EXPERIMENTAL INVESTIGATION Circ J 2004; 68: 580–586

Effects of Cellular Cardiomyoplasty on Ventricular Remodeling Assessed by Doppler and Topographic Immunohistochemistry

Minoru Yoshiyama, MD; Tetsuya Hayashi, MD**; Yasuhiro Nakamura, MD; Takashi Omura, MD; Yasukatsu Izumi, MD*; Ryo Matsumoto, MD; Kazuhide Takeuchi, MD; Yasushi Kitaura, MD**; Junichi Yoshikawa, MD

Background Myocardial infarction (MI) promotes deleterious remodeling of the myocardium, resulting in ventricular dilation and pump dysfunction. Supplementing infarcted myocardium with neonatal myocyte would attenuate deleterious remodeling and so the present study used and histology to analyze the cardiac function and histological regeneration of the damaged myocardium after cellular cardiomyo- plasty. Methods and Results Experimental MI was induced by 24-h coronary ligation followed by reperfusion in adult male Lewis rats and neonatal myocytes were injected directly into the infarct and peri-infarct regions. Three groups of animals were studied at 4 weeks after cellular cardiomyoplasty: noninfarcted control (control), MI plus sham injection (MI), and MI plus cell injection (MI+cell). Ventricular remodeling and cardiac perfor- mance were assessed by Doppler echocardiography or contrast echocardiography. At 4 weeks after cellular car- diomyoplasty, MI+cell exhibited attenuation of global ventricular dilation and cardiac function compared with MI hearts not receiving cellular cardiomyoplasty. Immunohistochemically, connexin-43-positive small cells were observed in the vicinity of the infarction in MI+cell . By electron microscopy, these cells con- tained myofilaments with Z-bands and poorly developed intercalated disks, suggesting neonatal myocardial cells. Furthermore, the myocardial cells were often making close contact with interstitial cells. Conclusions Implanted neonatal myocytes form viable grafts after MI, resulting in attenuated ventricular dila- tion and enhanced contractile function. Echocardiography, electron microscopy, and immunohistochemistry are useful methods for assessing the functional and histological regeneration of the damaged myocardium. (Circ J 2004; 68: 580–586) Key Words: Cell therapy; Echocardiography; Myocardial infarction; Remodeling

espite advances in the treatment of myocardial The recent development of cellular cardiomyoplasty infarction (MI), congestive heart failure secondary offers a new approach to restore impaired heart function8 D to infarction continues to be a major complication. and evaluation of the survival of the transplanted cells and MI promotes acute and chronic transformation of both the estimation of their ultimate effect on left ventricular func- necrotic infarct zone and the nonnecrotic, peri-infarct tion are very important issues. Doppler echocardiography tissue, leading to global alterations that have collectively is a useful tool for measuring cardiac systolic and diastolic been termed ‘ventricular remodeling’.1,2 The cardiomyo- function. In this study, we used invasive hemodynamic cytes lost during an MI cannot be regenerated, and the studies and non-invasive Doppler echocardiography to extent of the loss is inversely related to cardiac output, evaluate the cardiac function of rats with myocardial pressure-generating capacity, and, ultimately, survival.3,4 infarcts after the injection of neonatal cardiomyocytes or Cellular cardiomyoplasty, or the supplementation of tissue culture media alone. In addition, cell survival and cell –cell with exogenous cells, has previously been used in the treat- attachment were analyzed by electron microscopy, and ment of disease in which terminally differentiated cells are immunohistochemistry. irreparably damaged5 and supplementing infarcted myocar- dium with fetal or neonatal myocytes would result in the formation of viable muscle grafts capable of attenuating Methods deleterious post-MI remodeling.6,7 Myocardial Infarction Model Lewis strain male rats (300g, 8 weeks old; Seac (Received December 8, 2003; revised manuscript received March 3, Yoshitomi Ltd, Fukuoka, Japan) were cared for humanely, 2004; accepted March 12, 2004) in compliance with the ‘Principles of Laboratory Animal Departments of Internal Medicine and Cardiology, and *Pharmacol- Care’ formulated by the National Society for Medical Re- ogy, Graduate School of Medicine, Osaka City University, Osaka and search and the ‘Guide for the Care and Use of Laboratory **Third Department of Medicine, Osaka Medical College, Takatsuki, Japan Animals’ prepared by the Institute of Laboratory Animal Mailing address: Minoru Yoshiyama, MD, Department of Internal Resource and published by the National Institutes of Health Medicine and Cardiology, Graduate School of Medicine, Osaka City (NIH Publication No. 86-23, revised 1985). Acute MI was University, 1-4-3 Asahi-machi, Abeno-ku, Osaka 545-8585, Japan induced as described elsewhere.9

