- Email: [email protected]
Pioglitazone, a peroxisome proliferator-activated receptor-gamma activator, attenuates atrial fibrosis and atrial fibrillation promotion in rabbits with congestive heart failure
Pioglitazone, a peroxisome proliferator-activated receptor-gamma activator, attenuates atrial fibrosis and atrial fibrillation promotion in rabbits with congestive heart failure Masayuki Shimano, MD,* Yukiomi Tsuji, MD,† Yasuya Inden, MD,* Kazuhisa Kitamura, MD,* Tomohiro Uchikawa, MD,* Shuji Harata, MD,* Stanley Nattel, MD, FHRS,‡ Toyoaki Murohara, MD* From the *Department of Cardiology, Nagoya University Graduate School of Medicine, Nagoya, Japan, †Department of Cardiovascular Research, Research Institute of Environmental Medicine (RIEM), Nagoya University, Nagoya, Japan, and ‡ Research Center and Department of Medicine, Montreal Heart Institute and University of Montreal and Department of Pharmacology, McGill University, Montreal, Canada. BACKGROUND The peroxisome proliferator-activated receptor-␥ (PPAR-␥) activator pioglitazone antagonizes angiotensin II actions and possesses anti-inflammatory and antioxidant properties in vitro. There is evidence that pioglitazone improves ventricular remodeling in some experimental models. OBJECTIVE The purpose of this study was to assess the effects of pioglitazone on arrhythmogenic atrial structural remodeling versus the effects of the angiotensin II type 1 receptor blocker candesartan in a rabbit model of congestive heart failure. METHODS Rabbits subjected to ventricular tachypacing at 380 to 400 bpm for 4 weeks in the absence and presence of treatment with pioglitazone, candesartan, and combined pioglitazone and candesartan were assessed by electrophysiologic study, atrial fibrosis measurements, and cytokine expression analyses. RESULTS Atrial fibrillation (AF) lasting longer than 2 seconds was induced in no nonpaced controls but in all ventricular tachypacing-only rabbits (mean duration of AF: 8.0 ⫾ 1.4 seconds). Pioglitazone reduced the duration of AF (3.5 ⫾ 0.2 seconds, P ⬍.05) and attenuated atrial structural remodeling, with significant re-
Introduction Atrial fibrillation (AF) is a common arrhythmia seen in clinical practice. Congestive heart failure (CHF) is a common cause of AF.1 Experimental CHF causes atrial structural remodeling characterized by atrial interstitial fibrosis and local conduction abnormalities, which create a substrate for AF maintenance.2 Angiotensin II is an important mediator of the pathologic process.3,4 Both angiotensin-converting enzyme inhibitors and angiotensin II type 1 receptor blockers reduce CHF-related structural remodeling and AF
Supported by Grant-in-Aid for Scientific Research (18590766-00) from Japan Society for the Promotion of Science to Dr. Inden. Address reprint requests and correspondence: Dr. Yukiomi Tsuji, Department of Cardiovascular Research, RIEM, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8601 Japan. E-mail address: [email protected]. (Received October 13, 2007; accepted December 8, 2007.)
ductions in interatrial activation time (50 ⫾ 2 ms vs 41 ⫾ 2 ms, P ⬍.05) and atrial fibrosis (16.8% ⫾ 0.8% vs 10.9% ⫾ 0.7%, P ⬍.05; control 1.6% ⫾ 0.2%), effects comparable to those of candesartan (duration of AF: 3.0 ⫾ 0.2 seconds; activation time 44 ⫾ 2 ms; fibrosis: 9.4% ⫾ 0.6%). Both pioglitazone and candesartan reduced transforming growth factor-1, tumor necrosis factor-␣, and activated extracellular signal-regulated kinase expression similarly, but neither affected p38-kinase or c-Jun Nterminal kinase activation. The effects of combined pioglitazone and candesartan therapy were not significantly different from the effects of pioglitazone or candesartan alone. CONCLUSION Pioglitazone can attenuate congestive heart failure-induced atrial structural remodeling and AF promotion, with effects similar to those of candesartan. PPAR-␥ may be a potential therapeutic target for human AF. KEYWORDS Arrhythmia; Atrium; Heart failure; Remodeling; Antiarrhythmic agents (Heart Rhythm 2008;5:451– 459) © 2008 Heart Rhythm Society. All rights reserved.
promotion in a canine model3,5 and prevent AF in patients with left ventricular (LV) dysfunction6,7 and hypertension.8 The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-activated transcription factor that modulates gene expression for pathways involved in lipid storage and adipogenesis.9 The mechanism by which PPAR-␥ regulates myocardial metabolism is unclear. Studies have shown that PPAR-␥ activators such as pioglitazone antagonize the actions of angiotensin II and possess anti-inflammatory and antioxidant properties in vitro,10 –15 and that pioglitazone prevents cardiac remodeling in some experimental models.16,17 Shiomi et al16 reported that pioglitazone improves LV remodeling and function in mice with CHF after myocardial infarction, with a concomitant decrease in interstitial fibrosis and reduced expression of inflammatory cytokines. Iglarz et al17 demonstrated that pioglitazone suppresses cardiac fibrosis in mineralocorticoid-dependent hypertensive
1547-5271/$ -see front matter © 2008 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2007.12.010
452 rats. However, studies investigating the effects of pioglitazone in atrial remodeling have not been reported. This study was designed to assess the effects of pioglitazone on atrial structural remodeling and AF promotion and to compare them directly with the effects of the angiotensin II type 1 receptor blocker candesartan in a rabbit model of CHF. The study also tested the hypothesis that a combination of pioglitazone and candesartan would show greater effects on atrial structural remodeling than either agent alone.
Methods Animal handling followed the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23), and procedures were approved by the Animal Experimentation Ethics Committee of Nagoya University.
Animal model and experimental protocol Male New Zealand white rabbits (weight 2.8 –3.3 kg) were used. Programmable right ventricular pacemakers (Medtronic, Inc., Minneapolis, MN, USA) were implanted as described previously.18 After the animals recovered from surgery, the pacemaker was programmed to pace at 380 to 400 bpm for 4 weeks. ECGs were monitored once per week to adjust the pacing rate to the maximum rate, thus allowing for 1:1 capture in each rabbit. The mean pacing rate achieved for all rabbits was 386 ⫾ 0.9 bpm. At the end of 4 weeks of ventricular tachypacing, echocardiographic and open chest electrophysiologic studies were performed under anesthesia. After electrophysiologic studies, rabbits were euthanized with pentobarbital 40 mg/kg IV, and whole hearts were immersed in 10% formaldehyde for subsequent histologic study.
Groups Results from 15 ventricular tachypaced rabbits without any treatment (ventricular tachypacing-only group) and 15 nonpaced control rabbits were compared with results from rabbits subjected to ventricular tachypacing during daily oral administration of pioglitazone 5 mg/kg (PIO, n ⫽ 15), candesartan 2 mg/kg (CAN, n ⫽ 15), or combined pioglitazone 5 mg/kg and candesartan 2 mg/kg (PIO⫹CAN, n ⫽ 15), beginning 14 days before ventricular tachypacing onset and continuing throughout the study period.
Electrophysiologic study For electrophysiologic study on day 28, rabbits were anesthetized with ketamine hydrochloride 35 mg/kg IM and xylazine 3 mg/kg IM and ventilated mechanically with isoflurane-containing room air (0.5%/2 L/min). The pacemaker was deactivated, and body temperature was maintained at 37°C. Median sternotomy was performed, and custom-made bipolar electrodes were attached to the right atrial appendage (RAA) and left atrial appendage (LAA) for recording and stimulation. A programmable stimulator (Nihon Kohden, Tokyo, Japan) was used to deliver twice threshold currents at 2-ms pulse duration. Effective refractory periods (ERPs) were measured at the RAA and LAA at basic cycle lengths of 300, 200, and 150 ms with a train of
Heart Rhythm, Vol 5, No 3, March 2008 20 basic stimuli (S1) followed by a premature extrastimulus (S2) at 2-ms decrements. The effective refractory period was defined as the longest S1-S2 interval that failed to capture. The interatrial activation time was defined as the time from RAA pacing spike to LAA activation. AF induction was tested by applying burst pacing (cycle length 70 – 140 ms) at 5-ms increments and four times threshold current. AF was defined as a rapid (⬎500/min) irregular atrial rhythm lasting longer than 2 seconds. Durations of AF from all episodes were averaged in each rabbit to provide a representative mean duration of AF value.
