INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE ANNEXINS ARE A UNIQUE family of calcium binding proteins characterized by calcium-dependent binding to phospholipids (for a review, see Ref. 18). They contain a tetrarepeat of a conserved core domain (except annexin VI, which has an octarepeat) and therefore share important structural homologies (up to 80-90%). In contrast, the NH₂-terminal domain is specific to each annexin and is involved in their functional diversity. The different annexins have been shown to mediate a variety of biological functions, such as exocytosis and endocytosis, membrane trafficking, regulation of phospholipase A₂ activity, and cell differentiation (3, 5, 18, 19, 27). Annexin II has been mainly reported to participate in signal transduction pathways and in membrane-fusion events (28). In vitro studies have indicated that annexins V and VI might be involved in the formation and regulation of calcium channels (4, 15). In the heart, annexin II has been detected in rat (11) and bovine (33) ventricular myocardium but not in atrial myocytes (10). Its presence has been also reported in endothelial (31) and vascular smooth muscle cells (21). Annexins V and VI are the major annexins in mammalian heart (6, 16). However, their subcellular localization is still controversial: annexin V has been shown to be absent from myocytes and present in nonmyocyte cells (27) or localized in cardiac myocytes (8, 11, 12, 16, 30) associated with cell membranes (8, 16) or in association with the Z line of the cells (30). Annexin VI has been mainly located in the sarcolemma and intercalated disks of cardiomyocytes (13). Although the functions of annexins within the heart are unknown, their properties and abundance suggest that they might play a role in hypertension and heart failure, particularly in Ca²⁺-mediated events and cardiac remodeling (23). In fact, it has been reported that the overexpression of annexin VI in the heart of transgenic mice leads to cardiac dilatation and altered calcium homeostasis (9). Furthermore, the expression of annexin II was increased in vascular smooth muscle cells by dexamethasone treatment, suggesting that it may be involved in the elevated systemic vascular resistance observed in glucocorticoid-induced hypertension (21). Moreover, Song et al. (22) recently reported that the amount of annexins II and V was increased whereas that of annexin VI was decreased in the human failing heart, suggesting that annexins may be important in cardiac remodeling.

Knowing where a protein is expressed often provides a strong clue as to its biological function; therefore we decided to determine the localization of annexins II, V, and VI in guinea pig myocardium and to investigate whether the cardiac levels and cellular localization of annexins II, V, and VI were modified during the onset of cardiac failure in hypertensive animals. The guinea pig was chosen because calcium cycling in this species is similar to that in the human heart (7, 24). Hypertension and left ventricular hypertrophy were induced by suprarenal coarctation of the abdominal aorta, and the first signs of heart failure were identified 20 mo later. Western blot analysis indicated a large increase in the steady-state concentration of annexins II, V, and VI in left ventricle (LV), and immunofluorescent histochemistry strongly suggested that annexins are involved in myocardial remodeling, particularly in fibrotic areas.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Animal model and tissues. Hypertension and LV hypertrophy were induced in adult female guinea pigs (350 ± 50 g) from Charles River (Cléon, France) by stenosis of the abdominal aorta above the renal artery (17). Sham-operated animals underwent an identical procedure except that the hemoclip was not placed around the aorta. All animal procedures were in accordance with institutional guidelines and those formulated by the European Community for use of experimental animals. Twenty months after surgery, animals were anesthetized with successive intraperitoneal injections of ketamine hydrochloride (20 mg/kg) followed by 2% xylazine (0.25 ml/kg) 10 min later. Hemodynamic parameters were measured in vivo in the closed-chest anesthetized state. An ultraminiature catheter pressure transducer was inserted into the right carotid artery and passed into the LV. The catheter was attached to a transducer control unit (model TC50, Millar Instruments) connected to a recorder (2000 series, Gould Electronic). Arterial pressures; left ventricular systolic, diastolic, and end-diastolic pressures; and maximal rates of pressure development were measured. At the end of recording period, hearts were excised, weighed, and rapidly rinsed in ice-cold saline solution. Hearts from operated animals were paired with those from sham-operated controls. An equatorial section (5 mm) was cut for immunofluorescent histochemical analysis. LVs with septum and right ventricles (RVs) were separated, minced into 200- to 300-mg pieces, and frozen in liquid nitrogen. Tissue samples were stored at 80°C until use.

