An Anthelmintic Drug, Pyrvinium Pamoate, Thwarts Fibrosis and Ameliorates Myocardial Contractile Dysfunction in a Mouse Model of Myocardial Infarction

Metabolic adaptation to limited supplies of oxygen and nutrients plays a pivotal role in health and disease. Heart attack results from insufficient delivery of oxygen and nutrients to the heart, where cardiomyocytes die and cardiac fibroblasts proliferate – the latter causing scar formation, which impedes regeneration and impairs contractility of the heart. We postulated that cardiac fibroblasts survive metabolic stress by adapting their intracellular metabolism to low oxygen and nutrients, and impeding this metabolic adaptation would thwart their survival and facilitate the repair of scarred heart. Herein, we show that an anthelmintic drug, Pyrvinium pamoate, which has been previously shown to compromise cancer cell survival under glucose starvation condition, also disables cardiac fibroblast survival specifically under glucose deficient condition. Furthermore, Pyrvinium pamoate reduces scar formation and improves cardiac contractility in a mouse model of myocardial infarction. As Pyrvinium pamoate is an FDA-approved drug, our results suggest a therapeutic use of this or other related drugs to repair scarred heart and possibly other organs.


Introduction
Ischemic heart disease accounts for ~13 % of deaths worldwide and is the leading cause of death for both men and women in all developed countries. Myocardial infarction, commonly known as a heart attack, is primarily caused by the occlusion of blood supply to a part of the heart, causing the cardiomyocytes to die. Cardiomyocyte death is followed by fibrosis (i.e., scar formation) resulting from active proliferation and migration of cardiac fibroblasts and excessive deposition of extracellular collagen fibers [1,2]. Cardiac fibrosis has adverse effects on cardiac function, and furthermore, interferes with regeneration of cardiomyocytes and cardiac vascularization, the essential processes for restoring function of scarred heart [1,2]. Thus, an effective therapeutic intervention of cardiac fibrosis is very much in need for full regeneration of the heart injured by heart attack and other cardiac diseases. However, this therapeutic goal remains unmet despite several decades of extensive studies.
Conventional therapeutic approaches towards interfering with cardiac fibrosis have been to inhibit proliferation and/or migration of cardiac fibroblasts [1,2]. These include targeted inhibition of activities of fibrogenic cytokines/proteins such as transforming growth factor beta (TGF-β), connective tissue growth factor (CTGF/CCN2), platelet-derived growth factor (PDGF) and endothelin-1. Furthermore, unusually elevated intracardial level of angiotensin II (Ang II) was found in overloaded hearts with fibrosis and that drugs inhibiting angiotensin signaling pathway has been used to reduce cardiac fibrosis [3,4]. However, none have lead to desirable therapeutic consequences. Thus, as a completely new class of therapeutic targets, we considered altered intracellular metabolism in cardiac fibroblasts under ischemia. We hypothesized that, in ischemic heart diseases, cardiac fibroblasts survive and actively proliferate in a metabolically challenging microenvironment with limited amounts of oxygen and nutrients. Based on this hypothesis, we postulate that impeding such metabolic adaptation in cardiac fibroblasts serves as a novel and effective therapeutic target to attenuate cardiac fibrosis in ischemic heart diseases including myocardial infarction. If successful, this would become an important step towards curing this life threating and most prevalent human disease.
An anthelmintic drug, pyrvinium pamoate (PP), is an inhibitor of NADH-fumarate reductase (NADH-FR) activity in the anaerobic respiratory chain in mitochondria of parasitic worm [5]. In cancer cells, it has been recently reported that PP also inhibits canonical Wnt, unfolded protein response (UPR), androgen receptor and autophagy signals [6][7][8][9][10]. It has been also shown that PP can compromise the survival of cancer cells under glucose-starvation condition [11][12][13]. These previously studies suggested to us that PP could potentially starve cardiac fibroblasts to death in glucose-deficient microenvironment and thwart cardiac fibrosis.
In this study, we show that PP can in fact disables the survival of cardiac fibroblasts specifically under glucosedeficient media in vitro and also thwarts fibrosis and ameliorate myocardial contractile dysfunction in a mouse model of myocardial infarction.