Histology Transverse sections cut at 5-m intervals and stained with Masson trichrome at the midatrial level were used to evaluate fibrotic deposition in rabbit atria. Microscopic images were digitized with Viewfinder Light Image (Mitani Corp., Tokyo, Japan). To quantify the percent fibrosis area at right atrial (RA) and left atrial (LA) free walls, the blue pixel content of digitized images was measured relative to total tissue area using the image analyzer Win ROOF (Mitani Corp.). Blood vessels and perivascular interstitial tissues were excluded from fibrosis quantification. The average number of transverse sections of LA free wall used for quantitative analysis was 54 ⫾ 1.0 for each rabbit.
Western blots Studies in a canine model of ventricular tachypacing-induced CHF showed that molecular changes related to atrial structural remodeling were observed early after initiation of ventricular tachypacing.3,4,19 To investigate effects of pharmacologic interventions on the expression of transforming growth factor-1 (TGF-1), tumor necrosis factor-␣ (TNF␣), and mitogen-activated protein kinase (MAPK) subfamilies including extracellular signal-regulated kinase (ERK), p38-kinase, and c-Jun N-terminal kinase (JNK), we prepared 12 rabbits subjected to ventricular tachypacing for 5 days in each group. Atrial tissue samples were fast-frozen in liquid nitrogen and stored at – 80°C. Protein extracts were prepared and immunoblotting was performed as previously described.18 Commercially available primary antibodies were used. Antibodies to TNF-␣ (sc1348; goat), phosphorylated (sc7973; mouse) and total p38 (sc7972; mouse), phosphorylated ERK (sc16982; goat), and phosphorylated JNK (sc12882; goat) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA); TGF-1 (T0438; mouse) from Sigma Chemical (St. Louis, MO); total ERK (9102; rabbit) and total JNK (9252; rabbit) from Cell Signaling Technology (Danvers, MA); and glyceraldehyde 3-phosphate dehydrogenase ([GAPDH] ab8245; mouse) from Abcam plc (Camdridge, UK). Band densities were quantified by densitometry, standardized to GAPDH, and normalized to the control sample.
Statistical analysis Data are expressed as mean ⫾ SEM. Statistical significance was assessed by one-way analysis of variance with Bonferroni correction. P ⬍.05 was considered significant.
Shimano et al Table 1 and CAN
Pioglitazone and Atrial Structural Remodeling
453
Characteristics of ventricular tachypacing rabbits in the absence and presence of treatment with PIO, CAN, or combined PIO NP (n ⫽ 12)
FS (%) LVEDd (mm) LVESd (mm) IVSd (mm) PWd (mm) LAd (mm) Systolic BP (mmHg) Blood Glucose (mg/dL)
42.0 14.4 8.4 3.1 3.1 8.1 119 334
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.0 0.3 0.3 0.1 0.1 0.2 4 12
VTP alone (n ⫽ 12) 9.8 18.7 16.9 2.8 2.8 13.8 109 355
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.7* 0.3* 0.3* 0.1 0.1 0.2* 5 16
PIO (n ⫽ 12) 10.2 18.7 16.8 2.8 2.8 13.9 110 300
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.9* 0.3* 0.2* 0.1 0.1 0.2* 4 15
CAN (n ⫽ 12) 10.6 18.1 16.2 2.9 2.9 13.1 97 347
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.7* 0.3* 0.2* 0.1 0.1 0.3* 5† 14
PIO⫹CAN (n ⫽ 12) 10.4 18.2 16.3 2.9 2.9 13.5 98 310
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.8* 0.4* 0.4* 0.1 0.1 0.2* 5† 14
Values are given as mean ⫾ SEM. NP indicates non-paced control; n, number of rabbits; VTP, ventricular tachypacing; PIO, pioglitazone; CAN, candesartan; FS, fractional shortening; LVEDd, left ventricular end-diastolic diameter; LVESd, left ventricular end-systolic diameter; IVSd, intraventricular septum diameter; PWd, posterior wall diameter of left ventricle; LAd, left atrial diameter. *P ⬍.05 vs NP; †P ⬍.05 vs VTP alone.
Results Animal characteristics Table 1 summarizes echocardiographic parameters, systolic blood pressure, and blood glucose concentrations in nonpaced control and ventricular tachypacing-induced CHF rabbits in the absence and presence of treatment with PIO, CAN, or PIO⫹CAN. Ventricular tachypacing for 4 weeks decreased fractional shortening and increased LV end-systolic and enddiastolic diameter and LA diameter, regardless of treatment with PIO, CAN, or PIO⫹CAN. Systolic blood pressure was reduced in rabbits treated with CAN or PIO⫹CAN. No significant differences in blood glucose concentration were observed among groups. All rabbits subjected to ventricular tachypacing had considerable amounts of ascites and pleural effusion. Sixty of 100 rabbits survived for 28 days and 40 died prematurely. Mortality rate for each group was 50% (15/30) for ventricular tachypacing-only, 40% (10/25) for PIO, 35% (8/23) for CAN, and 32% (7/22) for PIO⫹CAN rabbits (onetailed probability of Chi-square distribution [0.388] ⬎.05 for intergroup differences). Failing hearts displayed marked enlargement of all cardiac chambers (Figure 1).
Electrophysiologic study Figure 2A and B illustrates effective refractory periods at the RAA and LAA, respectively. Effective refractory
Figure 1
periods had a tendency to be increased by ventricular tachypacing but were not significantly different among groups. Interatrial activation time was increased by ⬎70% with ventricular tachypacing, from 29 ⫾ 1 ms in nonpaced controls to 50 ⫾ 2 ms in ventricular tachypacing-only rabbits. Activation times were significantly reduced by treatment with PIO, CAN, and PIO⫹CAN (from 50 ⫾ 2 ms in ventricular tachypacing-only group to 41 ⫾ 2 ms, 44 ⫾ 2 ms, and 41 ⫾ 1 ms, respectively, P ⬍.05 for each vs ventricular tachypacing-only group, Figure 2C), although they remained significantly greater than control values (approximately 45% larger, P ⬍.05). Examples of atrial arrhythmia induced by LAA burst pacing and mean data of duration of AF are shown in Figure 3A and B, respectively. Duration of AF was 8.0 ⫾ 1.4 seconds in ventricular tachypaced rabbits in the absence of drug therapy, whereas AF was not induced in nonpaced controls. Duration of AF was significantly reduced in PIO- and CAN-treated rabbits (3.5 ⫾ 0.2 seconds and 3.0 ⫾ 0.2 seconds, respectively) compared with ventricular tachypacing-only rabbits (P ⬍.05). PIO⫹ CAN–treated rabbits had duration of AF averaging 2.9 ⫾ 0.2 seconds, which was not significantly different from PIO alone or CAN alone.
Control heart and failing heart from a sham-operated and a ventricular tachypaced rabbit, respectively.
454
Heart Rhythm, Vol 5, No 3, March 2008
Figure 2 Electrophysiologic parameters measured at basic cycle lengths (BCLs) of 150, 200m and 300 ms in nonpaced control rabbits (NP, n ⫽ 15), ventricular tachypacing-alone rabbits (VTP, n ⫽ 15), and VTP rabbits treated with pioglitazone (PIO, n ⫽ 15), candesartan (CAN, n ⫽ 15), or a combination of pioglitazone and candesartan (PIO⫹CAN, n ⫽ 15). A, B: Effective refractory periods (ERPs) at right (RAA) and left atrial appendage (LAA). C: Interatrial activation time between LAA and RAA. *P ⬍.05, **P ⬍.01 vs NP, †P ⬍.05 vs VTP.
Atrial fibrosis Representative LA tissue sections from each group stained with Masson trichrome are shown in Figure 4. Atria appeared grossly normal in nonpaced controls (panel A), whereas thick fibrous tissue separated bundles of myofibers in ventricular tachypacing-only rabbits (panel B). Atrial fibrosis was clearly attenuated in rabbits treated with PIO, CAN, and PIO⫹CAN (panels C, D, and E, respectively). Panel F summarizes quantitative data from RA and LA. The area of fibrosis was considerably increased in both atria in ventricular tachypacing-only rabbits (11.4% ⫾ 0.9% and 16.8% ⫾ 0.8% in RA and LA, respectively) compared with nonpaced controls (2.0% ⫾ 0.2% and 1.6% ⫾ 0.2%). PIO suppressed fibrosis (7.6% ⫾ 0.6% and 10.9% ⫾ 0.7%) to a
level comparable with CAN therapy (7.2% ⫾ 0.6% and 9.4% ⫾ 0.6%). The combination of PIO⫹CAN decreased fibrosis to 5.9% ⫾ 0.6% and 8.6% ⫾ 0.6% in RA and LA, with mean fibrosis values that were slightly less than for PIO or CAN alone but not significantly different from either. Fibrosis was more extensive in the LA, which is in agreement with the previous study in the canine CHF model.2 This difference may be due to greater hemodynamic stress in the LA.
