independent determinants of cardiac natriuretic
peptide production. Circulation 93: 2059–2067, 1996.
34. Oudot F, Grynberg A, and Sergiel JP. Eicosanoid synthesis in cardiomyocytes: influence of hypoxia, reoxygenation, and polyunsaturated fatty
acids. Am J Physiol Heart Circ Physiol 268: H308–H315, 1995.
35. Palmer RM, Delday MI, McMillan DN, Noble BS, Bain P, and Maltin
CA. Effects of the cyclo-oxygenase inhibitor, fenbufen, on clenbuterolinduced hypertrophy of cardiac and skeletal muscle of rats. Br J Pharmacol 101: 835–838, 1990.
36. Saito T, Rodger IW, Hu F, Shennib H, and Giaid A. Inhibition of
cyclooxygenase-2 improves cardiac function in myocardial infarction.
Biochem Biophys Res Commun 273: 772–775, 2000.
37. Sanchez PL, Salgado LM, Ferreri NR, and Escalante B. Effect of
cyclooxygenase-2 inhibition on renal function after renal ablation. Hypertension 34: 848–853, 1999.
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
1. Adams JW, Migita DS, Yu MK, Young R, Hellickson MS, CastroVargas FE, Domingo JD, Lee PH, Bui JS, and Henderson SA. Prostaglandin F2 alpha stimulates hypertrophic growth of cultured neonatal rat
ventricular myocytes. J Biol Chem 271: 1179–1186, 1996.
2. Agha AM, El Khatib AS, and Al Zuhair H. Modulation of oxidant status
by meloxicam in experimentally induced arthritis. Pharmacol Res 40:
385–392, 1999.
3. Bennett A and Villa G. Nimesulide: an NSAID that preferentially inhibits
COX-2, and has various unique pharmacological activities. Expert Opin
Pharmacother 1: 277–286, 2000.
4. Bolli R, Goldstein RE, Davenport N, and Epstein SE. Influence of
sulfinpyrazone and naproxen on infarct size in the dog. Am J Cardiol 47:
841–847, 1981.
5. Bombardier C, Laine L, Reicin A, Shapiro D, Burgos-Vargas R, Davis
B, Day R, Ferraz MB, Hawkey CJ, Hochberg MC, Kvien TK,
Schnitzer TJ, and the VIGOR Study Group. Comparison of upper
gastrointestinal toxicity of rofecoxib and naproxen in patients with rheumatoid arthritis. VIGOR Study Group. N Engl J Med 343: 1520–1530,
2000.
6. Bonow RO, Lipson LC, Sheehan FH, Capurro NL, Isner JM, Roberts
WC, Goldstein RE, and Epstein SE. Lack of effect of aspirin on
myocardial infarct size in the dog. Am J Cardiol 47: 258–264, 1981.
7. Chenevard R, Hurlimann D, Bechir M, Enseleit F, Spieker L, Hermann M, Riesen W, Gay S, Gay RE, Neidhart M, Michel B, Luscher
TF, Noll G, and Ruschitzka F. Selective COX-2 inhibition improves
endothelial function in coronary artery disease. Circulation 107: 405–409,
2003.
8. Cheng HF, Wang CJ, Moeckel GW, Zhang MZ, McKanna JA, and
Harris RC. Cyclooxygenase-2 inhibitor blocks expression of mediators of
renal injury in a model of diabetes and hypertension. Kidney Int 62:
929–939, 2002.
9. Cheng Y, Austin SC, Rocca B, Koller BH, Coffman TM, Grosser T,
Lawson JA, and FitzGerald GA. Role of prostacyclin in the cardiovascular response to thromboxane A2. Science 296: 539–541, 2002.
10. Cipollone F, Prontera C, Pini B, Marini M, Fazia M, De Cesare D,
Iezzi A, Ucchino S, Boccoli G, Saba V, Chiarelli F, Cuccurullo F, and
Mezzetti A. Overexpression of functionally coupled cyclooxygenase-2
and prostaglandin E synthase in symptomatic atherosclerotic plaques as a
basis of prostaglandin E2-dependent plaque instability. Circulation 104:
921–927, 2001.
11. Domoki F, Veltkamp R, Thrikawala N, Robins G, Bari F, Louis TM,
and Busija DW. Ischemia-reperfusion rapidly increases COX-2 expression in piglet cerebral arteries. Am J Physiol Heart Circ Physiol 277:
H1207–H1214, 1999.
12. Dubois RN, Abramson SB, Crofford L, Gupta RA, Simon LS, Van De
Putte LB, and Lipsky PE. Cyclooxygenase in biology and disease.
FASEB J 12: 1063–1073, 1998.
