Author + information
- Received October 29, 2002
- Revision received March 22, 2003
- Accepted March 27, 2003
- Published online July 2, 2003.
- Hiroshi Nakamura, MD, PhD*,* (, )
- Seiji Umemoto, MD, PhD*,
- George Naik, BSc†,
- Gordon Moe, MD, FACC†,
- Satoko Takata, MD*,
- Peter Liu, MD, FACC‡ and
- Masunori Matsuzaki, MD, PhD, FACC*
- ↵*Reprint requests and correspondence:
Dr. Hiroshi Nakamura, Department of Cardiovascular Medicine, Yamaguchi University School of Medicine, 1-1-1 Minami-kogushi, Ube, Yamaguchi, 755-8505 Japan.
Objectives The present study investigated the effects of tumor necrosis factor (TNF)-alpha and angiotensin II (ANG II) on cardiac remodeling and dysfunction at the early stage of acute myocardial infarction (MI) by using a novel heterotopic cardiac transplantation–coronary ligation model.
Background A recent clinical study has demonstrated a possible role of monocytosis in the development of left ventricular (LV) remodeling in patients with acute MI reperfusion.
Methods We performed isogenic heterotopic cardiac transplantation and simultaneous coronary ligation to produce MI in the donor heart and to evaluate the hearts of both donors and recipients in Lewis rats.
Results A significant decrease in LV fractional shortening and positive rate of rise in LV pressure and a significant increase in LV end-diastolic dimension/body weight and LV end-diastolic pressure were observed in the recipient hearts in the ligation group on day 7. TNF-alpha was significantly elevated not only in the plasma but also in the recipient hearts in the ligation group. In contrast, ANG II was significantly increased only in the infarct region of the donor hearts, but not in the plasma. Furthermore, the recipients’ transient LV remodeling and dysfunction were completely abolished by the intravenous administration of a TNF-alpha antagonist.
Conclusions We developed a novel cardiac transplantation–coronary ligation model capable of inducing MI in the absence of downstream hemodynamic effects and allowing differential quantification of indexes of cardiac remodeling in vivo, including the local and remote effects of ANG II and TNF-alpha on cardiac remodeling.
Myocardial infarction (MI) induces global changes in ventricular architecture that are collectively referred to as “post-MI remodeling” (1). A variety of peptides, including norepinephrine, angiotensin II (ANG II), endothelin, aldosterone, and tumor necrosis factor (TNF)-alpha, a pro-inflammatory cytokine, substantially contribute to disease progression in the failing heart (2). In this highly complex environment after MI, it is quite difficult to differentiate the effects of individual neurohormones and cytokines on cardiac remodeling, even when using the conventional MI model. Previous studies have shown that TNF-alpha is expressed by cardiac myocytes contralateral to the infarcted lesion, where myocardial remodeling is ongoing; such production has been shown to play a role in an important intrinsic myocardial stress response to injury (3). Although clinical data have shown that circulating levels of TNF-alpha are elevated in heart failure (HF), the exact biochemical mechanisms responsible for the transition from hemodynamic compensation to decompensation remain unknown (4). A recent clinical study has demonstrated a possible role of monocytosis in the development of left ventricular (LV) remodeling in patients with acute MI reperfusion (5). In the present study, to examine the effects of TNF-alpha and ANG II on cardiac remodeling and dysfunction at the early phase of acute MI, we performed isogenic heterotopic cardiac transplantation and simultaneous coronary ligation in Lewis rats and evaluated the donor and recipient hearts, which were beating independent of each other.
The study protocol was approved by the Institutional Animal Care and Use Committee of Yamaguchi University School of Medicine, in accordance with the guidelines of the American Heart Association.
A total of 256 inbred male Lewis rats (175 to 225 g) were purchased from Charles River, Inc. (Montreal, Canada). Isogenic heterotopic cardiac transplantation was performed according to the method of Ono and Lindsey (6). Proximal left anterior descending coronary artery ligation in the donor heart was performed immediately after the transplantation and at the same surgical intervention site (Fig. 1). Myocardial ischemia was confirmed by regional cyanosis. As a sham operation, nonligated hearts were transplanted into the same strain of rats. The viability of the donor heart was assessed based on the beating of the transplanted heart. To examine the effects of the TNF-alpha receptor antagonist (TNFR:Fc), a TNF-alpha antagonist (Amgen, Thousand Oaks, California) that is a chimeric fusion protein consisting of the extracellular domain of the Fc portion of the immunoglobulin G1molecule, on both recipient and donor cardiac function, six rats in the ligation group were post-surgically treated with a single subcutaneous injection of TNFR:Fc at a dose of 1.5 mg/kg body weight (7). The effect of recipient heart MI on the donor heart was not measured, owing to high mortality.