Changes in TGF-1, TNF-␣, and MAPK subfamilies
Figure 5 shows representative Western blots for TGF-1 and TNF-␣ as well as mean data in the LA, respectively. Treatment with either PIO or CAN similarly suppressed ventricular tachypacing-induced increases in TGF-1 and TNF-␣ expres-
Figure 3 A: Examples of atrial arrhythmias induced by burst pacing in a nonpaced control rabbit (NP) and a ventricular tachypacing (VTP)-induced CHF rabbit. B: Duration of atrial fibrillation (DAF) in NP, VTP, pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 15 per group). v ⫽ ventricular electrogram. †P ⬍.05 vs VTP.
Shimano et al
Pioglitazone and Atrial Structural Remodeling
A
B
NP
455
F
VTP
18
㽙400
㽙400
C
D
PIO
CAN
%Fibrosis in RA
16 14 12
**
10
*†
8
*† *†
6 4 2
G
0 18
**
㽙400
E
PIO+CAN
㽙400
%Fibrosis in LA
16 14 12 10
*† *†
*†
8 6 4 2 0
㽙400
PIO NP VTP PIO CAN +CAN
Figure 4 A–E: Representative histologic sections (Masson trichrome stain) of the LA free wall from each group. F, G: Percent fibrosis area (mean ⫾ SEM) at RA and LA in nonpaced control (NP), ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN treated rabbits (n ⫽ 12 hearts per group). *P ⬍.05, **P ⬍.01 vs NP, †P ⬍.05 vs VTP.
sion. The combination of PIO⫹CAN was not significantly different from either alone. Figure 6 shows representative Western blots with corresponding band density measurements for ERK, p38, and JNK (panels A, C, and E, respectively).
Band intensities of total and phosphorylated forms of all MAPKs were increased by ventricular tachypacing, but the ratio of phosphorylated to total MAPK (an indicator of phosphorylation-induced activation) was increased for ERK only
Figure 5 A, B: Examples of transforming growth factor-1 (TGF-1, ⬃13 kDa) and tumor necrosis factor-␣ (TNF-␣, ⬃17 kDa) bands with corresponding GAPDH bands, and band intensities (mean ⫾ SEM) normalized to nonpaced (NP) control values in NP, ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 12 hearts per group). *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
456
Heart Rhythm, Vol 5, No 3, March 2008
Figure 6 Examples of immunoblotting for extracellular signal-regulated kinase (ERK, 42- and 44-kDa isoforms (A), p38 (38 kDa) (C), and c-Jun N-terminal kinase (JNK, 46- and 54-kDa isoforms (E)) expression with corresponding GAPDH band (G), band intensities (mean ⫾ SEM) normalized to nonpaced (NP) control values, and ratio of phosphorylated/total ERK (B), p38 (D), and JNK (F) in the left atrium from NP, ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 12 hearts per group). *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
(panel B). Treatment with either PIO or CAN significantly reduced ERK activation similarly, in association with a decrease in the ratio. The combination of PIO⫹CAN had no additive effect on ERK activation. None of the interventions affected the ratio for p38 or JNK (panels D and E). The inhibitory effects of PIO or CAN treatment on TGF-1, TNF-␣, and phosphorylated ERK expression also were observed in the RA (Figure 7) and were qualitatively similar to LA changes. In addition, increases in TGF-1, TNF-␣, and MAPK subfamilies expression with 28-day ventricular tachy-
pacing were smaller than those with 5-day ventricular tachypacing (data not shown), which are consistent with previous findings in the canine model3,19 that molecular changes related to atrial structural remodeling are maximal within the first few days after ventricular tachypacing initiation.
Discussion Main findings This study showed that pioglitazone attenuated arrhythmogenic atrial structural remodeling in a rabbit model of CHF
Shimano et al
Pioglitazone and Atrial Structural Remodeling
457
Figure 7 Transforming growth factor- (TGF-1), TNF-␣ (A) and total and phosphorylated MAPK (B) expression changes in the right atrium (n ⫽ 12 hearts per group). Abbreviations and format as in Figure 6. *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
while significantly reducing TGF-1and TNF-␣ expression and ERK activation. These actions were comparable to those of candesartan. The additional effects of a combination of pioglitazone and candesartan were small, and differences were not statistically significant from those of pioglitazone or candesartan alone.
CHF-related atrial structural remodeling prevention in animal models Studies in canine models have indicated that atrial interstitial fibrosis is an important contributor to the substrate for CHF-related AF.2–5 Increased expression of tissue angiotensin II, MAPKs (ERK, p38, and JNK), TGF-1 and TNF-␣ has been reported as molecular alterations related to development of AF-promoting fibrosis.3,19,20 Our data from a rabbit CHF model are consistent with these findings. Treatment with candesartan decreased the fibrosis in LA by 45% in our rabbits, similar to studies in canine models showing approximately 40% and 50% reduction by enalapril3 and candesartan,5 respectively. Li et al3 reported that enalapril reduced ERK activation but did not affect activation of p38 and JNK in canine atria. We observed similar effects in candesartan-treated rabbits. Moreover, our study showed that candesartan attenuated increased expression of
TGF-1 and TNF-␣, which has been suggested to be an important profibrotic mediator19,21 and an atrial myofibroblast stimulator,22 respectively. Based on evidence that oxidative stress and inflammation are involved in AF-related remodeling, pharmacologic approaches targeting the AF substrate have been highlighted as potential “upstream” therapy for AF.23,24 ShiroshitaTakeshita et al25 found that simvastatin inhibited atrial fibroblast proliferation and attenuated LV dysfunction, preventing CHF-induced atrial structural remodeling and AF promotion in dogs. Lee et al20 reported that pirfenidone, which reportedly suppresses lung fibrosis by attenuating inflammatory cell infiltration and decreasing TGF-1 expression in a bleomycin-hamster model, reduced atrial fibrosis and AF vulnerability in CHF dogs. This study is the first to investigate the potential role of PPAR-␥ in atrial remodeling. We found that the PPAR-␥ agonist pioglitazone suppresses arrhythmogenic atrial structural remodeling. The effect of pioglitazone was observed to be equivalent to that of candesartan, but a combination of pioglitazone and candesartan did not appear to provide additional benefit over either agent alone. The lack of additional benefit from the combination is consistent with the
458 observation that both pioglitazone and candesartan seem to share common molecular signatures, that is, inhibition of CHF-induced upregulation of TGF-, TNF-␣, and phosphorylated ERK expression. If both drugs act to attenuate the same molecular pathways, it is not surprising that applying them together does not provide added benefit over either alone.
Potential underlying mechanisms Pharmacologic actions of PPAR-␥ activators have been extensively studied in vitro and in vivo. PPAR-␥ activators reduce the production of proinflammatory cytokines such as interleukin-1, IL-6, inducible nitric oxide, TNF-␣, and matrix metalloproteinase-9 by inhibiting transcription factors that bind to AP-1, STAT, and nuclear factor-B (NF-B) in monocytes and macrophages.10,11 Transrepression of NF-B and decreased TNF-␣ expression by pioglitazone are observed in cardiac myocytes,12 fibroblasts,14,15 and tissues.16 PPAR-␥ activators antagonize the actions of angiotensin II and possess antioxidant properties. Pioglitazone suppresses angiotensin II type 1 receptor expression.26 Angiotensin II-induced NADPH oxidase expression, and reactive oxygen species production are reduced in cardiac fibroblasts pretreated with pioglitazone.14,15 There is experimental evidence that pioglitazone reduces TGF-1 expression in the LVs of mice with CHF after myocardial infarction16 as well as angiotensin II–mediated procollagen-1 and collagen type I expression in cultured cardiac fibroblasts.14,15 Although the effects of PPAR-␥ activation on ERK, p38, and JNK signals involved in cardiac hypertrophy seemingly differ among PPAR-␥ activators, it has been consistently reported that pioglitazone inhibits cardiac hypertrophy.13,27 Because angiotensin II– dependent ERK activation plays an important role in atrial remodeling,4,23 it is possible that pioglitazone attenuates ERK activation indirectly by antagonizing angiotensin II– dependent pathways. A study demonstrated that pioglitazone promotes adiponectin secretion from adipocytes, which in turn attenuates cardiac hypertrophy in association with a reduction in phosphorylated ERK expression in mice.28 Therefore, it is conceivable that pioglitazone suppresses AF by antagonizing inflammatory, oxidative, and hypertrophic signaling pathways involved in atrial remodeling. The present study found that pioglitazone attenuated the increased expression of TNF-␣, TGF-1, and ERK but left unaffected p38 and JNK activation in atria from failing hearts. These changes were observed similarly in candesartan-treated rabbits. These findings suggest that pioglitazone modulates the same downstream effectors as candesartan. This notion is supported by the electrophysiologic and histologic data from this study showing no significant differences in duration of AF and percent fibrosis between rabbits treated with pioglitazone or candesartan alone or rabbits treated with the combination.