Circulation Journal Vol.68, June 2004 Effects of Cellular Cardiomyoplasty on Ventricular Remodeling 581

Fig1. A, M-mode echocardiograms from control rats, and untreated, cellular cardiomyoplasty therapy treated rats with MI at 4 weeks. Note left ventricular cavity dilatation, thin- ning and akinesis of the anterior wall in the rats with MI. Cellular cardiomyoplasty prevented left ventricular dilata- tion and improved anterior wall motion. AW indicates ante- rior wall; and PW, posterior wall. B, Pulsed-wave Doppler spectra of mitral inflow from control rats, and untreated, cellular cardiomyoplasty treated rats with MI at 4 weeks. The mitral inflow pattern of rats with MI shows increased peak E-wave velocity, rapid deceleration of the E wave, and decreased peak A-wave velocity compared with controls. Cellular cardiomyoplasty visually changed a relatively normal transmitral flow pattern in rats with MI.

Fig2. Two-dimensional echocardiograms (apical view) from rats 4 weeks after MI. Four-chamber view is shown at diastole and systole.

Cell Isolation and Purification py (MI + cell). Cells or medium were injected into the Rat neonatal ventricular myocytes were isolated as infarct area with a 26-gauge needle bent at an angle of 45°, previously reported. In brief, apical halves of the cardiac 3 mm from the tip. This configuration was designed to ventricles from 1- to 2-day-old Lewis rats were separated, facilitate injection into the infarct. Rats that survived to minced, and dispersed with 80 U/ml collagenase IV and 24h after coronary artery occlusion were randomized to 0.6mg/ml pancreatin. A discontinuous gradient of Percoll receive either injection of cells (3–5×106/50μl) or medium was used to segregate myocytes from nonmyocytes. After (50μl). Cardiac echocardiography was performed at 24h centrifugation for 30min at room temperature, the upper after MI and 4 weeks after cellular cardiomyoplasty. layer consisted of a mixed population of nonmyocyte cell types, and the lower layer consisted almost exclusively of Hemodynamic and Doppler Echocardiographic Studies cardiac myocytes. After the myocytes were incubated on Rats were lightly anesthetized with ketamine HCl uncoated 10cm culture dishes for 30min to remove any (25mg/kg, ip) and xylazine (10mg/kg, ip). Echocardiogra- remaining nonmyocytes, the nonattached viable cells were phy was performed using a commercially available echocar- used for cell transplantation. diographic system equipped with a 12-MHz phased-array transducer (SONOS 5500, Phillips, Andover, MA, USA). Animal Groups and Cardiomyocyte Transplantation A 2-dimensional short-axis view of the LV was obtained at (Cellular Cardiomyoplasty) the level of the papillary muscles. Ejection fraction was Three groups of animals were studied: control animals measured by modifying Simpson’s method, which uses 4- receiving neither infarction nor implantation and only chamber views (Fig1).10 If the left ventricular chamber injected with culture medium (control), infarcted animals could not be gained clearly, we used contrast echocardio- not receiving neonatal cardiomyocyte therapy (MI), and graphy. The contrast agent, optison, was injected into the infarcted animals receiving neonatal cardiomyocyte thera- left ventricular cavity via the femoral vein (Fig2). Pulsed-

Circulation Journal Vol.68, June 2004 582 YOSHIYAMA M et al.

