Premedication with pioglitazone prevents doxorubicin-induced left ventricular dysfunction in mice

Abstract

Background: Doxorubicin (DOX) is widely used as an effective chemotherapeutic agent for cancers; however, DOX induces cardiac toxicity, called DOX-induced cardiomyopathy. Although DOX-induced cardiomyopathy is known to be associated with a high cumulative dose of DOX, the mechanisms of its long-term effects have not been completely elucidated. Pioglitazone (Pio) is presently contraindicated in patients with symptomatic heart failure owing to the side effects. The concept of drug repositioning led us to hypothesize the potential effects of Pio as a premedication before DOX treatment, and to analyze this hypothesis in mice.

Methods: First, for the hyperacute (day 1) and acute (day 7) DOX-induced dysfunction models, mice were fed a standard diet with or without 0.02% (wt/wt) Pio for 5 days before DOX treatment (15 mg/kg body weight [BW] via intraperitoneal [i.p.] administration). The following 3 treatment groups were analyzed: standard diet + vehicle (Vehicle), standard diet + DOX (DOX), and Pio + DOX. Next, for the chronic model (day 35), the mice were administrated DOX once a week for 5 weeks (5 mg/kg BW/week, i.p.).

Results: In the acute phase after DOX treatment, the percent fractional shortening of the left ventricle (LV) was significantly decreased in DOX mice. This cardiac malfunction was improved in Pio + DOX mice. In the chronic phase, we observed that LV function was preserved in Pio + DOX mice.

Conclusions: Our findings may provide a new pathophysiological explanation by which Pio plays a role in the treatment of DOX-induced cardiomyopathy, but the molecular links between Pio and DOX-induced LV dysfunction remain largely elusive.

Keywords: Anticancer agent; Cardiac toxicity; LV dysfunction; Thiazolidinediones.

Fig. 1

Pio pretreatment partially prevented DOX-induced LV dysfunction in the acute phase (day 7), but not in the hyperacute phase (day 1). a Diagram showing the treatment course of each group. For the hyperacute and acute phases, mice were divided into 3 groups. Vehicle mice were fed a standard diet and treated with vehicle instead of DOX. DOX mice were fed a standard diet and treated with a single DOX bolus (15 mg/kg of body weight [BW]) on day 0. Pio + DOX mice were orally administered with Pio (0.02% [wt/wt]) from day − 5 as premedication, and then treated with a DOX bolus on day 0. All measurements were acquired and hearts were excised on both day 1 (hyperacute phase) and day 7 (acute phase). The horizontal open arrows represent standard diet intake, and the horizontal closed arrow indicates Pio intake. The vertical arrows indicate treatment with vehicle or DOX. b Kaplan-Meier survival curves of Vehicle (n = 15, black broken line), DOX (n = 15, blue line), or Pio + DOX mice (n = 15, red line). c Representative echocardiographs of Vehicle (left panel), DOX (middle panel), and Pio + DOX (right panel) mice. d, e, f, g, h Summary data of LVEDD, LVESD, %FS, HR, and LV wall thickness. Day 1: Vehicle (open upward triangles, n = 10), DOX (open circles, n = 10), and Pio + DOX (closed circles, n = 10). Day 7: Vehicle (n = 25), DOX (n = 26), and Pio + DOX (n = 25). i, j, k Summary data of BW, LV/BW, and lung/BW. Data are shown as means ± SEs. *P < 0.05 vs. Vehicle, †P < 0.05 vs. DOX. DOX, doxorubicin; Pio, pioglitazone; LV, left ventricle; LVEDD, LV end-diastolic diameter, LVESD, LV end-systolic diameter; %FS, percent fractional shortening

Fig. 2

Pio improved vacuolization but showed no significant biochemical or physiological changes in the acute phase of DOX-induced LV dysfunction mice (day 7). a Representative images of hematoxylin-eosin (HE) staining of a heart from each group. Scale bar, 50 μm. b Summary data of the cross-sectional area (CSA) of the left ventricle on HE staining. n = 3 for each group. c Number of vacuolization in cells in the HE stained samples of each sample, calculated as the percentage of cells containing vacuoles, i.e., 0: none; 1: less than 25%; 2: 25–50%; and 3: more than 50%. d Representative images of Masson trichrome (MT) staining of a heart from each group. Scale bar, 50 μm. e Summary data of fibrosis area in the MT-stained samples. n = 3 for each group. f, g Quantitative analysis of the gene expression levels of interleukin-1beta (Il1b) and tumor necrosis factor-alpha (Tnfa) in the heart on day 7 (n = 7–10 for each group). h Hydrogen peroxide (H₂O₂) release originating from non-fatty-acid substrates from isolated mitochondria in hearts during state 3 respiration (i.e. oxidative phosphorylation) using the complex I- and II-linked substrates glutamate, malate, and succinate (n = 3 for each group). i Levels of mitochondrial iron contents (n = 4–8 for each group). j Mitochondrial respiration using non-fatty-acid substrates in isolated mitochondria from the heart during state 3 respiration using the complex I- and II-linked substrates glutamate, malate, and succinate (n = 3 for each group). k Mitochondrial respiration with fatty-acid substrate (octanoyl-l-carnitine) in isolated mitochondria from the heart during state 3 respiration (n = 3 for each group). Data are shown as means ± SEs. *P < 0.05 vs. Vehicle. NS, not significant

Fig. 3

Pio pretreatment prevented LV dysfunction in the chronic phase of DOX-induced LV dysfunction mice (35 days). a Diagram showing the treatment course of each group. For the chronic phase, mice were divided into 3 groups. Vehicle mice were fed a standard diet and treated with vehicle instead of DOX. DOX mice were fed a standard diet and treated with weekly DOX injections for 35 days (25 mg/kg cumulative dose). Pio + DOX mice were orally administered with Pio (0.02% [wt/wt]) from day − 5 as premedication and were treated with weekly DOX injections for 35 days. Measurements of BW and food intake were performed every week. Cardiac function was evaluated by echocardiography and hearts were excised on day 35. Open horizontal arrows represent standard diet intake, and the closed horizontal arrow indicates Pio intake. The vertical arrows indicate treatment with vehicle or DOX. b Representative echocardiographs from a mouse in each group; Vehicle (left panel), DOX (middle panel), and Pio + DOX (right panel). The vertical arrows in the echocardiograph indicate LVEDD (left arrow) and LVESD (right arrow). c, d, e, f, g Summary data of LVEDD, LVESD, %FS, HR, and LV wall thickness. Vehicle (open upward triangles, n = 7), DOX (open circles n = 14), and Pio + DOX (closed circles, n = 14). h, i, j Summary data of BW, LV/BW, and lung/BW (n = 7–14 for each group). Data are shown as means ± SEs. *P < 0.05 vs. Vehicle; †P < 0.05 vs. DOX