Heart Rhythm, Vol 5, No 3, March 2008
Study limitations The present study did not measure the plasma concentrations of pioglitazone. The dose of 5 mg/kg used was based on the report by Shiomi et al16 demonstrating that pioglitazone and its metabolite concentrations in mice treated with 3 mg/kg were within the range achieved in humans after oral administration of the clinical dose of 30 mg/day. The actions of PPAR-␥ activators are complex and incompletely resolved. The precise mechanisms underlying pioglitazone attenuation of TGF-1, TNF-␣, and phosphorylated ERK expression in atria remain to be determined. The downstream effectors shared by pioglitazone and candesartan were not identified. Future studies on cellular signal transduction in in vitro experimental systems are needed to understand the mechanisms underlying the observation of no additional effect of combined PIO and CAN despite targeting of different receptors. Because we did not check routinely for retrograde VA conduction in rabbits, the possibility that retrograde atrial activation during ventricular tachypacing occurred and affected the electrophysiologic data cannot be excluded. The durations of induced AF, even in ventricular tachypacingonly rabbits, were short, probably due to the small size of rabbit atria. However, duration of AF was reproducibly increased in CHF rabbits and reproducibly reduced by the drugs. In combination with the fibrosis data, the results are consistent in showing protection against structural remodelling and its AF-promoting consequences with the drugs studied. Additional mechanisms of increased atrial activation time, such as Na⫹ channel and connexin remodeling, or additional mechanisms of fibrosis-unrelated AF cannot be excluded. Studies of the cellular and molecular electrophysiology of this model would be of interest but are beyond the scope of the present study. Histologic study for quantification of fibrosis was performed in atrial free wall only. A heterogenous spatial distribution of fibrosis at the LA posterior wall governing AF wave dynamics in a sheep CHF model has recently been reported.29
Clinical implications The experimental data showing an inhibitory effect of pioglitazone on atrial fibrotic remodeling, along with a recently published meta-analysis30 demonstrating decreased risk of death, heart attack, and stroke among diabetic patients, suggest that a significant population of patients may benefit from prevention of structural remodeling-related AF with use of pioglitazone.
Conclusion Pioglitazone treatment suppresses atrial arrhythmogenic structural remodeling. Atrial fibrotic remodeling appears to have broad clinical relevance as an AF substrate, including forms unassociated with CHF. PPAR-␥ may be a novel therapeutic target for human AF.
Shimano et al
Pioglitazone and Atrial Structural Remodeling
Acknowledgments We thank Takeda Pharmaceutical Company Limited for providing pioglitazone and candesartan, Medtronic Japan for pacemakers, and St. Jude Medical Japan for technical support.
References 1. Murgatroyd FD, Camm AJ. Atrial arrhythmias. Lancet 1993;341:1317–1322. 2. Li D, Fareh S, Leung TK, et al. Promotion of atrial fibrillation by heart failure in dogs: atrial remodeling of a different sort. Circulation 1999;100:87–95. 3. Li D, Shinagawa K, Pang L, et al. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillation substrate in dogs with ventricular tachypacing-induced congestive heart failure. Circulation 2001;104: 2608 –2614. 4. Nattel S, Shiroshita-Takeshita A, Brundel BJ, et al. Mechanisms of atrial fibrillation: lessons from animal models. Prog Cardiovasc Dis 2005;48:9 –28. 5. Kumagai K, Nakashima H, Urata H, et al. Effects of angiotensin II type 1 receptor antagonist on electrical and structural remodeling in atrial fibrillation. J Am Coll Cardiol 2003;41:2197–2204. 6. Pedersen OD, Bagger H, Kober L, et al. Trandolapril reduces the incidence of atrial fibrillation after acute myocardial infarction in patients with left ventricular dysfunction. Circulation 1999;100:376 –380. 7. Vermes E, Tardif JC, Bourassa MG, et al. Enalapril decreases the incidence of atrial fibrillation in patients with left ventricular dysfunction: insight from the Studies Of Left Ventricular Dysfunction (SOLVD) trials. Circulation 2003;107: 2926 –2931. 8. Wachtell K, Lehto M, Gerdts E, et al. Angiotensin II receptor blockade reduces new-onset atrial fibrillation and subsequent stroke compared to atenolol: the Losartan Intervention For End Point Reduction in Hypertension (LIFE) study. J Am Coll Cardiol 2005;45:712–719. 9. Yki-Jarvinen H. Thiazolidinediones. N Engl J Med 2004;351:1106 –1118. 10. Ricote M, Li AC, Willson TM, et al. The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature 1998; 391:79 – 82. 11. Jiang C, Ting AT, Seed B. PPAR-gamma agonists inhibit production of monocyte inflammatory cytokines. Nature 1998;391:82– 86. 12. Takano H, Nagai T, Asakawa M, et al. Peroxisome proliferator-activated receptor activators inhibit lipopolysaccharide-induced tumor necrosis factor-alpha expression in neonatal rat cardiac myocytes. Circ Res 2000;87:596 – 602. 13. Asakawa M, Takano H, Nagai T, et al. Peroxisome proliferator-activated receptor gamma plays a critical role in inhibition of cardiac hypertrophy in vitro and in vivo. Circulation 2002;105:1240 –1246. 14. Chen J, Mehta JL. Angiotensin II-mediated oxidative stress and procollagen-1 expression in cardiac fibroblasts: blockade by pravastatin and pioglitazone. Am J Physiol Heart Circ Physiol 2006;291:H1738 –H1745.
459 15. Chen K, Chen J, Li D, et al. Angiotensin II regulation of collagen type I expression in cardiac fibroblasts: modulation by PPAR-gamma ligand pioglitazone. Hypertension 2004;44:655– 661. 16. Shiomi T, Tsutsui H, Hayashidani S, et al. Pioglitazone, a peroxisome proliferator-activated receptor-gamma agonist, attenuates left ventricular remodeling and failure after experimental myocardial infarction. Circulation 2002;106: 3126 –3132. 17. Iglarz M, Touyz RM, Viel EC, et al. Peroxisome proliferator-activated receptoralpha and receptor-gamma activators prevent cardiac fibrosis in mineralocorticoid-dependent hypertension. Hypertension 2003;42:737–743. 18. Tsuji Y, Zicha S, Qi XY, et al. Potassium channel subunit remodeling in rabbits exposed to long-term bradycardia or tachycardia: discrete arrhythmogenic consequences related to differential delayed-rectifier changes. Circulation 2006;113: 345–355. 19. Hanna N, Cardin S, Leung TK, et al. Differences in atrial vs ventricular remodeling in dogs with ventricular tachypacing-induced congestive heart failure. Cardiovasc Res 2004;63:236 –244. 20. Lee KW, Everett TH, Rahmutula D, et al. Pirfenidone prevents the development of a vulnerable substrate for atrial fibrillation in a canine model of heart failure. Circulation 2006;114:1703–1712. 21. Verheule S, Sato T, Everett T, et al. Increased vulnerability to atrial fibrillation in transgenic mice with selective atrial fibrosis caused by overexpression of TGF-beta1. Circ Res 2004;94:1458 –1465. 22. Porter KE, Turner NA, O’Regan DJ, et al. Tumor necrosis factor alpha induces human atrial myofibroblast proliferation, invasion and MMP-9 secretion: inhibition by simvastatin. Cardiovasc Res 2004;64:507–515. 23. Goette A, Bukowska A, Lendeckel U. Non-ion channel blockers as anti-arrhythmic drugs (reversal of structural remodeling). Curr Opin Pharmacol 2007;7: 219 –224. 24. Nattel S, Carlsson L. Innovative approaches to anti-arrhythmic drug therapy. Nat Rev Drug Discov 2006;5:1034 –1049. 25. Shiroshita-Takeshita A, Brundel BJ, Burstein B, et al. Effects of simvastatin on the development of the atrial fibrillation substrate in dogs with congestive heart failure. Cardiovasc Res 2007;74:75– 84. 26. Diep QN, El Mabrouk M, Cohn JS, et al. Structure, endothelial function, cell growth, and inflammation in blood vessels of angiotensin ii-infused rats: role of peroxisome proliferator-activated receptor-gamma. Circulation 2002;105:2296 – 2302. 27. Sakai S, Miyauchi T, Irukayama-Tomobe Y, et al. Peroxisome proliferatoractivated receptor-gamma activators inhibit endothelin-1-related cardiac hypertrophy in rats. Clin Sci (Lond) 2002;103(Suppl 48):16S–20S. 28. Shibata R, Ouchi N, Ito M, et al. Adiponectin-mediated modulation of hypertrophic signals in the heart. Nat Med 2004;10:1384 –1389. 29. Tanaka K, Zlochiver S, Vikstrom KL, et al. Spatial distribution of fibrosis governs fibrillation wave dynamics in the posterior left atrium during heart failure. Circ Res 2007;101:839 – 847. 30. Lincoff AM, Wolski K, Nicholls SJ, et al. Pioglitazone and risk of cardiovascular events in patients with type 2 diabetes mellitus: a meta-analysis of randomized trials. JAMA 2007;298:1180 –1188.