Crude particulate preparation. Crude particulate preparations (CPP) were obtained according to the method of Rannou et al. (17). Briefly, tissue samples (200 mg) were thawed and homogenized in 10 ml of buffer (200 mmol/l sucrose, 20 mmol/l HEPES, pH 7.4) containing protease inhibitors (1.1 µmol/l leupeptin, 0.7 µmol/l aprotinin, 120 µmol/l phenylmethylsulfonyl fluoride, 1 mmol/l iodoacetamide, 0.7 µmol/l pepstatin, 1 mmol/l diisopropylfluorophosphate). The homogenate was centrifuged at 41,000 g for 45 min in a Sorvall SS34 rotor. The pellet was suspended in 0.1 mol/l NaCl, 30 mmol/l imidazole, 8% sucrose, pH 6.8, in the presence of the protease inhibitors. Protein content was determined by the method of Lowry using bovine serum albumin as a standard (12).

Western blot analysis. The amount of annexin protein contained in CPP was assessed using the Western blot technique (25). Anti-annexin antibodies were obtained by immunization of a rabbit with recombinant human annexins. Anti-annexin II and V antibodies did not cross-react with any other annexins. Anti-annexin VI antibody was purified by immunoaffinity and was highly specific. Heart proteins from CPP (30 µg) were loaded on a 10% polyacrylamide SDS gel. After electrophoretic separation, proteins were transferred to a nitrocellulose sheet (Hybond ECL, Amersham) at 150 V for 1.5 h. The membranes were then successively incubated in 5% fat-free milk, 0.1% Tween 20 in PBS for 2 h at room temperature, specific antiserum in blocking solution (1:5,000) overnight at 4°C, then with horseradish peroxidase-conjugated anti-rabbit IgG for 2 h at room temperature. The reactive proteins were detected by chemiluminescent reaction (ECL⁺ kit, Amersham) followed by exposition of the membranes to Hyperfilm ECL film and serial washes (5 × 5 min) with a medium composed of 0.1 M sodium acetate (pH = 4) and 0.5 M sodium chloride to eliminate the antibodies. The proteins were then stained with Coomassie blue R250 (0.1% in 1% CH₃COOH, 40% methanol) for 5 min. Nonspecific staining was then suppressed by a 5-min wash in 10% CH₃COOH, 50% methanol. Quantification of both the specific bands on autoradiography and the transferred MyHC band stained with Coomassie blue was performed by densitometric analysis after scanning. Each individual value represents the mean of three independent determinations.

Immunofluorescent histochemistry. Cardiac samples were placed in mounting solution (Ystosystem) and frozen at 155°C in isopentane. Equatorial frozen sections (7-8 µm) were fixed with acetone-methanol (1:1 for 20 min at 20°C), saturated with 5% BSA in PBS (30 min at room temperature), incubated with mouse monoclonal anti--actinin antibodies (1:50 in PBS for 30 min at 37°C), washed, and incubated with appropriate dilutions of either anti-annexin II, anti-annexin V, anti-annexin VI, or anti-Na⁺-K⁺-ATPase (2) antiserum (1:40, 1:20, 1:20, and 1:50, respectively, in PBS for 30 min at 37°C). After three rinses in PBS, the sections were successively incubated for 30 min at room temperature with a 1:50 dilution of anti-mouse immunoglobulins conjugated with Texas red and a 1:50 dilution of anti-rabbit immunoglobulins conjugated with fluorescein isothiocyanate fluorochrome. Fluorescence was visualized using a Leitz DM RD microscope equipped with epifluorescence optics.

Statistical analysis. Results are expressed as means ± SE. The statistical significance of differences among the groups was determined by one-way ANOVA, and comparisons between groups were performed by Scheffé's test. A value of P < 0.05 was considered to be statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Anatomic and hemodynamic studies. Abdominal aortic banding in 3-wk-old guinea pigs led to a progressive increase in pressure overload, left ventricular hypertrophy, and hypertension as the banded animals grew. Anatomic and hemodynamic data for banded and sham-operated animals 20 mo after operation are summarized in Table 1. By this time, guinea pigs were adult animals and did not show any signs of senescence (life duration is ~5 yr).

