NADPH oxidase following drug-induced myocardial injury promotes cardiac dysfunction and fibrosis

Author’s Accepted Manuscript Up-regulation of NOX1/NADPH oxidase following drug-induced myocardial injury promotes cardiac dysfunction and fibrosis Kazumi Iwata, Kuniharu Matsuno, Ayumi Murata, Kai Zhu, Hitomi Fukui, Keiko Ikuta, Masato Katsuyama, Masakazu Ibi, Misaki Matsumoto, Makoto Ohigashi, Xiaopeng Wen, Jia Zhang, Wenhao Cui, Chihiro Yabe-Nishimura

PII: DOI: Reference:

www.elsevier.com

S0891-5849(18)30164-3 https://doi.org/10.1016/j.freeradbiomed.2018.03.053 FRB13700

To appear in: Free Radical Biology and Medicine Received date: 13 December 2017 Revised date: 19 March 2018 Accepted date: 29 March 2018 Cite this article as: Kazumi Iwata, Kuniharu Matsuno, Ayumi Murata, Kai Zhu, Hitomi Fukui, Keiko Ikuta, Masato Katsuyama, Masakazu Ibi, Misaki Matsumoto, Makoto Ohigashi, Xiaopeng Wen, Jia Zhang, Wenhao Cui and Chihiro Yabe-Nishimura, Up-regulation of NOX1/NADPH oxidase following drug-induced myocardial injury promotes cardiac dysfunction and fibrosis, Free Radical Biology and Medicine, https://doi.org/10.1016/j.freeradbiomed.2018.03.053 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting galley proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Up-regulation of NOX1/NADPH oxidase following drug-induced myocardial injury promotes cardiac dysfunction and fibrosis

Kazumi Iwata a*, Kuniharu Matsuno a*, Ayumi Murata a, Kai Zhu a,b, Hitomi Fukui a, Keiko Ikuta a, Masato Katsuyama c, Masakazu Ibi a, Misaki Matsumoto a, Makoto Ohigashi a, Xiaopeng Wen a, Jia Zhang a, Wenhao Cui a, Chihiro Yabe-Nishimura a.

a

Department of Pharmacology, Kyoto Prefectural University of Medicine, 465 Kajii-cho

Kawaramachi-Hirokoji Kamigyo-ku Kyoto 602-8566, Japan b

Department of Nephrology, Renmin Hospital of Wuhan University, 238 Jiefang Rd.,

Wuchang District, Wuhan 430060, China. c

Radioisotope Center, Kyoto Prefectural University of Medicine, Kyoto, Japan, 465

Kajii-cho Kawaramachi-Hirokoji Kamigyo-ku Kyoto 602-8566, Japan

*These authors contributed equally to this work.

Correspondence to: Chihiro Yabe-Nishimura, MD, PhD, Department of Pharmacology, Kyoto Prefectural University of Medicine, 465 Kajii-cho Kawaramachi-Hirokoji Kamigyo-ku Kyoto 602-8566, Japan Fax:

+81-75-251-5348

Tel:

+81-75-251-5333

e-mail: [email protected]

Keywords: cardiac fibrosis, myocardial injury, NOX1/NADPH oxidase, reactive oxygen species

2

Abstract Cardiac fibrosis is a common feature in failing heart and therapeutic strategy to halt the progression of fibrosis is highly needed. We here report on NOX1, a non-phagocytic isoform of superoxide-producing NADPH oxidase, which promotes cardiac fibrosis in a drug-induced myocardial injury model. A single-dose administration of doxorubicin (DOX) elicited cardiac dysfunction accompanied by increased production of reactive oxygen species and marked elevation of NOX1 mRNA in the heart. In mice deficient in Nox1 (Nox1-/Y), cardiac functions were well retained and overall survival was significantly improved. However, increased level of serum creatine kinase was equivalent to that of wild-type mice (Nox1+/Y). At 4 days after DOX treatment, severe cardiac fibrosis accompanied by increased hydroxyproline content and activation of matrix metalloproteinase-9 was demonstrated in Nox1+/Y, but it was significantly attenuated in Nox1-/Y. When H9c2 cardiomyocytes were exposed to their homogenate, a dose-dependent increase in NOX1 mRNA was observed. Up-regulation of NOX1 mRNA in H9c2 co-incubated with their homogenate was abolished in the presence of TAK242, a TLR4 inhibitor. When isolated cardiac fibroblasts were exposed to H9c2 homogenates, increased proliferation and up-regulation of collagen 3a1 mRNA were demonstrated. These changes were significantly attenuated in cardiac fibroblasts exposed to homogenates from H9c2 harboring disrupted Nox1. These findings suggest that up-regulation of NOX1 following cellular damage promotes cardiac dysfunction and fibrosis by aggravating the pro-fibrotic response of cardiac fibroblasts. Modulation of the NOX1/NADPH oxidase signaling pathway may be a novel therapeutic strategy for preventing heart failure after myocardial injury.

3

Introduction Heart failure (HF) is one of the leading causes of mortality worldwide. Previous clinical trials demonstrated that morbidity and mortality in patients with HF with reduced ejection fraction (HFrEF) were diminished by treatment with angiotensin-converting enzyme inhibitors, angiotensin II receptor blockers, mineralocorticoid-receptor antagonists, and β-blockers [1]. On the other hand, in patients with HF with preserved ejection fraction (HFpEF), which accounts for 50% of HF cases, there is no effective therapy [2]. Given that decreased dilation in the left ventricle following the development of cardiac fibrosis is one of the causes of HFpEF [2], novel therapeutic strategy for cardiac fibrosis is highly needed. Doxorubicin (DOX) is a broad-spectrum antitumor anthracycline agent widely used to treat various hematopoietic and solid tumors. However, its therapeutic use is limited due to serious cardiotoxicity including structural and functional changes, often leading to congestive heart failure with poor prognosis [3]. Despite its cardiotoxicity, DOX is still included in most chemotherapeutic regimens because of its high efficacy. Investigations on the mechanism underlying DOX-induced cardiotoxicity demonstrated that production of reactive oxygen species (ROS) in cardiomyocytes is a principal cause of cardiotoxicity, eventually leading to cell death through apoptosis or necrosis. While redox cycling of DOX-derived quinone– semiquinone and a nonenzymic reaction involving ferric iron were speculated to generate ROS [4], the exact contribution of these factors in DOX-induced cardiotoxicity remains elusive. More recently, a superoxide-generating enzyme NADPH oxidase has been proposed as a novel source of ROS associated with DOX-induced cardiotoxicity [5-7]. Association of genetic polymorphisms of NADPH oxidase subunits with the risk of 4

DOX-induced cardiotoxicity has been documented in patients with lymphoma [5, 6]. NADPH oxidase is a multi-subunit enzyme composed of two membrane-bound subunits and several cytosolic regulatory subunits. So far, five homologs of the membrane bound catalytic subunit, NOX, have been identified [8, 9]. Among them, NOX2 (gp91phox), the phagocytic isoform of NOX, was reported to take part in cardiac dysfunction and remodeling following low-dose DOX treatment [6, 7]. Modulation of NOX2/NADPH oxidase activity could then be a possible strategy to prevent heart failure following therapeutic use of DOX. However, phagocyte NADPH oxidase is well-known to have a significant role in host defense [10] and inhibition of NOX2 may elicit various adverse effects in immune system [11]. As the non-phagocytic homolog of NOX2, NOX1 has been known to play pleiotropic roles in various organs including the cardiovascular system [8, 12]. We previously demonstrated that angiotensin II-induced hypertension was significantly suppressed in Nox1-deficient mice by maintainingnitric oxide (NO) bioavailability in the vessel [12]. Involvement of NOX1 was also documented in angiotensin II-dependent aortic dissection and diabetes mellitus-accelerated atherosclerosis in mice deficient in Nox1[13, 14]. On the other hand, NOX1 maintains the homeostasis of pulmonary vessels by regulating the turnover of smooth muscle cells [15]. While the expression of NOX1 is very low compared to other NOX isoforms expressed in the heart, we previously showed that up-regulation of NOX1 accelerated endotoxin-induced cardiomyocyte apoptosis and exacerbated cardiac dysfunction in a murine model of sepsis [16]. On the other hand, protective roles of NOX1 were revealed in hypoxia-induced bradycardia as well as in late phase of myocardial ischemic preconditioning[17, 18]. Accordingly, pathophysiological roles of NOX1 in cardiovascular diseases remain controversial. 5

It thus appeared relevant to examine whether NOX1/NADPH oxidase contributes to another model of cardiac disorder. We here report that up-regulation of NOX1 following DOX-induced cellular damage promotes proliferation of fibroblasts, leading to cardiac fibrosis. Modulation of the NOX1/NADPH oxidase signaling pathway may be a novel therapeutic strategy for the treatment of cardiac fibrosis, that underlies the development of HFpEF.