Both donor and recipient hearts were weighed, and 5-μm serial sections stained with hematoxylin and eosin were used for quantitative histologic analysis to determine the LV dimension, wall thickness, and infarct size (%MI), according to the method of Pfeffer et al. (1). The perivascular index of the recipient heart was examined and scored from 0 to 2, using all cross sections stained with Masson’s trichrome solution, according to the method of Matsusaka et al. (8). The average score was calculated for each heart.
On days 7 and 21 after surgery, body weight and body temperature were measured, and systolic blood pressure and heart rate (HR) were obtained by the tail-cuff method. Both recipient and donor hearts underwent echocardiographic and hemodynamic evaluation.
Left ventricular end-diastolic dimension (LVEDD), LV end-systolic dimension (LVESD), and fractional shortening (%FS) were obtained by averaging three cardiac cycles using an echocardiographic system (HP SONOS 5500; Hewlett Packard, Palo Alto, California) with a dynamically focused 12-MHz linear-array transducer under light anesthesia with ether. Next, LV function (intracavitary LV systolic and diastolic pressures) in both hearts was recorded as previously described (9).
White blood cell (WBC) counts were measured with an auto cell counter on days 7 and 21 after surgery. To test for the presence of intestinal ischemia due to surgery, we measured plasma endotoxin concentrations using an endotoxin-specific chromogenic test (ES test; Seikagaku Kogyo, Tokyo, Japan) on day 7, according to the method of Obayashi et al. (10). To evaluate monocyte activity, the plasma concentrations of monocyte chemoattractant protein (MCP)-1 were determined with an enzyme-linked immunosorbent assay kit for rat MCP-1 (Biosource, Camarillo, California). Immunoenzymatic staining was performed with a DAKO LSAB kit (DAKO, Carpinteria, California) according to the manufacturer’s instructions. Antibody against rat TNF-alpha (Cedarlae, Hornby, Ontario, Canada) was used as the primary antibody.
Tissue and plasma concentrations of TNF-alpha were determined as described previously (3). The tissue ANG II content was determined by the method of Naik et al. (11). Plasma ANG II concentrations were measured by radioimmunoassay with florisil (12). Expression of TNF-alpha messenger ribonucleic acid (mRNA) in the recipient heart was measured by semiquantitative polymerase chain reaction analysis (3). GAPDH was amplified using the primers GAPDHF 5′-GCCAAGTATGATGACATCAA-3′ and GAPDHR 5′-CCATATTCATTGTCATACCA-3′ (product size, 202 base pair). TNF-alpha was amplified using the primers TNF-alphaF 5′-ATGAGCACAGAAAGCATGATCC-3′ and TNF-alphaR 5′-GAAGATGATCTGAGTGTG-3′ (product size, 251 base pair). For zonal quantification, the donor heart with MI was divided into three pieces (infarct, border, and contralateral zone), and in the case of donor hearts in the sham group and recipient hearts, a whole heart was used because there were no regional differences comparable to those measured in the MI group.
Data are expressed as the mean value ± SD. The Student ttest was used to compare the differences between the two groups. A p value <0.05 was considered statistically significant.
The operation yielded 48 cardiac transplantation–infarction model rats from an initial total of 50 rats (96%). A total of 39 (98%) of 40 rats survived in the sham group. No graft failure of the donor heart or gut ischemia occurred in either group. Rats of the ligation group showed a significant decrease in body weight and systolic blood pressure compared with those of the sham group on day 7 (Table 1). In contrast, body temperature and HR were significantly increased on day 7 in the ligation group compared with the sham group. There were no significant differences in these parameters on day 21 between the sham and ligation groups. Although the WBC counts and plasma MCP-1 levels were significantly increased in the ligation group on day 7, compared with the sham group, no statistical difference was seen in the plasma endotoxin levels.
In the donor heart of the ligation group on day 7, the heart weight, LVEDD, and LVESD were significantly increased, whereas the peak systolic pressure, +dP/dt, and −dP/dt were significantly decreased compared with those values in the sham group (Table 2). There were no significant differences in the wall thickness of the noninfarcted area, %FS, HR, or LV end-diastolic pressure (LVEDP) in the donor heart between the two groups. On day 21, the donor hearts in the ligation group demonstrated significant LV scar formation. The heart weight, LVEDD, and LVEDP were significantly higher, and %FS significantly lower, in the ligation group than in the sham group. No significant differences in the wall thickness of the noninfarcted area, LVESD, HR, peak systolic pressure, +dP/dt, or −dP/dt were seen between the two groups.