Introduction Atrial fibrillation (AF) is a common arrhythmia seen in clinical practice. Congestive heart failure (CHF) is a common cause of AF.1 Experimental CHF causes atrial structural remodeling characterized by atrial interstitial fibrosis and local conduction abnormalities, which create a substrate for AF maintenance.2 Angiotensin II is an important mediator of the pathologic process.3,4 Both angiotensin-converting enzyme inhibitors and angiotensin II type 1 receptor blockers reduce CHF-related structural remodeling and AF
Supported by Grant-in-Aid for Scientific Research (18590766-00) from Japan Society for the Promotion of Science to Dr. Inden. Address reprint requests and correspondence: Dr. Yukiomi Tsuji, Department of Cardiovascular Research, RIEM, Nagoya University, Furo-cho, Chikusa-ku, Nagoya, 464-8601 Japan. E-mail address: [email protected]. (Received October 13, 2007; accepted December 8, 2007.)
ductions in interatrial activation time (50 ⫾ 2 ms vs 41 ⫾ 2 ms, P ⬍.05) and atrial fibrosis (16.8% ⫾ 0.8% vs 10.9% ⫾ 0.7%, P ⬍.05; control 1.6% ⫾ 0.2%), effects comparable to those of candesartan (duration of AF: 3.0 ⫾ 0.2 seconds; activation time 44 ⫾ 2 ms; fibrosis: 9.4% ⫾ 0.6%). Both pioglitazone and candesartan reduced transforming growth factor-1, tumor necrosis factor-␣, and activated extracellular signal-regulated kinase expression similarly, but neither affected p38-kinase or c-Jun Nterminal kinase activation. The effects of combined pioglitazone and candesartan therapy were not significantly different from the effects of pioglitazone or candesartan alone. CONCLUSION Pioglitazone can attenuate congestive heart failure-induced atrial structural remodeling and AF promotion, with effects similar to those of candesartan. PPAR-␥ may be a potential therapeutic target for human AF. KEYWORDS Arrhythmia; Atrium; Heart failure; Remodeling; Antiarrhythmic agents (Heart Rhythm 2008;5:451– 459) © 2008 Heart Rhythm Society. All rights reserved.
promotion in a canine model3,5 and prevent AF in patients with left ventricular (LV) dysfunction6,7 and hypertension.8 The peroxisome proliferator-activated receptor-␥ (PPAR-␥) is a ligand-activated transcription factor that modulates gene expression for pathways involved in lipid storage and adipogenesis.9 The mechanism by which PPAR-␥ regulates myocardial metabolism is unclear. Studies have shown that PPAR-␥ activators such as pioglitazone antagonize the actions of angiotensin II and possess anti-inflammatory and antioxidant properties in vitro,10 –15 and that pioglitazone prevents cardiac remodeling in some experimental models.16,17 Shiomi et al16 reported that pioglitazone improves LV remodeling and function in mice with CHF after myocardial infarction, with a concomitant decrease in interstitial fibrosis and reduced expression of inflammatory cytokines. Iglarz et al17 demonstrated that pioglitazone suppresses cardiac fibrosis in mineralocorticoid-dependent hypertensive
1547-5271/$ -see front matter © 2008 Heart Rhythm Society. All rights reserved.
doi:10.1016/j.hrthm.2007.12.010
452 rats. However, studies investigating the effects of pioglitazone in atrial remodeling have not been reported. This study was designed to assess the effects of pioglitazone on atrial structural remodeling and AF promotion and to compare them directly with the effects of the angiotensin II type 1 receptor blocker candesartan in a rabbit model of CHF. The study also tested the hypothesis that a combination of pioglitazone and candesartan would show greater effects on atrial structural remodeling than either agent alone.
Methods Animal handling followed the Guide for the Care and Use of Laboratory Animals (NIH Publication No. 85-23), and procedures were approved by the Animal Experimentation Ethics Committee of Nagoya University.
Animal model and experimental protocol Male New Zealand white rabbits (weight 2.8 –3.3 kg) were used. Programmable right ventricular pacemakers (Medtronic, Inc., Minneapolis, MN, USA) were implanted as described previously.18 After the animals recovered from surgery, the pacemaker was programmed to pace at 380 to 400 bpm for 4 weeks. ECGs were monitored once per week to adjust the pacing rate to the maximum rate, thus allowing for 1:1 capture in each rabbit. The mean pacing rate achieved for all rabbits was 386 ⫾ 0.9 bpm. At the end of 4 weeks of ventricular tachypacing, echocardiographic and open chest electrophysiologic studies were performed under anesthesia. After electrophysiologic studies, rabbits were euthanized with pentobarbital 40 mg/kg IV, and whole hearts were immersed in 10% formaldehyde for subsequent histologic study.
Groups Results from 15 ventricular tachypaced rabbits without any treatment (ventricular tachypacing-only group) and 15 nonpaced control rabbits were compared with results from rabbits subjected to ventricular tachypacing during daily oral administration of pioglitazone 5 mg/kg (PIO, n ⫽ 15), candesartan 2 mg/kg (CAN, n ⫽ 15), or combined pioglitazone 5 mg/kg and candesartan 2 mg/kg (PIO⫹CAN, n ⫽ 15), beginning 14 days before ventricular tachypacing onset and continuing throughout the study period.
Electrophysiologic study For electrophysiologic study on day 28, rabbits were anesthetized with ketamine hydrochloride 35 mg/kg IM and xylazine 3 mg/kg IM and ventilated mechanically with isoflurane-containing room air (0.5%/2 L/min). The pacemaker was deactivated, and body temperature was maintained at 37°C. Median sternotomy was performed, and custom-made bipolar electrodes were attached to the right atrial appendage (RAA) and left atrial appendage (LAA) for recording and stimulation. A programmable stimulator (Nihon Kohden, Tokyo, Japan) was used to deliver twice threshold currents at 2-ms pulse duration. Effective refractory periods (ERPs) were measured at the RAA and LAA at basic cycle lengths of 300, 200, and 150 ms with a train of
Heart Rhythm, Vol 5, No 3, March 2008 20 basic stimuli (S1) followed by a premature extrastimulus (S2) at 2-ms decrements. The effective refractory period was defined as the longest S1-S2 interval that failed to capture. The interatrial activation time was defined as the time from RAA pacing spike to LAA activation. AF induction was tested by applying burst pacing (cycle length 70 – 140 ms) at 5-ms increments and four times threshold current. AF was defined as a rapid (⬎500/min) irregular atrial rhythm lasting longer than 2 seconds. Durations of AF from all episodes were averaged in each rabbit to provide a representative mean duration of AF value.
Histology Transverse sections cut at 5-m intervals and stained with Masson trichrome at the midatrial level were used to evaluate fibrotic deposition in rabbit atria. Microscopic images were digitized with Viewfinder Light Image (Mitani Corp., Tokyo, Japan). To quantify the percent fibrosis area at right atrial (RA) and left atrial (LA) free walls, the blue pixel content of digitized images was measured relative to total tissue area using the image analyzer Win ROOF (Mitani Corp.). Blood vessels and perivascular interstitial tissues were excluded from fibrosis quantification. The average number of transverse sections of LA free wall used for quantitative analysis was 54 ⫾ 1.0 for each rabbit.