6

Materials and methods

Animal model Mice deficient in Nox1 gene (Nox1-/Y) [12] and their control littermates (Nox1+/Y) were maintained on a 12-hour light/dark cycle and fed ad libitum. A single dose of doxorubicin hydrochloride (20 mg kg-1; Wako Pure Chemical Industries Ltd., Osaka, Japan) was administrated to 8- to 12-week-old male mice by intraperitoneal injection. The present study was carried out with the approval of the Committee for Animal Research at Kyoto Prefectural University of Medicine.

Cell culture The H9c2 cell line, obtained from American Type Culture Collection (Rockville, MD, USA), was cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% FBS. Mouse cardiac fibroblasts were isolated and cultured as described earlier by a modification [19]. In brief, hearts removed from male Nox1+/Y or Nox1-/Y were washed with phosphate-buffered saline (PBS), and minced. After digestion in 0.2% collagenase type I solution (Wako Pure Chemical Industries Ltd.) at 37°C for 75 min, cells were pelleted by centrifugation at 300 g for 5 min and suspended in DMEM supplemented with 1% penicillin, 1% streptomycin, and 10% FBS. The suspension was transferred to a culture dish and incubated at 37°C. The culture dish was slowly shaken every 30 min for 3 hours. Cells weakly attached or unattached were removed by rinsing with PBS. The attached cells were cultured in the dish with DMEM. Fibroblasts cultured to the second passage were used in all experiments. 7

CRISPR-Cas9-mediated genome editing CRISPR design (http://crispr.mit.edu/) was used to construct sgRNAs. The gRNA cloning vector was a gift from Church G. (Addgene plasmid # 41824) [20] and used as a cloning backbone. pCAG-hCas9, a Cas9 expression vector, was a gift from Hatada I. (Addgene plasmid # 51142) [21]. Preparation of gRNA expression vector was carried out as previously described [20]. Briefly, oligonucleotides were annealed and extended to make a 100 bp double stranded DNA fragment using KOD-plus neo (TOYOBO, Osaka, Japan). The oligonucleotide sequences used are shown in the Data Supplement, Table I. The gRNA cloning vector was linearized using AflII and the 100bp DNA fragment was incorporated into the vector using In-Fusion HD cloning kit (TAKARA BIO㸪Shiga, Japan). Following sequence verification of the gRNA expression vectors, they were co-transfected with pCAG-hCas9 and pCDH-CMV-MCS-EF1-GFP-T2A-Puro (System biosciences, CA, USA) into H9c2 cells with Xfect transfection reagent (TAKARA BIO). Positive cells were selected using puromycin (4 μg/ml; Nacalai tesque, Kyoto, Japan) at one day prior to cell cloning. Empty gRNA cloning vector was used as a mock control.

Detection of cell proliferation Cell proliferation was measured by using a CellTiter-Glo® luminescent cell viability assay kit (Promega, WI, USA) according to the manufacturer's instructions. H9c2 cells were harvested with trypsin and centrifuged (200 g, 5  min). Collected H9c2 cells were washed with PBS twice and kept at -80 °C until use. Stored H9c2 was sonicated for 60 s by ultrasonic disruptor UD-201 (TOMY, Tokyo, Japan) at the concentration of 106cells/ml

8

with serum-free DMEM prior to incubation. Cardiac fibroblasts (5x103 cells/well) were grown in 96-well plates for 24 hours. For cell proliferation assay, cardiac fibroblasts were serum-starved for 24 hours and co-incubated with sonicated H9c2 cell suspensions for 72 hours without FBS.

Quantitative PCR Expression levels of mRNAs were evaluated by quantitative PCR (qPCR) as previously described with a modification [15]. Total RNA was isolated by the acid guanidinium thiocyanate/phenol/chloroform method. RNA was reverse transcribed using PrimeScript RT reagent Kit (TAKARA BIO). qPCR was performed using the GeneAmp 7300 Real-Time PCR System (Applied Biosystems, Foster City, CA, USA) with the SYBR Premix Ex Taq II (TAKARA BIO). The primer sequences used are shown in the Data Supplement, Table II. Dissociation curves were monitored to check the aberrant formation of primer dimers. PCR-amplified products were electrophoresed on 2% agarose gels to confirm the presence of a single band.

Detection of superoxide Superoxide (O2-) production was evaluated by dihydroethidium (DHE) staining and lucigenin chemiluminescence. DHE staining was performed as described previously [12]. Lucigenin chemiluminescence was measured as described with a minor modification [22]. Heart tissue was homogenized in Krebs-HEPES buffer and centrifuged at 4,500 g for 15 minutes. The supernatant fractions were collected and transferred to scintillation vials containing 5 µM lucigenin (Sigma, MO, USA) in Krebs-HEPES buffer, and incubated in the dark for 10 9

minutes at 37°C. After incubation, 200 µM NADPH was added and the chemiluminescence was measured by luminometer (AB-2200 Luminescencer-PSN, ATTO Co., Aichi, Japan) at 1-minute intervals over a 5 min period. Data are expressed as relative light units (RLU) per minute per milligram protein.

Measurement of cardiac functions Left ventricular systolic pressure (LVSP) and positive maximal values of the first derivative of left ventricular (LV) pressure (+dp/dt max) were determined by left ventricular catheterization as previously described [15].

Biochemical analysis Levels of creatine kinase (CK), creatine kinase-MB isozyme (CK-MB), alanine aminotransferase (ALT), aspartate aminotransferase (AST), blood urea nitrogen (BUN), and creatinine (Cre) were measured by Bioindustry Division of Oriental Yeast Co., Ltd. (Tokyo, Japan) according to standard methods. Activities of matrix metalloproteinase (MMP) 2 and MMP9 were assessed using gelatin zymography as described previously [23]. Hydroxyproline content was measured by colorimetric assay [24].

Histological analysis Mice were anesthetized with pentobarbital and transcardially perfused with 10 mL of PBS followed by 10 mL of 4% paraformaldehyde phosphate buffer. Paraffin-embedded sections were stained with Sirius red as previously described [24].

10

Immunohistochemical analyses Tissues isolated from mice under pentobarbital anesthesia (50 mg/kg, i.p.) were fixed in 4% paraformaldehyde/PBS overnight, washed three times with PBS, and embedded in paraffin. The 5-µm sections were incubated with primary antibodies overnight at 4°C. After washing with PBS containing 0.1% Tween 20, the sections were further incubated with a horseradish peroxidase-conjugated goat anti-rat or rabbit IgG (PK-4004 or PK-6101; VECTASTAIN Elite ABC HRP Kit). Counterstaining was carried out using hematoxylin. Primary antibodies against CD45 (1:200, 550539; BD Biosciences, San Diego, CA), CD68 (1:500, ab1252121; Abcam, Cambridge, UK), and F4/80 (1:100, ab100790; Abcam) were utilized.

Western blot analysis Proteins were extracted from cardiac tissue by mincing, homogenizing in lysis buffer (1% NP-40, 0.5% Sodium deoxycholate, 0.1% SDS, 1mM DTT, protease inhibitor cocktail in PBS, protease inhibitor cocktail; Sigma-Aldrich, St. Louis, MO, USA), and centrifugation at 8,000 g for 20 min. The supernatant fraction was subjected to Western blot analysis. An aliquot of 30 µg protein was separated by electrophoresis and transferred onto a nitrocellulose membrane. After blocking for 1 hour at room temperature, membranes were incubated overnight at 4°C with primary antibodies. Immunoreactive bands were detected by chemiluminescence following incubation with horseradish peroxidase-conjugated secondary antibodies. The band intensity was quantified densitometrically using ImageJ software. Primary antibodies used were against: cleaved caspase 3 (1:500, 9661S; Cell signaling Technology, Danvers, MA, USA), a-actin (1:20,000, A2172; Sigma-Aldrich), fibronectin (1:2,000, F3648; Sigma-Aldrich), HMGB1(1:10,000, ab79823; Abcam), alpha smooth 11

muscle actin (1:10,000, 14-9760-80; Thermo Fisher Scientific, Waltham, MA), and GAPDH (1:10,000, 2118S; Cell signaling Technology). Anti-rabbit IgG HRP linked antibody (1:20,000, 7074S; Cell signaling Technology) and anti-mouse IgG HRP linked antibody (1:20,000, 7076S; Cell signaling Technology) were used as secondary antibodies.

