INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

HEART FAILURE OF both ischemic and nonischemic origin is characterized by cardiomyocyte loss and scarring. The loss of myocardial mass can result in ventricular remodeling that is known to affect patient prognosis (20, 26). Several studies have attempted to identify the mechanism for the progression from myocardial infarction to end-stage heart failure (1, 8, 24). However, the critical events that lead to the progressive deterioration of left ventricular function are unknown.

Apoptosis is a mechanism of cell death that differs significantly from necrosis in morphological characteristics, regulation, and potential for pharmacological manipulation (19). Recently, many studies have demonstrated the occurrence of apoptosis during or after myocardial infarction (6), end-stage heart failure (22, 23), and acute ischemia-reperfusion (12) as well as in hypoperfused hibernating myocardium (5). Although it is speculated that myocyte apoptosis is likely involved in the progression of end-stage heart failure (22), apoptosis in patients with less severe heart failure has not been extensively examined. Furthermore, little is known about the signals that induce and/or modulate myocardial apoptosis. If apoptosis does indeed contribute to the pathophysiology of cardiomyopathy, the therapeutic manipulation of these signals will comprise an important future strategy to delay or prevent disease progression (7).

The receptor FAS is a type I membrane receptor protein and a member of the tumor necrosis factor receptor superfamily (17). FAS is expressed by a variety of cells including T cells, hepatocytes, keratinocytes, and other cell types of the thymus, lung, and ovary (11, 16, 28). In cells that express FAS and downstream components of FAS signal transduction, activation of the receptor by the FAS ligand or by activating antibodies to FAS can induce apoptosis in vitro and in vivo (13, 20). Although the expression of FAS by the myocardium in vivo has not been extensively investigated, recent studies in vitro have demonstrated induction of FAS expression in cultured neonatal rat cardiomyocytes by exposure to hypoxia (31).

In both the rat and human heart (6, 25), apoptosis has been observed in viable myocytes that remain after myocardial infarction. Recent examinations of dogs with chronic heart failure (30) have found myocyte apoptosis in regions immediately adjacent to fibrous scars or old infarcts. For these reasons, we hypothesized that the induction of apoptosis in analogous regions of the human heart might be facilitated by the expression of FAS by the cardiomyocytes or other supporting cells within those regions. We report here the finding of FAS expression adjacent to fibrotic foci and its relationship to DNA end labeling in biopsies of failing human heart.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tissue samples. Human heart tissue was obtained during coronary artery bypass grafting (CABG) performed on individuals with coronary artery disease who required this procedure as part of their treatment at Evangelismos General Hospital, Athens, Greece. Patients who were undergoing CABG and had left ventricular systolic dysfunction were selected for the study. Biopsies were obtained from the left ventricle of seven male patients. The sample was taken during the operation while the patient was on extracorporeal circulation; the site of the biopsy was near the apex in an area far from the site of a visible scar. A small sphenoid biopsy was excised, and the incision was closed with a single suture. All tissues were immediately cooled to 4°C and within 5 min were fixed in 10% neutral buffered Formalin. The fixed tissues were embedded in paraffin and sectioned at 5-µm thickness. All tissue samples were coded, and clinical data of the patients were not made available until after the study. The sampling protocol was approved by the Ethical Committee and the Scientific Board of Evangelismos General Hospital, and informed consent was obtained from all patients.

In situ end labeling. Tissue sections were deparaffinized by passing through three changes of xylene, then xylene-alcohol (1:1) for 15 min and 100% ethanol and 70% ethanol for 10 min each. In situ end labeling (ISEL) of fragmented DNA was performed by a modification of the method of Wijsman et al. (35). Briefly, the samples were incubated with 3% hydrogen peroxide to block endogenous peroxidase activity and then with proteinase K to permit access of reagents to intracellular DNA. The samples were then incubated for 2 h at 18°C in ISEL solution (0.001 mM digoxigenin-11-dUTP, 20 U/ml DNA polymerase I, and 0.01 mM each of dATP, dCTP, and dGTP). Afterward, the heart sections were treated with antidigoxigenin alkaline phosphatase, which was detected with a Fast Blue chromogen system.