The in vivo hemodynamic data show that aortic diastolic and systolic pressures were increased by 53 and 46%, respectively. LV systolic pressure was increased by 85% and LV diastolic pressure was unchanged. However, 20 mo after the operation an increase in LV end-diastolic pressure (34%) and a decrease in positive and negative dP/dt were observed. These changes were not observed at earlier time points (unpublished results). They represent the first signs of cardiac decompensation.

Body weight was not significantly different between the banded and sham-operated groups. Left ventricular weight and left ventricular-to-body weight ratio were significantly increased in banded animals, indicating LV hypertrophy, whereas right ventricular weight was not significantly different although a tendency toward an increase was observed. Moreover, lung and liver wet-to-dry weight ratios were unchanged in the banded animals (data not shown), demonstrating that there were no pathological features of established heart failure in this model.

Immunoblot analysis of annexin II, annexin V, and annexin VI. Protein levels of annexin II, V and VI were measured in CPP from sham-operated and banded animals with the immunoblotting technique. As previously observed (17), the yield of proteins in homogenate or CPP was similar in the two groups (71 ± 4.4 and 65.1 ± 4.2 mg protein/g of tissue in CPP, respectively). An example of immunoblot analysis is shown in Fig. 1. The specific bands labeled with anti-annexin antibodies correspond to 36 kDa for annexin II, 32.5 kDa for annexin V, and 68 kDa for annexin VI. Increased amounts of annexins were evident in LV from banded guinea pigs. Quantitative results obtained after densitometric analysis of each band and normalization by the protein content of each lane measured after Coomassie blue staining are shown in Fig. 1B. Significant increases in annexin II (2.6-fold), annexin V (1.45-fold), and annexin VI (2.3-fold) levels were observed in LV of the banded group. No significant modification of the accumulation of the annexins was observed in RV of banded guinea pigs compared with sham-operated guinea pigs.

View larger version (50K): [in this window] [in a new window]   Fig. 1.   Expression of annexin II, annexin V, and annexin VI in sham-operated and banded guinea pigs. A: example of immunoblot analysis. Proteins from crude particulate preparations (30 µg) of sham-operated animals (lanes 1, 3, 5, and 7) and of banded animals (lanes 2, 4, 6, and 8) were separated by 10% SDS-PAGE and detected by polyclonal antibodies against annexins II, V, and VI as described in text. B: quantitation of annexin II, annexin V, and annexin VI in left ventricle (LV) and right ventricle (RV) from sham-operated and banded guinea pigs. Bar graphs show protein levels (obtained by immunoblot analysis and normalization to total protein level of each lane) of annexins II, V, and VI in sham-operated animals (open bars) and in LV (solid bars) and RV (hatched bars) of banded animals. Data are means ± SE; n = 10 animals per group. Each value is mean of 3 different determinations. *** P < 0.001.

Immunohistochemical localization of annexin II, annexin V, and annexin VI. Double immunofluorescence using annexin-specific polyclonal antibodies and -actinin-specific monoclonal antibody was employed to determine the myocyte or nonmyocyte distribution of each annexin in the myocardium of sham-operated and banded guinea pigs and to compare the annexin labeling to the striated pattern of sarcomeric filaments. In addition, sarcolemma and T tubules were specifically labeled with anti-Na⁺-K⁺-ATPase antibodies (30).

In sham-operated animals, annexin II (Fig. 2A) showed a striking preponderance in the capillaries and in the endocardial endothelium, and immunoreactivity was faint in the interstitial tissue between myocytes. In banded animals, distribution of annexin II was unchanged, although labeling was reinforced between myocytes (Fig. 2B). Double labeling with -actinin or phase-contrast imaging (not shown) demonstrated that the most intense immunoreactivity corresponded to tissue areas that were devoid of myocytes and might further represent fibrotic areas. Annexin II localization in RV (Fig. 2C) from both sham-operated and banded guinea pigs was similar to that of the LV from sham-operated group.

View larger version (166K): [in this window] [in a new window]   View larger version (141K): [in this window] [in a new window]   View larger version (150K): [in this window] [in a new window]   Fig. 2.   Localization of annexin II by immunofluorescent staining in frozen ventricular sections from sham-operated (A) and banded (B and C) guinea pigs. A and B: LV. C: RV. Bars represent 50 µm.