Statistical analysis The results are expressed as the mean ± SEM. Statistical analysis was performed with the Kruskal-Wallis test followed by the Bonferroni test or two-way ANOVA followed by application of the Tukey-Kramer. Survival data were analyzed by the Kaplan-Meier test.

12

Results

Administration of DOX increased expression of NOX1 and production of ROS in mouse heart Three days after a single dose of DOX administration, a significant increase in NOX1 mRNA was demonstrated in the heart (Fig. 1A). DHE staining and lucigenin chemiluminescence assay indicated that absence of Nox1 almost completely abolished increased production of ROS in the cardiac tissue (Fig. 1B and C). Simultaneous analyses on other NOX isoforms expressed in the heart showed a transient elevation in NOX4 levels on Day 2, while NOX2 did not show any specific trend during the experimental period. Nox1 deficiency did not affect expression levels of other isoforms (Fig. 1D).

Nox1-deficiency attenuated cardiac dysfunction and improved the survival rate in DOX-treated mice To examine the involvement of NOX1-derived ROS in DOX-induced cellular damage, cardiac functions were evaluated. Before DOX administration, there was no difference in LVSP and +dp/dt max between the two genotypes. These parameters were significantly attenuated in wild-type mice (Nox1+/Y) at 4 days after DOX administration, but not in Nox1-/Y (Fig. 1E). When survival rates were examined for 7 days, Nox1-/Y survived significantly longer than Nox1+/Y (Fig. 1F).

Nox1-deficiency did not affect the severity of myocardial injury induced by DOX

13

We first evaluated the levels of cleaved caspase 3, because apoptotic loss of cardiomyocytes by DOX is known as a major factor leading to cardiac dysfunction [3]. At the time when NOX1 was significantly up-regulated, activation of caspase 3 was not detected in the heart at 4 days after DOX treatment (Fig. 2A). These findings were further verified by TUNEL staining (data not shown). When biochemical parameters of cardiac injury were assessed, serum levels of CK and CK-MB were significantly elevated at 4 days after DOX treatment without any difference between the genotypes (Fig. 2B). These findings suggested that Nox1-deficiency did not affect the severity of myocardial injury induced by DOX. To examine direct effects of DOX administration on other organs, biochemical parameters were measured in the serum of mice treated with DOX. As shown in the Supplemental Fig. 1, the levels of ALT, AST, and BUN/Cre ratio were equivalent between the two genotypes.

Nox1-deficiency blunted the expression of proinflammatory cytokines in DOX-treated heart. It is known that proinflammatory cytokines act synergistically to depress cardiac function [25]. As shown in Fig. 2C, expression levels of IL-1b, IL-6, and TNF-a were significantly elevated in Nox1+/Y following DOX administration, whereas in Nox1-/Y heart, the levels of IL-1b and IL-6 were significantly suppressed. Proinflammatory cytokines are expressed in various cells, including cardiomyocytes, fibroblasts, and migrated inflammatory cells. When infiltration of leukocytes was examined in cardiac tissue, the number of CD45 positive cells was significantly reduced in both genotypes (Supplemental Fig. 2A). Similarly,

14

infiltration of CD68 or F4/80 positive macrophages was not apparent in the tissue (Supplemental Fig. 2B).

Nox1-deficiency alleviated the development of DOX-induced cardiac fibrosis Cardiac fibrosis is among the major factors contributing to the development of cardiac dysfunction [26]. As shown in Fig. 3A, Sirius red staining clearly illustrated the development of fibrosis in the heart of Nox1+/Y at 4 days after DOX treatment. Simultaneously, the level of hydroxyproline, an amino acid specifically contained in collagen, as well as fibronectin expression were significantly elevated in Nox1+/Y, whereas these changes were significantly suppressed in Nox1-/Y (Fig. 3A-C). When the expression of pro-fibrotic factors, CTGF and TGF-b, were examined, up-regulation of CTGF by DOX treatment was significantly blunted in Nox1-/Y (Fig. 3D). Matrix metalloproteinases (MMPs) are enzymes responsible for extracellular matrix degradation. Excessive synthesis of MMP2 and MMP9 is involved in cardiac remodeling and progression of cardiac fibrosis. In the present animal model, the activity of MMP2 was almost unchanged in all groups (Supplemental Fig. 3A). On the other hand, the mRNA level and the activity of MMP9 were significantly increased in Nox1+/Y, but not in Nox1-/Y (Fig. 3E). Simultaneously, the expression of tissue inhibitor of metalloproteinases (TIMPs), main regulators of MMPs activity, was significantly augmented in Nox1+/Y, whereas it was significantly suppressed in Nox1-/Y heart (Supplemental Fig. 3B).

NOX1 was up-regulated in H9c2 cardiomyocytes exposed to their homogenate via toll-like receptor 4 15

To clarify the molecular mechanisms underlying DOX-induced NOX1 expression in the heart, a rat cardiomyoblast cell line H9c2 was exposed to a different concentration of DOX. Unexpectedly, expression of NOX1 was significantly decreased in cells treated with DOX (Fig. 4A). It has been reported that damage-associated molecular pattern molecules (DAMPs) released from injured cells induce inflammation and fibrosis via pattern recognition receptors (PRRs) such as toll-like receptors (TLRs) [27, 28]. In fact, we previously reported the TLR4-mediated up-regulation of NOX1 in cardiomyocytes by lipopolysaccharides (LPS) [16]. When intact H9c2 cells were exposed to their cell homogenates, a dose-dependent increase in NOX1 mRNA was observed (Fig. 4B and C). Increased NOX1 expression was completely abolished in the presence of TAK242, a TLR4 inhibitor (Fig. 4D). These results suggest that DOX does not directly up-regulate NOX1 expression in cardiomyocytes, but DAMPs released from injured cells may induce NOX1 expression. It is known that inflammatory cytokines and MMPs are expressed in cardiomyocytes. We examined whether induction of NOX1 in H9c2 cells affects the expression of IL-6 and MMP9. As shown in Supplemental Fig. 4, however, no significant change was observed in H9c2 exposed to cell homogenates.

Proliferation of cardiac fibroblasts was modulated by NOX1 in cardiomyocytes It is known that cardiac fibroblasts act as a sensor of myocardial damage by responding to DAMPs released from injured cells by proliferation and expression of inflammatory cytokines and extracellular matrix (ECM) [27]. When primary cultured cardiac fibroblasts isolated from Nox1+/Y and Nox1-/Y were exposed to homogenate of H9c2, significant 16

increases in cell proliferation and expression of IL-6 were observed in both genotype-derived fibroblasts (Fig. 5A and B; Supplemental Fig. 5A and B). Similarly, there was no consistent effect of Nox1-deficiency on the level of MMP9 mRNA in fibroblasts exposed to H9c2 homogenate (Supplemental Fig. 5B). A dose-dependent increase in fibroblast proliferation was significantly blunted when fibroblasts were co-incubated with homogenates derived from H9c2 clones harboring disrupted Nox1 (Fig. 5B; Supplemental Fig. 6). Among ECM components examined, a significant increase in collagen 3a1 mRNA was demonstrated in cardiac fibroblasts exposed to H9c2 homogenate. In either genotype of fibroblasts, up-regulation of collagen 3a1 was significantly suppressed when co-incubated with homogenates of H9c2 harboring disrupted Nox1 (Fig. 5C; Supplemental Fig. 7A and B). When expression of alpha smooth muscle actin (a-SMA), a marker of myofibroblast differentiation, was examined in cardiac fibroblasts, no significant difference was observed among fibroblasts exposed to mock- or Nox1-disrupted H9c2 homogenates. (Supplemental Fig. 8). Extracellular superoxide dismutase was reported to protect the heart from oxidant-induced fibrosis [29]. We therefore examined the effect of hydrogen peroxide on proliferation of cardiac fibroblasts. No effect was observed in fibroblasts isolated from both genotypes (Supplemental Fig. 9). We then generated H9c2 clones harboring disrupted Nox2 or Nox4 and co-incubated with fibroblasts (Supplemental Fig. 10). Homogenate derived from either of these clones induced significant proliferation of fibroblasts similar to wild-type H9c2 (Supplemental Fig. 11). It is reported that DAMPs released from damaged cardiac cells induced cardiac fibrosis via TLR4 and RAGE, receptor for advanced glycation end products [27]. Treatment of 17