Immunohistochemistry. For detection of FAS, deparaffinized heart sections were incubated with 1% BSA in PBS to block nonspecific binding. The sections were then incubated for 24 h at 4°C with monoclonal anti-human FAS (clone CH-11, Upstate Biotechnology, Saranec Lake, NY) at 1:200 dilution in PBS containing 1% BSA. Secondary antibody (biotinylated anti-mouse IgM) was applied for 2 h at 37°C in a humidified chamber; the diluent was 1% BSA in PBS containing 2 µg/ml DNase-free ribonuclease (RNase, Boehringer Mannheim, Indianapolis, IN). The biotinylated secondary antibody was detected with ABC detection system (Vector Laboratories, Burlingame, CA) and diaminobenzidine. The stained sections were mounted in Fluoromount solution (Southern Biotechnology, Birmingham, AL) containing 0.5 µg/ml propidium iodide (Sigma, St. Louis, MO) for fluorescence detection of chromatin.

Collagen detection and microscopy. Collagen was detected under polarized light on heart sections prepared by the Picrosirius red technique performed as previously described (10). Hematoxylin and eosin staining was performed by standard protocols. Photomicroscopy was performed with an Olympus BH2 fluorescence/phase-contrast microscope fitted with the PM10ADS automatic photography system. Propidium iodide fluorescence was detected under epifluorescence illumination with 475- to 510-nm excitation filter and >590-nm emission filter.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Patients. Clinical data of the patients are shown in Table 1. All patients with the exception of patient 2 had angina before operation. Patient 2 suffered an acute inferior myocardial infarction and acute mitral regurgitation due to rupture of chordae tendinae and was operated on 10 days after the acute event. All patients except patient 3 had previously documented myocardial infarction; all but patient 5 were hypertensive. As discussed further below, no clear relationship was observed between the clinical data and the degree of labeling by ISEL, FAS immunohistochemistry, or Picrosirius red procedures individually; however, a spatial association between two of the labels was found in all patient samples. Depending on location within the biopsy, samples exhibited varying numbers of leukocytes (data not shown), scarring, and regions devoid of nuclei.

ISEL of the myocardium. All samples were subjected to ISEL of fragmented DNA to identify cells undergoing apoptosis or DNA repair (Fig. 1). Cells were scored as apoptotic if the nucleus was ISEL positive (blue) and also displayed chromatin condensation against nuclear envelop and/or nuclear fragmentation (Fig. 1B, arrows denote representative examples). ISEL-positive nuclei were observed in all samples without apparent association to clinical findings (data not shown). In agreement with earlier studies, isolated regions of heavy ISEL were observed in some samples, usually near scars detected by Picrosirius red (see below). No ISEL was observed in regions of myocardium devoid of nuclei (data not shown), nor in control samples prepared without biotinylated dUTP (Fig. 1D).

View larger version (146K): [in this window] [in a new window]   Fig. 1.   Identification of fragmented DNA and collagen in human left ventricular biopsies. A: hematoxylin and eosin staining of a biopsy from patient 2. Note nuclear size and morphology. Bar = 50 µm. B: in situ end labeling (ISEL) (see MATERIALS AND METHODS) of biopsy from patient 2; same region as in A is shown. Positive reaction is dark blue due to chromogen Fast Blue. Note labeling of chromatin condensed against inner surface of nuclear envelope (arrows), consistent with apoptosis. C: region of less frequent labeling, also from patient 2. D: negative labeling of replicate section prepared in identical fashion but without digoxigenin-labeled dUTP. Bar = 50 µm; where not shown, magnification is same as in A. E and F: detection of collagen in left ventricular biopsies. Heart sections were prepared by Picrosirius red technique (10) and viewed under transmitted (E) and polarized light (F). Under polarized light, collagen appears as yellow-white. Note perivascular collagen within biopsy from patient 2 (same field in both panels). Bar = 200 µm.

Fibrosis and nuclear morphology. Collagen was detected as yellow-white color when viewed under polarized light in samples prepared by the Picrosirius red technique (10). Collagen was found in perivascular fibrotic foci (Fig. 1, E and F) scattered throughout the myocardium and in and around isolated scars (Fig. 2A). Fibrosis was detected in all patient samples, albeit to varying degrees.

View larger version (87K): [in this window] [in a new window]   Fig. 2.   Relationship of fibrosis and nuclear morphology in failing human myocardium. Sections were stained by Picrosirius red technique or with DNA intercalator propidium iodide following digestion of RNA with DNase-free RNase (see MATERIALS AND METHODS). A: Picrosirius red detection of collagen (white) identifies border between fibrotic (left) and nonfibrotic myocardium (right). B: propidium iodide staining of same region reveals nuclei in nonfibrotic tissue (right) but absent or decayed nuclear remnants in fibrotic area (left). Intensity of propidium iodide red fluorescence (>570 nm) is proportional to nuclear DNA content (32). Bar = 50 µm. C: higher magnification of nuclei immediately adjacent to border reveals chromatin morphologies suggestive of apoptotic bodies (arrowhead). Bar = 20 µm.