13. Harding P, Sigmon DH, Alfie ME, Huang PL, Fishman MC, Beierwaltes WH, and Carretero OA. Cyclooxygenase-2 mediates increased
renal renin content induced by low-sodium diet. Hypertension 29: 297–
302, 1997.
14. Harris RC. Cyclooxygenase-2 in the kidney. J Am Soc Nephrol 11:
2387–2394, 2000.
15. He Q, Wang D, Yang XP, Carretero OA, and LaPointe MC. Inducible
regulation of the human BNP promoter in transgenic mice. Am J Physiol
Heart Circ Physiol 280: H368–H376, 2001.
16. Hocherl K, Dreher F, Kurtz A, and Bucher M. Cyclooxygenase-2
inhibition attenuates lipopolysaccharide-induced cardiovascular failure.
Hypertension 40: 947–953, 2002.
H1423
H1424
COX-2 AND CARDIAC REMODELING
AJP-Heart Circ Physiol • VOL
45. Williams C, Shattuck-Brandt RL, and DuBois RN. The role of COX-2
in intestinal cancer. Ann NY Acad Sci 889: 72–83, 1999.
46. Williams CS, Mann M, and DuBois RN. The role of cyclooxygenases in
inflammation, cancer, and development. Oncogene 18: 7908–7916, 1999.
47. Wong SC, Fukuchi M, Melnyk P, Rodger I, and Giaid A. Induction
of cyclooxygenase-2 and activation of nuclear factor-kappaB in myocardium of patients with congestive heart failure. Circulation 98:
100–103, 1998.
48. Yamamoto T, Kakar NR, Vina ER, Johnson PE, and Bing RJ. Effect
of cyclooxygenase-2 inhibitor (celecoxib) on the infarcted heart in situ.
Pharmacology 63: 28–33, 2001.
49. Yang F, Liu YH, Yang XP, Xu J, Kapke A, and Carretero OA.
Myocardial infarction and cardiac remodelling in mice. Exp Physiol 87:
547–555, 2002.
50. Yang X, Ma N, Szabolcs MJ, Zhong J, Athan E, Sciacca RR, Michler
RE, Anderson GD, Wiese JF, Leahy KM, Gregory S, and Cannon PJ.
Upregulation of COX-2 during cardiac allograft rejection. Circulation
101: 430–438, 2000.
51. Yang XP, Liu YH, Mehta D, Cavasin MA, Shesely E, Xu J, Liu F, and
Carretero OA. Diminished cardioprotective response to inhibition of
angiotensin-converting enzyme and angiotensin II type 1 receptor in B2
kinin receptor gene knockout mice. Circ Res 88: 1072–1079, 2001.
286 • APRIL 2004 •
www.ajpheart.org
Downloaded from http://ajpheart.physiology.org/ by 10.220.32.247 on May 10, 2017
38. Scheuren N, Jacobs M, Ertl G, and Schorb W. Cyclooxygenase-2 in
myocardium stimulation by angiotensin-II in cultured cardiac fibroblasts
and role at acute myocardial infarction. J Mol Cell Cardiol 34: 29–37,
2002.
39. Shinmura K, Tang XL, Wang Y, Xuan YT, Liu SQ, Takano H,
Bhatnagar A, and Bolli R. Cyclooxygenase-2 mediates the cardioprotective effects of the late phase of ischemic preconditioning in conscious
rabbits. Proc Natl Acad Sci USA 97: 10197–10202, 2000.
40. Smith WL, DeWitt DL, and Garavito RM. Cyclooxygenases: structural,
cellular, and molecular biology. Annu Rev Biochem 69: 145–182, 2000.
41. Tardieu D, Jaeg JP, Deloly A, Corpet DE, Cadet J, and Petit CR. The
COX-2 inhibitor nimesulide suppresses superoxide and 8-hydroxy-deoxyguanosine formation, and stimulates apoptosis in mucosa during early
colonic inflammation in rats. Carcinogenesis 21: 973–976, 2000.
42. Tsuruda T, Boerrigter G, Huntley BK, Noser JA, Cataliotti A,
Costello-Boerrigter LC, Chen HH, and Burnett JC Jr. Brain natriuretic
peptide is produced in cardiac fibroblasts and induces matrix metalloproteinases. Circ Res 91: 1127–1134, 2002.
43. Vane JR, Bakhle YS, and Botting RM. Cyclooxygenases 1 and 2. Annu
Rev Pharmacol Toxicol 38: 97–120, 1998.
44. Wang JL, Cheng HF, Shappell S, and Harris RC. A selective cyclooxygenase-2 inhibitor decreases proteinuria and retards progressive renal
injury in rats. Kidney Int 57: 2334–2342, 2000.