TAK-242, a TLR-4 inhibitor, but not FPS-ZM1, an antagonist of RAGE, partially but significantly suppressed proliferation of cardiac fibroblasts induced by H9c2 homogenate (Supplemental Fig. 12A). Among of DAMPs, HMGB1, fibronectin-EDA and S100A1 are known to induce cardiac fibrosis via TLR4. When effects of Hmgb1-, fibronectin- (Fn) or S100a1-disrupted H9c2 (Supplemental Fig. 12B) were examined, increased proliferation of cardiac fibroblasts induced by H9c2 homogenate (Mock) was partially but significantly suppressed in cells co-incubated with homogenate of S100a1-, but not of Hmgb1- or Fn-disrupted H9c2 (Supplemental Fig. 12C). Because expression of S100A1 was significantly reduced in Nox1-disrupted H9c2 (Supplemental Fig. 12D), S100A1 seems to be one of the candidate molecules regulated by NOX1 to develop cardiac fibrosis following myocardial injury. To further characterize whether myocyte-derived factors are secreted signaling molecules that activate fibroblasts, isolated fibroblasts were exposed to the supernatant fraction of the serum-free medium of cultured H9c2 cells (Fig. 6A). As shown in Fig. 6B, the serum-free supernatant significantly augmented proliferation of fibroblasts isolated from either Nox1+/Y or Nox1-/Y. Proliferation of fibroblasts was significantly suppressed when co-incubated with the supernatant derived from Nox1-disrupted H9c2 clones. To reproduce the sequence of events assumed to occur after cardiac injury, where NOX1 in myocytes adjacent to injured cells is up-regulated, myocyte homogenate was first placed on intact H9c2 cells and cultured in the serum-free medium for 24 hours. The culture supernatant of these H9c2 cells containing cell homogenate was then placed on cardiac fibroblasts (Fig. 6C). Proliferation of fibroblasts isolated from either genotype was further

18

augmented by the supernatant of wild-type H9c2 containing co-incubated myocyte homogenate, but not by the supernatant derived from Nox1-disrupted H9c2 clones (Fig. 6D). Accordingly, these findings suggested that proliferation of cardiac fibroblasts could be further promoted by NOX1-dependent paracrine signaling from cardiomyocytes (Fig. 7).

19

Discussion In this study, we demonstrated for the first time the involvement of NOX1 in the development of cardiac dysfunction and fibrosis in a drug-induced cardiomyopathy model. It has been known that increased production of ROS leads to myocardial cell death and cardiac fibrosis. However, less attention has been paid to how myocyte injury and ROS promote fibrosis. Present findings depicted the role for NOX1, a minor isoform of NADPH oxidase expressed in the heart, in mediating the development of cardiac fibrosis. In addition to NOX1, there are two other NOX isoforms, NOX2 and NOX4, predominantly expressed in cardiac tissue. NOX2 was previously reported to take part in myocardial cell death, cardiac dysfunction, and remodeling following repeated administration of low-dose DOX [7]. More recently, single-dose administration of DOX was documented to induce cardiomyocyte atrophy by increased ROS production derived from NOX2 [30]. In our experimental model, however, there was no substantial alteration in the expression of cardiac NOX2 following DOX administration. We previously reported that NOX1 was rapidly and markedly up-regulated following administration of LPS and caused cardiomyocyte apoptosis [16]. In the present animal model, on the other hand, apoptotic cell death was not apparent and serum creatine kinase, a marker of myocardial cell death, was similarly elevated in both genotypes. Instead of affecting cellular damage, NOX1 therefore modulated the progress of cardiac fibrosis following DOX-induced myocyte injury. The difference in the role of NOX1 in these two animal models may be at least partly attributed to its distinct induction pattern after the treatment. Compared with the LPS model, DOX-induced up-regulation of NOX1 was slow and less.

20

ROS derived from NOX1 may therefore act as intracellular signaling molecules to promote cardiac fibrosis, rather than cellular apoptosis. While NOX1 was first identified as a non-phagocytic isoform of NADPH oxidase, we previously reported that it is expressed and up-regulated in peritoneal macrophages treated with LPS [31]. Inflammatory cells infiltrating cardiac tissue might therefore contribute to the increased level of NOX1 mRNA demonstrated in DOX-treated mice. However, the number of CD45 positive leukocytes was significantly reduced in both genotypes. Similarly, infiltration of CD68 or F4/80 positive macrophages was not apparent in the tissue. The limited accumulation of inflammatory cells in our animal model may be due to myelosuppression by a high-dose of DOX administration. Despite the limited accumulation of inflammatory cells in DOX-treated mouse heart, expression of inflammatory cytokines was significantly up-regulated in cardiac tissue. It is known that inflammatory cytokines are also expressed in cardiomyocytes and cardiac fibroblasts. However, NOX1 in either H9c2 cells or cardiac fibroblasts did not affect the expression of IL-6 in cultured cell experiments. The reason why we could not reproduce the findings obtained in animal studies is presently unclear. A speculative possibility is that cytokines may be induced in cell types other than cardiac myocytes or fibroblasts in DOX-treated heart. Previously, induction of NOX1 was reported in H9c2 cells treated with DOX [32]. In contrast, we demonstrated that DOX rather down-regulated NOX1 expression in the same cell line. The reason for this discrepancy might be due to the difference in culture conditions, because damaged cells currently up-regulated NOX1 expression in myocytes when co-incubated in the same medium. DAMPs released from damaged cells activate TLRs [27], 21

and TLR4-mediated up-regulation of NOX1 was demonstrated in this study, similar to our previous LPS model [16]. It is therefore conceivable that DAMPs derived from drug-induced myocardial injury up-regulated NOX1 in adjacent cardiomyocytes. In this context, our present data were in line with the earlier findings indicating that cardiac dysfunction and oxidative stress induced by DOX were attenuated in TLR4-deficient mice [33]. Given that involvement of high-mobility group protein 1 (HMGB1), one of the TLR4 ligands in DAMPs, was documented in DOX-induced cardiac dysfunction [34], the HMGB1/TLR4 pathway may be upstream of increased NOX1 expression in cardiomyocytes. The next question is how increased ROS derived from NOX1 induce cardiac fibrosis. Because hydrogen peroxide did not directly affect the proliferation of fibroblasts (Supplemental Fig. 9), ROS derived from myocyte NOX1 may modulate the pro-fibrotic response of cardiac fibroblasts. In the preceding study, we observed that the level of NOX1 mRNA in isolated cardiac fibroblasts was limited compared with cardiomyocytes [16]. Unlike in H9c2 cells, no induction of NOX1 mRNA was observed in isolated cardiac fibroblasts exposed to myocyte homogenate (data not shown). Furthermore, cardiac fibroblasts isolated from both Nox1+/Y and Nox1-/Y similarly proliferated upon exposure to myocyte homogenate. While proliferation of fibroblasts was attenuated by disruption of Nox1 in myocytes, homogenate from Nox2- and Nox4-disrupted H9c2 clones induced proliferation of fibroblasts similar to wild-type H9c2 (Supplemental Fig. 11). Accordingly, among NOX isoforms expressed in the heart, cardiomyocyte NOX1 regulates production and/or activation of profibrotic factors that induce proliferation and expression of ECM in cardiac fibroblasts. Because increased proliferation of fibroblasts induced by H9c2 homogenate was partially

22

blunted by S100a1 disruption, S100A1 may be one of the candidate profibrotic factors regulated by NOX1. It is still unclear whether NOX1 is involved in activation of cardiac fibroblasts. When expression of a-SMA was examined in cultured cardiac fibroblasts that may have already differentiated into myofibroblasts, no difference was observed among those exposed to mock- or Nox1-disrupted H9c2 homogenates. Because epithelial to mesenchymal and/or endothelial to mesenchymal transitions also contribute to the development of cardiac fibrosis [35], there is a possibility that NOX1-mediated mesenchymal transition takes part in fibrosis induced by myocardial injury. In fact, intense Sirius red staining for collagen was demonstrated in perivascular location of DOX-treated heart, and NOX1 is also expressed in endothelial cells [36]. Likewise, the involvement of NOX1 was recently reported in epithelial to mesenchymal transition in cancer cells [37, 38]. Certainly further study is required to clarify the underlying mechanism of cardiac remodeling mediated by NOX1.