Western blots Studies in a canine model of ventricular tachypacing-induced CHF showed that molecular changes related to atrial structural remodeling were observed early after initiation of ventricular tachypacing.3,4,19 To investigate effects of pharmacologic interventions on the expression of transforming growth factor-1 (TGF-1), tumor necrosis factor-␣ (TNF␣), and mitogen-activated protein kinase (MAPK) subfamilies including extracellular signal-regulated kinase (ERK), p38-kinase, and c-Jun N-terminal kinase (JNK), we prepared 12 rabbits subjected to ventricular tachypacing for 5 days in each group. Atrial tissue samples were fast-frozen in liquid nitrogen and stored at – 80°C. Protein extracts were prepared and immunoblotting was performed as previously described.18 Commercially available primary antibodies were used. Antibodies to TNF-␣ (sc1348; goat), phosphorylated (sc7973; mouse) and total p38 (sc7972; mouse), phosphorylated ERK (sc16982; goat), and phosphorylated JNK (sc12882; goat) were purchased from Santa Cruz Biotechnology (Santa Cruz, CA); TGF-1 (T0438; mouse) from Sigma Chemical (St. Louis, MO); total ERK (9102; rabbit) and total JNK (9252; rabbit) from Cell Signaling Technology (Danvers, MA); and glyceraldehyde 3-phosphate dehydrogenase ([GAPDH] ab8245; mouse) from Abcam plc (Camdridge, UK). Band densities were quantified by densitometry, standardized to GAPDH, and normalized to the control sample.
Statistical analysis Data are expressed as mean ⫾ SEM. Statistical significance was assessed by one-way analysis of variance with Bonferroni correction. P ⬍.05 was considered significant.
Shimano et al Table 1 and CAN
Pioglitazone and Atrial Structural Remodeling
453
Characteristics of ventricular tachypacing rabbits in the absence and presence of treatment with PIO, CAN, or combined PIO NP (n ⫽ 12)
FS (%) LVEDd (mm) LVESd (mm) IVSd (mm) PWd (mm) LAd (mm) Systolic BP (mmHg) Blood Glucose (mg/dL)
42.0 14.4 8.4 3.1 3.1 8.1 119 334
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
1.0 0.3 0.3 0.1 0.1 0.2 4 12
VTP alone (n ⫽ 12) 9.8 18.7 16.9 2.8 2.8 13.8 109 355
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.7* 0.3* 0.3* 0.1 0.1 0.2* 5 16
PIO (n ⫽ 12) 10.2 18.7 16.8 2.8 2.8 13.9 110 300
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.9* 0.3* 0.2* 0.1 0.1 0.2* 4 15
CAN (n ⫽ 12) 10.6 18.1 16.2 2.9 2.9 13.1 97 347
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.7* 0.3* 0.2* 0.1 0.1 0.3* 5† 14
PIO⫹CAN (n ⫽ 12) 10.4 18.2 16.3 2.9 2.9 13.5 98 310
⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾ ⫾
0.8* 0.4* 0.4* 0.1 0.1 0.2* 5† 14
Values are given as mean ⫾ SEM. NP indicates non-paced control; n, number of rabbits; VTP, ventricular tachypacing; PIO, pioglitazone; CAN, candesartan; FS, fractional shortening; LVEDd, left ventricular end-diastolic diameter; LVESd, left ventricular end-systolic diameter; IVSd, intraventricular septum diameter; PWd, posterior wall diameter of left ventricle; LAd, left atrial diameter. *P ⬍.05 vs NP; †P ⬍.05 vs VTP alone.
Results Animal characteristics Table 1 summarizes echocardiographic parameters, systolic blood pressure, and blood glucose concentrations in nonpaced control and ventricular tachypacing-induced CHF rabbits in the absence and presence of treatment with PIO, CAN, or PIO⫹CAN. Ventricular tachypacing for 4 weeks decreased fractional shortening and increased LV end-systolic and enddiastolic diameter and LA diameter, regardless of treatment with PIO, CAN, or PIO⫹CAN. Systolic blood pressure was reduced in rabbits treated with CAN or PIO⫹CAN. No significant differences in blood glucose concentration were observed among groups. All rabbits subjected to ventricular tachypacing had considerable amounts of ascites and pleural effusion. Sixty of 100 rabbits survived for 28 days and 40 died prematurely. Mortality rate for each group was 50% (15/30) for ventricular tachypacing-only, 40% (10/25) for PIO, 35% (8/23) for CAN, and 32% (7/22) for PIO⫹CAN rabbits (onetailed probability of Chi-square distribution [0.388] ⬎.05 for intergroup differences). Failing hearts displayed marked enlargement of all cardiac chambers (Figure 1).
Electrophysiologic study Figure 2A and B illustrates effective refractory periods at the RAA and LAA, respectively. Effective refractory
Figure 1
periods had a tendency to be increased by ventricular tachypacing but were not significantly different among groups. Interatrial activation time was increased by ⬎70% with ventricular tachypacing, from 29 ⫾ 1 ms in nonpaced controls to 50 ⫾ 2 ms in ventricular tachypacing-only rabbits. Activation times were significantly reduced by treatment with PIO, CAN, and PIO⫹CAN (from 50 ⫾ 2 ms in ventricular tachypacing-only group to 41 ⫾ 2 ms, 44 ⫾ 2 ms, and 41 ⫾ 1 ms, respectively, P ⬍.05 for each vs ventricular tachypacing-only group, Figure 2C), although they remained significantly greater than control values (approximately 45% larger, P ⬍.05). Examples of atrial arrhythmia induced by LAA burst pacing and mean data of duration of AF are shown in Figure 3A and B, respectively. Duration of AF was 8.0 ⫾ 1.4 seconds in ventricular tachypaced rabbits in the absence of drug therapy, whereas AF was not induced in nonpaced controls. Duration of AF was significantly reduced in PIO- and CAN-treated rabbits (3.5 ⫾ 0.2 seconds and 3.0 ⫾ 0.2 seconds, respectively) compared with ventricular tachypacing-only rabbits (P ⬍.05). PIO⫹ CAN–treated rabbits had duration of AF averaging 2.9 ⫾ 0.2 seconds, which was not significantly different from PIO alone or CAN alone.
Control heart and failing heart from a sham-operated and a ventricular tachypaced rabbit, respectively.
454
Heart Rhythm, Vol 5, No 3, March 2008
Figure 2 Electrophysiologic parameters measured at basic cycle lengths (BCLs) of 150, 200m and 300 ms in nonpaced control rabbits (NP, n ⫽ 15), ventricular tachypacing-alone rabbits (VTP, n ⫽ 15), and VTP rabbits treated with pioglitazone (PIO, n ⫽ 15), candesartan (CAN, n ⫽ 15), or a combination of pioglitazone and candesartan (PIO⫹CAN, n ⫽ 15). A, B: Effective refractory periods (ERPs) at right (RAA) and left atrial appendage (LAA). C: Interatrial activation time between LAA and RAA. *P ⬍.05, **P ⬍.01 vs NP, †P ⬍.05 vs VTP.
Atrial fibrosis Representative LA tissue sections from each group stained with Masson trichrome are shown in Figure 4. Atria appeared grossly normal in nonpaced controls (panel A), whereas thick fibrous tissue separated bundles of myofibers in ventricular tachypacing-only rabbits (panel B). Atrial fibrosis was clearly attenuated in rabbits treated with PIO, CAN, and PIO⫹CAN (panels C, D, and E, respectively). Panel F summarizes quantitative data from RA and LA. The area of fibrosis was considerably increased in both atria in ventricular tachypacing-only rabbits (11.4% ⫾ 0.9% and 16.8% ⫾ 0.8% in RA and LA, respectively) compared with nonpaced controls (2.0% ⫾ 0.2% and 1.6% ⫾ 0.2%). PIO suppressed fibrosis (7.6% ⫾ 0.6% and 10.9% ⫾ 0.7%) to a
level comparable with CAN therapy (7.2% ⫾ 0.6% and 9.4% ⫾ 0.6%). The combination of PIO⫹CAN decreased fibrosis to 5.9% ⫾ 0.6% and 8.6% ⫾ 0.6% in RA and LA, with mean fibrosis values that were slightly less than for PIO or CAN alone but not significantly different from either. Fibrosis was more extensive in the LA, which is in agreement with the previous study in the canine CHF model.2 This difference may be due to greater hemodynamic stress in the LA.
Changes in TGF-1, TNF-␣, and MAPK subfamilies
Figure 5 shows representative Western blots for TGF-1 and TNF-␣ as well as mean data in the LA, respectively. Treatment with either PIO or CAN similarly suppressed ventricular tachypacing-induced increases in TGF-1 and TNF-␣ expres-
Figure 3 A: Examples of atrial arrhythmias induced by burst pacing in a nonpaced control rabbit (NP) and a ventricular tachypacing (VTP)-induced CHF rabbit. B: Duration of atrial fibrillation (DAF) in NP, VTP, pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 15 per group). v ⫽ ventricular electrogram. †P ⬍.05 vs VTP.