Fig3. Contrast echocardiograms (apical view) from rats 4 weeks after MI. Untreated and cellular cardiomyoplasty treated rats with MI at 4 weeks were shown. Cellular cardiomyoplasty prevented left ventricular dilatation.

Table 1 Doppler Echocardiographic Measurements lead citrate, and were examined with a Hitachi H-7000 electron microscope.13 Control MI MI + cell n 101112 Immunohistochemistry LVDd (mm) 6.5±0.2 8.8±0.3** 8.2±0.3** Additional sections were obtained from the paraffin LVDs (mm) 4.2±0.1 6.6±0.2** 5.7±0.2**‡ block for the immunohistochemical microscopic study. The FS (%) 35±2 24±2.9** 30±1.7 EDV (μl) 427±15 636±24** 532±22**‡ sections were placed on silanized glass slides (No.S3003, EF (%) 72±5 49±3** 62±3*† DAKO Japan, Kyoto, Japan), which were then dried, E wave (cm/s) 70±5 93±3** 84±5* dewaxed in xylene, and rehydrated in graded concentra- A wave (cm/s) 46±3 13±1** 32±4**† tions of ethanol. After incubation with normal blocking E/A 1.6±0.2 7.3±0.4** 3.1±0.4**‡ serum, the sections were incubated overnight at 4°C with EDV, end-diastolic volume; EF, ejection fraction; FS, fractional shorten- monoclonal antibody against connexin-43 (ZYM#71-0700, ing; LVDd, LV end-diastolic dimension; LVDs, LV end-systolic dimension. Zymed Laboratories, Inc, South San Francisco, CA, USA). *p<0.05 vs Control, **p<0.01 vs Control; †p<0.05 vs MI, ‡p<0.01 vs MI. After incubation with secondary antibody, the sections were allowed to react with Vectastain Elite ABC reagent (Vector Laboratories, Burlingame, CA, USA), then reacted wave Doppler spectra of mitral inflow velocities were with the peroxidase substrate solution (Vectastain 3’3’- recorded from the apical 4-chamber view, with the sample diaminobenzidine substrate kit; Vector Laboratories).14 volume placed near the tips of the mitral leaflets and As a positive control for the immunohistochemical pro- adjusted to the position at which velocity was maximal and cedure, neonatal rat heart fixed in paraformaldehyde was the flow pattern laminar.11,12 Sample volume was set at the treated in the same way. As a negative control, some smallest size available. All Doppler spectra were recorded sections were first incubated with mouse ascites fluid, on paper speed at 200 mm/s and analyzed off-line. The which was obtained from animals injected with non-secret- hemodynamic studies are performed at 4 weeks after MI. ing hybridoma cells, and the treated with the second-stage reagent as described. Light and Electron Microscopy Studies For the light microscopic study, the specimens were Statistics fixed in 10% formaldehyde, embedded in paraffin, and cut All results are expressed as mean± SEM. Statistical into 4-μm-thick sections, which were stained with hema- significance was determined using ANOVA. Differences toxylin and eosin, and Mallory-azan. were considered significant at p<0.05. For the electron microscopic study, the specimens were fixed in 4% paraformaldehyde containing 0.25% glutaralde- hyde and 4.5% sucrose, postfixed in 1% osmium tetroxide, Results dehydrated through passage in a series of graded ethanol, Functional Assessment of the Infarcted Myocardium and embedded in Epon. Ultrathin sections obtained from Before and After Cellular Cardiomyoplasty the embedded blocks were stained with uranyl acetate and The left ventricular end-diastolic dimension (LVDd),