Conclusions Up-regulation of NOX1 following drug-induced cellular damage promotes cardiac dysfunction and fibrosis by aggravating the pro-fibrotic response of cardiac fibroblasts. Modulation of the NOX1/NADPH oxidase signaling pathway may be a novel therapeutic strategy for cardioprotection after myocardial injury.

23

Sources of Funding This work was supported in part by a Grant-in-Aid for Scientific Research (C) 15K08239 (to K. Iwata) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan.

Disclosure CY-N is a cofounder of a startup company developing NOX inhibitors.

24

References [1] C.W. Yancy, M. Jessup, B. Bozkurt, J. Butler, D.E. Casey, Jr., M.M. Colvin, M.H. Drazner, G.S. Filippatos, G.C. Fonarow, M.M. Givertz, S.M. Hollenberg, J. Lindenfeld, F.A. Masoudi, P.E. McBride, P.N. Peterson, L.W. Stevenson, C. Westlake, 2017 ACC/AHA/HFSA Focused Update of the 2013 ACCF/AHA Guideline for the Management of Heart Failure: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines and the Heart Failure Society of America, Circulation 136(6) (2017) e137-e161. [2] S.J. Shah, D.W. Kitzman, B.A. Borlaug, L. van Heerebeek, M.R. Zile, D.A. Kass, W.J. Paulus, Phenotype-Specific Treatment of Heart Failure With Preserved Ejection Fraction: A Multiorgan Roadmap, Circulation 134(1) (2016) 73-90. [3] G. Minotti, P. Menna, E. Salvatorelli, G. Cairo, L. Gianni, Anthracyclines: molecular advances and pharmacologic developments in antitumor activity and cardiotoxicity, Pharmacological reviews 56(2) (2004) 185-229. [4] M.S. Horenstein, R.S. Vander Heide, T.J. L'Ecuyer, Molecular basis of anthracycline-induced cardiotoxicity and its prevention, Mol Genet Metab 71(1-2) (2000) 436-44.

25

[5] A. Reichwagen, M. Ziepert, M. Kreuz, U. Godtel-Armbrust, T. Rixecker, V. Poeschel, M. Reza Toliat, P. Nurnberg, M. Tzvetkov, S. Deng, L. Trumper, G. Hasenfuss, M. Pfreundschuh, L. Wojnowski, Association of NADPH oxidase polymorphisms with anthracycline-induced cardiotoxicity in the RICOVER-60 trial of patients with aggressive CD20(+) B-cell lymphoma, Pharmacogenomics 16(4) (2015) 361-72. [6] L. Wojnowski, B. Kulle, M. Schirmer, G. Schluter, A. Schmidt, A. Rosenberger, S. Vonhof, H. Bickeboller, M.R. Toliat, E.K. Suk, M. Tzvetkov, A. Kruger, S. Seifert, M. Kloess, H. Hahn, M. Loeffler, P. Nurnberg, M. Pfreundschuh, L. Trumper, J. Brockmoller, G. Hasenfuss, NAD(P)H oxidase and multidrug resistance protein genetic polymorphisms are associated with doxorubicin-induced cardiotoxicity, Circulation 112(24) (2005) 3754-62. [7] Y. Zhao, D. McLaughlin, E. Robinson, A.P. Harvey, M.B. Hookham, A.M. Shah, B.J. McDermott, D.J. Grieve, Nox2 NADPH oxidase promotes pathologic cardiac remodeling associated with Doxorubicin chemotherapy, Cancer research 70(22) (2010) 9287-97. [8] K. Bedard, K.H. Krause, The NOX family of ROS-generating NADPH oxidases: physiology and pathophysiology, Physiological reviews 87(1) (2007) 245-313.

26

[9] Y.A. Suh, R.S. Arnold, B. Lassegue, J. Shi, X. Xu, D. Sorescu, A.B. Chung, K.K. Griendling, J.D. Lambeth, Cell transformation by the superoxide-generating oxidase Mox1, Nature 401(6748) (1999) 79-82. [10] J. Cachat, C. Deffert, S. Hugues, K.H. Krause, Phagocyte NADPH oxidase and specific immunity, Clin Sci (Lond) 128(10) (2015) 635-48. [11] Z. Wen, Y. Shimojima, T. Shirai, Y. Li, J. Ju, Z. Yang, L. Tian, J.J. Goronzy, C.M. Weyand, NADPH oxidase deficiency underlies dysfunction of aged CD8+ Tregs, J Clin Invest 126(5) (2016) 1953-67. [12] K. Matsuno, H. Yamada, K. Iwata, D. Jin, M. Katsuyama, M. Matsuki, S. Takai, K. Yamanishi, M. Miyazaki, H. Matsubara, C. Yabe-Nishimura, Nox1 is involved in angiotensin II-mediated hypertension: a study in Nox1-deficient mice, Circulation 112(17) (2005) 2677-85. [13] S.P. Gray, E. Di Marco, J. Okabe, C. Szyndralewiez, F. Heitz, A.C. Montezano, J.B. de Haan, C. Koulis, A. El-Osta, K.L. Andrews, J.P. Chin-Dusting, R.M. Touyz, K. Wingler, M.E. Cooper, H.H. Schmidt, K.A. Jandeleit-Dahm, NADPH oxidase 1 plays a key role in diabetes mellitus-accelerated atherosclerosis, Circulation 127(18) (2013) 1888-902.

27

[14] G. Gavazzi, C. Deffert, C. Trocme, M. Schappi, F.R. Herrmann, K.H. Krause, NOX1 deficiency protects from aortic dissection in response to angiotensin II, Hypertension 50(1) (2007) 189-96. [15] K. Iwata, K. Ikami, K. Matsuno, T. Yamashita, D. Shiba, M. Ibi, M. Matsumoto, M. Katsuyama, W. Cui, J. Zhang, K. Zhu, N. Takei, Y. Kokai, O. Ohneda, T. Yokoyama, C. Yabe-Nishimura, Deficiency of NOX1/nicotinamide adenine dinucleotide phosphate, reduced form oxidase leads to pulmonary vascular remodeling, Arteriosclerosis, thrombosis, and vascular biology 34(1) (2014) 110-9. [16] K. Matsuno, K. Iwata, M. Matsumoto, M. Katsuyama, W. Cui, A. Murata, H. Nakamura, M. Ibi, K. Ikami, J. Zhang, S. Matoba, D. Jin, S. Takai, H. Matsubara, N. Matsuda, C. Yabe-Nishimura, NOX1/NADPH oxidase is involved in endotoxin-induced cardiomyocyte apoptosis, Free radical biology & medicine 53(9) (2012) 1718-28. [17] A. Kojima, A. Matsumoto, H. Nishida, Y. Reien, K. Iwata, T. Shirayama, C. Yabe-Nishimura, H. Nakaya, A protective role of Nox1/NADPH oxidase in a mouse model with hypoxia-induced bradycardia, J Pharmacol Sci 127(3) (2015) 370-6.