Annexin V in LV (Fig. 3, A and C) and RV (not shown) from sham-operated guinea pigs was mainly observed at the level of the sarcolemma and intercalated disks and penetrating into the myocytes. Immunofluorescence was also detected in striation parallel to the long axis of the myocyte. The labeling of Na⁺-K⁺-ATPase (Figs. 3E and 4E), used as a marker of sarcolemma, was detected as expected at the level of myocyte membranes and invaginating into myocytes. In LV from banded animals, there was an overall increase in the labeling of annexin V, particularly evident between the myocytes (Fig. 3, B and D), whereas there was no detectable change in RV (Fig. 3F).

View larger version (187K): [in this window] [in a new window]   Fig. 3.   Localization of annexin V by immunofluorescent staining in frozen ventricular sections from sham-operated (A and C) and banded (B, D, and F) guinea pigs. A-E: LV. F: RV. Arrows indicate zones of increased magnification shown in insets (A: 2-fold; E: 1.4-fold). Na+-K+-ATPase labeling in sham-operated guinea pigs is shown in E. Bars represent 50 µm.

As shown in Fig. 4, annexin VI was located in the external sarcolemma and intercalated disks of myocytes from sham-operated guinea pigs; a similar labeling was observed for the Na⁺-K⁺-ATPase (Fig. 4E). In LV (Fig. 4, B and D) and RV (Fig. 4F) from banded guinea pigs, the localization was unchanged but the fluorescence was enhanced in LV.

View larger version (192K): [in this window] [in a new window]   Fig. 4.   Localization of annexin VI by immunofluorescence staining in frozen ventricular sections from sham-operated (A and C) and banded (B, D, and F) guinea pigs. A-E: LV. F: RV. Na+-K+-ATPase labeling in sham-operated guinea pigs is shown in E. Arrowheads indicate intercalated disks. Bars represent 50 µm.

Localization of annexins II, V, and VI in coronary arteries is represented in Fig. 5. It shows that annexins II and V were found in the endothelium and the adventitia of large vessels. Immunoreactivity of annexin VI in coronary artery was particularly intense in the endothelium and media. In banded animals, the localization of annexins II, V, and VI was similar to that observed in sham-operated guinea pigs.

View larger version (187K): [in this window] [in a new window]   Fig. 5.   Localization of annexins II, V, and VI in coronary arteries from frozen ventricular sections of sham-operated (A, C, and E) and banded (B, D, and F) guinea pigs. Na+-K+-ATPase labeling is shown in E. Bars represent 50 µm.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study we examined the cardiac changes in steady-state concentration and cellular distribution of annexin II, annexin V, and annexin VI during the onset of cardiac failure. The findings demonstrate that the three annexins are expressed in guinea pig hearts and that their expression is highly increased in LV of banded animals whereas it is unchanged in RV. In sham-operated guinea pigs, annexin II is mainly localized in the membranes of nonmuscle cells whereas strong staining specific to annexins V and VI is observed in the sarcolemma and intercalated disks of myocytes. Interstitial tissue between myocytes shows strong immunoreactivity to all three annexins in the banded guinea pig hearts.

In this model of progressive overload induced by aortic stenosis in guinea pigs, we have previously shown that during compensatory hypertrophy, which is observed after 6 and 12 mo of aortic stenosis, left ventricular end-diastolic pressure and positive and negative dP/dt were similar in sham-operated and banded guinea pigs (unpublished results). The significant changes in these two parameters after 20 mo of aortic stenosis are indicative of the first signs of decompensatory hypertrophy and of the transition from compensatory hypertrophy toward heart failure.

The presence of annexins II, V, and VI in guinea pig hearts has been established by immunoblot analysis of crude particulate preparations obtained in the absence of Ca²⁺ chelating agents to prevent the loss of annexins in the cytosol. The result is in line with studies in other species (16). It also demonstrates the specificity of the antibodies used in the immunofluorescent histochemical analysis. At the onset of heart failure, the content of annexins II, V, and VI is largely increased whereas in human heart failure, only annexins II and V are increased (22). Whether the difference in annexin VI level is due to the species or to the severity of heart failure is unknown.