Circulation Journal Vol.68, June 2004 Effects of Cellular Cardiomyoplasty on Ventricular Remodeling 583 fractional shortening (FS) and E/A (E wave=peak early Table 2 Hemodynamics and Ventricular Weights rapid filling wave/A wave=late filling wave because of Control MI MI + cell atrial contraction) at baseline were not significantly differ- ent between the 2 groups of MI rats (non-treated and n 101112 treated groups; LVDd: 7.4±0.2mm, 7.5±0.3mm, NS; %FS: HR (beats/min) 265±7 260±6 267±7 Mean BP (mmHg) 118±4 112±4 110±4 29±3, 30±2, NS; E/A: 3.2±0.9, 3.1±0.8, NS). Left ventricu- LVEDP (mmHg) 4±0 19±1** 11±1**‡ lar cavity size significantly increased in rats with MI at 4 RV (mg/g) 0.58±0.02 0.76±0.03* 0.62±0.04‡ weeks and the rats with MI had significant systolic dys- LV (mg/g) 2.18±0.03 2.16±0.05 2.08±0.04 function, as evidenced by decreased %FS and ejection frac- BW (g) 358±8 352±11 351±7 tion (EF). Cellular cardiomyoplasty significantly prevented MI size (%) – 42±2 28±3‡ the left ventricular cavity dilatation, decreases in %FS and BP, blood pressure; BW, body weight; HR, heart rate; LV, left ventricular EF, and increase in E/A at 4 weeks (Fig3, Table1). wegiht; LVEDP, LV end-diastolic pressure; MI, myocardial infarction; RV, right ventricular weight. Effects of Cellular Cardiomyoplasty on Hemodynamics and *p<0.05 vs Control, **p<0.01 vs Control; †p<0.05 vs MI, ‡p<0.01 vs MI. Heart Weight (Table2) In treated rats, the left ventricular end-diastolic pressure (LVEDP) decreased (MI: 19±1 mmHg, MI + cell: 11± untreated MI group. 1mmHg; p<0.01). Right ventricular weight was significant- ly higher in rats with MI (0.76±0.03mg/g) than in controls Electron Microscopy (0.58±0.02mg/g). Cellular cardiomyoplasty prevented an Under electron microscopy, myocardial cells far from increase in right ventricular weight (0.62±0.04mg/g). the infarcted area had a normal configuration, and fully developed intercalated discs and gap junctions were Light Microscopy and Immunohistochemistry observed (Fig5A,B). On the other hand, those in the vicin- In the MI rats that did not receive cellular cardiomyo- ity of the infarcted area were in various stages of degenera- plasty, replacement fibrosis and interstitial cells were tion, including vacuolar formation, destruction of the mito- prominent, corresponding to the histological findings of an chondrial cristae, partial dissociation of the intercalated old MI (Fig 4). In the cellular cardiomyoplasty group, discs, and necrosis with contraction bands (Fig5C,D). however, the extent of fibrosis tended to be small. Intrigu- In the cellular therapy group, clusters of small cells (2– ingly, spindle-shaped cells, apparently different from fibro- 3μm in width and 6–7μm in length) were observed in the cytes, were often found in the interstitium adjacent to the infarcted area, some of which contained myofilaments with infarcted area. Immunohistochemically, connexin-43 was Z-bands in the cytoplasm suggesting myocardial cells only observed in the cytoplasm of cells in the cellular (Figs5C, 6). The intercalated discs of the myocardial cells therapy group (Fig4) and these connexin-43-positive cells were poorly developed and no gap junctions were observed, were scattered around capillaries preserved in the infarcted which meant such cells were less mature myocardial cells area; they were not found in the control group or the and might be implanted neonatal myocytes. Intriguingly,

Fig4. Light micrographs. Upper and left lower panel (A-D): hematoxylin-eosin staining. Right lower panel (E): immno- histochemistry for connexin-43. Myocardial cells in a sham operated rat showed normal morphology (A). In rats with myocardial infarction, replacement fibrosis and interstitial cells were prominent (B). In rats with cellular cardiomyo- plasty, the extent of fibrosis looks small (C). Spindle-shaped cell (arrow) was often observed only in the cellular cardiomyoplasty therapy group (D). Immunohistochemically, connexin-43 was observed in the cytoplasm of cells (arrows) in the cellular cardiomyoplasty therapy group (Fig4). Scale bar, 10μm.