28

[18] S. Jiang, J. Streeter, B.M. Schickling, K. Zimmerman, R.M. Weiss, F.J. Miller, Jr., Nox1 NADPH oxidase is necessary for late but not early myocardial ischaemic preconditioning, Cardiovascular research 102(1) (2014) 79-87. [19] C.R. Kliment, T.D. Oury, Extracellular superoxide dismutase protects cardiovascular syndecan-1 from oxidative shedding, Free radical biology & medicine 50(9) (2011) 1075-80. [20] P. Mali, L. Yang, K.M. Esvelt, J. Aach, M. Guell, J.E. DiCarlo, J.E. Norville, G.M. Church, RNA-guided human genome engineering via Cas9, Science 339(6121) (2013) 823-6. [21] T. Horii, D. Tamura, S. Morita, M. Kimura, I. Hatada, Generation of an ICF syndrome model by efficient genome editing of human induced pluripotent stem cells using the CRISPR system, International journal of molecular sciences 14(10) (2013) 19774-81. [22] J.K. Bendall, N.J. Alp, N. Warrick, S. Cai, D. Adlam, K. Rockett, M. Yokoyama, S. Kawashima, K.M. Channon, Stoichiometric relationships between endothelial tetrahydrobiopterin, endothelial NO synthase (eNOS) activity, and eNOS coupling in vivo: insights from transgenic mice with endothelial-targeted GTP cyclohydrolase 1 and eNOS overexpression, Circ Res 97(9) (2005) 864-71. [23] K. Ishida, S. Takai, M. Murano, T. Nishikawa, T. Inoue, N. Murano, N. Inoue, D. Jin, E. Umegaki, K. Higuchi, M. Miyazaki, Role of chymase-dependent matrix metalloproteinase-9

29

activation in mice with dextran sodium sulfate-induced colitis, The Journal of pharmacology and experimental therapeutics 324(2) (2008) 422-6. [24] W. Cui, K. Matsuno, K. Iwata, M. Ibi, M. Matsumoto, J. Zhang, K. Zhu, M. Katsuyama, N.J. Torok, C. Yabe-Nishimura, NOX1/nicotinamide adenine dinucleotide phosphate, reduced form (NADPH) oxidase promotes proliferation of stellate cells and aggravates liver fibrosis induced by bile duct ligation, Hepatology 54(3) (2011) 949-58. [25] D.L. Maass, J. White, J.W. Horton, IL-1beta and IL-6 act synergistically with TNF-alpha to alter cardiac contractile function after burn trauma, Shock 18(4) (2002) 360-6. [26] D. Cappetta, F. Rossi, E. Piegari, F. Quaini, L. Berrino, K. Urbanek, A. De Angelis, Doxorubicin targets multiple players: A new view of an old problem, Pharmacological research (2017). [27] N.A. Turner, Inflammatory and fibrotic responses of cardiac fibroblasts to myocardial damage associated molecular patterns (DAMPs), Journal of molecular and cellular cardiology 94 (2016) 189-200. [28] W. Zhang, K.J. Lavine, S. Epelman, S.A. Evans, C.J. Weinheimer, P.M. Barger, D.L. Mann, Necrotic myocardial cells release damage-associated molecular patterns that provoke

30

fibroblast activation in vitro and trigger myocardial inflammation and fibrosis in vivo, Journal of the American Heart Association 4(6) (2015) e001993. [29] C.R. Kliment, H.B. Suliman, J.M. Tobolewski, C.M. Reynolds, B.J. Day, X. Zhu, C.F. McTiernan, K.R. McGaffin, C.A. Piantadosi, T.D. Oury, Extracellular superoxide dismutase regulates cardiac function and fibrosis, Journal of molecular and cellular cardiology 47(5) (2009) 730-42. [30] T. Shimauchi, T. Numaga-Tomita, T. Ito, A. Nishimura, R. Matsukane, S. Oda, S. Hoka, T. Ide, N. Koitabashi, K. Uchida, H. Sumimoto, Y. Mori, M. Nishida, TRPC3-Nox2 complex mediates doxorubicin-induced myocardial atrophy, JCI insight 2(15) (2017). [31] H. Yokota, A. Tsuzuki, Y. Shimada, A. Imai, D. Utsumi, T. Tsukahara, M. Matsumoto, K. Amagase, K. Iwata, A. Nakamura, C. Yabe-Nishimura, S. Kato, NOX1/NADPH Oxidase Expressed in Colonic Macrophages Contributes to the Pathogenesis of Colonic Inflammation in Trinitrobenzene Sulfonic Acid-Induced Murine Colitis, The Journal of pharmacology and experimental therapeutics 360(1) (2017) 192-200. [32] P. Spallarossa, P. Altieri, S. Garibaldi, G. Ghigliotti, C. Barisione, V. Manca, P. Fabbi, A. Ballestrero, C. Brunelli, A. Barsotti, Matrix metalloproteinase-2 and -9 are induced

31

differently by doxorubicin in H9c2 cells: The role of MAP kinases and NAD(P)H oxidase, Cardiovascular research 69(3) (2006) 736-45. [33] A. Riad, S. Bien, M. Gratz, F. Escher, D. Westermann, M.M. Heimesaat, S. Bereswill, T. Krieg, S.B. Felix, H.P. Schultheiss, H.K. Kroemer, C. Tschope, Toll-like receptor-4 deficiency attenuates doxorubicin-induced cardiomyopathy in mice, Eur J Heart Fail 10(3) (2008) 233-43. [34] Y. Yao, X. Xu, G. Zhang, Y. Zhang, W. Qian, T. Rui, Role of HMGB1 in doxorubicin-induced myocardial apoptosis and its regulation pathway, Basic Res Cardiol 107(3) (2012) 267. [35] P. Kong, P. Christia, N.G. Frangogiannis, The pathogenesis of cardiac fibrosis, Cellular and molecular life sciences : CMLS 71(4) (2014) 549-74. [36] G.R. Drummond, C.G. Sobey, Endothelial NADPH oxidases: which NOX to target in vascular disease?, Trends Endocrinol Metab 25(9) (2014) 452-63. [37] D. Kesanakurti, D. Maddirela, Y.K. Banasavadi-Siddegowda, T.H. Lai, Z. Qamri, N.K. Jacob, D. Sampath, S. Mohanam, B. Kaur, V.K. Puduvalli, A novel interaction of PAK4 with PPARgamma to regulate Nox1 and radiation-induced epithelial-to-mesenchymal transition in glioma, Oncogene

(2017).

32

[38] F. Liu, A.M. Gomez Garcia, F.L. Meyskens, Jr., NADPH oxidase 1 overexpression enhances invasion via matrix metalloproteinase-2 and epithelial-mesenchymal transition in melanoma cells, J Invest Dermatol 132(8) (2012) 2033-41.

33

Figure legends

Fig. 1. Nox1-deficiency attenuated cardiac dysfunction and improved the survival rate in DOX-treated mice. (A) Levels of NOX1 mRNA in cardiac tissue after administration of DOX. Data are expressed relative to GAPDH mRNA. N = 4 to 8 per group. *P < 0.05 vs. Day 0. (B) In situ detection of superoxide with dihydroethidium (DHE). Cross sections of the left ventricle were obtained from mice treated with DOX (Day 4). Images are representative of the results obtained from five animals. (C) Levels of superoxide in cardiac tissue determined by lucigenin chemiluminescence (Day 4). N = 5 per group. *P < 0.05, #P < 0.05. (D) Levels of NOX2 and NOX4 mRNAs expressed in cardiac tissue of Nox1+/Y (open circles) and Nox1-/Y (closed circles) after administration of DOX. Data are expressed relative to GAPDH mRNA. N = 4 to 8 per group. (E) Left ventricular systolic pressure (LSDP) and positive maximal values of the first derivative of left ventricular pressure (+dp/dt max) measured by left ventricular catheterization at 4 days after DOX treatment. N = 4 to 7 per group. * P < 0.05. (F) Survival rate after administration of DOX in Nox1+/Y (open circles, n = 17) and Nox1-/Y (closed circles, n = 17). The overall survival rate of Nox1-/Y was significantly higher than that of Nox1+/Y (*P < 0.05).

Fig. 2. Nox1-deficiency did not affect the severity of myocardial injury induced by DOX, but blunted the expression of proinflammatory cytokines. (A) Representative immunoblots and densitometric analysis of cleaved caspase 3 and a-actin in cardiac tissue (Day 4). N = 4 to 6 per group. (B) Levels of serum creatine kinase (CK) and creatine kinase-MB isozyme (CK-MB) (Day 4). N = 3 to 7 per group. *P < 0.05. (C) Levels of IL-1b,

34

IL-6 and TNF-a mRNAs in the heart at 3 days after DOX treatment. Data are expressed as number of copies per microgram of total RNA. N = 9 to 11 per group. *P < 0.05; #P < 0.05.