In the recipient hearts of the ligation group on day 7, LVEDD/body weight, HR, and LVEDP were significantly increased, whereas %FS and +dP/dt were significantly decreased compared with those values in the sham group (Table 3). No significant differences in the heart weight/body weight, perivascular index, LVESD/body weight, wall thickness, peak systolic pressure, or −dP/dt of the recipient heart were observed between the two groups on day 7. On day 21, only the perivascular index was significantly increased in the recipient heart of the ligation group compared with the sham group. Plasma TNF-alpha concentrations in the ligation group were significantly higher than those in the sham group on days 2, 4, and 7 after surgery, with a peak occurring on day 7. In contrast, no significant difference was seen in plasma ANG II concentrations between the two groups throughout the experiment (Fig. 2). There were no relationships between plasma TNF-alpha and %MI (r = 0.45, p = 0.17) or between ANG II and %MI (r = 0.48, p = 0.17) on day 7.
We further analyzed the zonal production of TNF-alpha and ANG II protein in the heart tissues of the sham and ligation groups on day 7 (Fig. 3). Tissue TNF-alpha concentrations in the recipient hearts of the ligation group were significantly elevated compared with the recipient hearts in the sham group. In contrast, in the donor heart, the tissue ANG II content was significantly higher in the infarct region than in other regions after surgery. There were no differences in tissue TNF-alpha concentrations among all regions in the donor hearts, and the tissue ANG II content did not differ between the recipient hearts of the two groups. In contrast, TNF-alpha mRNA in the recipient heart was significantly lower in the ligation group than in the sham group by 19% (Fig. 4).
Immunohistochemical analysis showed that TNF-alpha was positively stained in the cytoplasm of monocytes (yielding a brown color with a pale blue background) taken from the vessels of recipient hearts in the ligation group (Fig. 5). Neither monocytes nor myocytes were positively stained with TNF-alpha antibody in the sham group.
We further analyzed the effects of TNFR:Fc on the function of recipient hearts of the ligation group. A single injection of TNFR:Fc at the time of surgery significantly decreased LVEDD and LVEDP levels, whereas −dP/dt increased compared with the TNFR:Fc nontreated ligation group on day 7 (LVEDD: 3.20 ± 0.47 vs. 2.11 ± 0.17 mm/10−2g [nontreated vs. treated group]; LVEDP: 16.1 ± 2.5 vs. 3.0 ± 2.7 mm Hg; −dP/dt: 4,146 ± 1,718 vs. 6,039 ± 530 mm Hg/s; n = 6; p < 0.05). There were no significant differences in the morphometric, echocardiographic, or hemodynamic assessment between the sham group treated with TNFR:Fc and the untreated sham group.
In our study, we found no significant difference in plasma ANG II concentrations between the two groups at any time during the experiment, whereas the tissue ANG II content in the donor heart was significantly higher in the infarct region than in other regions on day 7 after surgery. Left ventricular remodeling and infarct healing after acute MI appeared to be dynamic and time-dependent processes that progressed in parallel to each other. Healing after acute MI involves inflammatory cell infiltrations followed by fibroblast proliferation, collagen deposition, and remodeling in the infarct zone. Structural LV remodeling after acute MI involves early regional infarct expansion and progressive global LV dilation (13). There is a general consensus that acute MI is associated with an overexpression of the renin-angiotensin system, which promotes structural LV remodeling, with ANG II as the primary mediator. Inhibition of ANG II formation with angiotensin-converting enzyme (ACE) inhibitors after acute MI has proven to be effective overall as an anti-remodeling therapy (14). One of the more interesting features in our study was the lack of a role for ANG II. Most of the effects of ANG II are autocrine/paracrine in nature (8), possibly because ligation of the donor heart does not seem to perturb systemic hemodynamics in a major way. Accordingly, systemic renin-angiotensin system activation in the setting of acute coronary ligation may be a reflection of a perturbation of systemic hemodynamics, and the activation of pro-inflammatory cytokines may be more of a response to ischemic tissue injury and not necessarily related to systemic hemodynamics.