Immunohistochemistry for FAS. Monoclonal antibody CH-11 raised against human FAS antigen labeled both large and small isolated regions of myocardium (Fig. 3, A and B). When the same region of tissue was localized in serial sections prepared simultaneously for evaluation of histology, fibrosis, and FAS (Fig. 4), the expression of FAS was found to occur adjacent to the fibrotic foci. This relationship held regardless of whether the fibrosis was perivascular (Fig. 4, B and C, left) or whether it was associated with a larger area of scar distal to vessels (Fig. 4, B and C, right). Simultaneous FAS and propidium iodide labeling of the same tissue section indicated that immunoreactivity to the anti-FAS monoclonal antibody (Figs. 5A, bottom) was detectable on viable heart cells adjacent to fibrosis (compare with Fig. 5B, bottom), but not within the fibrotic focus (Fig. 5, A and B, top).

View larger version (59K): [in this window] [in a new window]   Fig. 3.   Immunoreactivity of failing human myocardium to anti-FAS antibodies. Biopsies were immunostained with monoclonal antibodies to human FAS (clone CH-11) and were treated simultaneously with DNase-free RNase and propidium iodide (32). A: regions of intense reactivity and small isolated foci were found in all samples; positive reactivity is black by diaminobenzidine reaction (arrows). Biopsy shown is from patient 7; note vessels within regions of intense reactivity (bottom right). B: larger diffuse foci of immunoreactivity also were observed without apparent associations with vessels. Bar = 200 µm.

View larger version (94K): [in this window] [in a new window]   Fig. 4.   Relationship between fibrosis and immunolabeling for FAS. A: hematoxylin and eosin preparation of first of 3 serial sections of patient 4 biopsy; in all 3 panels, arrow denotes same vessel. B: Picrosirius red preparation of second adjacent serial section from patient 4, same area as in A. Note collagen surrounding vessel and within vertical scar at far right edge of image. C: FAS immunoreactivity of third of 3 serial sections, same area; note FAS labeling adjacent to fibrosis associated with either scar or vessels (arrow). Bar = 100 µm.

View larger version (108K): [in this window] [in a new window]   Fig. 5.   Relationship of immunohistochemistry for FAS and nuclear morphology. A: anti-FAS immunoreactivity of same field depicted in B. B: propidium iodide fluorescence of "border" between fibrotic (top-most area devoid of nuclei) and nonfibrotic myocardium, as defined in Fig. 4. Bar = 30 µm.

View larger version (90K): [in this window] [in a new window]   Fig. 6.   Relationship of FAS immunolabeling and fragmented DNA. A: anti-FAS immunoreactivity in first of 2 serial sections of patient 3 biopsy; positive labeling is dark brown from diaminobenzidine. B: ISEL of fragmented DNA in second adjacent serial section, same region as in A. Black rectangle denotes same region of FAS-positive cells in all 3 panels. C: higher magnification of ISEL-positive nuclei (dark blue) within second serial section. Note ISEL-positive nuclei within FAS-positive cells denoted by black rectangle (compare with A). Bar = 25 µm. See text for quantitation of ISEL-positive nuclei.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The results of ISEL studies confirm that DNA fragmentation is present in patients with systolic dysfunction. Our results are comparable to those of Narula et al. (22) and Olivetti et al. (25), who found >5% and <1%, respectively, of apoptotic myocytes in patients with heart failure or with myocardial infarction. In contrast, our ISEL data found fragmented DNA in 60-70% of total nuclei. As discussed by Grasl-Kraup et al. (13), DNA labeling techniques such as ISEL and terminal deoxynucleotidyl transferase-mediated nick end labeling do not discriminate between apoptotic, necrotic, or autolytic cells, and thus the number of cells actually undergoing apoptosis is likely much less than that suggested by the ISEL data. Although the numbers of apoptotic cells detected in these studies differ, the methods of detection and scoring clearly differed; in addition, differences in the hemodynamic, hypoxic, or other signals that may trigger apoptosis in the heart, as well as the timing and location of biopsy in relation to these signals, all could account for disparities in the percentage of apoptotic cells observed by various research groups.