Shimano et al
Pioglitazone and Atrial Structural Remodeling
A
B
NP
455
F
VTP
18
㽙400
㽙400
C
D
PIO
CAN
%Fibrosis in RA
16 14 12
**
10
*†
8
*† *†
6 4 2
G
0 18
**
㽙400
E
PIO+CAN
㽙400
%Fibrosis in LA
16 14 12 10
*† *†
*†
8 6 4 2 0
㽙400
PIO NP VTP PIO CAN +CAN
Figure 4 A–E: Representative histologic sections (Masson trichrome stain) of the LA free wall from each group. F, G: Percent fibrosis area (mean ⫾ SEM) at RA and LA in nonpaced control (NP), ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN treated rabbits (n ⫽ 12 hearts per group). *P ⬍.05, **P ⬍.01 vs NP, †P ⬍.05 vs VTP.
sion. The combination of PIO⫹CAN was not significantly different from either alone. Figure 6 shows representative Western blots with corresponding band density measurements for ERK, p38, and JNK (panels A, C, and E, respectively).
Band intensities of total and phosphorylated forms of all MAPKs were increased by ventricular tachypacing, but the ratio of phosphorylated to total MAPK (an indicator of phosphorylation-induced activation) was increased for ERK only
Figure 5 A, B: Examples of transforming growth factor-1 (TGF-1, ⬃13 kDa) and tumor necrosis factor-␣ (TNF-␣, ⬃17 kDa) bands with corresponding GAPDH bands, and band intensities (mean ⫾ SEM) normalized to nonpaced (NP) control values in NP, ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 12 hearts per group). *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
456
Heart Rhythm, Vol 5, No 3, March 2008
Figure 6 Examples of immunoblotting for extracellular signal-regulated kinase (ERK, 42- and 44-kDa isoforms (A), p38 (38 kDa) (C), and c-Jun N-terminal kinase (JNK, 46- and 54-kDa isoforms (E)) expression with corresponding GAPDH band (G), band intensities (mean ⫾ SEM) normalized to nonpaced (NP) control values, and ratio of phosphorylated/total ERK (B), p38 (D), and JNK (F) in the left atrium from NP, ventricular tachypacing (VTP), pioglitazone-treated (PIO), candesartan-treated (CAN), and PIO⫹CAN rabbits (n ⫽ 12 hearts per group). *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
(panel B). Treatment with either PIO or CAN significantly reduced ERK activation similarly, in association with a decrease in the ratio. The combination of PIO⫹CAN had no additive effect on ERK activation. None of the interventions affected the ratio for p38 or JNK (panels D and E). The inhibitory effects of PIO or CAN treatment on TGF-1, TNF-␣, and phosphorylated ERK expression also were observed in the RA (Figure 7) and were qualitatively similar to LA changes. In addition, increases in TGF-1, TNF-␣, and MAPK subfamilies expression with 28-day ventricular tachy-
pacing were smaller than those with 5-day ventricular tachypacing (data not shown), which are consistent with previous findings in the canine model3,19 that molecular changes related to atrial structural remodeling are maximal within the first few days after ventricular tachypacing initiation.
Discussion Main findings This study showed that pioglitazone attenuated arrhythmogenic atrial structural remodeling in a rabbit model of CHF
Shimano et al
Pioglitazone and Atrial Structural Remodeling
457
Figure 7 Transforming growth factor- (TGF-1), TNF-␣ (A) and total and phosphorylated MAPK (B) expression changes in the right atrium (n ⫽ 12 hearts per group). Abbreviations and format as in Figure 6. *P ⬍.05 vs NP, †P ⬍.05 vs VTP.
while significantly reducing TGF-1and TNF-␣ expression and ERK activation. These actions were comparable to those of candesartan. The additional effects of a combination of pioglitazone and candesartan were small, and differences were not statistically significant from those of pioglitazone or candesartan alone.
CHF-related atrial structural remodeling prevention in animal models Studies in canine models have indicated that atrial interstitial fibrosis is an important contributor to the substrate for CHF-related AF.2–5 Increased expression of tissue angiotensin II, MAPKs (ERK, p38, and JNK), TGF-1 and TNF-␣ has been reported as molecular alterations related to development of AF-promoting fibrosis.3,19,20 Our data from a rabbit CHF model are consistent with these findings. Treatment with candesartan decreased the fibrosis in LA by 45% in our rabbits, similar to studies in canine models showing approximately 40% and 50% reduction by enalapril3 and candesartan,5 respectively. Li et al3 reported that enalapril reduced ERK activation but did not affect activation of p38 and JNK in canine atria. We observed similar effects in candesartan-treated rabbits. Moreover, our study showed that candesartan attenuated increased expression of
TGF-1 and TNF-␣, which has been suggested to be an important profibrotic mediator19,21 and an atrial myofibroblast stimulator,22 respectively. Based on evidence that oxidative stress and inflammation are involved in AF-related remodeling, pharmacologic approaches targeting the AF substrate have been highlighted as potential “upstream” therapy for AF.23,24 ShiroshitaTakeshita et al25 found that simvastatin inhibited atrial fibroblast proliferation and attenuated LV dysfunction, preventing CHF-induced atrial structural remodeling and AF promotion in dogs. Lee et al20 reported that pirfenidone, which reportedly suppresses lung fibrosis by attenuating inflammatory cell infiltration and decreasing TGF-1 expression in a bleomycin-hamster model, reduced atrial fibrosis and AF vulnerability in CHF dogs. This study is the first to investigate the potential role of PPAR-␥ in atrial remodeling. We found that the PPAR-␥ agonist pioglitazone suppresses arrhythmogenic atrial structural remodeling. The effect of pioglitazone was observed to be equivalent to that of candesartan, but a combination of pioglitazone and candesartan did not appear to provide additional benefit over either agent alone. The lack of additional benefit from the combination is consistent with the
458 observation that both pioglitazone and candesartan seem to share common molecular signatures, that is, inhibition of CHF-induced upregulation of TGF-, TNF-␣, and phosphorylated ERK expression. If both drugs act to attenuate the same molecular pathways, it is not surprising that applying them together does not provide added benefit over either alone.
Potential underlying mechanisms Pharmacologic actions of PPAR-␥ activators have been extensively studied in vitro and in vivo. PPAR-␥ activators reduce the production of proinflammatory cytokines such as interleukin-1, IL-6, inducible nitric oxide, TNF-␣, and matrix metalloproteinase-9 by inhibiting transcription factors that bind to AP-1, STAT, and nuclear factor-B (NF-B) in monocytes and macrophages.10,11 Transrepression of NF-B and decreased TNF-␣ expression by pioglitazone are observed in cardiac myocytes,12 fibroblasts,14,15 and tissues.16 PPAR-␥ activators antagonize the actions of angiotensin II and possess antioxidant properties. Pioglitazone suppresses angiotensin II type 1 receptor expression.26 Angiotensin II-induced NADPH oxidase expression, and reactive oxygen species production are reduced in cardiac fibroblasts pretreated with pioglitazone.14,15 There is experimental evidence that pioglitazone reduces TGF-1 expression in the LVs of mice with CHF after myocardial infarction16 as well as angiotensin II–mediated procollagen-1 and collagen type I expression in cultured cardiac fibroblasts.14,15 Although the effects of PPAR-␥ activation on ERK, p38, and JNK signals involved in cardiac hypertrophy seemingly differ among PPAR-␥ activators, it has been consistently reported that pioglitazone inhibits cardiac hypertrophy.13,27 Because angiotensin II– dependent ERK activation plays an important role in atrial remodeling,4,23 it is possible that pioglitazone attenuates ERK activation indirectly by antagonizing angiotensin II– dependent pathways. A study demonstrated that pioglitazone promotes adiponectin secretion from adipocytes, which in turn attenuates cardiac hypertrophy in association with a reduction in phosphorylated ERK expression in mice.28 Therefore, it is conceivable that pioglitazone suppresses AF by antagonizing inflammatory, oxidative, and hypertrophic signaling pathways involved in atrial remodeling. The present study found that pioglitazone attenuated the increased expression of TNF-␣, TGF-1, and ERK but left unaffected p38 and JNK activation in atria from failing hearts. These changes were observed similarly in candesartan-treated rabbits. These findings suggest that pioglitazone modulates the same downstream effectors as candesartan. This notion is supported by the electrophysiologic and histologic data from this study showing no significant differences in duration of AF and percent fibrosis between rabbits treated with pioglitazone or candesartan alone or rabbits treated with the combination.