It has been reported that early use of certain anti-inflammatory agents promotes LV remodeling in patients with MI (15). Clinical studies have demonstrated the presence of activated circulating neutrophils, lymphocytes, and monocytes, as well as increased levels of pro-inflammatory cytokines such as TNF-alpha and C-reactive protein (16). Studies of LV dysfunction have revealed that baseline WBC counts are an independent predictor of mortality, and WBC counts after acute MI have been proposed as a means of risk stratification in patients with LV dysfunction (17). After the early appearance of neutrophils subsequent to acute MI, monocytes migrate, interacting with adhesion molecules, into the infarct zone, where they transform into macrophages, which are then activated. The macrophages and monocytes secrete several factors, such as macrophage colony–stimulating factor and MCP-1, which lead to peripheral monocytosis and monocyte infiltration into the infarct zone, indicating that acute monocyte recruitment may be involved in infarct healing (5,18). Recently, Maekawa et al. (5)have found that the peak monocyte count is an independent predictor of pump failure and that the count correlates positively with LV end-diastolic volume and negatively with the ejection fraction in patients with acute MI reperfusion, suggesting a possible role of monocytosis in the development of LV remodeling. Although we did not measure monocyte counts in this study, the WBC count and plasma MCP-1 concentrations, but not the plasma endotoxin levels (a marker of exopyrogen), were significantly increased in the ligation group on day 7, suggesting the activation of lymphocytes but not neutrophils. As such, the monocyte/macrophage-rich pathologic process may be involved in the development of cardiac remodeling in our study model.
Using the conventional MI model, Irwin et al. (3)demonstrated that TNF-alpha is not confined strictly to the infarct or peri-infarct zone but is expressed by cardiac myocytes within the myocardium contralateral to the infarct. They also suggested that expression of TNF-alpha mRNA in the contralateral noninfarct zone persists until 35 days after surgery. In our study, tissue TNF-alpha in the recipient heart, but not in the donor heart, and plasma TNF-alpha concentrations were significantly higher in the ligation group than in the sham group on day 7, whereas TNF-alpha mRNA in the recipient heart was significantly lower in the ligation group than in the sham group. Although we did not examine the expression of inflammatory cytokines such as intercellular adhesion molecule-1 or vascular cell adhesion molecule, TNF-alpha-expressing monocytes were observed only in the lumen of the small vessels in the recipient heart of the ligation group. We also observed little migration of monocytes into the recipient myocardium, suggesting that TNF-alpha in the recipient heart may be derived from circulating monocytes. These cytokines trigger the release and production of other cytokines (19). Innate immunity could also provide a response to ischemic stress, even in the same body, which is known as a “systemic inflammatory response” (20). Activation of autoantibodies may be another candidate for the systemic response to ischemic stress (21). Conversely, the expression of TNF-alpha mRNA in the recipient heart of the ligation group might be suppressed by elevated plasma TNF-alpha levels produced by peripheral inflammatory cells such as monocytes activated by myocardial ischemic stress, and TNF-alpha was not positively stained in the recipient myocardium of the ligation group, suggesting that recipient heart tissue may not contribute to the TNF-alpha production in this model. In contrast to the recipient heart, strong TNF-alpha staining was revealed in the infarct and peri-infarct zones of the ligated donor heart, which is in agreement with the findings in the conventional MI model (3; data not shown). Additionally, although the tissue TNF-alpha concentration did not vary in different regions of the donor heart, we cannot exclude the possibility that TNF-alpha mRNA levels may vary from region to region.
Recent studies have focused on the role of TNF-alpha in LV function and remodeling, as well as the development of HF (22). In the present study, we showed that LV remodeling and dysfunction of recipients on day 7 can be prevented by a single intravenous infusion of TNFR:Fc at the time of surgery. These results suggest that TNFR:Fc may bind to circulating plasma TNF-alpha derived from both circulating monocytes and the donor heart in which ischemia is induced, rendering TNF-alpha less active (7). In addition, elevated circulating TNF-alpha on day 7, which is fairly late in the healing process in the conventional MI model, may be due to differences in activation of the neurohumoral and cytokine system and may play an important role in LV dilation and dysfunction of the recipient heart, which is remotely located from the remodeling donor heart. On day 21, we observed no differences in plasma TNF-alpha or ANG II, recipient cardiac function, or LV dimension between the two groups, indicating that the remote suppression effects of TNF-alpha on recipient cardiac function might be transient. One exception was that the perivascular index in the recipient heart was significantly higher in the ligation group than in the sham group. Macrophages secrete cytokines that stimulate fibroblast proliferation and collagen deposition, which are important during infarct healing (5). In our study, overproduction of TNF-alpha, but not of ANG II, in the peripheral cells around day 7 may have contributed to the peri-arterial fibrosis seen in the recipient hearts on day 21. In this study, we did not measure fibrosis and collagen in the infarct and noninfarct zones and its correlation with remodeling, or the effects of the TNF-alpha antagonist on fibrosis and collagen, although collagen and the extracellular matrix are important determinants of remodeling after MI (8,9).