Fig. 3. Nox1-deficiency alleviated the development of DOX-induced cardiac fibrosis. (A) Representative photographs of Sirius red staining of cardiac sections (Day 4). Collagen is stained red. (B) Hydroxyproline content in the heart (Day 4). N = 8 to 12 per group. *P < 0.05; #P < 0.05. (C) Representative immunoblots and densitometric analysis of fibronectin and a-actin in cardiac tissue (Day 4). N = 4 to 6 per group. *P < 0.05; #P < 0.05. (D) Levels of CTGF and TGF-b mRNAs in the heart at 3 days after DOX treatment. Data are expressed as number of copies per microgram of total RNA. N = 5 to 12 per group. *P < 0.05; #P < 0.05. (E) Expression and activity of MMP9 in cardiac tissue. Levels of MMP9 mRNA were determined at 3 days after DOX treatment. N = 10 to 12 per group. Activities of MMP9 were determined by gelatin zymography at 4 days after DOX treatment. N = 4 to 6 per group. *P < 0.05; #P < 0.05.

Fig. 4. NOX1 was up-regulated in H9c2 myocytes exposed to their homogenate via toll-like receptor 4. (A) Levels of NOX1 mRNA in H9c2 treated with DOX. N = 3 to 4 per group. *P < 0.05. (B) A diagram for experimental protocol. Various concentrations of homogenate were placed on H9c2 and incubated for 6-24 hours. Levels of NOX1 mRNA in H9c2 treated with their homogenate in the absence (C) or presence of a TLR4 inhibitor TAK242 (D). N = 3 - 4 per group. *P < 0.05; #P < 0.05.

35

Fig. 5. Disruption of Nox1 in H9c2 blunted proliferation and expression of collagen 3a1 in cardiac fibroblasts. (A) A diagram for experimental protocol. Various concentrations of homogenate prepared from H9c2 clones harboring wild-type or disrupted Nox1 were placed on cardiac fibroblasts isolated from Nox1+/Y or Nox1-/Y. (B) Proliferation of fibroblasts were determined by quantitation of the ATP levels. Homogenate (Homo) of wild-type H9c2 similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % of control. N =4 per group. (C) Levels of fibronectin (FN), collagen 1a1 (Col1a1), and collagen 3a1 (Col3a1) mRNAs in fibroblasts co-incubated with homogenate from H9c2 clones harboring wild-type (Mock) or disrupted Nox1 (Clone 1-3). N = 3 per group. Similar results were obtained in three independent experiments. *P < 0.05 vs. control; #P < 0.05.

Fig. 6. Disruption of Nox1 in H9c2 blunted proliferation of cardiac fibroblasts induced by culture supernatant. (A) A diagram for experimental protocol. The supernatant fraction of the serum-free medium of cultured H9c2 clones harboring wild-type or disrupted Nox1 was placed on cardiac fibroblasts isolated from Nox1+/Y or Nox1-/Y. (B) Culture supernatant of wild-type H9c2 similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % of control. N =4-5 per group. (C) A diagram for experimental protocol. Wild-type myocyte homogenate was first placed on intact H9c2 clones harboring wild-type or disrupted Nox1 for 24 hours. The culture supernatant of these H9c2 cells containing wild-type myocyte homogenate was then placed on cardiac fibroblasts isolated from either genotype. (D) Culture supernatant (Sup) of wild-type H9c2 (Mock) similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % 36

of control. N =4-5 per group. Similar results were obtained in four independent experiments. *P < 0.05 vs. control. #P < 0.05

Fig. 7. A schematic diagram of NOX1-mediated cardiac remodeling. Doxorubicin (DOX), damage-associated molecular pattern molecules (DAMPs), toll-like receptor 4 (TLR4), extracellular matrix (ECM), reactive oxygen species (ROS).

Highlights

Cardiac fibrosis induced by doxorubicin was ameliorated in mice deficient in Nox1. NOX1 was up-regulated via TLR4 by DAMPs released from damaged cardiomyocytes. NOX1 in the heart positively regulates pro-fibrotic response of cardiac fibroblasts.

37

B

A

Nox1 (-/Y)

*

40

NOX1 mRNA expresion(×10-5) (relative to GAPDH)

Nox1 (+/Y)

Cont 30

*

*

20

DOX

10

0 0

1

2

3

4

5

Days

C

D

#

*

Nox1+/Y Nox1-/Y

6000

RLU/min/mg protein

4500

3000

1500

0

(+/Y)

(-/Y)

Cont

E

(+/Y)

400

NOX4 mRNA expression [% of Day 0 (+/Y)]

NOX2 mRNA expression [% of Day 0 (+/Y)]

150

100

50

0

(-/Y)

300

200

100

0 0

2

DOX

5

0

2

Days

*

F

*

12000

120

5

Days

Nox1+/Y Nox1-/Y 120

60

Survival rate (%)

+dp/dt max

LVSP (mmHg)

100

9000

90

6000

3000

30

80 60

*

40 20 0

0

0

(+/Y) Cont

(-/Y)

(+/Y) DOX

(-/Y)

(+/Y) Cont

(-/Y)

(+/Y)

(-/Y)

0

1

2

3

4

5

6

7

Days

DOX

Fig. 1. Nox1-deficiency attenuated cardiac dysfunction and improved the survival rate in DOXtreated mice. (A) Levels of NOX1 mRNA in cardiac tissue after administration of DOX. Data are expressed relative to GAPDH mRNA. N = 4 to 8 per group. *P < 0.05 vs. Day 0. (B) In situ detection of superoxide with dihydroethidium (DHE). Cross sections of the left ventricle were obtained from mice treated with DOX (Day 4). Images are representative of the results obtained from five animals. (C) Levels of superoxide in cardiac tissue determined by lucigenin chemiluminescence (Day 4). N = 5 per group. *P < 0.05, #P < 0.05. (D) Levels of NOX2 and NOX4 mRNAs expressed in cardiac tissue of Nox1+/Y (open circles) and Nox1-/Y (closed circles) after administration of DOX. Data are expressed relative to GAPDH mRNA. N = 4 to 8 per group. (E) Left ventricular systolic pressure (LSDP) and positive maximal values of the first derivative of left ventricular pressure (+dp/dt max) measured by left ventricular catheterization at 4 days after DOX treatment. N = 4 to 7 per group. * P < 0.05. (F) Survival rate after administration of DOX in Nox1+/Y (open circles, n = 17) and Nox1-/Y (closed circles, n = 17). The overall survival rate of Nox1-/Y was significantly higher than that of Nox1+/Y (*P < 0.05).

Cleaved caspase 3/α-actin [% of Cont (+/Y)]

150

A cleaved caspase 3 α-actin (+/Y)

(-/Y)

(+/Y)

Cont

(-/Y)

100

50

DOX 0

(+/Y)

( /Y) (-/Y)

(+/Y)

Cont

*

B

*

*

5000

( /Y) (-/Y)

DOX

*

6000

4000

CK-MB (IU/L)

CK (IU/L)

4500 3000

2000

3000

1500 1000

0

(+/Y)

(-/Y)

Cont

(+/Y)

0

(-/Y)

(+/Y)

DOX

(-/Y)

Cont

(+/Y)

(-/Y)

DOX

C

*

30

IL-6 (×103 copies/µg RNA)

IL-1β (×103 copies/µg RNA)

40

*

#

30

20

10

#

*

25

20

15

10

5

0

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

0

* 150

TNF-α (×10 TNF( 103 copies/µg RNA)

*

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

120

90

60

30

0

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

Fig. 2. Nox1-deficiency did not affect the severity of myocardial injury induced by DOX, but blunted the expression of proinflammatory cytokines. (A) Representative immunoblots and densitometric analysis of cleaved caspase 3 and a-actin in cardiac tissue (Day 4). N = 4 to 6 per group. (B) Levels of serum creatine kinase (CK) and creatine kinase-MB isozyme (CK-MB) (Day 4). N = 3 to 7 per group. *P < 0.05. (C) Levels of IL-1b, IL-6 and TNF-a mRNAs in the heart at 3 days after DOX treatment. Data are expressed as number of copies per microgram of total RNA. N = 9 to 11 per group. *P < 0.05; #P < 0.05.