Regardless, the occurrence of apoptosis within cells of any type comprising the heart parenchyma may have an important impact on the remodeling of the left ventricle. Apoptosis is believed to play an important role in myocardial cell death in the acute phase of myocardial infarction, not only near the infarct but also in remote areas (18, 25). However, no data exist to indicate whether this situation persists chronically. Interestingly, the biopsy that displayed the greatest percentage of ISEL-labeled nuclei was from the single patient (patient 2) who was biopsied within days, rather than years, after infarction (Table 1). Whether or not the incidence of apoptosis is related to time from myocardial infarction is not clear but will constitute an interesting topic for future study. Although our data suggest that apoptosis persists for many years postinfarction in patients with systolic dysfunction, it is unknown if apoptosis persists in myocardial infarction patients with small infarcts without systolic dysfunction. After infarction, the surviving portion of the left ventricle undergoes a lengthening with a secondary volume overload and progressive hypertrophy (26). Wall thinning and myocyte hypertrophy are greater near the infarcted tissue (2). Cell loss, myocardial fibrosis, myocyte lengthening, and slippage of cells all participate in the compensation of the heart in this setting (4). In ischemic cardiomyopathy, fibrotic foci are found with or without a healed infarct; these foci can include scarring from previous infarcts, perivascular/interstitial fibrosis, as well as microscopic fibrotic foci remote from the area of the infarction (4).

In the present study, expression of FAS was observed in association with perivascular fibrosis and with fibrous scars. Although individual microscopic foci of fibrosis were not observed in association with microscopic foci of FAS expression, it is possible that such an association exists but was difficult to discriminate in large or diffuse regions of FAS expression. Regardless, our finding of FAS expression adjacent to either perivascular fibrosis or adjacent to scars suggests that a common mechanism may underlie the association. With regard to noncardiac cell types, the role of extracellular matrix components in the regulation of cell death is being investigated actively (3, 15). Although the influence of extracellular matrix components on myocyte death is largely unknown, it seems reasonable to speculate that direct interaction of myocytes with the altered extracellular matrix of fibrotic foci may act to modulate or induce FAS expression and/or the subsequent susceptibility of myocytes to apoptotic signals.

Circulating and tissue hormones and/or hemodynamic overload may trigger fibroblast proliferation and accumulation of collagens in the absence of cell loss (9, 25, 34). On the other hand, death of individual myocytes occurs in ischemic heart disease and leads to the formation of foci of "replacement fibrosis"; recently, an increased collagen accumulation has been described in the left and right ventricle of patients with ischemic cardiomyopathy (29). Interestingly, the in vitro studies of Tanaka et al. (31) have shown that embryonic cardiomyocytes in culture begin to express FAS after exposure to hypoxia. It is not yet known if hypoxia also induces FAS expression in the adult human myocardium; however, cells adjacent to or overlying fibrotic foci might reasonably be expected to become hypoxic if perfusion is compromised due to capillary reorganization near the scar. Expression of functional FAS by the affected cells would then be expected to predispose those cells to apoptosis inducible by FAS ligand present in serum or on infiltrating inflammatory cells.

Our observations are consistent with this hypothesis; variable numbers of leukocytes were observed in all patient samples (data not shown). In addition, some parenchymal cells that expressed FAS were found to label also by the ISEL technique (Fig. 6), suggesting that at least some of the cells expressing FAS also were undergoing apoptosis. On the other hand, regions of the parenchyma that were positive for FAS but negative for ISEL also were observed. Furthermore, quantitation revealed that ISEL-positive nuclei were present in the same abundance within regions that were FAS negative as in those that were FAS positive (see RESULTS). The latter results suggest that other initiators of apoptosis may be equally if not more important than FAS in the failing heart. This interpretation assumes that the expression of FAS is maintained by cells throughout the later stages of apoptosis that are detected by techniques such as ISEL. Whether or not this assumption is correct will require kinetic studies of both FAS expression and DNA end labeling in isolated heart cells.

Regardless, these observations are consistent with the notion that expression of FAS by myocardial cells in itself is not sufficient to induce apoptosis but may render cells "competent" for induction of the death pathway pending contact with activators of the receptor and/or other stimuli. A confirmation of this hypothesis may be possible by careful sequential studies of an animal model of heart failure supplemented with exogenously administered ligators of FAS. The feasibility of such an approach is supported by the recent demonstration that intratracheal administration of anti-FAS antibodies induced lung epithelial cell apoptosis and pulmonary fibrosis in mice (14).

In summary, our study of biopsies of failing human heart found that FAS was expressed by remaining heart cells adjacent to fibrotic foci but not in areas distal to fibrosis. We observed the same pattern associated with fibrosis near (perivascular) or far from vessels and in all patient samples. We speculate that expansion of fibrosis in the failing heart may be accelerated through induction of myocardial FAS expression, subsequent activation of apoptosis, further loss of parenchymal cells, and more fibrosis, leading to a vicious cycle and continuous deterioration of ventricular function.