Although clinical trials of patients with post-MI and progressive chronic HF have indicated that ACE inhibitors have a significant long-term benefit on morbidity and mortality (23), clinical studies of Etanercept, a TNFR:Fc, have failed to show additional benefits in patients with chronic HF who have already been given an ACE inhibitor (24). There are several explanations for the negative results of anti-TNF-alpha therapies in clinical trails of HF. First, the biologic agents used in the trials had intrinsic toxicity. Second, TNF-alpha antagonists have adverse effects in the setting of HF (21). Finally, all of the trials using the TNF-alpha antagonist focused on the chronic phase, when cardiac remodeling had already been established, but not on the acute phase after MI. Our results with the TNF-alpha antagonist are consistent with previous reports (7,21)and, taken together with the results of a recent clinical study (24), indicate that using the TNF-alpha antagonist to neutralize activated cytokines during the acute phase of ischemic stress/reperfusion injury might be a beneficial strategy for the treatment of cardiac dysfunction and subsequent cardiac remodeling after acute MI.
Recently, Depre et al. (25)introduced an unloaded heart model in which the donor heart was isolated from the recipient heart and showed a parallel-circuit circulation distinct from the series-circuit circulation of the recipient. In their study, when donor hearts of sham rats were compared with recipient hearts of sham rats, the donor cardiac function was less than the recipient cardiac function, which maintains the systemic blood supply. In our experiment, we performed an additional coronary ligation and obtained MI in the donor heart. Infarct wall thinning and expansion in the absence of an increase in noninfarcted wall thickness are consistent with remodeling in the donor heart with MI. This phenomenon may have caused some differences in the results between our current model and the conventional MI model (1,3). The most important difference between these two models is not the loading condition but the ischemia in the donor heart. In our model, the donor heart is located parallel to the recipient circulation, so that the donor heart continues to beat independent of the recipient hemodynamics, with a higher survival rate. Moreover, our model allows one to examine the difference between the local and remote effects of cytokines and neurohormones—as activated by cardiac ischemia—on cardiac dysfunction and remodeling after acute MI. Note that practitioners should be cautions about extrapolating the results from our model and applying them directly to humans.
We have developed a novel heterotopic cardiac transplantation–coronary ligation model capable of inducing MI in the absence of downstream hemodynamic effects and allowing differential quantification of indexes of cardiac remodeling in vivo,such as the local and remote effects of ANG II and TNF-alpha on cardiac remodeling. We believe that our model could be used in trials of cytokine modulation of post-MI remodeling and will provide new insights into LV remodeling and dysfunction.
☆ This study was supported in part by the Idiopathic Cardiomyopathy Research Group of the Ministry of Health, Labor, and Welfare of Japan; the Heart and Stroke Foundation Award; and the Evelyn McGloin Fellowship Award in Canada.
- angiotensin-converting enzyme
- ANG II
- angiotensin II
- rate of rise in left ventricular pressure
- fractional shortening
- heart failure
- heart rate
- left ventricular
- left ventricular end-diastolic dimension
- left ventricular end-diastolic pressure
- left ventricular end-systolic dimension
- monocyte chemoattractant protein
- myocardial infarction
- tumor necrosis factor
- tumor necrosis factor-alpha receptor antagonist
- white blood cell
- Received October 29, 2002.
- Revision received March 22, 2003.
- Accepted March 27, 2003.
- American College of Cardiology Foundation
- Pfeffer M.A.,
- Pfeffer J.M.,
- Fishbein M.C.,
- et al.
- Mann D.L.
- Irwin W.M.,
- Mak S.,
- Mann D.L.,
- et al.
- Maekawa Y.,
- Anzai T.,
- Yoshikawa T.,
- et al.
- Bozkurt B.,
- Kribbs S.B.,
- Clubb F.J. Jr.,
- et al.
- Obayashi T.,
- Tamura H.,
- Tanaka S.,
- et al.
- Morimoto T.,
- Aoyama M.,
- Gotoh E.,
- et al.
- Barron H.V.,
- Cannon C.P.,
- Murphy S.A.,
- et al.
- Mann D.L.
- Feldman M.A.,
- Combes A.,
- Wagner D.,
- et al.
- Louis A.,
- Cleland G.F.J.,
- Crabbe S.,
- et al.