Circulation Journal Vol.68, June 2004 584 YOSHIYAMA M et al.

Fig5. Light micrograph (C) and electron micrographs (A, B, and D). Ultrastructure of cardiomyocytes far from the infracted area showed normal configuration (A), and developed intercalated disc and gap junction were observed (B). Semi-thin (1μm) section stained with toluidine blue was obtained to make an orientation before the electron microscopy (C). In the cellular therapy group, a cluster of small cells was observed within a circle. One of residual cardiomyocytes after the infarction (arrow) showed a variety of degeneration, including vacuolar formation, disarrangement of myofila- ments, and contraction bands (D). Scale bar, 1μm.

Fig6. Electron micrographs. In the cellular therapy group, a cluster of small cells was observed in the infracted area, one of which contained myofilaments with Z-bands (arrows) and poorly developed intercalated discs suggesting neonatal myocardial cell (A). The rectangular part is enlarged (B). The less mature state of myocardial cell is making close contact with an interstitial cell (arrows). Three or 4 cells making close contact each other in the interstitium, one of which might be a myocardial cell as myofilaments are observed in the cytoplasm (asterisk) (C). An interstitial cell has protruded a part of cytoplasm making contact via pseudopodia with a cardiomyocyte (D).

Circulation Journal Vol.68, June 2004 Effects of Cellular Cardiomyoplasty on Ventricular Remodeling 585 these immature myocardial cells were occasionally making in the MI+cell group, and some of the small cells contained point-to-point contact with interstitial cells (Fig6). myofilaments with Z-bands in the cytoplasm suggesting myocardial cells (Figs5C, 6). Furthermore, the intercalated discs of these cells were poorly developed and no gap junc- Discussion tions were observed, which mean the small cells observed Remodeling of the left after MI, which is a in the MI + cell group were less mature (Fig 5E). Thus, major cause of infarct-related heart failure and death, neonatal myocardial cells implanted on the day following depends on acute and chronic transformation of both the MI might have survived for 4 weeks in the rat heart. necrotic infarct region and the non-necrotic, peri-infarct Another important issue is the cell-cell attachment in the tissue. Despite pharmacotherapeutics and mechanical inter- cellular therapy group. This is the first report to detect ventions, the cardiomyocytes lost during MI cannot be immature myocardial cells making close contact with inter- regenerated. The recent finding that a small population of stitial cells. Under electron microscopy, some interstitial the cells is able to replicate itself is encour- cells were observed to be protruding part of their cytoplasm aging, but is still consistent with the concept that such to make contact with cardiomyocytes (Fig6). It might be regeneration is restricted to viable myocardium.15 Over the that transplanted cardiomyocytes need to make close past several years, increasing awareness of the shortcom- contact with interstitial cells to adjust to the microenviron- ings of and left ventricular assist ment and thus survive in the infarcted area. Further exami- system implantation has led to the consideration of alterna- nation focusing on the cellular contact in cardiomyoplasty tive means of treating end-stage heart failure. In this study, should be done. we showed by echocardiography that cellular cardio- We do not have any data to show that the neonatal myo- myoplasty prevented left ventricular remodeling and that cytes survived in the myocardium. In a previous study, transplanted neonatal cardiomyocyte formed viable grafts 15% of neonatal cardiomyocytes survived for at least 12 in infarcted hearts. weeks when transplanted into healthy rat myocardium.19 In As indicated by echocardiography, the rats with MI had the infarcted myocardium, based on quantitative TaqMan significant systolic and diastolic dysfunction, as shown by PCR analysis, approximately 60% of injected neonatal the significant decrease in %FS, EF and the increase in the cardiac myocytes would have survived, but the actual num- E/A ratio. We found that cellular cardiomyoplasty signifi- ber will be lower because >10% of the injected cells were cantly improved cardiac