In this study, annexin II is localized to endothelial cells of the coronary arteries and intramyocardial capillaries, in agreement with the established presence of annexin II in vascular endothelium (14). It is undetectable in ventricular myocytes (Fig. 2) and in atrial myocytes (16). Annexin II in LV from banded guinea pigs is increased 2.6-fold. Immunofluorescence staining suggests that this increase is correlated to the presence of annexin II in the interstitial tissue, between myocytes (Fig. 2). The increased accumulation of annexin II is in agreement with the report of Song et al. (22) in heart failure, suggesting that it is maintained whatever the severity of heart failure. We suggest that, among the different properties of annexin II, its ability to bind to components of the extracellular matrix such as collagen (32) and to mediate cell-cell adhesion (3, 28) might be an important feature involved in the remodeling process of interstitial and perivascular fibrosis.

The localization of annexin V in mammalian myocardium is controversial. As shown in Fig. 3, we found annexin V in the sarcolemma and in the intercalated disks of myocytes, in agreement with other reports (6, 8, 16). Invaginations of the sarcolemma into the cells, similar to those shown for the Na⁺-K⁺-ATPase (29), likely represent T tubules, as reported by Luckcuck et al. (13). However, we did not find, as reported by Wang et al. (30) in isolated cardiomyocytes from rat, that the labeling of annexin V is similar to the striated pattern observed for -actinin in guinea pig LV. Whether this discrepancy is related to a difference between species or to the specificity of the antibody for a splicing isoform of annexin V (30) is undetermined. Furthermore, the faint longitudinal striation pattern observed in longitudinal sections (Fig. 3B) strongly resembles that of cardiotin, a sarcoplasmic reticulum-specific protein (20), suggesting that annexin V is also present in sarcoplasmic reticulum. In LV from banded guinea pigs, accumulation of annexin V is increased by 1.45-fold and immunofluorescent analysis demonstrated that annexin V is mainly present in between the myocytes. Annexin V, like annexin II, is a collagen-binding protein that might promote interactions between cells and proteins of the extracellular matrix (32), suggesting a role for annexin V in the development of fibrosis. Furthermore, in myocytes from banded guinea pigs, the longitudinal pattern of annexin V labeling was less prominent than in control myocytes and even absent in some areas. If this is related to relocation of annexin V from sarcoplasmic reticulum, it would suggest that in myocytes, annexin V might also contribute to the alterations in Ca²⁺ homeostasis and sarcoplasmic reticulum Ca²⁺ handling reported in hypertension and heart failure (1, 26).

Luckcuck et al. (13) reported a similar localization of annexins V and VI in transverse sections from porcine LV. In this study, we partly confirm these findings and found strong labeling of annexin VI in the sarcolemma and intercalated disks (Figs. 3 and 4) but we did not find annexin VI in T tubules. In LV from banded guinea pigs, the 2.3-fold increase in accumulation of annexin VI, the unchanged localization of annexin VI in myocytes, and the presence of annexin VI in interstitial tissue suggest that annexin VI, like annexins II and V, might be involved in the development of fibrotic areas. Moreover, it has been shown in transgenic mice that overexpression of annexin VI leads to dilated cardiomyopathy (9); this suggests that at the onset of heart failure, annexin VI may participate in the remodeling and altered properties of the myocardium.

In this study, we show that annexins II, V, and VI are present in coronary arteries (Fig. 5) and annexin II in capillary vessels. Endothelial cells are strongly labeled by the three annexins and media by annexin VI only. We do not find any obvious difference in the labeling of coronary arteries by annexins between LV from sham-operated and banded animals, but the very strong staining of annexins in both tissues argues for a role of annexins in vascular remodeling during hypertension.

In the present study, we have analyzed the expression and localization of annexins II, V, and VI in the hearts of hypertensive animals during the onset of heart failure and have shown, for the first time, that the accumulation of each annexin is increased in LV whereas it remains unchanged in RV. These results suggest a mechanical and not an hormonal regulation of their expression during hypertension. The present study also reports for the first time that the respective localizations of annexins II, V, and VI are modified in the diseased heart. In particular, a prominent labeling of annexins in the interstitial tissue is observed. According to the potential roles of annexins in the regulation of association between proteins of the extracellular matrix, we propose that they are involved in the myocardial and vascular remodeling process occurring during heart failure.