Results and Discussion
First, we tested cytotoxicity of PP on cardiac fibroblasts cultured under the limited oxygen, glucose and glutamine (3 % oxygen, 300 μM glucose, <100 μM glutamine) (referred to as O 2 low /Glc low /Gln low or ischemia) and normal (20 -21 % oxygen, 25 mM glucose, 4 mM glutamine) (referred to as "normal") conditions for 72 hr ( Figure 1A). PP was found to impose specific toxicity on cardiac fibroblasts in ischemia (IC50=9.5 nM) ( Figure 1A). Furthermore, we found that the normal level of O 2 failed to rescue the survival of cardiac fibroblasts treated with PP ( Figure 1B). The survival of cardiac fibroblasts treated with PP in ischemia could not be rescued by normal levels of O 2 (20 -21%) and glutamine (4 mM) ( Figure 1B). In contrast, their survival was rescued partially or completely by supplementing O 2 (20 -21%)/glucose (25 mM) ( Figure 1C) or glucose (25 mM)/glutamine (4 mM) ( Figure 1D), respectively. These results demonstrate that the cytotoxic effect of PP on cardiac fibroblasts specifically under glucose-and glutaminedeficient condition (Glc low /Gln low ). Such PP effect on cardiac fibroblasts is similar to that on cancer cells [11].
We next tested the effectiveness of PP on cardiac fibroblasts in vivo by examining whether PP could thwart fibrosis by attenuating fibroblast survival in a mouse model of myocardial infarction (Figures 2 -8). In this model, left coronary artery was permanently ligated by surgical ligation. Within 24 hr following the occlusion, most of the cardiomyocytes in left ventricle die due to ischemia, and cardiac fibroblasts begin to proliferate and migrate at 3 -4 days-post-ligation, and the number of cardiac fibroblasts peaks at 7 days-post-ligation ( Figure 2). Secretion and deposition of collagen fibers by these activated cardiac fibroblasts cause fibrosis ( Figure 2). The oral administration of PP was initiated at 1 day-post-ligation (i.e. after most of the cardiomyocytes die due to ischemia) and continued daily (Figure 2). At 4 days-post-ligation, the number of proliferating cells in the ischemic region (i.e. infarcted zone) was quantified by staining a nuclear marker, Ki67, associated with cell proliferation using anti-Ki67 antibodies ( Figure 3A, B). The result shows significant reduction of Ki67 + cells in the infarcted zones of the heart treated with PP ( Figure 3A, B). This result indicates that PP suppresses cell proliferation in the ischemic region at the beginning of fibrosis phase. The number of cardiac myofibroblasts that actively participates in fibrosis was identified and quantified by staining them with anti-α-smooth muscle actin (αSMA) antibodies in both border and infarcted zones ( Figure 4A, B). The result demonstrated a significant reduction in the αSMA + cardiac myofibroblasts at 7 days-postligation in both border and infarcted zones where active cardiac fibrosis were present ( Figure 4A, B). By 14 days-post-ligation, we found the reduced extent of fibrosis in the heart treated with PP ( Figure 5A, B). This reduced fibrosis was not due to the reduced infarct area (due to reduced cardiomyocyte death/ necrosis), as no differences in the infarct size were found between the hearts treated with vehicle or PP at 4-days-postligation ( Figure 5A, B). The reduced fibrosis in the infarcted heart treated with PP was further confirmed by the decreased level of hydroxyproline (quantitative indicator of mature collagen molecules) content in this heart ( Figure 6). These results demonstrate that PP reduces the number of proliferating cardiac fibroblasts in the ischemic heart tissue in vivo, which results in the reduced fibrosis. Furthermore, by 14 days-post-ligation, the infarcted heart treated with PP exhibited improved left ventricular function as indicated by increased ejection fraction, as compared to that treated with vehicle only (Figure 7). These results show that the PP-treatment, not only reduces fibrosis, but also ameliorates myocardial contractile dysfunction, instead of causing cardiac rupture.
One possible explanation for this beneficial (i.e., improved cardiac function), but not adverse (i.e., cardiac rupture), effect of the PP-induced reduction of fibrosis is that the decreased scar formation provided a permissive tissue environment for more efficient angiogenesis that is a favorable condition for contractile function of the remaining cardiomyocytes. In support of this possibility, we found a statistically significant increase in vascular density in the scarred heart treated with PP ( Figure  8A, B).