dysfunction in these rats. Doppler noncardiomyocytes such as fibroblasts or endothelial cells, echocardiography is the technique for evaluating left ven- which may be more resistant to the physical strain of injec- tricular diastolic function16 and increased E wave velocity, tion and and may also have proliferated.20 decreased peak A wave velocity (or absent A wave), and Although global contractile function was increased after rapid E wave deceleration were observed in the rats, and cellular implantation, the reason remains uncertain and these flow patterns were similar to the transmitral flow pro- there may be several mechanisms responsible for the files observed in patients with heart failure with restrictive improved cardiac function. One is that implanted neonatal patterns. Improvement of E/A ratio is caused by preload, cardiomyocyte are actively responsible for force generation afterload reduction, improvement in left ventricular relaxa- during the cardiac cycle. Another is that enhanced cardiac tion, or a decrease in passive elastic properties. Our data do function is more likely to be a result of overall attenuation not directly answer the question of whether the changes in of deleterious ventricular remodeling within the infarcted the E/A ratio are caused by changes in myocardial proper- and viable myocardium rather than an active force genera- ties, changes in left ventricle loading conditions, or both. tion by cardiomyocytes.6 In addition, it is possible that We suppose that the improvements in the E/A ratio resulted growth factors, released by implanted cells, may exert a from a combination of improvements in hemodyamics, as protective effect through stimulation of angiogenesis well as in left ventricular properties. within the infarct and noninfarct regions.21 However, the Cellular cardiomyoplasty might overcome the shortcom- physiology underlying the improved cardiac function ings of each therapy and facilitate superior myocardial remains speculative. regeneration. There are, however, several remaining prob- In conclusion, we have demonstrated that neonatal car- lems with this treatment.17 Even when the myocytes are diomyocytes existed in the damaged myocardium and introduced by needle injection, cell survival is not clear and prevented the dilatation of the impaired myocardial wall. cell transplantation may not be able to recreate the neces- Doppler echocardiography, electron microscopy, and sary microenvironment around the impaired myocardium. immunohistochemistry are useful tools for assessing the We found the expression of connexin-43 in the cytoplasm effect of cellular cardiomyoplasty on the course of MI. of interstitial cells in the cellular cardiomyoplasty group only (Fig4). From the neonatal ventricle, in which the gap References junctions of myocytes are spread over the entire cell sur- 1. Anversa P, Li P, Zhang X, Olivetti G, Capasso JM. Ischaemic face, there is a progressive redistribution toward the adult myocardial injury and ventricular remodelling. Cardiovasc Res pattern in which junctions are confined to the transversely 1993; 27: 145–157. oriented terminals of the myocytes, within the complex 2. Pfeffer MA, Braunwald E. Ventricular remodeling after myocardial structure of the intercalated disk.18 Therefore, connexin-43- infarction: Experimental observations and clinical implications. Circulation 1990; 81: 1161–1172. positive cells observed around capillaries in the cellular 3. Fletcher PJ, Pfeffer JM, Pfeffer MA, Braunwald E. Left ventricular therapy group could be transplanted neonatal myocardial diastolic pressure–volume relations in rats with healed myocardial cells. infarction. Effects on systolic function. Circ Res 1981; 49: 618–626. Ultrastructural observation supported the possibility that 4. Pfeffer MA, Pfeffer JM, Fishbein MC, Fletcher PJ, Spadaro J, Kloner RA, et al. Myocardial infarct size and ventricular function in the spindle-shaped cells observed in the cellular therapy rats. Circ Res 1979; 44: 503–512. group were neonatal myocardial cells; that is, clusters of 5. Deacon T, Schumacher J, Dinsmore J, Thomas C, Palmer P, Kott S, small cells (2–3μm wide and 6–7μm long) were observed et al. Histological evidence of fetal pig neural cell survival after

Circulation Journal Vol.68, June 2004 586 YOSHIYAMA M et al.