A

*

B

Nox1(-/Y)

200

Hydroxyproline [% of Cont (+/Y)]

Nox1(+/Y)

Cont

DOX

*

150

100

50

0

C

#

(+/Y)

(-/Y)

(+/Y)

Cont

(-/Y)

DOX

D

Fibronectin α-actin (+/Y)

Cont

#

*

*

40

DOX

200

Fibronection [% of Cont (+/Y)]

(-/Y)

150

100

# 2.0

* TGF-β (×107 copies/µg RNA)

(-/Y)

CTGF (×107 copies/µg RNA)

(+/Y)

30

20

10

*

1.5

1.0

0.5

50

0 0

0.0

(+/Y) (-/Y) (+/Y) (-/Y) (+/Y)

(-/Y)

Cont

(+/Y)

(-/Y)

Cont

DOX

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

DOX

E

MMP9 (+/Y) (-/Y) (+/Y) Cont

*

# 300

MMP9 activity [% of Cont (+/Y)]

MMP9 (×103 copies/µg RNA)

400

300

200

100

0

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

(-/Y) DOX

*

#

200

100

0

(+/Y) (-/Y) (+/Y) (-/Y) Cont

DOX

Fig. 3. Nox1-deficiency alleviated the development of DOX-induced cardiac fibrosis. (A) Representative photographs of Sirius red staining of cardiac sections (Day 4). Collagen is stained red. (B) Hydroxyproline content in the heart (Day 4). N = 8 to 12 per group. *P < 0.05; #P < 0.05. (C) Representative immunoblots and densitometric analysis of fibronectin and a-actin in cardiac tissue (Day 4). N = 4 to 6 per group. *P < 0.05; #P < 0.05. (D) Levels of CTGF and TGF-b mRNAs in the heart at 3 days after DOX treatment. Data are expressed as number of copies per microgram of total RNA. N = 5 to 12 per group. *P < 0.05; #P < 0.05. (E) Expression and activity of MMP9 in cardiac tissue. Levels of MMP9 mRNA were determined at 3 days after DOX treatment. N = 10 to 12 per group. Activities of MMP9 were determined by gelatin zymography at 4 days after DOX treatment. N = 4 to 6 per group. *P < 0.05; #P < 0.05.

A

*

B

*

30

NOX1 mRNA expresion (×105 copies/µg RNA )

25

Homogenate

20

Collect cells Gene expression

15

WT H9c2

10

10 -510 6 /ml

WT H9c2 w/wo TAK-242

5

0 Cont 0.15

1.5

Cont 0.15

DOX

C

1.5 (µM)

DOX

6 hr

24 hr

*

*

D

* 160

NOX1/HPRT mRNA(% of Cont)

NOX1/β-actin mRNA(% of Cont)

350

#

300 250 200 150 100 50 0 Cont

10 5 10 6 Homo 6 hr

Cont

10 5 10 6(cells/ml) Homo 24 hr

140 120 100 80 60 40 20 0 Cont

Homo

DMSO

Cont

Homo

TAK-242

6 hr

Fig. 4. NOX1 was up-regulated in H9c2 myocytes exposed to their homogenate via toll-like receptor 4. (A) Levels of NOX1 mRNA in H9c2 treated with DOX. N = 3 to 4 per group. *P < 0.05. (B) A diagram for experimental protocol. Various concentrations of homogenate were placed on H9c2 and incubated for 6-24 hours. Levels of NOX1 mRNA in H9c2 treated with their homogenate (Homo) in the absence (C) or presence of a TLR4 inhibitor TAK242 (D). N = 3 - 4 per group. *P < 0.05; #P < 0.05.

A Homogenate 72 hr

Collect cells or Mock (WT) Nox1-disrupted

Cardiac fibroblast (CF)

10 -4 - 10 6 /ml

H9c2

B

Cell proliferation & Gene expression

CF:Nox1(+/Y) CF:Nox1(-/Y)

* *

200

150

100

50

200

*

180 Proliferation (% of Cont)

Proliferation (% of Cont)

250

160

#

140

*

#

#

120 100 80 60 40 20

0

Cont

Homo

0

Cont 106

105 Mock

104

106

105

104

106

Clone1

105

104

106

104 (cells/ml)

105 Clone3

Clone2

H9c2 Homo

C 200

140

600

*

FN/β-actin (% of Cont)

150

500 Col3a1/β-actin (% of Cont)

Col1a1/β-actin (% of Cont)

120 100

100

50

80 60 40

0 Cont Mock Clone1 Clone2 Clone3

H9c2 Homo

#

300

* 200

#

#

100

20 0

400

0 Cont

Mock Clone1 Clone2 Clone3

H9c2 Homo

Cont

Mock Clone1 Clone2 Clone3

H9c2 Homo

Fig. 5. Disruption of Nox1 in H9c2 blunted proliferation and expression of collagen 3a1 in cardiac fibroblasts. (A) A diagram for experimental protocol. Various concentrations of homogenate prepared from H9c2 clones harboring wild-type or disrupted Nox1 were placed on cardiac fibroblasts isolated from Nox1+/Y or Nox1-/Y. (B) Proliferation of fibroblasts were determined by quantitation of the ATP levels. Homogenate (Homo) of wild-type H9c2 similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % of control. N =4 per group. (C) Levels of fibronectin (FN), collagen 1a1 (Col1a1), and collagen 3a1 (Col3a1) mRNAs in fibroblasts co-incubated with homogenate from H9c2 clones harboring wild-type (Mock) or disrupted Nox1 (Clone 1-3). N = 3 per group. Similar results were obtained in three independent experiments. *P < 0.05 vs. control; #P < 0.05.

CF:Nox1(+/Y) CF:Nox1(-/Y)

B

180

**

Proliferation (% of Cont)

160 Culture supernatant Serum-free 72hr or Mock (WT) Nox1-disrupted

Cell proliferation

Cardiac fibroblast (CF)

H9c2 24 hr

*

180 Proliferation (% of Cont)

A

140 120 100 80 60 40 20

160

#

*

140 120

#

*

#

*

100 80 60 40 20

0

0 Cont

C

Mock

Cont

Mock Clone 1 Clone 2 Clone 3

Culture supernatant Homogenate

WT H9c2

Serum-free

106 /ml

72hr

or Mock (WT) Nox1-disrupted

Cardiac fibroblast (CF)

H9c2 24 hr

D

Cell proliferation

CF:Nox1(+/Y) CF:Nox1(-/Y) 300

*

300

*

250 200 150 100 50

0 Sup H9c2 Homo

(-)

Mock

(-)

(+)

Proliferation (% of Cont)

Proliferation (% of Cont)

350

*

250

#

*

200

#

*

#

*

#

*

150 100 50 0 Sup

(-)

(-)

H9c2 Homo

(-)

(+)

Mock Clone 1 Clone 2 Clone 3 (+)

(+)

(+)

(+)

Fig. 6. Disruption of Nox1 in H9c2 blunted proliferation of cardiac fibroblasts induced by culture supernatant. (A) A diagram for experimental protocol. The supernatant fraction of the serum-free medium of cultured H9c2 clones harboring wild-type or disrupted Nox1 was placed on cardiac fibroblasts isolated from Nox1+/Y or Nox1-/Y. (B) Culture supernatant of wild-type H9c2 similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % of control. N =4-5 per group. (C) A diagram for experimental protocol. Wild-type myocyte homogenate was first placed on intact H9c2 clones harboring wild-type or disrupted Nox1 for 24 hours. The culture supernatant of these H9c2 cells containing wild-type myocyte homogenate was then placed on cardiac fibroblasts isolated from either genotype. (D) Culture supernatant (Sup) of wild-type H9c2 (Mock) similarly increased proliferation of fibroblasts isolated from both genotypes. Data are expressed as % of control. N =4-5 per group. Similar results were obtained in four independent experiments. *P < 0.05 vs. control. #P < 0.05

Injury

DAMPs

Nox1

(DOX)

TLR4 ROS

Cardiomyocyte

S100A1 DAMPs?

Cardiomyocyte TLR4

Cardiac fibroblast

ECM production

Proliferation

Cardiac remodeling

Fig. 7. A schematic diagram of NOX1-mediated cardiac remodeling. Doxorubicin (DOX), damage-associated molecular pattern molecules (DAMPs), toll-like receptor 4 (TLR4), extracellular matrix (ECM), reactive oxygen species (ROS).

Injury

DAMPs

Nox1

TLR4 ROS

Cardiomyocyte

S100A1 DAMPs?

Cardiomyocyte

TLR4

Cardiac fibroblast st

ECM production

Proliferation

Remodeling