It was previously shown that PP had very little effect on cardiac remodeling, and was unable to reduce cardiac fibrosis or improve cardiac function [14]. The discrepancy between their result and ours may be due to the different protocols of PP administration to the mice. They directly injected PP into the heart at the same time when the coronary artery was ligated, thus several days prior to the proliferation of cardiac fibroblast and the initiation of fibrosis [14]. In our protocol, PP was orally administered beginning at 1-day-post-ligation, the time when most of the cardiomyocytes had already died, and the PPadministration was continued daily throughout the period of cardiac fibroblasts proliferation and fibrosis ( Figure 2). It is possible that the beginning the PP-administration at one or two days later after the cardiomyocytes are dead, combined with the continuous daily administration therefore, may be the key to bring the beneficial effect of PP on thwarting fibrosis and improving cardiac function after myocardial infarction. It is important to note that the infarcted heart be treated by PP one or two days later from the time when the coronary artery was ligated to minimize any adverse effects on cardiomyocytes in ischemic region. We did not unnecessarily want to enhance the death of cardiomyocytes by PP by administrating this drug while the cardiomyocytes were still dying. For this reason, we began the PP-administration on the next day after the ligation (i.e., 1-day-post-ligation) (Figure 2).
In a previous study, it has been shown that the oral administration of PP with the same dosage as ours resulted in the clearance of the drug from the circulation within 12 hours [7]. Therefore, it would be of interest to quantitatively measure precise pharmacodynamics of PP in the infarcted mice, as such future study may allow pinpoint exactly the mechanism by which PP attenuate cardiac fibrosis following myocardial infarction in mice.
PP inhibits NADH-FR activities, and four non-NADH-FR pathways: canonical Wnt, UPR, androgen receptor and autophagy signaling pathways [6][7][8][9][10]15]. Which of these signaling pathways play the central role(s) in thwarting fibrosis in vivo remains for the future studies. In our study, we demonstrate that the inhibition of cardiac fibroblast proliferation by PP is dependent on glucose, and glutamine with lesser degree, but not oxygen concentrations ( Figure 1B, C, D). Hence, it is less likely that the inhibitory effect of PP on cardiac fibrosis is mediated by the blocking of NADH-FR activity, as the activity is a part of the anaerobic respiratory chain in mitochondria. Androgen receptor knockout mice were previously shown to exhibit increased fibrosis [16], suggesting that the inhibition of androgen receptor signaling could not account for the effect of PP in a mouse model of myocardial infarction shown in this paper. While cancer cells are widely known as "glucose-addicted", they are also relatively resistant to glucose-starvation, the characteristics that they take advantage of for tumorigenesis [17][18][19][20]. Cardiac fibroblasts are also capable of surviving and proliferating under glucose-deficient condition and they exploit this unique ability to form fibrosis that impedes cardiac regeneration and interferes with normal cardiac contractility. Cancer cells also take advantage of their ability to survive under glucose-deficient condition for tumorigenesis [17][18][19][20]. These similar behaviors of cancer cells and cardiac fibroblasts may suggest that they share a related mechanism underlying their metabolic adaptation to nutrient stress. In this context, it is intriguing that PP also effectively kill several types of glucosestarved cancer cells, but its mode of action has been controversial [6,9,[11][12][13]. Therefore, it is of interest to investigate a putative mechanism that might be shared between cardiac and cancer cells for their survival under glucose-starvation, which is targeted and inhibited by PP.
In this paper, we demonstrate that PP compromise the survival of cardiac fibroblasts under glucose-deficient condition in vitro (Figure 1), and also attenuates fibrosis and ameliorates cardiac contractility without causing cardiac rupture in a mouse model of myocardial infarction (Figures 3 -8). As PP is an FDA-approved anthelmintic drug, its therapeutic applications are likely to pose fewer problems. Hence, PP or other related drugs may serve as a novel therapeutics to thwart fibrosis and to facilitate the repair of scarred heart and possibly other scarred organs/tissues.