transplantation into a patient with Parkinson’s disease. Nat Med histochemistry of the coronary chemoreceptor in human and canine 1997; 3: 350–353. hearts. Am Heart J 1995; 129: 946–959. 6. Li RK, Mickle DA, Weisel RD, Mohabeer MK, Zhang J, Rao V, et 14. Hayashi T, Ijiri Y, Toko H, Shimomura H, Okabe M, Terasaki F, et al. Natural history of fetal rat cardiomyocytes transplanted into adult al. Increased digitalis-like immunoreactive substances in patients rat myocardial scar tissue. Circulation 1997; 96: II-179– II-186; with hypertrophic cardiomyopathy. Eur Heart J 2000; 21: 296–305. Discussion 186–187. 15. Quaini F, Urbanek K, Beltrami AP, Finato N, Beltrami CA, Nadal- 7. Muller-Ehmsen J, Peterson KL, Kedes L, Whittaker P, Dow JS, Ginard B, et al. Chimerism of the transplanted heart. N Engl J Med Long TI, et al. Rebuilding a damaged heart: Long-term survival of 2002; 346: 5–15. transplanted neonatal rat cardiomyocytes after myocardial infarction 16. Nishimura RA, Tajik AJ. Evaluation of diastolic filling of left ventri- and effect on cardiac function. Circulation 2002; 105: 1720–1726. cle in health and disease: Doppler echocardiography is the clinician’s 8. Taylor DA, Atkins BZ, Hungspreugs P, Jones TR, Reedy MC, Rosetta Stone. J Am Coll Cardiol 1997; 30: 8–18. Hutcheson KA, et al. Regenerating functional myocardium: 17. Reinlib L, Field L. Cell transplantation as future therapy for cardio- Improved performance after skeletal myoblast transplantation. Nat vascular disease?: A workshop of the National Heart, Lung, and Med 1998; 4: 929–933. Blood Institute. Circulation 2000; 101: E182–E187. 9. Weisman HF, Bush DE, Mannisi JA, Weisfeldt ML, Healy B. 18. Peters NS, Severs NJ, Rothery SM, Lincoln C, Yacoub MH, Green Cellular mechanisms of myocardial infarct expansion. Circulation CR. Spatiotemporal relation between gap junctions and fascia 1988; 78: 186–201. adherens junctions during postnatal development of human ventricu- 10. Izutani S, Yoshiyama M, Omura T, Yoshida K, Nakamura Y, Kim S, lar myocardium. Circulation 1994; 90: 713–725. et al. Nipradilol can prevent left ventricular systolic and diastolic 19. Muller-Ehmsen J, Whittaker P, Kloner RA, Dow JS, Sakoda T, Long dysfunction after myocardial infarction in rats. Circ J 2002; 66: TI, et al. Survival and development of neonatal rat cardiomyocytes 289–293. transplanted into adult myocardium. J Mol Cell Cardiol 2002; 34: 11. Yoshiyama M, Takeuchi K, Omura T, Kim S, Yamagishi H, Toda I, 107–116. et al. Effects of candesartan and cilazapril on rats with myocardial 20. Muller-Ehmsen J, Peterson KL, Kedes L, Whittaker P, Dow JS, Long infarction assessed by echocardiography. Hypertension 1999; 33: TI, et al. Rebuilding a damaged heart: Long-term survival of trans- 961–968. planted neonatal rat cardiomyocytes after myocardial infarction and 12. Nakamura Y, Yoshiyama M, Omura T, Yoshida K, Kim S, Takeuchi effect on cardiac function. Circulation 2002; 105: 1720–1726. K, et al. Transmitral inflow pattern assessed by Doppler echocardiog- 21. Leor J, Prentice H, Sartorelli V, Quinones MJ, Patterson M, Kedes raphy in angiotensin II type 1A receptor knockout mice with myocar- LK, et al. Gene transfer and cell transplant: An experimental dial infarction. Circ J 2002; 66: 192–196. approach to repair a ‘broken heart’. Cardiovasc Res 1997; 35: 431– 13. Hayashi T, James TN, Buckingham DC. Ultrastructure and immuno- 441.

Circulation Journal Vol.68, June 2004