Heart Rhythm, Vol 5, No 3, March 2008
Study limitations The present study did not measure the plasma concentrations of pioglitazone. The dose of 5 mg/kg used was based on the report by Shiomi et al16 demonstrating that pioglitazone and its metabolite concentrations in mice treated with 3 mg/kg were within the range achieved in humans after oral administration of the clinical dose of 30 mg/day. The actions of PPAR-␥ activators are complex and incompletely resolved. The precise mechanisms underlying pioglitazone attenuation of TGF-1, TNF-␣, and phosphorylated ERK expression in atria remain to be determined. The downstream effectors shared by pioglitazone and candesartan were not identified. Future studies on cellular signal transduction in in vitro experimental systems are needed to understand the mechanisms underlying the observation of no additional effect of combined PIO and CAN despite targeting of different receptors. Because we did not check routinely for retrograde VA conduction in rabbits, the possibility that retrograde atrial activation during ventricular tachypacing occurred and affected the electrophysiologic data cannot be excluded. The durations of induced AF, even in ventricular tachypacingonly rabbits, were short, probably due to the small size of rabbit atria. However, duration of AF was reproducibly increased in CHF rabbits and reproducibly reduced by the drugs. In combination with the fibrosis data, the results are consistent in showing protection against structural remodelling and its AF-promoting consequences with the drugs studied. Additional mechanisms of increased atrial activation time, such as Na⫹ channel and connexin remodeling, or additional mechanisms of fibrosis-unrelated AF cannot be excluded. Studies of the cellular and molecular electrophysiology of this model would be of interest but are beyond the scope of the present study. Histologic study for quantification of fibrosis was performed in atrial free wall only. A heterogenous spatial distribution of fibrosis at the LA posterior wall governing AF wave dynamics in a sheep CHF model has recently been reported.29
Clinical implications The experimental data showing an inhibitory effect of pioglitazone on atrial fibrotic remodeling, along with a recently published meta-analysis30 demonstrating decreased risk of death, heart attack, and stroke among diabetic patients, suggest that a significant population of patients may benefit from prevention of structural remodeling-related AF with use of pioglitazone.
Conclusion Pioglitazone treatment suppresses atrial arrhythmogenic structural remodeling. Atrial fibrotic remodeling appears to have broad clinical relevance as an AF substrate, including forms unassociated with CHF. PPAR-␥ may be a novel therapeutic target for human AF.
Shimano et al
Pioglitazone and Atrial Structural Remodeling
Acknowledgments We thank Takeda Pharmaceutical Company Limited for providing pioglitazone and candesartan, Medtronic Japan for pacemakers, and St. Jude Medical Japan for technical support.
References 1. Murgatroyd FD, Camm AJ. Atrial arrhythmias. Lancet 1993;341:1317–1322. 2. Li D, Fareh S, Leung TK, et al. Promotion of atrial fibrillation by heart failure in dogs: atrial remodeling of a different sort. Circulation 1999;100:87–95. 3. Li D, Shinagawa K, Pang L, et al. Effects of angiotensin-converting enzyme inhibition on the development of the atrial fibrillation substrate in dogs with ventricular tachypacing-induced congestive heart failure. Circulation 2001;104: 2608 –2614. 4. Nattel S, Shiroshita-Takeshita A, Brundel BJ, et al. Mechanisms of atrial fibrillation: lessons from animal models. Prog Cardiovasc Dis 2005;48:9 –28. 5. Kumagai K, Nakashima H, Urata H, et al. Effects of angiotensin II type 1 receptor antagonist on electrical and structural remodeling in atrial fibrillation. J Am Coll Cardiol 2003;41:2197–2204. 6. Pedersen OD, Bagger H, Kober L, et al. Trandolapril reduces the incidence of atrial fibrillation after acute myocardial infarction in patients with left ventricular dysfunction. Circulation 1999;100:376 –380. 7. Vermes E, Tardif JC, Bourassa MG, et al. Enalapril decreases the incidence of atrial fibrillation in patients with left ventricular dysfunction: insight from the Studies Of Left Ventricular Dysfunction (SOLVD) trials. Circulation 2003;107: 2926 –2931. 8. Wachtell K, Lehto M, Gerdts E, et al. Angiotensin II receptor blockade reduces new-onset atrial fibrillation and subsequent stroke compared to atenolol: the Losartan Intervention For End Point Reduction in Hypertension (LIFE) study. J Am Coll Cardiol 2005;45:712–719. 9. Yki-Jarvinen H. Thiazolidinediones. N Engl J Med 2004;351:1106 –1118. 10. Ricote M, Li AC, Willson TM, et al. The peroxisome proliferator-activated receptor-gamma is a negative regulator of macrophage activation. Nature 1998; 391:79 – 82. 11. Jiang C, Ting AT, Seed B. PPAR-gamma agonists inhibit production of monocyte inflammatory cytokines. Nature 1998;391:82– 86. 12. Takano H, Nagai T, Asakawa M, et al. Peroxisome proliferator-activated receptor activators inhibit lipopolysaccharide-induced tumor necrosis factor-alpha expression in neonatal rat cardiac myocytes. Circ Res 2000;87:596 – 602. 13. Asakawa M, Takano H, Nagai T, et al. Peroxisome proliferator-activated receptor gamma plays a critical role in inhibition of cardiac hypertrophy in vitro and in vivo. Circulation 2002;105:1240 –1246. 14. Chen J, Mehta JL. Angiotensin II-mediated oxidative stress and procollagen-1 expression in cardiac fibroblasts: blockade by pravastatin and pioglitazone. Am J Physiol Heart Circ Physiol 2006;291:H1738 –H1745.
459 15. Chen K, Chen J, Li D, et al. Angiotensin II regulation of collagen type I expression in cardiac fibroblasts: modulation by PPAR-gamma ligand pioglitazone. Hypertension 2004;44:655– 661. 16. Shiomi T, Tsutsui H, Hayashidani S, et al. Pioglitazone, a peroxisome proliferator-activated receptor-gamma agonist, attenuates left ventricular remodeling and failure after experimental myocardial infarction. Circulation 2002;106: 3126 –3132. 17. Iglarz M, Touyz RM, Viel EC, et al. Peroxisome proliferator-activated receptoralpha and receptor-gamma activators prevent cardiac fibrosis in mineralocorticoid-dependent hypertension. Hypertension 2003;42:737–743. 18. Tsuji Y, Zicha S, Qi XY, et al. Potassium channel subunit remodeling in rabbits exposed to long-term bradycardia or tachycardia: discrete arrhythmogenic consequences related to differential delayed-rectifier changes. Circulation 2006;113: 345–355. 19. Hanna N, Cardin S, Leung TK, et al. Differences in atrial vs ventricular remodeling in dogs with ventricular tachypacing-induced congestive heart failure. Cardiovasc Res 2004;63:236 –244. 20. Lee KW, Everett TH, Rahmutula D, et al. Pirfenidone prevents the development of a vulnerable substrate for atrial fibrillation in a canine model of heart failure. Circulation 2006;114:1703–1712. 21. Verheule S, Sato T, Everett T, et al. Increased vulnerability to atrial fibrillation in transgenic mice with selective atrial fibrosis caused by overexpression of TGF-beta1. Circ Res 2004;94:1458 –1465. 22. Porter KE, Turner NA, O’Regan DJ, et al. Tumor necrosis factor alpha induces human atrial myofibroblast proliferation, invasion and MMP-9 secretion: inhibition by simvastatin. Cardiovasc Res 2004;64:507–515. 23. Goette A, Bukowska A, Lendeckel U. Non-ion channel blockers as anti-arrhythmic drugs (reversal of structural remodeling). Curr Opin Pharmacol 2007;7: 219 –224. 24. Nattel S, Carlsson L. Innovative approaches to anti-arrhythmic drug therapy. Nat Rev Drug Discov 2006;5:1034 –1049. 25. Shiroshita-Takeshita A, Brundel BJ, Burstein B, et al. Effects of simvastatin on the development of the atrial fibrillation substrate in dogs with congestive heart failure. Cardiovasc Res 2007;74:75– 84. 26. Diep QN, El Mabrouk M, Cohn JS, et al. Structure, endothelial function, cell growth, and inflammation in blood vessels of angiotensin ii-infused rats: role of peroxisome proliferator-activated receptor-gamma. Circulation 2002;105:2296 – 2302. 27. Sakai S, Miyauchi T, Irukayama-Tomobe Y, et al. Peroxisome proliferatoractivated receptor-gamma activators inhibit endothelin-1-related cardiac hypertrophy in rats. Clin Sci (Lond) 2002;103(Suppl 48):16S–20S. 28. Shibata R, Ouchi N, Ito M, et al. Adiponectin-mediated modulation of hypertrophic signals in the heart. Nat Med 2004;10:1384 –1389. 29. Tanaka K, Zlochiver S, Vikstrom KL, et al. Spatial distribution of fibrosis governs fibrillation wave dynamics in the posterior left atrium during heart failure. Circ Res 2007;101:839 – 847. 30. Lincoff AM, Wolski K, Nicholls SJ, et al. Pioglitazone and risk of cardiovascular events in patients with type 2 diabetes mellitus: a meta-analysis of randomized trials. JAMA 2007;298:1180 –1188.