Isolation of Cardiac Fibroblasts
Cardiac fibroblasts were prepared as previously described [21]. Heart tissues were minced, washed by cold D-PBS (Ca 2+ , Mg 2+ free), and suspended in 5 ml of 1 % collagenase (Worthington CLS-2) /DPBS (with Ca 2+ , Mg 2+ ). Collagenase treatment was performed at 37 °C for 30 min. Each collagenase-digested sample was treated with 4 ml of 0.25 % trypsin (Invitrogen) at 37 °C for 5 min. Collagenase/trypsintreated samples were further dissociated by pipetting up and down 50 times. Each tissue suspension was transferred to a 50 ml tube containing 20 ml of DMEM (Nacalai) supplemented with 10 % FBS, 1 % penicillin, 1 % streptomycin. Dissociated cells were centrifuged at 100 x g for 5 min at 4 °C and were suspended into DMEM supplemented with 20 % FBS, 1 % penicillin, 1 % streptomycin and cultured in the same media for approximately one week before being stored.

Cell Culture Experiments
Cardiac fibroblasts were plated on Collagen-Coated Dish (IWAKI). After 24 hr culture in the culture media (high glucose   A. An effect of PP on αSMA + myofibroblasts in the infarcted heart. αSMA + myofibroblasts (Red) and DAPI + cells (dark blue) in the heart at day 4 and day 7 after the coronary artery ligation. The representative images from the mouse heart treated with vehicle (DMSO) (-PP) or PP (+PP) are shown. Scale bars: 1 mm. B. Quantitation of αSMA + myofibroblasts. The number of αSMA + myofibroblasts was counted from sections from vehicle treated (-PP) (n=3 mice for day 4, n=5 mice for day 7) and PP treated (+PP) (n=4 mice for day 4 and day 7) mice was counted in both border and infarcted zones and shown as % (αSMA + / DAPI + ). The αSMA + area was also measured and shown as % (αSMA + area / whole section area).  Masson's trichorme stained heart sections at day 4 and day 14 post-ligation. Scale bars: 1 mm. B. Quantitation of infarct size and fibrosis area. The infarct size was measured by quantitating the necrotic area of the heart sections prepared from the vehicle treated (-PP) (n=3 mice) and PP treated (+PP) (n=4 mice) mice at day 4 post-ligation, and shown as % (necrotic area / whole heart area) (left). The fibrosis size was measured by Masson's trichrome staining method using the heart sections prepared from the vehicle treated (-PP) (n=3 mice) and PP treated (+PP) (n=4 mice) mice at day 14 postligation, and shown as % (fibrosis area stained blue / whole heart area). doi: 10.1371/journal.pone.0079374.g005 glutamine were present, of which final concentration was determined as 300 μM glucose, <100 μM glutamine. Low oxygen condition was accomplished by using Hypoxia chamber (Veritas) and premixed gas chamber (3 % O 2 , 10 % CO 2 , 87 % N 2 ). Viable cells were identified by exclusion of Trypan Blue dye (GIBCO) and counted.

Animals
All animal protocols were approved by the Animal Care and Use Committee of Nara Institute of Science and Technology (Permit Numbers: 1123, 1223 and 1312). All experiments were conducted with 8-to 12-weeks old male ICR mice (CLEA Japan, Inc., Tokyo, Japan).

Myocardial Infarction
Coronary artery ligation was performed on ICR male mice (8-12 weeks-old) produce myocardial infarction as previously described [21]. Mice were anaesthetized by inhalation of 2 to 2.5% vaporized isoflurane (Abbott Japan) and intubated with a 20-gauge intravenous catheter. Animals were ventilated with a volume-controlled respirator (Harvard Apparatus) with 200 μl per cycle at a respiratory rate of 110 cycles min -1 . After thoracotomy, the left coronary artery was ligated with an 8-0 nylon suture 1-2 mm below the tip of the left auricle. Occlusion was confirmed by the change of color (pallor) of the anterior   operated mice underwent the same procedure without tying the suture, but passing it below the coronary artery. Pyrvinium pamoate (PP) was suspended to a final concentration of 400 μg/ml in 4 % DMSO in saline, and the mice were force-fed intragastrically with 0.5 ml of the suspension or vehicle (4 % DMSO in saline). The oral administration was initiated at 1 daypost-ligation and continued daily up to 14 days-post-ligation.

Echocardiography
Echocardiography was performed on infarcted and shamoperated mice on days 1, 7 and 14 using Toshiba Diagnostic Ultrasound System machine (Aplio MX SSA-780A). In each instance, mice were anaesthetized with 2 to 2.5 % isoflurane and parasternal short-axis 2D-mode and M-mode views were recorded. The short-axis view was used to measure anteroposterior internal diameter, anterior wall thickness and posterior wall thickness at end-diastole and end-systole at the mid-papillary level. Left ventricular systolic function was assessed by ejection fraction: %EF =[(EDV -ESV) / EDV] × 100. All measurements and evaluations were performed in a blind controlled study.

Hydroxyproline Assay
Hydroxyproline content of heart tissues was measured as previously described [21].

Statistical Analysis
Statistically analyses were performed with Excel using Student's t-test. Significance levels were set at and indicated as ★ p<0.05, ★★ p<0.01, ★★★ p<0.001.